Clinical Review

Psoriasis is a chronic, autoimmune-mediated disease estimated to affect 2.8% of the US population.¹ The pathogenesis of psoriasis is thought to involve a complex process triggered by a combination of genetic and environmental factors that induce tumor necrosis factor (TNF) α secretion by keratinocytes, which in turn activates dendritic cells. Activated dendritic cells produce IL-23, leading to helper T cell (T_H17) differentiation.^2,3 T_H17 cells secrete IL-17A, which has been shown to promote psoriatic skin changes.⁴ Therefore, TNF-α, IL-23, and IL-17A have been recognized as key targets for psoriasis therapy.

The newest biologic agents targeting IL-17–mediated pathways include ixekizumab, brodalumab, and bimekizumab. Secukinumab, the first US Food and Drug Administration (FDA)–approved IL-17 inhibitor, has been available since 2015 and therefore is not included in this review. IL-23 inhibitors that are FDA approved or being evaluated in clinical trials include guselkumab, tildrakizumab, and risankizumab. In addition, certolizumab pegol, a TNF-α inhibitor, is being studied for use in psoriasis.

METHODS

We reviewed the published results of phase 3 clinical trials for ixekizumab, brodalumab, bimekizumab, guselkumab, tildrakizumab, risankizumab, and certolizumab pegol. We performed an English-language literature search (January 1, 2012 to October 15, 2017) of articles indexed for PubMed/MEDLINE using the following combinations of keywords: IL-23 and psoriasis; IL-17 and psoriasis; tumor necrosis factor and psoriasis; [drug name] and psoriasis. If data from phase 3 clinical trials were not yet available, data from phase 2 clinical trials were incorporated in our analysis. We also reviewed citations within articles to identify relevant sources.

RESULTS

Phase 3 clinical trial design, efficacy, and adverse events (AEs) for ixekizumab and brodalumab are reported in eTable 1^5-10 and for guselkumab and tildrakizumab in eTable 2.^11-14 Phase 2 clinical trial design, efficacy, and AEs are presented for risankizumab in eTable 3^15-18 and for certolizumab pegol in eTable 4.^17,19 No published clinical trial data were found for bimekizumab.

IL-17 Inhibitors

Ixekizumab
This recombinant, high-affinity IgG4κ antibody selectively binds and neutralizes IL-17A.^5,6 Three phase 3 clinical trials—UNCOVER-1, UNCOVER-2, and UNCOVER-3—evaluated ixekizumab for moderate to severe plaque psoriasis.⁷

The 3 UNCOVER trials were randomized, double-blind, phase 3 trials of 1296, 1224, and 1346 patients, respectively, assigned to a placebo group; a group treated with ixekizumab 80 mg every 2 weeks; and a group treated with ixekizumab 80 mg every 4 weeks. Both ixekizumab groups received a loading dose of 160 mg at week 0.^5,6 UNCOVER-2 and UNCOVER-3 also included a comparator group of patients on etanercept 50 mg.⁵ Co-primary end points included the percentage of patients reaching a psoriasis area and severity index (PASI) of 75 and with a static physician global assessment (PGA) score of clear (0) or almost clear (1) at week 12.^5,6

Ixekizumab achieved greater efficacy than placebo: 89.1%, 89.7%, and 87.3% of patients achieved PASI 75 in the every 2-week dosing group, and 82.6%, 77.5% and 84.2% achieved PASI 75 in the every 4-week dosing group in UNCOVER-1, UNCOVER-2, and UNCOVER-3, respectively (P<.001 for both treatment arms compared to placebo in all trials). The percentage of patients achieving a static PGA score of 0 or 1 also was higher in the ixekizumab groups in the 2-week and 4-week dosing groups in all UNCOVER trials—81.8% and 76.4% in UNCOVER-1, 83.2% and 72.9% in UNCOVER-2, and 80.5% and 75.4% in UNCOVER-3—compared to 3.2%, 2.4%, and 6.7% in the placebo groups of the 3 trials (P<.001 for both ixekizumab groups compared to placebo in all trials).^5,6 Ixekizumab also was found to be more effective than etanercept for both co-primary end points in both UNCOVER-2 and UNCOVER-3 (eTable 1).⁵

Safety data for all UNCOVER trials were pooled and reported.⁶ At week 12 the rate of at least 1 AE was 58.4% in patients on ixekizumab every 2 weeks and 58.8% in patients on ixekizumab every 4 weeks compared to 54.0% in the etanercept group in UNCOVER-2 and UNCOVER-3 and 46.8% in the placebo group. At week 12, 72 nonfatal serious AEs were reported: 12 in the placebo group, 14 in the etanercept group, 20 in the ixekizumab every 2 weeks group, and 26 in the ixekizumab every 4 weeks group.⁶

The most common AE across all groups was nasopharyngitis. Overall, infections were more frequent in patients treated with ixekizumab than in patients treated with placebo or etanercept. Specifically, oral candidiasis occurred more frequently in the ixekizumab groups, with a higher rate in the 2-week dosing group than in the 4-week dosing group.⁶ Two myocardial infarctions (MIs) occurred: 1 in the etanercept group and 1 in the placebo group.⁵

Brodalumab
This human monoclonal antibody binds to IL-17ra.^8,9 Three double-blind, placebo-controlled, phase 3 trials—AMAGINE-1, AMAGINE-2, and AMAGINE-3—evaluated its use for plaque psoriasis.¹⁰

In AMAGINE-1 (N=661), patients were randomized to receive brodalumab 140 mg or 210 mg (every 2 weeks for 12 weeks), or placebo.⁸ In AMAGINE-2 (N=1831) and AMAGINE-3 (N=1881), patients were randomized to receive brodalumab 140 mg or 210 mg (every 2 weeks for 12 weeks), ustekinumab 45 mg or 90 mg by weight (at weeks 0 and 4, then every 12 weeks thereafter), or placebo. In all trials, patients on brodalumab received a dose at week 0 and week 1. Co-primary end points were PASI 75 and a static PGA score of 0 or 1 at 12 weeks compared to placebo and to ustekinumab (in AMAGINE-2 and AMAGINE-3 only).⁸

At week 12, 83.3%, 86.3%, and 85.1% of patients on brodalumab 210 mg, and 60.3%, 66.6%, and 69.2% of patients on brodalumab 140 mg, achieved PASI 75 in AMAGINE-1, AMAGINE-2, and AMAGINE-3, respectively, compared to 2.7%, 8.1%, and 6.0% in the placebo groups (P<.001 between both brodalumab groups and placebo in all trials).⁸ Both brodalumab groups were noninferior but not significantly superior to ustekinumab, which achieved a PASI 75 of 70.0% in AMAGINE-2 and 69.3% in AMAGINE-3. The PASI 90 rate was higher, however, in both brodalumab groups compared to ustekinumab but significance was not reported (eTable 1).⁹ For both brodalumab groups, significantly more patients achieved a static PGA value of 0 or 1 compared to placebo (P<.001 across all trials). However, only the brodalumab 210-mg group achieved a significantly higher rate of static PGA 0 or 1 compared to ustekinumab in AMAGINE-2 and AMAGINE-3 (P<.001).⁹

After 12 weeks, the percentage of patients reporting at least 1 AE was 59.0%, 57.8%, and 56.8% in the brodalumab 210-mg group in AMAGINE-1, AMAGINE-2, and AMAGINE-3, respectively; 58.0%, 60.1%, and 52.6% in the brodalumab 140-mg group; and 51.0%, 53.4%, and 48.6% in the placebo group. Patients taking ustekinumab had an AE rate of 59.0% in AMAGINE-2 and 53.7% in AMAGINE-3. The most common AE was nasopharyngitis, followed by upper respiratory infection (URI) and headache across all trials.^8,9 Serious AEs were rare: 10 in AMAGINE-1, 31 in AMAGINE-2, and 24 in AMAGINE-3 across all groups. One death occurred from stroke in the brodalumab 210-mg group in AMAGINE-2.⁹

IL-23 Inhibitors

Guselkumab
This drug is a human IgG1κ antibody that binds to the p19 subunit of IL-23, thereby inhibiting IL-23 signaling.^11,12 Guselkumab was approved by the FDA in July 2017 for moderate to severe plaque psoriasis.¹³

VOYAGE 1 and VOYAGE 2 were phase 3, double-blind, placebo- and active comparator–controlled trials of 837 and 992 patients, respectively, randomized to receive adalimumab (80 mg at week 0 and 40 mg at week 1, then at 40 mg every 2 weeks thereafter), guselkumab 100 mg at weeks 0, 4, and 12, or placebo.¹¹ Co-primary end points for both trials were the percentage of patients reaching PASI 90 and an investigator global assessment (IGA) score of cleared (0) or minimal (1) at week 16.¹¹

By week 16 of both trials, PASI 90 values were statistically superior for guselkumab (VOYAGE 1, 73.3%; VOYAGE 2, 70.0%) compared to adalimumab (VOYAGE 1, 49.7%; VOYAGE 2, 46.8%) and placebo (VOYAGE 1, 2.9%; VOYAGE 2, 2.4%)(P<.001). Moreover, patients on guselkumab achieved a higher rate of IGA values of 0 and 1 at week 12 (85.1% in VOYAGE 1 and 84.1% in VOYAGE 2) than patients on adalimumab (65.9% in VOYAGE 1 and 67.7% in VOYAGE 2) and placebo (6.9% in VOYAGE 1 and 8.5% in VOYAGE 2)(P<.001).^11,12

The frequency of AEs was comparable across all groups in both trials.^11,12 During the 16-week treatment period, 51.7% and 47.6% of the guselkumab groups in VOYAGE 1 and VOYAGE 2, respectively; 51.1% and 48.4% of the adalimumab groups; and 49.4% and 44.8% of the placebo groups reported at least 1 AE. The most common AEs in all groups were nasopharyngitis, headache, and URI.^11,12

Serious AEs also occurred at similar rates: 2.4% and 1.6% in the guselkumab group in VOYAGE 1 and VOYAGE 2, respectively; 2.4% and 1.8% in the adalimumab group; and 1.7% and 1.2% in the placebo group.^11,12 One case of malignancy occurred in the VOYAGE 1 trial: basal cell carcinoma in the guselkumab group.¹¹ Three major cardiovascular events occurred across both trials: 1 MI in the guselkumab group in each trial and 1 MI in the adalimumab group in VOYAGE 1.^11,12

Tildrakizumab
A high-affinity, humanized IgG1κ antibody, tildrakizumab targets the p19 subunit of IL-23. As of February 2018, 2 double-blind, randomized phase 3 trials have studied tildrakizumab with published results: reSURFACE 1 and reSURFACE 2.¹⁴

reSURFACE 1 (N=772) and reSURFACE 2 (N=1090) randomized patients to receive tildrakizumab 100 or 200 mg (at weeks 0 and 4), etanercept 50 mg (twice weekly) for 12 weeks (reSURFACE 2 only), or placebo. Co-primary end points were the percentage of patients achieving PASI 75 and the percentage of patients achieving a PGA score of 0 or 1 at week 12.¹⁴

In reSURFACE 1, significantly more patients receiving tildrakizumab attained PASI 75 at week 12 compared to placebo: 200 mg, 62.0%; 100 mg, 64.0%; and placebo, 6.0% (P<.001 for tildrakizumab groups compared to placebo). Moreover, significantly proportionally more patients received a PGA score of 0 or 1 compared to placebo: 100 mg, 59%; 200 mg, 58.0%; placebo, 7.0% (P<.001 for both tildrakizumab groups compared to placebo).¹⁴

In reSURFACE 2, significantly more patients receiving tildrakizumab achieved PASI 75 compared to etanercept and placebo at week 12: 200 mg, 66.0%; 100mg, 61.0%; etanercept, 48.0%; placebo, 6.0% (P<.001 for both tildrakizumab groups compared to placebo; P<.05 for both tildrakizumab groups compared to etanercept). Additionally, significantly more patients in the tildrakizumab groups experienced a PGA score of 0 or 1 at week 12 compared to placebo: 200 mg, 59%; 100 mg, 55.0%; placebo, 5% (P<.001 for both tildrakizumab groups compared to placebo).¹⁴

Adverse events were reported at a similar rate across all groups. For reSURFACE 1 and reSURFACE 2, at least 1 AE by week 12 was reported by 42.2% and 45.2% of patients in the 200-mg group; 47.2% and 45.9% in the 100-mg group; and 48.1% and 55.1% in the placebo groups.¹⁴The most common AEs were nasopharyngitis, URI (reSURFACE 1), and erythema at the injection site (reSURFACE 2). One case of serious infection was reported in each of the tildrakizumab groups: 1 case of drug-related hypersensitivity reaction in the 200-mg group, and 1 major cardiovascular event in the 100-mg group of reSURFACE 1. There was 1 serious AE in reSURFACE 2 that led to death in which the cause was undetermined.¹⁴

Risankizumab
This humanized IgG1 antibody binds the p19 unit of IL-23.^15,16 The drug is undergoing 3 phase 3 trials—ultIMMa-1, ultIMMa-2, and IMMvent—for which only preliminary data have been published and are reported here.^16,17 There is 1 phase 2 randomized, dose-ranging trial with published data.¹⁵

ultIMMa-1 and ultIMMa-2 comprised 506 and 491 patients, respectively, randomized to receive risankizumab (150 mg at weeks 0, 4, and 16), ustekinumab (45 mg or 90 mg, by weight, at weeks 0, 4, and 16), or placebo. Co-primary end points were PASI 90 and a PGA score of 0 or 1 at week 16.¹⁷

In ultIMMa-1 and ultIMMa-2, 75.0% and 75.0% of patients on risankizumab 150 mg achieved PASI 90 compared to 42.0% and 48.0% on ustekinumab and 5.0% and 2.0% on placebo at 16 weeks (P<.001 between both placebo and ustekinumab in both trials).¹⁷ In both trials, patients receiving risankizumab achieved higher rates of a static PGA score of 0 or 1 (88.0% and 84.0%) compared to ustekinumab (63.0% and 62.0%) and placebo (8.0% and 5.0%) at 16 weeks (P<.001 for both trials).¹⁸

At week 16, 2.0% of patients on risankizumab reported a serious AE in both trials, compared to 8.0% and 3.0% of patients on ustekinumab and 3.0% and 1.0% on placebo. No new safety concerns were noted.¹⁷

In the phase 3 IMMvent trial, 605 patients were randomized to receive risankizumab (150 mg at weeks 0, 4, and 16) or adalimumab (80 mg at week 0, 40 mg at week 1, then 40 mg every 2 weeks). Co-primary end points were PASI 90 and a static PGA score of 0 or 1 at week 16.¹⁷

In IMMvent, risankizumab was significantly more effective than adalimumab for PASI 75 (risankizumab, 72.0%; adalimumab, 47.0%) and a static PGA score of 0 or 1 (risankizumab 84.0%; adalimumab, 60.0%) (P<.001 risankizumab compared to adalimumab for both end points).¹⁷

At week 16, serious AEs were reported in 3.0% of patients on risankizumab and 3.0% of patients on adalimumab. One patient receiving risankizumab died of an acute MI during the treatment phase.¹⁷

TNF Inhibitor

Certolizumab Pegol
Certolizumab pegol is a human PEGylated anti-TNF agent. In vitro studies have shown that certolizumab binds to soluble and membrane-bound TNF.¹⁹ Unlike other TNF inhibitors, certolizumab pegol is a Fab‘ portion of anti-TNF conjugated to a molecule of polyethylene glycol.¹⁹ The drug is approved in the United States for treating psoriatic arthritis, Crohn disease, and rheumatoid arthritis; its potential for treating psoriasis has been confirmed. Results of 1 phase 2 trial have been published¹⁹; data from 3 phase 3 trials are forthcoming.

This randomized, placebo-controlled, double-blind phase 2 study comprised 176 patients who received certolizumab 200 mg, certolizumab 400 mg, or placebo. The dosing schedule was 400 mg at week 0, followed by either 200 or 400 mg every other week until week 10. Co-primary end points were PASI 75 and a PGA score of 0 or 1 at week 12.¹⁹

Certolizumab was significantly more effective than placebo at week 12: 74.6% of the 200-mg group and 82.8% of the 400-mg group achieved PASI 75 compared to 6.8% of the placebo group (P<.001). Certolizumab also performed better for the PGA score: 52.5% and 72.4% of patients attained a score of 0 or 1 in the 200-mg and 400-mg groups compared to 1.7% in the placebo group.¹⁹

Adverse events were reported equally across all groups: 72% of patients in the 200-mg group, 70% in the 400-mg group, and 71% in the placebo group reported at least 1 AE, most commonly nasopharyngitis, headache, and pruritis.¹⁹

COMMENT

With the development of new insights into the pathogenesis of psoriasis, therapies that are targeted toward key cytokines may contribute to improved management of the disease. The results of these clinical trials demonstrate numerous promising options for psoriatic patients.

IL-17 Inhibitors Ixekizumab and Brodalumab

When comparing these 2 biologics, it is important to consider that these studies were not performed head to head, thereby inhibiting direct comparisons. Moreover, dosage ranges of the investigative drugs were not identical, which also makes comparisons challenging. However, when looking at the highest dosages of ixekizumab and brodalumab, results indicate that ixekizumab may be slightly more effective than brodalumab based on the percentage of patients who achieved a PASI 75 and a static PGA score of 0 or 1 (eTable 1).

Phase 3 trials have shown ixekizumab to maintain efficacy over 60 weeks of treatment.⁶ Ixekizumab also has been shown to alleviate other symptoms of psoriasis, such as itching, pain, and nail involvement.^20,21 Furthermore, ixekizumab appears to be equally effective in patients with or without prior exposure to biologics²²; therefore, ixekizumab may benefit patients who have not experienced success with other biologics.

Across the UNCOVER trials, 11 cases of inflammatory bowel disease were reported in patients receiving ixekizumab (ulcerative colitis in 7; Crohn disease in 4)⁶; it appears that at least 3 of these cases were new diagnoses. In light of a study suggesting that IL-17A might have a protective function in the intestine,²³ these findings may have important clinical implications and require follow-up studies.

Brodalumab also has been shown to maintain efficacy and acceptable safety for as long as 120 weeks.²⁴ In the extension period of the AMAGINE-1 trial, patients who experienced a return of disease during a withdrawal period recaptured static PGA success with re-treatment for 12 weeks (re-treatment was successful in 97% of those given a dosage of 210 mg and in 84% of those given 140 mg).⁸

Furthermore, phase 2 trials also have shown that brodalumab is effective in patients with a history of biologic use.²⁵ Across all AMAGINE trials, only 1 case of Crohn disease was reported in a patient taking brodalumab.⁹ There are concerns about depression, despite data from AMAGINE-1 stating patients on brodalumab actually had greater improvements in Hospital Anxiety and Depression Scale scores after 12 weeks of treatment (P<.001) for both brodalumab 140 mg and 210 mg compared to placebo.⁸ Regardless, brodalumab has a black-box warning for suicidal ideation and behavior, and availability is restricted through a Risk Evaluation and Mitigation Strategy (REMS) program.²⁶

Bimekizumab

Although no phase 2 or phase 3 clinical trial data have been published for bimekizumab (phase 2 trials are underway), it has been shown in a phase 1 trial to be effective for psoriasis. Bimekizumab also is unique; it is the first dual inhibitor of IL-17A and IL-17F.¹⁸

IL-23 Inhibitors Guselkumab, Tildrakizumab, and Risankizumab

Making comparisons among the IL-23 inhibitors also is difficult; studies were not head-to-head comparison trials, and the VOYAGE and reSURFACE studies used different time points for primary end points. Furthermore, only phase 2 trial data are available for risankizumab. Despite these limitations, results of these trials suggest that guselkumab and risankizumab may be slightly more efficacious than tildrakizumab. However, future studies, including head-to-head studies, would ultimately provide further information on how these agents compare.

Guselkumab was shown to remain efficacious at 48 weeks, though patients on maintenance dosing had better results than those who were re-treated.¹² Moreover, guselkumab was found to be effective in hard-to-treat areas, such as the scalp,¹¹ and in patients who did not respond to adalimumab. Guselkumab may therefore benefit patients who have experienced limited clinical improvement on other biologics.¹²

Tildrakizumab was shown to improve PASI 75 and PGA scores through week 28 of treatment. Moreover, a higher percentage of patients taking tildrakizumab scored 0 or 1 on the dermatology life quality index, suggesting that the drug improves quality of life.¹⁴ No specific safety concerns arose in either reSURFACE trial; however, long-term studies are needed for further evaluation.

Risankizumab appears to be a promising new therapy based on phase 2 trial results. Improvements also were seen in dermatology life quality index scores, scalp and fingernail symptoms, and palmoplantar psoriasis.¹⁵ Of note, neutralizing antidrug antibodies were found in 3 patients during this study,¹⁵ which may present potential problems for long-term efficacy. However, preliminary data from 3 phase 3 trials—ultIMMa-1, ultIMMa-2, and IMMvent—are promising.¹⁷

CONCLUSION

Advances in the understanding of psoriasis have led to new targeted therapies. Ongoing clinical trials have shown encouraging results for treating physical and psychological symptoms of psoriasis. The findings of these trials support the idea that therapies targeting IL-23, specifically its p19 subunit, are effective against psoriasis while sparing IL-12. Long-term data from open-label extension studies would help guide clinical recommendations regarding the safety profiles of these agents and determine their long-term utility.

Psoriasis is a chronic, autoimmune-mediated disease estimated to affect 2.8% of the US population.¹ The pathogenesis of psoriasis is thought to involve a complex process triggered by a combination of genetic and environmental factors that induce tumor necrosis factor (TNF) α secretion by keratinocytes, which in turn activates dendritic cells. Activated dendritic cells produce IL-23, leading to helper T cell (T_H17) differentiation.^2,3 T_H17 cells secrete IL-17A, which has been shown to promote psoriatic skin changes.⁴ Therefore, TNF-α, IL-23, and IL-17A have been recognized as key targets for psoriasis therapy.

The newest biologic agents targeting IL-17–mediated pathways include ixekizumab, brodalumab, and bimekizumab. Secukinumab, the first US Food and Drug Administration (FDA)–approved IL-17 inhibitor, has been available since 2015 and therefore is not included in this review. IL-23 inhibitors that are FDA approved or being evaluated in clinical trials include guselkumab, tildrakizumab, and risankizumab. In addition, certolizumab pegol, a TNF-α inhibitor, is being studied for use in psoriasis.

METHODS

We reviewed the published results of phase 3 clinical trials for ixekizumab, brodalumab, bimekizumab, guselkumab, tildrakizumab, risankizumab, and certolizumab pegol. We performed an English-language literature search (January 1, 2012 to October 15, 2017) of articles indexed for PubMed/MEDLINE using the following combinations of keywords: IL-23 and psoriasis; IL-17 and psoriasis; tumor necrosis factor and psoriasis; [drug name] and psoriasis. If data from phase 3 clinical trials were not yet available, data from phase 2 clinical trials were incorporated in our analysis. We also reviewed citations within articles to identify relevant sources.

RESULTS

Phase 3 clinical trial design, efficacy, and adverse events (AEs) for ixekizumab and brodalumab are reported in eTable 1^5-10 and for guselkumab and tildrakizumab in eTable 2.^11-14 Phase 2 clinical trial design, efficacy, and AEs are presented for risankizumab in eTable 3^15-18 and for certolizumab pegol in eTable 4.^17,19 No published clinical trial data were found for bimekizumab.

IL-17 Inhibitors

Ixekizumab
This recombinant, high-affinity IgG4κ antibody selectively binds and neutralizes IL-17A.^5,6 Three phase 3 clinical trials—UNCOVER-1, UNCOVER-2, and UNCOVER-3—evaluated ixekizumab for moderate to severe plaque psoriasis.⁷

The 3 UNCOVER trials were randomized, double-blind, phase 3 trials of 1296, 1224, and 1346 patients, respectively, assigned to a placebo group; a group treated with ixekizumab 80 mg every 2 weeks; and a group treated with ixekizumab 80 mg every 4 weeks. Both ixekizumab groups received a loading dose of 160 mg at week 0.^5,6 UNCOVER-2 and UNCOVER-3 also included a comparator group of patients on etanercept 50 mg.⁵ Co-primary end points included the percentage of patients reaching a psoriasis area and severity index (PASI) of 75 and with a static physician global assessment (PGA) score of clear (0) or almost clear (1) at week 12.^5,6

Ixekizumab achieved greater efficacy than placebo: 89.1%, 89.7%, and 87.3% of patients achieved PASI 75 in the every 2-week dosing group, and 82.6%, 77.5% and 84.2% achieved PASI 75 in the every 4-week dosing group in UNCOVER-1, UNCOVER-2, and UNCOVER-3, respectively (P<.001 for both treatment arms compared to placebo in all trials). The percentage of patients achieving a static PGA score of 0 or 1 also was higher in the ixekizumab groups in the 2-week and 4-week dosing groups in all UNCOVER trials—81.8% and 76.4% in UNCOVER-1, 83.2% and 72.9% in UNCOVER-2, and 80.5% and 75.4% in UNCOVER-3—compared to 3.2%, 2.4%, and 6.7% in the placebo groups of the 3 trials (P<.001 for both ixekizumab groups compared to placebo in all trials).^5,6 Ixekizumab also was found to be more effective than etanercept for both co-primary end points in both UNCOVER-2 and UNCOVER-3 (eTable 1).⁵

Safety data for all UNCOVER trials were pooled and reported.⁶ At week 12 the rate of at least 1 AE was 58.4% in patients on ixekizumab every 2 weeks and 58.8% in patients on ixekizumab every 4 weeks compared to 54.0% in the etanercept group in UNCOVER-2 and UNCOVER-3 and 46.8% in the placebo group. At week 12, 72 nonfatal serious AEs were reported: 12 in the placebo group, 14 in the etanercept group, 20 in the ixekizumab every 2 weeks group, and 26 in the ixekizumab every 4 weeks group.⁶

The most common AE across all groups was nasopharyngitis. Overall, infections were more frequent in patients treated with ixekizumab than in patients treated with placebo or etanercept. Specifically, oral candidiasis occurred more frequently in the ixekizumab groups, with a higher rate in the 2-week dosing group than in the 4-week dosing group.⁶ Two myocardial infarctions (MIs) occurred: 1 in the etanercept group and 1 in the placebo group.⁵

Brodalumab
This human monoclonal antibody binds to IL-17ra.^8,9 Three double-blind, placebo-controlled, phase 3 trials—AMAGINE-1, AMAGINE-2, and AMAGINE-3—evaluated its use for plaque psoriasis.¹⁰

In AMAGINE-1 (N=661), patients were randomized to receive brodalumab 140 mg or 210 mg (every 2 weeks for 12 weeks), or placebo.⁸ In AMAGINE-2 (N=1831) and AMAGINE-3 (N=1881), patients were randomized to receive brodalumab 140 mg or 210 mg (every 2 weeks for 12 weeks), ustekinumab 45 mg or 90 mg by weight (at weeks 0 and 4, then every 12 weeks thereafter), or placebo. In all trials, patients on brodalumab received a dose at week 0 and week 1. Co-primary end points were PASI 75 and a static PGA score of 0 or 1 at 12 weeks compared to placebo and to ustekinumab (in AMAGINE-2 and AMAGINE-3 only).⁸

At week 12, 83.3%, 86.3%, and 85.1% of patients on brodalumab 210 mg, and 60.3%, 66.6%, and 69.2% of patients on brodalumab 140 mg, achieved PASI 75 in AMAGINE-1, AMAGINE-2, and AMAGINE-3, respectively, compared to 2.7%, 8.1%, and 6.0% in the placebo groups (P<.001 between both brodalumab groups and placebo in all trials).⁸ Both brodalumab groups were noninferior but not significantly superior to ustekinumab, which achieved a PASI 75 of 70.0% in AMAGINE-2 and 69.3% in AMAGINE-3. The PASI 90 rate was higher, however, in both brodalumab groups compared to ustekinumab but significance was not reported (eTable 1).⁹ For both brodalumab groups, significantly more patients achieved a static PGA value of 0 or 1 compared to placebo (P<.001 across all trials). However, only the brodalumab 210-mg group achieved a significantly higher rate of static PGA 0 or 1 compared to ustekinumab in AMAGINE-2 and AMAGINE-3 (P<.001).⁹

After 12 weeks, the percentage of patients reporting at least 1 AE was 59.0%, 57.8%, and 56.8% in the brodalumab 210-mg group in AMAGINE-1, AMAGINE-2, and AMAGINE-3, respectively; 58.0%, 60.1%, and 52.6% in the brodalumab 140-mg group; and 51.0%, 53.4%, and 48.6% in the placebo group. Patients taking ustekinumab had an AE rate of 59.0% in AMAGINE-2 and 53.7% in AMAGINE-3. The most common AE was nasopharyngitis, followed by upper respiratory infection (URI) and headache across all trials.^8,9 Serious AEs were rare: 10 in AMAGINE-1, 31 in AMAGINE-2, and 24 in AMAGINE-3 across all groups. One death occurred from stroke in the brodalumab 210-mg group in AMAGINE-2.⁹

IL-23 Inhibitors

Guselkumab
This drug is a human IgG1κ antibody that binds to the p19 subunit of IL-23, thereby inhibiting IL-23 signaling.^11,12 Guselkumab was approved by the FDA in July 2017 for moderate to severe plaque psoriasis.¹³

VOYAGE 1 and VOYAGE 2 were phase 3, double-blind, placebo- and active comparator–controlled trials of 837 and 992 patients, respectively, randomized to receive adalimumab (80 mg at week 0 and 40 mg at week 1, then at 40 mg every 2 weeks thereafter), guselkumab 100 mg at weeks 0, 4, and 12, or placebo.¹¹ Co-primary end points for both trials were the percentage of patients reaching PASI 90 and an investigator global assessment (IGA) score of cleared (0) or minimal (1) at week 16.¹¹

By week 16 of both trials, PASI 90 values were statistically superior for guselkumab (VOYAGE 1, 73.3%; VOYAGE 2, 70.0%) compared to adalimumab (VOYAGE 1, 49.7%; VOYAGE 2, 46.8%) and placebo (VOYAGE 1, 2.9%; VOYAGE 2, 2.4%)(P<.001). Moreover, patients on guselkumab achieved a higher rate of IGA values of 0 and 1 at week 12 (85.1% in VOYAGE 1 and 84.1% in VOYAGE 2) than patients on adalimumab (65.9% in VOYAGE 1 and 67.7% in VOYAGE 2) and placebo (6.9% in VOYAGE 1 and 8.5% in VOYAGE 2)(P<.001).^11,12

The frequency of AEs was comparable across all groups in both trials.^11,12 During the 16-week treatment period, 51.7% and 47.6% of the guselkumab groups in VOYAGE 1 and VOYAGE 2, respectively; 51.1% and 48.4% of the adalimumab groups; and 49.4% and 44.8% of the placebo groups reported at least 1 AE. The most common AEs in all groups were nasopharyngitis, headache, and URI.^11,12

Serious AEs also occurred at similar rates: 2.4% and 1.6% in the guselkumab group in VOYAGE 1 and VOYAGE 2, respectively; 2.4% and 1.8% in the adalimumab group; and 1.7% and 1.2% in the placebo group.^11,12 One case of malignancy occurred in the VOYAGE 1 trial: basal cell carcinoma in the guselkumab group.¹¹ Three major cardiovascular events occurred across both trials: 1 MI in the guselkumab group in each trial and 1 MI in the adalimumab group in VOYAGE 1.^11,12

Tildrakizumab
A high-affinity, humanized IgG1κ antibody, tildrakizumab targets the p19 subunit of IL-23. As of February 2018, 2 double-blind, randomized phase 3 trials have studied tildrakizumab with published results: reSURFACE 1 and reSURFACE 2.¹⁴

reSURFACE 1 (N=772) and reSURFACE 2 (N=1090) randomized patients to receive tildrakizumab 100 or 200 mg (at weeks 0 and 4), etanercept 50 mg (twice weekly) for 12 weeks (reSURFACE 2 only), or placebo. Co-primary end points were the percentage of patients achieving PASI 75 and the percentage of patients achieving a PGA score of 0 or 1 at week 12.¹⁴

In reSURFACE 1, significantly more patients receiving tildrakizumab attained PASI 75 at week 12 compared to placebo: 200 mg, 62.0%; 100 mg, 64.0%; and placebo, 6.0% (P<.001 for tildrakizumab groups compared to placebo). Moreover, significantly proportionally more patients received a PGA score of 0 or 1 compared to placebo: 100 mg, 59%; 200 mg, 58.0%; placebo, 7.0% (P<.001 for both tildrakizumab groups compared to placebo).¹⁴

In reSURFACE 2, significantly more patients receiving tildrakizumab achieved PASI 75 compared to etanercept and placebo at week 12: 200 mg, 66.0%; 100mg, 61.0%; etanercept, 48.0%; placebo, 6.0% (P<.001 for both tildrakizumab groups compared to placebo; P<.05 for both tildrakizumab groups compared to etanercept). Additionally, significantly more patients in the tildrakizumab groups experienced a PGA score of 0 or 1 at week 12 compared to placebo: 200 mg, 59%; 100 mg, 55.0%; placebo, 5% (P<.001 for both tildrakizumab groups compared to placebo).¹⁴

Adverse events were reported at a similar rate across all groups. For reSURFACE 1 and reSURFACE 2, at least 1 AE by week 12 was reported by 42.2% and 45.2% of patients in the 200-mg group; 47.2% and 45.9% in the 100-mg group; and 48.1% and 55.1% in the placebo groups.¹⁴The most common AEs were nasopharyngitis, URI (reSURFACE 1), and erythema at the injection site (reSURFACE 2). One case of serious infection was reported in each of the tildrakizumab groups: 1 case of drug-related hypersensitivity reaction in the 200-mg group, and 1 major cardiovascular event in the 100-mg group of reSURFACE 1. There was 1 serious AE in reSURFACE 2 that led to death in which the cause was undetermined.¹⁴

Risankizumab
This humanized IgG1 antibody binds the p19 unit of IL-23.^15,16 The drug is undergoing 3 phase 3 trials—ultIMMa-1, ultIMMa-2, and IMMvent—for which only preliminary data have been published and are reported here.^16,17 There is 1 phase 2 randomized, dose-ranging trial with published data.¹⁵

ultIMMa-1 and ultIMMa-2 comprised 506 and 491 patients, respectively, randomized to receive risankizumab (150 mg at weeks 0, 4, and 16), ustekinumab (45 mg or 90 mg, by weight, at weeks 0, 4, and 16), or placebo. Co-primary end points were PASI 90 and a PGA score of 0 or 1 at week 16.¹⁷

In ultIMMa-1 and ultIMMa-2, 75.0% and 75.0% of patients on risankizumab 150 mg achieved PASI 90 compared to 42.0% and 48.0% on ustekinumab and 5.0% and 2.0% on placebo at 16 weeks (P<.001 between both placebo and ustekinumab in both trials).¹⁷ In both trials, patients receiving risankizumab achieved higher rates of a static PGA score of 0 or 1 (88.0% and 84.0%) compared to ustekinumab (63.0% and 62.0%) and placebo (8.0% and 5.0%) at 16 weeks (P<.001 for both trials).¹⁸

At week 16, 2.0% of patients on risankizumab reported a serious AE in both trials, compared to 8.0% and 3.0% of patients on ustekinumab and 3.0% and 1.0% on placebo. No new safety concerns were noted.¹⁷

In the phase 3 IMMvent trial, 605 patients were randomized to receive risankizumab (150 mg at weeks 0, 4, and 16) or adalimumab (80 mg at week 0, 40 mg at week 1, then 40 mg every 2 weeks). Co-primary end points were PASI 90 and a static PGA score of 0 or 1 at week 16.¹⁷

In IMMvent, risankizumab was significantly more effective than adalimumab for PASI 75 (risankizumab, 72.0%; adalimumab, 47.0%) and a static PGA score of 0 or 1 (risankizumab 84.0%; adalimumab, 60.0%) (P<.001 risankizumab compared to adalimumab for both end points).¹⁷

At week 16, serious AEs were reported in 3.0% of patients on risankizumab and 3.0% of patients on adalimumab. One patient receiving risankizumab died of an acute MI during the treatment phase.¹⁷

TNF Inhibitor

Certolizumab Pegol
Certolizumab pegol is a human PEGylated anti-TNF agent. In vitro studies have shown that certolizumab binds to soluble and membrane-bound TNF.¹⁹ Unlike other TNF inhibitors, certolizumab pegol is a Fab‘ portion of anti-TNF conjugated to a molecule of polyethylene glycol.¹⁹ The drug is approved in the United States for treating psoriatic arthritis, Crohn disease, and rheumatoid arthritis; its potential for treating psoriasis has been confirmed. Results of 1 phase 2 trial have been published¹⁹; data from 3 phase 3 trials are forthcoming.

This randomized, placebo-controlled, double-blind phase 2 study comprised 176 patients who received certolizumab 200 mg, certolizumab 400 mg, or placebo. The dosing schedule was 400 mg at week 0, followed by either 200 or 400 mg every other week until week 10. Co-primary end points were PASI 75 and a PGA score of 0 or 1 at week 12.¹⁹

Certolizumab was significantly more effective than placebo at week 12: 74.6% of the 200-mg group and 82.8% of the 400-mg group achieved PASI 75 compared to 6.8% of the placebo group (P<.001). Certolizumab also performed better for the PGA score: 52.5% and 72.4% of patients attained a score of 0 or 1 in the 200-mg and 400-mg groups compared to 1.7% in the placebo group.¹⁹

Adverse events were reported equally across all groups: 72% of patients in the 200-mg group, 70% in the 400-mg group, and 71% in the placebo group reported at least 1 AE, most commonly nasopharyngitis, headache, and pruritis.¹⁹

COMMENT

With the development of new insights into the pathogenesis of psoriasis, therapies that are targeted toward key cytokines may contribute to improved management of the disease. The results of these clinical trials demonstrate numerous promising options for psoriatic patients.

IL-17 Inhibitors Ixekizumab and Brodalumab

When comparing these 2 biologics, it is important to consider that these studies were not performed head to head, thereby inhibiting direct comparisons. Moreover, dosage ranges of the investigative drugs were not identical, which also makes comparisons challenging. However, when looking at the highest dosages of ixekizumab and brodalumab, results indicate that ixekizumab may be slightly more effective than brodalumab based on the percentage of patients who achieved a PASI 75 and a static PGA score of 0 or 1 (eTable 1).

Phase 3 trials have shown ixekizumab to maintain efficacy over 60 weeks of treatment.⁶ Ixekizumab also has been shown to alleviate other symptoms of psoriasis, such as itching, pain, and nail involvement.^20,21 Furthermore, ixekizumab appears to be equally effective in patients with or without prior exposure to biologics²²; therefore, ixekizumab may benefit patients who have not experienced success with other biologics.

Across the UNCOVER trials, 11 cases of inflammatory bowel disease were reported in patients receiving ixekizumab (ulcerative colitis in 7; Crohn disease in 4)⁶; it appears that at least 3 of these cases were new diagnoses. In light of a study suggesting that IL-17A might have a protective function in the intestine,²³ these findings may have important clinical implications and require follow-up studies.

Brodalumab also has been shown to maintain efficacy and acceptable safety for as long as 120 weeks.²⁴ In the extension period of the AMAGINE-1 trial, patients who experienced a return of disease during a withdrawal period recaptured static PGA success with re-treatment for 12 weeks (re-treatment was successful in 97% of those given a dosage of 210 mg and in 84% of those given 140 mg).⁸

Furthermore, phase 2 trials also have shown that brodalumab is effective in patients with a history of biologic use.²⁵ Across all AMAGINE trials, only 1 case of Crohn disease was reported in a patient taking brodalumab.⁹ There are concerns about depression, despite data from AMAGINE-1 stating patients on brodalumab actually had greater improvements in Hospital Anxiety and Depression Scale scores after 12 weeks of treatment (P<.001) for both brodalumab 140 mg and 210 mg compared to placebo.⁸ Regardless, brodalumab has a black-box warning for suicidal ideation and behavior, and availability is restricted through a Risk Evaluation and Mitigation Strategy (REMS) program.²⁶

Bimekizumab

Although no phase 2 or phase 3 clinical trial data have been published for bimekizumab (phase 2 trials are underway), it has been shown in a phase 1 trial to be effective for psoriasis. Bimekizumab also is unique; it is the first dual inhibitor of IL-17A and IL-17F.¹⁸

IL-23 Inhibitors Guselkumab, Tildrakizumab, and Risankizumab

Making comparisons among the IL-23 inhibitors also is difficult; studies were not head-to-head comparison trials, and the VOYAGE and reSURFACE studies used different time points for primary end points. Furthermore, only phase 2 trial data are available for risankizumab. Despite these limitations, results of these trials suggest that guselkumab and risankizumab may be slightly more efficacious than tildrakizumab. However, future studies, including head-to-head studies, would ultimately provide further information on how these agents compare.

Guselkumab was shown to remain efficacious at 48 weeks, though patients on maintenance dosing had better results than those who were re-treated.¹² Moreover, guselkumab was found to be effective in hard-to-treat areas, such as the scalp,¹¹ and in patients who did not respond to adalimumab. Guselkumab may therefore benefit patients who have experienced limited clinical improvement on other biologics.¹²

Tildrakizumab was shown to improve PASI 75 and PGA scores through week 28 of treatment. Moreover, a higher percentage of patients taking tildrakizumab scored 0 or 1 on the dermatology life quality index, suggesting that the drug improves quality of life.¹⁴ No specific safety concerns arose in either reSURFACE trial; however, long-term studies are needed for further evaluation.

Risankizumab appears to be a promising new therapy based on phase 2 trial results. Improvements also were seen in dermatology life quality index scores, scalp and fingernail symptoms, and palmoplantar psoriasis.¹⁵ Of note, neutralizing antidrug antibodies were found in 3 patients during this study,¹⁵ which may present potential problems for long-term efficacy. However, preliminary data from 3 phase 3 trials—ultIMMa-1, ultIMMa-2, and IMMvent—are promising.¹⁷

CONCLUSION

Advances in the understanding of psoriasis have led to new targeted therapies. Ongoing clinical trials have shown encouraging results for treating physical and psychological symptoms of psoriasis. The findings of these trials support the idea that therapies targeting IL-23, specifically its p19 subunit, are effective against psoriasis while sparing IL-12. Long-term data from open-label extension studies would help guide clinical recommendations regarding the safety profiles of these agents and determine their long-term utility.

Psoriasis is a chronic, autoimmune-mediated disease estimated to affect 2.8% of the US population.¹ The pathogenesis of psoriasis is thought to involve a complex process triggered by a combination of genetic and environmental factors that induce tumor necrosis factor (TNF) α secretion by keratinocytes, which in turn activates dendritic cells. Activated dendritic cells produce IL-23, leading to helper T cell (T_H17) differentiation.^2,3 T_H17 cells secrete IL-17A, which has been shown to promote psoriatic skin changes.⁴ Therefore, TNF-α, IL-23, and IL-17A have been recognized as key targets for psoriasis therapy.

The newest biologic agents targeting IL-17–mediated pathways include ixekizumab, brodalumab, and bimekizumab. Secukinumab, the first US Food and Drug Administration (FDA)–approved IL-17 inhibitor, has been available since 2015 and therefore is not included in this review. IL-23 inhibitors that are FDA approved or being evaluated in clinical trials include guselkumab, tildrakizumab, and risankizumab. In addition, certolizumab pegol, a TNF-α inhibitor, is being studied for use in psoriasis.

METHODS

We reviewed the published results of phase 3 clinical trials for ixekizumab, brodalumab, bimekizumab, guselkumab, tildrakizumab, risankizumab, and certolizumab pegol. We performed an English-language literature search (January 1, 2012 to October 15, 2017) of articles indexed for PubMed/MEDLINE using the following combinations of keywords: IL-23 and psoriasis; IL-17 and psoriasis; tumor necrosis factor and psoriasis; [drug name] and psoriasis. If data from phase 3 clinical trials were not yet available, data from phase 2 clinical trials were incorporated in our analysis. We also reviewed citations within articles to identify relevant sources.

RESULTS

Phase 3 clinical trial design, efficacy, and adverse events (AEs) for ixekizumab and brodalumab are reported in eTable 1^5-10 and for guselkumab and tildrakizumab in eTable 2.^11-14 Phase 2 clinical trial design, efficacy, and AEs are presented for risankizumab in eTable 3^15-18 and for certolizumab pegol in eTable 4.^17,19 No published clinical trial data were found for bimekizumab.

IL-17 Inhibitors

Ixekizumab
This recombinant, high-affinity IgG4κ antibody selectively binds and neutralizes IL-17A.^5,6 Three phase 3 clinical trials—UNCOVER-1, UNCOVER-2, and UNCOVER-3—evaluated ixekizumab for moderate to severe plaque psoriasis.⁷

The 3 UNCOVER trials were randomized, double-blind, phase 3 trials of 1296, 1224, and 1346 patients, respectively, assigned to a placebo group; a group treated with ixekizumab 80 mg every 2 weeks; and a group treated with ixekizumab 80 mg every 4 weeks. Both ixekizumab groups received a loading dose of 160 mg at week 0.^5,6 UNCOVER-2 and UNCOVER-3 also included a comparator group of patients on etanercept 50 mg.⁵ Co-primary end points included the percentage of patients reaching a psoriasis area and severity index (PASI) of 75 and with a static physician global assessment (PGA) score of clear (0) or almost clear (1) at week 12.^5,6

Ixekizumab achieved greater efficacy than placebo: 89.1%, 89.7%, and 87.3% of patients achieved PASI 75 in the every 2-week dosing group, and 82.6%, 77.5% and 84.2% achieved PASI 75 in the every 4-week dosing group in UNCOVER-1, UNCOVER-2, and UNCOVER-3, respectively (P<.001 for both treatment arms compared to placebo in all trials). The percentage of patients achieving a static PGA score of 0 or 1 also was higher in the ixekizumab groups in the 2-week and 4-week dosing groups in all UNCOVER trials—81.8% and 76.4% in UNCOVER-1, 83.2% and 72.9% in UNCOVER-2, and 80.5% and 75.4% in UNCOVER-3—compared to 3.2%, 2.4%, and 6.7% in the placebo groups of the 3 trials (P<.001 for both ixekizumab groups compared to placebo in all trials).^5,6 Ixekizumab also was found to be more effective than etanercept for both co-primary end points in both UNCOVER-2 and UNCOVER-3 (eTable 1).⁵

Safety data for all UNCOVER trials were pooled and reported.⁶ At week 12 the rate of at least 1 AE was 58.4% in patients on ixekizumab every 2 weeks and 58.8% in patients on ixekizumab every 4 weeks compared to 54.0% in the etanercept group in UNCOVER-2 and UNCOVER-3 and 46.8% in the placebo group. At week 12, 72 nonfatal serious AEs were reported: 12 in the placebo group, 14 in the etanercept group, 20 in the ixekizumab every 2 weeks group, and 26 in the ixekizumab every 4 weeks group.⁶

The most common AE across all groups was nasopharyngitis. Overall, infections were more frequent in patients treated with ixekizumab than in patients treated with placebo or etanercept. Specifically, oral candidiasis occurred more frequently in the ixekizumab groups, with a higher rate in the 2-week dosing group than in the 4-week dosing group.⁶ Two myocardial infarctions (MIs) occurred: 1 in the etanercept group and 1 in the placebo group.⁵

Brodalumab
This human monoclonal antibody binds to IL-17ra.^8,9 Three double-blind, placebo-controlled, phase 3 trials—AMAGINE-1, AMAGINE-2, and AMAGINE-3—evaluated its use for plaque psoriasis.¹⁰

In AMAGINE-1 (N=661), patients were randomized to receive brodalumab 140 mg or 210 mg (every 2 weeks for 12 weeks), or placebo.⁸ In AMAGINE-2 (N=1831) and AMAGINE-3 (N=1881), patients were randomized to receive brodalumab 140 mg or 210 mg (every 2 weeks for 12 weeks), ustekinumab 45 mg or 90 mg by weight (at weeks 0 and 4, then every 12 weeks thereafter), or placebo. In all trials, patients on brodalumab received a dose at week 0 and week 1. Co-primary end points were PASI 75 and a static PGA score of 0 or 1 at 12 weeks compared to placebo and to ustekinumab (in AMAGINE-2 and AMAGINE-3 only).⁸

At week 12, 83.3%, 86.3%, and 85.1% of patients on brodalumab 210 mg, and 60.3%, 66.6%, and 69.2% of patients on brodalumab 140 mg, achieved PASI 75 in AMAGINE-1, AMAGINE-2, and AMAGINE-3, respectively, compared to 2.7%, 8.1%, and 6.0% in the placebo groups (P<.001 between both brodalumab groups and placebo in all trials).⁸ Both brodalumab groups were noninferior but not significantly superior to ustekinumab, which achieved a PASI 75 of 70.0% in AMAGINE-2 and 69.3% in AMAGINE-3. The PASI 90 rate was higher, however, in both brodalumab groups compared to ustekinumab but significance was not reported (eTable 1).⁹ For both brodalumab groups, significantly more patients achieved a static PGA value of 0 or 1 compared to placebo (P<.001 across all trials). However, only the brodalumab 210-mg group achieved a significantly higher rate of static PGA 0 or 1 compared to ustekinumab in AMAGINE-2 and AMAGINE-3 (P<.001).⁹

After 12 weeks, the percentage of patients reporting at least 1 AE was 59.0%, 57.8%, and 56.8% in the brodalumab 210-mg group in AMAGINE-1, AMAGINE-2, and AMAGINE-3, respectively; 58.0%, 60.1%, and 52.6% in the brodalumab 140-mg group; and 51.0%, 53.4%, and 48.6% in the placebo group. Patients taking ustekinumab had an AE rate of 59.0% in AMAGINE-2 and 53.7% in AMAGINE-3. The most common AE was nasopharyngitis, followed by upper respiratory infection (URI) and headache across all trials.^8,9 Serious AEs were rare: 10 in AMAGINE-1, 31 in AMAGINE-2, and 24 in AMAGINE-3 across all groups. One death occurred from stroke in the brodalumab 210-mg group in AMAGINE-2.⁹

IL-23 Inhibitors

Guselkumab
This drug is a human IgG1κ antibody that binds to the p19 subunit of IL-23, thereby inhibiting IL-23 signaling.^11,12 Guselkumab was approved by the FDA in July 2017 for moderate to severe plaque psoriasis.¹³

VOYAGE 1 and VOYAGE 2 were phase 3, double-blind, placebo- and active comparator–controlled trials of 837 and 992 patients, respectively, randomized to receive adalimumab (80 mg at week 0 and 40 mg at week 1, then at 40 mg every 2 weeks thereafter), guselkumab 100 mg at weeks 0, 4, and 12, or placebo.¹¹ Co-primary end points for both trials were the percentage of patients reaching PASI 90 and an investigator global assessment (IGA) score of cleared (0) or minimal (1) at week 16.¹¹

By week 16 of both trials, PASI 90 values were statistically superior for guselkumab (VOYAGE 1, 73.3%; VOYAGE 2, 70.0%) compared to adalimumab (VOYAGE 1, 49.7%; VOYAGE 2, 46.8%) and placebo (VOYAGE 1, 2.9%; VOYAGE 2, 2.4%)(P<.001). Moreover, patients on guselkumab achieved a higher rate of IGA values of 0 and 1 at week 12 (85.1% in VOYAGE 1 and 84.1% in VOYAGE 2) than patients on adalimumab (65.9% in VOYAGE 1 and 67.7% in VOYAGE 2) and placebo (6.9% in VOYAGE 1 and 8.5% in VOYAGE 2)(P<.001).^11,12

The frequency of AEs was comparable across all groups in both trials.^11,12 During the 16-week treatment period, 51.7% and 47.6% of the guselkumab groups in VOYAGE 1 and VOYAGE 2, respectively; 51.1% and 48.4% of the adalimumab groups; and 49.4% and 44.8% of the placebo groups reported at least 1 AE. The most common AEs in all groups were nasopharyngitis, headache, and URI.^11,12

Serious AEs also occurred at similar rates: 2.4% and 1.6% in the guselkumab group in VOYAGE 1 and VOYAGE 2, respectively; 2.4% and 1.8% in the adalimumab group; and 1.7% and 1.2% in the placebo group.^11,12 One case of malignancy occurred in the VOYAGE 1 trial: basal cell carcinoma in the guselkumab group.¹¹ Three major cardiovascular events occurred across both trials: 1 MI in the guselkumab group in each trial and 1 MI in the adalimumab group in VOYAGE 1.^11,12

Tildrakizumab
A high-affinity, humanized IgG1κ antibody, tildrakizumab targets the p19 subunit of IL-23. As of February 2018, 2 double-blind, randomized phase 3 trials have studied tildrakizumab with published results: reSURFACE 1 and reSURFACE 2.¹⁴

reSURFACE 1 (N=772) and reSURFACE 2 (N=1090) randomized patients to receive tildrakizumab 100 or 200 mg (at weeks 0 and 4), etanercept 50 mg (twice weekly) for 12 weeks (reSURFACE 2 only), or placebo. Co-primary end points were the percentage of patients achieving PASI 75 and the percentage of patients achieving a PGA score of 0 or 1 at week 12.¹⁴

In reSURFACE 1, significantly more patients receiving tildrakizumab attained PASI 75 at week 12 compared to placebo: 200 mg, 62.0%; 100 mg, 64.0%; and placebo, 6.0% (P<.001 for tildrakizumab groups compared to placebo). Moreover, significantly proportionally more patients received a PGA score of 0 or 1 compared to placebo: 100 mg, 59%; 200 mg, 58.0%; placebo, 7.0% (P<.001 for both tildrakizumab groups compared to placebo).¹⁴

In reSURFACE 2, significantly more patients receiving tildrakizumab achieved PASI 75 compared to etanercept and placebo at week 12: 200 mg, 66.0%; 100mg, 61.0%; etanercept, 48.0%; placebo, 6.0% (P<.001 for both tildrakizumab groups compared to placebo; P<.05 for both tildrakizumab groups compared to etanercept). Additionally, significantly more patients in the tildrakizumab groups experienced a PGA score of 0 or 1 at week 12 compared to placebo: 200 mg, 59%; 100 mg, 55.0%; placebo, 5% (P<.001 for both tildrakizumab groups compared to placebo).¹⁴

Adverse events were reported at a similar rate across all groups. For reSURFACE 1 and reSURFACE 2, at least 1 AE by week 12 was reported by 42.2% and 45.2% of patients in the 200-mg group; 47.2% and 45.9% in the 100-mg group; and 48.1% and 55.1% in the placebo groups.¹⁴The most common AEs were nasopharyngitis, URI (reSURFACE 1), and erythema at the injection site (reSURFACE 2). One case of serious infection was reported in each of the tildrakizumab groups: 1 case of drug-related hypersensitivity reaction in the 200-mg group, and 1 major cardiovascular event in the 100-mg group of reSURFACE 1. There was 1 serious AE in reSURFACE 2 that led to death in which the cause was undetermined.¹⁴

Risankizumab
This humanized IgG1 antibody binds the p19 unit of IL-23.^15,16 The drug is undergoing 3 phase 3 trials—ultIMMa-1, ultIMMa-2, and IMMvent—for which only preliminary data have been published and are reported here.^16,17 There is 1 phase 2 randomized, dose-ranging trial with published data.¹⁵

ultIMMa-1 and ultIMMa-2 comprised 506 and 491 patients, respectively, randomized to receive risankizumab (150 mg at weeks 0, 4, and 16), ustekinumab (45 mg or 90 mg, by weight, at weeks 0, 4, and 16), or placebo. Co-primary end points were PASI 90 and a PGA score of 0 or 1 at week 16.¹⁷

In ultIMMa-1 and ultIMMa-2, 75.0% and 75.0% of patients on risankizumab 150 mg achieved PASI 90 compared to 42.0% and 48.0% on ustekinumab and 5.0% and 2.0% on placebo at 16 weeks (P<.001 between both placebo and ustekinumab in both trials).¹⁷ In both trials, patients receiving risankizumab achieved higher rates of a static PGA score of 0 or 1 (88.0% and 84.0%) compared to ustekinumab (63.0% and 62.0%) and placebo (8.0% and 5.0%) at 16 weeks (P<.001 for both trials).¹⁸

At week 16, 2.0% of patients on risankizumab reported a serious AE in both trials, compared to 8.0% and 3.0% of patients on ustekinumab and 3.0% and 1.0% on placebo. No new safety concerns were noted.¹⁷

In the phase 3 IMMvent trial, 605 patients were randomized to receive risankizumab (150 mg at weeks 0, 4, and 16) or adalimumab (80 mg at week 0, 40 mg at week 1, then 40 mg every 2 weeks). Co-primary end points were PASI 90 and a static PGA score of 0 or 1 at week 16.¹⁷

In IMMvent, risankizumab was significantly more effective than adalimumab for PASI 75 (risankizumab, 72.0%; adalimumab, 47.0%) and a static PGA score of 0 or 1 (risankizumab 84.0%; adalimumab, 60.0%) (P<.001 risankizumab compared to adalimumab for both end points).¹⁷

At week 16, serious AEs were reported in 3.0% of patients on risankizumab and 3.0% of patients on adalimumab. One patient receiving risankizumab died of an acute MI during the treatment phase.¹⁷

TNF Inhibitor

Certolizumab Pegol
Certolizumab pegol is a human PEGylated anti-TNF agent. In vitro studies have shown that certolizumab binds to soluble and membrane-bound TNF.¹⁹ Unlike other TNF inhibitors, certolizumab pegol is a Fab‘ portion of anti-TNF conjugated to a molecule of polyethylene glycol.¹⁹ The drug is approved in the United States for treating psoriatic arthritis, Crohn disease, and rheumatoid arthritis; its potential for treating psoriasis has been confirmed. Results of 1 phase 2 trial have been published¹⁹; data from 3 phase 3 trials are forthcoming.

This randomized, placebo-controlled, double-blind phase 2 study comprised 176 patients who received certolizumab 200 mg, certolizumab 400 mg, or placebo. The dosing schedule was 400 mg at week 0, followed by either 200 or 400 mg every other week until week 10. Co-primary end points were PASI 75 and a PGA score of 0 or 1 at week 12.¹⁹

Certolizumab was significantly more effective than placebo at week 12: 74.6% of the 200-mg group and 82.8% of the 400-mg group achieved PASI 75 compared to 6.8% of the placebo group (P<.001). Certolizumab also performed better for the PGA score: 52.5% and 72.4% of patients attained a score of 0 or 1 in the 200-mg and 400-mg groups compared to 1.7% in the placebo group.¹⁹

Adverse events were reported equally across all groups: 72% of patients in the 200-mg group, 70% in the 400-mg group, and 71% in the placebo group reported at least 1 AE, most commonly nasopharyngitis, headache, and pruritis.¹⁹

COMMENT

With the development of new insights into the pathogenesis of psoriasis, therapies that are targeted toward key cytokines may contribute to improved management of the disease. The results of these clinical trials demonstrate numerous promising options for psoriatic patients.

IL-17 Inhibitors Ixekizumab and Brodalumab

When comparing these 2 biologics, it is important to consider that these studies were not performed head to head, thereby inhibiting direct comparisons. Moreover, dosage ranges of the investigative drugs were not identical, which also makes comparisons challenging. However, when looking at the highest dosages of ixekizumab and brodalumab, results indicate that ixekizumab may be slightly more effective than brodalumab based on the percentage of patients who achieved a PASI 75 and a static PGA score of 0 or 1 (eTable 1).

Phase 3 trials have shown ixekizumab to maintain efficacy over 60 weeks of treatment.⁶ Ixekizumab also has been shown to alleviate other symptoms of psoriasis, such as itching, pain, and nail involvement.^20,21 Furthermore, ixekizumab appears to be equally effective in patients with or without prior exposure to biologics²²; therefore, ixekizumab may benefit patients who have not experienced success with other biologics.

Across the UNCOVER trials, 11 cases of inflammatory bowel disease were reported in patients receiving ixekizumab (ulcerative colitis in 7; Crohn disease in 4)⁶; it appears that at least 3 of these cases were new diagnoses. In light of a study suggesting that IL-17A might have a protective function in the intestine,²³ these findings may have important clinical implications and require follow-up studies.

Brodalumab also has been shown to maintain efficacy and acceptable safety for as long as 120 weeks.²⁴ In the extension period of the AMAGINE-1 trial, patients who experienced a return of disease during a withdrawal period recaptured static PGA success with re-treatment for 12 weeks (re-treatment was successful in 97% of those given a dosage of 210 mg and in 84% of those given 140 mg).⁸

Furthermore, phase 2 trials also have shown that brodalumab is effective in patients with a history of biologic use.²⁵ Across all AMAGINE trials, only 1 case of Crohn disease was reported in a patient taking brodalumab.⁹ There are concerns about depression, despite data from AMAGINE-1 stating patients on brodalumab actually had greater improvements in Hospital Anxiety and Depression Scale scores after 12 weeks of treatment (P<.001) for both brodalumab 140 mg and 210 mg compared to placebo.⁸ Regardless, brodalumab has a black-box warning for suicidal ideation and behavior, and availability is restricted through a Risk Evaluation and Mitigation Strategy (REMS) program.²⁶

Bimekizumab

Although no phase 2 or phase 3 clinical trial data have been published for bimekizumab (phase 2 trials are underway), it has been shown in a phase 1 trial to be effective for psoriasis. Bimekizumab also is unique; it is the first dual inhibitor of IL-17A and IL-17F.¹⁸

IL-23 Inhibitors Guselkumab, Tildrakizumab, and Risankizumab

Making comparisons among the IL-23 inhibitors also is difficult; studies were not head-to-head comparison trials, and the VOYAGE and reSURFACE studies used different time points for primary end points. Furthermore, only phase 2 trial data are available for risankizumab. Despite these limitations, results of these trials suggest that guselkumab and risankizumab may be slightly more efficacious than tildrakizumab. However, future studies, including head-to-head studies, would ultimately provide further information on how these agents compare.

Guselkumab was shown to remain efficacious at 48 weeks, though patients on maintenance dosing had better results than those who were re-treated.¹² Moreover, guselkumab was found to be effective in hard-to-treat areas, such as the scalp,¹¹ and in patients who did not respond to adalimumab. Guselkumab may therefore benefit patients who have experienced limited clinical improvement on other biologics.¹²

Tildrakizumab was shown to improve PASI 75 and PGA scores through week 28 of treatment. Moreover, a higher percentage of patients taking tildrakizumab scored 0 or 1 on the dermatology life quality index, suggesting that the drug improves quality of life.¹⁴ No specific safety concerns arose in either reSURFACE trial; however, long-term studies are needed for further evaluation.

Risankizumab appears to be a promising new therapy based on phase 2 trial results. Improvements also were seen in dermatology life quality index scores, scalp and fingernail symptoms, and palmoplantar psoriasis.¹⁵ Of note, neutralizing antidrug antibodies were found in 3 patients during this study,¹⁵ which may present potential problems for long-term efficacy. However, preliminary data from 3 phase 3 trials—ultIMMa-1, ultIMMa-2, and IMMvent—are promising.¹⁷

CONCLUSION

Advances in the understanding of psoriasis have led to new targeted therapies. Ongoing clinical trials have shown encouraging results for treating physical and psychological symptoms of psoriasis. The findings of these trials support the idea that therapies targeting IL-23, specifically its p19 subunit, are effective against psoriasis while sparing IL-12. Long-term data from open-label extension studies would help guide clinical recommendations regarding the safety profiles of these agents and determine their long-term utility.

Deepithelialized flaps and grafts have been widely used by reconstructive surgeons in a diverse range of medical specialties since the early 20th century. ¹ These reconstructive modalities have more recently been applied to dermatologic surgery. Deepithelialized flaps and grafts involve removal of the epidermis from the dermis for a variety of surgical purposes. Although these techniques play an important role in dermatologic surgery, reports of application of deepithelialized flaps and grafts in the dermatology literature is limited. This article includes a presentation of the applications of deepithelialized flaps and grafts in procedural dermatology.

DEEPITHELIALIZATION TECHNIQUES

There are a variety of techniques for deepithelialization, although sharp deepithelialization generally is preferred by dermatologic surgeons. The scalpel technique can be accomplished by making an intradermal incision with a No. 15 blade. Traction is an essential component of the deepthelialization process and facilitates sharp removal of the epidermis and superficial dermis in an even plane. The peeling orange technique, which has been described in reduction mammoplasty, is a variant of the scalpel technique used for creating a large area of deepithelialized tissue.² A No. 10 blade is used to make multiple partial-thickness intradermal incisions 1 to 2 cm apart along the pedicle. Traction facilitates rapid deepithelialization of the skin strips on the pedicle. A sharp curette is an alternative option for sharply removing the epithelium from a small area. Electric dermatome, laser, and electrocautery techniques for deepithelialization also can be considered.^2,3

APPLICATION OF DEEPITHELIALIZED FLAPS

Deepithelialized flaps may be considered for single-stage reconstruction with tunneled interpolation flaps, reconstruction requiring contour preservation, and reconstruction involving free margins.^4-17

Reconstruction With Single-Stage Tunneled Interpolated Flaps

Alar Base
A partially deepithelialized tunneled interpolated flap is an elegant reconstructive option for defects involving the upper cutaneous lip and alar base. The flap is elevated from the ipsilateral nasolabial fold, deepithelialized proximally, and tunneled under the intact portion of the cutaneous upper lip and ala. The flap is then deepithelialized superiorly to bolster the alar base and inset at the recipient site.⁴

Nasal Ala
The tunneled interpolated flap is useful for reconstruction of defects of the nasal ala. A flap with a superior deepithelialized pedicle and an anticipated inferior Burow triangle is designed along the axis of the nasolabial fold. The inferior Burow triangle and central flap are elevated at the level of the superficial subcutaneous fat and the pedicle is dissected. The donor and recipient sites are widely undermined, and the flap and pedicle pass through the tunnel. The donor site is closed primarily, the inferior Burow triangle is trimmed, and the flap is sutured into the defect.⁵ This flap allows for preservation of free margins and favorable placement of incision lines. Furthermore, pincushioning of the flap helps to recreate the rounded shape of the lateral ala.⁶

Nasal Tip
Nasal tip defects can be repaired with a retroangular flap, centered on the angular artery. The flap is elevated along the axis of the nasolabial fold, deepithelialized at its proximal base, and transferred through a subcutaneous tunnel to the nasal tip. The angular artery is ligated at the inferior aspect of the flap.⁷

Nasal Sidewall
A deepithelialized tunneled interpolated forehead flap, similar to the classic paramedian forehead flap, can be used to reconstruct nasal sidewall defects. A flap is elevated on the contralateral forehead and the proximal portion is deepithelialized. A tunnel is then bluntly dissected just above the periosteum, and the flap is introduced into the defect through the tunnel and inset. This flap has the advantages of being a single-stage procedure, restoring volume to the defect area, and maintaining excellent vascular supply.⁸

Eyelid
A tunneled interpolated forehead flap also can be used to repair medial canthal defects and for anterior lamellar repair of lower eyelid defects. In a study of 9 patients receiving a tunneled interpolated forehead flap in these anatomic locations, all flaps demonstrated viability, protection of the globe, and preservation of the concave architecture of the medial canthus.⁹

Earlobe
Earlobe defects may be repaired with a pull-through interpolated preauricular flap. A flap is elevated superiorly in the preauricular region and the proximal aspect of the flap is deepithelialized. The flap is pulled through a tunnel and inset at the anterior earlobe defect. The donor site is closed primarily.^10,11

Concha
Reconstruction of anterior conchal defects with exposed cartilage can be accomplished with a pull-through interpolated postauricular flap based on the auriculomastoid fossa. The postauricular flap is elevated, the base is deepithelialized, an incision is made in the medial aspect of the defect, and the flap is moved through a tunnel between the posterior and anterior surfaces of the ear. The flap is secured to the anterior surface of the concha.¹²

Reconstruction Requiring Contour Preservation

Central Face
The hinge flap is optimal for reconstruction of deep central facial defects (Figure 1). The hinge flap is planned at a site contiguous with a margin of the defect and can include the dermis, subcutaneous tissue, muscle, or a combination of these. The desired tissue is folded over on the pedicle to fill the defect. Cutaneous coverage is accomplished through a primary closure, separate flap, or skin graft. In addition to restoring contour and therefore the cosmetic subunit, the hinge flap is performed in a single stage, resists wound contracture, and provides a well-vascularized wound bed resulting in a low incidence of graft failure.^13,14 Muscular hinge flaps have been described for reconstruction of forehead defects with exposed bone based on the frontalis muscle.¹⁵

Lower Lip
A variant of a V-Y advancement flap has been described for reconstruction of defects greater than one-third the length of the lower lip. The top of the “V” is deepithelialized and the flap is advanced such that the top of the “V” abuts the inferior border of the defect. The “V” flap is inset at its advanced position, converting the “V”-shaped wound into a “Y.” An overlying buccal mucosal graft provides reconstruction of the lower red lip and labial mucosa.¹⁶

Helix of the Ear
Large defects of the scapha and helix of the ear can be reconstructed with the use of a staged interpolated postauricular flap. The postauricular flap is elevated into a subcutaneous plane. A full-thickness incision is made medial to the helical rim, and the flap is tunneled through and sutured into place. The pedicle is later divided, and the distal aspect of the flap is deepithelialized and inset into the helical rim for volume restoration.¹⁷

Reconstruction Involving Free Margins

Nasal Ala
For large defects involving the upper cutaneous lip with adjacent alar base involvement, a partially deepithelialized V-Y flap is a useful reconstructive option (Figure 2).

Infraorbital Region
A deepithelialized variant of a V-Y advancement flap can be used for closure of infraorbital defects. The limbs of the V-Y flap are deepithelialized and anchored to the medial and lateral canthal tendons or periosteum. Ectropion prevention is the primary advantage of this flap.¹⁸

APPLICATION OF DEEPITHELIALIZED GRAFTS

Deepithelialized grafts may be considered for volume replacement, reconstruction requiring contour preservation, and restoration of mechanical integrity in areas of high mechanical tension.^3,19-21

Reconstruction Requiring Contour Preservation

Deepithelialized grafts are used to improve depressed nasal scars and restore volume in deep nasal wounds. One method involves deepithelialization of 2 postauricular punch biopsies. An 18-gauge needle is used to make a small hole in the depressed nasal scar, the dermal grafts are inserted, and the defect is closed primarily.¹⁹ Dermal grafts may be harvested from excess full-thickness skin grafts (FTSGs) or dog-ear tissue. When used under flaps, the dermal graft is trimmed to the size of the defect. When used under FTSGs, thin dermal graft strips are placed in a gridlike pattern to allow for revascularization. A study of 15 patients with contour deformities reconstructed with dermal graft insertions demonstrated that 14 (94%) patients had no significant complications and improvement of scar depression was achieved.²⁰

Reconstruction in Areas of High Mechanical Tension

Plantar Foot
A combined dermal and full-thickness sandwich graft has been described for reconstruction of plantar foot defects.³ The graft is created by obtaining a FTSG twice the size of the wound defect and deepithelializing half of the graft. The graft is then defatted and the deepithelialized portion is folded beneath the other half, allowing the papillary dermis to make contact with the wound surface.

Scalp
Dermal graft reconstruction for scalp defects may be accomplished with a split-thickness skin flap. The flap is harvested using an electronic dermatome that ensures the proximal aspect is still attached to adjacent skin. The dermis is removed from the area underneath the back-folded split-thickness skin flap. The dermal graft is meshed and sutured into the recipient site. The split-thickness skin flap is replaced over the donor site. Meshed reversed dermal grafts have excellent survival rates, even with direct placement on bone without periosteum. Querings et al²¹ reported graft survival with no complications in 19 of 21 (90.4%) patients undergoing scalp or plantar sole reconstruction.

CONCLUSION

With the widespread adoption of the fresh-tissue technique for Mohs micrographic surgery and the establishment of the American Society for Dermatologic Surgery in 1970, the depth and scope of techniques used by dermatologic surgeons has dramatically expanded. Although the use of dermal flaps and grafts is not as widespread in dermatology as other reconstructive techniques, their unique advantages should be considered. Deepithelialized flaps and grafts should be considered when the following reconstructive goals are desired: (1) conversion of a 2-stage interpolation flap to a single-stage tunneled flap, (2) contour and cosmetic subunit preservation of deep defects through volume augmentation, (3) reconstruction in areas of high mechanical tension, and (4) free margin preservation. The multiple applications of deepithelialized flaps and grafts as described in this review demonstrate their continued applicability in dermatologic surgery.

Deepithelialized flaps and grafts have been widely used by reconstructive surgeons in a diverse range of medical specialties since the early 20th century. ¹ These reconstructive modalities have more recently been applied to dermatologic surgery. Deepithelialized flaps and grafts involve removal of the epidermis from the dermis for a variety of surgical purposes. Although these techniques play an important role in dermatologic surgery, reports of application of deepithelialized flaps and grafts in the dermatology literature is limited. This article includes a presentation of the applications of deepithelialized flaps and grafts in procedural dermatology.

DEEPITHELIALIZATION TECHNIQUES

There are a variety of techniques for deepithelialization, although sharp deepithelialization generally is preferred by dermatologic surgeons. The scalpel technique can be accomplished by making an intradermal incision with a No. 15 blade. Traction is an essential component of the deepthelialization process and facilitates sharp removal of the epidermis and superficial dermis in an even plane. The peeling orange technique, which has been described in reduction mammoplasty, is a variant of the scalpel technique used for creating a large area of deepithelialized tissue.² A No. 10 blade is used to make multiple partial-thickness intradermal incisions 1 to 2 cm apart along the pedicle. Traction facilitates rapid deepithelialization of the skin strips on the pedicle. A sharp curette is an alternative option for sharply removing the epithelium from a small area. Electric dermatome, laser, and electrocautery techniques for deepithelialization also can be considered.^2,3

APPLICATION OF DEEPITHELIALIZED FLAPS

Deepithelialized flaps may be considered for single-stage reconstruction with tunneled interpolation flaps, reconstruction requiring contour preservation, and reconstruction involving free margins.^4-17

Reconstruction With Single-Stage Tunneled Interpolated Flaps

Alar Base
A partially deepithelialized tunneled interpolated flap is an elegant reconstructive option for defects involving the upper cutaneous lip and alar base. The flap is elevated from the ipsilateral nasolabial fold, deepithelialized proximally, and tunneled under the intact portion of the cutaneous upper lip and ala. The flap is then deepithelialized superiorly to bolster the alar base and inset at the recipient site.⁴

Nasal Ala
The tunneled interpolated flap is useful for reconstruction of defects of the nasal ala. A flap with a superior deepithelialized pedicle and an anticipated inferior Burow triangle is designed along the axis of the nasolabial fold. The inferior Burow triangle and central flap are elevated at the level of the superficial subcutaneous fat and the pedicle is dissected. The donor and recipient sites are widely undermined, and the flap and pedicle pass through the tunnel. The donor site is closed primarily, the inferior Burow triangle is trimmed, and the flap is sutured into the defect.⁵ This flap allows for preservation of free margins and favorable placement of incision lines. Furthermore, pincushioning of the flap helps to recreate the rounded shape of the lateral ala.⁶

Nasal Tip
Nasal tip defects can be repaired with a retroangular flap, centered on the angular artery. The flap is elevated along the axis of the nasolabial fold, deepithelialized at its proximal base, and transferred through a subcutaneous tunnel to the nasal tip. The angular artery is ligated at the inferior aspect of the flap.⁷

Nasal Sidewall
A deepithelialized tunneled interpolated forehead flap, similar to the classic paramedian forehead flap, can be used to reconstruct nasal sidewall defects. A flap is elevated on the contralateral forehead and the proximal portion is deepithelialized. A tunnel is then bluntly dissected just above the periosteum, and the flap is introduced into the defect through the tunnel and inset. This flap has the advantages of being a single-stage procedure, restoring volume to the defect area, and maintaining excellent vascular supply.⁸

Eyelid
A tunneled interpolated forehead flap also can be used to repair medial canthal defects and for anterior lamellar repair of lower eyelid defects. In a study of 9 patients receiving a tunneled interpolated forehead flap in these anatomic locations, all flaps demonstrated viability, protection of the globe, and preservation of the concave architecture of the medial canthus.⁹

Earlobe
Earlobe defects may be repaired with a pull-through interpolated preauricular flap. A flap is elevated superiorly in the preauricular region and the proximal aspect of the flap is deepithelialized. The flap is pulled through a tunnel and inset at the anterior earlobe defect. The donor site is closed primarily.^10,11

Concha
Reconstruction of anterior conchal defects with exposed cartilage can be accomplished with a pull-through interpolated postauricular flap based on the auriculomastoid fossa. The postauricular flap is elevated, the base is deepithelialized, an incision is made in the medial aspect of the defect, and the flap is moved through a tunnel between the posterior and anterior surfaces of the ear. The flap is secured to the anterior surface of the concha.¹²

Reconstruction Requiring Contour Preservation

Central Face
The hinge flap is optimal for reconstruction of deep central facial defects (Figure 1). The hinge flap is planned at a site contiguous with a margin of the defect and can include the dermis, subcutaneous tissue, muscle, or a combination of these. The desired tissue is folded over on the pedicle to fill the defect. Cutaneous coverage is accomplished through a primary closure, separate flap, or skin graft. In addition to restoring contour and therefore the cosmetic subunit, the hinge flap is performed in a single stage, resists wound contracture, and provides a well-vascularized wound bed resulting in a low incidence of graft failure.^13,14 Muscular hinge flaps have been described for reconstruction of forehead defects with exposed bone based on the frontalis muscle.¹⁵

Lower Lip
A variant of a V-Y advancement flap has been described for reconstruction of defects greater than one-third the length of the lower lip. The top of the “V” is deepithelialized and the flap is advanced such that the top of the “V” abuts the inferior border of the defect. The “V” flap is inset at its advanced position, converting the “V”-shaped wound into a “Y.” An overlying buccal mucosal graft provides reconstruction of the lower red lip and labial mucosa.¹⁶

Helix of the Ear
Large defects of the scapha and helix of the ear can be reconstructed with the use of a staged interpolated postauricular flap. The postauricular flap is elevated into a subcutaneous plane. A full-thickness incision is made medial to the helical rim, and the flap is tunneled through and sutured into place. The pedicle is later divided, and the distal aspect of the flap is deepithelialized and inset into the helical rim for volume restoration.¹⁷

Reconstruction Involving Free Margins

Nasal Ala
For large defects involving the upper cutaneous lip with adjacent alar base involvement, a partially deepithelialized V-Y flap is a useful reconstructive option (Figure 2).

Infraorbital Region
A deepithelialized variant of a V-Y advancement flap can be used for closure of infraorbital defects. The limbs of the V-Y flap are deepithelialized and anchored to the medial and lateral canthal tendons or periosteum. Ectropion prevention is the primary advantage of this flap.¹⁸

APPLICATION OF DEEPITHELIALIZED GRAFTS

Deepithelialized grafts may be considered for volume replacement, reconstruction requiring contour preservation, and restoration of mechanical integrity in areas of high mechanical tension.^3,19-21

Reconstruction Requiring Contour Preservation

Deepithelialized grafts are used to improve depressed nasal scars and restore volume in deep nasal wounds. One method involves deepithelialization of 2 postauricular punch biopsies. An 18-gauge needle is used to make a small hole in the depressed nasal scar, the dermal grafts are inserted, and the defect is closed primarily.¹⁹ Dermal grafts may be harvested from excess full-thickness skin grafts (FTSGs) or dog-ear tissue. When used under flaps, the dermal graft is trimmed to the size of the defect. When used under FTSGs, thin dermal graft strips are placed in a gridlike pattern to allow for revascularization. A study of 15 patients with contour deformities reconstructed with dermal graft insertions demonstrated that 14 (94%) patients had no significant complications and improvement of scar depression was achieved.²⁰

Reconstruction in Areas of High Mechanical Tension

Plantar Foot
A combined dermal and full-thickness sandwich graft has been described for reconstruction of plantar foot defects.³ The graft is created by obtaining a FTSG twice the size of the wound defect and deepithelializing half of the graft. The graft is then defatted and the deepithelialized portion is folded beneath the other half, allowing the papillary dermis to make contact with the wound surface.

Scalp
Dermal graft reconstruction for scalp defects may be accomplished with a split-thickness skin flap. The flap is harvested using an electronic dermatome that ensures the proximal aspect is still attached to adjacent skin. The dermis is removed from the area underneath the back-folded split-thickness skin flap. The dermal graft is meshed and sutured into the recipient site. The split-thickness skin flap is replaced over the donor site. Meshed reversed dermal grafts have excellent survival rates, even with direct placement on bone without periosteum. Querings et al²¹ reported graft survival with no complications in 19 of 21 (90.4%) patients undergoing scalp or plantar sole reconstruction.

CONCLUSION

With the widespread adoption of the fresh-tissue technique for Mohs micrographic surgery and the establishment of the American Society for Dermatologic Surgery in 1970, the depth and scope of techniques used by dermatologic surgeons has dramatically expanded. Although the use of dermal flaps and grafts is not as widespread in dermatology as other reconstructive techniques, their unique advantages should be considered. Deepithelialized flaps and grafts should be considered when the following reconstructive goals are desired: (1) conversion of a 2-stage interpolation flap to a single-stage tunneled flap, (2) contour and cosmetic subunit preservation of deep defects through volume augmentation, (3) reconstruction in areas of high mechanical tension, and (4) free margin preservation. The multiple applications of deepithelialized flaps and grafts as described in this review demonstrate their continued applicability in dermatologic surgery.

Deepithelialized flaps and grafts have been widely used by reconstructive surgeons in a diverse range of medical specialties since the early 20th century. ¹ These reconstructive modalities have more recently been applied to dermatologic surgery. Deepithelialized flaps and grafts involve removal of the epidermis from the dermis for a variety of surgical purposes. Although these techniques play an important role in dermatologic surgery, reports of application of deepithelialized flaps and grafts in the dermatology literature is limited. This article includes a presentation of the applications of deepithelialized flaps and grafts in procedural dermatology.

DEEPITHELIALIZATION TECHNIQUES

There are a variety of techniques for deepithelialization, although sharp deepithelialization generally is preferred by dermatologic surgeons. The scalpel technique can be accomplished by making an intradermal incision with a No. 15 blade. Traction is an essential component of the deepthelialization process and facilitates sharp removal of the epidermis and superficial dermis in an even plane. The peeling orange technique, which has been described in reduction mammoplasty, is a variant of the scalpel technique used for creating a large area of deepithelialized tissue.² A No. 10 blade is used to make multiple partial-thickness intradermal incisions 1 to 2 cm apart along the pedicle. Traction facilitates rapid deepithelialization of the skin strips on the pedicle. A sharp curette is an alternative option for sharply removing the epithelium from a small area. Electric dermatome, laser, and electrocautery techniques for deepithelialization also can be considered.^2,3

APPLICATION OF DEEPITHELIALIZED FLAPS

Deepithelialized flaps may be considered for single-stage reconstruction with tunneled interpolation flaps, reconstruction requiring contour preservation, and reconstruction involving free margins.^4-17

Reconstruction With Single-Stage Tunneled Interpolated Flaps

Alar Base
A partially deepithelialized tunneled interpolated flap is an elegant reconstructive option for defects involving the upper cutaneous lip and alar base. The flap is elevated from the ipsilateral nasolabial fold, deepithelialized proximally, and tunneled under the intact portion of the cutaneous upper lip and ala. The flap is then deepithelialized superiorly to bolster the alar base and inset at the recipient site.⁴

Nasal Ala
The tunneled interpolated flap is useful for reconstruction of defects of the nasal ala. A flap with a superior deepithelialized pedicle and an anticipated inferior Burow triangle is designed along the axis of the nasolabial fold. The inferior Burow triangle and central flap are elevated at the level of the superficial subcutaneous fat and the pedicle is dissected. The donor and recipient sites are widely undermined, and the flap and pedicle pass through the tunnel. The donor site is closed primarily, the inferior Burow triangle is trimmed, and the flap is sutured into the defect.⁵ This flap allows for preservation of free margins and favorable placement of incision lines. Furthermore, pincushioning of the flap helps to recreate the rounded shape of the lateral ala.⁶

Nasal Tip
Nasal tip defects can be repaired with a retroangular flap, centered on the angular artery. The flap is elevated along the axis of the nasolabial fold, deepithelialized at its proximal base, and transferred through a subcutaneous tunnel to the nasal tip. The angular artery is ligated at the inferior aspect of the flap.⁷

Nasal Sidewall
A deepithelialized tunneled interpolated forehead flap, similar to the classic paramedian forehead flap, can be used to reconstruct nasal sidewall defects. A flap is elevated on the contralateral forehead and the proximal portion is deepithelialized. A tunnel is then bluntly dissected just above the periosteum, and the flap is introduced into the defect through the tunnel and inset. This flap has the advantages of being a single-stage procedure, restoring volume to the defect area, and maintaining excellent vascular supply.⁸

Eyelid
A tunneled interpolated forehead flap also can be used to repair medial canthal defects and for anterior lamellar repair of lower eyelid defects. In a study of 9 patients receiving a tunneled interpolated forehead flap in these anatomic locations, all flaps demonstrated viability, protection of the globe, and preservation of the concave architecture of the medial canthus.⁹

Earlobe
Earlobe defects may be repaired with a pull-through interpolated preauricular flap. A flap is elevated superiorly in the preauricular region and the proximal aspect of the flap is deepithelialized. The flap is pulled through a tunnel and inset at the anterior earlobe defect. The donor site is closed primarily.^10,11

Concha
Reconstruction of anterior conchal defects with exposed cartilage can be accomplished with a pull-through interpolated postauricular flap based on the auriculomastoid fossa. The postauricular flap is elevated, the base is deepithelialized, an incision is made in the medial aspect of the defect, and the flap is moved through a tunnel between the posterior and anterior surfaces of the ear. The flap is secured to the anterior surface of the concha.¹²

Reconstruction Requiring Contour Preservation

Central Face
The hinge flap is optimal for reconstruction of deep central facial defects (Figure 1). The hinge flap is planned at a site contiguous with a margin of the defect and can include the dermis, subcutaneous tissue, muscle, or a combination of these. The desired tissue is folded over on the pedicle to fill the defect. Cutaneous coverage is accomplished through a primary closure, separate flap, or skin graft. In addition to restoring contour and therefore the cosmetic subunit, the hinge flap is performed in a single stage, resists wound contracture, and provides a well-vascularized wound bed resulting in a low incidence of graft failure.^13,14 Muscular hinge flaps have been described for reconstruction of forehead defects with exposed bone based on the frontalis muscle.¹⁵

Lower Lip
A variant of a V-Y advancement flap has been described for reconstruction of defects greater than one-third the length of the lower lip. The top of the “V” is deepithelialized and the flap is advanced such that the top of the “V” abuts the inferior border of the defect. The “V” flap is inset at its advanced position, converting the “V”-shaped wound into a “Y.” An overlying buccal mucosal graft provides reconstruction of the lower red lip and labial mucosa.¹⁶

Helix of the Ear
Large defects of the scapha and helix of the ear can be reconstructed with the use of a staged interpolated postauricular flap. The postauricular flap is elevated into a subcutaneous plane. A full-thickness incision is made medial to the helical rim, and the flap is tunneled through and sutured into place. The pedicle is later divided, and the distal aspect of the flap is deepithelialized and inset into the helical rim for volume restoration.¹⁷

Reconstruction Involving Free Margins

Nasal Ala
For large defects involving the upper cutaneous lip with adjacent alar base involvement, a partially deepithelialized V-Y flap is a useful reconstructive option (Figure 2).

Infraorbital Region
A deepithelialized variant of a V-Y advancement flap can be used for closure of infraorbital defects. The limbs of the V-Y flap are deepithelialized and anchored to the medial and lateral canthal tendons or periosteum. Ectropion prevention is the primary advantage of this flap.¹⁸

APPLICATION OF DEEPITHELIALIZED GRAFTS

Deepithelialized grafts may be considered for volume replacement, reconstruction requiring contour preservation, and restoration of mechanical integrity in areas of high mechanical tension.^3,19-21

Reconstruction Requiring Contour Preservation

Deepithelialized grafts are used to improve depressed nasal scars and restore volume in deep nasal wounds. One method involves deepithelialization of 2 postauricular punch biopsies. An 18-gauge needle is used to make a small hole in the depressed nasal scar, the dermal grafts are inserted, and the defect is closed primarily.¹⁹ Dermal grafts may be harvested from excess full-thickness skin grafts (FTSGs) or dog-ear tissue. When used under flaps, the dermal graft is trimmed to the size of the defect. When used under FTSGs, thin dermal graft strips are placed in a gridlike pattern to allow for revascularization. A study of 15 patients with contour deformities reconstructed with dermal graft insertions demonstrated that 14 (94%) patients had no significant complications and improvement of scar depression was achieved.²⁰

Reconstruction in Areas of High Mechanical Tension

Plantar Foot
A combined dermal and full-thickness sandwich graft has been described for reconstruction of plantar foot defects.³ The graft is created by obtaining a FTSG twice the size of the wound defect and deepithelializing half of the graft. The graft is then defatted and the deepithelialized portion is folded beneath the other half, allowing the papillary dermis to make contact with the wound surface.

Scalp
Dermal graft reconstruction for scalp defects may be accomplished with a split-thickness skin flap. The flap is harvested using an electronic dermatome that ensures the proximal aspect is still attached to adjacent skin. The dermis is removed from the area underneath the back-folded split-thickness skin flap. The dermal graft is meshed and sutured into the recipient site. The split-thickness skin flap is replaced over the donor site. Meshed reversed dermal grafts have excellent survival rates, even with direct placement on bone without periosteum. Querings et al²¹ reported graft survival with no complications in 19 of 21 (90.4%) patients undergoing scalp or plantar sole reconstruction.

CONCLUSION

With the widespread adoption of the fresh-tissue technique for Mohs micrographic surgery and the establishment of the American Society for Dermatologic Surgery in 1970, the depth and scope of techniques used by dermatologic surgeons has dramatically expanded. Although the use of dermal flaps and grafts is not as widespread in dermatology as other reconstructive techniques, their unique advantages should be considered. Deepithelialized flaps and grafts should be considered when the following reconstructive goals are desired: (1) conversion of a 2-stage interpolation flap to a single-stage tunneled flap, (2) contour and cosmetic subunit preservation of deep defects through volume augmentation, (3) reconstruction in areas of high mechanical tension, and (4) free margin preservation. The multiple applications of deepithelialized flaps and grafts as described in this review demonstrate their continued applicability in dermatologic surgery.

Psoriasis is a chronic inflammatory disease that affects approximately 2% to 3% of the US population.¹ Patients with psoriasis are more likely to have cardiovascular risk factors (eg, obesity, metabolic syndrome) than individuals without psoriasis.² In fact, recent evidence has suggested that a diagnosis of psoriasis is an independent risk factor for cardiometabolic diseases including diabetes, major adverse cardiovascular events, and obesity.³ Given the well-recognized health benefits of physical activity and the associated reduction in coronary heart disease risk,⁴ patients with psoriasis specifically may benefit from regular participation in physical activity. Thus, an enhanced understanding of the relationship between psoriasis and vigorous physical activity would help determine the role of initiating and recommending interventions that implement physical activity for patients with psoriasis. A review was conducted to determine the relationship between psoriasis and vigorous physical activity.

Methods

An English-language literature search of PubMed articles indexed for MEDLINE (January 1, 1946–October 15, 2017) as well as articles in the Embase database (January 1, 1947–October 15, 2017) and Cochrane Library (January 1, 1992–October 15, 2017) using the terms psoriasis and physical activity was performed. The search strategy was established based on a prior review of vigorous physical activity in eczema.⁵ The article titles and/or abstracts were reviewed, and the studies were excluded if they did not evaluate physical activity in patients with psoriasis. Studies without a control group also were excluded. Articles on patients with psoriatic arthritis and studies that involved modification of dietary intake also were excluded.

Two reviewers (M.A. and E.B.L.) independently extracted data from the studies and compiled the results. The following factors were included in the data extracted: study year, location, and design; method of diagnosis of psoriasis; total number of patients included in the study; and age, gender, and level of physical activity of the study patients. Level of physical activity was the exposure, and diagnosis of psoriasis was the dependent variable. Physical activity was defined differently across the studies that were evaluated. To determine study quality, we implemented the Newcastle–Ottawa Scale (NOS), a 9-star scoring system that includes items such as selection criteria, comparability, and study outcome.⁶ Studies with an NOS score of 7 or higher were included in the meta-analysis.

Results

The literature search generated 353 nonduplicate articles. A thorough review of the articles yielded 4 studies that were incorporated in the final analysis.^7-10 We aimed to perform a meta-analysis; however, only 1 of the studies included in the final analysis had an NOS score of 7 or higher along with adequate data to be incorporated into our study.¹⁰ As a result, the meta-analysis was converted to a regular review.

The cross-sectional study we reviewed, which had an NOS score of 7, included males and females in the United States aged 20 to 59 years.¹⁰ Data were collected using the population-based National Health and Nutrition Examination Survey from 2003 to 2006. The survey measured the likelihood of participation in leisure-time moderate to vigorous physical activity (MVPA) and metabolic equivalent task (MET) minutes of MVPA in the past 30 days. Of 6549 participants, 385 were excluded from the analysis due to missing values for 1 or more of the study variables. Of the remaining 6164 participants, 84 (1.4%) reported having a diagnosis of psoriasis with few or no psoriasis patches at the time of the survey, and 71 (1.2%) reported having a diagnosis of psoriasis with few to extensive patches at the time of the survey.¹⁰

Participants with psoriasis were less likely to participate in MVPA in the previous 30 days compared to participants without psoriasis, but the association was not statistically significant.¹⁰ The study demonstrated that, on average, participants with psoriasis spent 31% (95% confidence interval [CI], −0.57 to −0.05) fewer MET minutes on leisure-time MVPA versus participants without psoriasis; however, this association was not statistically significant. It is important to note that the diagnosis of psoriasis was self-reported, and measures of disease duration or areas of involvement were not incorporated.

Comment

Our review revealed that vigorous physical activity may be reduced in patients with psoriasis compared to those without psoriasis. Initially, we aimed to perform a systematic review of the literature; however, only 1 study met the criteria for the systematic review, highlighting the need for more robust studies evaluating this subject.

Do et al¹⁰ demonstrated that psoriasis patients were less likely to participate in MVPA, but the findings were not statistically significant. Of those who participated in MVPA, MET minutes were fewer among patients with few to extensive skin lesions compared to those without psoriasis. The investigators suggested that psoriasis patients with more severe disease tend to exercise less and ultimately would benefit from regular vigorous physical activity.

Frankel et al⁷ performed a prospective cohort study in US women to evaluate the role of physical activity in preventing psoriasis. The investigators reported that the most physically active quintile had a lower multivariate relative risk of psoriasis (0.72; 95% CI, 0.59–0.89; P<.001 for trend) compared to the least active quintile.⁷ Additionally, vigorous physical activity, which was defined as 6 or more MET minutes, was associated with a significantly lower risk of incident psoriasis (0.66; 95% CI, 0.54–0.81; P<.001 for trend), which maintained significance after adjusting for body mass index (BMI). The investigators suggested that, by decreasing chronic inflammation and lowering levels of proinflammatory cytokines, vigorous physical activity may reduce the risk of psoriasis development in women.⁷ It is plausible that vigorous physical activity modifies the state of chronic inflammation, which could subsequently reduce the risk of developing psoriasis; however, further long-term, randomized, prospective studies are needed to verify the relationship between physical activity and development of psoriasis.

Torres et al⁸ performed a cross-sectional questionnaire study to assess physical activity in patients with severe psoriasis (defined as >10% body surface area involvement and/or disease requiring systemic therapy or phototherapy) versus healthy controls. Physical activity level was measured using the International Physical Activity Questionnaire. The odds ratio of low-level physical activity compared to non–low-level physical activity among psoriasis patients versus controls was 3.42 (95% CI, 1.47–7.91; P=.002). Additionally, the average total MET minutes of psoriasis patients were significantly reduced compared to those of the healthy controls (P=.001). Thus, the investigators suggested that vigorous physical activity is less likely in psoriasis patients, which may contribute to the increased risk of cardiovascular disease in this population.⁸ Vigorous physical activity would benefit patients with psoriasis to help lower the chronic state of inflammation and cardiometabolic comorbidities.

Demirel et al⁹ performed a study to compare aerobic exercise capacity and daily physical activity level in psoriasis patients (n=30) compared to controls (n=30). Daily physical activity, measured with an accelerometer, was significantly higher in male patients with psoriasis compared to controls (P=.021). No significant difference was reported in maximal aerobic capacity in both male and female psoriasis patients versus controls. The investigators suggested that the level of daily physical activity is not limited in psoriasis patients, yet the small sample size may limit the generalizability of the study.

The ability to dissipate heat during exercise seems to be diminished in patients with psoriasis. Specifically, it has been suggested that psoriasis lesions interfere with normal perspiration.¹¹ Moreover, joint involvement in patients with psoriatic arthritis may lead to physical functional disabilities that can interfere with the ability of these patients to participate in regular physical activity.^12-14 For this reason, our review excluded articles that evaluated patients with psoriatic arthritis. Despite this exclusion, it is important to consider that comorbid psoriatic arthritis in clinical practice may impede patients with psoriasis from participating in physical activity. Additionally, various social aspects also may limit physical activity in psoriasis patients; for instance, psoriasis patients often avoid activities that involve increased exposure of the skin (eg, communal showers, wearing sports attire).¹⁵

Furthermore, obese psoriasis patients are less likely to exercise compared to obese individuals without psoriasis.¹⁶ In patients with higher BMI, the risk of psoriasis is increased.¹⁷ A systematic review suggested that weight loss may improve psoriasis severity.¹⁸ Bariatric surgery also may improve psoriasis.¹⁹ Moreover, obesity may interfere with response to biologic therapies for psoriasis. Specifically, higher BMI is linked with lower response to fixed-dose biologic therapies compared to weight-based biologic options (eg, infliximab).^20,21

Conclusion

Given the increased risk of myocardial infarction in patients with psoriasis, it is important to recognize the barriers to physical activity that psoriasis patients face.²² Due to the considerable health benefits associated with regular physical activity, physicians should encourage patients with psoriasis to participate in physical activity as tolerated. Of note, the studies included in this review varied in their definitions of psoriasis disease severity and measures of physical activity level. Long-term, randomized, prospective studies are needed to clarify the relationship between psoriasis and physical activity. Evidence from these studies would help guide clinical recommendations regarding the role of physical activity for patients with psoriasis.

Psoriasis is a chronic inflammatory disease that affects approximately 2% to 3% of the US population.¹ Patients with psoriasis are more likely to have cardiovascular risk factors (eg, obesity, metabolic syndrome) than individuals without psoriasis.² In fact, recent evidence has suggested that a diagnosis of psoriasis is an independent risk factor for cardiometabolic diseases including diabetes, major adverse cardiovascular events, and obesity.³ Given the well-recognized health benefits of physical activity and the associated reduction in coronary heart disease risk,⁴ patients with psoriasis specifically may benefit from regular participation in physical activity. Thus, an enhanced understanding of the relationship between psoriasis and vigorous physical activity would help determine the role of initiating and recommending interventions that implement physical activity for patients with psoriasis. A review was conducted to determine the relationship between psoriasis and vigorous physical activity.

Methods

An English-language literature search of PubMed articles indexed for MEDLINE (January 1, 1946–October 15, 2017) as well as articles in the Embase database (January 1, 1947–October 15, 2017) and Cochrane Library (January 1, 1992–October 15, 2017) using the terms psoriasis and physical activity was performed. The search strategy was established based on a prior review of vigorous physical activity in eczema.⁵ The article titles and/or abstracts were reviewed, and the studies were excluded if they did not evaluate physical activity in patients with psoriasis. Studies without a control group also were excluded. Articles on patients with psoriatic arthritis and studies that involved modification of dietary intake also were excluded.

Two reviewers (M.A. and E.B.L.) independently extracted data from the studies and compiled the results. The following factors were included in the data extracted: study year, location, and design; method of diagnosis of psoriasis; total number of patients included in the study; and age, gender, and level of physical activity of the study patients. Level of physical activity was the exposure, and diagnosis of psoriasis was the dependent variable. Physical activity was defined differently across the studies that were evaluated. To determine study quality, we implemented the Newcastle–Ottawa Scale (NOS), a 9-star scoring system that includes items such as selection criteria, comparability, and study outcome.⁶ Studies with an NOS score of 7 or higher were included in the meta-analysis.

Results

The literature search generated 353 nonduplicate articles. A thorough review of the articles yielded 4 studies that were incorporated in the final analysis.^7-10 We aimed to perform a meta-analysis; however, only 1 of the studies included in the final analysis had an NOS score of 7 or higher along with adequate data to be incorporated into our study.¹⁰ As a result, the meta-analysis was converted to a regular review.

The cross-sectional study we reviewed, which had an NOS score of 7, included males and females in the United States aged 20 to 59 years.¹⁰ Data were collected using the population-based National Health and Nutrition Examination Survey from 2003 to 2006. The survey measured the likelihood of participation in leisure-time moderate to vigorous physical activity (MVPA) and metabolic equivalent task (MET) minutes of MVPA in the past 30 days. Of 6549 participants, 385 were excluded from the analysis due to missing values for 1 or more of the study variables. Of the remaining 6164 participants, 84 (1.4%) reported having a diagnosis of psoriasis with few or no psoriasis patches at the time of the survey, and 71 (1.2%) reported having a diagnosis of psoriasis with few to extensive patches at the time of the survey.¹⁰

Participants with psoriasis were less likely to participate in MVPA in the previous 30 days compared to participants without psoriasis, but the association was not statistically significant.¹⁰ The study demonstrated that, on average, participants with psoriasis spent 31% (95% confidence interval [CI], −0.57 to −0.05) fewer MET minutes on leisure-time MVPA versus participants without psoriasis; however, this association was not statistically significant. It is important to note that the diagnosis of psoriasis was self-reported, and measures of disease duration or areas of involvement were not incorporated.

Comment

Our review revealed that vigorous physical activity may be reduced in patients with psoriasis compared to those without psoriasis. Initially, we aimed to perform a systematic review of the literature; however, only 1 study met the criteria for the systematic review, highlighting the need for more robust studies evaluating this subject.

Do et al¹⁰ demonstrated that psoriasis patients were less likely to participate in MVPA, but the findings were not statistically significant. Of those who participated in MVPA, MET minutes were fewer among patients with few to extensive skin lesions compared to those without psoriasis. The investigators suggested that psoriasis patients with more severe disease tend to exercise less and ultimately would benefit from regular vigorous physical activity.

Frankel et al⁷ performed a prospective cohort study in US women to evaluate the role of physical activity in preventing psoriasis. The investigators reported that the most physically active quintile had a lower multivariate relative risk of psoriasis (0.72; 95% CI, 0.59–0.89; P<.001 for trend) compared to the least active quintile.⁷ Additionally, vigorous physical activity, which was defined as 6 or more MET minutes, was associated with a significantly lower risk of incident psoriasis (0.66; 95% CI, 0.54–0.81; P<.001 for trend), which maintained significance after adjusting for body mass index (BMI). The investigators suggested that, by decreasing chronic inflammation and lowering levels of proinflammatory cytokines, vigorous physical activity may reduce the risk of psoriasis development in women.⁷ It is plausible that vigorous physical activity modifies the state of chronic inflammation, which could subsequently reduce the risk of developing psoriasis; however, further long-term, randomized, prospective studies are needed to verify the relationship between physical activity and development of psoriasis.

Torres et al⁸ performed a cross-sectional questionnaire study to assess physical activity in patients with severe psoriasis (defined as >10% body surface area involvement and/or disease requiring systemic therapy or phototherapy) versus healthy controls. Physical activity level was measured using the International Physical Activity Questionnaire. The odds ratio of low-level physical activity compared to non–low-level physical activity among psoriasis patients versus controls was 3.42 (95% CI, 1.47–7.91; P=.002). Additionally, the average total MET minutes of psoriasis patients were significantly reduced compared to those of the healthy controls (P=.001). Thus, the investigators suggested that vigorous physical activity is less likely in psoriasis patients, which may contribute to the increased risk of cardiovascular disease in this population.⁸ Vigorous physical activity would benefit patients with psoriasis to help lower the chronic state of inflammation and cardiometabolic comorbidities.

Demirel et al⁹ performed a study to compare aerobic exercise capacity and daily physical activity level in psoriasis patients (n=30) compared to controls (n=30). Daily physical activity, measured with an accelerometer, was significantly higher in male patients with psoriasis compared to controls (P=.021). No significant difference was reported in maximal aerobic capacity in both male and female psoriasis patients versus controls. The investigators suggested that the level of daily physical activity is not limited in psoriasis patients, yet the small sample size may limit the generalizability of the study.

The ability to dissipate heat during exercise seems to be diminished in patients with psoriasis. Specifically, it has been suggested that psoriasis lesions interfere with normal perspiration.¹¹ Moreover, joint involvement in patients with psoriatic arthritis may lead to physical functional disabilities that can interfere with the ability of these patients to participate in regular physical activity.^12-14 For this reason, our review excluded articles that evaluated patients with psoriatic arthritis. Despite this exclusion, it is important to consider that comorbid psoriatic arthritis in clinical practice may impede patients with psoriasis from participating in physical activity. Additionally, various social aspects also may limit physical activity in psoriasis patients; for instance, psoriasis patients often avoid activities that involve increased exposure of the skin (eg, communal showers, wearing sports attire).¹⁵

Furthermore, obese psoriasis patients are less likely to exercise compared to obese individuals without psoriasis.¹⁶ In patients with higher BMI, the risk of psoriasis is increased.¹⁷ A systematic review suggested that weight loss may improve psoriasis severity.¹⁸ Bariatric surgery also may improve psoriasis.¹⁹ Moreover, obesity may interfere with response to biologic therapies for psoriasis. Specifically, higher BMI is linked with lower response to fixed-dose biologic therapies compared to weight-based biologic options (eg, infliximab).^20,21

Conclusion

Given the increased risk of myocardial infarction in patients with psoriasis, it is important to recognize the barriers to physical activity that psoriasis patients face.²² Due to the considerable health benefits associated with regular physical activity, physicians should encourage patients with psoriasis to participate in physical activity as tolerated. Of note, the studies included in this review varied in their definitions of psoriasis disease severity and measures of physical activity level. Long-term, randomized, prospective studies are needed to clarify the relationship between psoriasis and physical activity. Evidence from these studies would help guide clinical recommendations regarding the role of physical activity for patients with psoriasis.

Psoriasis is a chronic inflammatory disease that affects approximately 2% to 3% of the US population.¹ Patients with psoriasis are more likely to have cardiovascular risk factors (eg, obesity, metabolic syndrome) than individuals without psoriasis.² In fact, recent evidence has suggested that a diagnosis of psoriasis is an independent risk factor for cardiometabolic diseases including diabetes, major adverse cardiovascular events, and obesity.³ Given the well-recognized health benefits of physical activity and the associated reduction in coronary heart disease risk,⁴ patients with psoriasis specifically may benefit from regular participation in physical activity. Thus, an enhanced understanding of the relationship between psoriasis and vigorous physical activity would help determine the role of initiating and recommending interventions that implement physical activity for patients with psoriasis. A review was conducted to determine the relationship between psoriasis and vigorous physical activity.

Methods

An English-language literature search of PubMed articles indexed for MEDLINE (January 1, 1946–October 15, 2017) as well as articles in the Embase database (January 1, 1947–October 15, 2017) and Cochrane Library (January 1, 1992–October 15, 2017) using the terms psoriasis and physical activity was performed. The search strategy was established based on a prior review of vigorous physical activity in eczema.⁵ The article titles and/or abstracts were reviewed, and the studies were excluded if they did not evaluate physical activity in patients with psoriasis. Studies without a control group also were excluded. Articles on patients with psoriatic arthritis and studies that involved modification of dietary intake also were excluded.

Two reviewers (M.A. and E.B.L.) independently extracted data from the studies and compiled the results. The following factors were included in the data extracted: study year, location, and design; method of diagnosis of psoriasis; total number of patients included in the study; and age, gender, and level of physical activity of the study patients. Level of physical activity was the exposure, and diagnosis of psoriasis was the dependent variable. Physical activity was defined differently across the studies that were evaluated. To determine study quality, we implemented the Newcastle–Ottawa Scale (NOS), a 9-star scoring system that includes items such as selection criteria, comparability, and study outcome.⁶ Studies with an NOS score of 7 or higher were included in the meta-analysis.

Results

The literature search generated 353 nonduplicate articles. A thorough review of the articles yielded 4 studies that were incorporated in the final analysis.^7-10 We aimed to perform a meta-analysis; however, only 1 of the studies included in the final analysis had an NOS score of 7 or higher along with adequate data to be incorporated into our study.¹⁰ As a result, the meta-analysis was converted to a regular review.

The cross-sectional study we reviewed, which had an NOS score of 7, included males and females in the United States aged 20 to 59 years.¹⁰ Data were collected using the population-based National Health and Nutrition Examination Survey from 2003 to 2006. The survey measured the likelihood of participation in leisure-time moderate to vigorous physical activity (MVPA) and metabolic equivalent task (MET) minutes of MVPA in the past 30 days. Of 6549 participants, 385 were excluded from the analysis due to missing values for 1 or more of the study variables. Of the remaining 6164 participants, 84 (1.4%) reported having a diagnosis of psoriasis with few or no psoriasis patches at the time of the survey, and 71 (1.2%) reported having a diagnosis of psoriasis with few to extensive patches at the time of the survey.¹⁰

Participants with psoriasis were less likely to participate in MVPA in the previous 30 days compared to participants without psoriasis, but the association was not statistically significant.¹⁰ The study demonstrated that, on average, participants with psoriasis spent 31% (95% confidence interval [CI], −0.57 to −0.05) fewer MET minutes on leisure-time MVPA versus participants without psoriasis; however, this association was not statistically significant. It is important to note that the diagnosis of psoriasis was self-reported, and measures of disease duration or areas of involvement were not incorporated.

Comment

Our review revealed that vigorous physical activity may be reduced in patients with psoriasis compared to those without psoriasis. Initially, we aimed to perform a systematic review of the literature; however, only 1 study met the criteria for the systematic review, highlighting the need for more robust studies evaluating this subject.

Do et al¹⁰ demonstrated that psoriasis patients were less likely to participate in MVPA, but the findings were not statistically significant. Of those who participated in MVPA, MET minutes were fewer among patients with few to extensive skin lesions compared to those without psoriasis. The investigators suggested that psoriasis patients with more severe disease tend to exercise less and ultimately would benefit from regular vigorous physical activity.

Frankel et al⁷ performed a prospective cohort study in US women to evaluate the role of physical activity in preventing psoriasis. The investigators reported that the most physically active quintile had a lower multivariate relative risk of psoriasis (0.72; 95% CI, 0.59–0.89; P<.001 for trend) compared to the least active quintile.⁷ Additionally, vigorous physical activity, which was defined as 6 or more MET minutes, was associated with a significantly lower risk of incident psoriasis (0.66; 95% CI, 0.54–0.81; P<.001 for trend), which maintained significance after adjusting for body mass index (BMI). The investigators suggested that, by decreasing chronic inflammation and lowering levels of proinflammatory cytokines, vigorous physical activity may reduce the risk of psoriasis development in women.⁷ It is plausible that vigorous physical activity modifies the state of chronic inflammation, which could subsequently reduce the risk of developing psoriasis; however, further long-term, randomized, prospective studies are needed to verify the relationship between physical activity and development of psoriasis.

Torres et al⁸ performed a cross-sectional questionnaire study to assess physical activity in patients with severe psoriasis (defined as >10% body surface area involvement and/or disease requiring systemic therapy or phototherapy) versus healthy controls. Physical activity level was measured using the International Physical Activity Questionnaire. The odds ratio of low-level physical activity compared to non–low-level physical activity among psoriasis patients versus controls was 3.42 (95% CI, 1.47–7.91; P=.002). Additionally, the average total MET minutes of psoriasis patients were significantly reduced compared to those of the healthy controls (P=.001). Thus, the investigators suggested that vigorous physical activity is less likely in psoriasis patients, which may contribute to the increased risk of cardiovascular disease in this population.⁸ Vigorous physical activity would benefit patients with psoriasis to help lower the chronic state of inflammation and cardiometabolic comorbidities.

Demirel et al⁹ performed a study to compare aerobic exercise capacity and daily physical activity level in psoriasis patients (n=30) compared to controls (n=30). Daily physical activity, measured with an accelerometer, was significantly higher in male patients with psoriasis compared to controls (P=.021). No significant difference was reported in maximal aerobic capacity in both male and female psoriasis patients versus controls. The investigators suggested that the level of daily physical activity is not limited in psoriasis patients, yet the small sample size may limit the generalizability of the study.

The ability to dissipate heat during exercise seems to be diminished in patients with psoriasis. Specifically, it has been suggested that psoriasis lesions interfere with normal perspiration.¹¹ Moreover, joint involvement in patients with psoriatic arthritis may lead to physical functional disabilities that can interfere with the ability of these patients to participate in regular physical activity.^12-14 For this reason, our review excluded articles that evaluated patients with psoriatic arthritis. Despite this exclusion, it is important to consider that comorbid psoriatic arthritis in clinical practice may impede patients with psoriasis from participating in physical activity. Additionally, various social aspects also may limit physical activity in psoriasis patients; for instance, psoriasis patients often avoid activities that involve increased exposure of the skin (eg, communal showers, wearing sports attire).¹⁵

Furthermore, obese psoriasis patients are less likely to exercise compared to obese individuals without psoriasis.¹⁶ In patients with higher BMI, the risk of psoriasis is increased.¹⁷ A systematic review suggested that weight loss may improve psoriasis severity.¹⁸ Bariatric surgery also may improve psoriasis.¹⁹ Moreover, obesity may interfere with response to biologic therapies for psoriasis. Specifically, higher BMI is linked with lower response to fixed-dose biologic therapies compared to weight-based biologic options (eg, infliximab).^20,21

Conclusion

Given the increased risk of myocardial infarction in patients with psoriasis, it is important to recognize the barriers to physical activity that psoriasis patients face.²² Due to the considerable health benefits associated with regular physical activity, physicians should encourage patients with psoriasis to participate in physical activity as tolerated. Of note, the studies included in this review varied in their definitions of psoriasis disease severity and measures of physical activity level. Long-term, randomized, prospective studies are needed to clarify the relationship between psoriasis and physical activity. Evidence from these studies would help guide clinical recommendations regarding the role of physical activity for patients with psoriasis.

CASE Chronic pelvic pain from endometriosis

A 40-year-old woman (G0) has a 20-year history of chronic pelvic pain. Stage III endometriosis is diagnosed on laparoscopic excision of endometriotic tissue. Postoperative pain symptoms include dysmenorrhea and deep dyspareunia, and the patient is feeling anxious. Physical examination reveals a retroverted uterus, right adnexal fullness and tenderness, and tenderness on palpation of the right levator ani and right obturator internus; rectovaginal examination findings are unremarkable. The patient, though now engaged in a pelvic floor physical therapy program, has yet to achieve the pain control she desires. After reviewing the treatment strategies for endometriosis with the patient, she elects definitive surgical management with minimally invasive hysterectomy and salpingo-oophorectomy. What pre-, intra-, and postoperative pain management plan do you devise for this patient?

Chronic pelvic pain presents a unique clinical challenge, as pain typically is multifactorial, and several peripheral pain generators may be involved. Although surgery can be performed to manage anatomically based disease processes, it does not address pain from musculoskeletal or neuropathic sources. A complete medical history and a physical examination are of utmost importance in developing a comprehensive multimodal management plan that may include surgery as treatment for the pain.

The standard of care for surgery is a minimally invasive approach (vaginal, laparoscopic, or robot-assisted laparoscopic), as it causes the least amount of trauma. Benefits of minimally invasive surgery include shorter hospitalization and faster recovery, likely owing to improved perioperative pain control, decreased blood loss, and fewer infections. Although this approach minimizes surgical trauma and thereby helps decrease the surgical stress response, the patient experience can be optimized with use of enhanced recovery pathways (ERPs), a multimodal approach to perioperative care.

ERPs were initially proposed as a means of reducing the degree of surgical injury and the subsequent physiologic stress response.¹ This multimodal approach begins in the outpatient setting, includes preoperative and intraoperative modalities, and continues postoperatively. In patients with chronic pain, ERPs are even more important. Assigning “prehabilitation” and setting expectations for surgery goals are the first step in improving the patient experience. Intraoperative use of opioid-sparing anesthetics or regional anesthesia can improve recovery. After surgery, patients with chronic pain and/or opioid dependence receive medications on a schedule, along with short-interval follow-up. Ultimately, reducing acute postoperative pain may lower the risk of developing chronic pain.

In this article on patients with chronic pelvic pain, we highlight elements of ERPs within the framework of enhanced recovery after surgery. Many of the interventions proposed here also can be used to improve the surgical experience of patients without chronic pain.

Preadmission education, expectations, and optimization

Preoperative counseling for elective procedures generally occurs in the outpatient setting. Although discussion traditionally has covered the type of procedure and its associated risks, benefits, and alternatives, new guidelines suggest a more mindful and comprehensive approach is warranted. Individualized patient-centered education programs have a positive impact on the perioperative course, effecting reductions in preoperative anxiety, opioid requirements, and hospital length of stay.² From a pain management perspective, the clinician can take some time during preoperative counseling to inform the patient about the pain to be expected from surgery, the ways the pain will be managed intraoperatively and postoperatively, and the multimodal strategies that will be used throughout the patient’s stay² and that may allow for early discharge. Although preadmission counseling still should address expectations for the surgery, it also presents an opportunity both to assess the patient’s ability to cope with the physical and psychological stress of surgery and to offer the patient appropriate need-based interventions, such as prehabilitation and cognitive-behavioral therapy (CBT).

Prehabilitation is the process of increasing functional capacity before surgery in order to mitigate the stress of the surgery. Prehabilitation may involve aerobic exercise, strength training, or functional task training. The gynecologic surgery literature lacks prehabilitation data, but data in the colorectal literature support use of a prehabilitation program for patients having a scheduled colectomy, with improved postoperative recovery.³ Although the colectomy cohort predominantly included older men, the principle that guides program implementation is the same: improve recovery after the stress of abdominal surgery. Indeed, a patient who opts for an elective surgery may have to wait several weeks before undergoing the procedure, and during this period behavioral interventions can take effect. With postoperative complications occurring more often in patients with reduced functional capacity, the data support using prehabilitation to decrease the incidence of postoperative complications, particularly among the most vulnerable patients.⁴ However, a definitive recommendation on use of pelvic floor exercises as an adjunct to prehabilitation cannot be made.⁴ Successful prehabilitation takes at least 4 weeks and should be part of a multimodal program that addresses other behavioral risk factors that may negatively affect recovery.⁵ For example, current tobacco users have compromised pulmonary status and wound healing immediately after surgery, and use more opioids.⁶ Conversely, smoking cessation for as little as 4 weeks before surgery is associated with fewer complications.⁷ In addition, given that alcohol abuse may compromise the surgical stress response and increase the risk of opioid misuse, addressing alcohol abuse preoperatively may improve postoperative recovery.⁸

Treating mood disorders that coexist with chronic pain disorders is an important part of outpatient multimodal management—psychological intervention is a useful adjunct to prehabilitation in reducing perioperative anxiety and improving postoperative functional capacity.⁹ For patients who have chronic pain and are undergoing surgery, it is important to address any anxiety, depression, or poor coping skills (eg, pain catastrophizing) to try to reduce the postoperative pain experience and decrease the risk of chronic postsurgical pain (CPSP).^10,11

Before surgery, patients with chronic pain syndromes should be evaluated for emotional distress and pain coping ability. When possible, they should be referred to a pain psychologist, who can initiate CBT and other interventions. In addition, pain coping skills can be developed or reinforced to address preoperative anxiety and pain catastrophizing. These interventions, which may include use of visual imagery, breathing exercises, and other relaxation techniques, are applicable to the management of postoperative anxiety as well.

Read about preoperative multimodal analgesia and intra- and postoperative management.

Preoperative multimodal analgesia

Multimodal analgesia has several benefits. Simultaneous effects can be generated on multiple pain-related neurotransmitters, and a synergistic effect (eg, of acetaminophen and a nonsteroidal anti-inflammatory drug [NSAID]) can improve pain management. In addition, small doses of multiple medications can be given, instead of a large dose of a single medication. Of course, this strategy must be modified in elderly and patients with impaired renal function, who are at high risk for polypharmacy.

Preoperative administration of 3 medications—a selective cyclooxygenase 2 (COX-2) inhibitor, acetaminophen, and a gabapentinoid—is increasingly accepted as part of multimodal analgesia. The selective COX-2 inhibitor targets inflammatory prostaglandins and has anti-inflammatory and analgesic effects; acetaminophen, an effective analgesic with an unclear mechanism of action, can reduce postoperative opioid consumption¹² and works synergistically with NSAIDs¹³; and the gabapentinoid gabapentin has an analgesic effect likely contributing to decreased movement-related pain and subsequent improved functional recovery (data are mixed on whether continuing gabapentin after surgery prevents CPSP).¹⁴⁻¹⁶

Although serotonin and norepinephrine reuptake inhibitors (SNRIs) are commonly used in outpatient management of chronic pelvic pain, data suggest that their role in perioperative pain management is evolving. As SNRIs may reduce central nervous system (CNS) sensitization,¹⁷ their analgesic effect is thought to result from increased descending inhibitory tone in the CNS, which makes this class of medication ideal for patients with chronic neuropathic pain.¹⁵

Limited data also suggest a role for SNRIs in decreasing immediate postoperative pain and CPSP in high-risk patients. Studies of duloxetine use in the immediate perioperative period have found reduced postoperative acute pain and opioid use.^18,19 In addition, a short course of low-dose (37.5 mg) venlafaxine both before and after surgery has demonstrated a reduction in postoperative opioid use and a reduction in movement-related pain 6 months after surgery.²⁰

Intraoperative management

The surgical and anesthesia teams share the goal of optimizing both pain control and postoperative recovery. Surgical team members, who want longer-acting anesthetics for infiltration of incision sites, discuss with the anesthesiologist the appropriateness of using peripheral nerve blocks or neuraxial anesthesia, given the patient’s history and planned procedure. Anesthesia team members can improve anesthesia and minimize intraoperative opioid use through several methods, including total intravenous anesthesia,²¹ dexamethasone,²² ketorolac,²³ and intravenous ketamine. Ketamine, in particular, has a wide range of surgical applications and has been found to reduce postoperative pain, postoperative pain medication use, and the risk of CPSP.²

Incision sites should be infiltrated before and after surgery. Lidocaine traditionally is used for its rapid onset of action in reducing surgical site pain, but its short half-life may limit its applicability to postoperative pain. Recently, bupivacaine (half-life, 3.5 hours) and liposomal bupivacaine (24–34 hours) have gained more attention. Both of these medications appear to be as effective as lidocaine in reducing surgical site pain.²⁴

Transversus abdominis plane (TAP) blocks have been used as an adjunct in pain management during abdominopelvic surgery. Although initial data on postoperative pain and opioid use reductions with TAP blocks were inconclusive,²⁵ more recent data showed a role for TAP blocks in a multimodal approach for reducing opioid use during laparoscopic and open surgery.^26,27 Given the small number of studies on using liposomal bupivacaine for peripheral nerve blocks (eg, TAP blocks) in postoperative pain management, current data are inconclusive.²⁸

Postoperative management

The ERP approach calls for continuing multimodal analgesia after surgery—in most cases, scheduling early use of oral acetaminophen and ibuprofen, and providing short-acting, low-dose opioid analgesia as needed. All patients should be given a bowel regimen. Similar to undergoing prehabilitation for surgery, patients should prepare themselves for recovery. They should be encouraged to engage in early ambulation and oral intake and, when clinically appropriate, be given same-day discharge for minimally invasive surgical procedures.

Patients with chronic pain before surgery are at increased risk for suboptimal postoperative pain management, and those who are dependent on opioids require additional perioperative measures for adequate postoperative pain control. In these complicated cases, it is appropriate to enlist a pain specialist, potentially before surgery, to help plan perioperative and postoperative pain management.² Postoperative pain management for opioid-dependent patients should include pharmacologic and nonpharmacologic interventions, such as use of nonopioid medications (eg, gabapentin) and continuation of CBT. Patients with chronic pain should be closely followed up for assessment of postoperative pain control and recovery.

CASE Resolved

Surgical management is one aspect of the longer term multimodal pain management strategy for this patient. After preoperative pelvic floor physical therapy, she is receptive to starting a trial of an SNRI for her pain and mood symptoms. Both interventions allow for optimization of her preoperative physical and psychological status. Expectations are set that she will be discharged the day of surgery and that the surgery is but one component of her multimodal treatment plan. In addition, before surgery, she takes oral acetaminophen, gabapentin, and celecoxib—previously having had no contraindications to these medications. During surgery, bupivacaine is used for infiltration of all incision sites, and the anesthesia team administers ketamine and a TAP block. After surgery, the patient is prepared for same-day discharge and given the NSAIDs and acetaminophen she is scheduled to take over the next 72 hours. She is also given a limited prescription for oxycodone for breakthrough pain. An office visit 1 to 2 weeks after surgery is scheduled.

ERP strategies for surgical management of endometriosis have not only improved this patient’s postoperative recovery but also reduced her surgical stress response and subsequent transition to chronic postoperative pain. Many of the strategies used in this case are applicable to patients without chronic pain.

Share your thoughts! Send your Letter to the Editor to [email protected]. Please include your name and the city and state in which you practice.

CASE Chronic pelvic pain from endometriosis

A 40-year-old woman (G0) has a 20-year history of chronic pelvic pain. Stage III endometriosis is diagnosed on laparoscopic excision of endometriotic tissue. Postoperative pain symptoms include dysmenorrhea and deep dyspareunia, and the patient is feeling anxious. Physical examination reveals a retroverted uterus, right adnexal fullness and tenderness, and tenderness on palpation of the right levator ani and right obturator internus; rectovaginal examination findings are unremarkable. The patient, though now engaged in a pelvic floor physical therapy program, has yet to achieve the pain control she desires. After reviewing the treatment strategies for endometriosis with the patient, she elects definitive surgical management with minimally invasive hysterectomy and salpingo-oophorectomy. What pre-, intra-, and postoperative pain management plan do you devise for this patient?

Chronic pelvic pain presents a unique clinical challenge, as pain typically is multifactorial, and several peripheral pain generators may be involved. Although surgery can be performed to manage anatomically based disease processes, it does not address pain from musculoskeletal or neuropathic sources. A complete medical history and a physical examination are of utmost importance in developing a comprehensive multimodal management plan that may include surgery as treatment for the pain.

The standard of care for surgery is a minimally invasive approach (vaginal, laparoscopic, or robot-assisted laparoscopic), as it causes the least amount of trauma. Benefits of minimally invasive surgery include shorter hospitalization and faster recovery, likely owing to improved perioperative pain control, decreased blood loss, and fewer infections. Although this approach minimizes surgical trauma and thereby helps decrease the surgical stress response, the patient experience can be optimized with use of enhanced recovery pathways (ERPs), a multimodal approach to perioperative care.

ERPs were initially proposed as a means of reducing the degree of surgical injury and the subsequent physiologic stress response.¹ This multimodal approach begins in the outpatient setting, includes preoperative and intraoperative modalities, and continues postoperatively. In patients with chronic pain, ERPs are even more important. Assigning “prehabilitation” and setting expectations for surgery goals are the first step in improving the patient experience. Intraoperative use of opioid-sparing anesthetics or regional anesthesia can improve recovery. After surgery, patients with chronic pain and/or opioid dependence receive medications on a schedule, along with short-interval follow-up. Ultimately, reducing acute postoperative pain may lower the risk of developing chronic pain.

In this article on patients with chronic pelvic pain, we highlight elements of ERPs within the framework of enhanced recovery after surgery. Many of the interventions proposed here also can be used to improve the surgical experience of patients without chronic pain.

Preadmission education, expectations, and optimization

Preoperative counseling for elective procedures generally occurs in the outpatient setting. Although discussion traditionally has covered the type of procedure and its associated risks, benefits, and alternatives, new guidelines suggest a more mindful and comprehensive approach is warranted. Individualized patient-centered education programs have a positive impact on the perioperative course, effecting reductions in preoperative anxiety, opioid requirements, and hospital length of stay.² From a pain management perspective, the clinician can take some time during preoperative counseling to inform the patient about the pain to be expected from surgery, the ways the pain will be managed intraoperatively and postoperatively, and the multimodal strategies that will be used throughout the patient’s stay² and that may allow for early discharge. Although preadmission counseling still should address expectations for the surgery, it also presents an opportunity both to assess the patient’s ability to cope with the physical and psychological stress of surgery and to offer the patient appropriate need-based interventions, such as prehabilitation and cognitive-behavioral therapy (CBT).

Prehabilitation is the process of increasing functional capacity before surgery in order to mitigate the stress of the surgery. Prehabilitation may involve aerobic exercise, strength training, or functional task training. The gynecologic surgery literature lacks prehabilitation data, but data in the colorectal literature support use of a prehabilitation program for patients having a scheduled colectomy, with improved postoperative recovery.³ Although the colectomy cohort predominantly included older men, the principle that guides program implementation is the same: improve recovery after the stress of abdominal surgery. Indeed, a patient who opts for an elective surgery may have to wait several weeks before undergoing the procedure, and during this period behavioral interventions can take effect. With postoperative complications occurring more often in patients with reduced functional capacity, the data support using prehabilitation to decrease the incidence of postoperative complications, particularly among the most vulnerable patients.⁴ However, a definitive recommendation on use of pelvic floor exercises as an adjunct to prehabilitation cannot be made.⁴ Successful prehabilitation takes at least 4 weeks and should be part of a multimodal program that addresses other behavioral risk factors that may negatively affect recovery.⁵ For example, current tobacco users have compromised pulmonary status and wound healing immediately after surgery, and use more opioids.⁶ Conversely, smoking cessation for as little as 4 weeks before surgery is associated with fewer complications.⁷ In addition, given that alcohol abuse may compromise the surgical stress response and increase the risk of opioid misuse, addressing alcohol abuse preoperatively may improve postoperative recovery.⁸

Treating mood disorders that coexist with chronic pain disorders is an important part of outpatient multimodal management—psychological intervention is a useful adjunct to prehabilitation in reducing perioperative anxiety and improving postoperative functional capacity.⁹ For patients who have chronic pain and are undergoing surgery, it is important to address any anxiety, depression, or poor coping skills (eg, pain catastrophizing) to try to reduce the postoperative pain experience and decrease the risk of chronic postsurgical pain (CPSP).^10,11

Before surgery, patients with chronic pain syndromes should be evaluated for emotional distress and pain coping ability. When possible, they should be referred to a pain psychologist, who can initiate CBT and other interventions. In addition, pain coping skills can be developed or reinforced to address preoperative anxiety and pain catastrophizing. These interventions, which may include use of visual imagery, breathing exercises, and other relaxation techniques, are applicable to the management of postoperative anxiety as well.

Read about preoperative multimodal analgesia and intra- and postoperative management.

Preoperative multimodal analgesia

Multimodal analgesia has several benefits. Simultaneous effects can be generated on multiple pain-related neurotransmitters, and a synergistic effect (eg, of acetaminophen and a nonsteroidal anti-inflammatory drug [NSAID]) can improve pain management. In addition, small doses of multiple medications can be given, instead of a large dose of a single medication. Of course, this strategy must be modified in elderly and patients with impaired renal function, who are at high risk for polypharmacy.

Preoperative administration of 3 medications—a selective cyclooxygenase 2 (COX-2) inhibitor, acetaminophen, and a gabapentinoid—is increasingly accepted as part of multimodal analgesia. The selective COX-2 inhibitor targets inflammatory prostaglandins and has anti-inflammatory and analgesic effects; acetaminophen, an effective analgesic with an unclear mechanism of action, can reduce postoperative opioid consumption¹² and works synergistically with NSAIDs¹³; and the gabapentinoid gabapentin has an analgesic effect likely contributing to decreased movement-related pain and subsequent improved functional recovery (data are mixed on whether continuing gabapentin after surgery prevents CPSP).¹⁴⁻¹⁶

Although serotonin and norepinephrine reuptake inhibitors (SNRIs) are commonly used in outpatient management of chronic pelvic pain, data suggest that their role in perioperative pain management is evolving. As SNRIs may reduce central nervous system (CNS) sensitization,¹⁷ their analgesic effect is thought to result from increased descending inhibitory tone in the CNS, which makes this class of medication ideal for patients with chronic neuropathic pain.¹⁵

Limited data also suggest a role for SNRIs in decreasing immediate postoperative pain and CPSP in high-risk patients. Studies of duloxetine use in the immediate perioperative period have found reduced postoperative acute pain and opioid use.^18,19 In addition, a short course of low-dose (37.5 mg) venlafaxine both before and after surgery has demonstrated a reduction in postoperative opioid use and a reduction in movement-related pain 6 months after surgery.²⁰

Intraoperative management

The surgical and anesthesia teams share the goal of optimizing both pain control and postoperative recovery. Surgical team members, who want longer-acting anesthetics for infiltration of incision sites, discuss with the anesthesiologist the appropriateness of using peripheral nerve blocks or neuraxial anesthesia, given the patient’s history and planned procedure. Anesthesia team members can improve anesthesia and minimize intraoperative opioid use through several methods, including total intravenous anesthesia,²¹ dexamethasone,²² ketorolac,²³ and intravenous ketamine. Ketamine, in particular, has a wide range of surgical applications and has been found to reduce postoperative pain, postoperative pain medication use, and the risk of CPSP.²

Incision sites should be infiltrated before and after surgery. Lidocaine traditionally is used for its rapid onset of action in reducing surgical site pain, but its short half-life may limit its applicability to postoperative pain. Recently, bupivacaine (half-life, 3.5 hours) and liposomal bupivacaine (24–34 hours) have gained more attention. Both of these medications appear to be as effective as lidocaine in reducing surgical site pain.²⁴

Transversus abdominis plane (TAP) blocks have been used as an adjunct in pain management during abdominopelvic surgery. Although initial data on postoperative pain and opioid use reductions with TAP blocks were inconclusive,²⁵ more recent data showed a role for TAP blocks in a multimodal approach for reducing opioid use during laparoscopic and open surgery.^26,27 Given the small number of studies on using liposomal bupivacaine for peripheral nerve blocks (eg, TAP blocks) in postoperative pain management, current data are inconclusive.²⁸

Postoperative management

The ERP approach calls for continuing multimodal analgesia after surgery—in most cases, scheduling early use of oral acetaminophen and ibuprofen, and providing short-acting, low-dose opioid analgesia as needed. All patients should be given a bowel regimen. Similar to undergoing prehabilitation for surgery, patients should prepare themselves for recovery. They should be encouraged to engage in early ambulation and oral intake and, when clinically appropriate, be given same-day discharge for minimally invasive surgical procedures.

Patients with chronic pain before surgery are at increased risk for suboptimal postoperative pain management, and those who are dependent on opioids require additional perioperative measures for adequate postoperative pain control. In these complicated cases, it is appropriate to enlist a pain specialist, potentially before surgery, to help plan perioperative and postoperative pain management.² Postoperative pain management for opioid-dependent patients should include pharmacologic and nonpharmacologic interventions, such as use of nonopioid medications (eg, gabapentin) and continuation of CBT. Patients with chronic pain should be closely followed up for assessment of postoperative pain control and recovery.

CASE Resolved

Surgical management is one aspect of the longer term multimodal pain management strategy for this patient. After preoperative pelvic floor physical therapy, she is receptive to starting a trial of an SNRI for her pain and mood symptoms. Both interventions allow for optimization of her preoperative physical and psychological status. Expectations are set that she will be discharged the day of surgery and that the surgery is but one component of her multimodal treatment plan. In addition, before surgery, she takes oral acetaminophen, gabapentin, and celecoxib—previously having had no contraindications to these medications. During surgery, bupivacaine is used for infiltration of all incision sites, and the anesthesia team administers ketamine and a TAP block. After surgery, the patient is prepared for same-day discharge and given the NSAIDs and acetaminophen she is scheduled to take over the next 72 hours. She is also given a limited prescription for oxycodone for breakthrough pain. An office visit 1 to 2 weeks after surgery is scheduled.

ERP strategies for surgical management of endometriosis have not only improved this patient’s postoperative recovery but also reduced her surgical stress response and subsequent transition to chronic postoperative pain. Many of the strategies used in this case are applicable to patients without chronic pain.

Share your thoughts! Send your Letter to the Editor to [email protected]. Please include your name and the city and state in which you practice.

CASE Chronic pelvic pain from endometriosis

A 40-year-old woman (G0) has a 20-year history of chronic pelvic pain. Stage III endometriosis is diagnosed on laparoscopic excision of endometriotic tissue. Postoperative pain symptoms include dysmenorrhea and deep dyspareunia, and the patient is feeling anxious. Physical examination reveals a retroverted uterus, right adnexal fullness and tenderness, and tenderness on palpation of the right levator ani and right obturator internus; rectovaginal examination findings are unremarkable. The patient, though now engaged in a pelvic floor physical therapy program, has yet to achieve the pain control she desires. After reviewing the treatment strategies for endometriosis with the patient, she elects definitive surgical management with minimally invasive hysterectomy and salpingo-oophorectomy. What pre-, intra-, and postoperative pain management plan do you devise for this patient?

Chronic pelvic pain presents a unique clinical challenge, as pain typically is multifactorial, and several peripheral pain generators may be involved. Although surgery can be performed to manage anatomically based disease processes, it does not address pain from musculoskeletal or neuropathic sources. A complete medical history and a physical examination are of utmost importance in developing a comprehensive multimodal management plan that may include surgery as treatment for the pain.

The standard of care for surgery is a minimally invasive approach (vaginal, laparoscopic, or robot-assisted laparoscopic), as it causes the least amount of trauma. Benefits of minimally invasive surgery include shorter hospitalization and faster recovery, likely owing to improved perioperative pain control, decreased blood loss, and fewer infections. Although this approach minimizes surgical trauma and thereby helps decrease the surgical stress response, the patient experience can be optimized with use of enhanced recovery pathways (ERPs), a multimodal approach to perioperative care.

ERPs were initially proposed as a means of reducing the degree of surgical injury and the subsequent physiologic stress response.¹ This multimodal approach begins in the outpatient setting, includes preoperative and intraoperative modalities, and continues postoperatively. In patients with chronic pain, ERPs are even more important. Assigning “prehabilitation” and setting expectations for surgery goals are the first step in improving the patient experience. Intraoperative use of opioid-sparing anesthetics or regional anesthesia can improve recovery. After surgery, patients with chronic pain and/or opioid dependence receive medications on a schedule, along with short-interval follow-up. Ultimately, reducing acute postoperative pain may lower the risk of developing chronic pain.

In this article on patients with chronic pelvic pain, we highlight elements of ERPs within the framework of enhanced recovery after surgery. Many of the interventions proposed here also can be used to improve the surgical experience of patients without chronic pain.

Preadmission education, expectations, and optimization

Preoperative counseling for elective procedures generally occurs in the outpatient setting. Although discussion traditionally has covered the type of procedure and its associated risks, benefits, and alternatives, new guidelines suggest a more mindful and comprehensive approach is warranted. Individualized patient-centered education programs have a positive impact on the perioperative course, effecting reductions in preoperative anxiety, opioid requirements, and hospital length of stay.² From a pain management perspective, the clinician can take some time during preoperative counseling to inform the patient about the pain to be expected from surgery, the ways the pain will be managed intraoperatively and postoperatively, and the multimodal strategies that will be used throughout the patient’s stay² and that may allow for early discharge. Although preadmission counseling still should address expectations for the surgery, it also presents an opportunity both to assess the patient’s ability to cope with the physical and psychological stress of surgery and to offer the patient appropriate need-based interventions, such as prehabilitation and cognitive-behavioral therapy (CBT).

Prehabilitation is the process of increasing functional capacity before surgery in order to mitigate the stress of the surgery. Prehabilitation may involve aerobic exercise, strength training, or functional task training. The gynecologic surgery literature lacks prehabilitation data, but data in the colorectal literature support use of a prehabilitation program for patients having a scheduled colectomy, with improved postoperative recovery.³ Although the colectomy cohort predominantly included older men, the principle that guides program implementation is the same: improve recovery after the stress of abdominal surgery. Indeed, a patient who opts for an elective surgery may have to wait several weeks before undergoing the procedure, and during this period behavioral interventions can take effect. With postoperative complications occurring more often in patients with reduced functional capacity, the data support using prehabilitation to decrease the incidence of postoperative complications, particularly among the most vulnerable patients.⁴ However, a definitive recommendation on use of pelvic floor exercises as an adjunct to prehabilitation cannot be made.⁴ Successful prehabilitation takes at least 4 weeks and should be part of a multimodal program that addresses other behavioral risk factors that may negatively affect recovery.⁵ For example, current tobacco users have compromised pulmonary status and wound healing immediately after surgery, and use more opioids.⁶ Conversely, smoking cessation for as little as 4 weeks before surgery is associated with fewer complications.⁷ In addition, given that alcohol abuse may compromise the surgical stress response and increase the risk of opioid misuse, addressing alcohol abuse preoperatively may improve postoperative recovery.⁸

Treating mood disorders that coexist with chronic pain disorders is an important part of outpatient multimodal management—psychological intervention is a useful adjunct to prehabilitation in reducing perioperative anxiety and improving postoperative functional capacity.⁹ For patients who have chronic pain and are undergoing surgery, it is important to address any anxiety, depression, or poor coping skills (eg, pain catastrophizing) to try to reduce the postoperative pain experience and decrease the risk of chronic postsurgical pain (CPSP).^10,11

Before surgery, patients with chronic pain syndromes should be evaluated for emotional distress and pain coping ability. When possible, they should be referred to a pain psychologist, who can initiate CBT and other interventions. In addition, pain coping skills can be developed or reinforced to address preoperative anxiety and pain catastrophizing. These interventions, which may include use of visual imagery, breathing exercises, and other relaxation techniques, are applicable to the management of postoperative anxiety as well.

Read about preoperative multimodal analgesia and intra- and postoperative management.

Preoperative multimodal analgesia

Multimodal analgesia has several benefits. Simultaneous effects can be generated on multiple pain-related neurotransmitters, and a synergistic effect (eg, of acetaminophen and a nonsteroidal anti-inflammatory drug [NSAID]) can improve pain management. In addition, small doses of multiple medications can be given, instead of a large dose of a single medication. Of course, this strategy must be modified in elderly and patients with impaired renal function, who are at high risk for polypharmacy.

Preoperative administration of 3 medications—a selective cyclooxygenase 2 (COX-2) inhibitor, acetaminophen, and a gabapentinoid—is increasingly accepted as part of multimodal analgesia. The selective COX-2 inhibitor targets inflammatory prostaglandins and has anti-inflammatory and analgesic effects; acetaminophen, an effective analgesic with an unclear mechanism of action, can reduce postoperative opioid consumption¹² and works synergistically with NSAIDs¹³; and the gabapentinoid gabapentin has an analgesic effect likely contributing to decreased movement-related pain and subsequent improved functional recovery (data are mixed on whether continuing gabapentin after surgery prevents CPSP).¹⁴⁻¹⁶

Although serotonin and norepinephrine reuptake inhibitors (SNRIs) are commonly used in outpatient management of chronic pelvic pain, data suggest that their role in perioperative pain management is evolving. As SNRIs may reduce central nervous system (CNS) sensitization,¹⁷ their analgesic effect is thought to result from increased descending inhibitory tone in the CNS, which makes this class of medication ideal for patients with chronic neuropathic pain.¹⁵

Limited data also suggest a role for SNRIs in decreasing immediate postoperative pain and CPSP in high-risk patients. Studies of duloxetine use in the immediate perioperative period have found reduced postoperative acute pain and opioid use.^18,19 In addition, a short course of low-dose (37.5 mg) venlafaxine both before and after surgery has demonstrated a reduction in postoperative opioid use and a reduction in movement-related pain 6 months after surgery.²⁰

Intraoperative management

The surgical and anesthesia teams share the goal of optimizing both pain control and postoperative recovery. Surgical team members, who want longer-acting anesthetics for infiltration of incision sites, discuss with the anesthesiologist the appropriateness of using peripheral nerve blocks or neuraxial anesthesia, given the patient’s history and planned procedure. Anesthesia team members can improve anesthesia and minimize intraoperative opioid use through several methods, including total intravenous anesthesia,²¹ dexamethasone,²² ketorolac,²³ and intravenous ketamine. Ketamine, in particular, has a wide range of surgical applications and has been found to reduce postoperative pain, postoperative pain medication use, and the risk of CPSP.²

Incision sites should be infiltrated before and after surgery. Lidocaine traditionally is used for its rapid onset of action in reducing surgical site pain, but its short half-life may limit its applicability to postoperative pain. Recently, bupivacaine (half-life, 3.5 hours) and liposomal bupivacaine (24–34 hours) have gained more attention. Both of these medications appear to be as effective as lidocaine in reducing surgical site pain.²⁴

Transversus abdominis plane (TAP) blocks have been used as an adjunct in pain management during abdominopelvic surgery. Although initial data on postoperative pain and opioid use reductions with TAP blocks were inconclusive,²⁵ more recent data showed a role for TAP blocks in a multimodal approach for reducing opioid use during laparoscopic and open surgery.^26,27 Given the small number of studies on using liposomal bupivacaine for peripheral nerve blocks (eg, TAP blocks) in postoperative pain management, current data are inconclusive.²⁸

Postoperative management

The ERP approach calls for continuing multimodal analgesia after surgery—in most cases, scheduling early use of oral acetaminophen and ibuprofen, and providing short-acting, low-dose opioid analgesia as needed. All patients should be given a bowel regimen. Similar to undergoing prehabilitation for surgery, patients should prepare themselves for recovery. They should be encouraged to engage in early ambulation and oral intake and, when clinically appropriate, be given same-day discharge for minimally invasive surgical procedures.

Patients with chronic pain before surgery are at increased risk for suboptimal postoperative pain management, and those who are dependent on opioids require additional perioperative measures for adequate postoperative pain control. In these complicated cases, it is appropriate to enlist a pain specialist, potentially before surgery, to help plan perioperative and postoperative pain management.² Postoperative pain management for opioid-dependent patients should include pharmacologic and nonpharmacologic interventions, such as use of nonopioid medications (eg, gabapentin) and continuation of CBT. Patients with chronic pain should be closely followed up for assessment of postoperative pain control and recovery.

CASE Resolved

Surgical management is one aspect of the longer term multimodal pain management strategy for this patient. After preoperative pelvic floor physical therapy, she is receptive to starting a trial of an SNRI for her pain and mood symptoms. Both interventions allow for optimization of her preoperative physical and psychological status. Expectations are set that she will be discharged the day of surgery and that the surgery is but one component of her multimodal treatment plan. In addition, before surgery, she takes oral acetaminophen, gabapentin, and celecoxib—previously having had no contraindications to these medications. During surgery, bupivacaine is used for infiltration of all incision sites, and the anesthesia team administers ketamine and a TAP block. After surgery, the patient is prepared for same-day discharge and given the NSAIDs and acetaminophen she is scheduled to take over the next 72 hours. She is also given a limited prescription for oxycodone for breakthrough pain. An office visit 1 to 2 weeks after surgery is scheduled.

ERP strategies for surgical management of endometriosis have not only improved this patient’s postoperative recovery but also reduced her surgical stress response and subsequent transition to chronic postoperative pain. Many of the strategies used in this case are applicable to patients without chronic pain.

Share your thoughts! Send your Letter to the Editor to [email protected]. Please include your name and the city and state in which you practice.

Endometriosis is one of the more daunting diagnoses that gynecologists treat. In this roundtable discussion, moderated by OBG Management Board Member Arnold P. Advincula, MD, 2 leading surgeons discuss endometriosis diagnosis as well as medical and surgical management.

First-time evaluation

Arnold P. Advincula, MD: When a patient presents to your practice for the first time and you suspect endometriosis, what considerations tailor your evaluation, and what does that evaluation involve?

Hye-Chun Hur, MD, MPH: The diagnosis is contingent on a patient’s presenting profile. How symptomatic is she? How old is she? What are her reproductive goals? The gold standard for diagnosis is a histologic diagnosis, which is surgical. Depending on the age profile, however, and how close she is to menopause, the patient may be managed medically. Even women in the young reproductive age group may be managed medically if symptoms are responsive to medical treatment.

Douglas N. Brown, MD: I agree. When a patient presents without a laparoscopy, or a tissue diagnosis, but the symptoms are consistent with likely endometriosis (depending on where she is in her reproductive cycle and what her goals are), I think treating with a first-line therapy—hormonal treatments such as progestin-only oral contraceptive pills—is acceptable. I usually conduct a treatment trial period of 3 to 6 months to see if she obtains any symptom relief.

If that first-line treatment fails, generally you can move to a second-line treatment.

I have a discussion in which I either offer a second-line treatment, such as medroxyprogesterone (Depo-Provera) or leuprolide acetate (Lupron Depot), or get a tissue diagnosis, if possible, by performing laparoscopy. If first-line or even second-line therapy fails, you need to consider doing a diagnostic laparoscopy to confirm or deny the diagnosis.

Dr. Advincula: Are there any points in the evaluation of a patient who visits your practice for the first time where you would immediately offer a surgical approach, as opposed to starting with medical management?

Dr. Hur: A large percentage of my patients undergo surgical evaluation, as surgical diagnosis is the gold standard. If you look at the literature, even among surgeons, the accuracy of visual diagnosis is not great.^1,2 I target individuals who are either not responsive to medical treatment or who have never tried medical treatment but are trying to conceive, so they are not medical candidates, or individuals who genuinely want a diagnosis for surgical management—sometimes even before first-line medical treatment.

Dr. Brown: Your examination sometimes also dictates your approach. A patient may never have had a laparoscopy or hormone therapy, but if you find uterosacral ligament nodularity, extreme pain on examination, and suspicious findings on ultrasound or otherwise, a diagnostic laparoscopy may be warranted to confirm the diagnosis.

Endometrioma management

Dr. Advincula: Let’s jump ahead. You have decided to proceed with laparoscopy and you encounter an endometrioma. What is your management strategy, particularly in a fertility-desiring patient?

Dr. Hur: Even if a woman has not undergone first-line medical treatment, if she is trying to conceive or presents with infertility, it’s a different balancing act for approaching the patient. When a woman presents, either with an ultrasound finding or an intraoperative finding of an endometrioma, I am a strong advocate of treating symptomatic disease, which means complete cyst excision. Good clinical data suggest that reproductive outcomes are improved for spontaneous pregnancy rates when you excise an endometrioma.^3-6

Dr. Advincula: What are the risks of excision of an endometrioma cyst that patients need to know about?

Dr. Brown: Current standard of care is cystectomy, stripping the cyst wall away from the ovarian cortex. There is some concern that the stripping process, depending on how long the endometrioma has been present within the ovary, can cause some destruction to the underlying oocytes and perhaps impact that ovary’s ability to produce viable eggs.

Some studies, from France in particular, have investigated different energy sources, such as plasma energy, that make it possible to remove part of the cyst and then use the plasma energy to vaporize the rest of the cyst wall that may be lying on the cortex. Researchers looked at anti-Müllerian hormone levels, and there does seem to be a difference in terms of how you remove the cyst.^7-9 This energy source is not available to everyone; it’s similar to laser but does not have as much penetration. Standard of care is still ovarian stripping.

The conversation with the patient—if she is already infertile and this cyst is a problem—would be that it likely needs to be removed. There is a chance that she may need assisted reproduction; she might not be able to get pregnant on her own due either to the presence of the endometrioma or to the surgical process of removing it and stripping.

Dr. Advincula: How soon after surgery can a patient start to pursue trying to get pregnant?

Dr. Hur: I think there is no time restraint outside of recovery. As long as the patient has a routine postoperative course, she can try to conceive, spontaneously or with assisted reproduction. Some data suggest, however, that ovarian reserve is diminished immediately after surgery.^10–12 If you look at the spontaneous clinical pregnancy outcomes, they are comparable 3 to 6 months postsurgery.^4,12–14

Dr. Brown: I agree. Time is of the essence with a lot of patients, many of whom present after age 35.

Dr. Hur: It’s also important to highlight that there are 2 presentations with endometrioma: the symptomatic patient and the asymptomatic patient. In the asymptomatic patient, her age, reproductive goals, and the bilaterality (whether it is present on both sides or on one side) of the endometrioma are important in deciding on a patient-centered surgical plan. For someone with a smaller cyst, unilateral presentation, and maybe older age at presentation, it may or may not impact assisted reproductive outcomes.

If the patient is not symptomatic and she is older with bilateral endometriomas less than 4 cm, some data suggest that patient might be better served in a conservative fashion.^6,15–17 Then, once she is done with assisted reproduction, we might be more aggressive surgically by treating the finding that would not resolve spontaneously without surgical management. It is important to highlight that endometriomas do not resolve on their own; they require surgical management.

Read about managing endometriosis for the patient not seeking fertility

Endometriosis management for the patient not seeking fertility

Dr. Advincula: Let’s now consider a patient on whom you have performed laparoscopy not only to diagnose and confirm the evidence of endometriosis but also to treat endometriosis, an endometrioma, and potentially deeply infiltrative disease. But this person is not trying to get pregnant. Postoperatively, what is your approach?

Dr. Brown: Suppressive therapy for this patient could be first-line or second-line therapy, such as a Lupron Depot or Depo-Provera. We keep the patient on suppressive therapy (whatever treatments work for her), until she’s ready to get pregnant; then we take her off. Hopefully she gets pregnant. After she delivers, we reinitiate suppressive therapy. I will follow these women throughout their reproductive cycle, and I think having a team of physicians who are all on the same page can help this patient manage her disease through her reproductive years.

Dr. Hur: If a patient presented warranting surgical management once, and she is not menopausal, the likelihood that disease will recur is quite high. Understanding the nature and the pathology of the disease, hormonal suppression would be warranted. Suppression is not just for between pregnancies, it’s until the patient reaches natural menopause. It’s also in the hopes of suppressing the disease so she does not need recurrent surgeries.

We typically do not operate unless patients have recurrence of symptoms that no longer respond to medical therapy. Our hope is to buy them more time closer to the age of natural menopause so that medical repercussions do not result in hysterectomy and ovary removal, which have other nongynecologic manifestations, including negative impact on bone and cardiac health.

The role of the LNG-IUD

Dr. Advincula: Something that often comes up is the role of a levonorgestrel-releasing intrauterine device (LNG-IUD) as one therapy option, either preoperatively or postoperatively. What is your perspective?

Dr. Hur: I reserve the LNG-IUD as a second-line therapy for patients, predominantly because it allows direct delivery of the medication to the womb (rather than systemic exposure of the medication). For patients who experience adverse effects due to systemic exposure to first-line treatments, it might be a great option. However, I do not believe that it consistently suppresses the ovaries, which we understand feeds the pathology of the hormonal stimulation, and so typically I will reserve it as a second-line treatment.

Dr. Brown: I utilize the LNG-IUD in a similar fashion. I may have patients who have had a diagnostic laparoscopy somewhere else and were referred to me because they now have known stage 3 or 4 endometriosis without endometriomas. Those patients, if they are going to need suppressive therapy after surgery and are not ready to get pregnant, do very well with the LNG-IUD, and I will place it during surgery under anesthesia. If a patient has endometriomas seen at the time of surgery, we could still place an LNG-IUD at the time of surgery. We may need to add on an additional medication, however, like another oral progesterone. I do have patients that use both an IUD and either combined oral contraceptive pills and/or oral progestins. Those patients usually have complicated cases with very deep infiltrative disease.

Read about managing endometriosis involving the bowel

Managing endometriosis involving the bowel

Dr. Advincula: Patients often are quite concerned when the words “endometriosis” and “bowel” come together. How do you manage disease that involves the bowel?

Illustration: Kimberly Martens for OBG Management

Dr. Hur: A lot of patients with endometriosis have what I call neighboring disease—it’s not limited just to the pelvis, but it involves the neighboring organs including the bowel and bladder. Patients can present with symptoms related to those adjacent organs. However, not all disease involving the bowel or bladder manifests with symptoms, and patients with symptoms may not have visible disease.

Typically, when a patient presents with symptoms of bowel involvement, where the bowel lumen is narrowed to more than 50% and/or she has functional manifestations (signs of obstruction that result in abnormal bowel function), we have serious conversations about a bowel resection. If she has full-thickness disease without significant bowel dysfunction—other than blood in her stool—sometimes we talk about more conservative treatment because of the long-term manifestations that a bowel resection could have.

Dr. Brown: I agree completely. It is important to have a good relationship with our colorectal surgeons. If I suspect that the patient has narrowing of the lumen of the large bowel or she actually has symptoms such as bloody diarrhea during menstruation—which is suggestive of deep, infiltrative and penetrative disease—I will often order a colonoscopy ahead of time to get confirmed biopsies. Then the patient discussion occurs with our colorectal surgeon, who operates with me jointly if we decide to proceed with a bowel resection. It’s important to have subspecialty colleagues involved in this care, because a low anterior resection is a very big surgery and there can be down-the-stream complications.

The importance of multidisciplinary care

Dr. Advincula: What are your perspectives on a multidisciplinary or interdisciplinary approach to the patient with endometriosis?

Dr. Brown: As I previously mentioned, it is important to develop a good relationship with colorectal surgery/urology. In addition, behavioral therapists may be involved in the care of patients with endometriosis, for a number of reasons. The disease process is fluid. It will change during the patient’s reproductive years, and you need to manage it accordingly based on her symptoms. Sometimes the diagnosis is not made for 5 to 10 years, and that can lead to other issues: depression, fibromyalgia, or irritable bowel syndrome.

The patient may have multiple issues plus endometriosis. I think having specialists such as gastroenterologists and behavioral therapists on board, as well as colorectal and urological surgeons who can perform these complex surgeries, is very beneficial to the patient. That way, she benefits from the team’s focus and is cared for from start to finish.

Dr. Hur: I like to call the abdomen a studio. It does not have separate compartments for each organ system. It’s one big room, and often the neighboring organs are involved, including the bowel and bladder. I think Dr. Brown’s observation—the multidisciplinary approach to a patient’s comprehensive care—is critical. Like any surgery, preoperative planning and preoperative assessment are essential, and these steps should include the patient. The discussion should cover not only the surgical outcomes that the surgeons expect, but also what the patient expects to be improved. For example, for patients with extensive disease and bowel involvement, a bowel resection is not always the right approach because it can have potential long-term sequelae. Balancing the risks associated with surgery with the long-term benefits is an important part of the discussion.

Dr. Advincula: Those are both excellent perspectives. Endometriosis is a very complicated disease state, does require a multidisciplinary approach to management, and there are implications and strategies that involve both the medical approach to management and the surgical approach.

Share your thoughts! Send your Letter to the Editor to [email protected]. Please include your name and the city and state in which you practice.

Endometriosis is one of the more daunting diagnoses that gynecologists treat. In this roundtable discussion, moderated by OBG Management Board Member Arnold P. Advincula, MD, 2 leading surgeons discuss endometriosis diagnosis as well as medical and surgical management.

First-time evaluation

Arnold P. Advincula, MD: When a patient presents to your practice for the first time and you suspect endometriosis, what considerations tailor your evaluation, and what does that evaluation involve?

Hye-Chun Hur, MD, MPH: The diagnosis is contingent on a patient’s presenting profile. How symptomatic is she? How old is she? What are her reproductive goals? The gold standard for diagnosis is a histologic diagnosis, which is surgical. Depending on the age profile, however, and how close she is to menopause, the patient may be managed medically. Even women in the young reproductive age group may be managed medically if symptoms are responsive to medical treatment.

Douglas N. Brown, MD: I agree. When a patient presents without a laparoscopy, or a tissue diagnosis, but the symptoms are consistent with likely endometriosis (depending on where she is in her reproductive cycle and what her goals are), I think treating with a first-line therapy—hormonal treatments such as progestin-only oral contraceptive pills—is acceptable. I usually conduct a treatment trial period of 3 to 6 months to see if she obtains any symptom relief.

If that first-line treatment fails, generally you can move to a second-line treatment.

I have a discussion in which I either offer a second-line treatment, such as medroxyprogesterone (Depo-Provera) or leuprolide acetate (Lupron Depot), or get a tissue diagnosis, if possible, by performing laparoscopy. If first-line or even second-line therapy fails, you need to consider doing a diagnostic laparoscopy to confirm or deny the diagnosis.

Dr. Advincula: Are there any points in the evaluation of a patient who visits your practice for the first time where you would immediately offer a surgical approach, as opposed to starting with medical management?

Dr. Hur: A large percentage of my patients undergo surgical evaluation, as surgical diagnosis is the gold standard. If you look at the literature, even among surgeons, the accuracy of visual diagnosis is not great.^1,2 I target individuals who are either not responsive to medical treatment or who have never tried medical treatment but are trying to conceive, so they are not medical candidates, or individuals who genuinely want a diagnosis for surgical management—sometimes even before first-line medical treatment.

Dr. Brown: Your examination sometimes also dictates your approach. A patient may never have had a laparoscopy or hormone therapy, but if you find uterosacral ligament nodularity, extreme pain on examination, and suspicious findings on ultrasound or otherwise, a diagnostic laparoscopy may be warranted to confirm the diagnosis.

Endometrioma management

Dr. Advincula: Let’s jump ahead. You have decided to proceed with laparoscopy and you encounter an endometrioma. What is your management strategy, particularly in a fertility-desiring patient?

Dr. Hur: Even if a woman has not undergone first-line medical treatment, if she is trying to conceive or presents with infertility, it’s a different balancing act for approaching the patient. When a woman presents, either with an ultrasound finding or an intraoperative finding of an endometrioma, I am a strong advocate of treating symptomatic disease, which means complete cyst excision. Good clinical data suggest that reproductive outcomes are improved for spontaneous pregnancy rates when you excise an endometrioma.^3-6

Dr. Advincula: What are the risks of excision of an endometrioma cyst that patients need to know about?

Dr. Brown: Current standard of care is cystectomy, stripping the cyst wall away from the ovarian cortex. There is some concern that the stripping process, depending on how long the endometrioma has been present within the ovary, can cause some destruction to the underlying oocytes and perhaps impact that ovary’s ability to produce viable eggs.

Some studies, from France in particular, have investigated different energy sources, such as plasma energy, that make it possible to remove part of the cyst and then use the plasma energy to vaporize the rest of the cyst wall that may be lying on the cortex. Researchers looked at anti-Müllerian hormone levels, and there does seem to be a difference in terms of how you remove the cyst.^7-9 This energy source is not available to everyone; it’s similar to laser but does not have as much penetration. Standard of care is still ovarian stripping.

The conversation with the patient—if she is already infertile and this cyst is a problem—would be that it likely needs to be removed. There is a chance that she may need assisted reproduction; she might not be able to get pregnant on her own due either to the presence of the endometrioma or to the surgical process of removing it and stripping.

Dr. Advincula: How soon after surgery can a patient start to pursue trying to get pregnant?

Dr. Hur: I think there is no time restraint outside of recovery. As long as the patient has a routine postoperative course, she can try to conceive, spontaneously or with assisted reproduction. Some data suggest, however, that ovarian reserve is diminished immediately after surgery.^10–12 If you look at the spontaneous clinical pregnancy outcomes, they are comparable 3 to 6 months postsurgery.^4,12–14

Dr. Brown: I agree. Time is of the essence with a lot of patients, many of whom present after age 35.

Dr. Hur: It’s also important to highlight that there are 2 presentations with endometrioma: the symptomatic patient and the asymptomatic patient. In the asymptomatic patient, her age, reproductive goals, and the bilaterality (whether it is present on both sides or on one side) of the endometrioma are important in deciding on a patient-centered surgical plan. For someone with a smaller cyst, unilateral presentation, and maybe older age at presentation, it may or may not impact assisted reproductive outcomes.

If the patient is not symptomatic and she is older with bilateral endometriomas less than 4 cm, some data suggest that patient might be better served in a conservative fashion.^6,15–17 Then, once she is done with assisted reproduction, we might be more aggressive surgically by treating the finding that would not resolve spontaneously without surgical management. It is important to highlight that endometriomas do not resolve on their own; they require surgical management.

Read about managing endometriosis for the patient not seeking fertility

Endometriosis management for the patient not seeking fertility

Dr. Advincula: Let’s now consider a patient on whom you have performed laparoscopy not only to diagnose and confirm the evidence of endometriosis but also to treat endometriosis, an endometrioma, and potentially deeply infiltrative disease. But this person is not trying to get pregnant. Postoperatively, what is your approach?

Dr. Brown: Suppressive therapy for this patient could be first-line or second-line therapy, such as a Lupron Depot or Depo-Provera. We keep the patient on suppressive therapy (whatever treatments work for her), until she’s ready to get pregnant; then we take her off. Hopefully she gets pregnant. After she delivers, we reinitiate suppressive therapy. I will follow these women throughout their reproductive cycle, and I think having a team of physicians who are all on the same page can help this patient manage her disease through her reproductive years.

Dr. Hur: If a patient presented warranting surgical management once, and she is not menopausal, the likelihood that disease will recur is quite high. Understanding the nature and the pathology of the disease, hormonal suppression would be warranted. Suppression is not just for between pregnancies, it’s until the patient reaches natural menopause. It’s also in the hopes of suppressing the disease so she does not need recurrent surgeries.

We typically do not operate unless patients have recurrence of symptoms that no longer respond to medical therapy. Our hope is to buy them more time closer to the age of natural menopause so that medical repercussions do not result in hysterectomy and ovary removal, which have other nongynecologic manifestations, including negative impact on bone and cardiac health.

The role of the LNG-IUD

Dr. Advincula: Something that often comes up is the role of a levonorgestrel-releasing intrauterine device (LNG-IUD) as one therapy option, either preoperatively or postoperatively. What is your perspective?

Dr. Hur: I reserve the LNG-IUD as a second-line therapy for patients, predominantly because it allows direct delivery of the medication to the womb (rather than systemic exposure of the medication). For patients who experience adverse effects due to systemic exposure to first-line treatments, it might be a great option. However, I do not believe that it consistently suppresses the ovaries, which we understand feeds the pathology of the hormonal stimulation, and so typically I will reserve it as a second-line treatment.

Dr. Brown: I utilize the LNG-IUD in a similar fashion. I may have patients who have had a diagnostic laparoscopy somewhere else and were referred to me because they now have known stage 3 or 4 endometriosis without endometriomas. Those patients, if they are going to need suppressive therapy after surgery and are not ready to get pregnant, do very well with the LNG-IUD, and I will place it during surgery under anesthesia. If a patient has endometriomas seen at the time of surgery, we could still place an LNG-IUD at the time of surgery. We may need to add on an additional medication, however, like another oral progesterone. I do have patients that use both an IUD and either combined oral contraceptive pills and/or oral progestins. Those patients usually have complicated cases with very deep infiltrative disease.

Read about managing endometriosis involving the bowel

Managing endometriosis involving the bowel

Dr. Advincula: Patients often are quite concerned when the words “endometriosis” and “bowel” come together. How do you manage disease that involves the bowel?

Illustration: Kimberly Martens for OBG Management

Dr. Hur: A lot of patients with endometriosis have what I call neighboring disease—it’s not limited just to the pelvis, but it involves the neighboring organs including the bowel and bladder. Patients can present with symptoms related to those adjacent organs. However, not all disease involving the bowel or bladder manifests with symptoms, and patients with symptoms may not have visible disease.

Typically, when a patient presents with symptoms of bowel involvement, where the bowel lumen is narrowed to more than 50% and/or she has functional manifestations (signs of obstruction that result in abnormal bowel function), we have serious conversations about a bowel resection. If she has full-thickness disease without significant bowel dysfunction—other than blood in her stool—sometimes we talk about more conservative treatment because of the long-term manifestations that a bowel resection could have.

Dr. Brown: I agree completely. It is important to have a good relationship with our colorectal surgeons. If I suspect that the patient has narrowing of the lumen of the large bowel or she actually has symptoms such as bloody diarrhea during menstruation—which is suggestive of deep, infiltrative and penetrative disease—I will often order a colonoscopy ahead of time to get confirmed biopsies. Then the patient discussion occurs with our colorectal surgeon, who operates with me jointly if we decide to proceed with a bowel resection. It’s important to have subspecialty colleagues involved in this care, because a low anterior resection is a very big surgery and there can be down-the-stream complications.

The importance of multidisciplinary care

Dr. Advincula: What are your perspectives on a multidisciplinary or interdisciplinary approach to the patient with endometriosis?

Dr. Brown: As I previously mentioned, it is important to develop a good relationship with colorectal surgery/urology. In addition, behavioral therapists may be involved in the care of patients with endometriosis, for a number of reasons. The disease process is fluid. It will change during the patient’s reproductive years, and you need to manage it accordingly based on her symptoms. Sometimes the diagnosis is not made for 5 to 10 years, and that can lead to other issues: depression, fibromyalgia, or irritable bowel syndrome.

The patient may have multiple issues plus endometriosis. I think having specialists such as gastroenterologists and behavioral therapists on board, as well as colorectal and urological surgeons who can perform these complex surgeries, is very beneficial to the patient. That way, she benefits from the team’s focus and is cared for from start to finish.

Dr. Hur: I like to call the abdomen a studio. It does not have separate compartments for each organ system. It’s one big room, and often the neighboring organs are involved, including the bowel and bladder. I think Dr. Brown’s observation—the multidisciplinary approach to a patient’s comprehensive care—is critical. Like any surgery, preoperative planning and preoperative assessment are essential, and these steps should include the patient. The discussion should cover not only the surgical outcomes that the surgeons expect, but also what the patient expects to be improved. For example, for patients with extensive disease and bowel involvement, a bowel resection is not always the right approach because it can have potential long-term sequelae. Balancing the risks associated with surgery with the long-term benefits is an important part of the discussion.

Dr. Advincula: Those are both excellent perspectives. Endometriosis is a very complicated disease state, does require a multidisciplinary approach to management, and there are implications and strategies that involve both the medical approach to management and the surgical approach.

Share your thoughts! Send your Letter to the Editor to [email protected]. Please include your name and the city and state in which you practice.

Endometriosis is one of the more daunting diagnoses that gynecologists treat. In this roundtable discussion, moderated by OBG Management Board Member Arnold P. Advincula, MD, 2 leading surgeons discuss endometriosis diagnosis as well as medical and surgical management.

First-time evaluation

Arnold P. Advincula, MD: When a patient presents to your practice for the first time and you suspect endometriosis, what considerations tailor your evaluation, and what does that evaluation involve?

Hye-Chun Hur, MD, MPH: The diagnosis is contingent on a patient’s presenting profile. How symptomatic is she? How old is she? What are her reproductive goals? The gold standard for diagnosis is a histologic diagnosis, which is surgical. Depending on the age profile, however, and how close she is to menopause, the patient may be managed medically. Even women in the young reproductive age group may be managed medically if symptoms are responsive to medical treatment.

Douglas N. Brown, MD: I agree. When a patient presents without a laparoscopy, or a tissue diagnosis, but the symptoms are consistent with likely endometriosis (depending on where she is in her reproductive cycle and what her goals are), I think treating with a first-line therapy—hormonal treatments such as progestin-only oral contraceptive pills—is acceptable. I usually conduct a treatment trial period of 3 to 6 months to see if she obtains any symptom relief.

If that first-line treatment fails, generally you can move to a second-line treatment.

I have a discussion in which I either offer a second-line treatment, such as medroxyprogesterone (Depo-Provera) or leuprolide acetate (Lupron Depot), or get a tissue diagnosis, if possible, by performing laparoscopy. If first-line or even second-line therapy fails, you need to consider doing a diagnostic laparoscopy to confirm or deny the diagnosis.

Dr. Advincula: Are there any points in the evaluation of a patient who visits your practice for the first time where you would immediately offer a surgical approach, as opposed to starting with medical management?

Dr. Hur: A large percentage of my patients undergo surgical evaluation, as surgical diagnosis is the gold standard. If you look at the literature, even among surgeons, the accuracy of visual diagnosis is not great.^1,2 I target individuals who are either not responsive to medical treatment or who have never tried medical treatment but are trying to conceive, so they are not medical candidates, or individuals who genuinely want a diagnosis for surgical management—sometimes even before first-line medical treatment.

Dr. Brown: Your examination sometimes also dictates your approach. A patient may never have had a laparoscopy or hormone therapy, but if you find uterosacral ligament nodularity, extreme pain on examination, and suspicious findings on ultrasound or otherwise, a diagnostic laparoscopy may be warranted to confirm the diagnosis.

Endometrioma management

Dr. Advincula: Let’s jump ahead. You have decided to proceed with laparoscopy and you encounter an endometrioma. What is your management strategy, particularly in a fertility-desiring patient?

Dr. Hur: Even if a woman has not undergone first-line medical treatment, if she is trying to conceive or presents with infertility, it’s a different balancing act for approaching the patient. When a woman presents, either with an ultrasound finding or an intraoperative finding of an endometrioma, I am a strong advocate of treating symptomatic disease, which means complete cyst excision. Good clinical data suggest that reproductive outcomes are improved for spontaneous pregnancy rates when you excise an endometrioma.^3-6

Dr. Advincula: What are the risks of excision of an endometrioma cyst that patients need to know about?

Dr. Brown: Current standard of care is cystectomy, stripping the cyst wall away from the ovarian cortex. There is some concern that the stripping process, depending on how long the endometrioma has been present within the ovary, can cause some destruction to the underlying oocytes and perhaps impact that ovary’s ability to produce viable eggs.

Some studies, from France in particular, have investigated different energy sources, such as plasma energy, that make it possible to remove part of the cyst and then use the plasma energy to vaporize the rest of the cyst wall that may be lying on the cortex. Researchers looked at anti-Müllerian hormone levels, and there does seem to be a difference in terms of how you remove the cyst.^7-9 This energy source is not available to everyone; it’s similar to laser but does not have as much penetration. Standard of care is still ovarian stripping.

The conversation with the patient—if she is already infertile and this cyst is a problem—would be that it likely needs to be removed. There is a chance that she may need assisted reproduction; she might not be able to get pregnant on her own due either to the presence of the endometrioma or to the surgical process of removing it and stripping.

Dr. Advincula: How soon after surgery can a patient start to pursue trying to get pregnant?

Dr. Hur: I think there is no time restraint outside of recovery. As long as the patient has a routine postoperative course, she can try to conceive, spontaneously or with assisted reproduction. Some data suggest, however, that ovarian reserve is diminished immediately after surgery.^10–12 If you look at the spontaneous clinical pregnancy outcomes, they are comparable 3 to 6 months postsurgery.^4,12–14

Dr. Brown: I agree. Time is of the essence with a lot of patients, many of whom present after age 35.

Dr. Hur: It’s also important to highlight that there are 2 presentations with endometrioma: the symptomatic patient and the asymptomatic patient. In the asymptomatic patient, her age, reproductive goals, and the bilaterality (whether it is present on both sides or on one side) of the endometrioma are important in deciding on a patient-centered surgical plan. For someone with a smaller cyst, unilateral presentation, and maybe older age at presentation, it may or may not impact assisted reproductive outcomes.

If the patient is not symptomatic and she is older with bilateral endometriomas less than 4 cm, some data suggest that patient might be better served in a conservative fashion.^6,15–17 Then, once she is done with assisted reproduction, we might be more aggressive surgically by treating the finding that would not resolve spontaneously without surgical management. It is important to highlight that endometriomas do not resolve on their own; they require surgical management.

Read about managing endometriosis for the patient not seeking fertility

Endometriosis management for the patient not seeking fertility

Dr. Advincula: Let’s now consider a patient on whom you have performed laparoscopy not only to diagnose and confirm the evidence of endometriosis but also to treat endometriosis, an endometrioma, and potentially deeply infiltrative disease. But this person is not trying to get pregnant. Postoperatively, what is your approach?

Dr. Brown: Suppressive therapy for this patient could be first-line or second-line therapy, such as a Lupron Depot or Depo-Provera. We keep the patient on suppressive therapy (whatever treatments work for her), until she’s ready to get pregnant; then we take her off. Hopefully she gets pregnant. After she delivers, we reinitiate suppressive therapy. I will follow these women throughout their reproductive cycle, and I think having a team of physicians who are all on the same page can help this patient manage her disease through her reproductive years.

Dr. Hur: If a patient presented warranting surgical management once, and she is not menopausal, the likelihood that disease will recur is quite high. Understanding the nature and the pathology of the disease, hormonal suppression would be warranted. Suppression is not just for between pregnancies, it’s until the patient reaches natural menopause. It’s also in the hopes of suppressing the disease so she does not need recurrent surgeries.

We typically do not operate unless patients have recurrence of symptoms that no longer respond to medical therapy. Our hope is to buy them more time closer to the age of natural menopause so that medical repercussions do not result in hysterectomy and ovary removal, which have other nongynecologic manifestations, including negative impact on bone and cardiac health.

The role of the LNG-IUD

Dr. Advincula: Something that often comes up is the role of a levonorgestrel-releasing intrauterine device (LNG-IUD) as one therapy option, either preoperatively or postoperatively. What is your perspective?

Dr. Hur: I reserve the LNG-IUD as a second-line therapy for patients, predominantly because it allows direct delivery of the medication to the womb (rather than systemic exposure of the medication). For patients who experience adverse effects due to systemic exposure to first-line treatments, it might be a great option. However, I do not believe that it consistently suppresses the ovaries, which we understand feeds the pathology of the hormonal stimulation, and so typically I will reserve it as a second-line treatment.

Dr. Brown: I utilize the LNG-IUD in a similar fashion. I may have patients who have had a diagnostic laparoscopy somewhere else and were referred to me because they now have known stage 3 or 4 endometriosis without endometriomas. Those patients, if they are going to need suppressive therapy after surgery and are not ready to get pregnant, do very well with the LNG-IUD, and I will place it during surgery under anesthesia. If a patient has endometriomas seen at the time of surgery, we could still place an LNG-IUD at the time of surgery. We may need to add on an additional medication, however, like another oral progesterone. I do have patients that use both an IUD and either combined oral contraceptive pills and/or oral progestins. Those patients usually have complicated cases with very deep infiltrative disease.

Read about managing endometriosis involving the bowel

Managing endometriosis involving the bowel

Dr. Advincula: Patients often are quite concerned when the words “endometriosis” and “bowel” come together. How do you manage disease that involves the bowel?

Illustration: Kimberly Martens for OBG Management

Dr. Hur: A lot of patients with endometriosis have what I call neighboring disease—it’s not limited just to the pelvis, but it involves the neighboring organs including the bowel and bladder. Patients can present with symptoms related to those adjacent organs. However, not all disease involving the bowel or bladder manifests with symptoms, and patients with symptoms may not have visible disease.

Typically, when a patient presents with symptoms of bowel involvement, where the bowel lumen is narrowed to more than 50% and/or she has functional manifestations (signs of obstruction that result in abnormal bowel function), we have serious conversations about a bowel resection. If she has full-thickness disease without significant bowel dysfunction—other than blood in her stool—sometimes we talk about more conservative treatment because of the long-term manifestations that a bowel resection could have.

Dr. Brown: I agree completely. It is important to have a good relationship with our colorectal surgeons. If I suspect that the patient has narrowing of the lumen of the large bowel or she actually has symptoms such as bloody diarrhea during menstruation—which is suggestive of deep, infiltrative and penetrative disease—I will often order a colonoscopy ahead of time to get confirmed biopsies. Then the patient discussion occurs with our colorectal surgeon, who operates with me jointly if we decide to proceed with a bowel resection. It’s important to have subspecialty colleagues involved in this care, because a low anterior resection is a very big surgery and there can be down-the-stream complications.

The importance of multidisciplinary care

Dr. Advincula: What are your perspectives on a multidisciplinary or interdisciplinary approach to the patient with endometriosis?

Dr. Brown: As I previously mentioned, it is important to develop a good relationship with colorectal surgery/urology. In addition, behavioral therapists may be involved in the care of patients with endometriosis, for a number of reasons. The disease process is fluid. It will change during the patient’s reproductive years, and you need to manage it accordingly based on her symptoms. Sometimes the diagnosis is not made for 5 to 10 years, and that can lead to other issues: depression, fibromyalgia, or irritable bowel syndrome.

The patient may have multiple issues plus endometriosis. I think having specialists such as gastroenterologists and behavioral therapists on board, as well as colorectal and urological surgeons who can perform these complex surgeries, is very beneficial to the patient. That way, she benefits from the team’s focus and is cared for from start to finish.

Dr. Hur: I like to call the abdomen a studio. It does not have separate compartments for each organ system. It’s one big room, and often the neighboring organs are involved, including the bowel and bladder. I think Dr. Brown’s observation—the multidisciplinary approach to a patient’s comprehensive care—is critical. Like any surgery, preoperative planning and preoperative assessment are essential, and these steps should include the patient. The discussion should cover not only the surgical outcomes that the surgeons expect, but also what the patient expects to be improved. For example, for patients with extensive disease and bowel involvement, a bowel resection is not always the right approach because it can have potential long-term sequelae. Balancing the risks associated with surgery with the long-term benefits is an important part of the discussion.

Dr. Advincula: Those are both excellent perspectives. Endometriosis is a very complicated disease state, does require a multidisciplinary approach to management, and there are implications and strategies that involve both the medical approach to management and the surgical approach.

Share your thoughts! Send your Letter to the Editor to [email protected]. Please include your name and the city and state in which you practice.

In this Update, I report on the latest US Preventive Services Task Force (USPSTF) cervical cancer screening recommendations. In addition, I describe the results of 2 studies, a large prospective multicenter study of the accuracy of sentinel lymph node (SLN) biopsy in endometrial cancer, and a proof-of-concept review of use of checkpoint blockade to increase immune response and of its possible role in endometrial cancer.

hrHPV testing used alone as primary screening for cervical cancer: USPSTF recommendations

US Preventive Services Task Force. Draft recommendation statement: cervical cancer: screening. https://www.uspreventiveservicestaskforce.org/Page/Document/draft-recommendation-statement/cervical-cancer-screening2. Published October 2017. Accessed February 5, 2018.

‌

Despite our rapid advances in understanding the molecular underpinnings of cancer, gynecologic malignancies are still a major cause of morbidity and mortality among women. Cervical cancer stands as an example of how a cancer screening test can be implemented to reduce mortality. In this section, I report on the USPSTF cervical cancer screening recommendations, which were updated in October 2017.

Even with the widespread implementation of screening programs for cervical cancer in the United States, 13,240 women will be diagnosed with the disease in 2018, and 4,170 will die from cervical cancer.¹ Most often, cervical cancer occurs in women who have not been adequately screened. It is now recognized that the human papillomavirus (HPV) is the cause of cervical cancer.²

While cervical cytology has long been used as a screening test for cervical cancer, testing for high-risk HPV subtypes (hrHPV testing) also has been used as a screening modality. Traditionally, hrHPV testing is used in combination with cervical cytology, so called cotesting. There is convincing evidence that cervical cytology, as well as strategies that use hrHPV testing, can detect high-grade cervical precancers and cancers and thereby reduce mortality. However, cervical cancer screening is also associated with frequent follow-ups, invasive procedures performed to assess abnormal results, psychological distress, and adverse pregnancy outcomes of treatment for precancerous lesions.

The USPSTF based its new cervical cancer screening recommendations on clinical trial data and decision modeling of various screening strategies, and weighed the benefits and harms of each strategy.

Recommendations from the USPSTF

hrHPV screening for cervical cancer.  TheUSPSTF recommends screening with cervical cytology every 3 years for women 21 to 29 years of age. For women 30 to 65 years of age, screening with cytology every 3 years, or hrHPV testing alone used every 5 years, is recommended.

Data from large randomized trials suggest cytologic screening is slightly less sensitive than hrHPV testing in detecting high-grade (grade 2 or 3) cervical intraepithelial neoplasia (CIN). However, hrHPV testing results in more follow-up tests and colposcopies. In a decision model, the USPSTF found that cotesting increased the number of follow-up tests but did not increase detection of grade 3 CIN or invasive cancer. This is the first clinical guideline to recommend hrHPV testing used alone for screening. The American College of Obstetricians and Gynecologists (ACOG) continues to recommend cotesting (cytology in combination with hrHPV) as a primary screening modality in this population.³

Exceptions. According to the USPSTF, 3 populations should not be screened: women over 65 years of age with adequate prior screening who are not otherwise at high risk for cervical cancer; women under  21 years of age; and women who have had a hysterectomy and do not have a history of grade 2 or 3 CIN or cancer.

Summary. The USPSTF recommendations are intended for the general population and are not applicable to women with a history of high-grade CIN or cervical cancer, women with in utero exposure to diethylstilbestrol, and women who are immunocompromised. The remaining USPSTF recommendations are largely in line with guidelines published by ACOG and other groups.^3,4

Read about SLN biopsy to stage endometrial cancer

SLN biopsy for staging endometrial cancer

Rossi EC, Kowalski LD, Scalici J, et al. A comparison of sentinel lymph node biopsy to lymphadenectomy for endometrial cancer staging (FIRES trial): a multicentre, prospective, cohort study. Lancet Oncol. 2017;18(3):384-392.

‌

Surgery is the cornerstone of treatment for most gynecologic cancers. The widespread use of minimally invasive surgical techniques and the introduction of less radical procedures for gynecologic cancers have helped reduce surgical morbidity.

For endometrial cancer, the role of lymphadenectomy is controversial. Data from prospective trials of this procedure suggest an association with increased morbidity and long-term sequelae, such as lymphedema, and no association with improved survival.^5,6

SLN biopsy is an important advance and a potential alternative nodal evaluation method that may be associated with decreased morbidity. In this more limited assessment technique, the first nodal drainage basins of a tumor are identified and removed for pathologic evaluation.

Accuracy of SLN biopsy in endometrial cancer was the subject of Rossi and colleagues' recent large prospective multicenter study, the Fluorescence Imaging for Robotic Endometrial Sentinel lymph node biopsy (FIRES) trial.

Details of the study

Rossi and colleagues conducted the FIRES trial to estimate the sensitivity of SLN biopsy in detecting nodal metastases in women with stage I endometrial cancer. Patients (N = 385) from 10 US sites were enrolled in the study. SLN evaluation was performed after cervical injection of indocyanine green followed by robotic-assisted hysterectomy. After identification of the SLN, participants underwent pelvic lymphadenectomy. Performance of para-aortic lymphadenectomy was optional.

Mapping of the SLN was feasible in 86% of patients, including bilateral mapping in 52%. Twelve percent of the participants had nodal metastases. SLN biopsy had a sensitivity of 97% in women who had identification of the SLNs. Similarly, the negative predictive value was high, 99.6%. The procedure was associated with acceptable short-term toxicity with adverse events in 9% of study participants. Common complications included neurologic complications, respiratory distress, nausea and vomiting, and, in 3 patients, bowel injury.

Accurate detection of nodal metastases. Results of the study suggest SLN biopsy is accurate in detecting nodal metastases in women with endometrial cancer. Although long-term toxicity was not examined, other work suggests the lymphedema rates associated with SLN biopsy may be lower than those of lymphadenectomy. While the study described impressive performance characteristics, there remain technical challenges. Even among skilled surgeons trained for the protocol, there was no nodal mapping in nearly half of the women with endometrial cancer. Women without node mapping require full lymphadenectomy thus negating the possible benefits of the procedure.

Read about immunotherapy for gynecologic cancers

Immunotherapy for gynecologic cancers

Le DT, Durham JN, Smith KN, et al. Mismatch repair deficiency predicts response of solid tumors to PD-1 blockade. Science. 2017;357(6349):409-413.

‌

In oncology, precision medicine is rapidly becoming a standard treatment approach. Therapies are being used to target specific genetic alterations in tumors. In cancer immunotherapy, the immune system is being used to facilitate clearance of cancer cells.

The most common mechanism of action of clinically used immunotherapeutic agents is blockade of programmed cell death protein 1 (PD-1), a lymphocyte receptor that prevents the immune system from targeting the body's own cells.⁷ Cancers that have mutations in the DNA mismatch repair (MMR) proteins display microsatellite instability (MSI) and produce high levels of abnormal proteins.⁸ These abnormal proteins serve as tumor antigens that can be targeted by the body's normal immune system. 

In May 2017, the US Food and Drug Administration (FDA) granted accelerated approval of the PD-1 blocking antibody pembrolizumab for the treatment of unresectable or metastatic MSI-high (MSI-H) or MMR-deficient solid tumors.9 The approval was based on data from 149 patients treated in 5 studies that demonstrated a response rate of 39.6%, including responses that lasted at least 6 months in 78% of participants. This was the first ever cancer drug that received FDA approval based on a tumor's biomarker profile without regard to the site of origin. I describe the results of a study by Le and colleagues that examines the possible role of immunotherapy in a variety of solid tumors in this section.

Details of the study

This study examined the clinical efficacy of PD-1 blockade in 86 patients with advanced, MMR-deficient tumors from 12 different sites. Endometrial cancer was the second most frequent primary tumor site in 17% of patients. Within the cohort, the overall objective response rate was 53%, which included 21% of patients with complete radiographic response (no imaging evidence of cancer). Disease control, either complete or partial response or stable disease, was achieved in 77% of patients. After a median follow-up of 12.5 months, neither the median progression-free survival (PFS) nor median overall survival had been reached. The authors estimated that 2-year overall survival was 64%, substantially higher than expected for patients with advanced solid tumors.

Le and colleagues also performed several in vivo laboratory experiments to explore the mechanisms by which patients responded. In addition, they used sequencing to determine the prevalence of MMR deficiency in 12,019 cancer samples that included 32 distinct tumor types (FIGURE). Endometrial cancer had the highest frequency of MMR deficiency (17%). Four percent of cervical cancers and less than 2% of ovarian cancers were MMR-deficient.

The promise of immunotherapy for endometrial cancer. This study's data and other emerging data have important implications for women with gynecologic cancer, particularly endometrial cancer. First, given the frequency of MMR mutations among women with endometrial cancer, MMR testing should be strongly considered for these patients. Many institutions have protocols for reflex testing with immunohistochemistry for women with endometrial cancer. For women with positive test results, germline sequencing can be performed to determine if they have an inherited MMR deficiency, Lynch syndrome. Presence of an MMR deficiency is an important factor in cancer screening and potential treatment.

Second, the impressive results of PD-1 blockade in patients with MMR-deficient tumors suggest that this treatment strategy may be important for women with recurrent or metastatic endometrial cancer. The ideal timing of immunotherapy for women with endometrial cancer is an area of active ongoing study.

Share your thoughts! Send your Letter to the Editor to [email protected]. Please include your name and the city and state in which you practice.

In this Update, I report on the latest US Preventive Services Task Force (USPSTF) cervical cancer screening recommendations. In addition, I describe the results of 2 studies, a large prospective multicenter study of the accuracy of sentinel lymph node (SLN) biopsy in endometrial cancer, and a proof-of-concept review of use of checkpoint blockade to increase immune response and of its possible role in endometrial cancer.

hrHPV testing used alone as primary screening for cervical cancer: USPSTF recommendations

US Preventive Services Task Force. Draft recommendation statement: cervical cancer: screening. https://www.uspreventiveservicestaskforce.org/Page/Document/draft-recommendation-statement/cervical-cancer-screening2. Published October 2017. Accessed February 5, 2018.

‌

Despite our rapid advances in understanding the molecular underpinnings of cancer, gynecologic malignancies are still a major cause of morbidity and mortality among women. Cervical cancer stands as an example of how a cancer screening test can be implemented to reduce mortality. In this section, I report on the USPSTF cervical cancer screening recommendations, which were updated in October 2017.

Even with the widespread implementation of screening programs for cervical cancer in the United States, 13,240 women will be diagnosed with the disease in 2018, and 4,170 will die from cervical cancer.¹ Most often, cervical cancer occurs in women who have not been adequately screened. It is now recognized that the human papillomavirus (HPV) is the cause of cervical cancer.²

While cervical cytology has long been used as a screening test for cervical cancer, testing for high-risk HPV subtypes (hrHPV testing) also has been used as a screening modality. Traditionally, hrHPV testing is used in combination with cervical cytology, so called cotesting. There is convincing evidence that cervical cytology, as well as strategies that use hrHPV testing, can detect high-grade cervical precancers and cancers and thereby reduce mortality. However, cervical cancer screening is also associated with frequent follow-ups, invasive procedures performed to assess abnormal results, psychological distress, and adverse pregnancy outcomes of treatment for precancerous lesions.

The USPSTF based its new cervical cancer screening recommendations on clinical trial data and decision modeling of various screening strategies, and weighed the benefits and harms of each strategy.

Recommendations from the USPSTF

hrHPV screening for cervical cancer.  TheUSPSTF recommends screening with cervical cytology every 3 years for women 21 to 29 years of age. For women 30 to 65 years of age, screening with cytology every 3 years, or hrHPV testing alone used every 5 years, is recommended.

Data from large randomized trials suggest cytologic screening is slightly less sensitive than hrHPV testing in detecting high-grade (grade 2 or 3) cervical intraepithelial neoplasia (CIN). However, hrHPV testing results in more follow-up tests and colposcopies. In a decision model, the USPSTF found that cotesting increased the number of follow-up tests but did not increase detection of grade 3 CIN or invasive cancer. This is the first clinical guideline to recommend hrHPV testing used alone for screening. The American College of Obstetricians and Gynecologists (ACOG) continues to recommend cotesting (cytology in combination with hrHPV) as a primary screening modality in this population.³

Exceptions. According to the USPSTF, 3 populations should not be screened: women over 65 years of age with adequate prior screening who are not otherwise at high risk for cervical cancer; women under  21 years of age; and women who have had a hysterectomy and do not have a history of grade 2 or 3 CIN or cancer.

Summary. The USPSTF recommendations are intended for the general population and are not applicable to women with a history of high-grade CIN or cervical cancer, women with in utero exposure to diethylstilbestrol, and women who are immunocompromised. The remaining USPSTF recommendations are largely in line with guidelines published by ACOG and other groups.^3,4

Read about SLN biopsy to stage endometrial cancer

SLN biopsy for staging endometrial cancer

Rossi EC, Kowalski LD, Scalici J, et al. A comparison of sentinel lymph node biopsy to lymphadenectomy for endometrial cancer staging (FIRES trial): a multicentre, prospective, cohort study. Lancet Oncol. 2017;18(3):384-392.

‌

Surgery is the cornerstone of treatment for most gynecologic cancers. The widespread use of minimally invasive surgical techniques and the introduction of less radical procedures for gynecologic cancers have helped reduce surgical morbidity.

For endometrial cancer, the role of lymphadenectomy is controversial. Data from prospective trials of this procedure suggest an association with increased morbidity and long-term sequelae, such as lymphedema, and no association with improved survival.^5,6

SLN biopsy is an important advance and a potential alternative nodal evaluation method that may be associated with decreased morbidity. In this more limited assessment technique, the first nodal drainage basins of a tumor are identified and removed for pathologic evaluation.

Accuracy of SLN biopsy in endometrial cancer was the subject of Rossi and colleagues' recent large prospective multicenter study, the Fluorescence Imaging for Robotic Endometrial Sentinel lymph node biopsy (FIRES) trial.

Details of the study

Rossi and colleagues conducted the FIRES trial to estimate the sensitivity of SLN biopsy in detecting nodal metastases in women with stage I endometrial cancer. Patients (N = 385) from 10 US sites were enrolled in the study. SLN evaluation was performed after cervical injection of indocyanine green followed by robotic-assisted hysterectomy. After identification of the SLN, participants underwent pelvic lymphadenectomy. Performance of para-aortic lymphadenectomy was optional.

Mapping of the SLN was feasible in 86% of patients, including bilateral mapping in 52%. Twelve percent of the participants had nodal metastases. SLN biopsy had a sensitivity of 97% in women who had identification of the SLNs. Similarly, the negative predictive value was high, 99.6%. The procedure was associated with acceptable short-term toxicity with adverse events in 9% of study participants. Common complications included neurologic complications, respiratory distress, nausea and vomiting, and, in 3 patients, bowel injury.

Accurate detection of nodal metastases. Results of the study suggest SLN biopsy is accurate in detecting nodal metastases in women with endometrial cancer. Although long-term toxicity was not examined, other work suggests the lymphedema rates associated with SLN biopsy may be lower than those of lymphadenectomy. While the study described impressive performance characteristics, there remain technical challenges. Even among skilled surgeons trained for the protocol, there was no nodal mapping in nearly half of the women with endometrial cancer. Women without node mapping require full lymphadenectomy thus negating the possible benefits of the procedure.

Read about immunotherapy for gynecologic cancers

Immunotherapy for gynecologic cancers

Le DT, Durham JN, Smith KN, et al. Mismatch repair deficiency predicts response of solid tumors to PD-1 blockade. Science. 2017;357(6349):409-413.

‌

In oncology, precision medicine is rapidly becoming a standard treatment approach. Therapies are being used to target specific genetic alterations in tumors. In cancer immunotherapy, the immune system is being used to facilitate clearance of cancer cells.

The most common mechanism of action of clinically used immunotherapeutic agents is blockade of programmed cell death protein 1 (PD-1), a lymphocyte receptor that prevents the immune system from targeting the body's own cells.⁷ Cancers that have mutations in the DNA mismatch repair (MMR) proteins display microsatellite instability (MSI) and produce high levels of abnormal proteins.⁸ These abnormal proteins serve as tumor antigens that can be targeted by the body's normal immune system. 

In May 2017, the US Food and Drug Administration (FDA) granted accelerated approval of the PD-1 blocking antibody pembrolizumab for the treatment of unresectable or metastatic MSI-high (MSI-H) or MMR-deficient solid tumors.9 The approval was based on data from 149 patients treated in 5 studies that demonstrated a response rate of 39.6%, including responses that lasted at least 6 months in 78% of participants. This was the first ever cancer drug that received FDA approval based on a tumor's biomarker profile without regard to the site of origin. I describe the results of a study by Le and colleagues that examines the possible role of immunotherapy in a variety of solid tumors in this section.

Details of the study

This study examined the clinical efficacy of PD-1 blockade in 86 patients with advanced, MMR-deficient tumors from 12 different sites. Endometrial cancer was the second most frequent primary tumor site in 17% of patients. Within the cohort, the overall objective response rate was 53%, which included 21% of patients with complete radiographic response (no imaging evidence of cancer). Disease control, either complete or partial response or stable disease, was achieved in 77% of patients. After a median follow-up of 12.5 months, neither the median progression-free survival (PFS) nor median overall survival had been reached. The authors estimated that 2-year overall survival was 64%, substantially higher than expected for patients with advanced solid tumors.

Le and colleagues also performed several in vivo laboratory experiments to explore the mechanisms by which patients responded. In addition, they used sequencing to determine the prevalence of MMR deficiency in 12,019 cancer samples that included 32 distinct tumor types (FIGURE). Endometrial cancer had the highest frequency of MMR deficiency (17%). Four percent of cervical cancers and less than 2% of ovarian cancers were MMR-deficient.

The promise of immunotherapy for endometrial cancer. This study's data and other emerging data have important implications for women with gynecologic cancer, particularly endometrial cancer. First, given the frequency of MMR mutations among women with endometrial cancer, MMR testing should be strongly considered for these patients. Many institutions have protocols for reflex testing with immunohistochemistry for women with endometrial cancer. For women with positive test results, germline sequencing can be performed to determine if they have an inherited MMR deficiency, Lynch syndrome. Presence of an MMR deficiency is an important factor in cancer screening and potential treatment.

Second, the impressive results of PD-1 blockade in patients with MMR-deficient tumors suggest that this treatment strategy may be important for women with recurrent or metastatic endometrial cancer. The ideal timing of immunotherapy for women with endometrial cancer is an area of active ongoing study.

Share your thoughts! Send your Letter to the Editor to [email protected]. Please include your name and the city and state in which you practice.

In this Update, I report on the latest US Preventive Services Task Force (USPSTF) cervical cancer screening recommendations. In addition, I describe the results of 2 studies, a large prospective multicenter study of the accuracy of sentinel lymph node (SLN) biopsy in endometrial cancer, and a proof-of-concept review of use of checkpoint blockade to increase immune response and of its possible role in endometrial cancer.

hrHPV testing used alone as primary screening for cervical cancer: USPSTF recommendations

US Preventive Services Task Force. Draft recommendation statement: cervical cancer: screening. https://www.uspreventiveservicestaskforce.org/Page/Document/draft-recommendation-statement/cervical-cancer-screening2. Published October 2017. Accessed February 5, 2018.

‌

Despite our rapid advances in understanding the molecular underpinnings of cancer, gynecologic malignancies are still a major cause of morbidity and mortality among women. Cervical cancer stands as an example of how a cancer screening test can be implemented to reduce mortality. In this section, I report on the USPSTF cervical cancer screening recommendations, which were updated in October 2017.

Even with the widespread implementation of screening programs for cervical cancer in the United States, 13,240 women will be diagnosed with the disease in 2018, and 4,170 will die from cervical cancer.¹ Most often, cervical cancer occurs in women who have not been adequately screened. It is now recognized that the human papillomavirus (HPV) is the cause of cervical cancer.²

While cervical cytology has long been used as a screening test for cervical cancer, testing for high-risk HPV subtypes (hrHPV testing) also has been used as a screening modality. Traditionally, hrHPV testing is used in combination with cervical cytology, so called cotesting. There is convincing evidence that cervical cytology, as well as strategies that use hrHPV testing, can detect high-grade cervical precancers and cancers and thereby reduce mortality. However, cervical cancer screening is also associated with frequent follow-ups, invasive procedures performed to assess abnormal results, psychological distress, and adverse pregnancy outcomes of treatment for precancerous lesions.

The USPSTF based its new cervical cancer screening recommendations on clinical trial data and decision modeling of various screening strategies, and weighed the benefits and harms of each strategy.

Recommendations from the USPSTF

hrHPV screening for cervical cancer.  TheUSPSTF recommends screening with cervical cytology every 3 years for women 21 to 29 years of age. For women 30 to 65 years of age, screening with cytology every 3 years, or hrHPV testing alone used every 5 years, is recommended.

Data from large randomized trials suggest cytologic screening is slightly less sensitive than hrHPV testing in detecting high-grade (grade 2 or 3) cervical intraepithelial neoplasia (CIN). However, hrHPV testing results in more follow-up tests and colposcopies. In a decision model, the USPSTF found that cotesting increased the number of follow-up tests but did not increase detection of grade 3 CIN or invasive cancer. This is the first clinical guideline to recommend hrHPV testing used alone for screening. The American College of Obstetricians and Gynecologists (ACOG) continues to recommend cotesting (cytology in combination with hrHPV) as a primary screening modality in this population.³

Exceptions. According to the USPSTF, 3 populations should not be screened: women over 65 years of age with adequate prior screening who are not otherwise at high risk for cervical cancer; women under  21 years of age; and women who have had a hysterectomy and do not have a history of grade 2 or 3 CIN or cancer.

Summary. The USPSTF recommendations are intended for the general population and are not applicable to women with a history of high-grade CIN or cervical cancer, women with in utero exposure to diethylstilbestrol, and women who are immunocompromised. The remaining USPSTF recommendations are largely in line with guidelines published by ACOG and other groups.^3,4

Read about SLN biopsy to stage endometrial cancer

SLN biopsy for staging endometrial cancer

Rossi EC, Kowalski LD, Scalici J, et al. A comparison of sentinel lymph node biopsy to lymphadenectomy for endometrial cancer staging (FIRES trial): a multicentre, prospective, cohort study. Lancet Oncol. 2017;18(3):384-392.

‌

Surgery is the cornerstone of treatment for most gynecologic cancers. The widespread use of minimally invasive surgical techniques and the introduction of less radical procedures for gynecologic cancers have helped reduce surgical morbidity.

For endometrial cancer, the role of lymphadenectomy is controversial. Data from prospective trials of this procedure suggest an association with increased morbidity and long-term sequelae, such as lymphedema, and no association with improved survival.^5,6

SLN biopsy is an important advance and a potential alternative nodal evaluation method that may be associated with decreased morbidity. In this more limited assessment technique, the first nodal drainage basins of a tumor are identified and removed for pathologic evaluation.

Accuracy of SLN biopsy in endometrial cancer was the subject of Rossi and colleagues' recent large prospective multicenter study, the Fluorescence Imaging for Robotic Endometrial Sentinel lymph node biopsy (FIRES) trial.

Details of the study

Rossi and colleagues conducted the FIRES trial to estimate the sensitivity of SLN biopsy in detecting nodal metastases in women with stage I endometrial cancer. Patients (N = 385) from 10 US sites were enrolled in the study. SLN evaluation was performed after cervical injection of indocyanine green followed by robotic-assisted hysterectomy. After identification of the SLN, participants underwent pelvic lymphadenectomy. Performance of para-aortic lymphadenectomy was optional.

Mapping of the SLN was feasible in 86% of patients, including bilateral mapping in 52%. Twelve percent of the participants had nodal metastases. SLN biopsy had a sensitivity of 97% in women who had identification of the SLNs. Similarly, the negative predictive value was high, 99.6%. The procedure was associated with acceptable short-term toxicity with adverse events in 9% of study participants. Common complications included neurologic complications, respiratory distress, nausea and vomiting, and, in 3 patients, bowel injury.

Accurate detection of nodal metastases. Results of the study suggest SLN biopsy is accurate in detecting nodal metastases in women with endometrial cancer. Although long-term toxicity was not examined, other work suggests the lymphedema rates associated with SLN biopsy may be lower than those of lymphadenectomy. While the study described impressive performance characteristics, there remain technical challenges. Even among skilled surgeons trained for the protocol, there was no nodal mapping in nearly half of the women with endometrial cancer. Women without node mapping require full lymphadenectomy thus negating the possible benefits of the procedure.

Read about immunotherapy for gynecologic cancers

Immunotherapy for gynecologic cancers

Le DT, Durham JN, Smith KN, et al. Mismatch repair deficiency predicts response of solid tumors to PD-1 blockade. Science. 2017;357(6349):409-413.

‌

In oncology, precision medicine is rapidly becoming a standard treatment approach. Therapies are being used to target specific genetic alterations in tumors. In cancer immunotherapy, the immune system is being used to facilitate clearance of cancer cells.

The most common mechanism of action of clinically used immunotherapeutic agents is blockade of programmed cell death protein 1 (PD-1), a lymphocyte receptor that prevents the immune system from targeting the body's own cells.⁷ Cancers that have mutations in the DNA mismatch repair (MMR) proteins display microsatellite instability (MSI) and produce high levels of abnormal proteins.⁸ These abnormal proteins serve as tumor antigens that can be targeted by the body's normal immune system. 

In May 2017, the US Food and Drug Administration (FDA) granted accelerated approval of the PD-1 blocking antibody pembrolizumab for the treatment of unresectable or metastatic MSI-high (MSI-H) or MMR-deficient solid tumors.9 The approval was based on data from 149 patients treated in 5 studies that demonstrated a response rate of 39.6%, including responses that lasted at least 6 months in 78% of participants. This was the first ever cancer drug that received FDA approval based on a tumor's biomarker profile without regard to the site of origin. I describe the results of a study by Le and colleagues that examines the possible role of immunotherapy in a variety of solid tumors in this section.

Details of the study

This study examined the clinical efficacy of PD-1 blockade in 86 patients with advanced, MMR-deficient tumors from 12 different sites. Endometrial cancer was the second most frequent primary tumor site in 17% of patients. Within the cohort, the overall objective response rate was 53%, which included 21% of patients with complete radiographic response (no imaging evidence of cancer). Disease control, either complete or partial response or stable disease, was achieved in 77% of patients. After a median follow-up of 12.5 months, neither the median progression-free survival (PFS) nor median overall survival had been reached. The authors estimated that 2-year overall survival was 64%, substantially higher than expected for patients with advanced solid tumors.

Le and colleagues also performed several in vivo laboratory experiments to explore the mechanisms by which patients responded. In addition, they used sequencing to determine the prevalence of MMR deficiency in 12,019 cancer samples that included 32 distinct tumor types (FIGURE). Endometrial cancer had the highest frequency of MMR deficiency (17%). Four percent of cervical cancers and less than 2% of ovarian cancers were MMR-deficient.

The promise of immunotherapy for endometrial cancer. This study's data and other emerging data have important implications for women with gynecologic cancer, particularly endometrial cancer. First, given the frequency of MMR mutations among women with endometrial cancer, MMR testing should be strongly considered for these patients. Many institutions have protocols for reflex testing with immunohistochemistry for women with endometrial cancer. For women with positive test results, germline sequencing can be performed to determine if they have an inherited MMR deficiency, Lynch syndrome. Presence of an MMR deficiency is an important factor in cancer screening and potential treatment.

Second, the impressive results of PD-1 blockade in patients with MMR-deficient tumors suggest that this treatment strategy may be important for women with recurrent or metastatic endometrial cancer. The ideal timing of immunotherapy for women with endometrial cancer is an area of active ongoing study.

Share your thoughts! Send your Letter to the Editor to [email protected]. Please include your name and the city and state in which you practice.

Despite universal agreement that antibiotic overprescribing is a problem, the practice continues to vex us. Antibiotic use—whether appropriate or not—has been linked to rising rates of antimicrobial resistance, disruption of the gut microbiome leading to Clostridium difficile infections (CDI), allergic reactions, and increased health care costs (see Table 1).^1-6 And yet, clinicians continue to overprescribe this class of medication.

A 2016 report from the CDC estimates that at least 30% of antibiotics prescribed in US outpatient settings are unnecessary.⁷ Another report cites a slightly higher figure across a variety of health care settings.⁸ Pair these findings with the fact that there are currently few new drugs in development to target resistant bacteria, and you have the potential for a postantibiotic era in which common infections could become lethal.⁷

In 2003, the CDC launched its “Get Smart: Know When Antibiotics Work” program (now known as “Be Antibiotics Aware”), focused on decreasing inappropriate antibiotic use in the outpatient setting.⁹ In 2015, the White House released the National Action Plan for Combating Antibiotic-Resistant Bacteria, with a goal of decreasing inappropriate outpatient antibiotic use by 50% and inappropriate inpatient use by 20% by 2020.¹⁰ And, on an international level, the World Health Organization (WHO) in 2015 developed a five-year strategic framework for implementing its Global Action Plan on Antimicrobial Resistance.¹¹

Family practitioners are on the front lines of this battle. Here’s what we can do now.

WHEN AND WHERE ARE ANTIBIOTICS MOST OFTEN INAPPROPRIATELY PRESCRIBED?

The diagnosis leading to the most frequent inappropriate prescribing of antibiotics is acute respiratory tract infection (ARTI), which includes bronchitis, otitis media, pharyngitis, sinusitis, tonsillitis, the common cold, and pneumonia. Up to 40% of antibiotic prescriptions for these conditions are unnecessary.^8,12 Bronchitis is the most common ARTI diagnosis associated with inappropriate antibiotic prescriptions, while sinusitis, suppurative otitis media, and pharyngitis are the diagnoses associated with the lion’s share of all (appropriate and inappropriate) antibiotic prescriptions within the ARTI category.^8,9,12,13 Refer to national clinical guidelines, which delineate when antibiotic treatment is appropriate for these conditions.^14-16

With respect to setting, there are conflicting findings as to whether antibiotic prescribing differs in office-based versus emergency department (ED) settings.

• One study found a higher rate of antibiotic prescribing during ED visits than office visits (21% vs 9%), even though, between 2007 and 2009, more antibiotic prescriptions were written for adults in primary care offices than in either outpatient hospital clinics or EDs.¹⁷
• In a cross-sectional study using data from 2005 to 2010 National Ambulatory Medical Care Surveys (NAMCS) and National Hospital Ambulatory Medical Care Surveys (NHAMCS), more than half of patients with uncomplicated acute rhinosinu­sitis received a prescription for antibiotics, but there was no overall difference in antibiotic prescriptions between primary care and ED presentation.¹⁸
• A retrospective analysis found that between 2006 and 2010, outpatient hospital practices (56%) and community-practice offices (60%) prescribed more antibiotics for ARTIs than EDs did (51%).¹²

STICK TO NARROW-SPECTRUM AGENTS WHEN POSSIBLE

Using broad-spectrum antibiotics, such as quinolones or imipenem, firstline, contributes more to the problem of antibiotic resistance than does prescribing narrow-spectrum antibiotics such as amoxicillin, cephalexin, or trimethoprim-sulfamethoxazole.⁷ Yet between 2007 and 2009, broad-spectrum agents were prescribed for 61% of outpatient adult visits in which patients received an antibiotic prescription.¹⁷ Quinolones (25%), macrolides (20%), and aminopenicillins (12%) were most commonly prescribed, and antibiotic prescriptions were most often written for respiratory conditions, such as bronchitis, for which we now know antibiotics are rarely indicated.¹⁷

Between 2006 and 2008, pediatric patients who received antibiotic prescriptions were given broad-spectrum agents 50% of the time, of which macrolides were the class most commonly prescribed.¹³

More recently, researchers examined the frequency with which clinicians prescribe narrow-spectrum, firstline antibiotics for otitis media, sinusitis, and pharyngitis using 2010 to 2011 NAMCS/NHAMCS data. They found that providers used firstline agents recommended by professional guidelines 52% of the time, although it was estimated that they would have been appropriate in 80% of cases; pediatric patients were more likely to receive appropriate firstline antibiotics than adult patients.¹⁹ Macrolides, especially azithromycin, were the most common non-firstline antibiotics prescribed.^19,20 The bottom line is that when antibiotics are indicated for upper respiratory infections (otitis media, sinusitis, and pharyngitis), clinicians should prescribe a narrow-spectrum antibiotic first.

ANTIBIOTIC OVERPRESCIBING AFFECTS THE GUT AND BEYOND

The human intestinal microbiome is composed of a diverse array of bacteria, viruses, and parasites.²¹ The main functions of the gut microbiome include interacting with the immune system and participating in biochemical reactions in the gut, such as absorption of fat-soluble vitamins and the production of vitamin K.

As we know, antibiotics decrease the diversity of gut bacteria, which, in turn, can cause less efficient nutrient extraction, as well as vulnerability to enteric infections.²¹ It is well known, for example, that the bacterial gut microbiome can either inhibit or promote diarrheal illnesses such as those caused by CDI. CDI is now the most common health care-related infection, accounting for about a half-million health care facility infections per year.²² It extends hospital stays an average of almost 10 days and is estimated to cost the health care system $6.3 billion annually.²³

Antibiotics can also eliminate antibiotic-susceptible organisms, allowing resistant organisms to proliferate.⁴ They also promote the transmission of genes for antibiotic resistance between gut bacteria.⁴

Beyond the gut

Less well known is that gut bacteria can promote or inhibit extraintestinal infections.

Gut bacteria and HIV. In early HIV infections, for example, gut populations of Lactobacillus and Bifidobacteria are reduced, and the gut barrier becomes compromised.²⁴ Increasing translocation of bacterial products is associated with HIV disease progression. Preservation of Lactobacillus populations in the gut is associated with markers predictive of better HIV outcomes, including a higher CD4 count, a lower viral load, and less evidence of gut microbial translocation.²⁴ This underscores the importance of maintaining healthy gut flora in patients with HIV, using such steps as avoiding unnecessary antibiotics.

Gut bacteria and stress, depression. Antibiotics directly induce the expression of key genes that affect the stress response.²⁵ While causative studies are lacking, there is a growing body of evidence suggesting that the gut microbiome is involved in two-way communication with the brain and can affect, and be affected by, stress and depression.^21,26-30 Diseases and conditions that seem to have a putative connection to a disordered microbiome (dysbiosis) include depression, anxiety, Crohn disease, type 2 diabetes, and obesity. (For a discussion of the relationship between the gut microbiome and diabetes, see Endocrine Consult: The Gut Microbiome in Type 2 Diabetes.)

Gut bacteria and childhood obesity. Repeated use of broader-spectrum antibiotics in children younger than 24 months of age increases the risk for childhood obesity.^1,6 One theory for the association is that the effects of broad-spectrum antibiotics on the intestinal flora of young children may alter long-term energy homeostasis, resulting in a higher risk for obesity.¹

Gut bacteria and asthma. Studies demonstrate differences in the gut microbiomes of asthmatic and nonasthmatic patients. These differences affect the activities of helper T-cell subsets (Th1 and Th2), which in turn affect the development of immune tolerance.³¹

Although additional studies are needed to confirm these findings, the evidence collected thus far should make us all pause before prescribing drugs that can alter our microbiome in complex and only partially understood ways.

WHAT CAN WE DO RIGHT NOW?

The issues created by the inappropriate prescribing of antibiotics have been known for decades, and multiple attempts have been made to find solutions and implement change. Although some small successes have occurred, little overall progress has been made in reducing antibiotic prescribing in the general population. A historical review of why clinicians prescribe antibiotics inappropriately and the interventions that have successfully reduced this prescribing may prove valuable as we continue to look for new, effective answers.

Why do we overprescribe antibiotics? A 2015 systematic literature review found that patient demand, pharmaceutical company marketing activities, limited up-to-date information sources, and fear of losing patients are major reasons providers cite for prescribing antibiotics.³²

In a separate study that explored antibiotic prescribing habits for acute bronchitis, clinicians cited “patient demand” as the major reason for prescribing antibiotics. Respondents also reported that “other physicians were responsible for inappropriate antibiotic prescribing.”³³

Strategies that work

Some early intervention programs directed at reducing antibiotic prescribing demonstrated success (see Table 2).^34-36

One example comes from a 1996-1998 study of four primary care practices.³⁴ Researchers evaluated the impact of a multidimensional intervention effort targeted at clinicians and patients and aimed at lowering the use of antimicrobial agents for acute uncomplicated bronchitis in adults. It incorporated multiple elements, including office-based and household patient educational materials and a clinician intervention involving education, practice profiling, and academic detailing. Clinicians in this program reduced their rates of antibiotic prescribing for uncomplicated bronchitis from 74% to 48%.³⁴

Employing EMRs. A more recent study focused on use of electronic medical rec­ords (EMRs) and communications to modify clinician antibiotic prescribing.³⁵ By sending clinicians monthly emails comparing their prescribing patterns to those of peers and “typical top performers,” inappropriate antibiotic prescriptions for ARTIs went from 19.9% to 3.7%.³⁵

In another effort, the same researchers modified providers’ EMRs to detect when potentially inappropriate antibiotics were prescribed. The system then prompted the clinician to provide an “antibiotic justification note,” which remained visible in the patient’s chart. This approach, which encouraged providers to follow prescribing guidelines by capitalizing on their concerns about their reputations, produced a 77% reduction in antibiotic prescribing.³⁵

Focusing on the public. Studies have also examined the effectiveness of educating the public about when antibiotics are not likely to be helpful and of the harms of unnecessary antibiotics.

Studies conducted in Tennessee and Wisconsin that combined prescriber and community education about unnecessary antibiotics for children found that the intervention reduced antibiotic prescribing in both locations by about 19%, compared with about a 9% reduction in the control groups.^36,37

DOES PRESCRIBING ANTIBIOTICS AFFECT PATIENT SATISFACTION?

The results are mixed as to whether prescribing antibiotics affects patient satisfaction. Two studies in the early 2000s found that both patients and parents reported higher satisfaction with clinicians who explained why antibiotics were not indicated versus those who simply prescribed them—and that such explanations do not need to take a lot of time (see Table 3 for patient care tips).^37,38

A more recent study found that higher antibiotic prescribing practices in Britain were associated with modestly higher patient satisfaction ratings.³⁹ The authors of this study noted, however, that reduced antibiotic prescribing may be a proxy for other practice patterns that affected satisfaction ratings.

REDUCING ANTIBIOTIC PRESCRIBING REDUCES RESISTANCE

There is also strong evidence that when clinicians decrease antibiotic prescribing, antimicrobial resistance follows suit. One of the earlier landmark studies to demonstrate this was a Finnish study published in 1997.⁴⁰ The authors found that a reduction of macrolide antibiotic consumption in Finland led to a reduction in streptococci macrolide resistance from 16.5% to 8.6%.⁴⁰

Multiple studies have since demonstrated similar results for both respiratory and urinary tract infections.^41,42 A 2017 meta-analysis of 32 studies found that antibiotic stewardship programs reduced the incidence of infections and colonization with multidrug-resistant Gram-negative bacteria (by 51%), extended-spectrum beta-lactamase–producing Gram-negative bacteria (48%), and methicillin-resistant Staphylococcus aureus (37%). There was also a reduction in the incidence of CDI (32%).⁴³

Despite universal agreement that antibiotic overprescribing is a problem, the practice continues to vex us. Antibiotic use—whether appropriate or not—has been linked to rising rates of antimicrobial resistance, disruption of the gut microbiome leading to Clostridium difficile infections (CDI), allergic reactions, and increased health care costs (see Table 1).^1-6 And yet, clinicians continue to overprescribe this class of medication.

A 2016 report from the CDC estimates that at least 30% of antibiotics prescribed in US outpatient settings are unnecessary.⁷ Another report cites a slightly higher figure across a variety of health care settings.⁸ Pair these findings with the fact that there are currently few new drugs in development to target resistant bacteria, and you have the potential for a postantibiotic era in which common infections could become lethal.⁷

In 2003, the CDC launched its “Get Smart: Know When Antibiotics Work” program (now known as “Be Antibiotics Aware”), focused on decreasing inappropriate antibiotic use in the outpatient setting.⁹ In 2015, the White House released the National Action Plan for Combating Antibiotic-Resistant Bacteria, with a goal of decreasing inappropriate outpatient antibiotic use by 50% and inappropriate inpatient use by 20% by 2020.¹⁰ And, on an international level, the World Health Organization (WHO) in 2015 developed a five-year strategic framework for implementing its Global Action Plan on Antimicrobial Resistance.¹¹

Family practitioners are on the front lines of this battle. Here’s what we can do now.

WHEN AND WHERE ARE ANTIBIOTICS MOST OFTEN INAPPROPRIATELY PRESCRIBED?

The diagnosis leading to the most frequent inappropriate prescribing of antibiotics is acute respiratory tract infection (ARTI), which includes bronchitis, otitis media, pharyngitis, sinusitis, tonsillitis, the common cold, and pneumonia. Up to 40% of antibiotic prescriptions for these conditions are unnecessary.^8,12 Bronchitis is the most common ARTI diagnosis associated with inappropriate antibiotic prescriptions, while sinusitis, suppurative otitis media, and pharyngitis are the diagnoses associated with the lion’s share of all (appropriate and inappropriate) antibiotic prescriptions within the ARTI category.^8,9,12,13 Refer to national clinical guidelines, which delineate when antibiotic treatment is appropriate for these conditions.^14-16

With respect to setting, there are conflicting findings as to whether antibiotic prescribing differs in office-based versus emergency department (ED) settings.

• One study found a higher rate of antibiotic prescribing during ED visits than office visits (21% vs 9%), even though, between 2007 and 2009, more antibiotic prescriptions were written for adults in primary care offices than in either outpatient hospital clinics or EDs.¹⁷
• In a cross-sectional study using data from 2005 to 2010 National Ambulatory Medical Care Surveys (NAMCS) and National Hospital Ambulatory Medical Care Surveys (NHAMCS), more than half of patients with uncomplicated acute rhinosinu­sitis received a prescription for antibiotics, but there was no overall difference in antibiotic prescriptions between primary care and ED presentation.¹⁸
• A retrospective analysis found that between 2006 and 2010, outpatient hospital practices (56%) and community-practice offices (60%) prescribed more antibiotics for ARTIs than EDs did (51%).¹²

STICK TO NARROW-SPECTRUM AGENTS WHEN POSSIBLE

Using broad-spectrum antibiotics, such as quinolones or imipenem, firstline, contributes more to the problem of antibiotic resistance than does prescribing narrow-spectrum antibiotics such as amoxicillin, cephalexin, or trimethoprim-sulfamethoxazole.⁷ Yet between 2007 and 2009, broad-spectrum agents were prescribed for 61% of outpatient adult visits in which patients received an antibiotic prescription.¹⁷ Quinolones (25%), macrolides (20%), and aminopenicillins (12%) were most commonly prescribed, and antibiotic prescriptions were most often written for respiratory conditions, such as bronchitis, for which we now know antibiotics are rarely indicated.¹⁷

Between 2006 and 2008, pediatric patients who received antibiotic prescriptions were given broad-spectrum agents 50% of the time, of which macrolides were the class most commonly prescribed.¹³

More recently, researchers examined the frequency with which clinicians prescribe narrow-spectrum, firstline antibiotics for otitis media, sinusitis, and pharyngitis using 2010 to 2011 NAMCS/NHAMCS data. They found that providers used firstline agents recommended by professional guidelines 52% of the time, although it was estimated that they would have been appropriate in 80% of cases; pediatric patients were more likely to receive appropriate firstline antibiotics than adult patients.¹⁹ Macrolides, especially azithromycin, were the most common non-firstline antibiotics prescribed.^19,20 The bottom line is that when antibiotics are indicated for upper respiratory infections (otitis media, sinusitis, and pharyngitis), clinicians should prescribe a narrow-spectrum antibiotic first.

ANTIBIOTIC OVERPRESCIBING AFFECTS THE GUT AND BEYOND

The human intestinal microbiome is composed of a diverse array of bacteria, viruses, and parasites.²¹ The main functions of the gut microbiome include interacting with the immune system and participating in biochemical reactions in the gut, such as absorption of fat-soluble vitamins and the production of vitamin K.

As we know, antibiotics decrease the diversity of gut bacteria, which, in turn, can cause less efficient nutrient extraction, as well as vulnerability to enteric infections.²¹ It is well known, for example, that the bacterial gut microbiome can either inhibit or promote diarrheal illnesses such as those caused by CDI. CDI is now the most common health care-related infection, accounting for about a half-million health care facility infections per year.²² It extends hospital stays an average of almost 10 days and is estimated to cost the health care system $6.3 billion annually.²³

Antibiotics can also eliminate antibiotic-susceptible organisms, allowing resistant organisms to proliferate.⁴ They also promote the transmission of genes for antibiotic resistance between gut bacteria.⁴

Beyond the gut

Less well known is that gut bacteria can promote or inhibit extraintestinal infections.

Gut bacteria and HIV. In early HIV infections, for example, gut populations of Lactobacillus and Bifidobacteria are reduced, and the gut barrier becomes compromised.²⁴ Increasing translocation of bacterial products is associated with HIV disease progression. Preservation of Lactobacillus populations in the gut is associated with markers predictive of better HIV outcomes, including a higher CD4 count, a lower viral load, and less evidence of gut microbial translocation.²⁴ This underscores the importance of maintaining healthy gut flora in patients with HIV, using such steps as avoiding unnecessary antibiotics.

Gut bacteria and stress, depression. Antibiotics directly induce the expression of key genes that affect the stress response.²⁵ While causative studies are lacking, there is a growing body of evidence suggesting that the gut microbiome is involved in two-way communication with the brain and can affect, and be affected by, stress and depression.^21,26-30 Diseases and conditions that seem to have a putative connection to a disordered microbiome (dysbiosis) include depression, anxiety, Crohn disease, type 2 diabetes, and obesity. (For a discussion of the relationship between the gut microbiome and diabetes, see Endocrine Consult: The Gut Microbiome in Type 2 Diabetes.)

Gut bacteria and childhood obesity. Repeated use of broader-spectrum antibiotics in children younger than 24 months of age increases the risk for childhood obesity.^1,6 One theory for the association is that the effects of broad-spectrum antibiotics on the intestinal flora of young children may alter long-term energy homeostasis, resulting in a higher risk for obesity.¹

Gut bacteria and asthma. Studies demonstrate differences in the gut microbiomes of asthmatic and nonasthmatic patients. These differences affect the activities of helper T-cell subsets (Th1 and Th2), which in turn affect the development of immune tolerance.³¹

Although additional studies are needed to confirm these findings, the evidence collected thus far should make us all pause before prescribing drugs that can alter our microbiome in complex and only partially understood ways.

WHAT CAN WE DO RIGHT NOW?

The issues created by the inappropriate prescribing of antibiotics have been known for decades, and multiple attempts have been made to find solutions and implement change. Although some small successes have occurred, little overall progress has been made in reducing antibiotic prescribing in the general population. A historical review of why clinicians prescribe antibiotics inappropriately and the interventions that have successfully reduced this prescribing may prove valuable as we continue to look for new, effective answers.

Why do we overprescribe antibiotics? A 2015 systematic literature review found that patient demand, pharmaceutical company marketing activities, limited up-to-date information sources, and fear of losing patients are major reasons providers cite for prescribing antibiotics.³²

In a separate study that explored antibiotic prescribing habits for acute bronchitis, clinicians cited “patient demand” as the major reason for prescribing antibiotics. Respondents also reported that “other physicians were responsible for inappropriate antibiotic prescribing.”³³

Strategies that work

Some early intervention programs directed at reducing antibiotic prescribing demonstrated success (see Table 2).^34-36

One example comes from a 1996-1998 study of four primary care practices.³⁴ Researchers evaluated the impact of a multidimensional intervention effort targeted at clinicians and patients and aimed at lowering the use of antimicrobial agents for acute uncomplicated bronchitis in adults. It incorporated multiple elements, including office-based and household patient educational materials and a clinician intervention involving education, practice profiling, and academic detailing. Clinicians in this program reduced their rates of antibiotic prescribing for uncomplicated bronchitis from 74% to 48%.³⁴

Employing EMRs. A more recent study focused on use of electronic medical rec­ords (EMRs) and communications to modify clinician antibiotic prescribing.³⁵ By sending clinicians monthly emails comparing their prescribing patterns to those of peers and “typical top performers,” inappropriate antibiotic prescriptions for ARTIs went from 19.9% to 3.7%.³⁵

In another effort, the same researchers modified providers’ EMRs to detect when potentially inappropriate antibiotics were prescribed. The system then prompted the clinician to provide an “antibiotic justification note,” which remained visible in the patient’s chart. This approach, which encouraged providers to follow prescribing guidelines by capitalizing on their concerns about their reputations, produced a 77% reduction in antibiotic prescribing.³⁵

Focusing on the public. Studies have also examined the effectiveness of educating the public about when antibiotics are not likely to be helpful and of the harms of unnecessary antibiotics.

Studies conducted in Tennessee and Wisconsin that combined prescriber and community education about unnecessary antibiotics for children found that the intervention reduced antibiotic prescribing in both locations by about 19%, compared with about a 9% reduction in the control groups.^36,37

DOES PRESCRIBING ANTIBIOTICS AFFECT PATIENT SATISFACTION?

The results are mixed as to whether prescribing antibiotics affects patient satisfaction. Two studies in the early 2000s found that both patients and parents reported higher satisfaction with clinicians who explained why antibiotics were not indicated versus those who simply prescribed them—and that such explanations do not need to take a lot of time (see Table 3 for patient care tips).^37,38

A more recent study found that higher antibiotic prescribing practices in Britain were associated with modestly higher patient satisfaction ratings.³⁹ The authors of this study noted, however, that reduced antibiotic prescribing may be a proxy for other practice patterns that affected satisfaction ratings.

REDUCING ANTIBIOTIC PRESCRIBING REDUCES RESISTANCE

There is also strong evidence that when clinicians decrease antibiotic prescribing, antimicrobial resistance follows suit. One of the earlier landmark studies to demonstrate this was a Finnish study published in 1997.⁴⁰ The authors found that a reduction of macrolide antibiotic consumption in Finland led to a reduction in streptococci macrolide resistance from 16.5% to 8.6%.⁴⁰

Multiple studies have since demonstrated similar results for both respiratory and urinary tract infections.^41,42 A 2017 meta-analysis of 32 studies found that antibiotic stewardship programs reduced the incidence of infections and colonization with multidrug-resistant Gram-negative bacteria (by 51%), extended-spectrum beta-lactamase–producing Gram-negative bacteria (48%), and methicillin-resistant Staphylococcus aureus (37%). There was also a reduction in the incidence of CDI (32%).⁴³

Despite universal agreement that antibiotic overprescribing is a problem, the practice continues to vex us. Antibiotic use—whether appropriate or not—has been linked to rising rates of antimicrobial resistance, disruption of the gut microbiome leading to Clostridium difficile infections (CDI), allergic reactions, and increased health care costs (see Table 1).^1-6 And yet, clinicians continue to overprescribe this class of medication.

A 2016 report from the CDC estimates that at least 30% of antibiotics prescribed in US outpatient settings are unnecessary.⁷ Another report cites a slightly higher figure across a variety of health care settings.⁸ Pair these findings with the fact that there are currently few new drugs in development to target resistant bacteria, and you have the potential for a postantibiotic era in which common infections could become lethal.⁷

In 2003, the CDC launched its “Get Smart: Know When Antibiotics Work” program (now known as “Be Antibiotics Aware”), focused on decreasing inappropriate antibiotic use in the outpatient setting.⁹ In 2015, the White House released the National Action Plan for Combating Antibiotic-Resistant Bacteria, with a goal of decreasing inappropriate outpatient antibiotic use by 50% and inappropriate inpatient use by 20% by 2020.¹⁰ And, on an international level, the World Health Organization (WHO) in 2015 developed a five-year strategic framework for implementing its Global Action Plan on Antimicrobial Resistance.¹¹

Family practitioners are on the front lines of this battle. Here’s what we can do now.

WHEN AND WHERE ARE ANTIBIOTICS MOST OFTEN INAPPROPRIATELY PRESCRIBED?

The diagnosis leading to the most frequent inappropriate prescribing of antibiotics is acute respiratory tract infection (ARTI), which includes bronchitis, otitis media, pharyngitis, sinusitis, tonsillitis, the common cold, and pneumonia. Up to 40% of antibiotic prescriptions for these conditions are unnecessary.^8,12 Bronchitis is the most common ARTI diagnosis associated with inappropriate antibiotic prescriptions, while sinusitis, suppurative otitis media, and pharyngitis are the diagnoses associated with the lion’s share of all (appropriate and inappropriate) antibiotic prescriptions within the ARTI category.^8,9,12,13 Refer to national clinical guidelines, which delineate when antibiotic treatment is appropriate for these conditions.^14-16

With respect to setting, there are conflicting findings as to whether antibiotic prescribing differs in office-based versus emergency department (ED) settings.

• One study found a higher rate of antibiotic prescribing during ED visits than office visits (21% vs 9%), even though, between 2007 and 2009, more antibiotic prescriptions were written for adults in primary care offices than in either outpatient hospital clinics or EDs.¹⁷
• In a cross-sectional study using data from 2005 to 2010 National Ambulatory Medical Care Surveys (NAMCS) and National Hospital Ambulatory Medical Care Surveys (NHAMCS), more than half of patients with uncomplicated acute rhinosinu­sitis received a prescription for antibiotics, but there was no overall difference in antibiotic prescriptions between primary care and ED presentation.¹⁸
• A retrospective analysis found that between 2006 and 2010, outpatient hospital practices (56%) and community-practice offices (60%) prescribed more antibiotics for ARTIs than EDs did (51%).¹²

STICK TO NARROW-SPECTRUM AGENTS WHEN POSSIBLE

Using broad-spectrum antibiotics, such as quinolones or imipenem, firstline, contributes more to the problem of antibiotic resistance than does prescribing narrow-spectrum antibiotics such as amoxicillin, cephalexin, or trimethoprim-sulfamethoxazole.⁷ Yet between 2007 and 2009, broad-spectrum agents were prescribed for 61% of outpatient adult visits in which patients received an antibiotic prescription.¹⁷ Quinolones (25%), macrolides (20%), and aminopenicillins (12%) were most commonly prescribed, and antibiotic prescriptions were most often written for respiratory conditions, such as bronchitis, for which we now know antibiotics are rarely indicated.¹⁷

Between 2006 and 2008, pediatric patients who received antibiotic prescriptions were given broad-spectrum agents 50% of the time, of which macrolides were the class most commonly prescribed.¹³

More recently, researchers examined the frequency with which clinicians prescribe narrow-spectrum, firstline antibiotics for otitis media, sinusitis, and pharyngitis using 2010 to 2011 NAMCS/NHAMCS data. They found that providers used firstline agents recommended by professional guidelines 52% of the time, although it was estimated that they would have been appropriate in 80% of cases; pediatric patients were more likely to receive appropriate firstline antibiotics than adult patients.¹⁹ Macrolides, especially azithromycin, were the most common non-firstline antibiotics prescribed.^19,20 The bottom line is that when antibiotics are indicated for upper respiratory infections (otitis media, sinusitis, and pharyngitis), clinicians should prescribe a narrow-spectrum antibiotic first.

ANTIBIOTIC OVERPRESCIBING AFFECTS THE GUT AND BEYOND

The human intestinal microbiome is composed of a diverse array of bacteria, viruses, and parasites.²¹ The main functions of the gut microbiome include interacting with the immune system and participating in biochemical reactions in the gut, such as absorption of fat-soluble vitamins and the production of vitamin K.

As we know, antibiotics decrease the diversity of gut bacteria, which, in turn, can cause less efficient nutrient extraction, as well as vulnerability to enteric infections.²¹ It is well known, for example, that the bacterial gut microbiome can either inhibit or promote diarrheal illnesses such as those caused by CDI. CDI is now the most common health care-related infection, accounting for about a half-million health care facility infections per year.²² It extends hospital stays an average of almost 10 days and is estimated to cost the health care system $6.3 billion annually.²³

Antibiotics can also eliminate antibiotic-susceptible organisms, allowing resistant organisms to proliferate.⁴ They also promote the transmission of genes for antibiotic resistance between gut bacteria.⁴

Beyond the gut

Less well known is that gut bacteria can promote or inhibit extraintestinal infections.

Gut bacteria and HIV. In early HIV infections, for example, gut populations of Lactobacillus and Bifidobacteria are reduced, and the gut barrier becomes compromised.²⁴ Increasing translocation of bacterial products is associated with HIV disease progression. Preservation of Lactobacillus populations in the gut is associated with markers predictive of better HIV outcomes, including a higher CD4 count, a lower viral load, and less evidence of gut microbial translocation.²⁴ This underscores the importance of maintaining healthy gut flora in patients with HIV, using such steps as avoiding unnecessary antibiotics.

Gut bacteria and stress, depression. Antibiotics directly induce the expression of key genes that affect the stress response.²⁵ While causative studies are lacking, there is a growing body of evidence suggesting that the gut microbiome is involved in two-way communication with the brain and can affect, and be affected by, stress and depression.^21,26-30 Diseases and conditions that seem to have a putative connection to a disordered microbiome (dysbiosis) include depression, anxiety, Crohn disease, type 2 diabetes, and obesity. (For a discussion of the relationship between the gut microbiome and diabetes, see Endocrine Consult: The Gut Microbiome in Type 2 Diabetes.)

Gut bacteria and childhood obesity. Repeated use of broader-spectrum antibiotics in children younger than 24 months of age increases the risk for childhood obesity.^1,6 One theory for the association is that the effects of broad-spectrum antibiotics on the intestinal flora of young children may alter long-term energy homeostasis, resulting in a higher risk for obesity.¹

Gut bacteria and asthma. Studies demonstrate differences in the gut microbiomes of asthmatic and nonasthmatic patients. These differences affect the activities of helper T-cell subsets (Th1 and Th2), which in turn affect the development of immune tolerance.³¹

Although additional studies are needed to confirm these findings, the evidence collected thus far should make us all pause before prescribing drugs that can alter our microbiome in complex and only partially understood ways.

WHAT CAN WE DO RIGHT NOW?

The issues created by the inappropriate prescribing of antibiotics have been known for decades, and multiple attempts have been made to find solutions and implement change. Although some small successes have occurred, little overall progress has been made in reducing antibiotic prescribing in the general population. A historical review of why clinicians prescribe antibiotics inappropriately and the interventions that have successfully reduced this prescribing may prove valuable as we continue to look for new, effective answers.

Why do we overprescribe antibiotics? A 2015 systematic literature review found that patient demand, pharmaceutical company marketing activities, limited up-to-date information sources, and fear of losing patients are major reasons providers cite for prescribing antibiotics.³²

In a separate study that explored antibiotic prescribing habits for acute bronchitis, clinicians cited “patient demand” as the major reason for prescribing antibiotics. Respondents also reported that “other physicians were responsible for inappropriate antibiotic prescribing.”³³

Strategies that work

Some early intervention programs directed at reducing antibiotic prescribing demonstrated success (see Table 2).^34-36

One example comes from a 1996-1998 study of four primary care practices.³⁴ Researchers evaluated the impact of a multidimensional intervention effort targeted at clinicians and patients and aimed at lowering the use of antimicrobial agents for acute uncomplicated bronchitis in adults. It incorporated multiple elements, including office-based and household patient educational materials and a clinician intervention involving education, practice profiling, and academic detailing. Clinicians in this program reduced their rates of antibiotic prescribing for uncomplicated bronchitis from 74% to 48%.³⁴

Employing EMRs. A more recent study focused on use of electronic medical rec­ords (EMRs) and communications to modify clinician antibiotic prescribing.³⁵ By sending clinicians monthly emails comparing their prescribing patterns to those of peers and “typical top performers,” inappropriate antibiotic prescriptions for ARTIs went from 19.9% to 3.7%.³⁵

In another effort, the same researchers modified providers’ EMRs to detect when potentially inappropriate antibiotics were prescribed. The system then prompted the clinician to provide an “antibiotic justification note,” which remained visible in the patient’s chart. This approach, which encouraged providers to follow prescribing guidelines by capitalizing on their concerns about their reputations, produced a 77% reduction in antibiotic prescribing.³⁵

Focusing on the public. Studies have also examined the effectiveness of educating the public about when antibiotics are not likely to be helpful and of the harms of unnecessary antibiotics.

Studies conducted in Tennessee and Wisconsin that combined prescriber and community education about unnecessary antibiotics for children found that the intervention reduced antibiotic prescribing in both locations by about 19%, compared with about a 9% reduction in the control groups.^36,37

DOES PRESCRIBING ANTIBIOTICS AFFECT PATIENT SATISFACTION?

The results are mixed as to whether prescribing antibiotics affects patient satisfaction. Two studies in the early 2000s found that both patients and parents reported higher satisfaction with clinicians who explained why antibiotics were not indicated versus those who simply prescribed them—and that such explanations do not need to take a lot of time (see Table 3 for patient care tips).^37,38

A more recent study found that higher antibiotic prescribing practices in Britain were associated with modestly higher patient satisfaction ratings.³⁹ The authors of this study noted, however, that reduced antibiotic prescribing may be a proxy for other practice patterns that affected satisfaction ratings.

REDUCING ANTIBIOTIC PRESCRIBING REDUCES RESISTANCE

There is also strong evidence that when clinicians decrease antibiotic prescribing, antimicrobial resistance follows suit. One of the earlier landmark studies to demonstrate this was a Finnish study published in 1997.⁴⁰ The authors found that a reduction of macrolide antibiotic consumption in Finland led to a reduction in streptococci macrolide resistance from 16.5% to 8.6%.⁴⁰

Multiple studies have since demonstrated similar results for both respiratory and urinary tract infections.^41,42 A 2017 meta-analysis of 32 studies found that antibiotic stewardship programs reduced the incidence of infections and colonization with multidrug-resistant Gram-negative bacteria (by 51%), extended-spectrum beta-lactamase–producing Gram-negative bacteria (48%), and methicillin-resistant Staphylococcus aureus (37%). There was also a reduction in the incidence of CDI (32%).⁴³

From the University of North Dakota School of Medicine & Health Sciences, Fargo, ND.

Abstract

• Objective: To review the management of community-acquired pneumonia (CAP) in adults.
• Methods: Review of the literature.
• Results: Approximately 4 to 5 million cases of CAP are diagnosed in the United States annually, accounting for significant morbidity and mortality. While numerous studies have previously shown pneumococcus to be the most common causative pathogen, the 2015 EPIC study found that in nearly two-thirds of patients with CAP who required hospitalization, no pathogen was detected. Symptoms and signs of respiratory tract infection are useful in helping to diagnose pneumonia; however, they are less sensitive than chest imaging studies. Laboratory tests used in diagnosing pneumonia include sputum Gram stain and culture, blood culture, urinary antigen, polymerase chain reaction, and biologic markers. In empiric treatment of CAP, both the typical and atypical pathogens should be targeted. Influenza vaccine and pneumococcal polysaccharide and conjugate vaccines should be administered as recommended by the CDC to reduce risk of CAP.
• Conclusion: CAP is a common illness with high rates of morbidity and mortality. Treatment is for the most part empirical; diagnostic testing can be used to identify the causative organism and guide pathogen-specific therapy.

Key words: community-acquired pneumonia; adults; management; vaccines.

Despite advances in medical science, pneumonia remains a major cause of morbidity and mortality. In 2014, 50,620 patients in the United States died from the disease [1]. Pneumonia can be classified as community-acquired, hospital-acquired, or ventilator-associated. Another category, healthcare-associated pneumonia, was included in an earlier American Thoracic Society (ATS) and Infectious Diseases Society of America (IDSA) guideline but was removed from the 2016 guideline because there was no clear evidence that patients diagnosed with healthcare-associated pneumonia were at higher risk for harboring multidrug-resistant pathogens [2]. In this article, we review the epidemiology, microbiology, predisposing factors, diagnosis, treatment, and prevention of community-acquired pneumonia (CAP).

Definition and Epidemiology

CAP is defined as an acute infection of the lungs that develops in patients who have not been hospitalized recently and have not had regular exposure to the health care system [3]. A previously ambulatory patient who is diagnosed with pneumonia within 48 hours after admission also meets the criteria for CAP. Approximately 4 to 5 million cases of CAP are diagnosed in the United States annually [4]. About 25% of CAP patients require hospitalization, and about 5% to 10% of these patients are admitted to the intensive care unit (ICU) [5]. In-hospital mortality is considerable (~10% in population-based studies) [6] and 30-day mortality was found to be as high as 23% in a review by File and Marrie [7]. CAP also confers a high risk of long-term morbidity and mortality compared with the general population who have never had CAP, irrespective of age [8].

Causative Organisms

Numerous microorganisms can cause CAP. Common causes and less common causes are delineated in Table 1. 

Predisposing Factors

Most people diagnosed with CAP have one or more predisposing factors [12,13] (Table 2). 

Clinical Signs and Symptoms

Symptoms of CAP include fever, chills, rigors, fatigue, anorexia, diaphoresis, dyspnea, cough (with or without sputum production), and pleuritic chest pain. There is no individual symptom or cluster of symptoms that can absolutely differentiate pneumonia from other acute respiratory diseases, including upper and lower respiratory infections. However, if a patient presents with the constellation of symptoms of fever ≥ 100⁰F (37.8⁰C), productive cough, and tachycardia, it is more suggestive of pneumonia [14]. Abnormal vital signs include fever, hypothermia, tachypnea, tachycardia, and oxygen desaturation. Auscultation of the chest reveals crackles or other adventitious breath sounds. Elderly patients with pneumonia report a significantly lower number of both respiratory and nonrespiratory symptoms compared with younger patients. Clinicians should be aware of this phenomenon so it does not lead to delayed diagnosis and treatment [15].

Imaging Evaluation

The presence of a pulmonary consolidation or an infiltrate on chest radiograph is required to diagnose CAP, and a chest radiograph should be obtained when CAP is suspected [16]. It should be noted that there is no pattern of radiographic abnormalities reliable enough to differentiate infectious pneumonia from noninfectious causes [17].

There are case reports and case series demonstrating false-negative plain chest radiographs existing in dehydrated patients [18] or in neutropenic state. However, animal studies have shown that dogs challenged with pneumococcus showed abnormal pulmonary shadow, suggestive of pneumonia, regardless of hydration status [19]. There is also no reliable scientific evidence to support the notion that severe neutropenia can cause false-negative radiographs because of the inability to develop an acute inflammatory reaction in the lungs [20].

A chest CT scan is more sensitive than a plain chest radiograph in detecting pneumonia. Therefore, a chest CT should be performed in a patient with negative plain chest radiograph when pneumonia is still highly suspected [21]. A chest CT scan is also more sensitive in detecting cavitation, adenopathy, interstitial disease and empyema. It also has the advantage of better defining anatomical changes than plain films [22].

Because improvement of pulmonary opacities in patients with CAP lags behind clinical improvement, repeating chest imaging studies is not recommended in patients who demonstrate clinical improvement. Sometimes clearing of pulmonary infiltrate or consolidation can take 6 weeks or longer [23].

Laboratory Evaluation

Generally the etiologic agent of CAP cannot be determined solely on the basis of clinical signs and symptoms or imaging studies. Although routine microbiological testing for patients suspicious for CAP is not necessary for empirical treatment, by determining the etiologic agent of the pneumonia, a clinician will be able to narrow the antibiotics from a broad-spectrum empirical regimen to specific pathogen-directed therapy. Determination of certain etiologic agents causing the pneumonia can have important public health implications (eg, Mycobacterium tuberculosis and influenza virus) [24].

Sputum Gram Stain and Culture

Sputum Gram stain is an inexpensive test that may identify pathogens that cause CAP (eg, S. pneumonia and Haemophilus influenzae). A quality specimen is required. A sputum sample must contain > 25 neutrophils and < 10 squamous epithelial cells/low power field on Gram stain to be considered suitable for culture.

The sensitivity and specificity of sputum Gram stain and culture are highly variable in different clinical settings (eg, outpatient setting, nursing home, ICU). Reed et al’s meta-analysis of patients diagnosed with CAP in the United States showed the sensitivity and specificity of sputum Gram stain (compared with sputum culture) ranged from 15% to 100% and 11% to 100%, respectively [24]. In cases of proven bacteremic pneumococcal pneumonia, positive cultures from sputum samples were positive less than 50% of the time [25].

For patients who cannot provide sputum samples or are intubated, a deep-suction aspirate or bronchoalveolar lavage through a bronchoscopic procedure might be necessary to obtain pulmonary secretion for Gram stain and culture. Besides bacterial culture, sputum samples can also be sent for fungal and mycobacterial cultures and acid-fast stain if deemed clinically necessary.

Blood Culture

Because the positivity rate of blood culture in patients who are suspected to have pneumonia but not exposed to antimicrobial agents is disappointingly low (5%–14%), blood cultures are no longer recommended in patients hospitalized for CAP. Another reason for not recommending blood culture is positive culture rarely leads to changes in antibiotic regimen in patients without underlying diseases [26]. However, high-risk patients, including patients with severe CAP or in immunocompromised patients (eg, patients with neutropenia, asplenia or complement deficiencies) should have a blood culture done [24].

A multinational study published in 2008 examined 125 patients with pneumococcal bacteremic CAP versus 1847 patients with non-bacteremic CAP [27]. Analysis of the data demonstrated no association of pneumococcal bacteremic CAP and time to clinical stability, length of hospital stay, all-cause mortality or CAP-related mortality. The authors concluded that pneumococcal bacteremia does not increase the risk of poor outcomes in patients with CAP compared to non-bacteremic patients, and the presence of pneumococcal bacteremia should not deter de-escalation of therapy in clinically stable patients.

Urinary Antigen Tests

Urinary antigen tests may assist clinicians in narrowing antibiotic therapy when test results are positive. There are 2 U.S. Food and Drug Administration–approved tests available to clinicians for detecting pneumococcal and Legionella antigen in urine. The test for Legionella pneumophila detects disease due to serogroup 1 only, which accounts for 80% of community-acquired Legionnaires disease. The sensitivity and specificity of the Legionella urine antigen test are 90% and 99%, respectively. The pneumococcal urine antigen test is less sensitive and specific than the Legionella urine antigen test (sensitivity 80% and specificity > 90%) [28,29].

Advantages of the urinary antigen tests are that they are easily performed, results are available in less than an hour if done in-house, and results are not affected by prior exposure to antibiotics. However, the tests do not meet Clinical Laboratory Improvements Amendments criteria for waiver and must be performed by a technician in the laboratory.

Polymerase Chain Reaction

There are several FDA-approved polymerase chain reaction (PCR) tests commercially available to assist clinicians in diagnosing pneumonia. PCR test of nasopharyngeal swabs for diagnosing influenza have become standard in many medical U.S. facilities. The great advantage of using PCR to diagnose influenza is its high sensitivity and specificity and rapid turnaround time. PCR can also be used to detect Legionella species, S. pneumonia, Mycoplasma pneumoniae, Chlamydophila pneumonia and mycobacterial species [24].

One limitation of using PCR tests on respiratory specimens is that specimens can be contaminated with oral or upper airway flora, so the results must be interpreted with caution, bearing in mind that some of the pathogens isolated may be colonizers of the oral or upper airway flora [30].

Biologic Markers

Two biologic markers—procalcitonin and C-reactive protein (CRP)—can be used in conjunction with history, physical examination, laboratory tests and imaging studies to assist in the diagnosis and treatment of CAP [24]. Procalcitonin is a peptide precursor of the hormone calcitonin that is released by parenchymal cells into the bloodstream resulting in increased serum level in patients with bacterial infections. In contrast, there is no remarkable proclacitonin level increase with viral or noninfectious inflammation. The reference value of procalcitonin in the blood of an adult individual without infection or inflammation is < 0.15 ng/mL. In the blood, procalcitonin has a half-life of 25 to 30 hours. The quantitative immunoluminometric method (LUMI test, Brahms PCT, Berlin, Germany ) is the preferred test to use because of its high sensitivity [31].

A 2012 Cochrane meta-analysis that involved 4221 patients with acute respiratory infections (with half of the patients diagnosed with CAP) from 14 prospective trials found the use of procalcitonin test for antibiotic use significantly decreased median antibiotic exposure from 8 to 4 days without an increase in treatment failure, mortality rates in any clinical setting (eg, outpatient clinic, emergency room), or length of hospitalization [32]. A prospective study conducted in France on 100 ICU patients showed that increased procalcitonin from day 1 to day 3 has a poor prognosis factor for severe CAP whereas decreasing procalcitonin levels is associated with a favorable outcome [33].

CRP is an acute phase protein produced by the liver. CRP level in the blood increases in response to acute infection or inflammation. Use of CRP in assisting diagnosis and guiding treatment of CAP is more limited in part due to its poor specificity. A prospective study conducted on 168 consecutive patients presented with cough showed that a CRP > 40 mg/L had a sensitivity and specificity of 70% and 90%, respectively [34].

Treatment

Site of Care Decision

For patients with CAP, the clinician must decide whether the patient will be treated in an outpatient or inpatient setting, and for those in the inpatient setting, whether they can safely be treated on the general medical ward or should be the ICU. Two common scoring systems that can be used to aid the clinician in determining severity of the infection and guiding site-of-care decisions are the Pneumonia Severity Index (PSI) and CURB-65 scores.

The PSI score uses 20 different parameters, including comorbidities, laboratory parameters and radiographic findings to stratify patients into 5 mortality risk classes [35]. On the basis of associated mortality rates, it has been suggested that risk class I and II patients should be treated as outpatients, risk class III patients should be treated in an observation unit or with a short hospitalization, and risk class IV and V patients should be treated as inpatients [35].

The CURB-65 method of risk stratification is based on 5 clinical parameters: confusion, urea level, respiratory rate, systolic blood pressure and age ≥ 65 (Table 3) [36].

Patients with CURB-65 scores of 4 or 5 are considered to have severe pneumonia and admission to the ICU should be considered. Aside from the CURB-65 score, anyone requiring vasopressor support or mechanical ventilation merits admission to the ICU [16]. IDSA/ATS guidelines also recommend the use of “minor criteria” for making ICU admission decisions; these include respiratory rate ≥ 30 breaths / minute, PaO2 fraction ≤ 250, multilobar infiltrates, confusion, blood urea nitrogen ≥ 20 mg/dL, leukopenia, thrombocytopenia, hypothermia and hypotension [16]. These factors are associated with increased mortality due to CAP and admission to an ICU is indicated if 3 of the minor criteria for severe CAP are present.

Similar to CURB-65, another clinical calculator that can be used for assessing severity of CAP is SMART-COP [39]. This scoring system uses 8 weighted criteria to predict which patients will require intensive respiratory or vasopressor support. SMART-COP has a sensitivity of 79% and specificity 64% in predicting ICU admission, whereas CURB-65 had a pooled sensitivity of 57.2% and specificity of 77.2% [40].

Antibiotic Therapy

Antibiotics are the mainstay of treatment for CAP, with the majority of patients with CAP treated empirically taking into account the site of care, likely pathogen, and antimicrobial resistance issues. Patients with pneumonia who are treated as outpatients usually respond well to empiric antibiotic treatment and a causative pathogen is not usually sought. Patients who are hospitalized for treatment of CAP usually receive empiric antibiotic on admission. Once the etiology has been determined by microbiologic or serologic means, antimicrobial therapy should be adjusted accordingly. As noted previously, a CDC study found that the burden of viral etiologies was higher than previously thought, with rhinovirus and influenza accounting for 15% of cases and S. pneumoniae for only 5% [9]. This study highlighted the fact that despite advances in molecular techniques, most patients with pneumonia have no pathogen identified [9]. Given the lack of discernable pathogens in the majority of cases, unless a nonbacterial etiology is found patients should continue to be treated with antibiotics.

Outpatients without comorbidities or risk factors for drug-resistant S. pneumoniae (Table 4)

As previously mentioned, antibiotic therapy is typically empiric; neither clinical features nor radiographic features are sufficient to include or exclude infectious etiologies. Epidemiologic risk factors should be considered and, in certain cases, expanded antimicrobial coverage to include those entities; for example, treatment of anaerobes in the setting of lung abscess and antipseudomonal antibiotics for patients with bronchiectasis.

Of concern in the treatment of CAP is the increased prevalence of antimicrobial resistance among S. pneumoniae. The IDSA guidelines report that drug-resistant S. pneumoniae is more common in persons aged < 2 or > 65 years, and those with ß-lactam therapy within the previous 3 months, alcoholism, medical comorbidities, immunosuppressive illness or therapy, or exposure to a child who attends a day care center [16].

S. aureus should be considered during influenza outbreaks, with either vancomycin or linezolid being the recommended agents in the setting of methicillin-resistant S. aureus (MRSA). In a study comparing vancomycin versus linezolid for nosocomial pneumonia, the all-cause 60-day mortality was similar for both agents [41]. Datpomycin is another agent used against MRSA; however, its use in the setting of pneumonia is not indicated as daptomycin binds to surfactant, yielding it ineffective in the treatment of pneumonia [42]. Ceftaroline is a newer cephalosporin with activity against MRSA; its role in treatment of community-acquired MRSA pneumonia has not been fully elucidated, but it appears to be a useful agent for this indication [43,44].

Antibiotic Therapy for Selected Pathogens

S. pneumoniae

Patients with pneumococcal pneumonia who have penicillin-susceptible strains can be treated with intravenous penicillin (2 or 3 million units every 4 hours) or ceftriaxone. Once a patient meets criteria of stability, they can then be transitioned to oral penicillin, amoxicillin, or clarithromycin. Those with strains with reduced susceptibility can still be treated with penicillin but at a higher dose (4 million units IV every 4 hours) or a third-generation cephalosporin. Those whose pneumococcal pneumonia is complicated by bacteremia will benefit from dual therapy if severely ill, requiring ICU monitoring. Those not severely ill can be treated with monotherapy [46].

S. aureus

S. aureus is more commonly associated with hospital-acquired pneumonia but may also be seen during the influenza season and in those with severe necrotizing CAP. Both linezolid and vancomycin can be used to treat MRSA CAP. As noted above, ceftaroline has activity against MRSA and is approved for treatment of CAP, but is not approved by the FDA for MRSA CAP treatment. Similarly, tigecycline is approved for CAP and has activity against MRSA, but is not approved for MRSA CAP. Moreover, the FDA has warned of increased risk of death with tigecycline and has a black box warning to that effect [47].

Legionella

Treatment of legionellosis can be achieved with tetra­cyclines, macrolides, or fluoroquinolones. For nonimmunosuppressed patients with mild pneumonia, any of the listed antibiotics is considered appropriate. However, patients with severe infection or those with immunosuppression should be treated with either levofloxacin or azithromycin for 7 to 10 days [48].

C. pneumoniae

As with other atypical organisms, C. pneumoniae can be treated with doxycycline, a macrolide, or respiratory fluoroquinolones. However, length of therapy varies by regimen used; whereas treating with doxycycline 100 mg twice daily generally requires 14–21 days, moxifloxacin 400 mg daily only requires 10 days [49].

M. pneumoniae

As with C. pneumoniae, length of therapy of M. pneumoniae varies by antimicrobial used. Shortest courses are seen with the use of macrolides for 5 days, whereas 14 days is considered standard for doxycycline or a respiratory fluoroquinolone [50]. It should be noted that there has been increasingly documented resistance to macrolides, with known resistance of 8.2% in the United States [51].

Duration of Treatment

Most patients with CAP respond within 72 hours to appropriate therapy. IDSA/ATS guidelines recommend that patients be treated for a minimum of 5 days, and before discontinuing antibiotics patients should be afebrile a minimum of 48-72 hours and be clinically stable (Table 6) [16]. 

Hospitalized patients do not need to be monitored for an additional day once they have reached clinical stability (Table 6), are able to maintain oral intake, and have normal mentation, provided that other comorbidities are stable and social needs have been met [16]. Patients discharged from the hospital with instability have higher risk of readmission or death [55].

Transition to Oral Therapy

IDSA/ATS guidelines [16] recommend that patients should be transitioned from IV to oral antibiotics when they are improving clinically, have stable vital signs, and are able to ingest food/fluids and medications.

Management of Nonresponders

Although the majority of patients respond to antibiotics within 72 hours, treatment failure occurs in up to 15% of patients [45]. Nonresponding pneumonia is generally seen in 2 patterns: worsening of clinical status despite empiric antibiotics OR delay in achieving clinical stability as defined in Table 5 after 72 hours of treatment [13]. Risk factors associated with nonresponding pneumonia [56] are:

• Radiographic: multilobar infiltrates, pleural effusion, cavitation
• Bacteriologic: MRSA, gram-negative or Legionella pneumonia
• Severity index: PSI > 90
• Pharmacologic: incorrect antibiotic choice based on susceptibility

Patients with acute deterioration of clinical status will prompt transfer to a higher level of care and may require mechanical ventilator support. In those with delay in achieving clinical stability, question centers on whether the same antibiotics can be continued while doing further radiographic/microbiologic workup and/or changing antibiotics.

History should be reviewed with particular attention to exposures, travel history, and microbiologic and radiographic data. Clinicians should recall that viral causes account for up to 20% of pneumonias and there are also noninfectious causes that can mimic pyogenic infections [57]. If adequate initial cultures were not obtained, they should be obtained; however, care must be taken in reviewing new sets of cultures while on antibiotics as they may reveal colonization selected out by antibiotics and not a true pathogen. If repeat evaluation is unrevealing, then further evaluation with CT scan and bronchoscopy with bronchoalveolar lavage and biopsy is warranted. CT scans can show pleural effusions, bronchial obstructions or pattern suggestive of cryptogenic pneumonia. A bronchoscopy might yield a microbiologic diagnosis and with biopsy can also evaluate for noninfectious causes.

As with other infections, if escalation of antibiotics is undertaken, clinicians should be mindful to ensure that efforts are being made to elucidate the reason for nonresponse. To simply broaden antimicrobial therapy without attempts at establishing a microbiologic or radiographic cause for nonresponse may lead to inappropriate treatment recurrence of infection. Aside from patients who have bacteremic pneumococcal pneumonia in an ICU setting, there are no published reports pointing to superiority of combination antibiotics [46].

Other Treatment

Because of the inflammatory response associated with pneumonia, several agents have been evaluated as adjunctive treatment of pneumonia to decrease this inflammatory state; namely, steroids, macrolide antibiotics and statins. To date, only the use of steroids (methylprednisolone 0.5 mg/kg every 12 hours for 5 days) in those with severe CAP and high initial anti-inflammatory response (CRP > 150) was shown to decrease treatment failure, decreased risk of ARDS, possibly reduce length of stay, duration of intravenous antibiotics and clinical stability, without effect on mortality or adverse side effects [58,59].

Other adjunctive methods have not been found to have significant impact [16].

Prevention of Pneumonia

Prevention of pneumococcal pneumonia is twofold: prevention of infection caused by S. pneumoniae and prevention of influenza infection. As influenza infection is a risk factor for bacterial infection, specifically with S. pneumoniae, influenza vaccination can prevent bacterial pneumonia [60]. In their most recent recommendations, the CDC continues to recommend routine influenza vaccination for all persons aged greater than 6 months, unless otherwise contraindicated [61].

There are 2 vaccines for prevention of pneumococcal disease: the pneumococcal polysaccharide vaccine (PPSV23) and a conjugate vaccine (PCV13). Following vaccination with PPSV23, 80% of adults develop antibodies against at least 18 of the 23 serotypes [62]. Despite this response, PPSV23 is reported to be protective against invasive pneumococcal infection; yet there is no consensus regarding PPSV23 leading to decreased rates of pneumonia [63]. On the other hand, PCV13 vaccination was associated with prevention of both invasive disease and community-acquired pneumonia in adults 65 years or older [64]. The CDC recommends that all children aged 2 or under receive PCV13, whereas those aged 65 or older should receive PCV13 followed by a dose of PPSV23 [65]. The dose of PPSV23 should be given ≥1 year following the dose of PCV13 [66].Persons < 65 years of age with immunocompromising and certain other conditions should also receive vaccination [67] (Table 7). Full details, many scenarios, and timing of vaccinations can be found at www.cdc.gov/vaccines/schedules/downloads/adult/adult-schedule.pdf.

Cigarette smoking increases the risk of respiratory infections as evidenced by smokers accounting for almost half of all patients with invasive pneumococcal disease [11]. As this is a modifiable risk factor it should be a goal of a comprehensive approach towards prevention of pneumonia.

Summary

CAP remains a leading cause of hospitalization and death in the 21st century. Traditionally, pneumococcus has been considered the major pathogen causing CAP; however, the 2015 EPIC study found that in only 5% of patients diagnosed with CAP was S. pneumoniae detected. Despite the new findings, it is still recommended that empiric treatment for CAP target common typical bacteria (pneumococcus, H. influenzae, Moraxella catarrhalis) and atypical bacteria (M. pneumonia, C. pneumoniae, L. pneumophila).

Because diagnosing pneumonia through history and clinical examination is less than 50% sensitive, a chest imaging study (a plain chest radiograph or a chest CT scan) is usually required to make the diagnosis. Laboratory tests, such as sputum Gram stain/culture, blood culture, urinary antigen tests, PCR test, procalcitonin, and CRP are important adjunctive diagnostic modalities to assist in the diagnosis and management of CAP. However, no single test is sensitive and specific enough to be a stand-alone test. They should be used in conjunction with history, physical examination, and imaging studies. Because vaccination (PPSV23, PCV13, and influenza vaccine) remains the most effective tool in preventing the development of CAP, clinicians, should strive for 100% vaccination rates in appropriate persons.

Corresponding author: Tze Shein Lo, MD, University of North Dakota, 1919 Elm Street, Fargo, ND 58102, [email protected].

Financial disclosures: None.

Author contributions: drafting of article, PM, TSL; critical revision of the article, PM, TSL.

From the University of North Dakota School of Medicine & Health Sciences, Fargo, ND.

Abstract

• Objective: To review the management of community-acquired pneumonia (CAP) in adults.
• Methods: Review of the literature.
• Results: Approximately 4 to 5 million cases of CAP are diagnosed in the United States annually, accounting for significant morbidity and mortality. While numerous studies have previously shown pneumococcus to be the most common causative pathogen, the 2015 EPIC study found that in nearly two-thirds of patients with CAP who required hospitalization, no pathogen was detected. Symptoms and signs of respiratory tract infection are useful in helping to diagnose pneumonia; however, they are less sensitive than chest imaging studies. Laboratory tests used in diagnosing pneumonia include sputum Gram stain and culture, blood culture, urinary antigen, polymerase chain reaction, and biologic markers. In empiric treatment of CAP, both the typical and atypical pathogens should be targeted. Influenza vaccine and pneumococcal polysaccharide and conjugate vaccines should be administered as recommended by the CDC to reduce risk of CAP.
• Conclusion: CAP is a common illness with high rates of morbidity and mortality. Treatment is for the most part empirical; diagnostic testing can be used to identify the causative organism and guide pathogen-specific therapy.

Key words: community-acquired pneumonia; adults; management; vaccines.

Despite advances in medical science, pneumonia remains a major cause of morbidity and mortality. In 2014, 50,620 patients in the United States died from the disease [1]. Pneumonia can be classified as community-acquired, hospital-acquired, or ventilator-associated. Another category, healthcare-associated pneumonia, was included in an earlier American Thoracic Society (ATS) and Infectious Diseases Society of America (IDSA) guideline but was removed from the 2016 guideline because there was no clear evidence that patients diagnosed with healthcare-associated pneumonia were at higher risk for harboring multidrug-resistant pathogens [2]. In this article, we review the epidemiology, microbiology, predisposing factors, diagnosis, treatment, and prevention of community-acquired pneumonia (CAP).

Definition and Epidemiology

CAP is defined as an acute infection of the lungs that develops in patients who have not been hospitalized recently and have not had regular exposure to the health care system [3]. A previously ambulatory patient who is diagnosed with pneumonia within 48 hours after admission also meets the criteria for CAP. Approximately 4 to 5 million cases of CAP are diagnosed in the United States annually [4]. About 25% of CAP patients require hospitalization, and about 5% to 10% of these patients are admitted to the intensive care unit (ICU) [5]. In-hospital mortality is considerable (~10% in population-based studies) [6] and 30-day mortality was found to be as high as 23% in a review by File and Marrie [7]. CAP also confers a high risk of long-term morbidity and mortality compared with the general population who have never had CAP, irrespective of age [8].

Causative Organisms

Numerous microorganisms can cause CAP. Common causes and less common causes are delineated in Table 1. 

Predisposing Factors

Most people diagnosed with CAP have one or more predisposing factors [12,13] (Table 2). 

Clinical Signs and Symptoms

Symptoms of CAP include fever, chills, rigors, fatigue, anorexia, diaphoresis, dyspnea, cough (with or without sputum production), and pleuritic chest pain. There is no individual symptom or cluster of symptoms that can absolutely differentiate pneumonia from other acute respiratory diseases, including upper and lower respiratory infections. However, if a patient presents with the constellation of symptoms of fever ≥ 100⁰F (37.8⁰C), productive cough, and tachycardia, it is more suggestive of pneumonia [14]. Abnormal vital signs include fever, hypothermia, tachypnea, tachycardia, and oxygen desaturation. Auscultation of the chest reveals crackles or other adventitious breath sounds. Elderly patients with pneumonia report a significantly lower number of both respiratory and nonrespiratory symptoms compared with younger patients. Clinicians should be aware of this phenomenon so it does not lead to delayed diagnosis and treatment [15].

Imaging Evaluation

The presence of a pulmonary consolidation or an infiltrate on chest radiograph is required to diagnose CAP, and a chest radiograph should be obtained when CAP is suspected [16]. It should be noted that there is no pattern of radiographic abnormalities reliable enough to differentiate infectious pneumonia from noninfectious causes [17].

There are case reports and case series demonstrating false-negative plain chest radiographs existing in dehydrated patients [18] or in neutropenic state. However, animal studies have shown that dogs challenged with pneumococcus showed abnormal pulmonary shadow, suggestive of pneumonia, regardless of hydration status [19]. There is also no reliable scientific evidence to support the notion that severe neutropenia can cause false-negative radiographs because of the inability to develop an acute inflammatory reaction in the lungs [20].

A chest CT scan is more sensitive than a plain chest radiograph in detecting pneumonia. Therefore, a chest CT should be performed in a patient with negative plain chest radiograph when pneumonia is still highly suspected [21]. A chest CT scan is also more sensitive in detecting cavitation, adenopathy, interstitial disease and empyema. It also has the advantage of better defining anatomical changes than plain films [22].

Because improvement of pulmonary opacities in patients with CAP lags behind clinical improvement, repeating chest imaging studies is not recommended in patients who demonstrate clinical improvement. Sometimes clearing of pulmonary infiltrate or consolidation can take 6 weeks or longer [23].

Laboratory Evaluation

Generally the etiologic agent of CAP cannot be determined solely on the basis of clinical signs and symptoms or imaging studies. Although routine microbiological testing for patients suspicious for CAP is not necessary for empirical treatment, by determining the etiologic agent of the pneumonia, a clinician will be able to narrow the antibiotics from a broad-spectrum empirical regimen to specific pathogen-directed therapy. Determination of certain etiologic agents causing the pneumonia can have important public health implications (eg, Mycobacterium tuberculosis and influenza virus) [24].

Sputum Gram Stain and Culture

Sputum Gram stain is an inexpensive test that may identify pathogens that cause CAP (eg, S. pneumonia and Haemophilus influenzae). A quality specimen is required. A sputum sample must contain > 25 neutrophils and < 10 squamous epithelial cells/low power field on Gram stain to be considered suitable for culture.

The sensitivity and specificity of sputum Gram stain and culture are highly variable in different clinical settings (eg, outpatient setting, nursing home, ICU). Reed et al’s meta-analysis of patients diagnosed with CAP in the United States showed the sensitivity and specificity of sputum Gram stain (compared with sputum culture) ranged from 15% to 100% and 11% to 100%, respectively [24]. In cases of proven bacteremic pneumococcal pneumonia, positive cultures from sputum samples were positive less than 50% of the time [25].

For patients who cannot provide sputum samples or are intubated, a deep-suction aspirate or bronchoalveolar lavage through a bronchoscopic procedure might be necessary to obtain pulmonary secretion for Gram stain and culture. Besides bacterial culture, sputum samples can also be sent for fungal and mycobacterial cultures and acid-fast stain if deemed clinically necessary.

Blood Culture

Because the positivity rate of blood culture in patients who are suspected to have pneumonia but not exposed to antimicrobial agents is disappointingly low (5%–14%), blood cultures are no longer recommended in patients hospitalized for CAP. Another reason for not recommending blood culture is positive culture rarely leads to changes in antibiotic regimen in patients without underlying diseases [26]. However, high-risk patients, including patients with severe CAP or in immunocompromised patients (eg, patients with neutropenia, asplenia or complement deficiencies) should have a blood culture done [24].

A multinational study published in 2008 examined 125 patients with pneumococcal bacteremic CAP versus 1847 patients with non-bacteremic CAP [27]. Analysis of the data demonstrated no association of pneumococcal bacteremic CAP and time to clinical stability, length of hospital stay, all-cause mortality or CAP-related mortality. The authors concluded that pneumococcal bacteremia does not increase the risk of poor outcomes in patients with CAP compared to non-bacteremic patients, and the presence of pneumococcal bacteremia should not deter de-escalation of therapy in clinically stable patients.

Urinary Antigen Tests

Urinary antigen tests may assist clinicians in narrowing antibiotic therapy when test results are positive. There are 2 U.S. Food and Drug Administration–approved tests available to clinicians for detecting pneumococcal and Legionella antigen in urine. The test for Legionella pneumophila detects disease due to serogroup 1 only, which accounts for 80% of community-acquired Legionnaires disease. The sensitivity and specificity of the Legionella urine antigen test are 90% and 99%, respectively. The pneumococcal urine antigen test is less sensitive and specific than the Legionella urine antigen test (sensitivity 80% and specificity > 90%) [28,29].

Advantages of the urinary antigen tests are that they are easily performed, results are available in less than an hour if done in-house, and results are not affected by prior exposure to antibiotics. However, the tests do not meet Clinical Laboratory Improvements Amendments criteria for waiver and must be performed by a technician in the laboratory.

Polymerase Chain Reaction

There are several FDA-approved polymerase chain reaction (PCR) tests commercially available to assist clinicians in diagnosing pneumonia. PCR test of nasopharyngeal swabs for diagnosing influenza have become standard in many medical U.S. facilities. The great advantage of using PCR to diagnose influenza is its high sensitivity and specificity and rapid turnaround time. PCR can also be used to detect Legionella species, S. pneumonia, Mycoplasma pneumoniae, Chlamydophila pneumonia and mycobacterial species [24].

One limitation of using PCR tests on respiratory specimens is that specimens can be contaminated with oral or upper airway flora, so the results must be interpreted with caution, bearing in mind that some of the pathogens isolated may be colonizers of the oral or upper airway flora [30].

Biologic Markers

Two biologic markers—procalcitonin and C-reactive protein (CRP)—can be used in conjunction with history, physical examination, laboratory tests and imaging studies to assist in the diagnosis and treatment of CAP [24]. Procalcitonin is a peptide precursor of the hormone calcitonin that is released by parenchymal cells into the bloodstream resulting in increased serum level in patients with bacterial infections. In contrast, there is no remarkable proclacitonin level increase with viral or noninfectious inflammation. The reference value of procalcitonin in the blood of an adult individual without infection or inflammation is < 0.15 ng/mL. In the blood, procalcitonin has a half-life of 25 to 30 hours. The quantitative immunoluminometric method (LUMI test, Brahms PCT, Berlin, Germany ) is the preferred test to use because of its high sensitivity [31].

A 2012 Cochrane meta-analysis that involved 4221 patients with acute respiratory infections (with half of the patients diagnosed with CAP) from 14 prospective trials found the use of procalcitonin test for antibiotic use significantly decreased median antibiotic exposure from 8 to 4 days without an increase in treatment failure, mortality rates in any clinical setting (eg, outpatient clinic, emergency room), or length of hospitalization [32]. A prospective study conducted in France on 100 ICU patients showed that increased procalcitonin from day 1 to day 3 has a poor prognosis factor for severe CAP whereas decreasing procalcitonin levels is associated with a favorable outcome [33].

CRP is an acute phase protein produced by the liver. CRP level in the blood increases in response to acute infection or inflammation. Use of CRP in assisting diagnosis and guiding treatment of CAP is more limited in part due to its poor specificity. A prospective study conducted on 168 consecutive patients presented with cough showed that a CRP > 40 mg/L had a sensitivity and specificity of 70% and 90%, respectively [34].

Treatment

Site of Care Decision

For patients with CAP, the clinician must decide whether the patient will be treated in an outpatient or inpatient setting, and for those in the inpatient setting, whether they can safely be treated on the general medical ward or should be the ICU. Two common scoring systems that can be used to aid the clinician in determining severity of the infection and guiding site-of-care decisions are the Pneumonia Severity Index (PSI) and CURB-65 scores.

The PSI score uses 20 different parameters, including comorbidities, laboratory parameters and radiographic findings to stratify patients into 5 mortality risk classes [35]. On the basis of associated mortality rates, it has been suggested that risk class I and II patients should be treated as outpatients, risk class III patients should be treated in an observation unit or with a short hospitalization, and risk class IV and V patients should be treated as inpatients [35].

The CURB-65 method of risk stratification is based on 5 clinical parameters: confusion, urea level, respiratory rate, systolic blood pressure and age ≥ 65 (Table 3) [36].

Patients with CURB-65 scores of 4 or 5 are considered to have severe pneumonia and admission to the ICU should be considered. Aside from the CURB-65 score, anyone requiring vasopressor support or mechanical ventilation merits admission to the ICU [16]. IDSA/ATS guidelines also recommend the use of “minor criteria” for making ICU admission decisions; these include respiratory rate ≥ 30 breaths / minute, PaO2 fraction ≤ 250, multilobar infiltrates, confusion, blood urea nitrogen ≥ 20 mg/dL, leukopenia, thrombocytopenia, hypothermia and hypotension [16]. These factors are associated with increased mortality due to CAP and admission to an ICU is indicated if 3 of the minor criteria for severe CAP are present.

Similar to CURB-65, another clinical calculator that can be used for assessing severity of CAP is SMART-COP [39]. This scoring system uses 8 weighted criteria to predict which patients will require intensive respiratory or vasopressor support. SMART-COP has a sensitivity of 79% and specificity 64% in predicting ICU admission, whereas CURB-65 had a pooled sensitivity of 57.2% and specificity of 77.2% [40].

Antibiotic Therapy

Antibiotics are the mainstay of treatment for CAP, with the majority of patients with CAP treated empirically taking into account the site of care, likely pathogen, and antimicrobial resistance issues. Patients with pneumonia who are treated as outpatients usually respond well to empiric antibiotic treatment and a causative pathogen is not usually sought. Patients who are hospitalized for treatment of CAP usually receive empiric antibiotic on admission. Once the etiology has been determined by microbiologic or serologic means, antimicrobial therapy should be adjusted accordingly. As noted previously, a CDC study found that the burden of viral etiologies was higher than previously thought, with rhinovirus and influenza accounting for 15% of cases and S. pneumoniae for only 5% [9]. This study highlighted the fact that despite advances in molecular techniques, most patients with pneumonia have no pathogen identified [9]. Given the lack of discernable pathogens in the majority of cases, unless a nonbacterial etiology is found patients should continue to be treated with antibiotics.

Outpatients without comorbidities or risk factors for drug-resistant S. pneumoniae (Table 4)

As previously mentioned, antibiotic therapy is typically empiric; neither clinical features nor radiographic features are sufficient to include or exclude infectious etiologies. Epidemiologic risk factors should be considered and, in certain cases, expanded antimicrobial coverage to include those entities; for example, treatment of anaerobes in the setting of lung abscess and antipseudomonal antibiotics for patients with bronchiectasis.

Of concern in the treatment of CAP is the increased prevalence of antimicrobial resistance among S. pneumoniae. The IDSA guidelines report that drug-resistant S. pneumoniae is more common in persons aged < 2 or > 65 years, and those with ß-lactam therapy within the previous 3 months, alcoholism, medical comorbidities, immunosuppressive illness or therapy, or exposure to a child who attends a day care center [16].

S. aureus should be considered during influenza outbreaks, with either vancomycin or linezolid being the recommended agents in the setting of methicillin-resistant S. aureus (MRSA). In a study comparing vancomycin versus linezolid for nosocomial pneumonia, the all-cause 60-day mortality was similar for both agents [41]. Datpomycin is another agent used against MRSA; however, its use in the setting of pneumonia is not indicated as daptomycin binds to surfactant, yielding it ineffective in the treatment of pneumonia [42]. Ceftaroline is a newer cephalosporin with activity against MRSA; its role in treatment of community-acquired MRSA pneumonia has not been fully elucidated, but it appears to be a useful agent for this indication [43,44].

Antibiotic Therapy for Selected Pathogens

S. pneumoniae

Patients with pneumococcal pneumonia who have penicillin-susceptible strains can be treated with intravenous penicillin (2 or 3 million units every 4 hours) or ceftriaxone. Once a patient meets criteria of stability, they can then be transitioned to oral penicillin, amoxicillin, or clarithromycin. Those with strains with reduced susceptibility can still be treated with penicillin but at a higher dose (4 million units IV every 4 hours) or a third-generation cephalosporin. Those whose pneumococcal pneumonia is complicated by bacteremia will benefit from dual therapy if severely ill, requiring ICU monitoring. Those not severely ill can be treated with monotherapy [46].

S. aureus

S. aureus is more commonly associated with hospital-acquired pneumonia but may also be seen during the influenza season and in those with severe necrotizing CAP. Both linezolid and vancomycin can be used to treat MRSA CAP. As noted above, ceftaroline has activity against MRSA and is approved for treatment of CAP, but is not approved by the FDA for MRSA CAP treatment. Similarly, tigecycline is approved for CAP and has activity against MRSA, but is not approved for MRSA CAP. Moreover, the FDA has warned of increased risk of death with tigecycline and has a black box warning to that effect [47].

Legionella

Treatment of legionellosis can be achieved with tetra­cyclines, macrolides, or fluoroquinolones. For nonimmunosuppressed patients with mild pneumonia, any of the listed antibiotics is considered appropriate. However, patients with severe infection or those with immunosuppression should be treated with either levofloxacin or azithromycin for 7 to 10 days [48].

C. pneumoniae

As with other atypical organisms, C. pneumoniae can be treated with doxycycline, a macrolide, or respiratory fluoroquinolones. However, length of therapy varies by regimen used; whereas treating with doxycycline 100 mg twice daily generally requires 14–21 days, moxifloxacin 400 mg daily only requires 10 days [49].

M. pneumoniae

As with C. pneumoniae, length of therapy of M. pneumoniae varies by antimicrobial used. Shortest courses are seen with the use of macrolides for 5 days, whereas 14 days is considered standard for doxycycline or a respiratory fluoroquinolone [50]. It should be noted that there has been increasingly documented resistance to macrolides, with known resistance of 8.2% in the United States [51].

Duration of Treatment

Most patients with CAP respond within 72 hours to appropriate therapy. IDSA/ATS guidelines recommend that patients be treated for a minimum of 5 days, and before discontinuing antibiotics patients should be afebrile a minimum of 48-72 hours and be clinically stable (Table 6) [16]. 

Hospitalized patients do not need to be monitored for an additional day once they have reached clinical stability (Table 6), are able to maintain oral intake, and have normal mentation, provided that other comorbidities are stable and social needs have been met [16]. Patients discharged from the hospital with instability have higher risk of readmission or death [55].

Transition to Oral Therapy

IDSA/ATS guidelines [16] recommend that patients should be transitioned from IV to oral antibiotics when they are improving clinically, have stable vital signs, and are able to ingest food/fluids and medications.

Management of Nonresponders

Although the majority of patients respond to antibiotics within 72 hours, treatment failure occurs in up to 15% of patients [45]. Nonresponding pneumonia is generally seen in 2 patterns: worsening of clinical status despite empiric antibiotics OR delay in achieving clinical stability as defined in Table 5 after 72 hours of treatment [13]. Risk factors associated with nonresponding pneumonia [56] are:

• Radiographic: multilobar infiltrates, pleural effusion, cavitation
• Bacteriologic: MRSA, gram-negative or Legionella pneumonia
• Severity index: PSI > 90
• Pharmacologic: incorrect antibiotic choice based on susceptibility

Patients with acute deterioration of clinical status will prompt transfer to a higher level of care and may require mechanical ventilator support. In those with delay in achieving clinical stability, question centers on whether the same antibiotics can be continued while doing further radiographic/microbiologic workup and/or changing antibiotics.

History should be reviewed with particular attention to exposures, travel history, and microbiologic and radiographic data. Clinicians should recall that viral causes account for up to 20% of pneumonias and there are also noninfectious causes that can mimic pyogenic infections [57]. If adequate initial cultures were not obtained, they should be obtained; however, care must be taken in reviewing new sets of cultures while on antibiotics as they may reveal colonization selected out by antibiotics and not a true pathogen. If repeat evaluation is unrevealing, then further evaluation with CT scan and bronchoscopy with bronchoalveolar lavage and biopsy is warranted. CT scans can show pleural effusions, bronchial obstructions or pattern suggestive of cryptogenic pneumonia. A bronchoscopy might yield a microbiologic diagnosis and with biopsy can also evaluate for noninfectious causes.

As with other infections, if escalation of antibiotics is undertaken, clinicians should be mindful to ensure that efforts are being made to elucidate the reason for nonresponse. To simply broaden antimicrobial therapy without attempts at establishing a microbiologic or radiographic cause for nonresponse may lead to inappropriate treatment recurrence of infection. Aside from patients who have bacteremic pneumococcal pneumonia in an ICU setting, there are no published reports pointing to superiority of combination antibiotics [46].

Other Treatment

Because of the inflammatory response associated with pneumonia, several agents have been evaluated as adjunctive treatment of pneumonia to decrease this inflammatory state; namely, steroids, macrolide antibiotics and statins. To date, only the use of steroids (methylprednisolone 0.5 mg/kg every 12 hours for 5 days) in those with severe CAP and high initial anti-inflammatory response (CRP > 150) was shown to decrease treatment failure, decreased risk of ARDS, possibly reduce length of stay, duration of intravenous antibiotics and clinical stability, without effect on mortality or adverse side effects [58,59].

Other adjunctive methods have not been found to have significant impact [16].

Prevention of Pneumonia

Prevention of pneumococcal pneumonia is twofold: prevention of infection caused by S. pneumoniae and prevention of influenza infection. As influenza infection is a risk factor for bacterial infection, specifically with S. pneumoniae, influenza vaccination can prevent bacterial pneumonia [60]. In their most recent recommendations, the CDC continues to recommend routine influenza vaccination for all persons aged greater than 6 months, unless otherwise contraindicated [61].

There are 2 vaccines for prevention of pneumococcal disease: the pneumococcal polysaccharide vaccine (PPSV23) and a conjugate vaccine (PCV13). Following vaccination with PPSV23, 80% of adults develop antibodies against at least 18 of the 23 serotypes [62]. Despite this response, PPSV23 is reported to be protective against invasive pneumococcal infection; yet there is no consensus regarding PPSV23 leading to decreased rates of pneumonia [63]. On the other hand, PCV13 vaccination was associated with prevention of both invasive disease and community-acquired pneumonia in adults 65 years or older [64]. The CDC recommends that all children aged 2 or under receive PCV13, whereas those aged 65 or older should receive PCV13 followed by a dose of PPSV23 [65]. The dose of PPSV23 should be given ≥1 year following the dose of PCV13 [66].Persons < 65 years of age with immunocompromising and certain other conditions should also receive vaccination [67] (Table 7). Full details, many scenarios, and timing of vaccinations can be found at www.cdc.gov/vaccines/schedules/downloads/adult/adult-schedule.pdf.

Cigarette smoking increases the risk of respiratory infections as evidenced by smokers accounting for almost half of all patients with invasive pneumococcal disease [11]. As this is a modifiable risk factor it should be a goal of a comprehensive approach towards prevention of pneumonia.

Summary

CAP remains a leading cause of hospitalization and death in the 21st century. Traditionally, pneumococcus has been considered the major pathogen causing CAP; however, the 2015 EPIC study found that in only 5% of patients diagnosed with CAP was S. pneumoniae detected. Despite the new findings, it is still recommended that empiric treatment for CAP target common typical bacteria (pneumococcus, H. influenzae, Moraxella catarrhalis) and atypical bacteria (M. pneumonia, C. pneumoniae, L. pneumophila).

Because diagnosing pneumonia through history and clinical examination is less than 50% sensitive, a chest imaging study (a plain chest radiograph or a chest CT scan) is usually required to make the diagnosis. Laboratory tests, such as sputum Gram stain/culture, blood culture, urinary antigen tests, PCR test, procalcitonin, and CRP are important adjunctive diagnostic modalities to assist in the diagnosis and management of CAP. However, no single test is sensitive and specific enough to be a stand-alone test. They should be used in conjunction with history, physical examination, and imaging studies. Because vaccination (PPSV23, PCV13, and influenza vaccine) remains the most effective tool in preventing the development of CAP, clinicians, should strive for 100% vaccination rates in appropriate persons.

Corresponding author: Tze Shein Lo, MD, University of North Dakota, 1919 Elm Street, Fargo, ND 58102, [email protected].

Financial disclosures: None.

Author contributions: drafting of article, PM, TSL; critical revision of the article, PM, TSL.

From the University of North Dakota School of Medicine & Health Sciences, Fargo, ND.

Abstract

• Objective: To review the management of community-acquired pneumonia (CAP) in adults.
• Methods: Review of the literature.
• Results: Approximately 4 to 5 million cases of CAP are diagnosed in the United States annually, accounting for significant morbidity and mortality. While numerous studies have previously shown pneumococcus to be the most common causative pathogen, the 2015 EPIC study found that in nearly two-thirds of patients with CAP who required hospitalization, no pathogen was detected. Symptoms and signs of respiratory tract infection are useful in helping to diagnose pneumonia; however, they are less sensitive than chest imaging studies. Laboratory tests used in diagnosing pneumonia include sputum Gram stain and culture, blood culture, urinary antigen, polymerase chain reaction, and biologic markers. In empiric treatment of CAP, both the typical and atypical pathogens should be targeted. Influenza vaccine and pneumococcal polysaccharide and conjugate vaccines should be administered as recommended by the CDC to reduce risk of CAP.
• Conclusion: CAP is a common illness with high rates of morbidity and mortality. Treatment is for the most part empirical; diagnostic testing can be used to identify the causative organism and guide pathogen-specific therapy.

Key words: community-acquired pneumonia; adults; management; vaccines.

Despite advances in medical science, pneumonia remains a major cause of morbidity and mortality. In 2014, 50,620 patients in the United States died from the disease [1]. Pneumonia can be classified as community-acquired, hospital-acquired, or ventilator-associated. Another category, healthcare-associated pneumonia, was included in an earlier American Thoracic Society (ATS) and Infectious Diseases Society of America (IDSA) guideline but was removed from the 2016 guideline because there was no clear evidence that patients diagnosed with healthcare-associated pneumonia were at higher risk for harboring multidrug-resistant pathogens [2]. In this article, we review the epidemiology, microbiology, predisposing factors, diagnosis, treatment, and prevention of community-acquired pneumonia (CAP).

Definition and Epidemiology

CAP is defined as an acute infection of the lungs that develops in patients who have not been hospitalized recently and have not had regular exposure to the health care system [3]. A previously ambulatory patient who is diagnosed with pneumonia within 48 hours after admission also meets the criteria for CAP. Approximately 4 to 5 million cases of CAP are diagnosed in the United States annually [4]. About 25% of CAP patients require hospitalization, and about 5% to 10% of these patients are admitted to the intensive care unit (ICU) [5]. In-hospital mortality is considerable (~10% in population-based studies) [6] and 30-day mortality was found to be as high as 23% in a review by File and Marrie [7]. CAP also confers a high risk of long-term morbidity and mortality compared with the general population who have never had CAP, irrespective of age [8].

Causative Organisms

Numerous microorganisms can cause CAP. Common causes and less common causes are delineated in Table 1. 

Predisposing Factors

Most people diagnosed with CAP have one or more predisposing factors [12,13] (Table 2). 

Clinical Signs and Symptoms

Symptoms of CAP include fever, chills, rigors, fatigue, anorexia, diaphoresis, dyspnea, cough (with or without sputum production), and pleuritic chest pain. There is no individual symptom or cluster of symptoms that can absolutely differentiate pneumonia from other acute respiratory diseases, including upper and lower respiratory infections. However, if a patient presents with the constellation of symptoms of fever ≥ 100⁰F (37.8⁰C), productive cough, and tachycardia, it is more suggestive of pneumonia [14]. Abnormal vital signs include fever, hypothermia, tachypnea, tachycardia, and oxygen desaturation. Auscultation of the chest reveals crackles or other adventitious breath sounds. Elderly patients with pneumonia report a significantly lower number of both respiratory and nonrespiratory symptoms compared with younger patients. Clinicians should be aware of this phenomenon so it does not lead to delayed diagnosis and treatment [15].

Imaging Evaluation

The presence of a pulmonary consolidation or an infiltrate on chest radiograph is required to diagnose CAP, and a chest radiograph should be obtained when CAP is suspected [16]. It should be noted that there is no pattern of radiographic abnormalities reliable enough to differentiate infectious pneumonia from noninfectious causes [17].

There are case reports and case series demonstrating false-negative plain chest radiographs existing in dehydrated patients [18] or in neutropenic state. However, animal studies have shown that dogs challenged with pneumococcus showed abnormal pulmonary shadow, suggestive of pneumonia, regardless of hydration status [19]. There is also no reliable scientific evidence to support the notion that severe neutropenia can cause false-negative radiographs because of the inability to develop an acute inflammatory reaction in the lungs [20].

A chest CT scan is more sensitive than a plain chest radiograph in detecting pneumonia. Therefore, a chest CT should be performed in a patient with negative plain chest radiograph when pneumonia is still highly suspected [21]. A chest CT scan is also more sensitive in detecting cavitation, adenopathy, interstitial disease and empyema. It also has the advantage of better defining anatomical changes than plain films [22].

Because improvement of pulmonary opacities in patients with CAP lags behind clinical improvement, repeating chest imaging studies is not recommended in patients who demonstrate clinical improvement. Sometimes clearing of pulmonary infiltrate or consolidation can take 6 weeks or longer [23].

Laboratory Evaluation

Generally the etiologic agent of CAP cannot be determined solely on the basis of clinical signs and symptoms or imaging studies. Although routine microbiological testing for patients suspicious for CAP is not necessary for empirical treatment, by determining the etiologic agent of the pneumonia, a clinician will be able to narrow the antibiotics from a broad-spectrum empirical regimen to specific pathogen-directed therapy. Determination of certain etiologic agents causing the pneumonia can have important public health implications (eg, Mycobacterium tuberculosis and influenza virus) [24].

Sputum Gram Stain and Culture

Sputum Gram stain is an inexpensive test that may identify pathogens that cause CAP (eg, S. pneumonia and Haemophilus influenzae). A quality specimen is required. A sputum sample must contain > 25 neutrophils and < 10 squamous epithelial cells/low power field on Gram stain to be considered suitable for culture.

The sensitivity and specificity of sputum Gram stain and culture are highly variable in different clinical settings (eg, outpatient setting, nursing home, ICU). Reed et al’s meta-analysis of patients diagnosed with CAP in the United States showed the sensitivity and specificity of sputum Gram stain (compared with sputum culture) ranged from 15% to 100% and 11% to 100%, respectively [24]. In cases of proven bacteremic pneumococcal pneumonia, positive cultures from sputum samples were positive less than 50% of the time [25].

For patients who cannot provide sputum samples or are intubated, a deep-suction aspirate or bronchoalveolar lavage through a bronchoscopic procedure might be necessary to obtain pulmonary secretion for Gram stain and culture. Besides bacterial culture, sputum samples can also be sent for fungal and mycobacterial cultures and acid-fast stain if deemed clinically necessary.

Blood Culture

Because the positivity rate of blood culture in patients who are suspected to have pneumonia but not exposed to antimicrobial agents is disappointingly low (5%–14%), blood cultures are no longer recommended in patients hospitalized for CAP. Another reason for not recommending blood culture is positive culture rarely leads to changes in antibiotic regimen in patients without underlying diseases [26]. However, high-risk patients, including patients with severe CAP or in immunocompromised patients (eg, patients with neutropenia, asplenia or complement deficiencies) should have a blood culture done [24].

A multinational study published in 2008 examined 125 patients with pneumococcal bacteremic CAP versus 1847 patients with non-bacteremic CAP [27]. Analysis of the data demonstrated no association of pneumococcal bacteremic CAP and time to clinical stability, length of hospital stay, all-cause mortality or CAP-related mortality. The authors concluded that pneumococcal bacteremia does not increase the risk of poor outcomes in patients with CAP compared to non-bacteremic patients, and the presence of pneumococcal bacteremia should not deter de-escalation of therapy in clinically stable patients.

Urinary Antigen Tests

Urinary antigen tests may assist clinicians in narrowing antibiotic therapy when test results are positive. There are 2 U.S. Food and Drug Administration–approved tests available to clinicians for detecting pneumococcal and Legionella antigen in urine. The test for Legionella pneumophila detects disease due to serogroup 1 only, which accounts for 80% of community-acquired Legionnaires disease. The sensitivity and specificity of the Legionella urine antigen test are 90% and 99%, respectively. The pneumococcal urine antigen test is less sensitive and specific than the Legionella urine antigen test (sensitivity 80% and specificity > 90%) [28,29].

Advantages of the urinary antigen tests are that they are easily performed, results are available in less than an hour if done in-house, and results are not affected by prior exposure to antibiotics. However, the tests do not meet Clinical Laboratory Improvements Amendments criteria for waiver and must be performed by a technician in the laboratory.

Polymerase Chain Reaction

There are several FDA-approved polymerase chain reaction (PCR) tests commercially available to assist clinicians in diagnosing pneumonia. PCR test of nasopharyngeal swabs for diagnosing influenza have become standard in many medical U.S. facilities. The great advantage of using PCR to diagnose influenza is its high sensitivity and specificity and rapid turnaround time. PCR can also be used to detect Legionella species, S. pneumonia, Mycoplasma pneumoniae, Chlamydophila pneumonia and mycobacterial species [24].

One limitation of using PCR tests on respiratory specimens is that specimens can be contaminated with oral or upper airway flora, so the results must be interpreted with caution, bearing in mind that some of the pathogens isolated may be colonizers of the oral or upper airway flora [30].

Biologic Markers

Two biologic markers—procalcitonin and C-reactive protein (CRP)—can be used in conjunction with history, physical examination, laboratory tests and imaging studies to assist in the diagnosis and treatment of CAP [24]. Procalcitonin is a peptide precursor of the hormone calcitonin that is released by parenchymal cells into the bloodstream resulting in increased serum level in patients with bacterial infections. In contrast, there is no remarkable proclacitonin level increase with viral or noninfectious inflammation. The reference value of procalcitonin in the blood of an adult individual without infection or inflammation is < 0.15 ng/mL. In the blood, procalcitonin has a half-life of 25 to 30 hours. The quantitative immunoluminometric method (LUMI test, Brahms PCT, Berlin, Germany ) is the preferred test to use because of its high sensitivity [31].

A 2012 Cochrane meta-analysis that involved 4221 patients with acute respiratory infections (with half of the patients diagnosed with CAP) from 14 prospective trials found the use of procalcitonin test for antibiotic use significantly decreased median antibiotic exposure from 8 to 4 days without an increase in treatment failure, mortality rates in any clinical setting (eg, outpatient clinic, emergency room), or length of hospitalization [32]. A prospective study conducted in France on 100 ICU patients showed that increased procalcitonin from day 1 to day 3 has a poor prognosis factor for severe CAP whereas decreasing procalcitonin levels is associated with a favorable outcome [33].

CRP is an acute phase protein produced by the liver. CRP level in the blood increases in response to acute infection or inflammation. Use of CRP in assisting diagnosis and guiding treatment of CAP is more limited in part due to its poor specificity. A prospective study conducted on 168 consecutive patients presented with cough showed that a CRP > 40 mg/L had a sensitivity and specificity of 70% and 90%, respectively [34].

Treatment

Site of Care Decision

For patients with CAP, the clinician must decide whether the patient will be treated in an outpatient or inpatient setting, and for those in the inpatient setting, whether they can safely be treated on the general medical ward or should be the ICU. Two common scoring systems that can be used to aid the clinician in determining severity of the infection and guiding site-of-care decisions are the Pneumonia Severity Index (PSI) and CURB-65 scores.

The PSI score uses 20 different parameters, including comorbidities, laboratory parameters and radiographic findings to stratify patients into 5 mortality risk classes [35]. On the basis of associated mortality rates, it has been suggested that risk class I and II patients should be treated as outpatients, risk class III patients should be treated in an observation unit or with a short hospitalization, and risk class IV and V patients should be treated as inpatients [35].

The CURB-65 method of risk stratification is based on 5 clinical parameters: confusion, urea level, respiratory rate, systolic blood pressure and age ≥ 65 (Table 3) [36].

Patients with CURB-65 scores of 4 or 5 are considered to have severe pneumonia and admission to the ICU should be considered. Aside from the CURB-65 score, anyone requiring vasopressor support or mechanical ventilation merits admission to the ICU [16]. IDSA/ATS guidelines also recommend the use of “minor criteria” for making ICU admission decisions; these include respiratory rate ≥ 30 breaths / minute, PaO2 fraction ≤ 250, multilobar infiltrates, confusion, blood urea nitrogen ≥ 20 mg/dL, leukopenia, thrombocytopenia, hypothermia and hypotension [16]. These factors are associated with increased mortality due to CAP and admission to an ICU is indicated if 3 of the minor criteria for severe CAP are present.

Similar to CURB-65, another clinical calculator that can be used for assessing severity of CAP is SMART-COP [39]. This scoring system uses 8 weighted criteria to predict which patients will require intensive respiratory or vasopressor support. SMART-COP has a sensitivity of 79% and specificity 64% in predicting ICU admission, whereas CURB-65 had a pooled sensitivity of 57.2% and specificity of 77.2% [40].

Antibiotic Therapy

Antibiotics are the mainstay of treatment for CAP, with the majority of patients with CAP treated empirically taking into account the site of care, likely pathogen, and antimicrobial resistance issues. Patients with pneumonia who are treated as outpatients usually respond well to empiric antibiotic treatment and a causative pathogen is not usually sought. Patients who are hospitalized for treatment of CAP usually receive empiric antibiotic on admission. Once the etiology has been determined by microbiologic or serologic means, antimicrobial therapy should be adjusted accordingly. As noted previously, a CDC study found that the burden of viral etiologies was higher than previously thought, with rhinovirus and influenza accounting for 15% of cases and S. pneumoniae for only 5% [9]. This study highlighted the fact that despite advances in molecular techniques, most patients with pneumonia have no pathogen identified [9]. Given the lack of discernable pathogens in the majority of cases, unless a nonbacterial etiology is found patients should continue to be treated with antibiotics.

Outpatients without comorbidities or risk factors for drug-resistant S. pneumoniae (Table 4)

As previously mentioned, antibiotic therapy is typically empiric; neither clinical features nor radiographic features are sufficient to include or exclude infectious etiologies. Epidemiologic risk factors should be considered and, in certain cases, expanded antimicrobial coverage to include those entities; for example, treatment of anaerobes in the setting of lung abscess and antipseudomonal antibiotics for patients with bronchiectasis.

Of concern in the treatment of CAP is the increased prevalence of antimicrobial resistance among S. pneumoniae. The IDSA guidelines report that drug-resistant S. pneumoniae is more common in persons aged < 2 or > 65 years, and those with ß-lactam therapy within the previous 3 months, alcoholism, medical comorbidities, immunosuppressive illness or therapy, or exposure to a child who attends a day care center [16].

S. aureus should be considered during influenza outbreaks, with either vancomycin or linezolid being the recommended agents in the setting of methicillin-resistant S. aureus (MRSA). In a study comparing vancomycin versus linezolid for nosocomial pneumonia, the all-cause 60-day mortality was similar for both agents [41]. Datpomycin is another agent used against MRSA; however, its use in the setting of pneumonia is not indicated as daptomycin binds to surfactant, yielding it ineffective in the treatment of pneumonia [42]. Ceftaroline is a newer cephalosporin with activity against MRSA; its role in treatment of community-acquired MRSA pneumonia has not been fully elucidated, but it appears to be a useful agent for this indication [43,44].

Antibiotic Therapy for Selected Pathogens

S. pneumoniae

Patients with pneumococcal pneumonia who have penicillin-susceptible strains can be treated with intravenous penicillin (2 or 3 million units every 4 hours) or ceftriaxone. Once a patient meets criteria of stability, they can then be transitioned to oral penicillin, amoxicillin, or clarithromycin. Those with strains with reduced susceptibility can still be treated with penicillin but at a higher dose (4 million units IV every 4 hours) or a third-generation cephalosporin. Those whose pneumococcal pneumonia is complicated by bacteremia will benefit from dual therapy if severely ill, requiring ICU monitoring. Those not severely ill can be treated with monotherapy [46].

S. aureus

S. aureus is more commonly associated with hospital-acquired pneumonia but may also be seen during the influenza season and in those with severe necrotizing CAP. Both linezolid and vancomycin can be used to treat MRSA CAP. As noted above, ceftaroline has activity against MRSA and is approved for treatment of CAP, but is not approved by the FDA for MRSA CAP treatment. Similarly, tigecycline is approved for CAP and has activity against MRSA, but is not approved for MRSA CAP. Moreover, the FDA has warned of increased risk of death with tigecycline and has a black box warning to that effect [47].

Legionella

Treatment of legionellosis can be achieved with tetra­cyclines, macrolides, or fluoroquinolones. For nonimmunosuppressed patients with mild pneumonia, any of the listed antibiotics is considered appropriate. However, patients with severe infection or those with immunosuppression should be treated with either levofloxacin or azithromycin for 7 to 10 days [48].

C. pneumoniae

As with other atypical organisms, C. pneumoniae can be treated with doxycycline, a macrolide, or respiratory fluoroquinolones. However, length of therapy varies by regimen used; whereas treating with doxycycline 100 mg twice daily generally requires 14–21 days, moxifloxacin 400 mg daily only requires 10 days [49].

M. pneumoniae

As with C. pneumoniae, length of therapy of M. pneumoniae varies by antimicrobial used. Shortest courses are seen with the use of macrolides for 5 days, whereas 14 days is considered standard for doxycycline or a respiratory fluoroquinolone [50]. It should be noted that there has been increasingly documented resistance to macrolides, with known resistance of 8.2% in the United States [51].

Duration of Treatment

Most patients with CAP respond within 72 hours to appropriate therapy. IDSA/ATS guidelines recommend that patients be treated for a minimum of 5 days, and before discontinuing antibiotics patients should be afebrile a minimum of 48-72 hours and be clinically stable (Table 6) [16]. 

Hospitalized patients do not need to be monitored for an additional day once they have reached clinical stability (Table 6), are able to maintain oral intake, and have normal mentation, provided that other comorbidities are stable and social needs have been met [16]. Patients discharged from the hospital with instability have higher risk of readmission or death [55].

Transition to Oral Therapy

IDSA/ATS guidelines [16] recommend that patients should be transitioned from IV to oral antibiotics when they are improving clinically, have stable vital signs, and are able to ingest food/fluids and medications.

Management of Nonresponders

Although the majority of patients respond to antibiotics within 72 hours, treatment failure occurs in up to 15% of patients [45]. Nonresponding pneumonia is generally seen in 2 patterns: worsening of clinical status despite empiric antibiotics OR delay in achieving clinical stability as defined in Table 5 after 72 hours of treatment [13]. Risk factors associated with nonresponding pneumonia [56] are:

• Radiographic: multilobar infiltrates, pleural effusion, cavitation
• Bacteriologic: MRSA, gram-negative or Legionella pneumonia
• Severity index: PSI > 90
• Pharmacologic: incorrect antibiotic choice based on susceptibility

Patients with acute deterioration of clinical status will prompt transfer to a higher level of care and may require mechanical ventilator support. In those with delay in achieving clinical stability, question centers on whether the same antibiotics can be continued while doing further radiographic/microbiologic workup and/or changing antibiotics.

History should be reviewed with particular attention to exposures, travel history, and microbiologic and radiographic data. Clinicians should recall that viral causes account for up to 20% of pneumonias and there are also noninfectious causes that can mimic pyogenic infections [57]. If adequate initial cultures were not obtained, they should be obtained; however, care must be taken in reviewing new sets of cultures while on antibiotics as they may reveal colonization selected out by antibiotics and not a true pathogen. If repeat evaluation is unrevealing, then further evaluation with CT scan and bronchoscopy with bronchoalveolar lavage and biopsy is warranted. CT scans can show pleural effusions, bronchial obstructions or pattern suggestive of cryptogenic pneumonia. A bronchoscopy might yield a microbiologic diagnosis and with biopsy can also evaluate for noninfectious causes.

As with other infections, if escalation of antibiotics is undertaken, clinicians should be mindful to ensure that efforts are being made to elucidate the reason for nonresponse. To simply broaden antimicrobial therapy without attempts at establishing a microbiologic or radiographic cause for nonresponse may lead to inappropriate treatment recurrence of infection. Aside from patients who have bacteremic pneumococcal pneumonia in an ICU setting, there are no published reports pointing to superiority of combination antibiotics [46].

Other Treatment

Because of the inflammatory response associated with pneumonia, several agents have been evaluated as adjunctive treatment of pneumonia to decrease this inflammatory state; namely, steroids, macrolide antibiotics and statins. To date, only the use of steroids (methylprednisolone 0.5 mg/kg every 12 hours for 5 days) in those with severe CAP and high initial anti-inflammatory response (CRP > 150) was shown to decrease treatment failure, decreased risk of ARDS, possibly reduce length of stay, duration of intravenous antibiotics and clinical stability, without effect on mortality or adverse side effects [58,59].

Other adjunctive methods have not been found to have significant impact [16].

Prevention of Pneumonia

Prevention of pneumococcal pneumonia is twofold: prevention of infection caused by S. pneumoniae and prevention of influenza infection. As influenza infection is a risk factor for bacterial infection, specifically with S. pneumoniae, influenza vaccination can prevent bacterial pneumonia [60]. In their most recent recommendations, the CDC continues to recommend routine influenza vaccination for all persons aged greater than 6 months, unless otherwise contraindicated [61].

There are 2 vaccines for prevention of pneumococcal disease: the pneumococcal polysaccharide vaccine (PPSV23) and a conjugate vaccine (PCV13). Following vaccination with PPSV23, 80% of adults develop antibodies against at least 18 of the 23 serotypes [62]. Despite this response, PPSV23 is reported to be protective against invasive pneumococcal infection; yet there is no consensus regarding PPSV23 leading to decreased rates of pneumonia [63]. On the other hand, PCV13 vaccination was associated with prevention of both invasive disease and community-acquired pneumonia in adults 65 years or older [64]. The CDC recommends that all children aged 2 or under receive PCV13, whereas those aged 65 or older should receive PCV13 followed by a dose of PPSV23 [65]. The dose of PPSV23 should be given ≥1 year following the dose of PCV13 [66].Persons < 65 years of age with immunocompromising and certain other conditions should also receive vaccination [67] (Table 7). Full details, many scenarios, and timing of vaccinations can be found at www.cdc.gov/vaccines/schedules/downloads/adult/adult-schedule.pdf.

Cigarette smoking increases the risk of respiratory infections as evidenced by smokers accounting for almost half of all patients with invasive pneumococcal disease [11]. As this is a modifiable risk factor it should be a goal of a comprehensive approach towards prevention of pneumonia.

Summary

CAP remains a leading cause of hospitalization and death in the 21st century. Traditionally, pneumococcus has been considered the major pathogen causing CAP; however, the 2015 EPIC study found that in only 5% of patients diagnosed with CAP was S. pneumoniae detected. Despite the new findings, it is still recommended that empiric treatment for CAP target common typical bacteria (pneumococcus, H. influenzae, Moraxella catarrhalis) and atypical bacteria (M. pneumonia, C. pneumoniae, L. pneumophila).

Because diagnosing pneumonia through history and clinical examination is less than 50% sensitive, a chest imaging study (a plain chest radiograph or a chest CT scan) is usually required to make the diagnosis. Laboratory tests, such as sputum Gram stain/culture, blood culture, urinary antigen tests, PCR test, procalcitonin, and CRP are important adjunctive diagnostic modalities to assist in the diagnosis and management of CAP. However, no single test is sensitive and specific enough to be a stand-alone test. They should be used in conjunction with history, physical examination, and imaging studies. Because vaccination (PPSV23, PCV13, and influenza vaccine) remains the most effective tool in preventing the development of CAP, clinicians, should strive for 100% vaccination rates in appropriate persons.

Corresponding author: Tze Shein Lo, MD, University of North Dakota, 1919 Elm Street, Fargo, ND 58102, [email protected].

Financial disclosures: None.

Author contributions: drafting of article, PM, TSL; critical revision of the article, PM, TSL.

From the Division of Allergy, Pulmonary, and Critical Care Medicine, Department of Medicine, Vanderbilt University Medical Center, Nashville, TN (Dr. Marra), Division of Anesthesiology Critical Care Medicine, Vanderbilt University Medical Center, Nashville, TN (Dr. Hayhurst, Dr. Hughes, Dr. Pandharipande), Department of Clinical and Experimental Science, University of Brescia, Brescia, Italy (Dr. Marengoni), School of Medicine and Surgery,
University of Milano-Bicocca, Milan, Italy (Dr. Bellelli), and Rehabilitation and Aged Care Unit Hospital Ancelle, Cremona, Italy (Dr. Morandi).

Abstract

• Objective: To present an overview of the phenomenon of inappropriate medication prescription in older critically ill patients and examine possible strategies of intervention.
• Methods: Review of the literature.
• Results: Polypharmacy and inappropriate prescribing of medications in older persons may lead to a significant risk of adverse drug-related events and mortality. The intensive care unit (ICU) is often the place where potentially inappropriate medications (PIMs) are first prescribed. Common PIMs at ICU discharge are antipsychotics, benzodiazepines, opioids, anticholinergic medications, antidepressants, and drugs causing orthostatic hypotension. Different classes of medications, typically intended for short-term use, are sometimes inappropriately continued after discharge from the hospital. At admission, potential risk factors for PIM are multiple morbidities, polypharmacy, frailty and cognitive decline; at discharge, a high number of pre-admission PIMs, discharge to a location other than home, discharge from a surgical service, longer length of ICU and hospital stay, and mechanical ventilation. Inappropriate prescribing in older patients can be detected through either the use of explicit criteria, drug utilization reviews, and multidisciplinary teams, including a geriatrician and/or the involvement of a clinical pharmacist.
• Conclusion: Use of PIMs may be common in critical patients, both on admission and at discharge from ICU. Therapeutic reconciliation is recommended at every transition of care (eg, at hospital or ICU admission and discharge) in order to improve appropriateness of prescription.

Key words: elderly; intensive care unit; inappropriate medications; antipsychotics.

Since older persons are often affected by multiple chronic diseases and are prescribed several medications, the quality and safety of prescribing these medications has become a global health care issue [1–4]. Polypharmacy and inappropriate prescribing of medications among the elderly is receiving significant attention in the medical literature [5,6]. Inappropriate medications in the elderly can lead to falls, cognitive impairment and delirium, poorer health status, and higher mortality [7–10]. Medications are considered potentially inappropriate when (a) the risks of treatment outweigh the benefits [11], (b) they are prescribed for periods longer than clinically indicated or without any clear indication, (c) they are not prescribed when indicated [12], and (d) they are likely to interact with other drugs and diseases. Medications included in this category are often referred to as potentially inappropriate medications (PIMs), as in some situations their use is justified; however, if the risk of harm from the drug is judged to outweigh the potential clinical benefit after an individual patient’s clinical circumstances are considered, these drugs are considered “actually inappropriate medications” (AIMs) [6].

Advancing age is associated with substantial pharmacokinetic and pharmacodynamics changes, such as altered distribution volumes and altered permeability of the blood-brain barrier, impaired liver metabolism and renal capacity, up- and down-regulation of target receptors, transmitters, and signaling pathways changes, impaired homeostasis, and increased risk of adverse drug reactions (ADRs) that lead to increased mortality and morbidity and higher health care costs [2,11,13–19]. Studies show that ADRs cause approximately 5% of hospital admissions in the general population, but the percentage rises to 10% in older persons [20].

Avoiding PIMs represents a strategy aimed at reducing drug-related mortality and morbidity. This article provides an overview of the phenomenon of inappropriate medication prescription in older critically ill patients and examines available strategies of intervention.

Inappropriate Medications at ICU Discharge

Though PIMs and AIMs may be identified at the time of hospital discharge, the intensive care unit (ICU) is often the place where these medications are first prescribed [21]. Acute hospitalization may increase PIM prescribing because of newly prescribed medications, the presence of multiple prescribers, inadequate medication reconciliation, and a lack of care coordination among inpatient providers or in the transition back to outpatient care [22)].

A known complication of critical illness and ICU stay is a significant increase in psychological symptoms, sleep cycle alterations, delirium, and cognitive impairment, which may be associated with increased prescription of specific PIMs, such as antipsychotics or benzodiazepines [6,23,24]. Despite the lack of reliable evidence supporting their use in the ICU, antipsychotic agents are used routinely in ICU patients [25] to treat a variety of conditions, such as substance withdrawal, agitation not responding to other therapies, or delirium. Results from a multicenter study of 164,996 hospitalizations across 71 academic medical centers in the US showed that 1 out of 10 ICU patients received an antipsychotic during their hospital stay [25]. Jasiak et al estimated that one-third of patients initiated on an atypical antipsychotic therapy for ICU delirium received a hospital discharge prescription for these medications, with a potential annual outpatient medication cost of approximately $2255 per patient [26].

One potential consequence of antipsychotic use in the ICU is their continuation after the transition to other clinical settings, including discharge from the hospital [27] (Table 1). 

When examining the specific factors that may contribute to a patient being discharged on an antipsychotic, authors found that the specific antipsychotic used correlated with risk of continuation [27,30], with atypical antipsychotics having a greater likelihood of being continued than haloperidol [27,30]. Possible explanation for these results could be that physicians perceive less long-term risk from atypical agents, so may be more likely to continue them on discharge [30]. However, such an approach is not always safe. Indeed, although atypical antipsychotic agents tend to cause less tardive dyskinesia, they are known to be associated with similar rates of other adverse events compared with typical agents and have been linked to an increased risk of sudden cardiac death and pneumonia in the elderly [31,32].

Other factors independently associated with being discharged on a new antipsychotic medication were the severity of the acute illness as measured with the Acute Physiology and Chronic Health Evaluation II score at ICU admission (odds ratio [OR] 1 [95% confidence interval {CI}, 1.0–1.1]) and days treated with benzodiazepines (OR 1.1 [95% CI, 1.0–1.14]) [30]. Conversely, perhaps due to different practice patterns, Tomichek et al did not find an association between benzodiazepines administration and antipsychotic prescription at discharge in post hoc analyses [27].

Another possible reason for antipsychotic continuation may reside in the indication chosen [33]. Antipsychotic agents have sedative properties and they might be used to optimize sleep during hospitalization, despite the lack of evidence to support this indication [34]. Other factors potentially contributing to continuation of antipsychotics may include persistent delirium and agitation, newly diagnosed psychiatric illness, and difficulties experienced by physicians in deprescribing [35] with improper/incomplete medication reconciliation [33].

The continuation of antipsychotic therapy increased 30-day readmission rates in patients compared to those who had therapy stopped before discharge [33]. In addition to the well-described cardiac effects (prolonged QT interval), neuroleptic malignant syndrome and extrapyramidal symptoms may also occur, and longer-term use can predispose patients to metabolic disturbances, falls, and increase the risk of death in elderly patients with dementia [31].

Benzodiazepines and sedative hypnotics are commonly used to treat insomnia and agitation in older adults despite significant risk. Benzodiazepine administration was found to be an independent risk factor for a daily transition to delirium [36,37]. Pandharipande et al reported that every unit dose of lorazepam was associated with a higher risk for daily transition to delirium (OR 1.2, 95% CI 1.1–1.4, P = 0.003) [36] in critically ill patients. A more recent analysis found for every 5 mg of midazolam administered to a patient who is awake and without delirium, there is a 4% chance that this patient will develop delirium the next ICU day [37].

Given that the risk for benzodiazepine-associated delirium is dose-dependent, clinicians should use strategies known to reduce the daily number of benzodiazepines administered that often includes the use of a sedative associated with less delirium occurrence, such as dexmedetomidine or propofol [38]. Evidence has shown that long-term use of benzodiazepines has little benefit with many risks, including an increased susceptibility to spontaneous bacterial infection [39,40] and mortality in the setting of infection [41]. Nakafero et al showed that exposure to benzodiazepines was associated with increased occurrence of both influenza-like-illness–related pneumonia and mortality. Benzodiazepine use was associated also with increased occurrence of asthma exacerbation and with increased all-cause mortality during a median follow-up of 2 years in a cohort of asthmatic patients [42] as well with an increased risk of pneumonia and long-term mortality in patients with a prior diagnosis of community- acquired pneumonia [40]. Long-term use of benzodiazepines is also associated with increased risk of falls [43–45], cognitive impairment [46–48] and disability [49,50].

Other common types of PIMs at ICU discharge were opioids, anticholinergic medications, antidepressants, and drugs causing orthostatic hypotension [6]. Of the anticholinergic AIMs, H2 blockers (61%) and promethazine (15%) were the most common [6]. Only 16% of opioids, 23% of antidepressants, and 10% of drugs causing orthostatic hypotension were found to be actually inappropriate after the patient’s circumstances were considered (eg, postoperative pain control, a new diagnosis of major depressive disorder) [6].

Inappropriate Medications at Hospital Discharge

Medications typically intended for short-term use during acute illness are sometimes continued after discharge without documented indication [51]. Poudel et al found that in 206 patients 70 years of age and older discharged to residential aged care facilities from acute care, at least 1 PIM was identified in 112 (54.4%) patients on admission and 102 (49.5%) patients on discharge [11]. Commonly prescribed PIM categories, at both admission and discharge, were central nervous system, cardiovascular, gastrointestinal, and respiratory drugs and analgesics [6,11,52,53]. Of all medications prescribed at admission (1728), 10.8% were PIMs, and at discharge, of 1759 medications, 9.6% were PIMs. Of the total 187 PIMs on admission, 56 (30%) were stopped, and 131 (70%) were continued; 32 new PIMs were introduced [11].

Morandi et al in 2011 conducted a prospective cohort study including 120 patients age ≥ 60 who were discharged after receiving care in a medical, surgical, or cardiovascular ICU for shock or respiratory failure. The percentage of patients prescribed at least 1 PIM increased from 66% at pre-admission to 85% at discharge. The number of patients with 0 PIMs dropped from 34% at preadmission to 14% at discharge, and the number of patients with 3 or more PIMS increased from 16% at preadmission to 37% at discharge. While it is possible that these drugs may be appropriate when started during an acute illness in the ICU (eg, stress ulcer prophylaxis with H2-antagonists in mechanically ventilated patients), most should have been discontinued at ICU and/or hospital discharge [21].

Inappropriate prescriptions of proton pump inhibitors (PPIs) in hospital and primary care have been widely reported [54,55]. In a study conducted by Ahrens et al in 31 primary care practices, for 58% (263/506) of patients discharged from 35 hospitals with a PPI recommendation in hospital discharge letters, an appropriate indication was missing. In 57% of these cases general practitioners followed this recommendation and continued the prescription for more than 1 month [54]. The strongest factor associated with appropriate and inappropriate continuation of PPI after discharge was PPI prescription prior to hospitalization [54]. Although PPIs are safe, they can cause adverse effects. PPI intake has been found to have a significant association with risk of community-acquired pneumonia [56,57], hip fractures [58], Clostridium difficile-associated diarrhea [55,61,62], and to reduce the therapeutic effects of bisphosphonates [59] and low-dose aspirin [60].

Unintentional medication continuation is not a problem isolated to a single drug class or disease [63]. Scales et al evaluated rates of and risk factors for potentially unintentional medication continuation following hospitalization in a population of elderly patients (≥ 66 years) [51]. They created distinct cohorts by identifying seniors not previously receiving four classes of medications typically used to treat or prevent complications of acute illness: antipsychotic medications; gastric acid suppressants (ie, histamine-2 blockers and proton pump inhibitors); benzodiazepines; and inhaled bronchodilators and steroids [51]. Prescription without documented indication occurred across all medication classes, from 12,209 patients (1.4 %) for antipsychotic medications to 34,140 patients (6.1 %) for gastric acid suppressants [51].

Several potential risk factors were considered. The relationship between multimorbidity and polypharmacy is well described in the literature, and several studies have identified a positive association between the number of drugs and the use of PIMs [64–66]. Conversely, Poudel et al did not find any association between polypharmacy and PIM use [11]. Associations were found between the use of PIMs, frailty status, and cognitive decline of patients at admission and at discharge [11], while no association was observed with age, gender, in-hospital falls, delirium, and functional decline [11,67]. Other potential risk factors of a high number of PIMs at discharge were a high number of pre-admission PIMs, discharge to a location other than home, and discharge from a surgical service [1,6,68,69]. Length of ICU stay and mechanical ventilation had a positive influence on the number of PIMs used by acutely ill older patients [11,63,69]. In the study of Scales et al, the greatest absolute risk factor across all medication groups was longer hospitalization. The increased OR for medication continuation after a hospitalization lasting more than 7 days ranged from 2.03 (95% CI 1.94–2.11) for respiratory inhalers to 6.35 (95% CI 5.91–6.82) for antipsychotic medications [51].

Inappropriate Medications: Where and How to Intervene?

Early detection of PIMs may prevent adverse drug events and improve geriatric care in older adults [13,70]. PIM prevalence can often be a useful indicator of prescribing quality [2]. Appropriate interventions and an improved quality of prescribed medications require appropriate assessment tools to decrease the number of patients discharged on these medications [71,72]. Medication reconciliation is the process of avoiding inadvertent inconsistencies within a patient’s drug regimen, which can occur during transitions in different setting of care [73]. A multidisciplinary team should be involved in the medication reconciliation at each care transition to reevaluate medications use according to the clinical conditions, cognitive/functional status and the coexistence of geriatric syndromes (eg, dementia, malnutrition, delirium, urinary incontinence, frailty) (Figure).

Criteria for the Evaluation of Inappropriate Medications Prescription

Explicit criteria derived from expert reports or published reviews are available (Table 2).

Beers criteria PIMs have been found to be associated with poor health outcomes, including confusion, falls, and mortality [7,75,78]. The STOPP (Screening Tool of Older Person’s potentially inappropriate Prescriptions) and START (Screening Tool to Alert doctors to the Right Treatment) are evidence-based sets of criteria that were developed in Ireland and updated in October 2014, including some of the new criteria for direct oral anticoagulants, drugs affecting or affected by renal system and anti-muscarinic/anticholinergic agents [79]. 

Several other sets of criteria have been published to identify PIMs, such as the FORTA (Fit for the Aged) and the PRISCUS [86] criteria. FORTA allows a disease-related evaluation revealing over-treatment and under-treatment, and medications are graded as follows: A, indispensable drug, clear-cut benefit in terms of efficacy/safety ratio proven in elderly patients for a given indication; B, drugs with proven or obvious efficacy in the elderly, but limited extent of effect or safety concerns; C, drugs with questionable efficacy/safety profiles in the elderly which should be avoided or omitted in the presence of too many drugs or side effects; D, avoid in the elderly, omit first, refer also to negative listings. Negative lists such as PRISCUS, which provide an explicit listing of drugs, independent of the diagnosis, are easy to use. On the other hand, constant updates are needed, and such lists carry the risk of an assumption that drugs not listed would be appropriate in every case [87]. Both sets of criteria have in common that they refer to long-term medication and drugs frequently used during the inpatient stay, such as antibiotics, are hardly taken into account [87].

The Medication Appropriateness Index measures overall prescribing quality through 10 separate but interrelated domains [8]. Three components are used to detect PIMs: indication, effectiveness, and duplication. However, it does not give any precise guidance in relation to specific medicines and therefore has limited application for objectively defining PIMs.

Another prescribing quality assessment tool is the Inappropriate Prescribing in the Elderly Tool (IPET), which consists of a list of the 14 most prevalent prescription errors identified from an extensive list of inappropriate prescription instances drawn up by an expert Canadian Consensus Panel [88,89].

Another approach to assess the appropriateness of drugs prescribed for older people is the use of Drug Utilization Reviews (DURs) [16]. DURs use consensus opinion by drug therapy experts to define standards or explicit criteria for a single drug, class of drugs, or group of drugs [16]. DURs typically use retrospective information from large, nonclinical administrative databases to identify problems such as dosage range, duration, therapeutic duplication, and drug interactions [90, 91]. Monane et al [92] evaluated a program designed to decrease the use of PIMs among the elderly through a computerized online DUR database. Computer alerts triggered telephone calls to physicians by pharmacists to discuss a potential problem and any therapeutic substitution options. From a total of 43,007 telepharmacy calls generated by the alerts, they were able to reach 19,368 physicians regarding 24,266 alerts (56%). The rate of change to a more appropriate therapeutic agent was 24% (5860), but ranged from 40% for long half-life benzodiazepines to 2% to 7% for drugs that theoretically were contraindicated by patients’ self-reported history [92].

Computerized Support Systems to Reduce Inappropriate Prescribing in the Elderly

Other potential solutions for reducing inappropriate medications may include continuing medical education, electronic medical records surveillance, routine clinical evaluation, and/or improved hand-off communication between discharging and accepting providers. Incorporating this assessment of medication appropriateness into the medication reconciliation process when patients are discharged or transferred out of the ICU has the potential to enhance patient safety [21,93]. A randomized controlled trial conducted by Raebel et al [94] reported the effectiveness of a computerized pharmacy alert system plus collaboration between health care professionals for decreasing potentially inappropriate medication dispensing in elderly patients. Another study showed that computer-based access to complete drug profiles and alerts about potential prescribing problems reduced the occurrence of potentially inappropriate prescriptions [95]. A summary of these studies is shown in Table 3.

Interdisciplinary Teams to Reduce Inappropriate Prescribing in the Elderly

Some studies evaluated the effect of multidisciplinary teamwork in improving inappropriate medication prescribing in the elderly (Table 4).

Pharmacists in hospitals can play a significant role in the initiation of changes to patient’s therapy and management [11] (Table 5).

Mattison et al recently emphasized that studies of PIMs should determine scenarios in which it is appropriate to prescribe PIMs, moving beyond simply labeling some medications as “potentially inappropriate,” since some PIMs are appropriately prescribed in specific clinical situations [109]. Morandi et al showed that the positive predictive value (PPV) depends on the drug type. Thus, when developing a screening system, one cannot be concerned only with high negative predictive value (NPV), one must consider PPV as well [6]. Screening tools that include medication classes with low PPV will generate false positive “flags” or warnings, which could lead to misguided clinical decisions [6]. The fact that many PIMs are not AIMs also reveals the value of using a multidisciplinary team to identify AIMs from lists of PIMs generated when discharge medication lists are screened [6,110]. Thus, a multidisciplinary team is needed to consider the clinical context to distinguish PIMs from AIMs [6]. Of course, such a team is not available in some settings; when resources are limited, knowledge of which PIMs are most likely AIMs (ie, have high PPVs) could guide the development of computer-based decision support systems or other surveillance approaches that are efficient in that particular setting [6].

Approaches for optimizing prescribing in this population mainly depend on patient needs and comorbidities and most available data are derived from randomized controlled trials involving a single drug. Such trials do not take into account the confounding effects of multiple comorbidities and patient preferences. Therefore, approaches for optimizing prescription management that are available for and validated in younger patients are not applicable to elderly subjects [3,111].

Conclusion

Clinicians should seek to identify and discontinue AIMs at 3 important transitions during a critically ill elderly patient’s hospital course: at the time of hospital or ICU admission; at ICU discharge; and at hospital discharge. The patient’s clinical situation should be reviewed at every transition points, ideally by a multidisciplinary team of clinicians, to judge the appropriateness of each PIM [6]. After the hospital discharge, patient’s medications should be then reviewed by a multidisciplinary team and/or by the primary care physician according to the final discharge destination (ie, home, nursing home, rehabilitation) by using any of the validated tools. Regardless of the approach, it is clear that standardized care processes, including enhanced clinical decision support, are necessary to ensure that physicians do not continue exposing our patients to unnecessary medications and harm after discharge.

Corresponding author: Alessandro Morandi, MD, MPH, [email protected].

Funding/support: Dr. Pandiharipande is supported by National Institutes of Health HL111111 (Bethesda, MD) and by the VA Clinical Science Research and Development Service (Washington, DC) and the National Institutes of Health AG027472 and AG035117 (Bethesda, MD).

Financial disclosures: Dr. Pratik Pandharipande has received a research grant from Hospira Inc in collaboration with the NIH.

From the Division of Allergy, Pulmonary, and Critical Care Medicine, Department of Medicine, Vanderbilt University Medical Center, Nashville, TN (Dr. Marra), Division of Anesthesiology Critical Care Medicine, Vanderbilt University Medical Center, Nashville, TN (Dr. Hayhurst, Dr. Hughes, Dr. Pandharipande), Department of Clinical and Experimental Science, University of Brescia, Brescia, Italy (Dr. Marengoni), School of Medicine and Surgery,
University of Milano-Bicocca, Milan, Italy (Dr. Bellelli), and Rehabilitation and Aged Care Unit Hospital Ancelle, Cremona, Italy (Dr. Morandi).

Abstract

• Objective: To present an overview of the phenomenon of inappropriate medication prescription in older critically ill patients and examine possible strategies of intervention.
• Methods: Review of the literature.
• Results: Polypharmacy and inappropriate prescribing of medications in older persons may lead to a significant risk of adverse drug-related events and mortality. The intensive care unit (ICU) is often the place where potentially inappropriate medications (PIMs) are first prescribed. Common PIMs at ICU discharge are antipsychotics, benzodiazepines, opioids, anticholinergic medications, antidepressants, and drugs causing orthostatic hypotension. Different classes of medications, typically intended for short-term use, are sometimes inappropriately continued after discharge from the hospital. At admission, potential risk factors for PIM are multiple morbidities, polypharmacy, frailty and cognitive decline; at discharge, a high number of pre-admission PIMs, discharge to a location other than home, discharge from a surgical service, longer length of ICU and hospital stay, and mechanical ventilation. Inappropriate prescribing in older patients can be detected through either the use of explicit criteria, drug utilization reviews, and multidisciplinary teams, including a geriatrician and/or the involvement of a clinical pharmacist.
• Conclusion: Use of PIMs may be common in critical patients, both on admission and at discharge from ICU. Therapeutic reconciliation is recommended at every transition of care (eg, at hospital or ICU admission and discharge) in order to improve appropriateness of prescription.

Key words: elderly; intensive care unit; inappropriate medications; antipsychotics.

Since older persons are often affected by multiple chronic diseases and are prescribed several medications, the quality and safety of prescribing these medications has become a global health care issue [1–4]. Polypharmacy and inappropriate prescribing of medications among the elderly is receiving significant attention in the medical literature [5,6]. Inappropriate medications in the elderly can lead to falls, cognitive impairment and delirium, poorer health status, and higher mortality [7–10]. Medications are considered potentially inappropriate when (a) the risks of treatment outweigh the benefits [11], (b) they are prescribed for periods longer than clinically indicated or without any clear indication, (c) they are not prescribed when indicated [12], and (d) they are likely to interact with other drugs and diseases. Medications included in this category are often referred to as potentially inappropriate medications (PIMs), as in some situations their use is justified; however, if the risk of harm from the drug is judged to outweigh the potential clinical benefit after an individual patient’s clinical circumstances are considered, these drugs are considered “actually inappropriate medications” (AIMs) [6].

Advancing age is associated with substantial pharmacokinetic and pharmacodynamics changes, such as altered distribution volumes and altered permeability of the blood-brain barrier, impaired liver metabolism and renal capacity, up- and down-regulation of target receptors, transmitters, and signaling pathways changes, impaired homeostasis, and increased risk of adverse drug reactions (ADRs) that lead to increased mortality and morbidity and higher health care costs [2,11,13–19]. Studies show that ADRs cause approximately 5% of hospital admissions in the general population, but the percentage rises to 10% in older persons [20].

Avoiding PIMs represents a strategy aimed at reducing drug-related mortality and morbidity. This article provides an overview of the phenomenon of inappropriate medication prescription in older critically ill patients and examines available strategies of intervention.

Inappropriate Medications at ICU Discharge

Though PIMs and AIMs may be identified at the time of hospital discharge, the intensive care unit (ICU) is often the place where these medications are first prescribed [21]. Acute hospitalization may increase PIM prescribing because of newly prescribed medications, the presence of multiple prescribers, inadequate medication reconciliation, and a lack of care coordination among inpatient providers or in the transition back to outpatient care [22)].

A known complication of critical illness and ICU stay is a significant increase in psychological symptoms, sleep cycle alterations, delirium, and cognitive impairment, which may be associated with increased prescription of specific PIMs, such as antipsychotics or benzodiazepines [6,23,24]. Despite the lack of reliable evidence supporting their use in the ICU, antipsychotic agents are used routinely in ICU patients [25] to treat a variety of conditions, such as substance withdrawal, agitation not responding to other therapies, or delirium. Results from a multicenter study of 164,996 hospitalizations across 71 academic medical centers in the US showed that 1 out of 10 ICU patients received an antipsychotic during their hospital stay [25]. Jasiak et al estimated that one-third of patients initiated on an atypical antipsychotic therapy for ICU delirium received a hospital discharge prescription for these medications, with a potential annual outpatient medication cost of approximately $2255 per patient [26].

One potential consequence of antipsychotic use in the ICU is their continuation after the transition to other clinical settings, including discharge from the hospital [27] (Table 1). 

When examining the specific factors that may contribute to a patient being discharged on an antipsychotic, authors found that the specific antipsychotic used correlated with risk of continuation [27,30], with atypical antipsychotics having a greater likelihood of being continued than haloperidol [27,30]. Possible explanation for these results could be that physicians perceive less long-term risk from atypical agents, so may be more likely to continue them on discharge [30]. However, such an approach is not always safe. Indeed, although atypical antipsychotic agents tend to cause less tardive dyskinesia, they are known to be associated with similar rates of other adverse events compared with typical agents and have been linked to an increased risk of sudden cardiac death and pneumonia in the elderly [31,32].

Other factors independently associated with being discharged on a new antipsychotic medication were the severity of the acute illness as measured with the Acute Physiology and Chronic Health Evaluation II score at ICU admission (odds ratio [OR] 1 [95% confidence interval {CI}, 1.0–1.1]) and days treated with benzodiazepines (OR 1.1 [95% CI, 1.0–1.14]) [30]. Conversely, perhaps due to different practice patterns, Tomichek et al did not find an association between benzodiazepines administration and antipsychotic prescription at discharge in post hoc analyses [27].

Another possible reason for antipsychotic continuation may reside in the indication chosen [33]. Antipsychotic agents have sedative properties and they might be used to optimize sleep during hospitalization, despite the lack of evidence to support this indication [34]. Other factors potentially contributing to continuation of antipsychotics may include persistent delirium and agitation, newly diagnosed psychiatric illness, and difficulties experienced by physicians in deprescribing [35] with improper/incomplete medication reconciliation [33].

The continuation of antipsychotic therapy increased 30-day readmission rates in patients compared to those who had therapy stopped before discharge [33]. In addition to the well-described cardiac effects (prolonged QT interval), neuroleptic malignant syndrome and extrapyramidal symptoms may also occur, and longer-term use can predispose patients to metabolic disturbances, falls, and increase the risk of death in elderly patients with dementia [31].

Benzodiazepines and sedative hypnotics are commonly used to treat insomnia and agitation in older adults despite significant risk. Benzodiazepine administration was found to be an independent risk factor for a daily transition to delirium [36,37]. Pandharipande et al reported that every unit dose of lorazepam was associated with a higher risk for daily transition to delirium (OR 1.2, 95% CI 1.1–1.4, P = 0.003) [36] in critically ill patients. A more recent analysis found for every 5 mg of midazolam administered to a patient who is awake and without delirium, there is a 4% chance that this patient will develop delirium the next ICU day [37].

Given that the risk for benzodiazepine-associated delirium is dose-dependent, clinicians should use strategies known to reduce the daily number of benzodiazepines administered that often includes the use of a sedative associated with less delirium occurrence, such as dexmedetomidine or propofol [38]. Evidence has shown that long-term use of benzodiazepines has little benefit with many risks, including an increased susceptibility to spontaneous bacterial infection [39,40] and mortality in the setting of infection [41]. Nakafero et al showed that exposure to benzodiazepines was associated with increased occurrence of both influenza-like-illness–related pneumonia and mortality. Benzodiazepine use was associated also with increased occurrence of asthma exacerbation and with increased all-cause mortality during a median follow-up of 2 years in a cohort of asthmatic patients [42] as well with an increased risk of pneumonia and long-term mortality in patients with a prior diagnosis of community- acquired pneumonia [40]. Long-term use of benzodiazepines is also associated with increased risk of falls [43–45], cognitive impairment [46–48] and disability [49,50].

Other common types of PIMs at ICU discharge were opioids, anticholinergic medications, antidepressants, and drugs causing orthostatic hypotension [6]. Of the anticholinergic AIMs, H2 blockers (61%) and promethazine (15%) were the most common [6]. Only 16% of opioids, 23% of antidepressants, and 10% of drugs causing orthostatic hypotension were found to be actually inappropriate after the patient’s circumstances were considered (eg, postoperative pain control, a new diagnosis of major depressive disorder) [6].

Inappropriate Medications at Hospital Discharge

Medications typically intended for short-term use during acute illness are sometimes continued after discharge without documented indication [51]. Poudel et al found that in 206 patients 70 years of age and older discharged to residential aged care facilities from acute care, at least 1 PIM was identified in 112 (54.4%) patients on admission and 102 (49.5%) patients on discharge [11]. Commonly prescribed PIM categories, at both admission and discharge, were central nervous system, cardiovascular, gastrointestinal, and respiratory drugs and analgesics [6,11,52,53]. Of all medications prescribed at admission (1728), 10.8% were PIMs, and at discharge, of 1759 medications, 9.6% were PIMs. Of the total 187 PIMs on admission, 56 (30%) were stopped, and 131 (70%) were continued; 32 new PIMs were introduced [11].

Morandi et al in 2011 conducted a prospective cohort study including 120 patients age ≥ 60 who were discharged after receiving care in a medical, surgical, or cardiovascular ICU for shock or respiratory failure. The percentage of patients prescribed at least 1 PIM increased from 66% at pre-admission to 85% at discharge. The number of patients with 0 PIMs dropped from 34% at preadmission to 14% at discharge, and the number of patients with 3 or more PIMS increased from 16% at preadmission to 37% at discharge. While it is possible that these drugs may be appropriate when started during an acute illness in the ICU (eg, stress ulcer prophylaxis with H2-antagonists in mechanically ventilated patients), most should have been discontinued at ICU and/or hospital discharge [21].

Inappropriate prescriptions of proton pump inhibitors (PPIs) in hospital and primary care have been widely reported [54,55]. In a study conducted by Ahrens et al in 31 primary care practices, for 58% (263/506) of patients discharged from 35 hospitals with a PPI recommendation in hospital discharge letters, an appropriate indication was missing. In 57% of these cases general practitioners followed this recommendation and continued the prescription for more than 1 month [54]. The strongest factor associated with appropriate and inappropriate continuation of PPI after discharge was PPI prescription prior to hospitalization [54]. Although PPIs are safe, they can cause adverse effects. PPI intake has been found to have a significant association with risk of community-acquired pneumonia [56,57], hip fractures [58], Clostridium difficile-associated diarrhea [55,61,62], and to reduce the therapeutic effects of bisphosphonates [59] and low-dose aspirin [60].

Unintentional medication continuation is not a problem isolated to a single drug class or disease [63]. Scales et al evaluated rates of and risk factors for potentially unintentional medication continuation following hospitalization in a population of elderly patients (≥ 66 years) [51]. They created distinct cohorts by identifying seniors not previously receiving four classes of medications typically used to treat or prevent complications of acute illness: antipsychotic medications; gastric acid suppressants (ie, histamine-2 blockers and proton pump inhibitors); benzodiazepines; and inhaled bronchodilators and steroids [51]. Prescription without documented indication occurred across all medication classes, from 12,209 patients (1.4 %) for antipsychotic medications to 34,140 patients (6.1 %) for gastric acid suppressants [51].

Several potential risk factors were considered. The relationship between multimorbidity and polypharmacy is well described in the literature, and several studies have identified a positive association between the number of drugs and the use of PIMs [64–66]. Conversely, Poudel et al did not find any association between polypharmacy and PIM use [11]. Associations were found between the use of PIMs, frailty status, and cognitive decline of patients at admission and at discharge [11], while no association was observed with age, gender, in-hospital falls, delirium, and functional decline [11,67]. Other potential risk factors of a high number of PIMs at discharge were a high number of pre-admission PIMs, discharge to a location other than home, and discharge from a surgical service [1,6,68,69]. Length of ICU stay and mechanical ventilation had a positive influence on the number of PIMs used by acutely ill older patients [11,63,69]. In the study of Scales et al, the greatest absolute risk factor across all medication groups was longer hospitalization. The increased OR for medication continuation after a hospitalization lasting more than 7 days ranged from 2.03 (95% CI 1.94–2.11) for respiratory inhalers to 6.35 (95% CI 5.91–6.82) for antipsychotic medications [51].

Inappropriate Medications: Where and How to Intervene?

Early detection of PIMs may prevent adverse drug events and improve geriatric care in older adults [13,70]. PIM prevalence can often be a useful indicator of prescribing quality [2]. Appropriate interventions and an improved quality of prescribed medications require appropriate assessment tools to decrease the number of patients discharged on these medications [71,72]. Medication reconciliation is the process of avoiding inadvertent inconsistencies within a patient’s drug regimen, which can occur during transitions in different setting of care [73]. A multidisciplinary team should be involved in the medication reconciliation at each care transition to reevaluate medications use according to the clinical conditions, cognitive/functional status and the coexistence of geriatric syndromes (eg, dementia, malnutrition, delirium, urinary incontinence, frailty) (Figure).

Criteria for the Evaluation of Inappropriate Medications Prescription

Explicit criteria derived from expert reports or published reviews are available (Table 2).

Beers criteria PIMs have been found to be associated with poor health outcomes, including confusion, falls, and mortality [7,75,78]. The STOPP (Screening Tool of Older Person’s potentially inappropriate Prescriptions) and START (Screening Tool to Alert doctors to the Right Treatment) are evidence-based sets of criteria that were developed in Ireland and updated in October 2014, including some of the new criteria for direct oral anticoagulants, drugs affecting or affected by renal system and anti-muscarinic/anticholinergic agents [79]. 

Several other sets of criteria have been published to identify PIMs, such as the FORTA (Fit for the Aged) and the PRISCUS [86] criteria. FORTA allows a disease-related evaluation revealing over-treatment and under-treatment, and medications are graded as follows: A, indispensable drug, clear-cut benefit in terms of efficacy/safety ratio proven in elderly patients for a given indication; B, drugs with proven or obvious efficacy in the elderly, but limited extent of effect or safety concerns; C, drugs with questionable efficacy/safety profiles in the elderly which should be avoided or omitted in the presence of too many drugs or side effects; D, avoid in the elderly, omit first, refer also to negative listings. Negative lists such as PRISCUS, which provide an explicit listing of drugs, independent of the diagnosis, are easy to use. On the other hand, constant updates are needed, and such lists carry the risk of an assumption that drugs not listed would be appropriate in every case [87]. Both sets of criteria have in common that they refer to long-term medication and drugs frequently used during the inpatient stay, such as antibiotics, are hardly taken into account [87].

The Medication Appropriateness Index measures overall prescribing quality through 10 separate but interrelated domains [8]. Three components are used to detect PIMs: indication, effectiveness, and duplication. However, it does not give any precise guidance in relation to specific medicines and therefore has limited application for objectively defining PIMs.

Another prescribing quality assessment tool is the Inappropriate Prescribing in the Elderly Tool (IPET), which consists of a list of the 14 most prevalent prescription errors identified from an extensive list of inappropriate prescription instances drawn up by an expert Canadian Consensus Panel [88,89].

Another approach to assess the appropriateness of drugs prescribed for older people is the use of Drug Utilization Reviews (DURs) [16]. DURs use consensus opinion by drug therapy experts to define standards or explicit criteria for a single drug, class of drugs, or group of drugs [16]. DURs typically use retrospective information from large, nonclinical administrative databases to identify problems such as dosage range, duration, therapeutic duplication, and drug interactions [90, 91]. Monane et al [92] evaluated a program designed to decrease the use of PIMs among the elderly through a computerized online DUR database. Computer alerts triggered telephone calls to physicians by pharmacists to discuss a potential problem and any therapeutic substitution options. From a total of 43,007 telepharmacy calls generated by the alerts, they were able to reach 19,368 physicians regarding 24,266 alerts (56%). The rate of change to a more appropriate therapeutic agent was 24% (5860), but ranged from 40% for long half-life benzodiazepines to 2% to 7% for drugs that theoretically were contraindicated by patients’ self-reported history [92].

Computerized Support Systems to Reduce Inappropriate Prescribing in the Elderly

Other potential solutions for reducing inappropriate medications may include continuing medical education, electronic medical records surveillance, routine clinical evaluation, and/or improved hand-off communication between discharging and accepting providers. Incorporating this assessment of medication appropriateness into the medication reconciliation process when patients are discharged or transferred out of the ICU has the potential to enhance patient safety [21,93]. A randomized controlled trial conducted by Raebel et al [94] reported the effectiveness of a computerized pharmacy alert system plus collaboration between health care professionals for decreasing potentially inappropriate medication dispensing in elderly patients. Another study showed that computer-based access to complete drug profiles and alerts about potential prescribing problems reduced the occurrence of potentially inappropriate prescriptions [95]. A summary of these studies is shown in Table 3.

Interdisciplinary Teams to Reduce Inappropriate Prescribing in the Elderly

Some studies evaluated the effect of multidisciplinary teamwork in improving inappropriate medication prescribing in the elderly (Table 4).

Pharmacists in hospitals can play a significant role in the initiation of changes to patient’s therapy and management [11] (Table 5).

Mattison et al recently emphasized that studies of PIMs should determine scenarios in which it is appropriate to prescribe PIMs, moving beyond simply labeling some medications as “potentially inappropriate,” since some PIMs are appropriately prescribed in specific clinical situations [109]. Morandi et al showed that the positive predictive value (PPV) depends on the drug type. Thus, when developing a screening system, one cannot be concerned only with high negative predictive value (NPV), one must consider PPV as well [6]. Screening tools that include medication classes with low PPV will generate false positive “flags” or warnings, which could lead to misguided clinical decisions [6]. The fact that many PIMs are not AIMs also reveals the value of using a multidisciplinary team to identify AIMs from lists of PIMs generated when discharge medication lists are screened [6,110]. Thus, a multidisciplinary team is needed to consider the clinical context to distinguish PIMs from AIMs [6]. Of course, such a team is not available in some settings; when resources are limited, knowledge of which PIMs are most likely AIMs (ie, have high PPVs) could guide the development of computer-based decision support systems or other surveillance approaches that are efficient in that particular setting [6].

Approaches for optimizing prescribing in this population mainly depend on patient needs and comorbidities and most available data are derived from randomized controlled trials involving a single drug. Such trials do not take into account the confounding effects of multiple comorbidities and patient preferences. Therefore, approaches for optimizing prescription management that are available for and validated in younger patients are not applicable to elderly subjects [3,111].

Conclusion

Clinicians should seek to identify and discontinue AIMs at 3 important transitions during a critically ill elderly patient’s hospital course: at the time of hospital or ICU admission; at ICU discharge; and at hospital discharge. The patient’s clinical situation should be reviewed at every transition points, ideally by a multidisciplinary team of clinicians, to judge the appropriateness of each PIM [6]. After the hospital discharge, patient’s medications should be then reviewed by a multidisciplinary team and/or by the primary care physician according to the final discharge destination (ie, home, nursing home, rehabilitation) by using any of the validated tools. Regardless of the approach, it is clear that standardized care processes, including enhanced clinical decision support, are necessary to ensure that physicians do not continue exposing our patients to unnecessary medications and harm after discharge.

Corresponding author: Alessandro Morandi, MD, MPH, [email protected].

Funding/support: Dr. Pandiharipande is supported by National Institutes of Health HL111111 (Bethesda, MD) and by the VA Clinical Science Research and Development Service (Washington, DC) and the National Institutes of Health AG027472 and AG035117 (Bethesda, MD).

Financial disclosures: Dr. Pratik Pandharipande has received a research grant from Hospira Inc in collaboration with the NIH.

From the Division of Allergy, Pulmonary, and Critical Care Medicine, Department of Medicine, Vanderbilt University Medical Center, Nashville, TN (Dr. Marra), Division of Anesthesiology Critical Care Medicine, Vanderbilt University Medical Center, Nashville, TN (Dr. Hayhurst, Dr. Hughes, Dr. Pandharipande), Department of Clinical and Experimental Science, University of Brescia, Brescia, Italy (Dr. Marengoni), School of Medicine and Surgery,
University of Milano-Bicocca, Milan, Italy (Dr. Bellelli), and Rehabilitation and Aged Care Unit Hospital Ancelle, Cremona, Italy (Dr. Morandi).

Abstract

• Objective: To present an overview of the phenomenon of inappropriate medication prescription in older critically ill patients and examine possible strategies of intervention.
• Methods: Review of the literature.
• Results: Polypharmacy and inappropriate prescribing of medications in older persons may lead to a significant risk of adverse drug-related events and mortality. The intensive care unit (ICU) is often the place where potentially inappropriate medications (PIMs) are first prescribed. Common PIMs at ICU discharge are antipsychotics, benzodiazepines, opioids, anticholinergic medications, antidepressants, and drugs causing orthostatic hypotension. Different classes of medications, typically intended for short-term use, are sometimes inappropriately continued after discharge from the hospital. At admission, potential risk factors for PIM are multiple morbidities, polypharmacy, frailty and cognitive decline; at discharge, a high number of pre-admission PIMs, discharge to a location other than home, discharge from a surgical service, longer length of ICU and hospital stay, and mechanical ventilation. Inappropriate prescribing in older patients can be detected through either the use of explicit criteria, drug utilization reviews, and multidisciplinary teams, including a geriatrician and/or the involvement of a clinical pharmacist.
• Conclusion: Use of PIMs may be common in critical patients, both on admission and at discharge from ICU. Therapeutic reconciliation is recommended at every transition of care (eg, at hospital or ICU admission and discharge) in order to improve appropriateness of prescription.

Key words: elderly; intensive care unit; inappropriate medications; antipsychotics.

Since older persons are often affected by multiple chronic diseases and are prescribed several medications, the quality and safety of prescribing these medications has become a global health care issue [1–4]. Polypharmacy and inappropriate prescribing of medications among the elderly is receiving significant attention in the medical literature [5,6]. Inappropriate medications in the elderly can lead to falls, cognitive impairment and delirium, poorer health status, and higher mortality [7–10]. Medications are considered potentially inappropriate when (a) the risks of treatment outweigh the benefits [11], (b) they are prescribed for periods longer than clinically indicated or without any clear indication, (c) they are not prescribed when indicated [12], and (d) they are likely to interact with other drugs and diseases. Medications included in this category are often referred to as potentially inappropriate medications (PIMs), as in some situations their use is justified; however, if the risk of harm from the drug is judged to outweigh the potential clinical benefit after an individual patient’s clinical circumstances are considered, these drugs are considered “actually inappropriate medications” (AIMs) [6].

Advancing age is associated with substantial pharmacokinetic and pharmacodynamics changes, such as altered distribution volumes and altered permeability of the blood-brain barrier, impaired liver metabolism and renal capacity, up- and down-regulation of target receptors, transmitters, and signaling pathways changes, impaired homeostasis, and increased risk of adverse drug reactions (ADRs) that lead to increased mortality and morbidity and higher health care costs [2,11,13–19]. Studies show that ADRs cause approximately 5% of hospital admissions in the general population, but the percentage rises to 10% in older persons [20].

Avoiding PIMs represents a strategy aimed at reducing drug-related mortality and morbidity. This article provides an overview of the phenomenon of inappropriate medication prescription in older critically ill patients and examines available strategies of intervention.

Inappropriate Medications at ICU Discharge

Though PIMs and AIMs may be identified at the time of hospital discharge, the intensive care unit (ICU) is often the place where these medications are first prescribed [21]. Acute hospitalization may increase PIM prescribing because of newly prescribed medications, the presence of multiple prescribers, inadequate medication reconciliation, and a lack of care coordination among inpatient providers or in the transition back to outpatient care [22)].

A known complication of critical illness and ICU stay is a significant increase in psychological symptoms, sleep cycle alterations, delirium, and cognitive impairment, which may be associated with increased prescription of specific PIMs, such as antipsychotics or benzodiazepines [6,23,24]. Despite the lack of reliable evidence supporting their use in the ICU, antipsychotic agents are used routinely in ICU patients [25] to treat a variety of conditions, such as substance withdrawal, agitation not responding to other therapies, or delirium. Results from a multicenter study of 164,996 hospitalizations across 71 academic medical centers in the US showed that 1 out of 10 ICU patients received an antipsychotic during their hospital stay [25]. Jasiak et al estimated that one-third of patients initiated on an atypical antipsychotic therapy for ICU delirium received a hospital discharge prescription for these medications, with a potential annual outpatient medication cost of approximately $2255 per patient [26].

One potential consequence of antipsychotic use in the ICU is their continuation after the transition to other clinical settings, including discharge from the hospital [27] (Table 1). 

When examining the specific factors that may contribute to a patient being discharged on an antipsychotic, authors found that the specific antipsychotic used correlated with risk of continuation [27,30], with atypical antipsychotics having a greater likelihood of being continued than haloperidol [27,30]. Possible explanation for these results could be that physicians perceive less long-term risk from atypical agents, so may be more likely to continue them on discharge [30]. However, such an approach is not always safe. Indeed, although atypical antipsychotic agents tend to cause less tardive dyskinesia, they are known to be associated with similar rates of other adverse events compared with typical agents and have been linked to an increased risk of sudden cardiac death and pneumonia in the elderly [31,32].

Other factors independently associated with being discharged on a new antipsychotic medication were the severity of the acute illness as measured with the Acute Physiology and Chronic Health Evaluation II score at ICU admission (odds ratio [OR] 1 [95% confidence interval {CI}, 1.0–1.1]) and days treated with benzodiazepines (OR 1.1 [95% CI, 1.0–1.14]) [30]. Conversely, perhaps due to different practice patterns, Tomichek et al did not find an association between benzodiazepines administration and antipsychotic prescription at discharge in post hoc analyses [27].

Another possible reason for antipsychotic continuation may reside in the indication chosen [33]. Antipsychotic agents have sedative properties and they might be used to optimize sleep during hospitalization, despite the lack of evidence to support this indication [34]. Other factors potentially contributing to continuation of antipsychotics may include persistent delirium and agitation, newly diagnosed psychiatric illness, and difficulties experienced by physicians in deprescribing [35] with improper/incomplete medication reconciliation [33].

The continuation of antipsychotic therapy increased 30-day readmission rates in patients compared to those who had therapy stopped before discharge [33]. In addition to the well-described cardiac effects (prolonged QT interval), neuroleptic malignant syndrome and extrapyramidal symptoms may also occur, and longer-term use can predispose patients to metabolic disturbances, falls, and increase the risk of death in elderly patients with dementia [31].

Benzodiazepines and sedative hypnotics are commonly used to treat insomnia and agitation in older adults despite significant risk. Benzodiazepine administration was found to be an independent risk factor for a daily transition to delirium [36,37]. Pandharipande et al reported that every unit dose of lorazepam was associated with a higher risk for daily transition to delirium (OR 1.2, 95% CI 1.1–1.4, P = 0.003) [36] in critically ill patients. A more recent analysis found for every 5 mg of midazolam administered to a patient who is awake and without delirium, there is a 4% chance that this patient will develop delirium the next ICU day [37].

Given that the risk for benzodiazepine-associated delirium is dose-dependent, clinicians should use strategies known to reduce the daily number of benzodiazepines administered that often includes the use of a sedative associated with less delirium occurrence, such as dexmedetomidine or propofol [38]. Evidence has shown that long-term use of benzodiazepines has little benefit with many risks, including an increased susceptibility to spontaneous bacterial infection [39,40] and mortality in the setting of infection [41]. Nakafero et al showed that exposure to benzodiazepines was associated with increased occurrence of both influenza-like-illness–related pneumonia and mortality. Benzodiazepine use was associated also with increased occurrence of asthma exacerbation and with increased all-cause mortality during a median follow-up of 2 years in a cohort of asthmatic patients [42] as well with an increased risk of pneumonia and long-term mortality in patients with a prior diagnosis of community- acquired pneumonia [40]. Long-term use of benzodiazepines is also associated with increased risk of falls [43–45], cognitive impairment [46–48] and disability [49,50].

Other common types of PIMs at ICU discharge were opioids, anticholinergic medications, antidepressants, and drugs causing orthostatic hypotension [6]. Of the anticholinergic AIMs, H2 blockers (61%) and promethazine (15%) were the most common [6]. Only 16% of opioids, 23% of antidepressants, and 10% of drugs causing orthostatic hypotension were found to be actually inappropriate after the patient’s circumstances were considered (eg, postoperative pain control, a new diagnosis of major depressive disorder) [6].

Inappropriate Medications at Hospital Discharge

Medications typically intended for short-term use during acute illness are sometimes continued after discharge without documented indication [51]. Poudel et al found that in 206 patients 70 years of age and older discharged to residential aged care facilities from acute care, at least 1 PIM was identified in 112 (54.4%) patients on admission and 102 (49.5%) patients on discharge [11]. Commonly prescribed PIM categories, at both admission and discharge, were central nervous system, cardiovascular, gastrointestinal, and respiratory drugs and analgesics [6,11,52,53]. Of all medications prescribed at admission (1728), 10.8% were PIMs, and at discharge, of 1759 medications, 9.6% were PIMs. Of the total 187 PIMs on admission, 56 (30%) were stopped, and 131 (70%) were continued; 32 new PIMs were introduced [11].

Morandi et al in 2011 conducted a prospective cohort study including 120 patients age ≥ 60 who were discharged after receiving care in a medical, surgical, or cardiovascular ICU for shock or respiratory failure. The percentage of patients prescribed at least 1 PIM increased from 66% at pre-admission to 85% at discharge. The number of patients with 0 PIMs dropped from 34% at preadmission to 14% at discharge, and the number of patients with 3 or more PIMS increased from 16% at preadmission to 37% at discharge. While it is possible that these drugs may be appropriate when started during an acute illness in the ICU (eg, stress ulcer prophylaxis with H2-antagonists in mechanically ventilated patients), most should have been discontinued at ICU and/or hospital discharge [21].

Inappropriate prescriptions of proton pump inhibitors (PPIs) in hospital and primary care have been widely reported [54,55]. In a study conducted by Ahrens et al in 31 primary care practices, for 58% (263/506) of patients discharged from 35 hospitals with a PPI recommendation in hospital discharge letters, an appropriate indication was missing. In 57% of these cases general practitioners followed this recommendation and continued the prescription for more than 1 month [54]. The strongest factor associated with appropriate and inappropriate continuation of PPI after discharge was PPI prescription prior to hospitalization [54]. Although PPIs are safe, they can cause adverse effects. PPI intake has been found to have a significant association with risk of community-acquired pneumonia [56,57], hip fractures [58], Clostridium difficile-associated diarrhea [55,61,62], and to reduce the therapeutic effects of bisphosphonates [59] and low-dose aspirin [60].

Unintentional medication continuation is not a problem isolated to a single drug class or disease [63]. Scales et al evaluated rates of and risk factors for potentially unintentional medication continuation following hospitalization in a population of elderly patients (≥ 66 years) [51]. They created distinct cohorts by identifying seniors not previously receiving four classes of medications typically used to treat or prevent complications of acute illness: antipsychotic medications; gastric acid suppressants (ie, histamine-2 blockers and proton pump inhibitors); benzodiazepines; and inhaled bronchodilators and steroids [51]. Prescription without documented indication occurred across all medication classes, from 12,209 patients (1.4 %) for antipsychotic medications to 34,140 patients (6.1 %) for gastric acid suppressants [51].

Several potential risk factors were considered. The relationship between multimorbidity and polypharmacy is well described in the literature, and several studies have identified a positive association between the number of drugs and the use of PIMs [64–66]. Conversely, Poudel et al did not find any association between polypharmacy and PIM use [11]. Associations were found between the use of PIMs, frailty status, and cognitive decline of patients at admission and at discharge [11], while no association was observed with age, gender, in-hospital falls, delirium, and functional decline [11,67]. Other potential risk factors of a high number of PIMs at discharge were a high number of pre-admission PIMs, discharge to a location other than home, and discharge from a surgical service [1,6,68,69]. Length of ICU stay and mechanical ventilation had a positive influence on the number of PIMs used by acutely ill older patients [11,63,69]. In the study of Scales et al, the greatest absolute risk factor across all medication groups was longer hospitalization. The increased OR for medication continuation after a hospitalization lasting more than 7 days ranged from 2.03 (95% CI 1.94–2.11) for respiratory inhalers to 6.35 (95% CI 5.91–6.82) for antipsychotic medications [51].

Inappropriate Medications: Where and How to Intervene?

Early detection of PIMs may prevent adverse drug events and improve geriatric care in older adults [13,70]. PIM prevalence can often be a useful indicator of prescribing quality [2]. Appropriate interventions and an improved quality of prescribed medications require appropriate assessment tools to decrease the number of patients discharged on these medications [71,72]. Medication reconciliation is the process of avoiding inadvertent inconsistencies within a patient’s drug regimen, which can occur during transitions in different setting of care [73]. A multidisciplinary team should be involved in the medication reconciliation at each care transition to reevaluate medications use according to the clinical conditions, cognitive/functional status and the coexistence of geriatric syndromes (eg, dementia, malnutrition, delirium, urinary incontinence, frailty) (Figure).

Criteria for the Evaluation of Inappropriate Medications Prescription

Explicit criteria derived from expert reports or published reviews are available (Table 2).

Beers criteria PIMs have been found to be associated with poor health outcomes, including confusion, falls, and mortality [7,75,78]. The STOPP (Screening Tool of Older Person’s potentially inappropriate Prescriptions) and START (Screening Tool to Alert doctors to the Right Treatment) are evidence-based sets of criteria that were developed in Ireland and updated in October 2014, including some of the new criteria for direct oral anticoagulants, drugs affecting or affected by renal system and anti-muscarinic/anticholinergic agents [79]. 

Several other sets of criteria have been published to identify PIMs, such as the FORTA (Fit for the Aged) and the PRISCUS [86] criteria. FORTA allows a disease-related evaluation revealing over-treatment and under-treatment, and medications are graded as follows: A, indispensable drug, clear-cut benefit in terms of efficacy/safety ratio proven in elderly patients for a given indication; B, drugs with proven or obvious efficacy in the elderly, but limited extent of effect or safety concerns; C, drugs with questionable efficacy/safety profiles in the elderly which should be avoided or omitted in the presence of too many drugs or side effects; D, avoid in the elderly, omit first, refer also to negative listings. Negative lists such as PRISCUS, which provide an explicit listing of drugs, independent of the diagnosis, are easy to use. On the other hand, constant updates are needed, and such lists carry the risk of an assumption that drugs not listed would be appropriate in every case [87]. Both sets of criteria have in common that they refer to long-term medication and drugs frequently used during the inpatient stay, such as antibiotics, are hardly taken into account [87].

The Medication Appropriateness Index measures overall prescribing quality through 10 separate but interrelated domains [8]. Three components are used to detect PIMs: indication, effectiveness, and duplication. However, it does not give any precise guidance in relation to specific medicines and therefore has limited application for objectively defining PIMs.

Another prescribing quality assessment tool is the Inappropriate Prescribing in the Elderly Tool (IPET), which consists of a list of the 14 most prevalent prescription errors identified from an extensive list of inappropriate prescription instances drawn up by an expert Canadian Consensus Panel [88,89].

Another approach to assess the appropriateness of drugs prescribed for older people is the use of Drug Utilization Reviews (DURs) [16]. DURs use consensus opinion by drug therapy experts to define standards or explicit criteria for a single drug, class of drugs, or group of drugs [16]. DURs typically use retrospective information from large, nonclinical administrative databases to identify problems such as dosage range, duration, therapeutic duplication, and drug interactions [90, 91]. Monane et al [92] evaluated a program designed to decrease the use of PIMs among the elderly through a computerized online DUR database. Computer alerts triggered telephone calls to physicians by pharmacists to discuss a potential problem and any therapeutic substitution options. From a total of 43,007 telepharmacy calls generated by the alerts, they were able to reach 19,368 physicians regarding 24,266 alerts (56%). The rate of change to a more appropriate therapeutic agent was 24% (5860), but ranged from 40% for long half-life benzodiazepines to 2% to 7% for drugs that theoretically were contraindicated by patients’ self-reported history [92].

Computerized Support Systems to Reduce Inappropriate Prescribing in the Elderly

Other potential solutions for reducing inappropriate medications may include continuing medical education, electronic medical records surveillance, routine clinical evaluation, and/or improved hand-off communication between discharging and accepting providers. Incorporating this assessment of medication appropriateness into the medication reconciliation process when patients are discharged or transferred out of the ICU has the potential to enhance patient safety [21,93]. A randomized controlled trial conducted by Raebel et al [94] reported the effectiveness of a computerized pharmacy alert system plus collaboration between health care professionals for decreasing potentially inappropriate medication dispensing in elderly patients. Another study showed that computer-based access to complete drug profiles and alerts about potential prescribing problems reduced the occurrence of potentially inappropriate prescriptions [95]. A summary of these studies is shown in Table 3.

Interdisciplinary Teams to Reduce Inappropriate Prescribing in the Elderly

Some studies evaluated the effect of multidisciplinary teamwork in improving inappropriate medication prescribing in the elderly (Table 4).

Pharmacists in hospitals can play a significant role in the initiation of changes to patient’s therapy and management [11] (Table 5).

Mattison et al recently emphasized that studies of PIMs should determine scenarios in which it is appropriate to prescribe PIMs, moving beyond simply labeling some medications as “potentially inappropriate,” since some PIMs are appropriately prescribed in specific clinical situations [109]. Morandi et al showed that the positive predictive value (PPV) depends on the drug type. Thus, when developing a screening system, one cannot be concerned only with high negative predictive value (NPV), one must consider PPV as well [6]. Screening tools that include medication classes with low PPV will generate false positive “flags” or warnings, which could lead to misguided clinical decisions [6]. The fact that many PIMs are not AIMs also reveals the value of using a multidisciplinary team to identify AIMs from lists of PIMs generated when discharge medication lists are screened [6,110]. Thus, a multidisciplinary team is needed to consider the clinical context to distinguish PIMs from AIMs [6]. Of course, such a team is not available in some settings; when resources are limited, knowledge of which PIMs are most likely AIMs (ie, have high PPVs) could guide the development of computer-based decision support systems or other surveillance approaches that are efficient in that particular setting [6].

Approaches for optimizing prescribing in this population mainly depend on patient needs and comorbidities and most available data are derived from randomized controlled trials involving a single drug. Such trials do not take into account the confounding effects of multiple comorbidities and patient preferences. Therefore, approaches for optimizing prescription management that are available for and validated in younger patients are not applicable to elderly subjects [3,111].

Conclusion

Clinicians should seek to identify and discontinue AIMs at 3 important transitions during a critically ill elderly patient’s hospital course: at the time of hospital or ICU admission; at ICU discharge; and at hospital discharge. The patient’s clinical situation should be reviewed at every transition points, ideally by a multidisciplinary team of clinicians, to judge the appropriateness of each PIM [6]. After the hospital discharge, patient’s medications should be then reviewed by a multidisciplinary team and/or by the primary care physician according to the final discharge destination (ie, home, nursing home, rehabilitation) by using any of the validated tools. Regardless of the approach, it is clear that standardized care processes, including enhanced clinical decision support, are necessary to ensure that physicians do not continue exposing our patients to unnecessary medications and harm after discharge.

Corresponding author: Alessandro Morandi, MD, MPH, [email protected].

Funding/support: Dr. Pandiharipande is supported by National Institutes of Health HL111111 (Bethesda, MD) and by the VA Clinical Science Research and Development Service (Washington, DC) and the National Institutes of Health AG027472 and AG035117 (Bethesda, MD).

Financial disclosures: Dr. Pratik Pandharipande has received a research grant from Hospira Inc in collaboration with the NIH.