Drug Therapy

Medication to help prevent breast cancer is not recommended for women without increased risk, but could benefit women at increased risk for the disease, according to an update from the U.S. Preventive Services Task Force.

Dr. Cecil Fox/National Cancer Institute

SOURCEs: Owens DK et al. JAMA. 2019 Sept 3. doi: 10.1001/jama.2019.11885; Nelson HD et al. JAMA. 2019 Sept 3. doi: 10.1001/jama.2019.5780.

Medication to help prevent breast cancer is not recommended for women without increased risk, but could benefit women at increased risk for the disease, according to an update from the U.S. Preventive Services Task Force.

Dr. Cecil Fox/National Cancer Institute

SOURCEs: Owens DK et al. JAMA. 2019 Sept 3. doi: 10.1001/jama.2019.11885; Nelson HD et al. JAMA. 2019 Sept 3. doi: 10.1001/jama.2019.5780.

Medication to help prevent breast cancer is not recommended for women without increased risk, but could benefit women at increased risk for the disease, according to an update from the U.S. Preventive Services Task Force.

Dr. Cecil Fox/National Cancer Institute

SOURCEs: Owens DK et al. JAMA. 2019 Sept 3. doi: 10.1001/jama.2019.11885; Nelson HD et al. JAMA. 2019 Sept 3. doi: 10.1001/jama.2019.5780.

A number of medications commonly prescribed by rheumatologists may interact with cannabidiol oil, investigators at the Imperial College Healthcare NHS Trust, London, reported.

“Patients are increasingly requesting information concerning the safety of CBD oil,” Taryn Youngstein, MD, and associates said in letter to the editor in Rheumatology, but current guidelines on the use of medical cannabis do “not address the potential interactions between CBD oil and medicines frequently used in the rheumatology clinic.”

The most important potential CBD interaction, they suggested, may be with corticosteroids. Hydrocortisone and prednisolone both inhibit the cytochrome P450 enzyme CYP3A, but CBD is a potent inhibitor of CYP3A, so “concomitant use may decrease glucocorticoid clearance and increase risk of systemic [corticosteroid] side effects,” the investigators wrote.

CBD also is known to inhibit the cytochrome P450 isozymes CYP2C9, CYP2D6, CYP2C19, CYP3A4, and CYP1A2, which, alone or in combination, are involved in the metabolization of naproxen, tramadol, amitriptyline, and tofacitinib (Xeljanz), according to a literature search done via the college’s medicine information department that also used the British National Formulary and the Natural Medicines online interaction checker.

The Janus kinase inhibitor tofacitinib is included among the possible interactions, but the other Food and Drug Administration–approved JAK inhibitor, baricitinib (Olumiant), is primarily metabolized by the kidneys and should not have significant interaction with CBD, Dr. Youngstein and associates said. Most of the conventional synthetic and biologic disease-modifying antirheumatic drugs, including methotrexate, hydroxychloroquine, adalimumab (Humira), and abatacept (Orencia), also are expected to be relatively free from CBD interactions.

This first published report on interactions between CBD oil and common rheumatology medications “highlights the importance of taking comprehensive drug histories, by asking directly about drugs considered alternative medicines and food supplements,” they said.

The investigators declared no conflicts of interest, and there was no specific funding for the study.

[email protected]

SOURCE: Wilson-Morkeh H et al. Rheumatology. 2019 July 29. doi: 10.1093/rheumatology/kez304.

A number of medications commonly prescribed by rheumatologists may interact with cannabidiol oil, investigators at the Imperial College Healthcare NHS Trust, London, reported.

“Patients are increasingly requesting information concerning the safety of CBD oil,” Taryn Youngstein, MD, and associates said in letter to the editor in Rheumatology, but current guidelines on the use of medical cannabis do “not address the potential interactions between CBD oil and medicines frequently used in the rheumatology clinic.”

The most important potential CBD interaction, they suggested, may be with corticosteroids. Hydrocortisone and prednisolone both inhibit the cytochrome P450 enzyme CYP3A, but CBD is a potent inhibitor of CYP3A, so “concomitant use may decrease glucocorticoid clearance and increase risk of systemic [corticosteroid] side effects,” the investigators wrote.

CBD also is known to inhibit the cytochrome P450 isozymes CYP2C9, CYP2D6, CYP2C19, CYP3A4, and CYP1A2, which, alone or in combination, are involved in the metabolization of naproxen, tramadol, amitriptyline, and tofacitinib (Xeljanz), according to a literature search done via the college’s medicine information department that also used the British National Formulary and the Natural Medicines online interaction checker.

The Janus kinase inhibitor tofacitinib is included among the possible interactions, but the other Food and Drug Administration–approved JAK inhibitor, baricitinib (Olumiant), is primarily metabolized by the kidneys and should not have significant interaction with CBD, Dr. Youngstein and associates said. Most of the conventional synthetic and biologic disease-modifying antirheumatic drugs, including methotrexate, hydroxychloroquine, adalimumab (Humira), and abatacept (Orencia), also are expected to be relatively free from CBD interactions.

This first published report on interactions between CBD oil and common rheumatology medications “highlights the importance of taking comprehensive drug histories, by asking directly about drugs considered alternative medicines and food supplements,” they said.

The investigators declared no conflicts of interest, and there was no specific funding for the study.

[email protected]

SOURCE: Wilson-Morkeh H et al. Rheumatology. 2019 July 29. doi: 10.1093/rheumatology/kez304.

A number of medications commonly prescribed by rheumatologists may interact with cannabidiol oil, investigators at the Imperial College Healthcare NHS Trust, London, reported.

“Patients are increasingly requesting information concerning the safety of CBD oil,” Taryn Youngstein, MD, and associates said in letter to the editor in Rheumatology, but current guidelines on the use of medical cannabis do “not address the potential interactions between CBD oil and medicines frequently used in the rheumatology clinic.”

The most important potential CBD interaction, they suggested, may be with corticosteroids. Hydrocortisone and prednisolone both inhibit the cytochrome P450 enzyme CYP3A, but CBD is a potent inhibitor of CYP3A, so “concomitant use may decrease glucocorticoid clearance and increase risk of systemic [corticosteroid] side effects,” the investigators wrote.

CBD also is known to inhibit the cytochrome P450 isozymes CYP2C9, CYP2D6, CYP2C19, CYP3A4, and CYP1A2, which, alone or in combination, are involved in the metabolization of naproxen, tramadol, amitriptyline, and tofacitinib (Xeljanz), according to a literature search done via the college’s medicine information department that also used the British National Formulary and the Natural Medicines online interaction checker.

The Janus kinase inhibitor tofacitinib is included among the possible interactions, but the other Food and Drug Administration–approved JAK inhibitor, baricitinib (Olumiant), is primarily metabolized by the kidneys and should not have significant interaction with CBD, Dr. Youngstein and associates said. Most of the conventional synthetic and biologic disease-modifying antirheumatic drugs, including methotrexate, hydroxychloroquine, adalimumab (Humira), and abatacept (Orencia), also are expected to be relatively free from CBD interactions.

This first published report on interactions between CBD oil and common rheumatology medications “highlights the importance of taking comprehensive drug histories, by asking directly about drugs considered alternative medicines and food supplements,” they said.

The investigators declared no conflicts of interest, and there was no specific funding for the study.

[email protected]

SOURCE: Wilson-Morkeh H et al. Rheumatology. 2019 July 29. doi: 10.1093/rheumatology/kez304.

There has been increasing emphasis from drug regulatory agencies on collecting robust data on multiple outcomes from clinical trials in addition to the efficacy outcomes and usual safety data. For about a decade, the US Food and Drug Administration has required the collection of cardiovascular outcome data during the testing of new antidiabetic therapies. There are several potential consequences of this mandate, in addition to our now having a better understanding of cardiovascular risk. Studies are likely to be larger, longer, and more expensive. Patient cohorts are selected with this in mind, meaning that studies may be harder to compare, and labeled indications may be more specific. And we now have several drugs carrying a specific indication to reduce cardiovascular death in patients with diabetes!

But as we dig deeper into the reduction in cardiovascular deaths seen in clinical trials with some of the sodium-glucose cotransporter 2 (SGLT2) inhibitors, several questions arise. Why is their effect on mortality and cardiovascular events (and preservation of renal function) not a consistent drug class effect? All of these inhibitors decrease glucose reabsorption and thus cause glucosuria, resulting in lower blood glucose levels with modest caloric wasting and weight loss, as well as natriuresis with mild volume depletion. But the individual drugs behaved slightly differently in clinical trials. Perhaps this was due to slightly different trial populations, or chance (despite large trial numbers), or maybe molecular differences in the drugs despite their shared effect on glucosuria, resulting in distinct “off-target” effects. Perhaps the drugs differentially affect other transporters, on cells other than renal tubular cells, altering their function. An additional known effect of the drug class is uricosuria and mild relative hypouricemia. The differential effects of these drugs on urate transport into and out of different cells that may influence components of the metabolic syndrome and cardiovascular and renal outcomes has yet to be fully explored.

But one thing that seems to be true is that the effect of empagliflozin and canagliflozin on cardiac mortality is not due to simply lowering the blood glucose. Trials like the UK Prospective Diabetes Study¹ demonstrated that better glucose control reduced microvascular complications, but they did not initially show a reduction in myocardial infarction. It took long-term follow-up studies to indicate that more intensive initial glucose control could reduce cardiovascular events. But a beneficial effect of empagliflozin (compared with placebo) on cardiovascular mortality (but interestingly not on stroke or nonfatal myocardial infarction) was seen within 3 months.² This observation suggests unique properties of this drug and some others in the class, in addition to their glucose-lowering effect. Puzzling to me, looking at several of the SGLT2 inhibitor drug studies, is why they seemed to behave differently in terms of different cardiovascular outcomes (eg, heart failure, stroke, nonfatal myocardial infarction, need for limb amputation). While some of these seemingly paradoxical outcomes have also been seen in studies of other drugs, these differences are hard for me to understand on a biological basis: they do not seem consistent with simply differential drug effects on either acute thrombosis or chronic hypoperfusion. We have much more to learn.

For the moment, I suppose we should let our practice be guided by the results of specific clinical trials, hoping that at some point head-to-head comparator drug trials will be undertaken to provide us with even better guidance in drug selection.

We can also hope that our patients with diabetes will somehow be able to afford our increasingly complex and evidence-supported pharmacotherapy, as now not only can we lower the levels of blood glucose and biomarkers of comorbidity, we can also reduce adverse cardiovascular outcomes.

There has been increasing emphasis from drug regulatory agencies on collecting robust data on multiple outcomes from clinical trials in addition to the efficacy outcomes and usual safety data. For about a decade, the US Food and Drug Administration has required the collection of cardiovascular outcome data during the testing of new antidiabetic therapies. There are several potential consequences of this mandate, in addition to our now having a better understanding of cardiovascular risk. Studies are likely to be larger, longer, and more expensive. Patient cohorts are selected with this in mind, meaning that studies may be harder to compare, and labeled indications may be more specific. And we now have several drugs carrying a specific indication to reduce cardiovascular death in patients with diabetes!

But as we dig deeper into the reduction in cardiovascular deaths seen in clinical trials with some of the sodium-glucose cotransporter 2 (SGLT2) inhibitors, several questions arise. Why is their effect on mortality and cardiovascular events (and preservation of renal function) not a consistent drug class effect? All of these inhibitors decrease glucose reabsorption and thus cause glucosuria, resulting in lower blood glucose levels with modest caloric wasting and weight loss, as well as natriuresis with mild volume depletion. But the individual drugs behaved slightly differently in clinical trials. Perhaps this was due to slightly different trial populations, or chance (despite large trial numbers), or maybe molecular differences in the drugs despite their shared effect on glucosuria, resulting in distinct “off-target” effects. Perhaps the drugs differentially affect other transporters, on cells other than renal tubular cells, altering their function. An additional known effect of the drug class is uricosuria and mild relative hypouricemia. The differential effects of these drugs on urate transport into and out of different cells that may influence components of the metabolic syndrome and cardiovascular and renal outcomes has yet to be fully explored.

But one thing that seems to be true is that the effect of empagliflozin and canagliflozin on cardiac mortality is not due to simply lowering the blood glucose. Trials like the UK Prospective Diabetes Study¹ demonstrated that better glucose control reduced microvascular complications, but they did not initially show a reduction in myocardial infarction. It took long-term follow-up studies to indicate that more intensive initial glucose control could reduce cardiovascular events. But a beneficial effect of empagliflozin (compared with placebo) on cardiovascular mortality (but interestingly not on stroke or nonfatal myocardial infarction) was seen within 3 months.² This observation suggests unique properties of this drug and some others in the class, in addition to their glucose-lowering effect. Puzzling to me, looking at several of the SGLT2 inhibitor drug studies, is why they seemed to behave differently in terms of different cardiovascular outcomes (eg, heart failure, stroke, nonfatal myocardial infarction, need for limb amputation). While some of these seemingly paradoxical outcomes have also been seen in studies of other drugs, these differences are hard for me to understand on a biological basis: they do not seem consistent with simply differential drug effects on either acute thrombosis or chronic hypoperfusion. We have much more to learn.

For the moment, I suppose we should let our practice be guided by the results of specific clinical trials, hoping that at some point head-to-head comparator drug trials will be undertaken to provide us with even better guidance in drug selection.

We can also hope that our patients with diabetes will somehow be able to afford our increasingly complex and evidence-supported pharmacotherapy, as now not only can we lower the levels of blood glucose and biomarkers of comorbidity, we can also reduce adverse cardiovascular outcomes.

There has been increasing emphasis from drug regulatory agencies on collecting robust data on multiple outcomes from clinical trials in addition to the efficacy outcomes and usual safety data. For about a decade, the US Food and Drug Administration has required the collection of cardiovascular outcome data during the testing of new antidiabetic therapies. There are several potential consequences of this mandate, in addition to our now having a better understanding of cardiovascular risk. Studies are likely to be larger, longer, and more expensive. Patient cohorts are selected with this in mind, meaning that studies may be harder to compare, and labeled indications may be more specific. And we now have several drugs carrying a specific indication to reduce cardiovascular death in patients with diabetes!

But as we dig deeper into the reduction in cardiovascular deaths seen in clinical trials with some of the sodium-glucose cotransporter 2 (SGLT2) inhibitors, several questions arise. Why is their effect on mortality and cardiovascular events (and preservation of renal function) not a consistent drug class effect? All of these inhibitors decrease glucose reabsorption and thus cause glucosuria, resulting in lower blood glucose levels with modest caloric wasting and weight loss, as well as natriuresis with mild volume depletion. But the individual drugs behaved slightly differently in clinical trials. Perhaps this was due to slightly different trial populations, or chance (despite large trial numbers), or maybe molecular differences in the drugs despite their shared effect on glucosuria, resulting in distinct “off-target” effects. Perhaps the drugs differentially affect other transporters, on cells other than renal tubular cells, altering their function. An additional known effect of the drug class is uricosuria and mild relative hypouricemia. The differential effects of these drugs on urate transport into and out of different cells that may influence components of the metabolic syndrome and cardiovascular and renal outcomes has yet to be fully explored.

But one thing that seems to be true is that the effect of empagliflozin and canagliflozin on cardiac mortality is not due to simply lowering the blood glucose. Trials like the UK Prospective Diabetes Study¹ demonstrated that better glucose control reduced microvascular complications, but they did not initially show a reduction in myocardial infarction. It took long-term follow-up studies to indicate that more intensive initial glucose control could reduce cardiovascular events. But a beneficial effect of empagliflozin (compared with placebo) on cardiovascular mortality (but interestingly not on stroke or nonfatal myocardial infarction) was seen within 3 months.² This observation suggests unique properties of this drug and some others in the class, in addition to their glucose-lowering effect. Puzzling to me, looking at several of the SGLT2 inhibitor drug studies, is why they seemed to behave differently in terms of different cardiovascular outcomes (eg, heart failure, stroke, nonfatal myocardial infarction, need for limb amputation). While some of these seemingly paradoxical outcomes have also been seen in studies of other drugs, these differences are hard for me to understand on a biological basis: they do not seem consistent with simply differential drug effects on either acute thrombosis or chronic hypoperfusion. We have much more to learn.

For the moment, I suppose we should let our practice be guided by the results of specific clinical trials, hoping that at some point head-to-head comparator drug trials will be undertaken to provide us with even better guidance in drug selection.

We can also hope that our patients with diabetes will somehow be able to afford our increasingly complex and evidence-supported pharmacotherapy, as now not only can we lower the levels of blood glucose and biomarkers of comorbidity, we can also reduce adverse cardiovascular outcomes.

An 83-year-old woman with hypertension, hypothyroidism, and a history of depression presented to the emergency department with acute shortness of breath and hypoxia. She was found to have submassive pulmonary embolism, and a heparin infusion was started immediately.

Urgent nasopharyngeal laryngoscopy revealed a hematoma at the base of her tongue that extended into the vallecula, piriform sinuses, and aryepiglottic fold, causing acute airway obstruction. These features combined with the supratherapeutic aPTT led to the diagnosis of pseudo-Ludwig angina.

DANGER OF RAPID AIRWAY COMPROMISE

Pseudo-Ludwig angina is a rare condition in which over-anticoagulation causes sublingual swelling leading to airway obstruction, whereas true Ludwig angina is an infectious regional suppuration of the neck.

Most reported cases of pseudo-Ludwig angina have resulted from overanticogulation with warfarin or warfarin-like substances (rodenticides), or from coagulopathy due to liver disease.^1–3 Early recognition is essential to avoid airway compromise.

In our patient, all anticoagulation was discontinued, and she was intubated until the hematoma began to resolve, the aPTT returned to normal, and respiratory compromise improved. At follow-up 2 months later, the sublingual hematoma had completely resolved (Figure 1). And at a 6-month follow-up visit, the pulmonary embolism had resolved, and pulmonary pressures by 2-dimensional echocardiography were normal.

An 83-year-old woman with hypertension, hypothyroidism, and a history of depression presented to the emergency department with acute shortness of breath and hypoxia. She was found to have submassive pulmonary embolism, and a heparin infusion was started immediately.

Urgent nasopharyngeal laryngoscopy revealed a hematoma at the base of her tongue that extended into the vallecula, piriform sinuses, and aryepiglottic fold, causing acute airway obstruction. These features combined with the supratherapeutic aPTT led to the diagnosis of pseudo-Ludwig angina.

DANGER OF RAPID AIRWAY COMPROMISE

Pseudo-Ludwig angina is a rare condition in which over-anticoagulation causes sublingual swelling leading to airway obstruction, whereas true Ludwig angina is an infectious regional suppuration of the neck.

Most reported cases of pseudo-Ludwig angina have resulted from overanticogulation with warfarin or warfarin-like substances (rodenticides), or from coagulopathy due to liver disease.^1–3 Early recognition is essential to avoid airway compromise.

In our patient, all anticoagulation was discontinued, and she was intubated until the hematoma began to resolve, the aPTT returned to normal, and respiratory compromise improved. At follow-up 2 months later, the sublingual hematoma had completely resolved (Figure 1). And at a 6-month follow-up visit, the pulmonary embolism had resolved, and pulmonary pressures by 2-dimensional echocardiography were normal.

An 83-year-old woman with hypertension, hypothyroidism, and a history of depression presented to the emergency department with acute shortness of breath and hypoxia. She was found to have submassive pulmonary embolism, and a heparin infusion was started immediately.

Urgent nasopharyngeal laryngoscopy revealed a hematoma at the base of her tongue that extended into the vallecula, piriform sinuses, and aryepiglottic fold, causing acute airway obstruction. These features combined with the supratherapeutic aPTT led to the diagnosis of pseudo-Ludwig angina.

DANGER OF RAPID AIRWAY COMPROMISE

Pseudo-Ludwig angina is a rare condition in which over-anticoagulation causes sublingual swelling leading to airway obstruction, whereas true Ludwig angina is an infectious regional suppuration of the neck.

Most reported cases of pseudo-Ludwig angina have resulted from overanticogulation with warfarin or warfarin-like substances (rodenticides), or from coagulopathy due to liver disease.^1–3 Early recognition is essential to avoid airway compromise.

In our patient, all anticoagulation was discontinued, and she was intubated until the hematoma began to resolve, the aPTT returned to normal, and respiratory compromise improved. At follow-up 2 months later, the sublingual hematoma had completely resolved (Figure 1). And at a 6-month follow-up visit, the pulmonary embolism had resolved, and pulmonary pressures by 2-dimensional echocardiography were normal.

A 50-year-old man with Crohn disease and psoriatic arthritis treated with infliximab and methotrexate presented to a tertiary care hospital with fever, cough, and chest discomfort. The symptoms had first appeared 2 weeks earlier, and he had gone to an urgent care center, where he was prescribed a 5-day course of azithromycin and a corticosteroid, but this had not relieved his symptoms.

Bronchoscopy revealed edematous mucosa throughout, with minimal secretion. Specimens for bacterial, acid-fast bacillus, and fungal cultures were obtained from bronchoalveolar lavage. Endobronchial lymph node biopsy with ultrasonographic guidance revealed nonnecrotizing granuloma.

Bronchoalveolar lavage cultures showed no growth, but the patient’s serum histoplasma antigen was positive at 5.99 ng/dL (reference range: none detected), leading to the diagnosis of mediastinal granuloma due to histoplasmosis with possible dissemination. His immunosuppressant drugs were stopped, and oral itraconazole was started.

At a follow-up visit 2 months later, his serum antigen level had decreased to 0.68 ng/dL, and he had no symptoms whatsoever. At a visit 1 month after that, infliximab and methotrexate were restarted because of an exacerbation of Crohn disease. His oral itraconazole treatment was to be continued for at least 12 months, given the high suspicion for disseminated histoplasmosis while on immunosuppressant therapy.

DIFFERENTIAL DIAGNOSIS OF GRANULOMATOUS LUNG DISEASE AND LYMPHADENOPATHY

The differential diagnosis of granulomatous lung disease and lymphadenopathy is broad and includes noninfectious and infectious conditions.¹

Noninfectious causes include lymphoma, sarcoidosis, inflammatory bowel disease, hypersensitivity pneumonia, side effects of drugs (eg, methotrexate, etanercept), rheumatoid nodules, vasculitis (eg, Churg-Strauss syndrome, granulomatosis with polyangiitis, primary amyloidosis, pneumoconiosis (eg, beryllium, cobalt), and Castleman disease.

There is concern that tumor necrosis factor antagonists may increase the risk of lymphoma, but a 2017 study found no evidence of this.²

Infectious conditions associated with granulomatous lung disease include tuberculosis, nontuberculous mycobacterial infection, fungal infection (eg, Cryptococcus, Coccidioides, Histoplasma, Blastomyces), brucellosis, tularemia (respiratory type B), parasitic infection (eg, Toxocara, Leishmania, Echinococcus, Schistosoma), and Whipple disease.

HISTOPLASMOSIS

Histoplasmosis, caused by infection with Histoplasma capsulatum, is the most prevalent endemic mycotic disease in the United States.³ The fungus is commonly found in the Ohio and Mississippi River valleys in the United States, and also in Central and South America and Asia.

Risk factors for histoplasmosis include living in or traveling to an endemic area, exposure to aerosolized soil that contains spores, and exposure to bats or birds and their droppings.⁴

Fewer than 5% of exposed individuals develop symptoms, which include fever, chills, headache, myalgia, anorexia, cough, and chest pain.⁵ Patients may experience symptoms shortly after exposure or may remain free of symptoms for years, with intermittent relapses of symptoms.⁶ Hilar or mediastinal lymphadenopathy is common in acute pulmonary histoplasmosis.⁷

The risk of disseminated histoplasmosis is greater in patients with reduced cell-mediated immunity, such as in human immunodeficiency virus infection, acquired immunodeficiency syndrome, solid-organ or bone marrow transplant, hematologic malignancies, immunosuppression (corticosteroids, disease-modifying antirheumatic drugs, and tumor necrosis factor antagonists), and congenital T-cell deficiencies.⁸

In a retrospective study, infliximab was the tumor necrosis factor antagonist most commonly associated with histoplasmosis.⁹ In a study of patients with rheumatoid arthritis, the disease-modifying drug most commonly associated was methotrexate.¹⁰

Isolation of H capsulatum from clinical specimens remains the gold standard for confirmation of histoplasmosis. The sensitivity of culture to detect H capsulatum depends on the clinical manifestations: it is 74% in patients with disseminated histoplasmosis, but only 42% in patients with acute pulmonary histoplasmosis.¹¹ The serum histoplasma antigen test has a sensitivity of 91.8% in disseminated histoplasmosis, 87.5% in chronic pulmonary histoplasmosis, and 83% in acute pulmonary histoplasmosis.¹²

Urine testing for histoplasma antigen has generally proven to be slightly more sensitive than serum testing in all manifestations of histoplasmosis.¹³ Combining urine and serum testing increases the likelihood of antigen detection.

TREATMENT

Asymptomatic patients with mediastinal histoplasmosis do not require treatment. (Note: in some cases, lymphadenopathy is found incidentally, and biopsy is done to rule out malignancy.)

Standard treatment of symptomatic mediastinal histoplasmosis is oral itraconazole 200 mg, 3 times daily for 3 days, followed by 200 mg orally once or twice daily for 6 to 12 weeks.¹⁴

Although stopping immunosuppressant drugs is considered the standard of care in treating histoplasmosis in immunocompromised patients, there are no guidelines on when to resume them. However, a retrospective study of 98 cases of histoplasmosis in patients on tumor necrosis factor antagonists found that resuming immunosuppressants might be safe with close monitoring during the course of antifungal therapy.⁹ The role of long-term suppressive therapy with antifungal agents in patients on chronic immunosuppressive therapy is still unknown and needs further study.

TAKE-HOME MESSAGES

• Histoplasmosis is the most prevalent endemic mycotic disease in the United States, and mediastinal lymphadenopathy is commonly seen in acute pulmonary histoplasmosis.
• Histoplasmosis should be included in the differential diagnosis of granulomatous lung disease in patients from an endemic area or with a history of travel to an endemic area.
• Immunosuppressive agents such as tumor necrosis factor antagonists and disease-modifying antirheumatic drugs can predispose to invasive fungal infection, including histoplasmosis.
• While isolation of H capsulatum from culture remains the gold standard for the diagnosis of histoplasmosis, the histoplasma antigen tests (serum and urine) is more sensitive than culture.

A 50-year-old man with Crohn disease and psoriatic arthritis treated with infliximab and methotrexate presented to a tertiary care hospital with fever, cough, and chest discomfort. The symptoms had first appeared 2 weeks earlier, and he had gone to an urgent care center, where he was prescribed a 5-day course of azithromycin and a corticosteroid, but this had not relieved his symptoms.

Bronchoscopy revealed edematous mucosa throughout, with minimal secretion. Specimens for bacterial, acid-fast bacillus, and fungal cultures were obtained from bronchoalveolar lavage. Endobronchial lymph node biopsy with ultrasonographic guidance revealed nonnecrotizing granuloma.

Bronchoalveolar lavage cultures showed no growth, but the patient’s serum histoplasma antigen was positive at 5.99 ng/dL (reference range: none detected), leading to the diagnosis of mediastinal granuloma due to histoplasmosis with possible dissemination. His immunosuppressant drugs were stopped, and oral itraconazole was started.

At a follow-up visit 2 months later, his serum antigen level had decreased to 0.68 ng/dL, and he had no symptoms whatsoever. At a visit 1 month after that, infliximab and methotrexate were restarted because of an exacerbation of Crohn disease. His oral itraconazole treatment was to be continued for at least 12 months, given the high suspicion for disseminated histoplasmosis while on immunosuppressant therapy.

DIFFERENTIAL DIAGNOSIS OF GRANULOMATOUS LUNG DISEASE AND LYMPHADENOPATHY

The differential diagnosis of granulomatous lung disease and lymphadenopathy is broad and includes noninfectious and infectious conditions.¹

Noninfectious causes include lymphoma, sarcoidosis, inflammatory bowel disease, hypersensitivity pneumonia, side effects of drugs (eg, methotrexate, etanercept), rheumatoid nodules, vasculitis (eg, Churg-Strauss syndrome, granulomatosis with polyangiitis, primary amyloidosis, pneumoconiosis (eg, beryllium, cobalt), and Castleman disease.

There is concern that tumor necrosis factor antagonists may increase the risk of lymphoma, but a 2017 study found no evidence of this.²

Infectious conditions associated with granulomatous lung disease include tuberculosis, nontuberculous mycobacterial infection, fungal infection (eg, Cryptococcus, Coccidioides, Histoplasma, Blastomyces), brucellosis, tularemia (respiratory type B), parasitic infection (eg, Toxocara, Leishmania, Echinococcus, Schistosoma), and Whipple disease.

HISTOPLASMOSIS

Histoplasmosis, caused by infection with Histoplasma capsulatum, is the most prevalent endemic mycotic disease in the United States.³ The fungus is commonly found in the Ohio and Mississippi River valleys in the United States, and also in Central and South America and Asia.

Risk factors for histoplasmosis include living in or traveling to an endemic area, exposure to aerosolized soil that contains spores, and exposure to bats or birds and their droppings.⁴

Fewer than 5% of exposed individuals develop symptoms, which include fever, chills, headache, myalgia, anorexia, cough, and chest pain.⁵ Patients may experience symptoms shortly after exposure or may remain free of symptoms for years, with intermittent relapses of symptoms.⁶ Hilar or mediastinal lymphadenopathy is common in acute pulmonary histoplasmosis.⁷

The risk of disseminated histoplasmosis is greater in patients with reduced cell-mediated immunity, such as in human immunodeficiency virus infection, acquired immunodeficiency syndrome, solid-organ or bone marrow transplant, hematologic malignancies, immunosuppression (corticosteroids, disease-modifying antirheumatic drugs, and tumor necrosis factor antagonists), and congenital T-cell deficiencies.⁸

In a retrospective study, infliximab was the tumor necrosis factor antagonist most commonly associated with histoplasmosis.⁹ In a study of patients with rheumatoid arthritis, the disease-modifying drug most commonly associated was methotrexate.¹⁰

Isolation of H capsulatum from clinical specimens remains the gold standard for confirmation of histoplasmosis. The sensitivity of culture to detect H capsulatum depends on the clinical manifestations: it is 74% in patients with disseminated histoplasmosis, but only 42% in patients with acute pulmonary histoplasmosis.¹¹ The serum histoplasma antigen test has a sensitivity of 91.8% in disseminated histoplasmosis, 87.5% in chronic pulmonary histoplasmosis, and 83% in acute pulmonary histoplasmosis.¹²

Urine testing for histoplasma antigen has generally proven to be slightly more sensitive than serum testing in all manifestations of histoplasmosis.¹³ Combining urine and serum testing increases the likelihood of antigen detection.

TREATMENT

Asymptomatic patients with mediastinal histoplasmosis do not require treatment. (Note: in some cases, lymphadenopathy is found incidentally, and biopsy is done to rule out malignancy.)

Standard treatment of symptomatic mediastinal histoplasmosis is oral itraconazole 200 mg, 3 times daily for 3 days, followed by 200 mg orally once or twice daily for 6 to 12 weeks.¹⁴

Although stopping immunosuppressant drugs is considered the standard of care in treating histoplasmosis in immunocompromised patients, there are no guidelines on when to resume them. However, a retrospective study of 98 cases of histoplasmosis in patients on tumor necrosis factor antagonists found that resuming immunosuppressants might be safe with close monitoring during the course of antifungal therapy.⁹ The role of long-term suppressive therapy with antifungal agents in patients on chronic immunosuppressive therapy is still unknown and needs further study.

TAKE-HOME MESSAGES

• Histoplasmosis is the most prevalent endemic mycotic disease in the United States, and mediastinal lymphadenopathy is commonly seen in acute pulmonary histoplasmosis.
• Histoplasmosis should be included in the differential diagnosis of granulomatous lung disease in patients from an endemic area or with a history of travel to an endemic area.
• Immunosuppressive agents such as tumor necrosis factor antagonists and disease-modifying antirheumatic drugs can predispose to invasive fungal infection, including histoplasmosis.
• While isolation of H capsulatum from culture remains the gold standard for the diagnosis of histoplasmosis, the histoplasma antigen tests (serum and urine) is more sensitive than culture.

A 50-year-old man with Crohn disease and psoriatic arthritis treated with infliximab and methotrexate presented to a tertiary care hospital with fever, cough, and chest discomfort. The symptoms had first appeared 2 weeks earlier, and he had gone to an urgent care center, where he was prescribed a 5-day course of azithromycin and a corticosteroid, but this had not relieved his symptoms.

Bronchoscopy revealed edematous mucosa throughout, with minimal secretion. Specimens for bacterial, acid-fast bacillus, and fungal cultures were obtained from bronchoalveolar lavage. Endobronchial lymph node biopsy with ultrasonographic guidance revealed nonnecrotizing granuloma.

Bronchoalveolar lavage cultures showed no growth, but the patient’s serum histoplasma antigen was positive at 5.99 ng/dL (reference range: none detected), leading to the diagnosis of mediastinal granuloma due to histoplasmosis with possible dissemination. His immunosuppressant drugs were stopped, and oral itraconazole was started.

At a follow-up visit 2 months later, his serum antigen level had decreased to 0.68 ng/dL, and he had no symptoms whatsoever. At a visit 1 month after that, infliximab and methotrexate were restarted because of an exacerbation of Crohn disease. His oral itraconazole treatment was to be continued for at least 12 months, given the high suspicion for disseminated histoplasmosis while on immunosuppressant therapy.

DIFFERENTIAL DIAGNOSIS OF GRANULOMATOUS LUNG DISEASE AND LYMPHADENOPATHY

The differential diagnosis of granulomatous lung disease and lymphadenopathy is broad and includes noninfectious and infectious conditions.¹

Noninfectious causes include lymphoma, sarcoidosis, inflammatory bowel disease, hypersensitivity pneumonia, side effects of drugs (eg, methotrexate, etanercept), rheumatoid nodules, vasculitis (eg, Churg-Strauss syndrome, granulomatosis with polyangiitis, primary amyloidosis, pneumoconiosis (eg, beryllium, cobalt), and Castleman disease.

There is concern that tumor necrosis factor antagonists may increase the risk of lymphoma, but a 2017 study found no evidence of this.²

Infectious conditions associated with granulomatous lung disease include tuberculosis, nontuberculous mycobacterial infection, fungal infection (eg, Cryptococcus, Coccidioides, Histoplasma, Blastomyces), brucellosis, tularemia (respiratory type B), parasitic infection (eg, Toxocara, Leishmania, Echinococcus, Schistosoma), and Whipple disease.

HISTOPLASMOSIS

Histoplasmosis, caused by infection with Histoplasma capsulatum, is the most prevalent endemic mycotic disease in the United States.³ The fungus is commonly found in the Ohio and Mississippi River valleys in the United States, and also in Central and South America and Asia.

Risk factors for histoplasmosis include living in or traveling to an endemic area, exposure to aerosolized soil that contains spores, and exposure to bats or birds and their droppings.⁴

Fewer than 5% of exposed individuals develop symptoms, which include fever, chills, headache, myalgia, anorexia, cough, and chest pain.⁵ Patients may experience symptoms shortly after exposure or may remain free of symptoms for years, with intermittent relapses of symptoms.⁶ Hilar or mediastinal lymphadenopathy is common in acute pulmonary histoplasmosis.⁷

The risk of disseminated histoplasmosis is greater in patients with reduced cell-mediated immunity, such as in human immunodeficiency virus infection, acquired immunodeficiency syndrome, solid-organ or bone marrow transplant, hematologic malignancies, immunosuppression (corticosteroids, disease-modifying antirheumatic drugs, and tumor necrosis factor antagonists), and congenital T-cell deficiencies.⁸

In a retrospective study, infliximab was the tumor necrosis factor antagonist most commonly associated with histoplasmosis.⁹ In a study of patients with rheumatoid arthritis, the disease-modifying drug most commonly associated was methotrexate.¹⁰

Isolation of H capsulatum from clinical specimens remains the gold standard for confirmation of histoplasmosis. The sensitivity of culture to detect H capsulatum depends on the clinical manifestations: it is 74% in patients with disseminated histoplasmosis, but only 42% in patients with acute pulmonary histoplasmosis.¹¹ The serum histoplasma antigen test has a sensitivity of 91.8% in disseminated histoplasmosis, 87.5% in chronic pulmonary histoplasmosis, and 83% in acute pulmonary histoplasmosis.¹²

Urine testing for histoplasma antigen has generally proven to be slightly more sensitive than serum testing in all manifestations of histoplasmosis.¹³ Combining urine and serum testing increases the likelihood of antigen detection.

TREATMENT

Asymptomatic patients with mediastinal histoplasmosis do not require treatment. (Note: in some cases, lymphadenopathy is found incidentally, and biopsy is done to rule out malignancy.)

Standard treatment of symptomatic mediastinal histoplasmosis is oral itraconazole 200 mg, 3 times daily for 3 days, followed by 200 mg orally once or twice daily for 6 to 12 weeks.¹⁴

Although stopping immunosuppressant drugs is considered the standard of care in treating histoplasmosis in immunocompromised patients, there are no guidelines on when to resume them. However, a retrospective study of 98 cases of histoplasmosis in patients on tumor necrosis factor antagonists found that resuming immunosuppressants might be safe with close monitoring during the course of antifungal therapy.⁹ The role of long-term suppressive therapy with antifungal agents in patients on chronic immunosuppressive therapy is still unknown and needs further study.

TAKE-HOME MESSAGES

• Histoplasmosis is the most prevalent endemic mycotic disease in the United States, and mediastinal lymphadenopathy is commonly seen in acute pulmonary histoplasmosis.
• Histoplasmosis should be included in the differential diagnosis of granulomatous lung disease in patients from an endemic area or with a history of travel to an endemic area.
• Immunosuppressive agents such as tumor necrosis factor antagonists and disease-modifying antirheumatic drugs can predispose to invasive fungal infection, including histoplasmosis.
• While isolation of H capsulatum from culture remains the gold standard for the diagnosis of histoplasmosis, the histoplasma antigen tests (serum and urine) is more sensitive than culture.

Inhibition of the renin-angiotensin-aldosterone system with angiotensin-converting enzyme (ACE) inhibitors and angiotensin II receptor blockers (ARBs) is widely used in the treatment of heart failure, hypertension, chronic kidney disease, and coronary artery disease with left ventricular dysfunction.

See related article

In this issue, Momoniat et al¹ review the benefits of ACE inhibitors and ARBs and how to manage adverse effects. I would like to add some of my own observations.

ARE ACE INHIBITORS REALLY BETTER THAN ARBs?

ACE inhibitors have been the cornerstone of treatment for patients with heart failure with reduced ejection fraction (HFrEF), in whom their use is associated with reduced rates of morbidity and death.^2,3 The use of ARBs in these patients is also associated with decreased rates of morbidity and death^4,5; however, in early comparisons, ACE inhibitors were deemed more effective in decreasing the incidence of myocardial infarction, cardiovascular death, and all-cause mortality in patients with hypertension, diabetes, and increased cardiovascular risk,⁶ and all-cause mortality in patients with HFrEF.⁷

This presumed superiority of ACE inhibitors over ARBs was thought to be a result of a greater vasodilatory effect caused by inhibiting the degradation of bradykinin and leading to increased levels of nitric oxide and vasoactive prostaglandins.⁸ Another proposed explanation was that because ARBs block angiotensin II AT1 receptors but not AT2 receptors, the increased stimulation of markedly upregulated AT2 receptors in atheromatous plaques in response to elevated serum levels of angiotensin II was deleterious.⁶ Therefore, ACE inhibitors have been recommended as first-line therapy by most guidelines, whereas ARBs are recommended as second-line therapy, when patients are unable to tolerate ACE inhibitors.

Nevertheless, the much debated differences in outcomes between ACE inhibitors and ARBs do not seem to be real and may have originated from a generational gap in the trials.

The ACE inhibitor trials were performed a decade earlier than the ARB trials. Indirect comparisons of their respective placebo-controlled trials assumed that the placebo groups used for comparison in the 2 sets of trials were similar.^9,10 Actually, the rate of cardiovascular disease decreased nearly 50% between the decades of 1990 to 2000 and 2000 to 2010, the likely result of aggressive primary and secondary prevention strategies in clinical practice, including revascularization and lipid-lowering therapy.¹⁰

In fact, a meta-regression analysis showed that the differences between ACE inhibitors and ARBs compared with placebo were due to higher event rates in the placebo groups in the ACE inhibitor trials than in the ARB trials for the outcomes of death, cardiovascular death, and myocardial infarction.¹¹ Sensitivity analyses restricted to trials published after 2000 to control for this generational gap showed similar efficacy with ACE inhibitors vs placebo and with ARBs vs placebo for all clinical outcomes.¹¹ Moreover, recent studies have shown that ARBs produce a greater decrease in cardiovascular events than ACE inhibitors, especially in patients with established cardiovascular disease.^12,13

An advantage of ARBs over ACE inhibitors is fewer adverse effects: in general, ARBs are better tolerated than ACE inhibitors.¹⁴ There are also ethnic differences in the risks of adverse reactions to these medications. African Americans have a higher risk of developing angioedema with ACE inhibitors compared with the rest of the US population, and Chinese Americans have a higher risk than whites of developing cough with ACE inhibitors.^9,15

In my medical practice, I try to make sure patients with HFrEF, hypertension, chronic kidney disease, and coronary artery disease with left ventricular dysfunction receive an inhibitor of the renin-angiotensin-aldosterone system.

Which agent?

I prefer ARBs because patients tolerate them better. I continue ACE inhibitors in patients who are already taking them without adverse effects, and I change to ARBs in patients who later become unable to tolerate ACE inhibitors.

Most antihypertensive agents increase the risk of incident gout, except for calcium channel blockers and losartan.¹⁶ Losartan is the only ARB with a uricosuric effect, although a mild one,^17,18 due to inhibition of the urate transporter 1,¹⁹ and therefore I prefer to use it instead of other ARBs or ACE inhibitors in patients who have a concomitant diagnosis of gout.

Which combinations of agents?

The addition of beta-blockers and mineralocorticoid receptor blockers to ACE inhibitors or ARBs is associated with a further decrease in the mortality risk for patients with HFrEF,^20–22 but some patients cannot tolerate these combinations or optimized doses of these medications because of worsening hypotension or increased risk of developing acute kidney injury or hyperkalemia.

In most cases, I try not to combine ACE inhibitors with ARBs. This combination may be useful in nondiabetic patients with proteinuria refractory to maximum treatment with 1 class of these agents, but it is associated with an increased risk of hyperkalemia or acute kidney injury in patients with diabetic nephropathy without improving rates of the clinical outcomes of death or cardiovascular events.²³ I prefer adding a daily low dose of a mineralocorticoid receptor blocker to an ACE inhibitor or an ARB, which is more effective in controlling refractory proteinuria.²⁴ This regimen is associated with decreased rates of mortality, cardiovascular mortality, and hospitalization for heart failure in patients with HFrEF,²² although it can lead to a higher frequency of hyperkalemia,²⁵ and patients on it require frequent dietary education and monitoring of serum potassium.

I avoid combining direct renin inhibitors with ACE inhibitors or ARBs, since this combination has been contraindicated by the US Food and Drug Administration due to lack of reduction in target-organ damage and an associated increased risk of hypotension, hyperkalemia, and kidney failure, and a slight increase in the risk of stroke or death in patients with diabetic nephropathy.²⁶

Valsartan-sacubitril

Neprilysin is a membrane-bound endopeptidase that degrades vasoactive peptides, including B-type natriuretic peptide and atrial natriuretic peptide.²⁷ The combination of the ARB valsartan and the neprilysin inhibitor sacubitril is associated with a 20% further decrease in rates of cardiovascular mortality and hospitalization and a 16% decrease in total mortality for patients with HFrEF compared with an ACE inhibitor, although there can also be more hypotension and angioedema with the combination.^27,28

Very importantly, an ACE inhibitor cannot be used together with valsartan-sacubitril due to increased risk of angioedema and cough. I change ACE inhibitors or ARBs to valsartan-sacubitril in patients with HFrEF who still have symptoms of heart failure. Interestingly, a network meta-analysis showed that the combination of valsartan-sacubitril plus a mineralocorticoid receptor blocker and a beta-blocker resulted in the greatest mortality reduction in patients with HFrEF.⁷ A word of caution, though: one can also expect an increased risk of hypotension, hyperkalemia, and kidney failure.

Monitoring

It is crucial to monitor blood pressure, serum potassium, and renal function in patients receiving ACE inhibitors, ARBs, mineralocorticoid receptor blockers, valsartan-sacubitril, or combinations of these medications, particularly in elderly patients, who are more susceptible to complications. I use a multidisciplinary approach in my clinic: a patient educator, dietitian, pharmacist, and advanced practice nurse play key roles in educating and monitoring patients for the development of possible complications from this therapy or interactions with other medications.

A recent population-based cohort study found an association of ACE inhibitor use with a 14% relative increase in lung cancer incidence after 10 years of use, compared with ARBs,²⁹ but this may not represent a large absolute risk (calculated number needed to harm of 2,970 after 10 years of ACE inhibitor use) and should be balanced against the improvement in morbidity and mortality gained with use of an ACE inhibitor. Additional studies with long-term follow-up are needed to investigate this possible association.

TAKE-HOME POINTS

• Blockade of the renin-angiotensin-aldosterone system is a cornerstone in the therapy of cardiovascular disease.
• ARBs are as effective as ACE inhibitors and have a better tolerability profile.
• ACE inhibitors cause more angioedema in African Americans and more cough in Chinese Americans than in the rest of the population.
• ACE inhibitors and most ARBs (except for losartan) increase the risk of gout.
• The combination of beta-blockers and mineralocorticoid receptor blockers with ACE inhibitors or ARBs and, lately, the use of the valsartan-sacubitril combination have been increasingly beneficial for patients with HFrEF.

Inhibition of the renin-angiotensin-aldosterone system with angiotensin-converting enzyme (ACE) inhibitors and angiotensin II receptor blockers (ARBs) is widely used in the treatment of heart failure, hypertension, chronic kidney disease, and coronary artery disease with left ventricular dysfunction.

See related article

In this issue, Momoniat et al¹ review the benefits of ACE inhibitors and ARBs and how to manage adverse effects. I would like to add some of my own observations.

ARE ACE INHIBITORS REALLY BETTER THAN ARBs?

ACE inhibitors have been the cornerstone of treatment for patients with heart failure with reduced ejection fraction (HFrEF), in whom their use is associated with reduced rates of morbidity and death.^2,3 The use of ARBs in these patients is also associated with decreased rates of morbidity and death^4,5; however, in early comparisons, ACE inhibitors were deemed more effective in decreasing the incidence of myocardial infarction, cardiovascular death, and all-cause mortality in patients with hypertension, diabetes, and increased cardiovascular risk,⁶ and all-cause mortality in patients with HFrEF.⁷

This presumed superiority of ACE inhibitors over ARBs was thought to be a result of a greater vasodilatory effect caused by inhibiting the degradation of bradykinin and leading to increased levels of nitric oxide and vasoactive prostaglandins.⁸ Another proposed explanation was that because ARBs block angiotensin II AT1 receptors but not AT2 receptors, the increased stimulation of markedly upregulated AT2 receptors in atheromatous plaques in response to elevated serum levels of angiotensin II was deleterious.⁶ Therefore, ACE inhibitors have been recommended as first-line therapy by most guidelines, whereas ARBs are recommended as second-line therapy, when patients are unable to tolerate ACE inhibitors.

Nevertheless, the much debated differences in outcomes between ACE inhibitors and ARBs do not seem to be real and may have originated from a generational gap in the trials.

The ACE inhibitor trials were performed a decade earlier than the ARB trials. Indirect comparisons of their respective placebo-controlled trials assumed that the placebo groups used for comparison in the 2 sets of trials were similar.^9,10 Actually, the rate of cardiovascular disease decreased nearly 50% between the decades of 1990 to 2000 and 2000 to 2010, the likely result of aggressive primary and secondary prevention strategies in clinical practice, including revascularization and lipid-lowering therapy.¹⁰

In fact, a meta-regression analysis showed that the differences between ACE inhibitors and ARBs compared with placebo were due to higher event rates in the placebo groups in the ACE inhibitor trials than in the ARB trials for the outcomes of death, cardiovascular death, and myocardial infarction.¹¹ Sensitivity analyses restricted to trials published after 2000 to control for this generational gap showed similar efficacy with ACE inhibitors vs placebo and with ARBs vs placebo for all clinical outcomes.¹¹ Moreover, recent studies have shown that ARBs produce a greater decrease in cardiovascular events than ACE inhibitors, especially in patients with established cardiovascular disease.^12,13

An advantage of ARBs over ACE inhibitors is fewer adverse effects: in general, ARBs are better tolerated than ACE inhibitors.¹⁴ There are also ethnic differences in the risks of adverse reactions to these medications. African Americans have a higher risk of developing angioedema with ACE inhibitors compared with the rest of the US population, and Chinese Americans have a higher risk than whites of developing cough with ACE inhibitors.^9,15

In my medical practice, I try to make sure patients with HFrEF, hypertension, chronic kidney disease, and coronary artery disease with left ventricular dysfunction receive an inhibitor of the renin-angiotensin-aldosterone system.

Which agent?

I prefer ARBs because patients tolerate them better. I continue ACE inhibitors in patients who are already taking them without adverse effects, and I change to ARBs in patients who later become unable to tolerate ACE inhibitors.

Most antihypertensive agents increase the risk of incident gout, except for calcium channel blockers and losartan.¹⁶ Losartan is the only ARB with a uricosuric effect, although a mild one,^17,18 due to inhibition of the urate transporter 1,¹⁹ and therefore I prefer to use it instead of other ARBs or ACE inhibitors in patients who have a concomitant diagnosis of gout.

Which combinations of agents?

The addition of beta-blockers and mineralocorticoid receptor blockers to ACE inhibitors or ARBs is associated with a further decrease in the mortality risk for patients with HFrEF,^20–22 but some patients cannot tolerate these combinations or optimized doses of these medications because of worsening hypotension or increased risk of developing acute kidney injury or hyperkalemia.

In most cases, I try not to combine ACE inhibitors with ARBs. This combination may be useful in nondiabetic patients with proteinuria refractory to maximum treatment with 1 class of these agents, but it is associated with an increased risk of hyperkalemia or acute kidney injury in patients with diabetic nephropathy without improving rates of the clinical outcomes of death or cardiovascular events.²³ I prefer adding a daily low dose of a mineralocorticoid receptor blocker to an ACE inhibitor or an ARB, which is more effective in controlling refractory proteinuria.²⁴ This regimen is associated with decreased rates of mortality, cardiovascular mortality, and hospitalization for heart failure in patients with HFrEF,²² although it can lead to a higher frequency of hyperkalemia,²⁵ and patients on it require frequent dietary education and monitoring of serum potassium.

I avoid combining direct renin inhibitors with ACE inhibitors or ARBs, since this combination has been contraindicated by the US Food and Drug Administration due to lack of reduction in target-organ damage and an associated increased risk of hypotension, hyperkalemia, and kidney failure, and a slight increase in the risk of stroke or death in patients with diabetic nephropathy.²⁶

Valsartan-sacubitril

Neprilysin is a membrane-bound endopeptidase that degrades vasoactive peptides, including B-type natriuretic peptide and atrial natriuretic peptide.²⁷ The combination of the ARB valsartan and the neprilysin inhibitor sacubitril is associated with a 20% further decrease in rates of cardiovascular mortality and hospitalization and a 16% decrease in total mortality for patients with HFrEF compared with an ACE inhibitor, although there can also be more hypotension and angioedema with the combination.^27,28

Very importantly, an ACE inhibitor cannot be used together with valsartan-sacubitril due to increased risk of angioedema and cough. I change ACE inhibitors or ARBs to valsartan-sacubitril in patients with HFrEF who still have symptoms of heart failure. Interestingly, a network meta-analysis showed that the combination of valsartan-sacubitril plus a mineralocorticoid receptor blocker and a beta-blocker resulted in the greatest mortality reduction in patients with HFrEF.⁷ A word of caution, though: one can also expect an increased risk of hypotension, hyperkalemia, and kidney failure.

Monitoring

It is crucial to monitor blood pressure, serum potassium, and renal function in patients receiving ACE inhibitors, ARBs, mineralocorticoid receptor blockers, valsartan-sacubitril, or combinations of these medications, particularly in elderly patients, who are more susceptible to complications. I use a multidisciplinary approach in my clinic: a patient educator, dietitian, pharmacist, and advanced practice nurse play key roles in educating and monitoring patients for the development of possible complications from this therapy or interactions with other medications.

A recent population-based cohort study found an association of ACE inhibitor use with a 14% relative increase in lung cancer incidence after 10 years of use, compared with ARBs,²⁹ but this may not represent a large absolute risk (calculated number needed to harm of 2,970 after 10 years of ACE inhibitor use) and should be balanced against the improvement in morbidity and mortality gained with use of an ACE inhibitor. Additional studies with long-term follow-up are needed to investigate this possible association.

TAKE-HOME POINTS

• Blockade of the renin-angiotensin-aldosterone system is a cornerstone in the therapy of cardiovascular disease.
• ARBs are as effective as ACE inhibitors and have a better tolerability profile.
• ACE inhibitors cause more angioedema in African Americans and more cough in Chinese Americans than in the rest of the population.
• ACE inhibitors and most ARBs (except for losartan) increase the risk of gout.
• The combination of beta-blockers and mineralocorticoid receptor blockers with ACE inhibitors or ARBs and, lately, the use of the valsartan-sacubitril combination have been increasingly beneficial for patients with HFrEF.

Inhibition of the renin-angiotensin-aldosterone system with angiotensin-converting enzyme (ACE) inhibitors and angiotensin II receptor blockers (ARBs) is widely used in the treatment of heart failure, hypertension, chronic kidney disease, and coronary artery disease with left ventricular dysfunction.

See related article

In this issue, Momoniat et al¹ review the benefits of ACE inhibitors and ARBs and how to manage adverse effects. I would like to add some of my own observations.

ARE ACE INHIBITORS REALLY BETTER THAN ARBs?

ACE inhibitors have been the cornerstone of treatment for patients with heart failure with reduced ejection fraction (HFrEF), in whom their use is associated with reduced rates of morbidity and death.^2,3 The use of ARBs in these patients is also associated with decreased rates of morbidity and death^4,5; however, in early comparisons, ACE inhibitors were deemed more effective in decreasing the incidence of myocardial infarction, cardiovascular death, and all-cause mortality in patients with hypertension, diabetes, and increased cardiovascular risk,⁶ and all-cause mortality in patients with HFrEF.⁷

This presumed superiority of ACE inhibitors over ARBs was thought to be a result of a greater vasodilatory effect caused by inhibiting the degradation of bradykinin and leading to increased levels of nitric oxide and vasoactive prostaglandins.⁸ Another proposed explanation was that because ARBs block angiotensin II AT1 receptors but not AT2 receptors, the increased stimulation of markedly upregulated AT2 receptors in atheromatous plaques in response to elevated serum levels of angiotensin II was deleterious.⁶ Therefore, ACE inhibitors have been recommended as first-line therapy by most guidelines, whereas ARBs are recommended as second-line therapy, when patients are unable to tolerate ACE inhibitors.

Nevertheless, the much debated differences in outcomes between ACE inhibitors and ARBs do not seem to be real and may have originated from a generational gap in the trials.

The ACE inhibitor trials were performed a decade earlier than the ARB trials. Indirect comparisons of their respective placebo-controlled trials assumed that the placebo groups used for comparison in the 2 sets of trials were similar.^9,10 Actually, the rate of cardiovascular disease decreased nearly 50% between the decades of 1990 to 2000 and 2000 to 2010, the likely result of aggressive primary and secondary prevention strategies in clinical practice, including revascularization and lipid-lowering therapy.¹⁰

In fact, a meta-regression analysis showed that the differences between ACE inhibitors and ARBs compared with placebo were due to higher event rates in the placebo groups in the ACE inhibitor trials than in the ARB trials for the outcomes of death, cardiovascular death, and myocardial infarction.¹¹ Sensitivity analyses restricted to trials published after 2000 to control for this generational gap showed similar efficacy with ACE inhibitors vs placebo and with ARBs vs placebo for all clinical outcomes.¹¹ Moreover, recent studies have shown that ARBs produce a greater decrease in cardiovascular events than ACE inhibitors, especially in patients with established cardiovascular disease.^12,13

An advantage of ARBs over ACE inhibitors is fewer adverse effects: in general, ARBs are better tolerated than ACE inhibitors.¹⁴ There are also ethnic differences in the risks of adverse reactions to these medications. African Americans have a higher risk of developing angioedema with ACE inhibitors compared with the rest of the US population, and Chinese Americans have a higher risk than whites of developing cough with ACE inhibitors.^9,15

In my medical practice, I try to make sure patients with HFrEF, hypertension, chronic kidney disease, and coronary artery disease with left ventricular dysfunction receive an inhibitor of the renin-angiotensin-aldosterone system.

Which agent?

I prefer ARBs because patients tolerate them better. I continue ACE inhibitors in patients who are already taking them without adverse effects, and I change to ARBs in patients who later become unable to tolerate ACE inhibitors.

Most antihypertensive agents increase the risk of incident gout, except for calcium channel blockers and losartan.¹⁶ Losartan is the only ARB with a uricosuric effect, although a mild one,^17,18 due to inhibition of the urate transporter 1,¹⁹ and therefore I prefer to use it instead of other ARBs or ACE inhibitors in patients who have a concomitant diagnosis of gout.

Which combinations of agents?

The addition of beta-blockers and mineralocorticoid receptor blockers to ACE inhibitors or ARBs is associated with a further decrease in the mortality risk for patients with HFrEF,^20–22 but some patients cannot tolerate these combinations or optimized doses of these medications because of worsening hypotension or increased risk of developing acute kidney injury or hyperkalemia.

In most cases, I try not to combine ACE inhibitors with ARBs. This combination may be useful in nondiabetic patients with proteinuria refractory to maximum treatment with 1 class of these agents, but it is associated with an increased risk of hyperkalemia or acute kidney injury in patients with diabetic nephropathy without improving rates of the clinical outcomes of death or cardiovascular events.²³ I prefer adding a daily low dose of a mineralocorticoid receptor blocker to an ACE inhibitor or an ARB, which is more effective in controlling refractory proteinuria.²⁴ This regimen is associated with decreased rates of mortality, cardiovascular mortality, and hospitalization for heart failure in patients with HFrEF,²² although it can lead to a higher frequency of hyperkalemia,²⁵ and patients on it require frequent dietary education and monitoring of serum potassium.

I avoid combining direct renin inhibitors with ACE inhibitors or ARBs, since this combination has been contraindicated by the US Food and Drug Administration due to lack of reduction in target-organ damage and an associated increased risk of hypotension, hyperkalemia, and kidney failure, and a slight increase in the risk of stroke or death in patients with diabetic nephropathy.²⁶

Valsartan-sacubitril

Neprilysin is a membrane-bound endopeptidase that degrades vasoactive peptides, including B-type natriuretic peptide and atrial natriuretic peptide.²⁷ The combination of the ARB valsartan and the neprilysin inhibitor sacubitril is associated with a 20% further decrease in rates of cardiovascular mortality and hospitalization and a 16% decrease in total mortality for patients with HFrEF compared with an ACE inhibitor, although there can also be more hypotension and angioedema with the combination.^27,28

Very importantly, an ACE inhibitor cannot be used together with valsartan-sacubitril due to increased risk of angioedema and cough. I change ACE inhibitors or ARBs to valsartan-sacubitril in patients with HFrEF who still have symptoms of heart failure. Interestingly, a network meta-analysis showed that the combination of valsartan-sacubitril plus a mineralocorticoid receptor blocker and a beta-blocker resulted in the greatest mortality reduction in patients with HFrEF.⁷ A word of caution, though: one can also expect an increased risk of hypotension, hyperkalemia, and kidney failure.

Monitoring

It is crucial to monitor blood pressure, serum potassium, and renal function in patients receiving ACE inhibitors, ARBs, mineralocorticoid receptor blockers, valsartan-sacubitril, or combinations of these medications, particularly in elderly patients, who are more susceptible to complications. I use a multidisciplinary approach in my clinic: a patient educator, dietitian, pharmacist, and advanced practice nurse play key roles in educating and monitoring patients for the development of possible complications from this therapy or interactions with other medications.

A recent population-based cohort study found an association of ACE inhibitor use with a 14% relative increase in lung cancer incidence after 10 years of use, compared with ARBs,²⁹ but this may not represent a large absolute risk (calculated number needed to harm of 2,970 after 10 years of ACE inhibitor use) and should be balanced against the improvement in morbidity and mortality gained with use of an ACE inhibitor. Additional studies with long-term follow-up are needed to investigate this possible association.

TAKE-HOME POINTS

• Blockade of the renin-angiotensin-aldosterone system is a cornerstone in the therapy of cardiovascular disease.
• ARBs are as effective as ACE inhibitors and have a better tolerability profile.
• ACE inhibitors cause more angioedema in African Americans and more cough in Chinese Americans than in the rest of the population.
• ACE inhibitors and most ARBs (except for losartan) increase the risk of gout.
• The combination of beta-blockers and mineralocorticoid receptor blockers with ACE inhibitors or ARBs and, lately, the use of the valsartan-sacubitril combination have been increasingly beneficial for patients with HFrEF.

When scientists discovered the band of hemoglobin A_1c during electrophoresis in the 1950s and 1960s and discerned it was elevated in patients with diabetes, little did they know the important role it would play in the diagnosis and treatment of diabetes in the decades to come.^1–3 Despite some caveats, a hemoglobin A_1c level of 6.5% or higher is diagnostic of diabetes across most populations, and hemoglobin A_1c goals ranging from 6.5% to 7.5% have been set for different subsets of patients depending on comorbidities, complications, risk of hypoglycemia, life expectancy, disease duration, patient preferences, and available resources.⁴

With a growing number of medications for diabetes—insulin in its various formulations and 11 other classes—hemoglobin A_1c targets can now be tailored to fit individual patient profiles. Although helping patients attain their glycemic goals is paramount, other factors should be considered when prescribing or changing a drug treatment regimen, such as cardiovascular risk reduction, weight control, avoidance of hypoglycemia, and minimizing out-of-pocket drug costs (Table 1).

CARDIOVASCULAR BENEFIT

Patients with type 2 diabetes have a 2 to 3 times higher risk of clinical atherosclerotic disease, according to 20 years of surveillance data from the Framingham cohort.⁵

Mixed results with intensive treatment

Reducing cardiovascular risk remains an important goal in diabetes management, but unfortunately, data from the long-term clinical trials aimed at reducing macrovascular risk with intensive glycemic management have been conflicting.

The United Kingdom Prospective Diabetes Study (UKPDS),⁶ which enrolled more than 4,000 patients with newly diagnosed type 2 diabetes, did not initially show a statistically significant difference in the incidence of myocardial infarction with intensive control vs conventional control, although intensive treatment did reduce the incidence of microvascular disease. However, 10 years after the trial ended, the incidence was 15% lower in the intensive-treatment group than in the conventional-treatment group, and the difference was statistically significant.⁷

A 10-year follow-up analysis of the Veterans Affairs Diabetes Trial (VADT)⁸ showed that patients who had been randomly assigned to intensive glucose control for 5.6 years had 8.6 fewer major cardiovascular events per 1,000 person-years than those assigned to standard therapy, but no improvement in median overall survival. The hemoglobin A_1c levels achieved during the trial were 6.9% and 8.4%, respectively.

In 2008, the US Food and Drug Administration (FDA)⁹ mandated that all new applications for diabetes drugs must include cardiovascular outcome studies. Therefore, we now have data on the cardiovascular benefits of two antihyperglycemic drug classes—incretins and sodium-glucose cotransporter 2 (SGLT2) inhibitors, making them attractive medications to target both cardiac and glucose concerns.

Incretins

The incretin drugs comprise 2 classes, glucagon-like peptide 1 (GLP-1) receptor agonists and dipeptidyl peptidase 4 (DPP-4) inhibitors.

Liraglutide. The Liraglutide Effect and Action in Diabetes: Evaluation of Cardiovascular Outcome Results (LEADER) trial¹⁰ compared liraglutide (a GLP-1 receptor agonist) and placebo in 9,000 patients with diabetes who either had or were at high risk of cardiovascular disease. Patients in the liraglutide group had a lower risk of the primary composite end point of death from cardiovascular causes or the first episode of nonfatal (including silent) myocardial infarction or nonfatal stroke, and a lower risk of cardiovascular death, all-cause mortality, and microvascular events than those in the placebo group. The number of patients who would need to be treated to prevent 1 event in 3 years was 66 in the analysis of the primary outcome and 98 in the analysis of death from any cause.⁹

Lixisenatide. The Evaluation of Lixisenatide in Acute Coronary Syndrome (ELIXA) trial¹¹ studied the effect of the once-daily GLP-1 receptor agonist lixisenatide on cardiovascular outcomes in 6,000 patients with type 2 diabetes with a recent coronary event. In contrast to LEADER, ELIXA did not show a cardiovascular benefit over placebo.

Exenatide. The Exenatide Study of Cardiovascular Event Lowering (EXSCEL)¹² assessed another GLP-1 extended-release drug, exenatide, in 14,000 patients, 73% of whom had established cardiovascular disease. In those patients, the drug had a modest benefit in terms of first occurrence of any component of the composite outcome of death from cardiovascular causes, nonfatal myocardial infarction, or nonfatal stroke (3-component major adverse cardiac event [MACE] outcome) in a time-to-event analysis, but the results were not statistically significant. However, the drug did significantly reduce all-cause mortality.

Semaglutide, another GLP-1 receptor agonist recently approved by the FDA, also showed benefit in patients who had cardiovascular disease or were at high risk, with significant reduction in the primary composite end point of death from cardiovascular causes or the first occurrence of nonfatal myocardial infarction (including silent) or nonfatal stroke.¹³

Dulaglutide, a newer GLP-1 drug, was associated with significantly reduced major adverse cardiovascular events (a composite end point of cardiovascular death, nonfatal myocardial infarction, or nonfatal stroke) in about 9,900 patients with diabetes, with a median follow-up of more than 5 years. Only 31% of the patients in the trial had established cardiovascular disease.¹⁴

Comment. GLP-1 drugs as a class are a good option for patients with diabetes who require weight loss, and liraglutide is now FDA-approved for reduction of cardiovascular events in patients with type 2 diabetes with established cardiovascular disease. However, other factors should be considered when prescribing these drugs: they have adverse gastrointestinal effects, the cardiovascular benefit was not a class effect, they are relatively expensive, and they must be injected. Also, they should not be prescribed concurrently with a DPP-4 inhibitor because they target the same pathway.

The other class of diabetes drugs that have shown cardiovascular benefit are the SGLT2 inhibitors.

Empagliflozin. The Empagliflozin Cardiovascular Outcome Event Trial in Type 2 Diabetes Mellitus Patients (EMPA-REG)¹⁵ compared the efficacy of empagliflozin vs placebo in 7,000 patients with diabetes and cardiovascular disease and showed relative risk reductions of 38% in death from cardiovascular death, 31% in sudden death, and 35% in heart failure hospitalizations. Empagliflozin also showed benefit in terms of progression of kidney disease and occurrence of clinically relevant renal events in this population.¹⁶

Canagliflozin also has cardiovascular outcome data and showed significant benefit when compared with placebo in the primary outcome of the composite of death from cardiovascular causes, nonfatal myocardial infarction, or nonfatal stroke, but no significant effects on cardiovascular death or all-cause mortality.¹⁷ Data from this trial also suggested a nonsignificant benefit of canagliflozin in decreasing progression of albuminuria and in the composite outcome of a sustained 40% reduction in the estimated glomerular filtration rate (eGFR), the need for renal replacement therapy, or death from renal causes.

The above data led to an additional indication from the FDA for empagliflozin—and recently, canagliflozin—to prevent cardiovascular death in patients with diabetes with established disease, but other factors should be considered when prescribing them. Patients taking canagliflozin showed a significantly increased risk of amputation. SGLT2 inhibitors as a class also increase the risk of genital infections in men and women; this is an important consideration since patients with diabetes complain of vaginal fungal and urinary tract infections even without the use of these drugs. A higher incidence of fractures with canagliflozin should also be considered when using these medications in elderly and osteoporosis-prone patients at high risk of falling.

Dapagliflozin, the third drug in this class, was associated with a lower rate of hospitalization for heart failure in about 17,160 patients—including 10,186 without atherosclerotic cardiovascular disease—who were followed for a median of 4.2 years.¹⁸ It did not show benefit for the primary safety outcome, a composite of major adverse cardiovascular events defined as cardiovascular death, myocardial infarction, or ischemic stroke.

WEIGHT MANAGEMENT

Weight loss can help overweight patients reach their hemoglobin A_1c target.

Metformin should be continued as other drugs are added because it does not induce weight gain and may help with weight loss of up to 2 kg as shown in the Diabetes Prevention Program Outcomes Study.¹⁹

GLP-1 receptor agonists and SGLT2 inhibitors help with weight loss and are good additions to a basal insulin regimen to minimize weight gain.

Liraglutide was associated with a mean weight loss of 2.3 kg over 36 months of treatment compared with placebo in the LEADER trial.¹⁰

In the Trial to Evaluate Cardiovascular and Other Long-term Outcomes With Semaglutide in Subjects With Type 2 Diabetes (SUSTAIN-6),²⁰ the mean body weight in the semaglutide group, compared with the placebo group, was 2.9 kg lower in the group receiving a lower dose and 4.3 kg lower in the group receiving a higher dose of the drug.

In a 24-week trial in 182 patients with type 2 diabetes inadequately controlled on metformin, dapagliflozin produced a statistically significant weight reduction of 2.08 kg (95% confidence interval 2.84–1.31; P < .0001) compared with placebo.²¹

Lifestyle changes aimed at weight management should be emphasized and discussed at every visit.

HYPOGLYCEMIA RISK

Hypoglycemia is a major consideration when tailoring hemoglobin A_1c targets. In the Action to Control Cardiovascular Risk (ACCORD) trial,²² severe, symptomatic hypoglycemia increased the risk of death in both the intensive and conventional treatment groups. In VADT, the occurrence of a recent severe hypoglycemic event was the strongest independent predictor of death within 90 days. Further analysis showed that even though serious hypoglycemia occurred more often in the intensive therapy group, it was associated with progression of coronary artery calcification in the standard therapy group.²³ Hence, it is imperative that tight glycemic control not be achieved at the cost of severe or recurrent hypoglycemia.

In terms of hypoglycemia, metformin is an excellent medication. The American Diabetes Association²⁴ recommends metformin as the first-line therapy for newly diagnosed diabetes. Long-term follow-up data from UKPDS showed that metformin decreased mortality and the incidence of myocardial infarction and lowered treatment costs as well as the overall risk of hypoglycemia.²⁵ When prescribed, it should be titrated to the highest dose.

The FDA²⁶ has changed the prescribing information for metformin in patients with renal impairment. Metformin should not be started if the eGFR is less than 45 mL/min/1.73 m², but it can be continued if the patient is already receiving it and the eGFR is between 30 and 45. Previously, creatinine levels were used to define renal impairment and suitability for metformin. This change has increased the number of patients who can benefit from this medication.

In patients who have a contraindication to metformin, DPP-4 inhibitors can be considered, as they carry a low risk of hypoglycemia as well. Sulfonylureas should be used with caution in these patients, especially if their oral intake is variable. When sulfonylureas were compared to the DPP-4 inhibitor sitagliptin as an add-on to metformin, the rate of hypoglycemia was 32% in the sulfonylurea group vs 5% in the sitagliptin group.²⁷

Of the sulfonylureas, glipizide and glimepiride are better than glyburide because of a comparatively lower risk of hypoglycemia and a higher selectivity for binding the K_ATP channel on the pancreatic beta cell.²⁸

Meglitinides can be a good option for patients who skip meals, but they are more expensive than other generic oral hypoglycemic agents and require multiple daily dosing.

GLP-1 analogues also have a low risk of hypoglycemia but are only available in injectable formulations. Patients must be willing and able to perform the injections themselves.²⁹

In 2010, among US residents age 65 and older, 10.9 million (about 27%) had diabetes,³⁰ and this number is projected to increase to 26.7 million by 2050.³¹ This population is prone to hypoglycemia when treated with insulin and sulfonylureas. An injury sustained by a fall induced by hypoglycemia can be life-altering. In addition, no randomized clinical trials show the effect of tight glycemic control on complications in older patients with diabetes because patients older than 80 are often excluded.

A reasonable goal suggested by the European Diabetes Working Party for Older People 2011³² and reiterated by the American Geriatrics Society in 2013³³ is a hemoglobin A_1c between 7% and 7.5% for relatively healthy older patients and 7.5% to 8% or 8.5% in frail elderly patients with diabetes.

Consider prescribing medications that carry a low risk of hypoglycemia, can be dose-adjusted for kidney function, and do not rely on manual dexterity for administration (ie, do not require patients to give themselves injections). These include metformin and DPP-4 inhibitors.

DRUG COMBINATIONS

Polypharmacy is a concern for all patients with diabetes, especially since it increases the risk of drug interactions and adverse effects, increases out-of-pocket costs, and decreases the likelihood that patients will remain adherent to their treatment regimen. The use of combination medications can reduce the number of pills or injections required, as well as copayments.

Due to concern for multiple drug-drug interactions (and also due to the progressive nature of diabetes), many people with type 2 diabetes are given insulin in lieu of pills to lower their blood glucose. In addition to premixed insulin combinations (such as combinations of neutral protamine Hagedorn and regular insulin or combinations of insulin analogues), long-acting basal insulins can now be prescribed with a GLP-1 drug in fixed-dose combinations such as insulin glargine plus lixisenatide and insulin degludec plus liraglutide.

COST CONSIDERATIONS

It is important to discuss medication cost with patients, because many newer diabetic drugs are expensive and add to the financial burden of patients already paying for multiple medications, such as antihypertensives and statins.

Metformin and sulfonylureas are less expensive alternatives for patients who cannot afford GLP-1 analogues or SGLT2 inhibitors. Even within the same drug class, the formulary-preferred drug may be cheaper than the nonformulary alternative. Thus, it is helpful to research formulary alternatives before discussing treatment regimens with patients.

When scientists discovered the band of hemoglobin A_1c during electrophoresis in the 1950s and 1960s and discerned it was elevated in patients with diabetes, little did they know the important role it would play in the diagnosis and treatment of diabetes in the decades to come.^1–3 Despite some caveats, a hemoglobin A_1c level of 6.5% or higher is diagnostic of diabetes across most populations, and hemoglobin A_1c goals ranging from 6.5% to 7.5% have been set for different subsets of patients depending on comorbidities, complications, risk of hypoglycemia, life expectancy, disease duration, patient preferences, and available resources.⁴

With a growing number of medications for diabetes—insulin in its various formulations and 11 other classes—hemoglobin A_1c targets can now be tailored to fit individual patient profiles. Although helping patients attain their glycemic goals is paramount, other factors should be considered when prescribing or changing a drug treatment regimen, such as cardiovascular risk reduction, weight control, avoidance of hypoglycemia, and minimizing out-of-pocket drug costs (Table 1).

CARDIOVASCULAR BENEFIT

Patients with type 2 diabetes have a 2 to 3 times higher risk of clinical atherosclerotic disease, according to 20 years of surveillance data from the Framingham cohort.⁵

Mixed results with intensive treatment

Reducing cardiovascular risk remains an important goal in diabetes management, but unfortunately, data from the long-term clinical trials aimed at reducing macrovascular risk with intensive glycemic management have been conflicting.

The United Kingdom Prospective Diabetes Study (UKPDS),⁶ which enrolled more than 4,000 patients with newly diagnosed type 2 diabetes, did not initially show a statistically significant difference in the incidence of myocardial infarction with intensive control vs conventional control, although intensive treatment did reduce the incidence of microvascular disease. However, 10 years after the trial ended, the incidence was 15% lower in the intensive-treatment group than in the conventional-treatment group, and the difference was statistically significant.⁷

A 10-year follow-up analysis of the Veterans Affairs Diabetes Trial (VADT)⁸ showed that patients who had been randomly assigned to intensive glucose control for 5.6 years had 8.6 fewer major cardiovascular events per 1,000 person-years than those assigned to standard therapy, but no improvement in median overall survival. The hemoglobin A_1c levels achieved during the trial were 6.9% and 8.4%, respectively.

In 2008, the US Food and Drug Administration (FDA)⁹ mandated that all new applications for diabetes drugs must include cardiovascular outcome studies. Therefore, we now have data on the cardiovascular benefits of two antihyperglycemic drug classes—incretins and sodium-glucose cotransporter 2 (SGLT2) inhibitors, making them attractive medications to target both cardiac and glucose concerns.

Incretins

The incretin drugs comprise 2 classes, glucagon-like peptide 1 (GLP-1) receptor agonists and dipeptidyl peptidase 4 (DPP-4) inhibitors.

Liraglutide. The Liraglutide Effect and Action in Diabetes: Evaluation of Cardiovascular Outcome Results (LEADER) trial¹⁰ compared liraglutide (a GLP-1 receptor agonist) and placebo in 9,000 patients with diabetes who either had or were at high risk of cardiovascular disease. Patients in the liraglutide group had a lower risk of the primary composite end point of death from cardiovascular causes or the first episode of nonfatal (including silent) myocardial infarction or nonfatal stroke, and a lower risk of cardiovascular death, all-cause mortality, and microvascular events than those in the placebo group. The number of patients who would need to be treated to prevent 1 event in 3 years was 66 in the analysis of the primary outcome and 98 in the analysis of death from any cause.⁹

Lixisenatide. The Evaluation of Lixisenatide in Acute Coronary Syndrome (ELIXA) trial¹¹ studied the effect of the once-daily GLP-1 receptor agonist lixisenatide on cardiovascular outcomes in 6,000 patients with type 2 diabetes with a recent coronary event. In contrast to LEADER, ELIXA did not show a cardiovascular benefit over placebo.

Exenatide. The Exenatide Study of Cardiovascular Event Lowering (EXSCEL)¹² assessed another GLP-1 extended-release drug, exenatide, in 14,000 patients, 73% of whom had established cardiovascular disease. In those patients, the drug had a modest benefit in terms of first occurrence of any component of the composite outcome of death from cardiovascular causes, nonfatal myocardial infarction, or nonfatal stroke (3-component major adverse cardiac event [MACE] outcome) in a time-to-event analysis, but the results were not statistically significant. However, the drug did significantly reduce all-cause mortality.

Semaglutide, another GLP-1 receptor agonist recently approved by the FDA, also showed benefit in patients who had cardiovascular disease or were at high risk, with significant reduction in the primary composite end point of death from cardiovascular causes or the first occurrence of nonfatal myocardial infarction (including silent) or nonfatal stroke.¹³

Dulaglutide, a newer GLP-1 drug, was associated with significantly reduced major adverse cardiovascular events (a composite end point of cardiovascular death, nonfatal myocardial infarction, or nonfatal stroke) in about 9,900 patients with diabetes, with a median follow-up of more than 5 years. Only 31% of the patients in the trial had established cardiovascular disease.¹⁴

Comment. GLP-1 drugs as a class are a good option for patients with diabetes who require weight loss, and liraglutide is now FDA-approved for reduction of cardiovascular events in patients with type 2 diabetes with established cardiovascular disease. However, other factors should be considered when prescribing these drugs: they have adverse gastrointestinal effects, the cardiovascular benefit was not a class effect, they are relatively expensive, and they must be injected. Also, they should not be prescribed concurrently with a DPP-4 inhibitor because they target the same pathway.

The other class of diabetes drugs that have shown cardiovascular benefit are the SGLT2 inhibitors.

Empagliflozin. The Empagliflozin Cardiovascular Outcome Event Trial in Type 2 Diabetes Mellitus Patients (EMPA-REG)¹⁵ compared the efficacy of empagliflozin vs placebo in 7,000 patients with diabetes and cardiovascular disease and showed relative risk reductions of 38% in death from cardiovascular death, 31% in sudden death, and 35% in heart failure hospitalizations. Empagliflozin also showed benefit in terms of progression of kidney disease and occurrence of clinically relevant renal events in this population.¹⁶

Canagliflozin also has cardiovascular outcome data and showed significant benefit when compared with placebo in the primary outcome of the composite of death from cardiovascular causes, nonfatal myocardial infarction, or nonfatal stroke, but no significant effects on cardiovascular death or all-cause mortality.¹⁷ Data from this trial also suggested a nonsignificant benefit of canagliflozin in decreasing progression of albuminuria and in the composite outcome of a sustained 40% reduction in the estimated glomerular filtration rate (eGFR), the need for renal replacement therapy, or death from renal causes.

The above data led to an additional indication from the FDA for empagliflozin—and recently, canagliflozin—to prevent cardiovascular death in patients with diabetes with established disease, but other factors should be considered when prescribing them. Patients taking canagliflozin showed a significantly increased risk of amputation. SGLT2 inhibitors as a class also increase the risk of genital infections in men and women; this is an important consideration since patients with diabetes complain of vaginal fungal and urinary tract infections even without the use of these drugs. A higher incidence of fractures with canagliflozin should also be considered when using these medications in elderly and osteoporosis-prone patients at high risk of falling.

Dapagliflozin, the third drug in this class, was associated with a lower rate of hospitalization for heart failure in about 17,160 patients—including 10,186 without atherosclerotic cardiovascular disease—who were followed for a median of 4.2 years.¹⁸ It did not show benefit for the primary safety outcome, a composite of major adverse cardiovascular events defined as cardiovascular death, myocardial infarction, or ischemic stroke.

WEIGHT MANAGEMENT

Weight loss can help overweight patients reach their hemoglobin A_1c target.

Metformin should be continued as other drugs are added because it does not induce weight gain and may help with weight loss of up to 2 kg as shown in the Diabetes Prevention Program Outcomes Study.¹⁹

GLP-1 receptor agonists and SGLT2 inhibitors help with weight loss and are good additions to a basal insulin regimen to minimize weight gain.

Liraglutide was associated with a mean weight loss of 2.3 kg over 36 months of treatment compared with placebo in the LEADER trial.¹⁰

In the Trial to Evaluate Cardiovascular and Other Long-term Outcomes With Semaglutide in Subjects With Type 2 Diabetes (SUSTAIN-6),²⁰ the mean body weight in the semaglutide group, compared with the placebo group, was 2.9 kg lower in the group receiving a lower dose and 4.3 kg lower in the group receiving a higher dose of the drug.

In a 24-week trial in 182 patients with type 2 diabetes inadequately controlled on metformin, dapagliflozin produced a statistically significant weight reduction of 2.08 kg (95% confidence interval 2.84–1.31; P < .0001) compared with placebo.²¹

Lifestyle changes aimed at weight management should be emphasized and discussed at every visit.

HYPOGLYCEMIA RISK

Hypoglycemia is a major consideration when tailoring hemoglobin A_1c targets. In the Action to Control Cardiovascular Risk (ACCORD) trial,²² severe, symptomatic hypoglycemia increased the risk of death in both the intensive and conventional treatment groups. In VADT, the occurrence of a recent severe hypoglycemic event was the strongest independent predictor of death within 90 days. Further analysis showed that even though serious hypoglycemia occurred more often in the intensive therapy group, it was associated with progression of coronary artery calcification in the standard therapy group.²³ Hence, it is imperative that tight glycemic control not be achieved at the cost of severe or recurrent hypoglycemia.

In terms of hypoglycemia, metformin is an excellent medication. The American Diabetes Association²⁴ recommends metformin as the first-line therapy for newly diagnosed diabetes. Long-term follow-up data from UKPDS showed that metformin decreased mortality and the incidence of myocardial infarction and lowered treatment costs as well as the overall risk of hypoglycemia.²⁵ When prescribed, it should be titrated to the highest dose.

The FDA²⁶ has changed the prescribing information for metformin in patients with renal impairment. Metformin should not be started if the eGFR is less than 45 mL/min/1.73 m², but it can be continued if the patient is already receiving it and the eGFR is between 30 and 45. Previously, creatinine levels were used to define renal impairment and suitability for metformin. This change has increased the number of patients who can benefit from this medication.

In patients who have a contraindication to metformin, DPP-4 inhibitors can be considered, as they carry a low risk of hypoglycemia as well. Sulfonylureas should be used with caution in these patients, especially if their oral intake is variable. When sulfonylureas were compared to the DPP-4 inhibitor sitagliptin as an add-on to metformin, the rate of hypoglycemia was 32% in the sulfonylurea group vs 5% in the sitagliptin group.²⁷

Of the sulfonylureas, glipizide and glimepiride are better than glyburide because of a comparatively lower risk of hypoglycemia and a higher selectivity for binding the K_ATP channel on the pancreatic beta cell.²⁸

Meglitinides can be a good option for patients who skip meals, but they are more expensive than other generic oral hypoglycemic agents and require multiple daily dosing.

GLP-1 analogues also have a low risk of hypoglycemia but are only available in injectable formulations. Patients must be willing and able to perform the injections themselves.²⁹

In 2010, among US residents age 65 and older, 10.9 million (about 27%) had diabetes,³⁰ and this number is projected to increase to 26.7 million by 2050.³¹ This population is prone to hypoglycemia when treated with insulin and sulfonylureas. An injury sustained by a fall induced by hypoglycemia can be life-altering. In addition, no randomized clinical trials show the effect of tight glycemic control on complications in older patients with diabetes because patients older than 80 are often excluded.

A reasonable goal suggested by the European Diabetes Working Party for Older People 2011³² and reiterated by the American Geriatrics Society in 2013³³ is a hemoglobin A_1c between 7% and 7.5% for relatively healthy older patients and 7.5% to 8% or 8.5% in frail elderly patients with diabetes.

Consider prescribing medications that carry a low risk of hypoglycemia, can be dose-adjusted for kidney function, and do not rely on manual dexterity for administration (ie, do not require patients to give themselves injections). These include metformin and DPP-4 inhibitors.

DRUG COMBINATIONS

Polypharmacy is a concern for all patients with diabetes, especially since it increases the risk of drug interactions and adverse effects, increases out-of-pocket costs, and decreases the likelihood that patients will remain adherent to their treatment regimen. The use of combination medications can reduce the number of pills or injections required, as well as copayments.

Due to concern for multiple drug-drug interactions (and also due to the progressive nature of diabetes), many people with type 2 diabetes are given insulin in lieu of pills to lower their blood glucose. In addition to premixed insulin combinations (such as combinations of neutral protamine Hagedorn and regular insulin or combinations of insulin analogues), long-acting basal insulins can now be prescribed with a GLP-1 drug in fixed-dose combinations such as insulin glargine plus lixisenatide and insulin degludec plus liraglutide.

COST CONSIDERATIONS

It is important to discuss medication cost with patients, because many newer diabetic drugs are expensive and add to the financial burden of patients already paying for multiple medications, such as antihypertensives and statins.

Metformin and sulfonylureas are less expensive alternatives for patients who cannot afford GLP-1 analogues or SGLT2 inhibitors. Even within the same drug class, the formulary-preferred drug may be cheaper than the nonformulary alternative. Thus, it is helpful to research formulary alternatives before discussing treatment regimens with patients.

When scientists discovered the band of hemoglobin A_1c during electrophoresis in the 1950s and 1960s and discerned it was elevated in patients with diabetes, little did they know the important role it would play in the diagnosis and treatment of diabetes in the decades to come.^1–3 Despite some caveats, a hemoglobin A_1c level of 6.5% or higher is diagnostic of diabetes across most populations, and hemoglobin A_1c goals ranging from 6.5% to 7.5% have been set for different subsets of patients depending on comorbidities, complications, risk of hypoglycemia, life expectancy, disease duration, patient preferences, and available resources.⁴

With a growing number of medications for diabetes—insulin in its various formulations and 11 other classes—hemoglobin A_1c targets can now be tailored to fit individual patient profiles. Although helping patients attain their glycemic goals is paramount, other factors should be considered when prescribing or changing a drug treatment regimen, such as cardiovascular risk reduction, weight control, avoidance of hypoglycemia, and minimizing out-of-pocket drug costs (Table 1).

CARDIOVASCULAR BENEFIT

Patients with type 2 diabetes have a 2 to 3 times higher risk of clinical atherosclerotic disease, according to 20 years of surveillance data from the Framingham cohort.⁵

Mixed results with intensive treatment

Reducing cardiovascular risk remains an important goal in diabetes management, but unfortunately, data from the long-term clinical trials aimed at reducing macrovascular risk with intensive glycemic management have been conflicting.

The United Kingdom Prospective Diabetes Study (UKPDS),⁶ which enrolled more than 4,000 patients with newly diagnosed type 2 diabetes, did not initially show a statistically significant difference in the incidence of myocardial infarction with intensive control vs conventional control, although intensive treatment did reduce the incidence of microvascular disease. However, 10 years after the trial ended, the incidence was 15% lower in the intensive-treatment group than in the conventional-treatment group, and the difference was statistically significant.⁷

A 10-year follow-up analysis of the Veterans Affairs Diabetes Trial (VADT)⁸ showed that patients who had been randomly assigned to intensive glucose control for 5.6 years had 8.6 fewer major cardiovascular events per 1,000 person-years than those assigned to standard therapy, but no improvement in median overall survival. The hemoglobin A_1c levels achieved during the trial were 6.9% and 8.4%, respectively.

In 2008, the US Food and Drug Administration (FDA)⁹ mandated that all new applications for diabetes drugs must include cardiovascular outcome studies. Therefore, we now have data on the cardiovascular benefits of two antihyperglycemic drug classes—incretins and sodium-glucose cotransporter 2 (SGLT2) inhibitors, making them attractive medications to target both cardiac and glucose concerns.

Incretins

The incretin drugs comprise 2 classes, glucagon-like peptide 1 (GLP-1) receptor agonists and dipeptidyl peptidase 4 (DPP-4) inhibitors.

Liraglutide. The Liraglutide Effect and Action in Diabetes: Evaluation of Cardiovascular Outcome Results (LEADER) trial¹⁰ compared liraglutide (a GLP-1 receptor agonist) and placebo in 9,000 patients with diabetes who either had or were at high risk of cardiovascular disease. Patients in the liraglutide group had a lower risk of the primary composite end point of death from cardiovascular causes or the first episode of nonfatal (including silent) myocardial infarction or nonfatal stroke, and a lower risk of cardiovascular death, all-cause mortality, and microvascular events than those in the placebo group. The number of patients who would need to be treated to prevent 1 event in 3 years was 66 in the analysis of the primary outcome and 98 in the analysis of death from any cause.⁹

Lixisenatide. The Evaluation of Lixisenatide in Acute Coronary Syndrome (ELIXA) trial¹¹ studied the effect of the once-daily GLP-1 receptor agonist lixisenatide on cardiovascular outcomes in 6,000 patients with type 2 diabetes with a recent coronary event. In contrast to LEADER, ELIXA did not show a cardiovascular benefit over placebo.

Exenatide. The Exenatide Study of Cardiovascular Event Lowering (EXSCEL)¹² assessed another GLP-1 extended-release drug, exenatide, in 14,000 patients, 73% of whom had established cardiovascular disease. In those patients, the drug had a modest benefit in terms of first occurrence of any component of the composite outcome of death from cardiovascular causes, nonfatal myocardial infarction, or nonfatal stroke (3-component major adverse cardiac event [MACE] outcome) in a time-to-event analysis, but the results were not statistically significant. However, the drug did significantly reduce all-cause mortality.

Semaglutide, another GLP-1 receptor agonist recently approved by the FDA, also showed benefit in patients who had cardiovascular disease or were at high risk, with significant reduction in the primary composite end point of death from cardiovascular causes or the first occurrence of nonfatal myocardial infarction (including silent) or nonfatal stroke.¹³

Dulaglutide, a newer GLP-1 drug, was associated with significantly reduced major adverse cardiovascular events (a composite end point of cardiovascular death, nonfatal myocardial infarction, or nonfatal stroke) in about 9,900 patients with diabetes, with a median follow-up of more than 5 years. Only 31% of the patients in the trial had established cardiovascular disease.¹⁴

Comment. GLP-1 drugs as a class are a good option for patients with diabetes who require weight loss, and liraglutide is now FDA-approved for reduction of cardiovascular events in patients with type 2 diabetes with established cardiovascular disease. However, other factors should be considered when prescribing these drugs: they have adverse gastrointestinal effects, the cardiovascular benefit was not a class effect, they are relatively expensive, and they must be injected. Also, they should not be prescribed concurrently with a DPP-4 inhibitor because they target the same pathway.

The other class of diabetes drugs that have shown cardiovascular benefit are the SGLT2 inhibitors.

Empagliflozin. The Empagliflozin Cardiovascular Outcome Event Trial in Type 2 Diabetes Mellitus Patients (EMPA-REG)¹⁵ compared the efficacy of empagliflozin vs placebo in 7,000 patients with diabetes and cardiovascular disease and showed relative risk reductions of 38% in death from cardiovascular death, 31% in sudden death, and 35% in heart failure hospitalizations. Empagliflozin also showed benefit in terms of progression of kidney disease and occurrence of clinically relevant renal events in this population.¹⁶

Canagliflozin also has cardiovascular outcome data and showed significant benefit when compared with placebo in the primary outcome of the composite of death from cardiovascular causes, nonfatal myocardial infarction, or nonfatal stroke, but no significant effects on cardiovascular death or all-cause mortality.¹⁷ Data from this trial also suggested a nonsignificant benefit of canagliflozin in decreasing progression of albuminuria and in the composite outcome of a sustained 40% reduction in the estimated glomerular filtration rate (eGFR), the need for renal replacement therapy, or death from renal causes.

The above data led to an additional indication from the FDA for empagliflozin—and recently, canagliflozin—to prevent cardiovascular death in patients with diabetes with established disease, but other factors should be considered when prescribing them. Patients taking canagliflozin showed a significantly increased risk of amputation. SGLT2 inhibitors as a class also increase the risk of genital infections in men and women; this is an important consideration since patients with diabetes complain of vaginal fungal and urinary tract infections even without the use of these drugs. A higher incidence of fractures with canagliflozin should also be considered when using these medications in elderly and osteoporosis-prone patients at high risk of falling.

Dapagliflozin, the third drug in this class, was associated with a lower rate of hospitalization for heart failure in about 17,160 patients—including 10,186 without atherosclerotic cardiovascular disease—who were followed for a median of 4.2 years.¹⁸ It did not show benefit for the primary safety outcome, a composite of major adverse cardiovascular events defined as cardiovascular death, myocardial infarction, or ischemic stroke.

WEIGHT MANAGEMENT

Weight loss can help overweight patients reach their hemoglobin A_1c target.

Metformin should be continued as other drugs are added because it does not induce weight gain and may help with weight loss of up to 2 kg as shown in the Diabetes Prevention Program Outcomes Study.¹⁹

GLP-1 receptor agonists and SGLT2 inhibitors help with weight loss and are good additions to a basal insulin regimen to minimize weight gain.

Liraglutide was associated with a mean weight loss of 2.3 kg over 36 months of treatment compared with placebo in the LEADER trial.¹⁰

In the Trial to Evaluate Cardiovascular and Other Long-term Outcomes With Semaglutide in Subjects With Type 2 Diabetes (SUSTAIN-6),²⁰ the mean body weight in the semaglutide group, compared with the placebo group, was 2.9 kg lower in the group receiving a lower dose and 4.3 kg lower in the group receiving a higher dose of the drug.

In a 24-week trial in 182 patients with type 2 diabetes inadequately controlled on metformin, dapagliflozin produced a statistically significant weight reduction of 2.08 kg (95% confidence interval 2.84–1.31; P < .0001) compared with placebo.²¹

Lifestyle changes aimed at weight management should be emphasized and discussed at every visit.

HYPOGLYCEMIA RISK

Hypoglycemia is a major consideration when tailoring hemoglobin A_1c targets. In the Action to Control Cardiovascular Risk (ACCORD) trial,²² severe, symptomatic hypoglycemia increased the risk of death in both the intensive and conventional treatment groups. In VADT, the occurrence of a recent severe hypoglycemic event was the strongest independent predictor of death within 90 days. Further analysis showed that even though serious hypoglycemia occurred more often in the intensive therapy group, it was associated with progression of coronary artery calcification in the standard therapy group.²³ Hence, it is imperative that tight glycemic control not be achieved at the cost of severe or recurrent hypoglycemia.

In terms of hypoglycemia, metformin is an excellent medication. The American Diabetes Association²⁴ recommends metformin as the first-line therapy for newly diagnosed diabetes. Long-term follow-up data from UKPDS showed that metformin decreased mortality and the incidence of myocardial infarction and lowered treatment costs as well as the overall risk of hypoglycemia.²⁵ When prescribed, it should be titrated to the highest dose.

The FDA²⁶ has changed the prescribing information for metformin in patients with renal impairment. Metformin should not be started if the eGFR is less than 45 mL/min/1.73 m², but it can be continued if the patient is already receiving it and the eGFR is between 30 and 45. Previously, creatinine levels were used to define renal impairment and suitability for metformin. This change has increased the number of patients who can benefit from this medication.

In patients who have a contraindication to metformin, DPP-4 inhibitors can be considered, as they carry a low risk of hypoglycemia as well. Sulfonylureas should be used with caution in these patients, especially if their oral intake is variable. When sulfonylureas were compared to the DPP-4 inhibitor sitagliptin as an add-on to metformin, the rate of hypoglycemia was 32% in the sulfonylurea group vs 5% in the sitagliptin group.²⁷

Of the sulfonylureas, glipizide and glimepiride are better than glyburide because of a comparatively lower risk of hypoglycemia and a higher selectivity for binding the K_ATP channel on the pancreatic beta cell.²⁸

Meglitinides can be a good option for patients who skip meals, but they are more expensive than other generic oral hypoglycemic agents and require multiple daily dosing.

GLP-1 analogues also have a low risk of hypoglycemia but are only available in injectable formulations. Patients must be willing and able to perform the injections themselves.²⁹

In 2010, among US residents age 65 and older, 10.9 million (about 27%) had diabetes,³⁰ and this number is projected to increase to 26.7 million by 2050.³¹ This population is prone to hypoglycemia when treated with insulin and sulfonylureas. An injury sustained by a fall induced by hypoglycemia can be life-altering. In addition, no randomized clinical trials show the effect of tight glycemic control on complications in older patients with diabetes because patients older than 80 are often excluded.

A reasonable goal suggested by the European Diabetes Working Party for Older People 2011³² and reiterated by the American Geriatrics Society in 2013³³ is a hemoglobin A_1c between 7% and 7.5% for relatively healthy older patients and 7.5% to 8% or 8.5% in frail elderly patients with diabetes.

Consider prescribing medications that carry a low risk of hypoglycemia, can be dose-adjusted for kidney function, and do not rely on manual dexterity for administration (ie, do not require patients to give themselves injections). These include metformin and DPP-4 inhibitors.

DRUG COMBINATIONS

Polypharmacy is a concern for all patients with diabetes, especially since it increases the risk of drug interactions and adverse effects, increases out-of-pocket costs, and decreases the likelihood that patients will remain adherent to their treatment regimen. The use of combination medications can reduce the number of pills or injections required, as well as copayments.

Due to concern for multiple drug-drug interactions (and also due to the progressive nature of diabetes), many people with type 2 diabetes are given insulin in lieu of pills to lower their blood glucose. In addition to premixed insulin combinations (such as combinations of neutral protamine Hagedorn and regular insulin or combinations of insulin analogues), long-acting basal insulins can now be prescribed with a GLP-1 drug in fixed-dose combinations such as insulin glargine plus lixisenatide and insulin degludec plus liraglutide.

COST CONSIDERATIONS

It is important to discuss medication cost with patients, because many newer diabetic drugs are expensive and add to the financial burden of patients already paying for multiple medications, such as antihypertensives and statins.

Metformin and sulfonylureas are less expensive alternatives for patients who cannot afford GLP-1 analogues or SGLT2 inhibitors. Even within the same drug class, the formulary-preferred drug may be cheaper than the nonformulary alternative. Thus, it is helpful to research formulary alternatives before discussing treatment regimens with patients.

A highly active, water- and alcohol-soluble, basic pressor substance is formed when renin and renin-activator interact, for which we suggest the name “angiotonin.”

—Irvine H. Page and O.M. Helmer, 1940.¹

The renin-angiotensin-aldosterone system regulates salt and, in part, water homeostasis, and therefore blood pressure and fluid balance through its actions on the heart, kidneys, and blood vessels.² Drugs that target this system—angiotensin-converting enzyme (ACE) inhibitors and angiotensin II receptor blockers (ARBs)—are used primarily to treat hypertension and also to treat chronic kidney disease and heart failure with reduced ejection fraction.

See related editorial

Controlling blood pressure is important, as hypertension increases the risk of myocardial infarction, cerebrovascular events, and progression of chronic kidney disease, which itself is a risk factor for cardiovascular disease. However, the benefit of these drugs is only partly due to their effect on blood pressure. They also reduce proteinuria, which is a graded risk factor for progression of kidney disease as well as morbidity and death from vascular events.³

Despite the benefits of ACE inhibitors and ARBs, concern about their adverse effects—especially hyperkalemia and a decline in renal function—has led to their underuse in patients likely to derive the greatest benefit.³

ACE INHIBITORS AND ARBs

ACE inhibitors, as their name indicates, inhibit conversion of angiotensin I to angiotensin II by ACE, resulting in vasodilation of the efferent arteriole and a drop in blood pressure. Inhibition of ACE, a kininase, also results in a rise in kinins. One of these, bradykinin, is associated with some of the side effects of this class of drugs such as cough, which affects 5% to 20% of patients.⁴ Elevation of bradykinin is also believed to account for ACE inhibitor-induced angioedema, an uncommon but potentially serious side effect. Kinins are also associated with desirable effects such as lowering blood pressure, increasing insulin sensitivity, and dilating blood vessels.

ARBs were developed as an alternative for patients unable to tolerate the adverse effects of ACE inhibitors. While ACE inhibitors reduce the activity of angiotensin II at both the AT1 and AT2 receptors, ARBs block only the AT1 receptors, thereby inhibiting their vasoconstricting activity on smooth muscle. ARBs also raise the levels of renin, angiotensin I, and angiotensin II as a result of feedback inhibition. Angiotensin II is associated with release of inflammatory mediators such as tumor necrosis factor alpha, cytokines, and chemokines, the consequences of which are also inhibited by ARBs, further preventing renal fibrosis and scarring from chronic inflammation.³

What is the evidence supporting the use of ACE inhibitors and ARBs?

ACE inhibitors and ARBs, used singly, reduce blood pressure and proteinuria, slow progression of kidney disease, and improve outcomes in patients who have heart failure, diabetes mellitus, or a history of myocardial infarction.^5–11 

While dual blockade with the combination of an ACE inhibitor and an ARB lowers blood pressure and proteinuria to a greater degree than monotherapy, dual blockade has been associated with higher rates of complications, including hyperkalemia.^12–17

RISK FACTORS FOR HYPERKALEMIA

ACE inhibitors and ARBs raise potassium, especially when used in combination. Other risk factors for hyperkalemia include the following—and note that some of them are also indications for ACE inhibitors and ARBs:

Renal insufficiency. The kidneys are responsible for over 90% of potassium removal in healthy individuals,^18,19 and the lower the GFR, the higher the risk of hyperkalemia.^3,20,21

Heart failure

Diabetes mellitus^6,21–23

Endogenous potassium load due to hemolysis, rhabdomyolysis, insulin deficiency, lactic acidosis, or gastrointestinal bleeding

Exogenous potassium load due to dietary consumption or blood products

Other medications, eg, sacubitril-valsartan, aldosterone antagonists, mineralocorticoid receptor antagonists, potassium-sparing diuretics, beta-adrenergic antagonists, nonsteroidal anti-inflammatory drugs, heparin, cyclosporine, trimethoprim, digoxin

Hypertension

Hypoaldosteronism (including type 4 renal tubular acidosis)

Addison disease

Advanced age

Lower body mass index.

Both hypokalemia and hyperkalemia are associated with a higher risk of death,^20,21,24  but in patients with heart failure, the survival benefit from ACE inhibitors, ARBs, and mineralocorticoid receptor antagonists outweighs the risk of hyperkalemia.^25–27 Weir and Rolfe²⁸ concluded that patients with heart failure and chronic kidney disease are at greatest risk of hyperkalemia from renin-angiotensin-aldosterone system inhibition, but the increases in potassium levels are small (about 0.1 to 0.3 mmol/L) and unlikely to be clinically significant.

Hyperkalemia tends to recur. Einhorn et al²⁰ found that nearly half of patients with chronic kidney disease who had an episode of hyperkalemia had 1 or more recurrent episodes within a year.

Another concern about using ACE inhibitors and ARBs, especially in patients with chronic kidney disease, is that the serum creatinine level tends to rise when starting these drugs,²⁹ although several studies have shown that an acute rise in creatinine may demonstrate that the drug is actually protecting the kidney.^30,31 Hirsch³² described this phenomenon as “prerenal success,” proposing that the decline in GFR is hemodynamic, secondary to a fall in intraglomerular pressure as a result of efferent vasodilation, and therefore should not be reversed.

Schmidt et al,^33,34 in a study in 122,363 patients who began ACE inhibitor or ARB therapy, found that cardiorenal outcomes were worse, with higher rates of end-stage renal disease, myocardial infarction, heart failure, and death, in those in whom creatinine rose by 30% or more since starting treatment. This trend was also seen, to a lesser degree, in those with a smaller increase in creatinine, suggesting that even this group of patients should receive close monitoring.

Whether renin-angiotensin-aldosterone system inhibitors provide a benefit in advanced progressive chronic kidney disease remains unclear.^35–37  The Angiotensin Converting Enzyme Inhibitor (ACEi)/Angiotensin Receptor Blocker (ARB) Withdrawal in Advanced Renal Disease trial (STOP-ACEi),³⁸ currently under way, will provide valuable data to help close this gap in our knowledge. This open-label randomized controlled trial is testing the hypothesis that stopping ACE inhibitor or ARB treatment, or a combination of both, compared with continuing these treatments, will improve or stabilize renal function in patients with progressive stage 4 or 5 chronic kidney disease.

NEED FOR MONITORING

Taken together, the above data suggest close and regular monitoring is required in patients receiving these drugs. However, monitoring tends to be lax.^34,37,39 A 2017 study of adherence to the guidelines for monitoring serum creatinine and potassium after starting an ACE inhibitor or ARB and subsequent discontinuation found that fewer than 10% of patients had follow-up within the recommended 2 weeks after starting these drugs.³⁴ Most patients with a creatinine rise of 30% or more or a potassium level higher than 6.0 mmol/L continued treatment. There was also no evidence of increased monitoring in those deemed at higher risk of these complications.

WHAT DO THE GUIDELINES SUGGEST?

ACE inhibitors and ARBs in chronic kidney disease and hypertension

Target blood pressures vary in guidelines from different organizations.^4,40–45 The 2017 joint guidelines of the American College of Cardiology and American Heart Association (ACC/AHA)⁴⁰ recommend a target blood pressure of 130/80 mm Hg or less in all patients irrespective of the level of proteinuria and whether they have diabetes mellitus, based on several studies.^46–48 In the elderly, other factors such as the risk of hypotension and falls must be taken into consideration in establishing the most appropriate blood pressure target.

In general, a renin-angiotensin-aldosterone system inhibitor is recommended if the patient has diabetes, stage 1, 2, or 3 chronic kidney disease, or proteinuria. For example, the guidelines recommend a renin-angiotensin-aldosterone system inhibitor in diabetic patients with albuminuria.

None of the guidelines recommend routine use of combination therapy.

ACE inhibitors and ARBs in heart failure

The 2017 ACC/AHA and Heart Failure Society of America (HFSA) guidelines for heart failure⁴⁹ recommend an ACE inhibitor or ARB for patients with stage C (symptomatic) heart failure with reduced ejection fraction, in view of the known cardiovascular morbidity and mortality benefits.

The European Society of Cardiology⁵⁰ recommends ACE inhibitors for patients with symptomatic heart failure with reduced ejection fraction, as well as those with asymptomatic left ventricular systolic dysfunction. In patients with stable coronary artery disease, an ACE inhibitor should be considered even with normal left ventricular function.

ARBs should be used as alternatives in those unable to tolerate ACE inhibitors.

Combination therapy should be avoided due to the increased risk of renal impairment and hyperkalemia but may be considered in patients with heart failure and reduced ejection fraction in whom other treatments are unsuitable. These include patients on beta-blockers who cannot tolerate mineralocorticoid receptor antagonists such as spironolactone. Combination therapy should be done only under strict supervision.⁵⁰

Close monitoring of serum potassium is recommended during ACE inhibitor or ARB use. Those at greatest risk of hyperkalemia include elderly patients, those taking other medications associated with hyperkalemia, and diabetic patients, because of their higher risk of renovascular disease.

Caution is advised when starting ACE inhibitor or ARB therapy in these high-risk groups as well as in patients with potassium levels higher than 5.0 mmol/L at baseline, at high risk of prerenal acute kidney injury, with known renal insufficiency, and with previous deterioration in renal function on these medications.^3,41,51

Before starting therapy, ensure that patients are volume-replete and measure baseline serum electrolytes and creatinine.^41,51

The ACC/AHA and HFSA recommend starting at a low dose and titrating upward slowly. If maximal doses are not tolerated, then a lower dose should be maintained.⁴⁹ The European Society of Cardiology guidelines⁵² suggest increasing the dose at no less than every 2 weeks unless in an inpatient setting. Blood testing should be done 7 to 14 days after starting therapy, after any titration in dosage, and every 4 months thereafter.⁵³

The guidelines generally agree that a rise in creatinine of up to 30% and a fall in eGFR of up to 25% is acceptable, with the need for regular monitoring, particularly in high-risk groups.^{40–42,51,52}

What if serum potassium or creatinine rises during treatment?

If hyperkalemia arises or renal function declines by a significant amount, one should first address contributing factors. If no improvement is seen, then the dose of the ACE inhibitor or ARB should be reduced by 50% and blood work repeated in 1 to 2 weeks. If the laboratory values do not return to an acceptable level, reducing the dose further or stopping the drug is advised.

Give dietary advice to all patients with chronic kidney disease being considered for a renin-angiotensin-aldosterone system inhibitor or for an increase in dose with a potassium level higher than 4.5 mmol/L. A low-potassium diet should aim for potassium intake of less than 50 or 75 mmol/day and sodium intake of less than 60 mmol/day for hypertensive patients with chronic kidney disease.

Review the patient’s medications if the baseline potassium level is higher than 5.0 mmol/L. Consider stopping potassium-sparing agents, digoxin, trimethoprim, and nonsteroidal anti-inflammatory drugs. Also think about starting a non–potassium-sparing diuretic as well as sodium bicarbonate to reduce potassium levels. Blood work should be repeated within 2 weeks after these changes.

Do not start a renin-angiotensin-aldosterone system inhibitor, or do not increase the dose, if the potassium level is elevated until measures have been taken to reduce the degree of hyperkalemia.⁵¹

In renal transplant recipients, renin-angiotensin-aldosterone system inhibitors are often preferred to manage hypertension in those who have proteinuria or cardiovascular disease. However, the risk of hyperkalemia is also greater with concomitant use of immunosuppressive drugs such as tacrolimus and cyclosporine. Management of complications should be approached according to guidelines discussed above.⁵¹

Monitor renal function, potassium. The National Institute for Health and Care Excellence guideline⁵⁴ advocates that baseline renal function testing should be followed by repeat blood testing 1 to 2 weeks after starting renin-angiotensin-aldosterone system inhibitors in patients with ischemic heart disease. The advice is similar when starting therapy in patients with chronic heart failure, emphasizing the need to monitor after each dose increment and to use clinical judgment when deciding to start treatment. The AHA advises caution in patients with renal insufficiency or a potassium level above 5.0 mmol/L.⁴⁹

Sick day rules. The National Institute for Health and Care Excellence encourages discussing “sick day rules” with patients starting renin-angiotensin-aldosterone system inhibitors. This means patients should be advised to temporarily stop taking nephrotoxic medications, including over-the-counter nonsteroidal anti-inflammatory drugs, in any potential state of illness or dehydration, such as diarrhea and vomiting. There is, however, little evidence that this advice can actually reduce the incidence of acute kidney injury.^55,56

OUR RECOMMENDATIONS

Our advice for managing patients receiving ACE inhibitors or ARBs is summarized in Table 1.

A highly active, water- and alcohol-soluble, basic pressor substance is formed when renin and renin-activator interact, for which we suggest the name “angiotonin.”

—Irvine H. Page and O.M. Helmer, 1940.¹

The renin-angiotensin-aldosterone system regulates salt and, in part, water homeostasis, and therefore blood pressure and fluid balance through its actions on the heart, kidneys, and blood vessels.² Drugs that target this system—angiotensin-converting enzyme (ACE) inhibitors and angiotensin II receptor blockers (ARBs)—are used primarily to treat hypertension and also to treat chronic kidney disease and heart failure with reduced ejection fraction.

See related editorial

Controlling blood pressure is important, as hypertension increases the risk of myocardial infarction, cerebrovascular events, and progression of chronic kidney disease, which itself is a risk factor for cardiovascular disease. However, the benefit of these drugs is only partly due to their effect on blood pressure. They also reduce proteinuria, which is a graded risk factor for progression of kidney disease as well as morbidity and death from vascular events.³

Despite the benefits of ACE inhibitors and ARBs, concern about their adverse effects—especially hyperkalemia and a decline in renal function—has led to their underuse in patients likely to derive the greatest benefit.³

ACE INHIBITORS AND ARBs

ACE inhibitors, as their name indicates, inhibit conversion of angiotensin I to angiotensin II by ACE, resulting in vasodilation of the efferent arteriole and a drop in blood pressure. Inhibition of ACE, a kininase, also results in a rise in kinins. One of these, bradykinin, is associated with some of the side effects of this class of drugs such as cough, which affects 5% to 20% of patients.⁴ Elevation of bradykinin is also believed to account for ACE inhibitor-induced angioedema, an uncommon but potentially serious side effect. Kinins are also associated with desirable effects such as lowering blood pressure, increasing insulin sensitivity, and dilating blood vessels.

ARBs were developed as an alternative for patients unable to tolerate the adverse effects of ACE inhibitors. While ACE inhibitors reduce the activity of angiotensin II at both the AT1 and AT2 receptors, ARBs block only the AT1 receptors, thereby inhibiting their vasoconstricting activity on smooth muscle. ARBs also raise the levels of renin, angiotensin I, and angiotensin II as a result of feedback inhibition. Angiotensin II is associated with release of inflammatory mediators such as tumor necrosis factor alpha, cytokines, and chemokines, the consequences of which are also inhibited by ARBs, further preventing renal fibrosis and scarring from chronic inflammation.³

What is the evidence supporting the use of ACE inhibitors and ARBs?

ACE inhibitors and ARBs, used singly, reduce blood pressure and proteinuria, slow progression of kidney disease, and improve outcomes in patients who have heart failure, diabetes mellitus, or a history of myocardial infarction.^5–11 

While dual blockade with the combination of an ACE inhibitor and an ARB lowers blood pressure and proteinuria to a greater degree than monotherapy, dual blockade has been associated with higher rates of complications, including hyperkalemia.^12–17

RISK FACTORS FOR HYPERKALEMIA

ACE inhibitors and ARBs raise potassium, especially when used in combination. Other risk factors for hyperkalemia include the following—and note that some of them are also indications for ACE inhibitors and ARBs:

Renal insufficiency. The kidneys are responsible for over 90% of potassium removal in healthy individuals,^18,19 and the lower the GFR, the higher the risk of hyperkalemia.^3,20,21

Heart failure

Diabetes mellitus^6,21–23

Endogenous potassium load due to hemolysis, rhabdomyolysis, insulin deficiency, lactic acidosis, or gastrointestinal bleeding

Exogenous potassium load due to dietary consumption or blood products

Other medications, eg, sacubitril-valsartan, aldosterone antagonists, mineralocorticoid receptor antagonists, potassium-sparing diuretics, beta-adrenergic antagonists, nonsteroidal anti-inflammatory drugs, heparin, cyclosporine, trimethoprim, digoxin

Hypertension

Hypoaldosteronism (including type 4 renal tubular acidosis)

Addison disease

Advanced age

Lower body mass index.

Both hypokalemia and hyperkalemia are associated with a higher risk of death,^20,21,24  but in patients with heart failure, the survival benefit from ACE inhibitors, ARBs, and mineralocorticoid receptor antagonists outweighs the risk of hyperkalemia.^25–27 Weir and Rolfe²⁸ concluded that patients with heart failure and chronic kidney disease are at greatest risk of hyperkalemia from renin-angiotensin-aldosterone system inhibition, but the increases in potassium levels are small (about 0.1 to 0.3 mmol/L) and unlikely to be clinically significant.

Hyperkalemia tends to recur. Einhorn et al²⁰ found that nearly half of patients with chronic kidney disease who had an episode of hyperkalemia had 1 or more recurrent episodes within a year.

Another concern about using ACE inhibitors and ARBs, especially in patients with chronic kidney disease, is that the serum creatinine level tends to rise when starting these drugs,²⁹ although several studies have shown that an acute rise in creatinine may demonstrate that the drug is actually protecting the kidney.^30,31 Hirsch³² described this phenomenon as “prerenal success,” proposing that the decline in GFR is hemodynamic, secondary to a fall in intraglomerular pressure as a result of efferent vasodilation, and therefore should not be reversed.

Schmidt et al,^33,34 in a study in 122,363 patients who began ACE inhibitor or ARB therapy, found that cardiorenal outcomes were worse, with higher rates of end-stage renal disease, myocardial infarction, heart failure, and death, in those in whom creatinine rose by 30% or more since starting treatment. This trend was also seen, to a lesser degree, in those with a smaller increase in creatinine, suggesting that even this group of patients should receive close monitoring.

Whether renin-angiotensin-aldosterone system inhibitors provide a benefit in advanced progressive chronic kidney disease remains unclear.^35–37  The Angiotensin Converting Enzyme Inhibitor (ACEi)/Angiotensin Receptor Blocker (ARB) Withdrawal in Advanced Renal Disease trial (STOP-ACEi),³⁸ currently under way, will provide valuable data to help close this gap in our knowledge. This open-label randomized controlled trial is testing the hypothesis that stopping ACE inhibitor or ARB treatment, or a combination of both, compared with continuing these treatments, will improve or stabilize renal function in patients with progressive stage 4 or 5 chronic kidney disease.

NEED FOR MONITORING

Taken together, the above data suggest close and regular monitoring is required in patients receiving these drugs. However, monitoring tends to be lax.^34,37,39 A 2017 study of adherence to the guidelines for monitoring serum creatinine and potassium after starting an ACE inhibitor or ARB and subsequent discontinuation found that fewer than 10% of patients had follow-up within the recommended 2 weeks after starting these drugs.³⁴ Most patients with a creatinine rise of 30% or more or a potassium level higher than 6.0 mmol/L continued treatment. There was also no evidence of increased monitoring in those deemed at higher risk of these complications.

WHAT DO THE GUIDELINES SUGGEST?

ACE inhibitors and ARBs in chronic kidney disease and hypertension

Target blood pressures vary in guidelines from different organizations.^4,40–45 The 2017 joint guidelines of the American College of Cardiology and American Heart Association (ACC/AHA)⁴⁰ recommend a target blood pressure of 130/80 mm Hg or less in all patients irrespective of the level of proteinuria and whether they have diabetes mellitus, based on several studies.^46–48 In the elderly, other factors such as the risk of hypotension and falls must be taken into consideration in establishing the most appropriate blood pressure target.

In general, a renin-angiotensin-aldosterone system inhibitor is recommended if the patient has diabetes, stage 1, 2, or 3 chronic kidney disease, or proteinuria. For example, the guidelines recommend a renin-angiotensin-aldosterone system inhibitor in diabetic patients with albuminuria.

None of the guidelines recommend routine use of combination therapy.

ACE inhibitors and ARBs in heart failure

The 2017 ACC/AHA and Heart Failure Society of America (HFSA) guidelines for heart failure⁴⁹ recommend an ACE inhibitor or ARB for patients with stage C (symptomatic) heart failure with reduced ejection fraction, in view of the known cardiovascular morbidity and mortality benefits.

The European Society of Cardiology⁵⁰ recommends ACE inhibitors for patients with symptomatic heart failure with reduced ejection fraction, as well as those with asymptomatic left ventricular systolic dysfunction. In patients with stable coronary artery disease, an ACE inhibitor should be considered even with normal left ventricular function.

ARBs should be used as alternatives in those unable to tolerate ACE inhibitors.

Combination therapy should be avoided due to the increased risk of renal impairment and hyperkalemia but may be considered in patients with heart failure and reduced ejection fraction in whom other treatments are unsuitable. These include patients on beta-blockers who cannot tolerate mineralocorticoid receptor antagonists such as spironolactone. Combination therapy should be done only under strict supervision.⁵⁰

Close monitoring of serum potassium is recommended during ACE inhibitor or ARB use. Those at greatest risk of hyperkalemia include elderly patients, those taking other medications associated with hyperkalemia, and diabetic patients, because of their higher risk of renovascular disease.

Caution is advised when starting ACE inhibitor or ARB therapy in these high-risk groups as well as in patients with potassium levels higher than 5.0 mmol/L at baseline, at high risk of prerenal acute kidney injury, with known renal insufficiency, and with previous deterioration in renal function on these medications.^3,41,51

Before starting therapy, ensure that patients are volume-replete and measure baseline serum electrolytes and creatinine.^41,51

The ACC/AHA and HFSA recommend starting at a low dose and titrating upward slowly. If maximal doses are not tolerated, then a lower dose should be maintained.⁴⁹ The European Society of Cardiology guidelines⁵² suggest increasing the dose at no less than every 2 weeks unless in an inpatient setting. Blood testing should be done 7 to 14 days after starting therapy, after any titration in dosage, and every 4 months thereafter.⁵³

The guidelines generally agree that a rise in creatinine of up to 30% and a fall in eGFR of up to 25% is acceptable, with the need for regular monitoring, particularly in high-risk groups.^{40–42,51,52}

What if serum potassium or creatinine rises during treatment?

If hyperkalemia arises or renal function declines by a significant amount, one should first address contributing factors. If no improvement is seen, then the dose of the ACE inhibitor or ARB should be reduced by 50% and blood work repeated in 1 to 2 weeks. If the laboratory values do not return to an acceptable level, reducing the dose further or stopping the drug is advised.

Give dietary advice to all patients with chronic kidney disease being considered for a renin-angiotensin-aldosterone system inhibitor or for an increase in dose with a potassium level higher than 4.5 mmol/L. A low-potassium diet should aim for potassium intake of less than 50 or 75 mmol/day and sodium intake of less than 60 mmol/day for hypertensive patients with chronic kidney disease.

Review the patient’s medications if the baseline potassium level is higher than 5.0 mmol/L. Consider stopping potassium-sparing agents, digoxin, trimethoprim, and nonsteroidal anti-inflammatory drugs. Also think about starting a non–potassium-sparing diuretic as well as sodium bicarbonate to reduce potassium levels. Blood work should be repeated within 2 weeks after these changes.

Do not start a renin-angiotensin-aldosterone system inhibitor, or do not increase the dose, if the potassium level is elevated until measures have been taken to reduce the degree of hyperkalemia.⁵¹

In renal transplant recipients, renin-angiotensin-aldosterone system inhibitors are often preferred to manage hypertension in those who have proteinuria or cardiovascular disease. However, the risk of hyperkalemia is also greater with concomitant use of immunosuppressive drugs such as tacrolimus and cyclosporine. Management of complications should be approached according to guidelines discussed above.⁵¹

Monitor renal function, potassium. The National Institute for Health and Care Excellence guideline⁵⁴ advocates that baseline renal function testing should be followed by repeat blood testing 1 to 2 weeks after starting renin-angiotensin-aldosterone system inhibitors in patients with ischemic heart disease. The advice is similar when starting therapy in patients with chronic heart failure, emphasizing the need to monitor after each dose increment and to use clinical judgment when deciding to start treatment. The AHA advises caution in patients with renal insufficiency or a potassium level above 5.0 mmol/L.⁴⁹

Sick day rules. The National Institute for Health and Care Excellence encourages discussing “sick day rules” with patients starting renin-angiotensin-aldosterone system inhibitors. This means patients should be advised to temporarily stop taking nephrotoxic medications, including over-the-counter nonsteroidal anti-inflammatory drugs, in any potential state of illness or dehydration, such as diarrhea and vomiting. There is, however, little evidence that this advice can actually reduce the incidence of acute kidney injury.^55,56

OUR RECOMMENDATIONS

Our advice for managing patients receiving ACE inhibitors or ARBs is summarized in Table 1.

A highly active, water- and alcohol-soluble, basic pressor substance is formed when renin and renin-activator interact, for which we suggest the name “angiotonin.”

—Irvine H. Page and O.M. Helmer, 1940.¹

The renin-angiotensin-aldosterone system regulates salt and, in part, water homeostasis, and therefore blood pressure and fluid balance through its actions on the heart, kidneys, and blood vessels.² Drugs that target this system—angiotensin-converting enzyme (ACE) inhibitors and angiotensin II receptor blockers (ARBs)—are used primarily to treat hypertension and also to treat chronic kidney disease and heart failure with reduced ejection fraction.

See related editorial

Controlling blood pressure is important, as hypertension increases the risk of myocardial infarction, cerebrovascular events, and progression of chronic kidney disease, which itself is a risk factor for cardiovascular disease. However, the benefit of these drugs is only partly due to their effect on blood pressure. They also reduce proteinuria, which is a graded risk factor for progression of kidney disease as well as morbidity and death from vascular events.³

Despite the benefits of ACE inhibitors and ARBs, concern about their adverse effects—especially hyperkalemia and a decline in renal function—has led to their underuse in patients likely to derive the greatest benefit.³

ACE INHIBITORS AND ARBs

ACE inhibitors, as their name indicates, inhibit conversion of angiotensin I to angiotensin II by ACE, resulting in vasodilation of the efferent arteriole and a drop in blood pressure. Inhibition of ACE, a kininase, also results in a rise in kinins. One of these, bradykinin, is associated with some of the side effects of this class of drugs such as cough, which affects 5% to 20% of patients.⁴ Elevation of bradykinin is also believed to account for ACE inhibitor-induced angioedema, an uncommon but potentially serious side effect. Kinins are also associated with desirable effects such as lowering blood pressure, increasing insulin sensitivity, and dilating blood vessels.

ARBs were developed as an alternative for patients unable to tolerate the adverse effects of ACE inhibitors. While ACE inhibitors reduce the activity of angiotensin II at both the AT1 and AT2 receptors, ARBs block only the AT1 receptors, thereby inhibiting their vasoconstricting activity on smooth muscle. ARBs also raise the levels of renin, angiotensin I, and angiotensin II as a result of feedback inhibition. Angiotensin II is associated with release of inflammatory mediators such as tumor necrosis factor alpha, cytokines, and chemokines, the consequences of which are also inhibited by ARBs, further preventing renal fibrosis and scarring from chronic inflammation.³

What is the evidence supporting the use of ACE inhibitors and ARBs?

ACE inhibitors and ARBs, used singly, reduce blood pressure and proteinuria, slow progression of kidney disease, and improve outcomes in patients who have heart failure, diabetes mellitus, or a history of myocardial infarction.^5–11 

While dual blockade with the combination of an ACE inhibitor and an ARB lowers blood pressure and proteinuria to a greater degree than monotherapy, dual blockade has been associated with higher rates of complications, including hyperkalemia.^12–17

RISK FACTORS FOR HYPERKALEMIA

ACE inhibitors and ARBs raise potassium, especially when used in combination. Other risk factors for hyperkalemia include the following—and note that some of them are also indications for ACE inhibitors and ARBs:

Renal insufficiency. The kidneys are responsible for over 90% of potassium removal in healthy individuals,^18,19 and the lower the GFR, the higher the risk of hyperkalemia.^3,20,21

Heart failure

Diabetes mellitus^6,21–23

Endogenous potassium load due to hemolysis, rhabdomyolysis, insulin deficiency, lactic acidosis, or gastrointestinal bleeding

Exogenous potassium load due to dietary consumption or blood products

Other medications, eg, sacubitril-valsartan, aldosterone antagonists, mineralocorticoid receptor antagonists, potassium-sparing diuretics, beta-adrenergic antagonists, nonsteroidal anti-inflammatory drugs, heparin, cyclosporine, trimethoprim, digoxin

Hypertension

Hypoaldosteronism (including type 4 renal tubular acidosis)

Addison disease

Advanced age

Lower body mass index.

Both hypokalemia and hyperkalemia are associated with a higher risk of death,^20,21,24  but in patients with heart failure, the survival benefit from ACE inhibitors, ARBs, and mineralocorticoid receptor antagonists outweighs the risk of hyperkalemia.^25–27 Weir and Rolfe²⁸ concluded that patients with heart failure and chronic kidney disease are at greatest risk of hyperkalemia from renin-angiotensin-aldosterone system inhibition, but the increases in potassium levels are small (about 0.1 to 0.3 mmol/L) and unlikely to be clinically significant.

Hyperkalemia tends to recur. Einhorn et al²⁰ found that nearly half of patients with chronic kidney disease who had an episode of hyperkalemia had 1 or more recurrent episodes within a year.

Another concern about using ACE inhibitors and ARBs, especially in patients with chronic kidney disease, is that the serum creatinine level tends to rise when starting these drugs,²⁹ although several studies have shown that an acute rise in creatinine may demonstrate that the drug is actually protecting the kidney.^30,31 Hirsch³² described this phenomenon as “prerenal success,” proposing that the decline in GFR is hemodynamic, secondary to a fall in intraglomerular pressure as a result of efferent vasodilation, and therefore should not be reversed.

Schmidt et al,^33,34 in a study in 122,363 patients who began ACE inhibitor or ARB therapy, found that cardiorenal outcomes were worse, with higher rates of end-stage renal disease, myocardial infarction, heart failure, and death, in those in whom creatinine rose by 30% or more since starting treatment. This trend was also seen, to a lesser degree, in those with a smaller increase in creatinine, suggesting that even this group of patients should receive close monitoring.

Whether renin-angiotensin-aldosterone system inhibitors provide a benefit in advanced progressive chronic kidney disease remains unclear.^35–37  The Angiotensin Converting Enzyme Inhibitor (ACEi)/Angiotensin Receptor Blocker (ARB) Withdrawal in Advanced Renal Disease trial (STOP-ACEi),³⁸ currently under way, will provide valuable data to help close this gap in our knowledge. This open-label randomized controlled trial is testing the hypothesis that stopping ACE inhibitor or ARB treatment, or a combination of both, compared with continuing these treatments, will improve or stabilize renal function in patients with progressive stage 4 or 5 chronic kidney disease.

NEED FOR MONITORING

Taken together, the above data suggest close and regular monitoring is required in patients receiving these drugs. However, monitoring tends to be lax.^34,37,39 A 2017 study of adherence to the guidelines for monitoring serum creatinine and potassium after starting an ACE inhibitor or ARB and subsequent discontinuation found that fewer than 10% of patients had follow-up within the recommended 2 weeks after starting these drugs.³⁴ Most patients with a creatinine rise of 30% or more or a potassium level higher than 6.0 mmol/L continued treatment. There was also no evidence of increased monitoring in those deemed at higher risk of these complications.

WHAT DO THE GUIDELINES SUGGEST?

ACE inhibitors and ARBs in chronic kidney disease and hypertension

Target blood pressures vary in guidelines from different organizations.^4,40–45 The 2017 joint guidelines of the American College of Cardiology and American Heart Association (ACC/AHA)⁴⁰ recommend a target blood pressure of 130/80 mm Hg or less in all patients irrespective of the level of proteinuria and whether they have diabetes mellitus, based on several studies.^46–48 In the elderly, other factors such as the risk of hypotension and falls must be taken into consideration in establishing the most appropriate blood pressure target.

In general, a renin-angiotensin-aldosterone system inhibitor is recommended if the patient has diabetes, stage 1, 2, or 3 chronic kidney disease, or proteinuria. For example, the guidelines recommend a renin-angiotensin-aldosterone system inhibitor in diabetic patients with albuminuria.

None of the guidelines recommend routine use of combination therapy.

ACE inhibitors and ARBs in heart failure

The 2017 ACC/AHA and Heart Failure Society of America (HFSA) guidelines for heart failure⁴⁹ recommend an ACE inhibitor or ARB for patients with stage C (symptomatic) heart failure with reduced ejection fraction, in view of the known cardiovascular morbidity and mortality benefits.

The European Society of Cardiology⁵⁰ recommends ACE inhibitors for patients with symptomatic heart failure with reduced ejection fraction, as well as those with asymptomatic left ventricular systolic dysfunction. In patients with stable coronary artery disease, an ACE inhibitor should be considered even with normal left ventricular function.

ARBs should be used as alternatives in those unable to tolerate ACE inhibitors.

Combination therapy should be avoided due to the increased risk of renal impairment and hyperkalemia but may be considered in patients with heart failure and reduced ejection fraction in whom other treatments are unsuitable. These include patients on beta-blockers who cannot tolerate mineralocorticoid receptor antagonists such as spironolactone. Combination therapy should be done only under strict supervision.⁵⁰

Close monitoring of serum potassium is recommended during ACE inhibitor or ARB use. Those at greatest risk of hyperkalemia include elderly patients, those taking other medications associated with hyperkalemia, and diabetic patients, because of their higher risk of renovascular disease.

Caution is advised when starting ACE inhibitor or ARB therapy in these high-risk groups as well as in patients with potassium levels higher than 5.0 mmol/L at baseline, at high risk of prerenal acute kidney injury, with known renal insufficiency, and with previous deterioration in renal function on these medications.^3,41,51

Before starting therapy, ensure that patients are volume-replete and measure baseline serum electrolytes and creatinine.^41,51

The ACC/AHA and HFSA recommend starting at a low dose and titrating upward slowly. If maximal doses are not tolerated, then a lower dose should be maintained.⁴⁹ The European Society of Cardiology guidelines⁵² suggest increasing the dose at no less than every 2 weeks unless in an inpatient setting. Blood testing should be done 7 to 14 days after starting therapy, after any titration in dosage, and every 4 months thereafter.⁵³

The guidelines generally agree that a rise in creatinine of up to 30% and a fall in eGFR of up to 25% is acceptable, with the need for regular monitoring, particularly in high-risk groups.^{40–42,51,52}

What if serum potassium or creatinine rises during treatment?

If hyperkalemia arises or renal function declines by a significant amount, one should first address contributing factors. If no improvement is seen, then the dose of the ACE inhibitor or ARB should be reduced by 50% and blood work repeated in 1 to 2 weeks. If the laboratory values do not return to an acceptable level, reducing the dose further or stopping the drug is advised.

Give dietary advice to all patients with chronic kidney disease being considered for a renin-angiotensin-aldosterone system inhibitor or for an increase in dose with a potassium level higher than 4.5 mmol/L. A low-potassium diet should aim for potassium intake of less than 50 or 75 mmol/day and sodium intake of less than 60 mmol/day for hypertensive patients with chronic kidney disease.

Review the patient’s medications if the baseline potassium level is higher than 5.0 mmol/L. Consider stopping potassium-sparing agents, digoxin, trimethoprim, and nonsteroidal anti-inflammatory drugs. Also think about starting a non–potassium-sparing diuretic as well as sodium bicarbonate to reduce potassium levels. Blood work should be repeated within 2 weeks after these changes.

Do not start a renin-angiotensin-aldosterone system inhibitor, or do not increase the dose, if the potassium level is elevated until measures have been taken to reduce the degree of hyperkalemia.⁵¹

In renal transplant recipients, renin-angiotensin-aldosterone system inhibitors are often preferred to manage hypertension in those who have proteinuria or cardiovascular disease. However, the risk of hyperkalemia is also greater with concomitant use of immunosuppressive drugs such as tacrolimus and cyclosporine. Management of complications should be approached according to guidelines discussed above.⁵¹

Monitor renal function, potassium. The National Institute for Health and Care Excellence guideline⁵⁴ advocates that baseline renal function testing should be followed by repeat blood testing 1 to 2 weeks after starting renin-angiotensin-aldosterone system inhibitors in patients with ischemic heart disease. The advice is similar when starting therapy in patients with chronic heart failure, emphasizing the need to monitor after each dose increment and to use clinical judgment when deciding to start treatment. The AHA advises caution in patients with renal insufficiency or a potassium level above 5.0 mmol/L.⁴⁹

Sick day rules. The National Institute for Health and Care Excellence encourages discussing “sick day rules” with patients starting renin-angiotensin-aldosterone system inhibitors. This means patients should be advised to temporarily stop taking nephrotoxic medications, including over-the-counter nonsteroidal anti-inflammatory drugs, in any potential state of illness or dehydration, such as diarrhea and vomiting. There is, however, little evidence that this advice can actually reduce the incidence of acute kidney injury.^55,56

OUR RECOMMENDATIONS

Our advice for managing patients receiving ACE inhibitors or ARBs is summarized in Table 1.

PARIS – For patients with peripheral artery disease, statin therapy is a literal lifeline, nearly halving mortality risk, according to new research presented at the annual congress of the European Society of Cardiology.

Patients with peripheral manifestations of cardiovascular disease “are a population with an extremely high risk to suffer a heart attack or a stroke,” said Joern Dopheide, MD, during a press conference at the meeting. Despite the known benefits of statins, including the reduction of all-cause and cardiovascular death and the reduction of morbidity, adherence to guideline-directed statin therapy is far from optimal, said Dr. Dopheide of Bern (Switzerland) University Hospital.

Patients with peripheral artery disease (PAD) not taking statins had a mortality rate of 34%, more than three times that of patients adherent to an intensified statin regimen. More surprisingly, patients who had been on a statin and then stopped the medication also had a mortality rate of 33%, indistinguishable from those who had never been treated with a statin.

Although statin adherence is low in general, it’s especially low in patients with PAD, said Dr. Dopheide. Still, he said, “few systematic data exist on the prognostic value of statin adherence and the correlation between adherence and cardiovascular outcome in PAD patients.”

Accordingly, Dr. Dopheide and his coinvestigators sought to determine the association between statin adherence and survival in PAD patients. The researchers obtained baseline and follow-up data for a cohort of 691 symptomatic PAD patients seen at a single site, looking at statin dosage, LDL cholesterol levels, and survival.

The patients were followed for a period of 50 months. Dr. Dopheide said that “Over the time course, we were able to increase the statin adherence from about 73% to about 81%, and parallel to that, we were able to reduce the LDL cholesterol levels from about 97 to 83 mg/dL, and we were able to increase the intensity of patients on statin therapy.”

Dr. Dopheide said that he and his colleagues saw a dose-response effect, so that the biggest drop in cholesterol was seen in patients on high statin doses, on more potent statins, or both.

Intensity was increased in some cases by upping statin dose – the mean statin dose climbed from 50 to 58 mg daily during the study period. An alternative strategy was to switch to a more potent statin such as atorvastatin or rosuvastatin; sometimes both intensity and dose were boosted.

“We were able to see that patients who were always on their statin therapy had a pretty low mortality rate of about 20%,” a figure that was halved for patients on more intensive statin therapy, who had a mortality rate of 10% across the study period, said Dr. Dopheide. “Patients in whom we started a statin therapy still profited from it, and had only a 15% mortality,” he added.

Some of the most surprising – and disturbing – study findings involved those who reduced their statin dose: “When patients discontinued their usual dose and decreased it, they suffered an even higher mortality rate, of nearly 43%. So that was kind of surprising and shocking to us.”

Identifying these high-risk patients and keeping them adherent is a substantial clinical challenge, but an important goal, said Dr. Dopheide. “We know that patients with peripheral arterial disease are a little more underrepresented in daily practice; it’s hard to identify them, especially when they are asymptomatic,” he acknowledged. However, once a PAD patient is identified, “One should at least keep the patient on the statin dosage they have,” or initiate statins if needed.

Further, warned Dr. Dopheide, “One should never discontinue statin or decrease the dosage,” adding that PAD patients should be informed that they are at “very high risk for myocardial infarction or stroke.” These patients “should regard their statin therapy as one of the most important and life-saving medications they can take,” he said.

Dr. Dopheide reported no outside sources of funding and no conflicts of interest.

[email protected]

SOURCE: Dopheide, J., et al. ESC Congress 2019, Abstract P5363.

PARIS – For patients with peripheral artery disease, statin therapy is a literal lifeline, nearly halving mortality risk, according to new research presented at the annual congress of the European Society of Cardiology.

Patients with peripheral manifestations of cardiovascular disease “are a population with an extremely high risk to suffer a heart attack or a stroke,” said Joern Dopheide, MD, during a press conference at the meeting. Despite the known benefits of statins, including the reduction of all-cause and cardiovascular death and the reduction of morbidity, adherence to guideline-directed statin therapy is far from optimal, said Dr. Dopheide of Bern (Switzerland) University Hospital.

Patients with peripheral artery disease (PAD) not taking statins had a mortality rate of 34%, more than three times that of patients adherent to an intensified statin regimen. More surprisingly, patients who had been on a statin and then stopped the medication also had a mortality rate of 33%, indistinguishable from those who had never been treated with a statin.

Although statin adherence is low in general, it’s especially low in patients with PAD, said Dr. Dopheide. Still, he said, “few systematic data exist on the prognostic value of statin adherence and the correlation between adherence and cardiovascular outcome in PAD patients.”

Accordingly, Dr. Dopheide and his coinvestigators sought to determine the association between statin adherence and survival in PAD patients. The researchers obtained baseline and follow-up data for a cohort of 691 symptomatic PAD patients seen at a single site, looking at statin dosage, LDL cholesterol levels, and survival.

The patients were followed for a period of 50 months. Dr. Dopheide said that “Over the time course, we were able to increase the statin adherence from about 73% to about 81%, and parallel to that, we were able to reduce the LDL cholesterol levels from about 97 to 83 mg/dL, and we were able to increase the intensity of patients on statin therapy.”

Dr. Dopheide said that he and his colleagues saw a dose-response effect, so that the biggest drop in cholesterol was seen in patients on high statin doses, on more potent statins, or both.

Intensity was increased in some cases by upping statin dose – the mean statin dose climbed from 50 to 58 mg daily during the study period. An alternative strategy was to switch to a more potent statin such as atorvastatin or rosuvastatin; sometimes both intensity and dose were boosted.

“We were able to see that patients who were always on their statin therapy had a pretty low mortality rate of about 20%,” a figure that was halved for patients on more intensive statin therapy, who had a mortality rate of 10% across the study period, said Dr. Dopheide. “Patients in whom we started a statin therapy still profited from it, and had only a 15% mortality,” he added.

Some of the most surprising – and disturbing – study findings involved those who reduced their statin dose: “When patients discontinued their usual dose and decreased it, they suffered an even higher mortality rate, of nearly 43%. So that was kind of surprising and shocking to us.”

Identifying these high-risk patients and keeping them adherent is a substantial clinical challenge, but an important goal, said Dr. Dopheide. “We know that patients with peripheral arterial disease are a little more underrepresented in daily practice; it’s hard to identify them, especially when they are asymptomatic,” he acknowledged. However, once a PAD patient is identified, “One should at least keep the patient on the statin dosage they have,” or initiate statins if needed.

Further, warned Dr. Dopheide, “One should never discontinue statin or decrease the dosage,” adding that PAD patients should be informed that they are at “very high risk for myocardial infarction or stroke.” These patients “should regard their statin therapy as one of the most important and life-saving medications they can take,” he said.

Dr. Dopheide reported no outside sources of funding and no conflicts of interest.

[email protected]

SOURCE: Dopheide, J., et al. ESC Congress 2019, Abstract P5363.

PARIS – For patients with peripheral artery disease, statin therapy is a literal lifeline, nearly halving mortality risk, according to new research presented at the annual congress of the European Society of Cardiology.

Patients with peripheral manifestations of cardiovascular disease “are a population with an extremely high risk to suffer a heart attack or a stroke,” said Joern Dopheide, MD, during a press conference at the meeting. Despite the known benefits of statins, including the reduction of all-cause and cardiovascular death and the reduction of morbidity, adherence to guideline-directed statin therapy is far from optimal, said Dr. Dopheide of Bern (Switzerland) University Hospital.

Patients with peripheral artery disease (PAD) not taking statins had a mortality rate of 34%, more than three times that of patients adherent to an intensified statin regimen. More surprisingly, patients who had been on a statin and then stopped the medication also had a mortality rate of 33%, indistinguishable from those who had never been treated with a statin.

Although statin adherence is low in general, it’s especially low in patients with PAD, said Dr. Dopheide. Still, he said, “few systematic data exist on the prognostic value of statin adherence and the correlation between adherence and cardiovascular outcome in PAD patients.”

Accordingly, Dr. Dopheide and his coinvestigators sought to determine the association between statin adherence and survival in PAD patients. The researchers obtained baseline and follow-up data for a cohort of 691 symptomatic PAD patients seen at a single site, looking at statin dosage, LDL cholesterol levels, and survival.

The patients were followed for a period of 50 months. Dr. Dopheide said that “Over the time course, we were able to increase the statin adherence from about 73% to about 81%, and parallel to that, we were able to reduce the LDL cholesterol levels from about 97 to 83 mg/dL, and we were able to increase the intensity of patients on statin therapy.”

Dr. Dopheide said that he and his colleagues saw a dose-response effect, so that the biggest drop in cholesterol was seen in patients on high statin doses, on more potent statins, or both.

Intensity was increased in some cases by upping statin dose – the mean statin dose climbed from 50 to 58 mg daily during the study period. An alternative strategy was to switch to a more potent statin such as atorvastatin or rosuvastatin; sometimes both intensity and dose were boosted.

“We were able to see that patients who were always on their statin therapy had a pretty low mortality rate of about 20%,” a figure that was halved for patients on more intensive statin therapy, who had a mortality rate of 10% across the study period, said Dr. Dopheide. “Patients in whom we started a statin therapy still profited from it, and had only a 15% mortality,” he added.

Some of the most surprising – and disturbing – study findings involved those who reduced their statin dose: “When patients discontinued their usual dose and decreased it, they suffered an even higher mortality rate, of nearly 43%. So that was kind of surprising and shocking to us.”

Identifying these high-risk patients and keeping them adherent is a substantial clinical challenge, but an important goal, said Dr. Dopheide. “We know that patients with peripheral arterial disease are a little more underrepresented in daily practice; it’s hard to identify them, especially when they are asymptomatic,” he acknowledged. However, once a PAD patient is identified, “One should at least keep the patient on the statin dosage they have,” or initiate statins if needed.

Further, warned Dr. Dopheide, “One should never discontinue statin or decrease the dosage,” adding that PAD patients should be informed that they are at “very high risk for myocardial infarction or stroke.” These patients “should regard their statin therapy as one of the most important and life-saving medications they can take,” he said.

Dr. Dopheide reported no outside sources of funding and no conflicts of interest.

[email protected]

SOURCE: Dopheide, J., et al. ESC Congress 2019, Abstract P5363.

Major depressive disorder (MDD) is a significant pediatric health problem, with a lifetime prevalence as high as 20% by the end of adolescence.^1-3 Major depressive disorder in adolescence is associated with significant morbidity, including poor social functioning, school difficulties, early pregnancy, and increased risk of physical illness and substance abuse.^4-6 It is also linked with significant mortality, with increased risk for suicide, which is now the second leading cause of death in individuals age 10 to 24 years.^1,7,8

As their name suggests, antidepressants comprise a group of medications that are used to treat MDD; they are also, however, first-line agents for generalized anxiety disorder (GAD), posttraumatic stress disorder (PTSD), and obsessive-compulsive disorder (OCD) in adults. Anxiety disorders (including GAD and other anxiety diagnoses) and PTSD are also common in childhood and adolescence with a combined lifetime prevalence ranging from 15% to 30%.^9,10 These disorders are also associated with increased risk of suicide.¹¹ For all of these disorders, depending on the severity of presentation and the preference of the patient, treatments are often a combination of psychotherapy and psychopharmacology.

Clinicians face several challenges when considering antidepressants for pediatric patients. Pediatricians and psychiatrists need to understand whether these medications work in children and adolescents, and whether there are unique developmental safety and tolerability issues. The evidence base in child psychiatry is considerably smaller compared with that of adult psychiatry. From this more limited evidence base also came the controversial “black-box” warning regarding a risk of emergent suicidality when starting antidepressants that accompanies all antidepressants for pediatric, but not adult, patients. This warning has had major effects on clinical encounters with children experiencing depression, including altering clinician prescribing behavior.¹²

In this article, we review the current evidence for antidepressant efficacy, tolerability, and safety in pediatric patients. We also suggest ways in which clinicians might choose, start, and stop antidepressants in children, as well as how to talk with parents about benefits, risks, and the black-box warning.

Do antidepressants work in children?

Selective serotonin reuptake inhibitors. Selective serotonin reuptake inhibitors (SSRIs) are the most commonly used class of antidepressants in both children and adults.¹³ While only a few SSRIs are FDA-approved for pediatric indications, the lack of FDA approval is typically related to a lack of sufficient testing in randomized controlled trials (RCTs) for specific pediatric indications, rather than to demonstrable differences in efficacy between antidepressant agents. Since there is currently no data to suggest inferiority of one agent compared to another in children or adults,^14,15 efficacy data will be discussed here as applied to the class of SSRIs, generalizing from RCTs conducted on individual drugs. Table 1 lists FDA indications and dosing information for individual antidepressants.

There is strong evidence that SSRIs are effective for treating pediatric anxiety disorders (eg, social anxiety disorder and GAD)¹⁶ and OCD,¹⁷ with numbers needed to treat (NNT) between 3 and 5. For both of these disorders, SSRIs combined with cognitive-behavioral therapy (CBT) have the highest likelihood of improving symptoms or achieving remission.^17,18

Selective serotonin reuptake inhibitors are also effective for treating pediatric MDD; however, the literature is more complex for this disorder compared to GAD and OCD as there are considerable differences in effect sizes between National Institute of Mental Health (NIMH)–funded studies and industry-sponsored trials.¹³ The major NIMH-sponsored adolescent depression trial, TADS (Treatment for Adolescents and Depression Study), showed that SSRIs (fluoxetine in this case) were quite effective, with an NNT of 4 over the acute phase (12 weeks).¹⁹ Ultimately, approximately 80% of adolescents improved over 9 months. Many industry-sponsored trials for MDD in pediatric patients had large placebo response rates (approximately 60%), which resulted in smaller between-group differences, and estimates of an NNT closer to 12,¹³ which has muddied the waters in meta-analyses that include all trials.²⁰ Improvement in depressive symptoms also appears to be bolstered by concomitant CBT in MDD,¹⁹ but not as robustly as in GAD and OCD. While the full benefit of SSRIs for depression may take as long as 8 weeks, a meta-analysis of depression studies of pediatric patients suggests that significant benefits from placebo are observed as early as 2 weeks, and that further treatment gains are minimal after 4 weeks.¹⁵ Thus, we recommend at least a 4- to 6-week trial at therapeutic dosing before deeming a medication a treatment failure.

Continue to: Posttraumatic stress disorder...

Posttraumatic stress disorder is a fourth disorder in which SSRIs are a first-line treatment in adults. The data for using SSRIs to treat pediatric patients with PTSD is scant, with only a few RCTs, and no large NIMH-funded trials. Randomized controlled trials have not demonstrated significant differences between SSRIs and placebo^21,22 and thus the current first-line recommendation in pediatric PTSD remains trauma-focused therapy, with good evidence for trauma-focused CBT.²³ Practically speaking, there can be considerable overlap of PTSD, depression, and anxiety symptoms in children,²³ and children with a history of trauma who also have comorbid MDD may benefit from medication if their symptoms persist despite an adequate trial of psychotherapy.

Taken together, the current evidence suggests that SSRIs are often effective in pediatric GAD, OCD, and MDD, with low NNTs (ranging from 3 to 5 based on NIMH-funded trials) for all of these disorders; there is not yet sufficient evidence of efficacy in pediatric patients with PTSD.

Fluoxetine has been studied more intensively than other SSRIs (for example, it was the antidepressant used in the TADS trial), and thus has the largest evidence base. For this reason, fluoxetine is often considered the first of the first-line options. Additionally, fluoxetine has a longer half-life than other antidepressants, which may make it more effective in situations where patients are likely to miss doses, and results in a lower risk of withdrawal symptoms when stopped due to “self-tapering.”

SNRIs and atypical antidepressants. Other antidepressants commonly used in pediatric patients but with far less evidence of efficacy include serotonin-norepinephrine reuptake inhibitors (SNRIs) and the atypical antidepressants bupropion and mirtazapine. The SNRI duloxetine is FDA-approved for treating GAD in children age 7 to 17, but there are no other pediatric indications for duloxetine, or for the other SNRIs.

In general, adverse effect profiles are worse for SNRIs compared to SSRIs, further limiting their utility. While there are no pediatric studies demonstrating SNRI efficacy for neuropathic pain, good data exists in adults.²⁴ Thus, an SNRI could be a reasonable option if a pediatric patient has failed prior adequate SSRI trials and also has comorbid neuropathic pain.

Continue to: Neither bupropion nor mirtazapine...

Neither bupropion nor mirtazapine have undergone rigorous testing in pediatric patients, and therefore these agents should generally be considered only once other first-line treatments have failed. Bupropion has been evaluated for attention-deficit/hyperactivity disorder (ADHD)²⁵ and for adolescent smoking cessation.²⁶ However, the evidence is weak, and bupropion is not considered a first-line option for children and adolescents.

Tricyclic antidepressants. Randomized controlled trials have demonstrated that tricyclic antidepressants (TCAs) are efficacious for treating several pediatric conditions; however, their significant side effect profile, their monitoring requirements, as well as their lethality in overdose has left them replaced by SSRIs in most cases. That said, they can be appropriate in refractory ADHD (desipramine^27,28) and refractory OCD (clomipramine is FDA-approved for this indication²⁹); they are considered a third-line treatment for enuresis.³⁰

Why did my patient stop the medication?

Common adverse effects. Although the greatest benefit of antidepressant medications compared with placebo is achieved relatively early on in treatment, it generally takes time for these benefits to accrue and become clinically apparent.^15,31 By contrast, most adverse effects of antidepressants present and are at their most severe early in treatment. The combination of early adverse effects and delayed efficacy leads many patients, families, and clinicians to discontinue medications before they have an adequate chance to work. Thus, it is imperative to provide psychoeducation before starting a medication about the typical time-course of improvement and adverse effects (Table 2).

Adverse effects of SSRIs often appear or worsen transiently during initiation of a medication, during a dose increase,³² or, theoretically, with the addition of a medication that interferes with SSRI metabolism (eg, cimetidine inhibition of cytochrome P450 2D6).³³ If families are prepared for this phenomenon and the therapeutic alliance is adequate, adverse effects can be tolerated to allow for a full medication trial. Common adverse effects of SSRIs include sleep problems (insomnia/sedation), gastrointestinal upset, sexual dysfunction, dry mouth, and hyperhidrosis. Although SSRIs differ somewhat in the frequency of these effects, as a class, they are more similar than different. Adequate psychoeducation is especially imperative in the treatment of OCD and anxiety disorders, where there is limited evidence of efficacy for any non-serotonergic antidepressants.

Serotonin-norepinephrine reuptake inhibitors are not considered first-line medications because of the reduced evidence base compared to SSRIs and their enhanced adverse effect profiles. Because SNRIs partially share a mechanism of action with SSRIs, they also share portions of the adverse effects profile. However, SNRIs have the additional adverse effect of hypertension, which is related to their noradrenergic activity. Thus, it is reasonable to obtain a baseline blood pressure before initiating an SNRI, as well as periodically after initiation and during dose increases, particularly if the patient has other risk factors for hypertension.³⁴

Continue to: Although TCAs have efficacy...

Although TCAs have efficacy in some pediatric disorders,^27-29,35 their adverse effect profile limits their use. Tricyclic antidepressants are highly anticholinergic (causing dizziness secondary to orthostatic hypotension, dry mouth, and urinary retention) and antihistaminergic (causing sedation and weight gain). Additionally, TCAs lower the seizure threshold and have adverse cardiac effects relating to their anti-alpha-1 adrenergic activity, resulting in dose-dependent increases in the QTc and cardiac toxicity in overdose that could lead to arrhythmia and death. These medications have their place, but their use requires careful informed consent, clear treatment goals, and baseline and periodic cardiac monitoring (via electrocardiogram).

Serious adverse effects. Clinicians may be hesitant to prescribe antidepressants for pediatric patients because of the potential for more serious adverse effects, including severe behavioral activation syndromes, serotonin syndrome, and emergent suicidality. However, current FDA-approved antidepressants arguably have one of the most positive risk/benefit profiles of any orally-administered medication approved for pediatric patients. Having a strong understanding of the evidence is critical to evaluating when it is appropriate to prescribe an antidepressant, how to properly monitor the patient, and how to obtain accurate informed consent.

Pediatric behavioral activation syndrome. Many clinicians report that children receiving antidepressants experience a pediatric behavioral activation syndrome, which exists along a spectrum from mild activation, increased energy, insomnia, or irritability up through more severe presentations of agitation, hyperactivity, or possibly mania. A recent meta-analysis suggested a positive association between antidepressant use and activation events on the milder end of this spectrum in pediatric patients with non-OCD anxiety disorders,¹⁶ and it is thought that compared with adolescents, younger children are more susceptible to activation adverse effects.³⁶ The likelihood of activation events has been associated with higher antidepressant plasma levels,³⁷ suggesting that dose or individual differences in metabolism may play a role. At the severe end of the spectrum, the risk of induction of mania in pediatric patients with depression or anxiety is relatively rare (<2%) and not statistically different from placebo in RCTs of pediatric participants.³⁸ Meta-analyses of larger randomized, placebo-controlled trials of adults do not support the idea that SSRIs and other second-generation antidepressants carry an increased risk of mania compared with placebo.^39,40 Children or adolescents with bona fide bipolar disorder (ie, patients who have had observed mania that meets all DSM-5 criteria) should be treated with a mood-stabilizing agent or antipsychotic if prescribed an antidepressant.⁴¹ These clear-cut cases are, however, relatively rare, and more often clinicians are confronted with ambiguous cases that include a family history of bipolar disorder along with “softer” symptoms of irritability, intrusiveness, or aggression. In these children, SSRIs may be appropriate for depressive, OCD, or anxiety symptoms, and should be strongly considered before prescribing antipsychotics or mood stabilizers, as long as initiated with proper monitoring.

Serotonin syndrome is a life-threatening condition caused by excess synaptic serotonin. It is characterized by confusion, sweating, diarrhea, hypertension, hyperthermia, and tachycardia. At its most severe, serotonin syndrome can result in seizures, arrhythmias, and death. The risk of serotonin syndrome is very low when using an SSRI as monotherapy. Risk increases with polypharmacy, particularly unexamined polypharmacy when multiple serotonergic agents are inadvertently on board. Commonly used serotonergic agents include other antidepressants, migraine medications (eg, triptans), some pain medications, and the cough suppressant dextromethorphan.

The easiest way to mitigate the risk of serotonin syndrome is to use an interaction index computer program, which can help ensure that the interacting agents are not prescribed without first discussing the risks and benefits. It is important to teach adolescents that certain recreational drugs are highly serotonergic and can cause serious interactions with antidepressants. For example, recreational use of dextrometh­orphan or 3,4-methylenedioxymethamphetamine (MDMA; commonly known as “ecstasy”) has been associated with serotonin syndrome in adolescents taking antidepressant medications.^42,43

Continue to: Suicidality

Suicidality. The black-box warning regarding a risk of emergent suicidality when starting antidepressant treatment in children is controversial.⁴⁴ The prospect that a medication intended to ameliorate depression might instead risk increasing suicidal thinking is alarming to parents and clinicians alike. To appropriately weigh and discuss the risks and benefits with families, it is important to understand the data upon which the warning is based.

In 2004, the FDA commissioned a review of 23 antidepressant trials, both published and unpublished, pooling studies across multiple indications (MDD, OCD, anxiety, and ADHD) and multiple antidepressant classes. This meta-analysis, which included nearly 4,400 pediatric patients, found a small but statistically significant increase in spontaneously-reported suicidal thoughts or actions, with a risk difference of 1% (95% confidence interval [CI], 1% to 2%).⁴⁵ These data suggest that if one treats 100 pediatric patients, 1 to 2 of them may experience short-term increases in suicidal thinking or behavior.⁴⁵ There were no differences in suicidal thinking when assessed systematically (ie, when all subjects reported symptoms of suicidal ideation on structured rating scales), and there were no completed suicides.⁴⁵ A subsequent analysis that included 27 pediatric trials suggested an even lower, although still significant, risk difference (<1%), yielding a number needed to harm (NNH) of 143.⁴⁶ Thus, with low NNT for efficacy (3 to 6) and relatively high NNH for emergent suicidal thoughts or behaviors (100 to 143), for many patients the benefits will outweigh the risks.

Figure 1, Figure 2, and Figure 3 are Cates plots that depict the absolute benefits of antidepressants compared with the risk of suicidality for pediatric patients with MDD, OCD, and anxiety disorders. Recent meta-analyses have suggested that the increased risk of suicidality in antidepressant trials is specific to studies that included children and adolescents, and is not observed in adult studies. A meta-analysis of 70 trials involving 18,526 participants suggested that the odds ratio of suicidality in trials of children and adolescents was 2.39 (95% CI, 1.31 to 4.33) compared with 0.81 (95% CI, 0.51 to 1.28) in adults.⁴⁷ Additionally, a network meta-analysis exclusively focusing on pediatric antidepressant trials in MDD reported significantly higher suicidality-related adverse events in venlafaxine trials compared with placebo, duloxetine, and several SSRIs (fluoxetine, paroxetine, and escitalopram).²⁰ These data should be interpreted with caution as differences in suicidality detected between agents is quite possibly related to differences in the method of assessment between trials, as opposed to actual differences in risk between agents.

Epidemiologic data further support the use of antidepressants in pediatric patients, showing that antidepressant use is associated with decreased teen suicide attempts and completions,⁴⁸ and the decline in prescriptions that occurred following the black-box warning was accompanied by a 14% increase in teen suicides.⁴⁹ Multiple hypotheses have been proposed to explain the pediatric clinical trial findings. One idea is that potential adverse effects of activation, or the intended effects of restoring the motivation, energy, and social engagement that is often impaired in depression, increases the likelihood of thinking about suicide or acting on thoughts. Another theory is that reporting of suicidality may be increased, rather than increased de novo suicidality itself. Antidepressants are effective for treating pediatric anxiety disorders, including social anxiety disorder,¹⁶ which could result in more willingness to report. Also, the manner in which adverse effects are generally ascertained in trials might have led to increased spontaneous reporting. In many trials, investigators ask whether participants have any adverse effects in general, and inquire about specific adverse effects only if the family answers affirmatively. Thus, the increased rate of other adverse effects associated with antidepressants (sleep problems, gastrointestinal upset, dry mouth, etc.) might trigger a specific question regarding suicidal ideation, which the child or family then may be more likely to report. Alternatively, any type of psychiatric treatment could increase an individual’s propensity to report; in adolescent psychotherapy trials, non-medicated participants have reported emergent suicidality at similar frequencies as those described in drug trials.⁵⁰ Regardless of the mechanism, the possibility of treatment-emergent suicidality is a low-frequency but serious event that necessitates careful monitoring when starting medication. Current guidelines suggest seeing children weekly for the first month after medication initiation, every 2 weeks for the following month, and monthly thereafter.⁵¹

Continue to: How long should the antidepressant be continued?

How long should the antidepressant be continued?

Many patients are concerned about how long they may be taking medication, and whether they will be taking an antidepressant “forever.” A treatment course can be broken into an acute phase, wherein remission is achieved and maintained for 6 to 8 weeks. This is followed by a continuation phase, with the goal of relapse prevention, lasting 16 to 20 weeks. The length of the last phase—the maintenance phase—depends both on the child’s history, the underlying therapeutic indication, the adverse effect burden experienced, and the family’s preferences/values. In general, for a first depressive episode, after treating for 1 year, a trial of discontinuation can be attempted with close monitoring. For a second depressive episode, we recommend 2 years of remission on antidepressant therapy before attempting discontinuation. In patients who have had 3 depressive episodes, or have had episodes of high severity, we recommend continuing antidepressant treatment indefinitely. Although much less well studied, the risk of relapse following SSRI discontinuation appears much more significant in OCD, whereas anxiety disorders and MDD have a relatively comparable risk.⁵²

In general, stopping an antidepressant should be done carefully and slowly. The speed with which a specific antidepressant can be stopped is largely related to its half-life. Agents with very long half-lives, such as fluoxetine (half-life of 5 days for the parent compound and 9 days for active metabolite), can often be stopped altogether, being “auto-tapered” by the long half-life. One might still consider a taper if the patient were taking high doses. Medications with shorter half-lives must be more carefully tapered to avoid discontinuation syndromes. Discontinuation syndromes are characterized by flu-like symptoms (nausea, myalgias, fatigue, dizziness) and worsening mood. Medications with short half-lives (eg, paroxetine and venlafaxine) have the highest potential for this syndrome in children,⁵³ and thus are used less frequently.

What to do when first-line treatments fail

When a child does not experience sufficient improvement from first-line treatments, it is crucial to determine whether they have experienced an adequate dosing, duration, and quality of medication and psychotherapy.

Adequate psychotherapy? To determine whether children are receiving adequate CBT, ask:

1. if the child receives homework from psychotherapy
2. if the parents are included in treatment
3. if therapy has involved identifying thought patterns that may be contributing to the child’s illness, and
4. if the therapist has ever exposed the child to a challenge likely to produce anxiety or distress in a supervised environment and has developed an exposure hierarchy (for conditions with primarily exposure-based therapies, such as OCD or anxiety disorders).

If a family is not receiving most of these elements in psychotherapy, this is a good indicator that they may not be receiving evidence-based CBT.

Continue to: Adequate pharmacotherapy?

Adequate pharmacotherapy? Similarly, when determining the adequacy of previous pharmacotherapy, it is critical to determine whether the child received an adequate dose of medications (at least the FDA-recommended minimum dose) for an adequate duration of time at therapeutic dosing (at least 6 weeks for MDD, 8 weeks for anxiety disorders, and 8 to 12 weeks for pediatric patients with OCD), and that the child actually took the medication regularly during that period. Patient compliance can typically be tracked through checking refill requests or intervals through the patient’s pharmacy. Ensuring proper delivery of first-line treatments is imperative because (1) the adverse effects associated with second-line treatments are often more substantial; (2) the cost in terms of time and money is considerably higher with second-line treatments, and; (3) the evidence regarding the benefits of these treatments is much less certain.

Inadequate dosing is a common reason for non-response in pediatric patients. Therapeutic dose ranges for common antidepressants are displayed in Table 1. Many clinicians underdose antidepressants for pediatric patients initially (and often throughout treatment) because they fear that the typical dose titration used in clinical trials will increase the risk of adverse effects compared with more conservative dosing. There is limited evidence to suggest that this underdosing strategy is likely to be successful; adverse effects attributable to these medications are modest, and most that are experienced early in treatment (eg, headache, increased anxiety or irritability, sleep problems, gastrointestinal upset) are self-limiting and may be coincidental rather than medication-induced. Furthermore, there is no evidence for efficacy of subtherapeutic dosing in children in the acute phase of treatment or for preventing relapse.¹⁴ Thus, from an efficacy standpoint, a medication trial has not officially begun until the therapeutic dose range is reached.

Once dosing is within the therapeutic range, however, pediatric data differs from the adult literature. In most adult psychi­atric conditions, higher doses of SSRIs within the therapeutic range are associated with an increased response rate.^14,54 In pediatrics, there are few fixed dose trials, and once within the recommended therapeutic range, minimal data supports an association between higher dosing and higher efficacy.¹⁴ In general, pediatric guidelines are silent regarding optimal dosing of SSRIs within the recommended dose range, and higher antidepressant doses often result in a more significant adverse effect burden for children. One exception is pediatric OCD, where, similar to adults, the guidelines suggest that higher dosing of SSRIs often is required to induce a therapeutic response as compared to MDD and GAD.^31,55

If a child does not respond to adequate first-line treatment (or has a treatment history that cannot be fully verified), repeating these first-line interventions carries little risk and can be quite effective. For example, 60% of adolescents with MDD who did not initially respond to an SSRI demonstrated a significant response when prescribed a second SSRI or venlafaxine (with or without CBT).⁵⁶

When pediatric patients continue to experience significantly distressing and/or debilitating symptoms (particularly in MDD) despite multiple trials of antidepressants and psychotherapy, practitioners should consider a careful referral to interventional psychiatry services, which can include the more intensive treatments of electroconvulsive therapy, repetitive transcranial magnetic stimulation, or ketamine (see Box 1). Given the substantial morbidity and mortality associated with adolescent depression, interventional psychiatry treatments are under-researched and under-utilized clinically in pediatric populations.

Continue to: Antidepressants in general...

Antidepressants in general, and SSRIs in particular, are the first-line pharmacotherapy for pediatric anxiety, OCD, and MDD. For PTSD, although they are a first-line treatment in adults, their efficacy has not been demonstrated in children and adolescents. Antidepressants are generally safe, well-tolerated, and effective, with low NNTs (3 to 5 for anxiety and OCD; 4 to 12 in MDD, depending on whether industry trials are included). It is important that clinicians and families be educated about possible adverse effects and their time course in order to anticipate difficulties, ensure adequate informed consent, and monitor appropriately. The black-box warning regarding treatment-emergent suicidal thoughts or behaviors must be discussed (for suggested talking points, see Box 2). The NNH is large (100 to 143), and for many patients, the benefits will outweigh the risks. For pediatric patients who fail to respond to multiple adequate trials of antidepressants and psychotherapy, referrals for interventional psychiatry consultation should be considered.

Bottom Line

Although the evidence base for prescribing antidepressants for children and adolescents is smaller compared to the adult literature, properly understanding and prescribing these agents, and explaining their risks and benefits to families, can make a major difference in patient compliance, satisfaction, and outcomes. Antidepressants (particularly selective serotonin reuptake inhibitors) are the firstline pharmacologic intervention for pediatric patients with anxiety disorders, obsessive-compulsive disorder, or major depressive disorder.

Related Resource

• Bonin L, Moreland CS. Overview of prevention and treatment for pediatric depression. https://www.uptodate.com/contents/overview-of-prevention-and-treatment-for-pediatric-depression. Last reviewed July 2019.

Drug Brand Names

Bupropion • Wellbutrin, Zyban
Cimetidine • Tagamet
Citalopram • Celexa
Clomipramine • Anafranil
Desipramine • Norpramin
Desvenlafaxine • Pristiq
Duloxetine • Cymbalta
Escitalopram • Lexapro
Fluoxetine • Prozac
Fluvoxamine • Luvox
Imipramine • Tofranil
Mirtazapine • Remeron
Nortriptyline • Pamelor
Paroxetine • Paxil
Sertraline • Zoloft
Venlafaxine • Effexor
Vilazodone • Viibryd
Vortioxetine • Trintellix

Box 1

Box 2

Major depressive disorder (MDD) is a significant pediatric health problem, with a lifetime prevalence as high as 20% by the end of adolescence.^1-3 Major depressive disorder in adolescence is associated with significant morbidity, including poor social functioning, school difficulties, early pregnancy, and increased risk of physical illness and substance abuse.^4-6 It is also linked with significant mortality, with increased risk for suicide, which is now the second leading cause of death in individuals age 10 to 24 years.^1,7,8

As their name suggests, antidepressants comprise a group of medications that are used to treat MDD; they are also, however, first-line agents for generalized anxiety disorder (GAD), posttraumatic stress disorder (PTSD), and obsessive-compulsive disorder (OCD) in adults. Anxiety disorders (including GAD and other anxiety diagnoses) and PTSD are also common in childhood and adolescence with a combined lifetime prevalence ranging from 15% to 30%.^9,10 These disorders are also associated with increased risk of suicide.¹¹ For all of these disorders, depending on the severity of presentation and the preference of the patient, treatments are often a combination of psychotherapy and psychopharmacology.

Clinicians face several challenges when considering antidepressants for pediatric patients. Pediatricians and psychiatrists need to understand whether these medications work in children and adolescents, and whether there are unique developmental safety and tolerability issues. The evidence base in child psychiatry is considerably smaller compared with that of adult psychiatry. From this more limited evidence base also came the controversial “black-box” warning regarding a risk of emergent suicidality when starting antidepressants that accompanies all antidepressants for pediatric, but not adult, patients. This warning has had major effects on clinical encounters with children experiencing depression, including altering clinician prescribing behavior.¹²

In this article, we review the current evidence for antidepressant efficacy, tolerability, and safety in pediatric patients. We also suggest ways in which clinicians might choose, start, and stop antidepressants in children, as well as how to talk with parents about benefits, risks, and the black-box warning.

Do antidepressants work in children?

Selective serotonin reuptake inhibitors. Selective serotonin reuptake inhibitors (SSRIs) are the most commonly used class of antidepressants in both children and adults.¹³ While only a few SSRIs are FDA-approved for pediatric indications, the lack of FDA approval is typically related to a lack of sufficient testing in randomized controlled trials (RCTs) for specific pediatric indications, rather than to demonstrable differences in efficacy between antidepressant agents. Since there is currently no data to suggest inferiority of one agent compared to another in children or adults,^14,15 efficacy data will be discussed here as applied to the class of SSRIs, generalizing from RCTs conducted on individual drugs. Table 1 lists FDA indications and dosing information for individual antidepressants.

There is strong evidence that SSRIs are effective for treating pediatric anxiety disorders (eg, social anxiety disorder and GAD)¹⁶ and OCD,¹⁷ with numbers needed to treat (NNT) between 3 and 5. For both of these disorders, SSRIs combined with cognitive-behavioral therapy (CBT) have the highest likelihood of improving symptoms or achieving remission.^17,18

Selective serotonin reuptake inhibitors are also effective for treating pediatric MDD; however, the literature is more complex for this disorder compared to GAD and OCD as there are considerable differences in effect sizes between National Institute of Mental Health (NIMH)–funded studies and industry-sponsored trials.¹³ The major NIMH-sponsored adolescent depression trial, TADS (Treatment for Adolescents and Depression Study), showed that SSRIs (fluoxetine in this case) were quite effective, with an NNT of 4 over the acute phase (12 weeks).¹⁹ Ultimately, approximately 80% of adolescents improved over 9 months. Many industry-sponsored trials for MDD in pediatric patients had large placebo response rates (approximately 60%), which resulted in smaller between-group differences, and estimates of an NNT closer to 12,¹³ which has muddied the waters in meta-analyses that include all trials.²⁰ Improvement in depressive symptoms also appears to be bolstered by concomitant CBT in MDD,¹⁹ but not as robustly as in GAD and OCD. While the full benefit of SSRIs for depression may take as long as 8 weeks, a meta-analysis of depression studies of pediatric patients suggests that significant benefits from placebo are observed as early as 2 weeks, and that further treatment gains are minimal after 4 weeks.¹⁵ Thus, we recommend at least a 4- to 6-week trial at therapeutic dosing before deeming a medication a treatment failure.

Continue to: Posttraumatic stress disorder...

Posttraumatic stress disorder is a fourth disorder in which SSRIs are a first-line treatment in adults. The data for using SSRIs to treat pediatric patients with PTSD is scant, with only a few RCTs, and no large NIMH-funded trials. Randomized controlled trials have not demonstrated significant differences between SSRIs and placebo^21,22 and thus the current first-line recommendation in pediatric PTSD remains trauma-focused therapy, with good evidence for trauma-focused CBT.²³ Practically speaking, there can be considerable overlap of PTSD, depression, and anxiety symptoms in children,²³ and children with a history of trauma who also have comorbid MDD may benefit from medication if their symptoms persist despite an adequate trial of psychotherapy.

Taken together, the current evidence suggests that SSRIs are often effective in pediatric GAD, OCD, and MDD, with low NNTs (ranging from 3 to 5 based on NIMH-funded trials) for all of these disorders; there is not yet sufficient evidence of efficacy in pediatric patients with PTSD.

Fluoxetine has been studied more intensively than other SSRIs (for example, it was the antidepressant used in the TADS trial), and thus has the largest evidence base. For this reason, fluoxetine is often considered the first of the first-line options. Additionally, fluoxetine has a longer half-life than other antidepressants, which may make it more effective in situations where patients are likely to miss doses, and results in a lower risk of withdrawal symptoms when stopped due to “self-tapering.”

SNRIs and atypical antidepressants. Other antidepressants commonly used in pediatric patients but with far less evidence of efficacy include serotonin-norepinephrine reuptake inhibitors (SNRIs) and the atypical antidepressants bupropion and mirtazapine. The SNRI duloxetine is FDA-approved for treating GAD in children age 7 to 17, but there are no other pediatric indications for duloxetine, or for the other SNRIs.

In general, adverse effect profiles are worse for SNRIs compared to SSRIs, further limiting their utility. While there are no pediatric studies demonstrating SNRI efficacy for neuropathic pain, good data exists in adults.²⁴ Thus, an SNRI could be a reasonable option if a pediatric patient has failed prior adequate SSRI trials and also has comorbid neuropathic pain.

Continue to: Neither bupropion nor mirtazapine...

Neither bupropion nor mirtazapine have undergone rigorous testing in pediatric patients, and therefore these agents should generally be considered only once other first-line treatments have failed. Bupropion has been evaluated for attention-deficit/hyperactivity disorder (ADHD)²⁵ and for adolescent smoking cessation.²⁶ However, the evidence is weak, and bupropion is not considered a first-line option for children and adolescents.

Tricyclic antidepressants. Randomized controlled trials have demonstrated that tricyclic antidepressants (TCAs) are efficacious for treating several pediatric conditions; however, their significant side effect profile, their monitoring requirements, as well as their lethality in overdose has left them replaced by SSRIs in most cases. That said, they can be appropriate in refractory ADHD (desipramine^27,28) and refractory OCD (clomipramine is FDA-approved for this indication²⁹); they are considered a third-line treatment for enuresis.³⁰

Why did my patient stop the medication?

Common adverse effects. Although the greatest benefit of antidepressant medications compared with placebo is achieved relatively early on in treatment, it generally takes time for these benefits to accrue and become clinically apparent.^15,31 By contrast, most adverse effects of antidepressants present and are at their most severe early in treatment. The combination of early adverse effects and delayed efficacy leads many patients, families, and clinicians to discontinue medications before they have an adequate chance to work. Thus, it is imperative to provide psychoeducation before starting a medication about the typical time-course of improvement and adverse effects (Table 2).

Adverse effects of SSRIs often appear or worsen transiently during initiation of a medication, during a dose increase,³² or, theoretically, with the addition of a medication that interferes with SSRI metabolism (eg, cimetidine inhibition of cytochrome P450 2D6).³³ If families are prepared for this phenomenon and the therapeutic alliance is adequate, adverse effects can be tolerated to allow for a full medication trial. Common adverse effects of SSRIs include sleep problems (insomnia/sedation), gastrointestinal upset, sexual dysfunction, dry mouth, and hyperhidrosis. Although SSRIs differ somewhat in the frequency of these effects, as a class, they are more similar than different. Adequate psychoeducation is especially imperative in the treatment of OCD and anxiety disorders, where there is limited evidence of efficacy for any non-serotonergic antidepressants.

Serotonin-norepinephrine reuptake inhibitors are not considered first-line medications because of the reduced evidence base compared to SSRIs and their enhanced adverse effect profiles. Because SNRIs partially share a mechanism of action with SSRIs, they also share portions of the adverse effects profile. However, SNRIs have the additional adverse effect of hypertension, which is related to their noradrenergic activity. Thus, it is reasonable to obtain a baseline blood pressure before initiating an SNRI, as well as periodically after initiation and during dose increases, particularly if the patient has other risk factors for hypertension.³⁴

Continue to: Although TCAs have efficacy...

Although TCAs have efficacy in some pediatric disorders,^27-29,35 their adverse effect profile limits their use. Tricyclic antidepressants are highly anticholinergic (causing dizziness secondary to orthostatic hypotension, dry mouth, and urinary retention) and antihistaminergic (causing sedation and weight gain). Additionally, TCAs lower the seizure threshold and have adverse cardiac effects relating to their anti-alpha-1 adrenergic activity, resulting in dose-dependent increases in the QTc and cardiac toxicity in overdose that could lead to arrhythmia and death. These medications have their place, but their use requires careful informed consent, clear treatment goals, and baseline and periodic cardiac monitoring (via electrocardiogram).

Serious adverse effects. Clinicians may be hesitant to prescribe antidepressants for pediatric patients because of the potential for more serious adverse effects, including severe behavioral activation syndromes, serotonin syndrome, and emergent suicidality. However, current FDA-approved antidepressants arguably have one of the most positive risk/benefit profiles of any orally-administered medication approved for pediatric patients. Having a strong understanding of the evidence is critical to evaluating when it is appropriate to prescribe an antidepressant, how to properly monitor the patient, and how to obtain accurate informed consent.

Pediatric behavioral activation syndrome. Many clinicians report that children receiving antidepressants experience a pediatric behavioral activation syndrome, which exists along a spectrum from mild activation, increased energy, insomnia, or irritability up through more severe presentations of agitation, hyperactivity, or possibly mania. A recent meta-analysis suggested a positive association between antidepressant use and activation events on the milder end of this spectrum in pediatric patients with non-OCD anxiety disorders,¹⁶ and it is thought that compared with adolescents, younger children are more susceptible to activation adverse effects.³⁶ The likelihood of activation events has been associated with higher antidepressant plasma levels,³⁷ suggesting that dose or individual differences in metabolism may play a role. At the severe end of the spectrum, the risk of induction of mania in pediatric patients with depression or anxiety is relatively rare (<2%) and not statistically different from placebo in RCTs of pediatric participants.³⁸ Meta-analyses of larger randomized, placebo-controlled trials of adults do not support the idea that SSRIs and other second-generation antidepressants carry an increased risk of mania compared with placebo.^39,40 Children or adolescents with bona fide bipolar disorder (ie, patients who have had observed mania that meets all DSM-5 criteria) should be treated with a mood-stabilizing agent or antipsychotic if prescribed an antidepressant.⁴¹ These clear-cut cases are, however, relatively rare, and more often clinicians are confronted with ambiguous cases that include a family history of bipolar disorder along with “softer” symptoms of irritability, intrusiveness, or aggression. In these children, SSRIs may be appropriate for depressive, OCD, or anxiety symptoms, and should be strongly considered before prescribing antipsychotics or mood stabilizers, as long as initiated with proper monitoring.

Serotonin syndrome is a life-threatening condition caused by excess synaptic serotonin. It is characterized by confusion, sweating, diarrhea, hypertension, hyperthermia, and tachycardia. At its most severe, serotonin syndrome can result in seizures, arrhythmias, and death. The risk of serotonin syndrome is very low when using an SSRI as monotherapy. Risk increases with polypharmacy, particularly unexamined polypharmacy when multiple serotonergic agents are inadvertently on board. Commonly used serotonergic agents include other antidepressants, migraine medications (eg, triptans), some pain medications, and the cough suppressant dextromethorphan.

The easiest way to mitigate the risk of serotonin syndrome is to use an interaction index computer program, which can help ensure that the interacting agents are not prescribed without first discussing the risks and benefits. It is important to teach adolescents that certain recreational drugs are highly serotonergic and can cause serious interactions with antidepressants. For example, recreational use of dextrometh­orphan or 3,4-methylenedioxymethamphetamine (MDMA; commonly known as “ecstasy”) has been associated with serotonin syndrome in adolescents taking antidepressant medications.^42,43

Continue to: Suicidality

Suicidality. The black-box warning regarding a risk of emergent suicidality when starting antidepressant treatment in children is controversial.⁴⁴ The prospect that a medication intended to ameliorate depression might instead risk increasing suicidal thinking is alarming to parents and clinicians alike. To appropriately weigh and discuss the risks and benefits with families, it is important to understand the data upon which the warning is based.

In 2004, the FDA commissioned a review of 23 antidepressant trials, both published and unpublished, pooling studies across multiple indications (MDD, OCD, anxiety, and ADHD) and multiple antidepressant classes. This meta-analysis, which included nearly 4,400 pediatric patients, found a small but statistically significant increase in spontaneously-reported suicidal thoughts or actions, with a risk difference of 1% (95% confidence interval [CI], 1% to 2%).⁴⁵ These data suggest that if one treats 100 pediatric patients, 1 to 2 of them may experience short-term increases in suicidal thinking or behavior.⁴⁵ There were no differences in suicidal thinking when assessed systematically (ie, when all subjects reported symptoms of suicidal ideation on structured rating scales), and there were no completed suicides.⁴⁵ A subsequent analysis that included 27 pediatric trials suggested an even lower, although still significant, risk difference (<1%), yielding a number needed to harm (NNH) of 143.⁴⁶ Thus, with low NNT for efficacy (3 to 6) and relatively high NNH for emergent suicidal thoughts or behaviors (100 to 143), for many patients the benefits will outweigh the risks.

Figure 1, Figure 2, and Figure 3 are Cates plots that depict the absolute benefits of antidepressants compared with the risk of suicidality for pediatric patients with MDD, OCD, and anxiety disorders. Recent meta-analyses have suggested that the increased risk of suicidality in antidepressant trials is specific to studies that included children and adolescents, and is not observed in adult studies. A meta-analysis of 70 trials involving 18,526 participants suggested that the odds ratio of suicidality in trials of children and adolescents was 2.39 (95% CI, 1.31 to 4.33) compared with 0.81 (95% CI, 0.51 to 1.28) in adults.⁴⁷ Additionally, a network meta-analysis exclusively focusing on pediatric antidepressant trials in MDD reported significantly higher suicidality-related adverse events in venlafaxine trials compared with placebo, duloxetine, and several SSRIs (fluoxetine, paroxetine, and escitalopram).²⁰ These data should be interpreted with caution as differences in suicidality detected between agents is quite possibly related to differences in the method of assessment between trials, as opposed to actual differences in risk between agents.

Epidemiologic data further support the use of antidepressants in pediatric patients, showing that antidepressant use is associated with decreased teen suicide attempts and completions,⁴⁸ and the decline in prescriptions that occurred following the black-box warning was accompanied by a 14% increase in teen suicides.⁴⁹ Multiple hypotheses have been proposed to explain the pediatric clinical trial findings. One idea is that potential adverse effects of activation, or the intended effects of restoring the motivation, energy, and social engagement that is often impaired in depression, increases the likelihood of thinking about suicide or acting on thoughts. Another theory is that reporting of suicidality may be increased, rather than increased de novo suicidality itself. Antidepressants are effective for treating pediatric anxiety disorders, including social anxiety disorder,¹⁶ which could result in more willingness to report. Also, the manner in which adverse effects are generally ascertained in trials might have led to increased spontaneous reporting. In many trials, investigators ask whether participants have any adverse effects in general, and inquire about specific adverse effects only if the family answers affirmatively. Thus, the increased rate of other adverse effects associated with antidepressants (sleep problems, gastrointestinal upset, dry mouth, etc.) might trigger a specific question regarding suicidal ideation, which the child or family then may be more likely to report. Alternatively, any type of psychiatric treatment could increase an individual’s propensity to report; in adolescent psychotherapy trials, non-medicated participants have reported emergent suicidality at similar frequencies as those described in drug trials.⁵⁰ Regardless of the mechanism, the possibility of treatment-emergent suicidality is a low-frequency but serious event that necessitates careful monitoring when starting medication. Current guidelines suggest seeing children weekly for the first month after medication initiation, every 2 weeks for the following month, and monthly thereafter.⁵¹

Continue to: How long should the antidepressant be continued?

How long should the antidepressant be continued?

Many patients are concerned about how long they may be taking medication, and whether they will be taking an antidepressant “forever.” A treatment course can be broken into an acute phase, wherein remission is achieved and maintained for 6 to 8 weeks. This is followed by a continuation phase, with the goal of relapse prevention, lasting 16 to 20 weeks. The length of the last phase—the maintenance phase—depends both on the child’s history, the underlying therapeutic indication, the adverse effect burden experienced, and the family’s preferences/values. In general, for a first depressive episode, after treating for 1 year, a trial of discontinuation can be attempted with close monitoring. For a second depressive episode, we recommend 2 years of remission on antidepressant therapy before attempting discontinuation. In patients who have had 3 depressive episodes, or have had episodes of high severity, we recommend continuing antidepressant treatment indefinitely. Although much less well studied, the risk of relapse following SSRI discontinuation appears much more significant in OCD, whereas anxiety disorders and MDD have a relatively comparable risk.⁵²

In general, stopping an antidepressant should be done carefully and slowly. The speed with which a specific antidepressant can be stopped is largely related to its half-life. Agents with very long half-lives, such as fluoxetine (half-life of 5 days for the parent compound and 9 days for active metabolite), can often be stopped altogether, being “auto-tapered” by the long half-life. One might still consider a taper if the patient were taking high doses. Medications with shorter half-lives must be more carefully tapered to avoid discontinuation syndromes. Discontinuation syndromes are characterized by flu-like symptoms (nausea, myalgias, fatigue, dizziness) and worsening mood. Medications with short half-lives (eg, paroxetine and venlafaxine) have the highest potential for this syndrome in children,⁵³ and thus are used less frequently.

What to do when first-line treatments fail

When a child does not experience sufficient improvement from first-line treatments, it is crucial to determine whether they have experienced an adequate dosing, duration, and quality of medication and psychotherapy.

Adequate psychotherapy? To determine whether children are receiving adequate CBT, ask:

1. if the child receives homework from psychotherapy
2. if the parents are included in treatment
3. if therapy has involved identifying thought patterns that may be contributing to the child’s illness, and
4. if the therapist has ever exposed the child to a challenge likely to produce anxiety or distress in a supervised environment and has developed an exposure hierarchy (for conditions with primarily exposure-based therapies, such as OCD or anxiety disorders).

If a family is not receiving most of these elements in psychotherapy, this is a good indicator that they may not be receiving evidence-based CBT.

Continue to: Adequate pharmacotherapy?

Adequate pharmacotherapy? Similarly, when determining the adequacy of previous pharmacotherapy, it is critical to determine whether the child received an adequate dose of medications (at least the FDA-recommended minimum dose) for an adequate duration of time at therapeutic dosing (at least 6 weeks for MDD, 8 weeks for anxiety disorders, and 8 to 12 weeks for pediatric patients with OCD), and that the child actually took the medication regularly during that period. Patient compliance can typically be tracked through checking refill requests or intervals through the patient’s pharmacy. Ensuring proper delivery of first-line treatments is imperative because (1) the adverse effects associated with second-line treatments are often more substantial; (2) the cost in terms of time and money is considerably higher with second-line treatments, and; (3) the evidence regarding the benefits of these treatments is much less certain.

Inadequate dosing is a common reason for non-response in pediatric patients. Therapeutic dose ranges for common antidepressants are displayed in Table 1. Many clinicians underdose antidepressants for pediatric patients initially (and often throughout treatment) because they fear that the typical dose titration used in clinical trials will increase the risk of adverse effects compared with more conservative dosing. There is limited evidence to suggest that this underdosing strategy is likely to be successful; adverse effects attributable to these medications are modest, and most that are experienced early in treatment (eg, headache, increased anxiety or irritability, sleep problems, gastrointestinal upset) are self-limiting and may be coincidental rather than medication-induced. Furthermore, there is no evidence for efficacy of subtherapeutic dosing in children in the acute phase of treatment or for preventing relapse.¹⁴ Thus, from an efficacy standpoint, a medication trial has not officially begun until the therapeutic dose range is reached.

Once dosing is within the therapeutic range, however, pediatric data differs from the adult literature. In most adult psychi­atric conditions, higher doses of SSRIs within the therapeutic range are associated with an increased response rate.^14,54 In pediatrics, there are few fixed dose trials, and once within the recommended therapeutic range, minimal data supports an association between higher dosing and higher efficacy.¹⁴ In general, pediatric guidelines are silent regarding optimal dosing of SSRIs within the recommended dose range, and higher antidepressant doses often result in a more significant adverse effect burden for children. One exception is pediatric OCD, where, similar to adults, the guidelines suggest that higher dosing of SSRIs often is required to induce a therapeutic response as compared to MDD and GAD.^31,55

If a child does not respond to adequate first-line treatment (or has a treatment history that cannot be fully verified), repeating these first-line interventions carries little risk and can be quite effective. For example, 60% of adolescents with MDD who did not initially respond to an SSRI demonstrated a significant response when prescribed a second SSRI or venlafaxine (with or without CBT).⁵⁶

When pediatric patients continue to experience significantly distressing and/or debilitating symptoms (particularly in MDD) despite multiple trials of antidepressants and psychotherapy, practitioners should consider a careful referral to interventional psychiatry services, which can include the more intensive treatments of electroconvulsive therapy, repetitive transcranial magnetic stimulation, or ketamine (see Box 1). Given the substantial morbidity and mortality associated with adolescent depression, interventional psychiatry treatments are under-researched and under-utilized clinically in pediatric populations.

Continue to: Antidepressants in general...

Antidepressants in general, and SSRIs in particular, are the first-line pharmacotherapy for pediatric anxiety, OCD, and MDD. For PTSD, although they are a first-line treatment in adults, their efficacy has not been demonstrated in children and adolescents. Antidepressants are generally safe, well-tolerated, and effective, with low NNTs (3 to 5 for anxiety and OCD; 4 to 12 in MDD, depending on whether industry trials are included). It is important that clinicians and families be educated about possible adverse effects and their time course in order to anticipate difficulties, ensure adequate informed consent, and monitor appropriately. The black-box warning regarding treatment-emergent suicidal thoughts or behaviors must be discussed (for suggested talking points, see Box 2). The NNH is large (100 to 143), and for many patients, the benefits will outweigh the risks. For pediatric patients who fail to respond to multiple adequate trials of antidepressants and psychotherapy, referrals for interventional psychiatry consultation should be considered.

Bottom Line

Although the evidence base for prescribing antidepressants for children and adolescents is smaller compared to the adult literature, properly understanding and prescribing these agents, and explaining their risks and benefits to families, can make a major difference in patient compliance, satisfaction, and outcomes. Antidepressants (particularly selective serotonin reuptake inhibitors) are the firstline pharmacologic intervention for pediatric patients with anxiety disorders, obsessive-compulsive disorder, or major depressive disorder.

Related Resource

• Bonin L, Moreland CS. Overview of prevention and treatment for pediatric depression. https://www.uptodate.com/contents/overview-of-prevention-and-treatment-for-pediatric-depression. Last reviewed July 2019.

Drug Brand Names

Bupropion • Wellbutrin, Zyban
Cimetidine • Tagamet
Citalopram • Celexa
Clomipramine • Anafranil
Desipramine • Norpramin
Desvenlafaxine • Pristiq
Duloxetine • Cymbalta
Escitalopram • Lexapro
Fluoxetine • Prozac
Fluvoxamine • Luvox
Imipramine • Tofranil
Mirtazapine • Remeron
Nortriptyline • Pamelor
Paroxetine • Paxil
Sertraline • Zoloft
Venlafaxine • Effexor
Vilazodone • Viibryd
Vortioxetine • Trintellix

Box 1

Box 2

Major depressive disorder (MDD) is a significant pediatric health problem, with a lifetime prevalence as high as 20% by the end of adolescence.^1-3 Major depressive disorder in adolescence is associated with significant morbidity, including poor social functioning, school difficulties, early pregnancy, and increased risk of physical illness and substance abuse.^4-6 It is also linked with significant mortality, with increased risk for suicide, which is now the second leading cause of death in individuals age 10 to 24 years.^1,7,8

As their name suggests, antidepressants comprise a group of medications that are used to treat MDD; they are also, however, first-line agents for generalized anxiety disorder (GAD), posttraumatic stress disorder (PTSD), and obsessive-compulsive disorder (OCD) in adults. Anxiety disorders (including GAD and other anxiety diagnoses) and PTSD are also common in childhood and adolescence with a combined lifetime prevalence ranging from 15% to 30%.^9,10 These disorders are also associated with increased risk of suicide.¹¹ For all of these disorders, depending on the severity of presentation and the preference of the patient, treatments are often a combination of psychotherapy and psychopharmacology.

Clinicians face several challenges when considering antidepressants for pediatric patients. Pediatricians and psychiatrists need to understand whether these medications work in children and adolescents, and whether there are unique developmental safety and tolerability issues. The evidence base in child psychiatry is considerably smaller compared with that of adult psychiatry. From this more limited evidence base also came the controversial “black-box” warning regarding a risk of emergent suicidality when starting antidepressants that accompanies all antidepressants for pediatric, but not adult, patients. This warning has had major effects on clinical encounters with children experiencing depression, including altering clinician prescribing behavior.¹²

In this article, we review the current evidence for antidepressant efficacy, tolerability, and safety in pediatric patients. We also suggest ways in which clinicians might choose, start, and stop antidepressants in children, as well as how to talk with parents about benefits, risks, and the black-box warning.

Do antidepressants work in children?

Selective serotonin reuptake inhibitors. Selective serotonin reuptake inhibitors (SSRIs) are the most commonly used class of antidepressants in both children and adults.¹³ While only a few SSRIs are FDA-approved for pediatric indications, the lack of FDA approval is typically related to a lack of sufficient testing in randomized controlled trials (RCTs) for specific pediatric indications, rather than to demonstrable differences in efficacy between antidepressant agents. Since there is currently no data to suggest inferiority of one agent compared to another in children or adults,^14,15 efficacy data will be discussed here as applied to the class of SSRIs, generalizing from RCTs conducted on individual drugs. Table 1 lists FDA indications and dosing information for individual antidepressants.

There is strong evidence that SSRIs are effective for treating pediatric anxiety disorders (eg, social anxiety disorder and GAD)¹⁶ and OCD,¹⁷ with numbers needed to treat (NNT) between 3 and 5. For both of these disorders, SSRIs combined with cognitive-behavioral therapy (CBT) have the highest likelihood of improving symptoms or achieving remission.^17,18

Selective serotonin reuptake inhibitors are also effective for treating pediatric MDD; however, the literature is more complex for this disorder compared to GAD and OCD as there are considerable differences in effect sizes between National Institute of Mental Health (NIMH)–funded studies and industry-sponsored trials.¹³ The major NIMH-sponsored adolescent depression trial, TADS (Treatment for Adolescents and Depression Study), showed that SSRIs (fluoxetine in this case) were quite effective, with an NNT of 4 over the acute phase (12 weeks).¹⁹ Ultimately, approximately 80% of adolescents improved over 9 months. Many industry-sponsored trials for MDD in pediatric patients had large placebo response rates (approximately 60%), which resulted in smaller between-group differences, and estimates of an NNT closer to 12,¹³ which has muddied the waters in meta-analyses that include all trials.²⁰ Improvement in depressive symptoms also appears to be bolstered by concomitant CBT in MDD,¹⁹ but not as robustly as in GAD and OCD. While the full benefit of SSRIs for depression may take as long as 8 weeks, a meta-analysis of depression studies of pediatric patients suggests that significant benefits from placebo are observed as early as 2 weeks, and that further treatment gains are minimal after 4 weeks.¹⁵ Thus, we recommend at least a 4- to 6-week trial at therapeutic dosing before deeming a medication a treatment failure.

Continue to: Posttraumatic stress disorder...

Posttraumatic stress disorder is a fourth disorder in which SSRIs are a first-line treatment in adults. The data for using SSRIs to treat pediatric patients with PTSD is scant, with only a few RCTs, and no large NIMH-funded trials. Randomized controlled trials have not demonstrated significant differences between SSRIs and placebo^21,22 and thus the current first-line recommendation in pediatric PTSD remains trauma-focused therapy, with good evidence for trauma-focused CBT.²³ Practically speaking, there can be considerable overlap of PTSD, depression, and anxiety symptoms in children,²³ and children with a history of trauma who also have comorbid MDD may benefit from medication if their symptoms persist despite an adequate trial of psychotherapy.

Taken together, the current evidence suggests that SSRIs are often effective in pediatric GAD, OCD, and MDD, with low NNTs (ranging from 3 to 5 based on NIMH-funded trials) for all of these disorders; there is not yet sufficient evidence of efficacy in pediatric patients with PTSD.

Fluoxetine has been studied more intensively than other SSRIs (for example, it was the antidepressant used in the TADS trial), and thus has the largest evidence base. For this reason, fluoxetine is often considered the first of the first-line options. Additionally, fluoxetine has a longer half-life than other antidepressants, which may make it more effective in situations where patients are likely to miss doses, and results in a lower risk of withdrawal symptoms when stopped due to “self-tapering.”

SNRIs and atypical antidepressants. Other antidepressants commonly used in pediatric patients but with far less evidence of efficacy include serotonin-norepinephrine reuptake inhibitors (SNRIs) and the atypical antidepressants bupropion and mirtazapine. The SNRI duloxetine is FDA-approved for treating GAD in children age 7 to 17, but there are no other pediatric indications for duloxetine, or for the other SNRIs.

In general, adverse effect profiles are worse for SNRIs compared to SSRIs, further limiting their utility. While there are no pediatric studies demonstrating SNRI efficacy for neuropathic pain, good data exists in adults.²⁴ Thus, an SNRI could be a reasonable option if a pediatric patient has failed prior adequate SSRI trials and also has comorbid neuropathic pain.

Continue to: Neither bupropion nor mirtazapine...

Neither bupropion nor mirtazapine have undergone rigorous testing in pediatric patients, and therefore these agents should generally be considered only once other first-line treatments have failed. Bupropion has been evaluated for attention-deficit/hyperactivity disorder (ADHD)²⁵ and for adolescent smoking cessation.²⁶ However, the evidence is weak, and bupropion is not considered a first-line option for children and adolescents.

Tricyclic antidepressants. Randomized controlled trials have demonstrated that tricyclic antidepressants (TCAs) are efficacious for treating several pediatric conditions; however, their significant side effect profile, their monitoring requirements, as well as their lethality in overdose has left them replaced by SSRIs in most cases. That said, they can be appropriate in refractory ADHD (desipramine^27,28) and refractory OCD (clomipramine is FDA-approved for this indication²⁹); they are considered a third-line treatment for enuresis.³⁰

Why did my patient stop the medication?

Common adverse effects. Although the greatest benefit of antidepressant medications compared with placebo is achieved relatively early on in treatment, it generally takes time for these benefits to accrue and become clinically apparent.^15,31 By contrast, most adverse effects of antidepressants present and are at their most severe early in treatment. The combination of early adverse effects and delayed efficacy leads many patients, families, and clinicians to discontinue medications before they have an adequate chance to work. Thus, it is imperative to provide psychoeducation before starting a medication about the typical time-course of improvement and adverse effects (Table 2).

Adverse effects of SSRIs often appear or worsen transiently during initiation of a medication, during a dose increase,³² or, theoretically, with the addition of a medication that interferes with SSRI metabolism (eg, cimetidine inhibition of cytochrome P450 2D6).³³ If families are prepared for this phenomenon and the therapeutic alliance is adequate, adverse effects can be tolerated to allow for a full medication trial. Common adverse effects of SSRIs include sleep problems (insomnia/sedation), gastrointestinal upset, sexual dysfunction, dry mouth, and hyperhidrosis. Although SSRIs differ somewhat in the frequency of these effects, as a class, they are more similar than different. Adequate psychoeducation is especially imperative in the treatment of OCD and anxiety disorders, where there is limited evidence of efficacy for any non-serotonergic antidepressants.

Serotonin-norepinephrine reuptake inhibitors are not considered first-line medications because of the reduced evidence base compared to SSRIs and their enhanced adverse effect profiles. Because SNRIs partially share a mechanism of action with SSRIs, they also share portions of the adverse effects profile. However, SNRIs have the additional adverse effect of hypertension, which is related to their noradrenergic activity. Thus, it is reasonable to obtain a baseline blood pressure before initiating an SNRI, as well as periodically after initiation and during dose increases, particularly if the patient has other risk factors for hypertension.³⁴

Continue to: Although TCAs have efficacy...

Although TCAs have efficacy in some pediatric disorders,^27-29,35 their adverse effect profile limits their use. Tricyclic antidepressants are highly anticholinergic (causing dizziness secondary to orthostatic hypotension, dry mouth, and urinary retention) and antihistaminergic (causing sedation and weight gain). Additionally, TCAs lower the seizure threshold and have adverse cardiac effects relating to their anti-alpha-1 adrenergic activity, resulting in dose-dependent increases in the QTc and cardiac toxicity in overdose that could lead to arrhythmia and death. These medications have their place, but their use requires careful informed consent, clear treatment goals, and baseline and periodic cardiac monitoring (via electrocardiogram).

Serious adverse effects. Clinicians may be hesitant to prescribe antidepressants for pediatric patients because of the potential for more serious adverse effects, including severe behavioral activation syndromes, serotonin syndrome, and emergent suicidality. However, current FDA-approved antidepressants arguably have one of the most positive risk/benefit profiles of any orally-administered medication approved for pediatric patients. Having a strong understanding of the evidence is critical to evaluating when it is appropriate to prescribe an antidepressant, how to properly monitor the patient, and how to obtain accurate informed consent.

Pediatric behavioral activation syndrome. Many clinicians report that children receiving antidepressants experience a pediatric behavioral activation syndrome, which exists along a spectrum from mild activation, increased energy, insomnia, or irritability up through more severe presentations of agitation, hyperactivity, or possibly mania. A recent meta-analysis suggested a positive association between antidepressant use and activation events on the milder end of this spectrum in pediatric patients with non-OCD anxiety disorders,¹⁶ and it is thought that compared with adolescents, younger children are more susceptible to activation adverse effects.³⁶ The likelihood of activation events has been associated with higher antidepressant plasma levels,³⁷ suggesting that dose or individual differences in metabolism may play a role. At the severe end of the spectrum, the risk of induction of mania in pediatric patients with depression or anxiety is relatively rare (<2%) and not statistically different from placebo in RCTs of pediatric participants.³⁸ Meta-analyses of larger randomized, placebo-controlled trials of adults do not support the idea that SSRIs and other second-generation antidepressants carry an increased risk of mania compared with placebo.^39,40 Children or adolescents with bona fide bipolar disorder (ie, patients who have had observed mania that meets all DSM-5 criteria) should be treated with a mood-stabilizing agent or antipsychotic if prescribed an antidepressant.⁴¹ These clear-cut cases are, however, relatively rare, and more often clinicians are confronted with ambiguous cases that include a family history of bipolar disorder along with “softer” symptoms of irritability, intrusiveness, or aggression. In these children, SSRIs may be appropriate for depressive, OCD, or anxiety symptoms, and should be strongly considered before prescribing antipsychotics or mood stabilizers, as long as initiated with proper monitoring.

Serotonin syndrome is a life-threatening condition caused by excess synaptic serotonin. It is characterized by confusion, sweating, diarrhea, hypertension, hyperthermia, and tachycardia. At its most severe, serotonin syndrome can result in seizures, arrhythmias, and death. The risk of serotonin syndrome is very low when using an SSRI as monotherapy. Risk increases with polypharmacy, particularly unexamined polypharmacy when multiple serotonergic agents are inadvertently on board. Commonly used serotonergic agents include other antidepressants, migraine medications (eg, triptans), some pain medications, and the cough suppressant dextromethorphan.

The easiest way to mitigate the risk of serotonin syndrome is to use an interaction index computer program, which can help ensure that the interacting agents are not prescribed without first discussing the risks and benefits. It is important to teach adolescents that certain recreational drugs are highly serotonergic and can cause serious interactions with antidepressants. For example, recreational use of dextrometh­orphan or 3,4-methylenedioxymethamphetamine (MDMA; commonly known as “ecstasy”) has been associated with serotonin syndrome in adolescents taking antidepressant medications.^42,43

Continue to: Suicidality

Suicidality. The black-box warning regarding a risk of emergent suicidality when starting antidepressant treatment in children is controversial.⁴⁴ The prospect that a medication intended to ameliorate depression might instead risk increasing suicidal thinking is alarming to parents and clinicians alike. To appropriately weigh and discuss the risks and benefits with families, it is important to understand the data upon which the warning is based.

In 2004, the FDA commissioned a review of 23 antidepressant trials, both published and unpublished, pooling studies across multiple indications (MDD, OCD, anxiety, and ADHD) and multiple antidepressant classes. This meta-analysis, which included nearly 4,400 pediatric patients, found a small but statistically significant increase in spontaneously-reported suicidal thoughts or actions, with a risk difference of 1% (95% confidence interval [CI], 1% to 2%).⁴⁵ These data suggest that if one treats 100 pediatric patients, 1 to 2 of them may experience short-term increases in suicidal thinking or behavior.⁴⁵ There were no differences in suicidal thinking when assessed systematically (ie, when all subjects reported symptoms of suicidal ideation on structured rating scales), and there were no completed suicides.⁴⁵ A subsequent analysis that included 27 pediatric trials suggested an even lower, although still significant, risk difference (<1%), yielding a number needed to harm (NNH) of 143.⁴⁶ Thus, with low NNT for efficacy (3 to 6) and relatively high NNH for emergent suicidal thoughts or behaviors (100 to 143), for many patients the benefits will outweigh the risks.

Figure 1, Figure 2, and Figure 3 are Cates plots that depict the absolute benefits of antidepressants compared with the risk of suicidality for pediatric patients with MDD, OCD, and anxiety disorders. Recent meta-analyses have suggested that the increased risk of suicidality in antidepressant trials is specific to studies that included children and adolescents, and is not observed in adult studies. A meta-analysis of 70 trials involving 18,526 participants suggested that the odds ratio of suicidality in trials of children and adolescents was 2.39 (95% CI, 1.31 to 4.33) compared with 0.81 (95% CI, 0.51 to 1.28) in adults.⁴⁷ Additionally, a network meta-analysis exclusively focusing on pediatric antidepressant trials in MDD reported significantly higher suicidality-related adverse events in venlafaxine trials compared with placebo, duloxetine, and several SSRIs (fluoxetine, paroxetine, and escitalopram).²⁰ These data should be interpreted with caution as differences in suicidality detected between agents is quite possibly related to differences in the method of assessment between trials, as opposed to actual differences in risk between agents.

Epidemiologic data further support the use of antidepressants in pediatric patients, showing that antidepressant use is associated with decreased teen suicide attempts and completions,⁴⁸ and the decline in prescriptions that occurred following the black-box warning was accompanied by a 14% increase in teen suicides.⁴⁹ Multiple hypotheses have been proposed to explain the pediatric clinical trial findings. One idea is that potential adverse effects of activation, or the intended effects of restoring the motivation, energy, and social engagement that is often impaired in depression, increases the likelihood of thinking about suicide or acting on thoughts. Another theory is that reporting of suicidality may be increased, rather than increased de novo suicidality itself. Antidepressants are effective for treating pediatric anxiety disorders, including social anxiety disorder,¹⁶ which could result in more willingness to report. Also, the manner in which adverse effects are generally ascertained in trials might have led to increased spontaneous reporting. In many trials, investigators ask whether participants have any adverse effects in general, and inquire about specific adverse effects only if the family answers affirmatively. Thus, the increased rate of other adverse effects associated with antidepressants (sleep problems, gastrointestinal upset, dry mouth, etc.) might trigger a specific question regarding suicidal ideation, which the child or family then may be more likely to report. Alternatively, any type of psychiatric treatment could increase an individual’s propensity to report; in adolescent psychotherapy trials, non-medicated participants have reported emergent suicidality at similar frequencies as those described in drug trials.⁵⁰ Regardless of the mechanism, the possibility of treatment-emergent suicidality is a low-frequency but serious event that necessitates careful monitoring when starting medication. Current guidelines suggest seeing children weekly for the first month after medication initiation, every 2 weeks for the following month, and monthly thereafter.⁵¹

Continue to: How long should the antidepressant be continued?

How long should the antidepressant be continued?

Many patients are concerned about how long they may be taking medication, and whether they will be taking an antidepressant “forever.” A treatment course can be broken into an acute phase, wherein remission is achieved and maintained for 6 to 8 weeks. This is followed by a continuation phase, with the goal of relapse prevention, lasting 16 to 20 weeks. The length of the last phase—the maintenance phase—depends both on the child’s history, the underlying therapeutic indication, the adverse effect burden experienced, and the family’s preferences/values. In general, for a first depressive episode, after treating for 1 year, a trial of discontinuation can be attempted with close monitoring. For a second depressive episode, we recommend 2 years of remission on antidepressant therapy before attempting discontinuation. In patients who have had 3 depressive episodes, or have had episodes of high severity, we recommend continuing antidepressant treatment indefinitely. Although much less well studied, the risk of relapse following SSRI discontinuation appears much more significant in OCD, whereas anxiety disorders and MDD have a relatively comparable risk.⁵²

In general, stopping an antidepressant should be done carefully and slowly. The speed with which a specific antidepressant can be stopped is largely related to its half-life. Agents with very long half-lives, such as fluoxetine (half-life of 5 days for the parent compound and 9 days for active metabolite), can often be stopped altogether, being “auto-tapered” by the long half-life. One might still consider a taper if the patient were taking high doses. Medications with shorter half-lives must be more carefully tapered to avoid discontinuation syndromes. Discontinuation syndromes are characterized by flu-like symptoms (nausea, myalgias, fatigue, dizziness) and worsening mood. Medications with short half-lives (eg, paroxetine and venlafaxine) have the highest potential for this syndrome in children,⁵³ and thus are used less frequently.

What to do when first-line treatments fail

When a child does not experience sufficient improvement from first-line treatments, it is crucial to determine whether they have experienced an adequate dosing, duration, and quality of medication and psychotherapy.

Adequate psychotherapy? To determine whether children are receiving adequate CBT, ask:

1. if the child receives homework from psychotherapy
2. if the parents are included in treatment
3. if therapy has involved identifying thought patterns that may be contributing to the child’s illness, and
4. if the therapist has ever exposed the child to a challenge likely to produce anxiety or distress in a supervised environment and has developed an exposure hierarchy (for conditions with primarily exposure-based therapies, such as OCD or anxiety disorders).

If a family is not receiving most of these elements in psychotherapy, this is a good indicator that they may not be receiving evidence-based CBT.

Continue to: Adequate pharmacotherapy?

Adequate pharmacotherapy? Similarly, when determining the adequacy of previous pharmacotherapy, it is critical to determine whether the child received an adequate dose of medications (at least the FDA-recommended minimum dose) for an adequate duration of time at therapeutic dosing (at least 6 weeks for MDD, 8 weeks for anxiety disorders, and 8 to 12 weeks for pediatric patients with OCD), and that the child actually took the medication regularly during that period. Patient compliance can typically be tracked through checking refill requests or intervals through the patient’s pharmacy. Ensuring proper delivery of first-line treatments is imperative because (1) the adverse effects associated with second-line treatments are often more substantial; (2) the cost in terms of time and money is considerably higher with second-line treatments, and; (3) the evidence regarding the benefits of these treatments is much less certain.

Inadequate dosing is a common reason for non-response in pediatric patients. Therapeutic dose ranges for common antidepressants are displayed in Table 1. Many clinicians underdose antidepressants for pediatric patients initially (and often throughout treatment) because they fear that the typical dose titration used in clinical trials will increase the risk of adverse effects compared with more conservative dosing. There is limited evidence to suggest that this underdosing strategy is likely to be successful; adverse effects attributable to these medications are modest, and most that are experienced early in treatment (eg, headache, increased anxiety or irritability, sleep problems, gastrointestinal upset) are self-limiting and may be coincidental rather than medication-induced. Furthermore, there is no evidence for efficacy of subtherapeutic dosing in children in the acute phase of treatment or for preventing relapse.¹⁴ Thus, from an efficacy standpoint, a medication trial has not officially begun until the therapeutic dose range is reached.

Once dosing is within the therapeutic range, however, pediatric data differs from the adult literature. In most adult psychi­atric conditions, higher doses of SSRIs within the therapeutic range are associated with an increased response rate.^14,54 In pediatrics, there are few fixed dose trials, and once within the recommended therapeutic range, minimal data supports an association between higher dosing and higher efficacy.¹⁴ In general, pediatric guidelines are silent regarding optimal dosing of SSRIs within the recommended dose range, and higher antidepressant doses often result in a more significant adverse effect burden for children. One exception is pediatric OCD, where, similar to adults, the guidelines suggest that higher dosing of SSRIs often is required to induce a therapeutic response as compared to MDD and GAD.^31,55

If a child does not respond to adequate first-line treatment (or has a treatment history that cannot be fully verified), repeating these first-line interventions carries little risk and can be quite effective. For example, 60% of adolescents with MDD who did not initially respond to an SSRI demonstrated a significant response when prescribed a second SSRI or venlafaxine (with or without CBT).⁵⁶

When pediatric patients continue to experience significantly distressing and/or debilitating symptoms (particularly in MDD) despite multiple trials of antidepressants and psychotherapy, practitioners should consider a careful referral to interventional psychiatry services, which can include the more intensive treatments of electroconvulsive therapy, repetitive transcranial magnetic stimulation, or ketamine (see Box 1). Given the substantial morbidity and mortality associated with adolescent depression, interventional psychiatry treatments are under-researched and under-utilized clinically in pediatric populations.

Continue to: Antidepressants in general...

Antidepressants in general, and SSRIs in particular, are the first-line pharmacotherapy for pediatric anxiety, OCD, and MDD. For PTSD, although they are a first-line treatment in adults, their efficacy has not been demonstrated in children and adolescents. Antidepressants are generally safe, well-tolerated, and effective, with low NNTs (3 to 5 for anxiety and OCD; 4 to 12 in MDD, depending on whether industry trials are included). It is important that clinicians and families be educated about possible adverse effects and their time course in order to anticipate difficulties, ensure adequate informed consent, and monitor appropriately. The black-box warning regarding treatment-emergent suicidal thoughts or behaviors must be discussed (for suggested talking points, see Box 2). The NNH is large (100 to 143), and for many patients, the benefits will outweigh the risks. For pediatric patients who fail to respond to multiple adequate trials of antidepressants and psychotherapy, referrals for interventional psychiatry consultation should be considered.

Bottom Line

Although the evidence base for prescribing antidepressants for children and adolescents is smaller compared to the adult literature, properly understanding and prescribing these agents, and explaining their risks and benefits to families, can make a major difference in patient compliance, satisfaction, and outcomes. Antidepressants (particularly selective serotonin reuptake inhibitors) are the firstline pharmacologic intervention for pediatric patients with anxiety disorders, obsessive-compulsive disorder, or major depressive disorder.

Related Resource

• Bonin L, Moreland CS. Overview of prevention and treatment for pediatric depression. https://www.uptodate.com/contents/overview-of-prevention-and-treatment-for-pediatric-depression. Last reviewed July 2019.

Drug Brand Names

Bupropion • Wellbutrin, Zyban
Cimetidine • Tagamet
Citalopram • Celexa
Clomipramine • Anafranil
Desipramine • Norpramin
Desvenlafaxine • Pristiq
Duloxetine • Cymbalta
Escitalopram • Lexapro
Fluoxetine • Prozac
Fluvoxamine • Luvox
Imipramine • Tofranil
Mirtazapine • Remeron
Nortriptyline • Pamelor
Paroxetine • Paxil
Sertraline • Zoloft
Venlafaxine • Effexor
Vilazodone • Viibryd
Vortioxetine • Trintellix

Box 1

Box 2