Applied Evidence

Hit early, hit hard. That expression arose during the evolution of treatment for human immunodeficiency virus (HIV).¹ While this approach has not lived up to expectations for HIV treatment, it may have found its place in the treatment of recurrent herpes simplex virus (HSV) infections.

Our review focuses on episodic treatment of acute recurrent HSV outbreaks for immunocompetent persons. We do not discuss suppressive therapy, which may be indicated for frequent or severe recurrences (6 or more per year) in immunocompetent persons, for immunocompromised patients, or as an adjunctive measure to reduce genital herpes transmission.²

As we will describe in detail, the efficacy of the new short-course therapy is, at minimum, comparable to that seen with the older, longer-course trials of topical and oral antiviral therapy. In one head-to-head comparison, Leone et al compared a short-course regimen (3 days) of valacyclovir with 5 days of treatment; they found no difference in results.³ If the efficacy of short-course treatment is the same as that of longer courses, the increased convenience and expected improvement in patient adherence with these new regimens argue strongly in their favor. (See Scope of the problem.)

The strategy: Take advantage of a brief therapeutic window

The innate and acquired immune responses of chronically infected, immunocompetent persons rapidly limit cutaneous viral replication, thereby truncating the duration of recurrent HSV outbreaks.^13,14 In both recurrent herpes labialis and genital herpes, HSV viral titers peak in the first 24 hours following lesion onset (FIGURE 1A).^13-15

Herpes labialis lesion size and pain are also greatest in the first 24 hours.^13,16 Most herpes labialis lesions progress from the vesicle stage to the ulcer/soft crust stage within 48 hours, with a hard crust forming by day 2 or 3 (FIGURE 1B).¹⁷

With genital lesions, crust formation depends on whether the skin area is dry (3–4 days) or moist (8–9 days).¹⁴

The likely events are a burst of virus replication in the first 24 hours of outbreak that lyses basal keratinocytes in a discreet area of epidermis innervated by the infected neuron(s), followed by a vigorous immune response that curtails the infection and creates, in part, the clinical disease (erythema, swelling, vesiculation, and ulceration). The subsequent elements of the illness, which are the majority of the lesion course, are related to wound healing

Recognizing the window. Given the brief period of viral replication and the rapid evolution of lesions, the therapeutic window for treating HSV outbreaks with antiviral drugs is both early and short, making it problematic to effectively treat HSV recurrences. Patients often have mature lesions by the time they consult a physician, rendering subsequent anti-viral treatment less effective.¹⁸ However, before lesions appear, many patients experience prodromal symptoms such as pain, burning, or itching.^13,18 These symptoms can be a prompt to start treatment early, thereby taking advantage of the transient therapeutic window.

If a patient is able to self-administer therapy when prodromal symptoms occur, there may be a greater benefit to treatment. Giving patients drugs for self-administration is therefore an important strategy in managing HSV recurrences.

Traditionally, patient-initiated episodic therapy for recurrent genital herpes and herpes labialis has involved multiple daily doses of topical or oral antiviral agents for 4 to 5 days.^19-26 Studies of the pathogenesis of HSV recurrences, however, indicate—as said earlier—that the period of virus replication is early and brief, such that a shorter duration of treatment might be more appropriate and equally effective. Other recent clinical studies have indicated that patient-initiated, short-course, high-dose OAV treatment of recurrent HSV infections may be as effective as the traditional therapies.^3,27-30 In the section that follows, we examine and compare the results of these trials. (See The agents and how they work.)

FIGURE 1A
Lesion HSV-1 titer peaks within 24 hours of onset of herpes labialis lesions

FIGURE 1B
Hard crust formation occurs by 48 to 72 hours of onset of herpes labialis lesions

Clinical trials: Short-course, high-dose, patient-initiated episodic OAV therapy for recurrent genital herpes

Three-day vs 5-day valacyclovir therapy. The efficacy of 3-day treatment with oral valacyclovir was compared with that of 5-day treatment in immunocompetent adults with a history of ≥4 episodes of recurrent genital herpes and confirmed HSV infection.³ Eight hundred participants were randomized to receive 500 mg twice daily valacyclovir for 3 days (and placebo for the remaining 2 days) or 500 mg twice daily for 5 days, and were required to self-administer therapy no later than 24 hours after the onset of symptoms.

The primary endpoint was time to lesion healing (defined as the number of days from initiation of therapy to lesion reepithelialization). Secondary endpoints were pain duration, episode duration (defined as time from initiation of therapy to resolution of all symptoms) and percentage of patients with aborted lesions.

The 3-day valacyclovir treatment exhibited similar time to lesion healing, length of episode, and percentage of patients with aborted lesions as the 5-day treatment (TABLE 1), suggesting equal efficacy. Duration of pain was also similar (data not shown). Adverse events were similar for both treatment groups, with the most common being headache (10%), nausea (4%), and diarrhea (4%, 5-day treatment vs 2%, 3-day treatment).

Placebo-controlled trial of 2-day acyclovir therapy. Wald and coworkers examined the effect of a shorter treatment regimen of acyclovir (2 days) on recurrent genital herpes.²⁸ Eighty-four immunocompetent HSV-2–infected patients with a history of ≥3 recurrences in the previous 12 months were randomized to receive either 2 days of 800 mg 3 times daily acyclovir or matching placebo. Patients were asked to take their medication no later than 12 hours after the first sign or symptom of an episode.

Efficacy endpoints were time to lesion healing, episode duration, and percentage of patients with aborted lesions. Short-course acyclovir therapy was shown to decrease time to healing (P=.001) and episode duration (P<.001) by 2 days compared with placebo (TABLE 1). Short-course acyclovir therapy also increased the percentage of patients with aborted lesions compared with placebo (27% vs 11%; P=.029 (TABLE 1). Adverse events were not recorded in this analysis.

Placebo-controlled trial of single-day famciclovir therapy. Aoki and colleagues²⁹ performed a randomized, double-blind, patient-initiated, placebo-controlled trial to assess the efficacy and safety of patient-initiated, single-day famciclovir 1000 mg twice daily in immunocompetent adults with recurrent genital herpes. The 329 patients in the study were instructed to self-initiate therapy within 6 hours of the onset of prodromal symptoms or genital herpes lesions, and were asked to return to the clinic no later than 24 hours after initiation of therapy. Patients were followed until their lesions healed or for up to 14 days.

The primary endpoint was time to lesion healing of nonaborted lesions. Secondary endpoints were time to healing of all lesions (aborted and nonaborted), time to resolution of pain and other symptoms, and the percentage of patients who did not progress to a full outbreak.

Single-day treatment with famciclovir shortened the time to healing of nonaborted genital herpes lesions by approximately 2 days (P<.001), and the time to healing of all lesions by 1.5 days (P<.001) compared with placebo, and increased the percentage of patients who did not progress to a full outbreak (23% vs 13%;P=.003) (TABLE 1). Famciclovir also reduced the time to resolution of all symptoms by approximately 2 days (P<.001) (data not shown).

Adverse events were mild to moderate; the most common in the famciclovir and placebo groups, respectively, were headache (13.5% vs 5.4%), nausea (2.5% vs 3.6%), and diarrhea (4.9% vs 1.2%).

TABLE 1
Short-course, patient-initiated OAV therapy is effective for treating episodic genital herpes

Short-course, high-dose, patient-initiated episodic OAV therapy for recurrent herpes labialis

Placebo-controlled trial of single-day and 2-day valacyclovir therapy. Spruance and coworkers studied the efficacy of single-day and 2-day valacyclovir treatments in comparison with placebo for an episode of herpes labialis.²⁷ Two identical studies were performed on individuals who were at least 12 years old, had a clinical history of recurrent cold sores, and had experienced ≥3 episodes in the preceding year. Participants in both studies (study 1, N=1524; study 2, N=1627) were required to self-administer 2 g valacyclovir twice daily for 1 day (valacyclovir 1 day), 2 g valacyclovir twice daily for 1 day followed by 1 g twice daily for 1 day (valacyclovir 2 days), or matching placebo at the earliest onset of prodromal symptoms and before the appearance of lesions. Patients were asked to return to the clinic within 24 hours of initiation of therapy.

The primary endpoint in study 1 was clinician-observed duration of all herpes labialis lesions and the secondary endpoint was the percentage of subjects who had herpes labialis lesions that did not progress beyond the papule stage. In study 2, the endpoints were reversed: the primary endpoint was the percentage of patients with lesions that did not progress and the secondary endpoint was the duration of lesions. Other efficacy endpoints were time to healing of vesicular (classical) lesions and duration of pain and discomfort.

Both studies demonstrated that single-day valacyclovir treatment significantly decreased lesion healing time and the duration of herpes labialis episodes by 0.5 to 1.0 days compared with placebo (TABLE 2). A statistically significant decrease in the duration of pain and other symptoms was also seen with single-day valacyclovir compared with placebo (data not shown). In both studies, a higher percentage of patients in the valacyclovir group did not progress to full outbreak compared with placebo, but these differences were not statistically significant. The results with 2 days of valacyclovir treatment were similar. Adverse events were similar between the treatment groups and the placebo group.

Placebo-controlled trial of single-dose and single-day famciclovir therapy. Spruance and coworkers assessed patient-initiated famciclovir 1500 mg (single-dose) and 750 mg twice daily (single-day) in immunocompetent adults with recurrent cold sores.³⁰ Subjects (N=1376) were at least 18 years of age and had experienced ≥3 episodes of cold sores over the previous 12 months. Subjects were instructed to administer 1500 mg (single-dose), 750 mg twice daily (single-day), or matching placebo within 1 hour of the onset of prodromal symptoms and before the onset of lesions, and were asked to return to the clinic within 24 hours of initiating medication.

The primary endpoint was time to healing of primary vesicular lesions. Secondary endpoints included time to healing of all vesicular lesions (primary and secondary [secondary lesions are defined as lesions that developed in addition to and on 1 or more days after primary lesions and that were located at least 1 cm from primary lesions]), time to return to normal skin for all lesions (defined as loss of crust, swelling, and dry flaking), duration of lesion tenderness and pain, and proportion of patients with aborted lesions.

There was a statistically significant decrease in time to healing of primary vesicular lesions by approximately 2 days with both single-dose and single-day famciclovir compared with placebo, with no significant difference between the 2 famciclovir regimens in time to healing of primary vesicular lesions (TABLE 2). There was also a statistically significant decrease in the time to healing of all lesions (primary and secondary) by approximately 2 days with both famciclovir treatments compared with placebo, with no significant differences seen in healing between the famciclovir arms (data not shown).

However, only single-dose famciclovir had a statistically significant decrease in the duration of lesion tenderness and pain and the time to return to normal skin compared with placebo (data not shown). No difference was noted between the famciclovir arms in the percentage of patients with aborted lesions compared with placebo. Adverse events in both famciclovir groups were similar to those in the placebo group.

TABLE 2
Short-course, patient-initiated OAV therapy is effective against recurrent herpes labialis

CORRESPONDENCE
Spotswood Spruance MD Professor of Medicine, Division of Infectious Diseases, University of Utah School of Medicine, Room 4B319, 30 North 1900 East, Salt Lake City, UT 84132-2405. E-mail:[email protected]

Hit early, hit hard. That expression arose during the evolution of treatment for human immunodeficiency virus (HIV).¹ While this approach has not lived up to expectations for HIV treatment, it may have found its place in the treatment of recurrent herpes simplex virus (HSV) infections.

Our review focuses on episodic treatment of acute recurrent HSV outbreaks for immunocompetent persons. We do not discuss suppressive therapy, which may be indicated for frequent or severe recurrences (6 or more per year) in immunocompetent persons, for immunocompromised patients, or as an adjunctive measure to reduce genital herpes transmission.²

As we will describe in detail, the efficacy of the new short-course therapy is, at minimum, comparable to that seen with the older, longer-course trials of topical and oral antiviral therapy. In one head-to-head comparison, Leone et al compared a short-course regimen (3 days) of valacyclovir with 5 days of treatment; they found no difference in results.³ If the efficacy of short-course treatment is the same as that of longer courses, the increased convenience and expected improvement in patient adherence with these new regimens argue strongly in their favor. (See Scope of the problem.)

The strategy: Take advantage of a brief therapeutic window

The innate and acquired immune responses of chronically infected, immunocompetent persons rapidly limit cutaneous viral replication, thereby truncating the duration of recurrent HSV outbreaks.^13,14 In both recurrent herpes labialis and genital herpes, HSV viral titers peak in the first 24 hours following lesion onset (FIGURE 1A).^13-15

Herpes labialis lesion size and pain are also greatest in the first 24 hours.^13,16 Most herpes labialis lesions progress from the vesicle stage to the ulcer/soft crust stage within 48 hours, with a hard crust forming by day 2 or 3 (FIGURE 1B).¹⁷

With genital lesions, crust formation depends on whether the skin area is dry (3–4 days) or moist (8–9 days).¹⁴

The likely events are a burst of virus replication in the first 24 hours of outbreak that lyses basal keratinocytes in a discreet area of epidermis innervated by the infected neuron(s), followed by a vigorous immune response that curtails the infection and creates, in part, the clinical disease (erythema, swelling, vesiculation, and ulceration). The subsequent elements of the illness, which are the majority of the lesion course, are related to wound healing

Recognizing the window. Given the brief period of viral replication and the rapid evolution of lesions, the therapeutic window for treating HSV outbreaks with antiviral drugs is both early and short, making it problematic to effectively treat HSV recurrences. Patients often have mature lesions by the time they consult a physician, rendering subsequent anti-viral treatment less effective.¹⁸ However, before lesions appear, many patients experience prodromal symptoms such as pain, burning, or itching.^13,18 These symptoms can be a prompt to start treatment early, thereby taking advantage of the transient therapeutic window.

If a patient is able to self-administer therapy when prodromal symptoms occur, there may be a greater benefit to treatment. Giving patients drugs for self-administration is therefore an important strategy in managing HSV recurrences.

Traditionally, patient-initiated episodic therapy for recurrent genital herpes and herpes labialis has involved multiple daily doses of topical or oral antiviral agents for 4 to 5 days.^19-26 Studies of the pathogenesis of HSV recurrences, however, indicate—as said earlier—that the period of virus replication is early and brief, such that a shorter duration of treatment might be more appropriate and equally effective. Other recent clinical studies have indicated that patient-initiated, short-course, high-dose OAV treatment of recurrent HSV infections may be as effective as the traditional therapies.^3,27-30 In the section that follows, we examine and compare the results of these trials. (See The agents and how they work.)

FIGURE 1A
Lesion HSV-1 titer peaks within 24 hours of onset of herpes labialis lesions

FIGURE 1B
Hard crust formation occurs by 48 to 72 hours of onset of herpes labialis lesions

Clinical trials: Short-course, high-dose, patient-initiated episodic OAV therapy for recurrent genital herpes

Three-day vs 5-day valacyclovir therapy. The efficacy of 3-day treatment with oral valacyclovir was compared with that of 5-day treatment in immunocompetent adults with a history of ≥4 episodes of recurrent genital herpes and confirmed HSV infection.³ Eight hundred participants were randomized to receive 500 mg twice daily valacyclovir for 3 days (and placebo for the remaining 2 days) or 500 mg twice daily for 5 days, and were required to self-administer therapy no later than 24 hours after the onset of symptoms.

The primary endpoint was time to lesion healing (defined as the number of days from initiation of therapy to lesion reepithelialization). Secondary endpoints were pain duration, episode duration (defined as time from initiation of therapy to resolution of all symptoms) and percentage of patients with aborted lesions.

The 3-day valacyclovir treatment exhibited similar time to lesion healing, length of episode, and percentage of patients with aborted lesions as the 5-day treatment (TABLE 1), suggesting equal efficacy. Duration of pain was also similar (data not shown). Adverse events were similar for both treatment groups, with the most common being headache (10%), nausea (4%), and diarrhea (4%, 5-day treatment vs 2%, 3-day treatment).

Placebo-controlled trial of 2-day acyclovir therapy. Wald and coworkers examined the effect of a shorter treatment regimen of acyclovir (2 days) on recurrent genital herpes.²⁸ Eighty-four immunocompetent HSV-2–infected patients with a history of ≥3 recurrences in the previous 12 months were randomized to receive either 2 days of 800 mg 3 times daily acyclovir or matching placebo. Patients were asked to take their medication no later than 12 hours after the first sign or symptom of an episode.

Efficacy endpoints were time to lesion healing, episode duration, and percentage of patients with aborted lesions. Short-course acyclovir therapy was shown to decrease time to healing (P=.001) and episode duration (P<.001) by 2 days compared with placebo (TABLE 1). Short-course acyclovir therapy also increased the percentage of patients with aborted lesions compared with placebo (27% vs 11%; P=.029 (TABLE 1). Adverse events were not recorded in this analysis.

Placebo-controlled trial of single-day famciclovir therapy. Aoki and colleagues²⁹ performed a randomized, double-blind, patient-initiated, placebo-controlled trial to assess the efficacy and safety of patient-initiated, single-day famciclovir 1000 mg twice daily in immunocompetent adults with recurrent genital herpes. The 329 patients in the study were instructed to self-initiate therapy within 6 hours of the onset of prodromal symptoms or genital herpes lesions, and were asked to return to the clinic no later than 24 hours after initiation of therapy. Patients were followed until their lesions healed or for up to 14 days.

The primary endpoint was time to lesion healing of nonaborted lesions. Secondary endpoints were time to healing of all lesions (aborted and nonaborted), time to resolution of pain and other symptoms, and the percentage of patients who did not progress to a full outbreak.

Single-day treatment with famciclovir shortened the time to healing of nonaborted genital herpes lesions by approximately 2 days (P<.001), and the time to healing of all lesions by 1.5 days (P<.001) compared with placebo, and increased the percentage of patients who did not progress to a full outbreak (23% vs 13%;P=.003) (TABLE 1). Famciclovir also reduced the time to resolution of all symptoms by approximately 2 days (P<.001) (data not shown).

Adverse events were mild to moderate; the most common in the famciclovir and placebo groups, respectively, were headache (13.5% vs 5.4%), nausea (2.5% vs 3.6%), and diarrhea (4.9% vs 1.2%).

TABLE 1
Short-course, patient-initiated OAV therapy is effective for treating episodic genital herpes

Short-course, high-dose, patient-initiated episodic OAV therapy for recurrent herpes labialis

Placebo-controlled trial of single-day and 2-day valacyclovir therapy. Spruance and coworkers studied the efficacy of single-day and 2-day valacyclovir treatments in comparison with placebo for an episode of herpes labialis.²⁷ Two identical studies were performed on individuals who were at least 12 years old, had a clinical history of recurrent cold sores, and had experienced ≥3 episodes in the preceding year. Participants in both studies (study 1, N=1524; study 2, N=1627) were required to self-administer 2 g valacyclovir twice daily for 1 day (valacyclovir 1 day), 2 g valacyclovir twice daily for 1 day followed by 1 g twice daily for 1 day (valacyclovir 2 days), or matching placebo at the earliest onset of prodromal symptoms and before the appearance of lesions. Patients were asked to return to the clinic within 24 hours of initiation of therapy.

The primary endpoint in study 1 was clinician-observed duration of all herpes labialis lesions and the secondary endpoint was the percentage of subjects who had herpes labialis lesions that did not progress beyond the papule stage. In study 2, the endpoints were reversed: the primary endpoint was the percentage of patients with lesions that did not progress and the secondary endpoint was the duration of lesions. Other efficacy endpoints were time to healing of vesicular (classical) lesions and duration of pain and discomfort.

Both studies demonstrated that single-day valacyclovir treatment significantly decreased lesion healing time and the duration of herpes labialis episodes by 0.5 to 1.0 days compared with placebo (TABLE 2). A statistically significant decrease in the duration of pain and other symptoms was also seen with single-day valacyclovir compared with placebo (data not shown). In both studies, a higher percentage of patients in the valacyclovir group did not progress to full outbreak compared with placebo, but these differences were not statistically significant. The results with 2 days of valacyclovir treatment were similar. Adverse events were similar between the treatment groups and the placebo group.

Placebo-controlled trial of single-dose and single-day famciclovir therapy. Spruance and coworkers assessed patient-initiated famciclovir 1500 mg (single-dose) and 750 mg twice daily (single-day) in immunocompetent adults with recurrent cold sores.³⁰ Subjects (N=1376) were at least 18 years of age and had experienced ≥3 episodes of cold sores over the previous 12 months. Subjects were instructed to administer 1500 mg (single-dose), 750 mg twice daily (single-day), or matching placebo within 1 hour of the onset of prodromal symptoms and before the onset of lesions, and were asked to return to the clinic within 24 hours of initiating medication.

The primary endpoint was time to healing of primary vesicular lesions. Secondary endpoints included time to healing of all vesicular lesions (primary and secondary [secondary lesions are defined as lesions that developed in addition to and on 1 or more days after primary lesions and that were located at least 1 cm from primary lesions]), time to return to normal skin for all lesions (defined as loss of crust, swelling, and dry flaking), duration of lesion tenderness and pain, and proportion of patients with aborted lesions.

There was a statistically significant decrease in time to healing of primary vesicular lesions by approximately 2 days with both single-dose and single-day famciclovir compared with placebo, with no significant difference between the 2 famciclovir regimens in time to healing of primary vesicular lesions (TABLE 2). There was also a statistically significant decrease in the time to healing of all lesions (primary and secondary) by approximately 2 days with both famciclovir treatments compared with placebo, with no significant differences seen in healing between the famciclovir arms (data not shown).

However, only single-dose famciclovir had a statistically significant decrease in the duration of lesion tenderness and pain and the time to return to normal skin compared with placebo (data not shown). No difference was noted between the famciclovir arms in the percentage of patients with aborted lesions compared with placebo. Adverse events in both famciclovir groups were similar to those in the placebo group.

TABLE 2
Short-course, patient-initiated OAV therapy is effective against recurrent herpes labialis

CORRESPONDENCE
Spotswood Spruance MD Professor of Medicine, Division of Infectious Diseases, University of Utah School of Medicine, Room 4B319, 30 North 1900 East, Salt Lake City, UT 84132-2405. E-mail:[email protected]

Hit early, hit hard. That expression arose during the evolution of treatment for human immunodeficiency virus (HIV).¹ While this approach has not lived up to expectations for HIV treatment, it may have found its place in the treatment of recurrent herpes simplex virus (HSV) infections.

Our review focuses on episodic treatment of acute recurrent HSV outbreaks for immunocompetent persons. We do not discuss suppressive therapy, which may be indicated for frequent or severe recurrences (6 or more per year) in immunocompetent persons, for immunocompromised patients, or as an adjunctive measure to reduce genital herpes transmission.²

As we will describe in detail, the efficacy of the new short-course therapy is, at minimum, comparable to that seen with the older, longer-course trials of topical and oral antiviral therapy. In one head-to-head comparison, Leone et al compared a short-course regimen (3 days) of valacyclovir with 5 days of treatment; they found no difference in results.³ If the efficacy of short-course treatment is the same as that of longer courses, the increased convenience and expected improvement in patient adherence with these new regimens argue strongly in their favor. (See Scope of the problem.)

The strategy: Take advantage of a brief therapeutic window

The innate and acquired immune responses of chronically infected, immunocompetent persons rapidly limit cutaneous viral replication, thereby truncating the duration of recurrent HSV outbreaks.^13,14 In both recurrent herpes labialis and genital herpes, HSV viral titers peak in the first 24 hours following lesion onset (FIGURE 1A).^13-15

Herpes labialis lesion size and pain are also greatest in the first 24 hours.^13,16 Most herpes labialis lesions progress from the vesicle stage to the ulcer/soft crust stage within 48 hours, with a hard crust forming by day 2 or 3 (FIGURE 1B).¹⁷

With genital lesions, crust formation depends on whether the skin area is dry (3–4 days) or moist (8–9 days).¹⁴

The likely events are a burst of virus replication in the first 24 hours of outbreak that lyses basal keratinocytes in a discreet area of epidermis innervated by the infected neuron(s), followed by a vigorous immune response that curtails the infection and creates, in part, the clinical disease (erythema, swelling, vesiculation, and ulceration). The subsequent elements of the illness, which are the majority of the lesion course, are related to wound healing

Recognizing the window. Given the brief period of viral replication and the rapid evolution of lesions, the therapeutic window for treating HSV outbreaks with antiviral drugs is both early and short, making it problematic to effectively treat HSV recurrences. Patients often have mature lesions by the time they consult a physician, rendering subsequent anti-viral treatment less effective.¹⁸ However, before lesions appear, many patients experience prodromal symptoms such as pain, burning, or itching.^13,18 These symptoms can be a prompt to start treatment early, thereby taking advantage of the transient therapeutic window.

If a patient is able to self-administer therapy when prodromal symptoms occur, there may be a greater benefit to treatment. Giving patients drugs for self-administration is therefore an important strategy in managing HSV recurrences.

Traditionally, patient-initiated episodic therapy for recurrent genital herpes and herpes labialis has involved multiple daily doses of topical or oral antiviral agents for 4 to 5 days.^19-26 Studies of the pathogenesis of HSV recurrences, however, indicate—as said earlier—that the period of virus replication is early and brief, such that a shorter duration of treatment might be more appropriate and equally effective. Other recent clinical studies have indicated that patient-initiated, short-course, high-dose OAV treatment of recurrent HSV infections may be as effective as the traditional therapies.^3,27-30 In the section that follows, we examine and compare the results of these trials. (See The agents and how they work.)

FIGURE 1A
Lesion HSV-1 titer peaks within 24 hours of onset of herpes labialis lesions

FIGURE 1B
Hard crust formation occurs by 48 to 72 hours of onset of herpes labialis lesions

Clinical trials: Short-course, high-dose, patient-initiated episodic OAV therapy for recurrent genital herpes

Three-day vs 5-day valacyclovir therapy. The efficacy of 3-day treatment with oral valacyclovir was compared with that of 5-day treatment in immunocompetent adults with a history of ≥4 episodes of recurrent genital herpes and confirmed HSV infection.³ Eight hundred participants were randomized to receive 500 mg twice daily valacyclovir for 3 days (and placebo for the remaining 2 days) or 500 mg twice daily for 5 days, and were required to self-administer therapy no later than 24 hours after the onset of symptoms.

The primary endpoint was time to lesion healing (defined as the number of days from initiation of therapy to lesion reepithelialization). Secondary endpoints were pain duration, episode duration (defined as time from initiation of therapy to resolution of all symptoms) and percentage of patients with aborted lesions.

The 3-day valacyclovir treatment exhibited similar time to lesion healing, length of episode, and percentage of patients with aborted lesions as the 5-day treatment (TABLE 1), suggesting equal efficacy. Duration of pain was also similar (data not shown). Adverse events were similar for both treatment groups, with the most common being headache (10%), nausea (4%), and diarrhea (4%, 5-day treatment vs 2%, 3-day treatment).

Placebo-controlled trial of 2-day acyclovir therapy. Wald and coworkers examined the effect of a shorter treatment regimen of acyclovir (2 days) on recurrent genital herpes.²⁸ Eighty-four immunocompetent HSV-2–infected patients with a history of ≥3 recurrences in the previous 12 months were randomized to receive either 2 days of 800 mg 3 times daily acyclovir or matching placebo. Patients were asked to take their medication no later than 12 hours after the first sign or symptom of an episode.

Efficacy endpoints were time to lesion healing, episode duration, and percentage of patients with aborted lesions. Short-course acyclovir therapy was shown to decrease time to healing (P=.001) and episode duration (P<.001) by 2 days compared with placebo (TABLE 1). Short-course acyclovir therapy also increased the percentage of patients with aborted lesions compared with placebo (27% vs 11%; P=.029 (TABLE 1). Adverse events were not recorded in this analysis.

Placebo-controlled trial of single-day famciclovir therapy. Aoki and colleagues²⁹ performed a randomized, double-blind, patient-initiated, placebo-controlled trial to assess the efficacy and safety of patient-initiated, single-day famciclovir 1000 mg twice daily in immunocompetent adults with recurrent genital herpes. The 329 patients in the study were instructed to self-initiate therapy within 6 hours of the onset of prodromal symptoms or genital herpes lesions, and were asked to return to the clinic no later than 24 hours after initiation of therapy. Patients were followed until their lesions healed or for up to 14 days.

The primary endpoint was time to lesion healing of nonaborted lesions. Secondary endpoints were time to healing of all lesions (aborted and nonaborted), time to resolution of pain and other symptoms, and the percentage of patients who did not progress to a full outbreak.

Single-day treatment with famciclovir shortened the time to healing of nonaborted genital herpes lesions by approximately 2 days (P<.001), and the time to healing of all lesions by 1.5 days (P<.001) compared with placebo, and increased the percentage of patients who did not progress to a full outbreak (23% vs 13%;P=.003) (TABLE 1). Famciclovir also reduced the time to resolution of all symptoms by approximately 2 days (P<.001) (data not shown).

Adverse events were mild to moderate; the most common in the famciclovir and placebo groups, respectively, were headache (13.5% vs 5.4%), nausea (2.5% vs 3.6%), and diarrhea (4.9% vs 1.2%).

TABLE 1
Short-course, patient-initiated OAV therapy is effective for treating episodic genital herpes

Short-course, high-dose, patient-initiated episodic OAV therapy for recurrent herpes labialis

Placebo-controlled trial of single-day and 2-day valacyclovir therapy. Spruance and coworkers studied the efficacy of single-day and 2-day valacyclovir treatments in comparison with placebo for an episode of herpes labialis.²⁷ Two identical studies were performed on individuals who were at least 12 years old, had a clinical history of recurrent cold sores, and had experienced ≥3 episodes in the preceding year. Participants in both studies (study 1, N=1524; study 2, N=1627) were required to self-administer 2 g valacyclovir twice daily for 1 day (valacyclovir 1 day), 2 g valacyclovir twice daily for 1 day followed by 1 g twice daily for 1 day (valacyclovir 2 days), or matching placebo at the earliest onset of prodromal symptoms and before the appearance of lesions. Patients were asked to return to the clinic within 24 hours of initiation of therapy.

The primary endpoint in study 1 was clinician-observed duration of all herpes labialis lesions and the secondary endpoint was the percentage of subjects who had herpes labialis lesions that did not progress beyond the papule stage. In study 2, the endpoints were reversed: the primary endpoint was the percentage of patients with lesions that did not progress and the secondary endpoint was the duration of lesions. Other efficacy endpoints were time to healing of vesicular (classical) lesions and duration of pain and discomfort.

Both studies demonstrated that single-day valacyclovir treatment significantly decreased lesion healing time and the duration of herpes labialis episodes by 0.5 to 1.0 days compared with placebo (TABLE 2). A statistically significant decrease in the duration of pain and other symptoms was also seen with single-day valacyclovir compared with placebo (data not shown). In both studies, a higher percentage of patients in the valacyclovir group did not progress to full outbreak compared with placebo, but these differences were not statistically significant. The results with 2 days of valacyclovir treatment were similar. Adverse events were similar between the treatment groups and the placebo group.

Placebo-controlled trial of single-dose and single-day famciclovir therapy. Spruance and coworkers assessed patient-initiated famciclovir 1500 mg (single-dose) and 750 mg twice daily (single-day) in immunocompetent adults with recurrent cold sores.³⁰ Subjects (N=1376) were at least 18 years of age and had experienced ≥3 episodes of cold sores over the previous 12 months. Subjects were instructed to administer 1500 mg (single-dose), 750 mg twice daily (single-day), or matching placebo within 1 hour of the onset of prodromal symptoms and before the onset of lesions, and were asked to return to the clinic within 24 hours of initiating medication.

The primary endpoint was time to healing of primary vesicular lesions. Secondary endpoints included time to healing of all vesicular lesions (primary and secondary [secondary lesions are defined as lesions that developed in addition to and on 1 or more days after primary lesions and that were located at least 1 cm from primary lesions]), time to return to normal skin for all lesions (defined as loss of crust, swelling, and dry flaking), duration of lesion tenderness and pain, and proportion of patients with aborted lesions.

There was a statistically significant decrease in time to healing of primary vesicular lesions by approximately 2 days with both single-dose and single-day famciclovir compared with placebo, with no significant difference between the 2 famciclovir regimens in time to healing of primary vesicular lesions (TABLE 2). There was also a statistically significant decrease in the time to healing of all lesions (primary and secondary) by approximately 2 days with both famciclovir treatments compared with placebo, with no significant differences seen in healing between the famciclovir arms (data not shown).

However, only single-dose famciclovir had a statistically significant decrease in the duration of lesion tenderness and pain and the time to return to normal skin compared with placebo (data not shown). No difference was noted between the famciclovir arms in the percentage of patients with aborted lesions compared with placebo. Adverse events in both famciclovir groups were similar to those in the placebo group.

TABLE 2
Short-course, patient-initiated OAV therapy is effective against recurrent herpes labialis

CORRESPONDENCE
Spotswood Spruance MD Professor of Medicine, Division of Infectious Diseases, University of Utah School of Medicine, Room 4B319, 30 North 1900 East, Salt Lake City, UT 84132-2405. E-mail:[email protected]

How do FPs compare with dermatologists?

Dermatology literature boasts about the superiority of the dermatologist in diagnostic ability, cost savings, and cancer prevention when compared with primary care physicians.^6-10 Studies have evaluated the skill level of primary care physicians compared with dermatologists in identifying skin disorders when tested with color transparencies, computer images, and slides—however, rarely with actual patients.^7,9-16 Some studies have suggested a higher rate of referral for skin problems than for other non-dermatologic conditions.^14,17,18

Often the outcome of interest in these studies is disease-oriented, judging a physician’s diagnostic ability, rather than examining a patient-oriented outcome, such as resolution of lesion or patient satisfaction.

Thus, the secondary aims of our study were to observe how family physicians diagnose and treat the lesions, and to gauge their concordance with dermatologists’ assessments and plans. We hypothesized that, in an office setting, family physicians would provide effective and efficient treatment for most patients who present with new skin lesions, and that there is high diagnostic concordance between the 2 specialties.

We first share our study findings, and then provide details of our Methodology and Results.

Family physicians excel at dermatologic care

Our study demonstrates that most skin conditions diagnosed and managed by family physicians improve. At day 7, 84% of patients who were contacted reported their skin lesions were “better” or “much better.” Moreover, patients said they were highly satisfied with their care. Referrals to subspecialists were infrequent.

These findings counter those from previous studies questioning primary care physicians’ care of dermatologic conditions. We believe it is likely that patients in previous studies reflected different populations than are typically seen by family physicians.^18-20 Another difference may be that family physicians used other resources to assist with their diagnosis and treatment decisions. As we hypothesized, family physicians had good correlation with dermatologists in both diagnosis and treatment, and skin lesions improved.

Important study limitations

We relied on patient reports of improvement. While-self impression of degree of improvement is a patient-centered outcome, there may be instances in which inappropriate or insufficient treatment may produce temporary symptomatic relief and mask true improvement.

Although the patients’ primary care physicians were not involved in the follow-up process, it is possible they felt some social pressure to report higher levels of improvement or satisfaction.

Though we attempted to enroll all eligible patients, some patients seen for skin conditions may not have been captured. As we met our planned enrollment rates, we believe we captured most of the eligible encounters.

Some studies have questioned primary care physicians’ abilities to properly diagnose skin cancers.^21,22 Our study was not designed or powered to detect skin cancers or the number, if any, of missed diagnoses of skin cancer.

Cues for teachers of family medicine

Most diagnoses fell within a limited set of diagnostic categories that probably reflect a distribution of skin disorders more typical within family medicine than in dermatology clinics. This range of disease defines a set of diagnostic skills, information resources, and treatment plans required to make these diagnoses and manage these conditions in family practice settings. This information should help physicians involved in training family physicians to concentrate on these common categories of diagnoses. Most important, our study conducted with actual patients found that family physicians manage skin lesions effectively and efficiently, with high patient satisfaction.

Methods

Study design and participants

We conducted a multisite, 3-state (Maryland, Virginia, and Washington, DC) prospective cohort study under the auspices of the Capital Area Practice Based Research Network (CAPRICORN). Between May 24 and August 13, 2004, all patients with new skin lesions who were seen by participating physicians were expected to enter into the study. Institutional Review Board approval was obtained from Georgetown University prior to the study. Written informed consent was obtained from all physicians and patients.

Inclusion/exclusion criteria

A lesion was considered new if patients presented to a family physician with one or more skin lesion that had not been previously treated or examined by another physician.

Patients were ineligible if they: 1) had a lesion with unknown duration; 2) had no telephone for follow-up; 3) did not speak English or Spanish; or, 4) had a lesion resulting from trauma.

Interventions

The initial intervention consisted of 2 parts: 1) after examining a patient, family physicians completed a 10-question survey, recording diagnosis, treatment plan, and resources used in treatment; 2) research assistants completed a 14-question survey, consisting of general patient and lesion information. Follow-up patient surveys were completed by telephone on days 7, 28, and 84.

Two university-based dermatologists helped develop the photography protocol. They specifically requested 3 digital photos of lesions under incandescent light, specific information for diagnosis, and direction for how photographs should be taken. The photographs were taken using Olympus C-5000 5MP Digital Camera w/3x Optical Zoom and were developed with HP photo glossy paper. The dermatologists separately reviewed the photographs blinded to the family physician’s diagnosis and treatment. The dermatologists commented on diagnosis and treatment plan for the first 99 patients enrolled in the study.

Outcomes

The primary outcome was dichotomous: whether skin lesions improved or not at day 7. Secondary outcomes were measures of improvement at days 28 and 84. We also examined patients’ satisfaction on a scale of 1 to 5 (“How satisfied were you with your skin care provided by your family physician?” 1=very satisfied, 5=very unsatisfied).

The categorization of acute skin lesions was developed by a modified delphi process in order to classify the lesions into groups. The principal investigator initially categorized all diagnoses and treatments. Next, 3 other members of the study (AK, BP, and DM) individually reviewed and guided categorizations. The 2 dermatologists gave the final input. This resulted in 41 categories for diagnosis and 9 for treatment.

Statistical analysis

Descriptive statistics provided baseline characteristics for the group. Frequencies were computed on patient, visit, and lesion characteristics, including patient improvement at days 7, 28, and 84. We also computed patient satisfaction with the care provided by their physician at 7, 28, and 84 days. Agreement rates between the family physicians and the 2 dermatologists were obtained for the subset of cases where both dermatologists agreed on the diagnosis. Similarly, the agreement rates were computed for recommended treatment using only those cases where the 2 dermatologists agreed on treatment. All descriptive statistics were computed with SPSS (SPSS, Inc, Chicago, Ill).

RESULTS

A total of 244 patients with 267 skin lesions were recruited by 53 family physicians during the study period. The 7-day follow-up patient survey was completed for 234 lesions (88%), the 28-day survey was completed for 220 lesions (82%), and the 84-day survey was completed for 203 lesions (76%). Study participants ranged in age from 3 months to 86 years; adults were predominantly college-educated, non-Hispanic, and white (TABLE 1). The majority of study participants (73%) reported that their skin lesion was the primary reason for their appointment.

Characteristics of the clinical encounters are presented in TABLE 2. While most skin lesions were present for 30 days or less (62%), over one quarter had been present for more than 90 days. The family physicians made 40 general dermatologic diagnoses. Only 3 lesions (1%) were considered malignant (data not shown). Family physicians reported relatively high confidence with their diagnoses (mean confidence score of 8.4, with range 1 to 10, 1=not at all certain, 10=very certain).

Other characteristics of the clinical encounters not shown in TABLE 2 are the family physicians’ judgment on resolution of the lesions and diagnostic steps used in treating the lesions. In most cases, family physicians believed the lesion would resolve within 12 weeks (203 lesions received a score of ≥7, 0=no improvement expected, 10=complete resolution expected). There was a bimodal distribution with 144 lesions receiving a 10, while 36 received a grade of 0. To make their diagnosis, most family physicians examined other parts of the skin (70%), consulted a colleague (14%), or consulted an electronic resource (6%). Laboratory tests, skin scrapings, diagnostic cultures, Woods lamp exams, or skin biopsies were performed in a total of 10% of encounters.

TABLE 3 reports the primary outcome, patient-reported resolution of skin lesions. These data were restricted only to lesions that were expected to improve (defined as a clinician assigned resolution score ≥7).

Overall, patients were very satisfied with the dermatologic care provided by their family physician. On a 5-point satisfaction scale, 55% of patients reported 1, the highest satisfaction level and 34% reported 2, the next highest level at day 7. At days 28 and 84, 93% of the patients reported the 2 highest levels of satisfaction. These data exclude patients lost to follow-up. Including all participants in the denominator, the rates of either the 2 highest levels of satisfaction at day 7 was 78%, at day 28 was 76%, and at day 84 was 70%.

The overall agreements in diagnosis and treatment, respectively, between the family physicians and the dermatologists were 72% and 80%. We examined only the aspects where both of the dermatologists agreed. Interestingly, for the more common diagnoses, the agreement rates were above 80%; however, for less common diagnoses, the rates were 62%. This trend was not observed in the treatment agreements, primarily due to dermatologists recommending steroids much more often than family physicians prescribed steroids. See Table 4 and Table 5.

TABLE 1
Characteristics of study sample

TABLE 2
Skin lesions seen in study sites

TABLE 3
Patients reported high satisfaction

Acknowledgments

The views expressed are those of the authors. No official endorsement by the Agency for Healthcare Research and Quality is intended or should be inferred. We would like to thank the following medical students who played an integral role in recruitment, Aaron Baker, Richard Sisson and Giovina Lara Bomba. We would like to thank Haewon Park for editorial assistance. We would like to that the following practices for participation, Potomac Physician Associates of Kensington, La Clinica del Pueblo, Community of Hope, Fort Lincoln, Fairfax Family Practice of Vienna, Fair Oaks, and Prince William.

CORRESPONDENCE
Dan Merenstein, MD, 215 Kober Cogan Hall, 3750 Reservoir Road, NW, Washington, DC 20007. E-mail: [email protected]

How do FPs compare with dermatologists?

Dermatology literature boasts about the superiority of the dermatologist in diagnostic ability, cost savings, and cancer prevention when compared with primary care physicians.^6-10 Studies have evaluated the skill level of primary care physicians compared with dermatologists in identifying skin disorders when tested with color transparencies, computer images, and slides—however, rarely with actual patients.^7,9-16 Some studies have suggested a higher rate of referral for skin problems than for other non-dermatologic conditions.^14,17,18

Often the outcome of interest in these studies is disease-oriented, judging a physician’s diagnostic ability, rather than examining a patient-oriented outcome, such as resolution of lesion or patient satisfaction.

Thus, the secondary aims of our study were to observe how family physicians diagnose and treat the lesions, and to gauge their concordance with dermatologists’ assessments and plans. We hypothesized that, in an office setting, family physicians would provide effective and efficient treatment for most patients who present with new skin lesions, and that there is high diagnostic concordance between the 2 specialties.

We first share our study findings, and then provide details of our Methodology and Results.

Family physicians excel at dermatologic care

Our study demonstrates that most skin conditions diagnosed and managed by family physicians improve. At day 7, 84% of patients who were contacted reported their skin lesions were “better” or “much better.” Moreover, patients said they were highly satisfied with their care. Referrals to subspecialists were infrequent.

These findings counter those from previous studies questioning primary care physicians’ care of dermatologic conditions. We believe it is likely that patients in previous studies reflected different populations than are typically seen by family physicians.^18-20 Another difference may be that family physicians used other resources to assist with their diagnosis and treatment decisions. As we hypothesized, family physicians had good correlation with dermatologists in both diagnosis and treatment, and skin lesions improved.

Important study limitations

We relied on patient reports of improvement. While-self impression of degree of improvement is a patient-centered outcome, there may be instances in which inappropriate or insufficient treatment may produce temporary symptomatic relief and mask true improvement.

Although the patients’ primary care physicians were not involved in the follow-up process, it is possible they felt some social pressure to report higher levels of improvement or satisfaction.

Though we attempted to enroll all eligible patients, some patients seen for skin conditions may not have been captured. As we met our planned enrollment rates, we believe we captured most of the eligible encounters.

Some studies have questioned primary care physicians’ abilities to properly diagnose skin cancers.^21,22 Our study was not designed or powered to detect skin cancers or the number, if any, of missed diagnoses of skin cancer.

Cues for teachers of family medicine

Most diagnoses fell within a limited set of diagnostic categories that probably reflect a distribution of skin disorders more typical within family medicine than in dermatology clinics. This range of disease defines a set of diagnostic skills, information resources, and treatment plans required to make these diagnoses and manage these conditions in family practice settings. This information should help physicians involved in training family physicians to concentrate on these common categories of diagnoses. Most important, our study conducted with actual patients found that family physicians manage skin lesions effectively and efficiently, with high patient satisfaction.

Methods

Study design and participants

We conducted a multisite, 3-state (Maryland, Virginia, and Washington, DC) prospective cohort study under the auspices of the Capital Area Practice Based Research Network (CAPRICORN). Between May 24 and August 13, 2004, all patients with new skin lesions who were seen by participating physicians were expected to enter into the study. Institutional Review Board approval was obtained from Georgetown University prior to the study. Written informed consent was obtained from all physicians and patients.

Inclusion/exclusion criteria

A lesion was considered new if patients presented to a family physician with one or more skin lesion that had not been previously treated or examined by another physician.

Patients were ineligible if they: 1) had a lesion with unknown duration; 2) had no telephone for follow-up; 3) did not speak English or Spanish; or, 4) had a lesion resulting from trauma.

Interventions

The initial intervention consisted of 2 parts: 1) after examining a patient, family physicians completed a 10-question survey, recording diagnosis, treatment plan, and resources used in treatment; 2) research assistants completed a 14-question survey, consisting of general patient and lesion information. Follow-up patient surveys were completed by telephone on days 7, 28, and 84.

Two university-based dermatologists helped develop the photography protocol. They specifically requested 3 digital photos of lesions under incandescent light, specific information for diagnosis, and direction for how photographs should be taken. The photographs were taken using Olympus C-5000 5MP Digital Camera w/3x Optical Zoom and were developed with HP photo glossy paper. The dermatologists separately reviewed the photographs blinded to the family physician’s diagnosis and treatment. The dermatologists commented on diagnosis and treatment plan for the first 99 patients enrolled in the study.

Outcomes

The primary outcome was dichotomous: whether skin lesions improved or not at day 7. Secondary outcomes were measures of improvement at days 28 and 84. We also examined patients’ satisfaction on a scale of 1 to 5 (“How satisfied were you with your skin care provided by your family physician?” 1=very satisfied, 5=very unsatisfied).

The categorization of acute skin lesions was developed by a modified delphi process in order to classify the lesions into groups. The principal investigator initially categorized all diagnoses and treatments. Next, 3 other members of the study (AK, BP, and DM) individually reviewed and guided categorizations. The 2 dermatologists gave the final input. This resulted in 41 categories for diagnosis and 9 for treatment.

Statistical analysis

Descriptive statistics provided baseline characteristics for the group. Frequencies were computed on patient, visit, and lesion characteristics, including patient improvement at days 7, 28, and 84. We also computed patient satisfaction with the care provided by their physician at 7, 28, and 84 days. Agreement rates between the family physicians and the 2 dermatologists were obtained for the subset of cases where both dermatologists agreed on the diagnosis. Similarly, the agreement rates were computed for recommended treatment using only those cases where the 2 dermatologists agreed on treatment. All descriptive statistics were computed with SPSS (SPSS, Inc, Chicago, Ill).

RESULTS

A total of 244 patients with 267 skin lesions were recruited by 53 family physicians during the study period. The 7-day follow-up patient survey was completed for 234 lesions (88%), the 28-day survey was completed for 220 lesions (82%), and the 84-day survey was completed for 203 lesions (76%). Study participants ranged in age from 3 months to 86 years; adults were predominantly college-educated, non-Hispanic, and white (TABLE 1). The majority of study participants (73%) reported that their skin lesion was the primary reason for their appointment.

Characteristics of the clinical encounters are presented in TABLE 2. While most skin lesions were present for 30 days or less (62%), over one quarter had been present for more than 90 days. The family physicians made 40 general dermatologic diagnoses. Only 3 lesions (1%) were considered malignant (data not shown). Family physicians reported relatively high confidence with their diagnoses (mean confidence score of 8.4, with range 1 to 10, 1=not at all certain, 10=very certain).

Other characteristics of the clinical encounters not shown in TABLE 2 are the family physicians’ judgment on resolution of the lesions and diagnostic steps used in treating the lesions. In most cases, family physicians believed the lesion would resolve within 12 weeks (203 lesions received a score of ≥7, 0=no improvement expected, 10=complete resolution expected). There was a bimodal distribution with 144 lesions receiving a 10, while 36 received a grade of 0. To make their diagnosis, most family physicians examined other parts of the skin (70%), consulted a colleague (14%), or consulted an electronic resource (6%). Laboratory tests, skin scrapings, diagnostic cultures, Woods lamp exams, or skin biopsies were performed in a total of 10% of encounters.

TABLE 3 reports the primary outcome, patient-reported resolution of skin lesions. These data were restricted only to lesions that were expected to improve (defined as a clinician assigned resolution score ≥7).

Overall, patients were very satisfied with the dermatologic care provided by their family physician. On a 5-point satisfaction scale, 55% of patients reported 1, the highest satisfaction level and 34% reported 2, the next highest level at day 7. At days 28 and 84, 93% of the patients reported the 2 highest levels of satisfaction. These data exclude patients lost to follow-up. Including all participants in the denominator, the rates of either the 2 highest levels of satisfaction at day 7 was 78%, at day 28 was 76%, and at day 84 was 70%.

The overall agreements in diagnosis and treatment, respectively, between the family physicians and the dermatologists were 72% and 80%. We examined only the aspects where both of the dermatologists agreed. Interestingly, for the more common diagnoses, the agreement rates were above 80%; however, for less common diagnoses, the rates were 62%. This trend was not observed in the treatment agreements, primarily due to dermatologists recommending steroids much more often than family physicians prescribed steroids. See Table 4 and Table 5.

TABLE 1
Characteristics of study sample

TABLE 2
Skin lesions seen in study sites

TABLE 3
Patients reported high satisfaction

Acknowledgments

The views expressed are those of the authors. No official endorsement by the Agency for Healthcare Research and Quality is intended or should be inferred. We would like to thank the following medical students who played an integral role in recruitment, Aaron Baker, Richard Sisson and Giovina Lara Bomba. We would like to thank Haewon Park for editorial assistance. We would like to that the following practices for participation, Potomac Physician Associates of Kensington, La Clinica del Pueblo, Community of Hope, Fort Lincoln, Fairfax Family Practice of Vienna, Fair Oaks, and Prince William.

CORRESPONDENCE
Dan Merenstein, MD, 215 Kober Cogan Hall, 3750 Reservoir Road, NW, Washington, DC 20007. E-mail: [email protected]

How do FPs compare with dermatologists?

Dermatology literature boasts about the superiority of the dermatologist in diagnostic ability, cost savings, and cancer prevention when compared with primary care physicians.^6-10 Studies have evaluated the skill level of primary care physicians compared with dermatologists in identifying skin disorders when tested with color transparencies, computer images, and slides—however, rarely with actual patients.^7,9-16 Some studies have suggested a higher rate of referral for skin problems than for other non-dermatologic conditions.^14,17,18

Often the outcome of interest in these studies is disease-oriented, judging a physician’s diagnostic ability, rather than examining a patient-oriented outcome, such as resolution of lesion or patient satisfaction.

Thus, the secondary aims of our study were to observe how family physicians diagnose and treat the lesions, and to gauge their concordance with dermatologists’ assessments and plans. We hypothesized that, in an office setting, family physicians would provide effective and efficient treatment for most patients who present with new skin lesions, and that there is high diagnostic concordance between the 2 specialties.

We first share our study findings, and then provide details of our Methodology and Results.

Family physicians excel at dermatologic care

Our study demonstrates that most skin conditions diagnosed and managed by family physicians improve. At day 7, 84% of patients who were contacted reported their skin lesions were “better” or “much better.” Moreover, patients said they were highly satisfied with their care. Referrals to subspecialists were infrequent.

These findings counter those from previous studies questioning primary care physicians’ care of dermatologic conditions. We believe it is likely that patients in previous studies reflected different populations than are typically seen by family physicians.^18-20 Another difference may be that family physicians used other resources to assist with their diagnosis and treatment decisions. As we hypothesized, family physicians had good correlation with dermatologists in both diagnosis and treatment, and skin lesions improved.

Important study limitations

We relied on patient reports of improvement. While-self impression of degree of improvement is a patient-centered outcome, there may be instances in which inappropriate or insufficient treatment may produce temporary symptomatic relief and mask true improvement.

Although the patients’ primary care physicians were not involved in the follow-up process, it is possible they felt some social pressure to report higher levels of improvement or satisfaction.

Though we attempted to enroll all eligible patients, some patients seen for skin conditions may not have been captured. As we met our planned enrollment rates, we believe we captured most of the eligible encounters.

Some studies have questioned primary care physicians’ abilities to properly diagnose skin cancers.^21,22 Our study was not designed or powered to detect skin cancers or the number, if any, of missed diagnoses of skin cancer.

Cues for teachers of family medicine

Most diagnoses fell within a limited set of diagnostic categories that probably reflect a distribution of skin disorders more typical within family medicine than in dermatology clinics. This range of disease defines a set of diagnostic skills, information resources, and treatment plans required to make these diagnoses and manage these conditions in family practice settings. This information should help physicians involved in training family physicians to concentrate on these common categories of diagnoses. Most important, our study conducted with actual patients found that family physicians manage skin lesions effectively and efficiently, with high patient satisfaction.

Methods

Study design and participants

We conducted a multisite, 3-state (Maryland, Virginia, and Washington, DC) prospective cohort study under the auspices of the Capital Area Practice Based Research Network (CAPRICORN). Between May 24 and August 13, 2004, all patients with new skin lesions who were seen by participating physicians were expected to enter into the study. Institutional Review Board approval was obtained from Georgetown University prior to the study. Written informed consent was obtained from all physicians and patients.

Inclusion/exclusion criteria

A lesion was considered new if patients presented to a family physician with one or more skin lesion that had not been previously treated or examined by another physician.

Patients were ineligible if they: 1) had a lesion with unknown duration; 2) had no telephone for follow-up; 3) did not speak English or Spanish; or, 4) had a lesion resulting from trauma.

Interventions

The initial intervention consisted of 2 parts: 1) after examining a patient, family physicians completed a 10-question survey, recording diagnosis, treatment plan, and resources used in treatment; 2) research assistants completed a 14-question survey, consisting of general patient and lesion information. Follow-up patient surveys were completed by telephone on days 7, 28, and 84.

Two university-based dermatologists helped develop the photography protocol. They specifically requested 3 digital photos of lesions under incandescent light, specific information for diagnosis, and direction for how photographs should be taken. The photographs were taken using Olympus C-5000 5MP Digital Camera w/3x Optical Zoom and were developed with HP photo glossy paper. The dermatologists separately reviewed the photographs blinded to the family physician’s diagnosis and treatment. The dermatologists commented on diagnosis and treatment plan for the first 99 patients enrolled in the study.

Outcomes

The primary outcome was dichotomous: whether skin lesions improved or not at day 7. Secondary outcomes were measures of improvement at days 28 and 84. We also examined patients’ satisfaction on a scale of 1 to 5 (“How satisfied were you with your skin care provided by your family physician?” 1=very satisfied, 5=very unsatisfied).

The categorization of acute skin lesions was developed by a modified delphi process in order to classify the lesions into groups. The principal investigator initially categorized all diagnoses and treatments. Next, 3 other members of the study (AK, BP, and DM) individually reviewed and guided categorizations. The 2 dermatologists gave the final input. This resulted in 41 categories for diagnosis and 9 for treatment.

Statistical analysis

Descriptive statistics provided baseline characteristics for the group. Frequencies were computed on patient, visit, and lesion characteristics, including patient improvement at days 7, 28, and 84. We also computed patient satisfaction with the care provided by their physician at 7, 28, and 84 days. Agreement rates between the family physicians and the 2 dermatologists were obtained for the subset of cases where both dermatologists agreed on the diagnosis. Similarly, the agreement rates were computed for recommended treatment using only those cases where the 2 dermatologists agreed on treatment. All descriptive statistics were computed with SPSS (SPSS, Inc, Chicago, Ill).

RESULTS

A total of 244 patients with 267 skin lesions were recruited by 53 family physicians during the study period. The 7-day follow-up patient survey was completed for 234 lesions (88%), the 28-day survey was completed for 220 lesions (82%), and the 84-day survey was completed for 203 lesions (76%). Study participants ranged in age from 3 months to 86 years; adults were predominantly college-educated, non-Hispanic, and white (TABLE 1). The majority of study participants (73%) reported that their skin lesion was the primary reason for their appointment.

Characteristics of the clinical encounters are presented in TABLE 2. While most skin lesions were present for 30 days or less (62%), over one quarter had been present for more than 90 days. The family physicians made 40 general dermatologic diagnoses. Only 3 lesions (1%) were considered malignant (data not shown). Family physicians reported relatively high confidence with their diagnoses (mean confidence score of 8.4, with range 1 to 10, 1=not at all certain, 10=very certain).

Other characteristics of the clinical encounters not shown in TABLE 2 are the family physicians’ judgment on resolution of the lesions and diagnostic steps used in treating the lesions. In most cases, family physicians believed the lesion would resolve within 12 weeks (203 lesions received a score of ≥7, 0=no improvement expected, 10=complete resolution expected). There was a bimodal distribution with 144 lesions receiving a 10, while 36 received a grade of 0. To make their diagnosis, most family physicians examined other parts of the skin (70%), consulted a colleague (14%), or consulted an electronic resource (6%). Laboratory tests, skin scrapings, diagnostic cultures, Woods lamp exams, or skin biopsies were performed in a total of 10% of encounters.

TABLE 3 reports the primary outcome, patient-reported resolution of skin lesions. These data were restricted only to lesions that were expected to improve (defined as a clinician assigned resolution score ≥7).

Overall, patients were very satisfied with the dermatologic care provided by their family physician. On a 5-point satisfaction scale, 55% of patients reported 1, the highest satisfaction level and 34% reported 2, the next highest level at day 7. At days 28 and 84, 93% of the patients reported the 2 highest levels of satisfaction. These data exclude patients lost to follow-up. Including all participants in the denominator, the rates of either the 2 highest levels of satisfaction at day 7 was 78%, at day 28 was 76%, and at day 84 was 70%.

The overall agreements in diagnosis and treatment, respectively, between the family physicians and the dermatologists were 72% and 80%. We examined only the aspects where both of the dermatologists agreed. Interestingly, for the more common diagnoses, the agreement rates were above 80%; however, for less common diagnoses, the rates were 62%. This trend was not observed in the treatment agreements, primarily due to dermatologists recommending steroids much more often than family physicians prescribed steroids. See Table 4 and Table 5.

TABLE 1
Characteristics of study sample

TABLE 2
Skin lesions seen in study sites

TABLE 3
Patients reported high satisfaction

Acknowledgments

The views expressed are those of the authors. No official endorsement by the Agency for Healthcare Research and Quality is intended or should be inferred. We would like to thank the following medical students who played an integral role in recruitment, Aaron Baker, Richard Sisson and Giovina Lara Bomba. We would like to thank Haewon Park for editorial assistance. We would like to that the following practices for participation, Potomac Physician Associates of Kensington, La Clinica del Pueblo, Community of Hope, Fort Lincoln, Fairfax Family Practice of Vienna, Fair Oaks, and Prince William.

CORRESPONDENCE
Dan Merenstein, MD, 215 Kober Cogan Hall, 3750 Reservoir Road, NW, Washington, DC 20007. E-mail: [email protected]

A 51-year-old man of northern European descent who works outdoors for the city asks you to look at a “mole” on his face. The lesion does not have the classic appearance of a melanoma that you have seen before, and it is still fairly small. Should you advise a wait-and-see approach or perform a biopsy?

Opt for early detection

Given the patient’s likely genetic predisposition to skin cancer and his regular, lengthy exposure to sunlight (see Risk factors), you would be wise to follow your cutaneous examination with a biopsy.

The ABCDs of visual assessment. The classic clinical presentation of melanoma is well known. The most publicized means for identifying potentially atypical pigmented skin lesions is the “ABCD” mnemonic (asymmetry, border irregularity, color variegation, and lesion diameter >6 mm) (FIGURES 1 AND 2).^7,8 The ABCDs are primarily an educational tool for patients. This mnemonic was expanded to include E, representing evolution of a pigmented lesion.⁹ Any pigmented lesion observed to change significantly on serial examinations warrants biopsy to exclude melanoma.

Caveat: not all melanomas are pigmented. Amelanotic melanomas are a diagnostic challenge and may be lethal if left unattended.

Routine screening. Large-scale skin cancer screening has been performed and found to be a statistically ineffective means of detection. Moreover, the US Preventive Services Task Force found insufficient evidence to recommend for or against routine counseling by primary care physicians to prevent skin cancer.¹⁰

A large-scale educational and screening campaign was performed in Italy from 1991 to 1996. During this period, 90,000 educational leaflets were distributed to a target population of approximately 243,000. A total of 2050 individuals requested a skin examination, resulting in detection of 13 melanomas.¹¹ However, 92.3% of the melanomas were thin (<1.52 mm deep). Despite the lack of statistical significance for such screenings, many organizations do perform them and find melanomas, which can be life saving for those few individuals.

Anatomic areas to focus on. While melanoma can affect any anatomic region, it is especially common on sun-exposed areas, including the head, neck, and upper extremities. Acral lentiginous melanoma is found on palmar, plantar, and subungual regions. Include the scalp, ocular mucosa, and oral cavity in your examination.

FIGURE 1
Melanoma in situ

FIGURE 2
Melanoma in situ

Which biopsy technique to choose?

A properly performed biopsy is mandatory to accurately diagnose and microstage the tumor. No other test reliably surpasses routine histologic examination.

When possible, arrange for a full-thickness excisional biopsy with a narrow (2-mm) rim of normal skin, especially for large lesions in which sampling error may be a factor.

Tissue samples are submitted for histologic examination in formalin for permanent sections. Frozen sections are not recommended for the initial diagnosis of melanoma due to artifactual changes from the freezing process.

When lesions are too large for primary excision, or if an optimal cosmetic result is important to the patient, an incisional or punch biopsy may be taken from the area believed to be the deepest portion of the tumor.

For the largest lesions, multiple punch or incisional biopsies may help to reduce the risk of sampling error. Diagnostic and management problems arise when the initial biopsy does not sample the complete skin thickness or when large lesions are not sampled adequately.

Classification of melanoma

The traditional melanoma classification scheme includes several subtypes (see Detailed melanoma classifications). Though these terms have limited clinical utility, we include them for the sake of completeness and because the terminology may arise in consultation with colleagues.

Most important for the prognosis is a determination of the tumor’s depth of invasiveness.

In situ or invasive?

Melanoma in situ lesions are confined to the epidermis and may extend along hair appendages. It often occurs on sun-exposed areas (FIGURES 1 AND 2). Histologically, melanoma in situ is an asymmetric and poorly circumscribed proliferation of melanocytes usually larger than 6 mm in diameter (FIGURE 3).

Melanocytes form irregular nests that are not equidistant and have areas of confluence (FIGURE 3). Single melanocytes predominate over nests in some areas and may entirely replace the basal layer.

Additional histologic features include single melanocytes above the dermal-epidermal junction (Pagetoid spread) and uneven distribution of pigment. In many cases the dermis has inflammatory infiltrates, melanophages, and evidence of solar-damage. Particularly in anatomical areas rich in hair follicles, the neoplastic cells may spread into the epithelium of the hair follicles without extending to the dermis.

Melanoma in situ may be quite large in diameter (horizontal growth phase) without becoming invasive; however, always be concerned about invasion.

Invasive melanoma shares the same histologic features but invades the dermis or subcutaneous fat (vertical growth phase). Aside from the conventional criteria used for the histologic diagnosis of melanoma, acral melanomas may show an increased number of dendritic melanocytes loaded with melanin (FIGURE 4). Desmoplastic melanoma is most commonly found on the head and neck; however, it varies in clinical presentation (FIGURE 5). It is often associated with other types of melanoma, is common in older persons, and has a slight male predilection.

FIGURE 3
Melanocytes

FIGURE 4
Acral melanoma

FIGURE 5
Dermoplastic melanoma

Microstaging: The key to good management

The categories “superficial spreading” and “nodular” are based on the seminal work of Wallace Clark, who described putative growth phases of cutaneous melanoma.¹² Clark hypothesized that melanoma initially grows horizontally and only later begins an invasive vertical growth phase.

The horizontal growth phase is common in sun-exposed sites and often occurs over a long period of time. The vertical growth phase is a much more aggressive growth pattern that, if left unchecked, can be lethal. This is often seen in nodular melanoma (FIGURE 6).

Measurement of vertical growth (“microstaging”) is the most important prognostic indicator for localized cutaneous melanoma.¹³

Clark described 5 levels (“Clarks’ levels,” I–V) of invasion (FIGURE 7).

Independently of Clark, Breslow described tumor thickness as an important prognostic factor.^14,15 Breslow thickness is measured from the granular layer of the epidermis to the deepest area of invasion (FIGURE 7). Most reports indicate that, overall, Breslow tumor thickness more closely correlates with clinical outcome.¹⁶

Special circumstances that affect microstaging. Although Clark and Breslow measurements are both conceptually simple, a number of factors, including hair follicle involvement and ulceration, must be considered when measuring melanoma depth. It is often useful to have a Breslow thickness and a Clark’s level for primary cutaneous melanomas (both methods are used for staging thin primary lesions).

The location of the primary lesion affects interpretation of measurement results. For example, a Breslow thickness of 0.5 mm confers a different meaning on the eyelid (very thin skin, Clark’s level proportionately deeper than tumor thickness might indicate) than the back (thick skin, proportionately more superficial Clark’s level than tumor thickness would indicate).

Measurement of vertical involvement is used to stage the tumor with the TNM classification. This was recently updated as outlined by Balch et al¹³ in 2001 (TABLE 1).

FIGURE 6
Nodular melanoma

FIGURE 7
Classifying melanoma: Both Clark’s level and Breslow measurement are often used for staging

TABLE 1
Using the TNM staging system to determine prognosis for melanoma

Preferred management and contingencies

Primary treatment for cutaneous melanoma is wide local excision of the primary site.

How wide should excision margins be? The appropriate margin of normal skin (measured from the biopsy scar or lateral border of residual melanoma) varies with Breslow tumor thickness.¹⁷ With thin and intermediate thickness, a surgical margin encompassing 1 cm of normal skin is the consensus.¹⁷ For thicker lesions, many authorities recommend a 2-cm margin. A 1-cm margin may be inadequate due to the risk of local recurrence; however, whether a 2 cm or 3 cm margin is optimal remains unclear.^18,19TABLE 2 reviews treatment guidelines for surgical excision of primary cutaneous melanoma.^20-24

Biopsied tissue submitted as permanent sections. The planned primary excision site should be oriented and submitted to pathology in formalin for permanent section examination. Frozen sections are generally not recommended; when the pathologist assesses surgical margin status, scattered melanocytes in adjacent sun-damaged skin may lead to uncertainty. Moreover, artifacts from the freezing process make interpretation difficult.

2 qualifiers for the above advice: immunostains have increased the utility of frozen sections, and frozen sections were not used in the clinical trials from which current treatment options were derived. These considerations apply particularly to the use of frozen section analysis in Mohs micrographic surgery (see below).

The wide local excision specimen can be examined by multiple surgical pathology orientation methods. In many cases, if the primary lesion was completely excised by the initial biopsy, the pathologist will examine the area of the biopsy scar for residual melanoma and select portions of the lateral margins for examination. Alternatively, the entire lesion, all of the margins, or even the entire specimen may be submitted for examination depending on the clinical circumstances and concerns of the pathologist and surgeon, such as incomplete removal at the time of initial biopsy or close deep or lateral margins. Communication between surgeon and pathologist must be clear and unambiguous.

Mohs micrographic surgery increasingly used for melanoma.^25-31 This procedure may be performed with frozen or permanent sections, using concomitant immunohistochemical staining with either method of embedding. This has been especially useful in cosmetically sensitive areas such as the face, ear, nose, genitalia, and extremities.³⁰ Mohs micrographic surgery with permanent sections also makes possible the most definitive surgical margin analysis.

TABLE 2
Recommended incision margins

Sentinel lymph node biopsy

Lymphatic mapping and sentinel lymph node biopsy have significantly changed the initial approach to the patient with melanoma and have reinvigorated debate over the importance and management of regional lymph nodes in cutaneous malignant melanoma. The anatomic basis of, and techniques for, lymphatic mapping and sentinel lymph node biopsy are well-described elsewhere and are not reviewed here.³²

Histolopathologic data from regional lymph nodes is the most important prognostic factor for cutaneous melanoma,¹³ and sentinel lymph node biopsy is proven as a staging procedure—as reflected in the revised AJCC TNM Staging System published in 2001 (TABLES 1 AND 2).¹³ This staging system differs from the previous one in several ways:

1. Thickness and ulceration are the primary predictors of survival with localized melanoma (stages I and II)
2. Number of metastatic lymph nodes involved and tumor burden are most important predictors of survival in stage III melanoma
3. Anatomic site of metastasis (distant) and presence/absence of elevated lactate dehydrogenase (LDH) are primary predictors of survival with stage IV melanoma
4. Ulceration should prompt an overall upstaging for melanoma stages I–III
5. Satellite metastases around a primary melanoma and in-transit metastases together indicate a stage III melanoma
6. Staging decisions are based on sentinel node biopsy/lymphatic mapping.¹³

Multiple studies have reviewed the changes between the new and old TNM staging systems for melanoma.¹³

When to biopsy. Clinical experience supports reserving lymphatic mapping and sentinel lymph node mapping for patients with primary lesions 1 mm or thicker. Some authorities have recommended sentinel lymph node biopsy for lesions <1 mm thick if they are ulcerated or if a discrepancy exists between the Breslow tumor thickness and Clark’s level. For such lesions, biopsy has been recommended for Clark’s level >III.³³

Though sentinel lymph node biopsy can be performed successfully after wide local excision, doing it before excision will prevent disruption of lymphatics for node mapping.³⁴ The sentinel lymph node should be submitted to pathology for permanent section analysis. Frozen section analysis is not recommended for melanoma due to its lower sensitivity.³⁵

Though not widely adopted, some surgeons use intraoperative touch prep cytology as part of an initial intraoperative evaluation.³⁶ If the surgeon is prepared to proceed with regional node dissection and a “black” grossly involved node is encountered, intraoperative frozen section may be used to confirm metastastic melanoma. Neither touch preps nor frozen sections are likely to demonstrate the small, isolated metastatic tumor cells commonly identified in small, grossly normal lymph nodes.

What to expect from the pathologist. If blue dye is used for localization, the pathologist should document the presence of that color in the lymph nodes submitted. Outside of a clinical trial, fresh lymph node tissue is not usually preserved for polymerase chain reaction (PCR) or other molecular methods for detecting tumor cells.

Although examination techniques vary in some details among institutions, general principles are well accepted and include submission of the entire lymph node, step or serial sections, and, when initial sections do not reveal metastatic tumor, use of immunohistochemistry for melanocyte-associated antigens (see Immunohistochemical techniques and antibody testing).

Fine points regarding interpretation of sentinel lymph nodes. First, the rationale behind biopsy of sentinel lymph nodes is that the lymphatic system is a potential means of metastasis. However, it is not the only means of spread. Hematogenous and tissue spread are potential mechanisms as well.

Second, the lymphatic system acts as a drain but not a dam for lymphatic flow. Thus a negative lymph node biopsy does not guarantee that a metastasis has not already occurred.

Third, the increased sensitivity of identification of positive sentinel lymph nodes with more specific tests (ie, immunohistochemistry, PCR, rtPCR) calls a negative result into question. A sentinel lymph node that is “negative” with immunohistochemistry may in fact be positive with gene amplification techniques (PCR, rtPCR).

Fourth, surgeons performing sentinel node biopsies vary in thoroughness when identifying which node or nodes contain the dye. Lymphoscintigraphy is more accurate than using dyes; however, many surgeons use both.⁴⁴ Thus, while sentinel node biopsy has demonstrated a role in staging, its use and the interpretation of results must be performed with these variables in mind.

Regional lymph node dissection: Does it benefit patients?

Patients found to have metastatic melanoma in a sentinel lymph node are advised to undergo regional lymph node dissection. Unfortunately, few data address survival with this therapy. The role of sentinel lymph node biopsy and elective lymph node dissection in the management of cutaneous malignant melanoma generates debate, much of which focuses on alternate interpretations of the data from the era of elective lymph node dissection for intermediate-thickness melanoma.

The debate over elective lymph node dissection has been summarized by others in detail.⁴⁷

Proponents of an aggressive strategy correctly point out that data from trials are simply not comparable to what is seen in the current era of sentinel lymph node biopsy.

Opponents of an aggressive strategy, who are numerous and vocal,^48-50 correctly observe that no survival benefit has been demonstrated for any treatment strategy based on regional lymph node dissection. However, to be balanced, such long-term outcome data are simply not yet available. Certainly questions will continue to arise regarding the theoretical validity and cost-effectiveness of sentinel lymph node biopsy.^48,49,51

How good is adjuvant therapy?

Adjuvant therapy for cutaneous melanoma can include regional external beam radiation, systemic cytotoxic chemotherapy, or immunotherapy. Radiation has a limited role, being used postoperatively to treat regional lymph node dissection beds at high risk of recurrence and to treat locally recurrent and in-transit disease.⁵²

Systemic cytotoxin (often dacarbazine) has not been effective as adjuvant therapy.⁵³ When combined chemo- and immunotherapy is used for metastatic disease, it is with the hope of discovering an effective combination that might also be used in the adjuvant setting.⁵⁴ However, it has been the initial success with immunotherapy that frames the debate regarding adjuvant systemic therapy for cutaneous malignant melanoma.

High-dose interferon alpha 2b. For thick or node-positive cutaneous malignant melanoma, a statistically significant increase in overall survival with high-dose interferon alpha 2b was demonstrated in a landmark Eastern Cooperative Oncology Group trial.⁵⁵ Though the results have been confirmed in other trials, the clinical benefit is still debated. The expense and toxicity of the trial, as well as the absence of benefit in a second ECOG trial of high-dose systemic therapy, have contributed to the uncertainty.⁵⁶

In an effort to reduce toxicity, low-dose systemic interferon alpha 2b has been tested in numerous trials and found to be of no benefit.⁵⁶ High-dose alpha 2b interferon therapy continues to be a focus of active clinical investigation by national and international cooperative groups. In addition to high-dose systemic interferon, a large number of immunotherapeutic approaches for the systemic treatment of melanoma have been devised and are under active investigation, including non-specific immune adjuvants such as Bacile Calmette-Guerin, levaminsole, vaccines prepared from autologous melanoma cells and vaccines designed against chemically defined antigens (gangliocides).⁵⁷ Because of the controversies surrounding adjuvant systemic therapy, patients are best served by participating in an ongoing clinical trial if the expense and toxicity of high dose systemic interferon alpha 2b are deemed to be unacceptable.

Acknowledgments

We thank Brian C. Brockway, M.S. (Department of Medical Media, Veterans affairs Medical Center, Augusta, Georgia) for illustrations.

CORRESPONDENCE
Joshua E. Lane, MD, 308 Hospital Drive, Suite 200, Macon, GA 31217. E-mail: [email protected]

A 51-year-old man of northern European descent who works outdoors for the city asks you to look at a “mole” on his face. The lesion does not have the classic appearance of a melanoma that you have seen before, and it is still fairly small. Should you advise a wait-and-see approach or perform a biopsy?

Opt for early detection

Given the patient’s likely genetic predisposition to skin cancer and his regular, lengthy exposure to sunlight (see Risk factors), you would be wise to follow your cutaneous examination with a biopsy.

The ABCDs of visual assessment. The classic clinical presentation of melanoma is well known. The most publicized means for identifying potentially atypical pigmented skin lesions is the “ABCD” mnemonic (asymmetry, border irregularity, color variegation, and lesion diameter >6 mm) (FIGURES 1 AND 2).^7,8 The ABCDs are primarily an educational tool for patients. This mnemonic was expanded to include E, representing evolution of a pigmented lesion.⁹ Any pigmented lesion observed to change significantly on serial examinations warrants biopsy to exclude melanoma.

Caveat: not all melanomas are pigmented. Amelanotic melanomas are a diagnostic challenge and may be lethal if left unattended.

Routine screening. Large-scale skin cancer screening has been performed and found to be a statistically ineffective means of detection. Moreover, the US Preventive Services Task Force found insufficient evidence to recommend for or against routine counseling by primary care physicians to prevent skin cancer.¹⁰

A large-scale educational and screening campaign was performed in Italy from 1991 to 1996. During this period, 90,000 educational leaflets were distributed to a target population of approximately 243,000. A total of 2050 individuals requested a skin examination, resulting in detection of 13 melanomas.¹¹ However, 92.3% of the melanomas were thin (<1.52 mm deep). Despite the lack of statistical significance for such screenings, many organizations do perform them and find melanomas, which can be life saving for those few individuals.

Anatomic areas to focus on. While melanoma can affect any anatomic region, it is especially common on sun-exposed areas, including the head, neck, and upper extremities. Acral lentiginous melanoma is found on palmar, plantar, and subungual regions. Include the scalp, ocular mucosa, and oral cavity in your examination.

FIGURE 1
Melanoma in situ

FIGURE 2
Melanoma in situ

Which biopsy technique to choose?

A properly performed biopsy is mandatory to accurately diagnose and microstage the tumor. No other test reliably surpasses routine histologic examination.

When possible, arrange for a full-thickness excisional biopsy with a narrow (2-mm) rim of normal skin, especially for large lesions in which sampling error may be a factor.

Tissue samples are submitted for histologic examination in formalin for permanent sections. Frozen sections are not recommended for the initial diagnosis of melanoma due to artifactual changes from the freezing process.

When lesions are too large for primary excision, or if an optimal cosmetic result is important to the patient, an incisional or punch biopsy may be taken from the area believed to be the deepest portion of the tumor.

For the largest lesions, multiple punch or incisional biopsies may help to reduce the risk of sampling error. Diagnostic and management problems arise when the initial biopsy does not sample the complete skin thickness or when large lesions are not sampled adequately.

Classification of melanoma

The traditional melanoma classification scheme includes several subtypes (see Detailed melanoma classifications). Though these terms have limited clinical utility, we include them for the sake of completeness and because the terminology may arise in consultation with colleagues.

Most important for the prognosis is a determination of the tumor’s depth of invasiveness.

In situ or invasive?

Melanoma in situ lesions are confined to the epidermis and may extend along hair appendages. It often occurs on sun-exposed areas (FIGURES 1 AND 2). Histologically, melanoma in situ is an asymmetric and poorly circumscribed proliferation of melanocytes usually larger than 6 mm in diameter (FIGURE 3).

Melanocytes form irregular nests that are not equidistant and have areas of confluence (FIGURE 3). Single melanocytes predominate over nests in some areas and may entirely replace the basal layer.

Additional histologic features include single melanocytes above the dermal-epidermal junction (Pagetoid spread) and uneven distribution of pigment. In many cases the dermis has inflammatory infiltrates, melanophages, and evidence of solar-damage. Particularly in anatomical areas rich in hair follicles, the neoplastic cells may spread into the epithelium of the hair follicles without extending to the dermis.

Melanoma in situ may be quite large in diameter (horizontal growth phase) without becoming invasive; however, always be concerned about invasion.

Invasive melanoma shares the same histologic features but invades the dermis or subcutaneous fat (vertical growth phase). Aside from the conventional criteria used for the histologic diagnosis of melanoma, acral melanomas may show an increased number of dendritic melanocytes loaded with melanin (FIGURE 4). Desmoplastic melanoma is most commonly found on the head and neck; however, it varies in clinical presentation (FIGURE 5). It is often associated with other types of melanoma, is common in older persons, and has a slight male predilection.

FIGURE 3
Melanocytes

FIGURE 4
Acral melanoma

FIGURE 5
Dermoplastic melanoma

Microstaging: The key to good management

The categories “superficial spreading” and “nodular” are based on the seminal work of Wallace Clark, who described putative growth phases of cutaneous melanoma.¹² Clark hypothesized that melanoma initially grows horizontally and only later begins an invasive vertical growth phase.

The horizontal growth phase is common in sun-exposed sites and often occurs over a long period of time. The vertical growth phase is a much more aggressive growth pattern that, if left unchecked, can be lethal. This is often seen in nodular melanoma (FIGURE 6).

Measurement of vertical growth (“microstaging”) is the most important prognostic indicator for localized cutaneous melanoma.¹³

Clark described 5 levels (“Clarks’ levels,” I–V) of invasion (FIGURE 7).

Independently of Clark, Breslow described tumor thickness as an important prognostic factor.^14,15 Breslow thickness is measured from the granular layer of the epidermis to the deepest area of invasion (FIGURE 7). Most reports indicate that, overall, Breslow tumor thickness more closely correlates with clinical outcome.¹⁶

Special circumstances that affect microstaging. Although Clark and Breslow measurements are both conceptually simple, a number of factors, including hair follicle involvement and ulceration, must be considered when measuring melanoma depth. It is often useful to have a Breslow thickness and a Clark’s level for primary cutaneous melanomas (both methods are used for staging thin primary lesions).

The location of the primary lesion affects interpretation of measurement results. For example, a Breslow thickness of 0.5 mm confers a different meaning on the eyelid (very thin skin, Clark’s level proportionately deeper than tumor thickness might indicate) than the back (thick skin, proportionately more superficial Clark’s level than tumor thickness would indicate).

Measurement of vertical involvement is used to stage the tumor with the TNM classification. This was recently updated as outlined by Balch et al¹³ in 2001 (TABLE 1).

FIGURE 6
Nodular melanoma

FIGURE 7
Classifying melanoma: Both Clark’s level and Breslow measurement are often used for staging

TABLE 1
Using the TNM staging system to determine prognosis for melanoma

Preferred management and contingencies

Primary treatment for cutaneous melanoma is wide local excision of the primary site.

How wide should excision margins be? The appropriate margin of normal skin (measured from the biopsy scar or lateral border of residual melanoma) varies with Breslow tumor thickness.¹⁷ With thin and intermediate thickness, a surgical margin encompassing 1 cm of normal skin is the consensus.¹⁷ For thicker lesions, many authorities recommend a 2-cm margin. A 1-cm margin may be inadequate due to the risk of local recurrence; however, whether a 2 cm or 3 cm margin is optimal remains unclear.^18,19TABLE 2 reviews treatment guidelines for surgical excision of primary cutaneous melanoma.^20-24

Biopsied tissue submitted as permanent sections. The planned primary excision site should be oriented and submitted to pathology in formalin for permanent section examination. Frozen sections are generally not recommended; when the pathologist assesses surgical margin status, scattered melanocytes in adjacent sun-damaged skin may lead to uncertainty. Moreover, artifacts from the freezing process make interpretation difficult.

2 qualifiers for the above advice: immunostains have increased the utility of frozen sections, and frozen sections were not used in the clinical trials from which current treatment options were derived. These considerations apply particularly to the use of frozen section analysis in Mohs micrographic surgery (see below).

The wide local excision specimen can be examined by multiple surgical pathology orientation methods. In many cases, if the primary lesion was completely excised by the initial biopsy, the pathologist will examine the area of the biopsy scar for residual melanoma and select portions of the lateral margins for examination. Alternatively, the entire lesion, all of the margins, or even the entire specimen may be submitted for examination depending on the clinical circumstances and concerns of the pathologist and surgeon, such as incomplete removal at the time of initial biopsy or close deep or lateral margins. Communication between surgeon and pathologist must be clear and unambiguous.

Mohs micrographic surgery increasingly used for melanoma.^25-31 This procedure may be performed with frozen or permanent sections, using concomitant immunohistochemical staining with either method of embedding. This has been especially useful in cosmetically sensitive areas such as the face, ear, nose, genitalia, and extremities.³⁰ Mohs micrographic surgery with permanent sections also makes possible the most definitive surgical margin analysis.

TABLE 2
Recommended incision margins

Sentinel lymph node biopsy

Lymphatic mapping and sentinel lymph node biopsy have significantly changed the initial approach to the patient with melanoma and have reinvigorated debate over the importance and management of regional lymph nodes in cutaneous malignant melanoma. The anatomic basis of, and techniques for, lymphatic mapping and sentinel lymph node biopsy are well-described elsewhere and are not reviewed here.³²

Histolopathologic data from regional lymph nodes is the most important prognostic factor for cutaneous melanoma,¹³ and sentinel lymph node biopsy is proven as a staging procedure—as reflected in the revised AJCC TNM Staging System published in 2001 (TABLES 1 AND 2).¹³ This staging system differs from the previous one in several ways:

1. Thickness and ulceration are the primary predictors of survival with localized melanoma (stages I and II)
2. Number of metastatic lymph nodes involved and tumor burden are most important predictors of survival in stage III melanoma
3. Anatomic site of metastasis (distant) and presence/absence of elevated lactate dehydrogenase (LDH) are primary predictors of survival with stage IV melanoma
4. Ulceration should prompt an overall upstaging for melanoma stages I–III
5. Satellite metastases around a primary melanoma and in-transit metastases together indicate a stage III melanoma
6. Staging decisions are based on sentinel node biopsy/lymphatic mapping.¹³

Multiple studies have reviewed the changes between the new and old TNM staging systems for melanoma.¹³

When to biopsy. Clinical experience supports reserving lymphatic mapping and sentinel lymph node mapping for patients with primary lesions 1 mm or thicker. Some authorities have recommended sentinel lymph node biopsy for lesions <1 mm thick if they are ulcerated or if a discrepancy exists between the Breslow tumor thickness and Clark’s level. For such lesions, biopsy has been recommended for Clark’s level >III.³³

Though sentinel lymph node biopsy can be performed successfully after wide local excision, doing it before excision will prevent disruption of lymphatics for node mapping.³⁴ The sentinel lymph node should be submitted to pathology for permanent section analysis. Frozen section analysis is not recommended for melanoma due to its lower sensitivity.³⁵

Though not widely adopted, some surgeons use intraoperative touch prep cytology as part of an initial intraoperative evaluation.³⁶ If the surgeon is prepared to proceed with regional node dissection and a “black” grossly involved node is encountered, intraoperative frozen section may be used to confirm metastastic melanoma. Neither touch preps nor frozen sections are likely to demonstrate the small, isolated metastatic tumor cells commonly identified in small, grossly normal lymph nodes.

What to expect from the pathologist. If blue dye is used for localization, the pathologist should document the presence of that color in the lymph nodes submitted. Outside of a clinical trial, fresh lymph node tissue is not usually preserved for polymerase chain reaction (PCR) or other molecular methods for detecting tumor cells.

Although examination techniques vary in some details among institutions, general principles are well accepted and include submission of the entire lymph node, step or serial sections, and, when initial sections do not reveal metastatic tumor, use of immunohistochemistry for melanocyte-associated antigens (see Immunohistochemical techniques and antibody testing).

Fine points regarding interpretation of sentinel lymph nodes. First, the rationale behind biopsy of sentinel lymph nodes is that the lymphatic system is a potential means of metastasis. However, it is not the only means of spread. Hematogenous and tissue spread are potential mechanisms as well.

Second, the lymphatic system acts as a drain but not a dam for lymphatic flow. Thus a negative lymph node biopsy does not guarantee that a metastasis has not already occurred.

Third, the increased sensitivity of identification of positive sentinel lymph nodes with more specific tests (ie, immunohistochemistry, PCR, rtPCR) calls a negative result into question. A sentinel lymph node that is “negative” with immunohistochemistry may in fact be positive with gene amplification techniques (PCR, rtPCR).

Fourth, surgeons performing sentinel node biopsies vary in thoroughness when identifying which node or nodes contain the dye. Lymphoscintigraphy is more accurate than using dyes; however, many surgeons use both.⁴⁴ Thus, while sentinel node biopsy has demonstrated a role in staging, its use and the interpretation of results must be performed with these variables in mind.

Regional lymph node dissection: Does it benefit patients?

Patients found to have metastatic melanoma in a sentinel lymph node are advised to undergo regional lymph node dissection. Unfortunately, few data address survival with this therapy. The role of sentinel lymph node biopsy and elective lymph node dissection in the management of cutaneous malignant melanoma generates debate, much of which focuses on alternate interpretations of the data from the era of elective lymph node dissection for intermediate-thickness melanoma.

The debate over elective lymph node dissection has been summarized by others in detail.⁴⁷

Proponents of an aggressive strategy correctly point out that data from trials are simply not comparable to what is seen in the current era of sentinel lymph node biopsy.

Opponents of an aggressive strategy, who are numerous and vocal,^48-50 correctly observe that no survival benefit has been demonstrated for any treatment strategy based on regional lymph node dissection. However, to be balanced, such long-term outcome data are simply not yet available. Certainly questions will continue to arise regarding the theoretical validity and cost-effectiveness of sentinel lymph node biopsy.^48,49,51

How good is adjuvant therapy?

Adjuvant therapy for cutaneous melanoma can include regional external beam radiation, systemic cytotoxic chemotherapy, or immunotherapy. Radiation has a limited role, being used postoperatively to treat regional lymph node dissection beds at high risk of recurrence and to treat locally recurrent and in-transit disease.⁵²

Systemic cytotoxin (often dacarbazine) has not been effective as adjuvant therapy.⁵³ When combined chemo- and immunotherapy is used for metastatic disease, it is with the hope of discovering an effective combination that might also be used in the adjuvant setting.⁵⁴ However, it has been the initial success with immunotherapy that frames the debate regarding adjuvant systemic therapy for cutaneous malignant melanoma.

High-dose interferon alpha 2b. For thick or node-positive cutaneous malignant melanoma, a statistically significant increase in overall survival with high-dose interferon alpha 2b was demonstrated in a landmark Eastern Cooperative Oncology Group trial.⁵⁵ Though the results have been confirmed in other trials, the clinical benefit is still debated. The expense and toxicity of the trial, as well as the absence of benefit in a second ECOG trial of high-dose systemic therapy, have contributed to the uncertainty.⁵⁶

In an effort to reduce toxicity, low-dose systemic interferon alpha 2b has been tested in numerous trials and found to be of no benefit.⁵⁶ High-dose alpha 2b interferon therapy continues to be a focus of active clinical investigation by national and international cooperative groups. In addition to high-dose systemic interferon, a large number of immunotherapeutic approaches for the systemic treatment of melanoma have been devised and are under active investigation, including non-specific immune adjuvants such as Bacile Calmette-Guerin, levaminsole, vaccines prepared from autologous melanoma cells and vaccines designed against chemically defined antigens (gangliocides).⁵⁷ Because of the controversies surrounding adjuvant systemic therapy, patients are best served by participating in an ongoing clinical trial if the expense and toxicity of high dose systemic interferon alpha 2b are deemed to be unacceptable.

Acknowledgments

We thank Brian C. Brockway, M.S. (Department of Medical Media, Veterans affairs Medical Center, Augusta, Georgia) for illustrations.

CORRESPONDENCE
Joshua E. Lane, MD, 308 Hospital Drive, Suite 200, Macon, GA 31217. E-mail: [email protected]

A 51-year-old man of northern European descent who works outdoors for the city asks you to look at a “mole” on his face. The lesion does not have the classic appearance of a melanoma that you have seen before, and it is still fairly small. Should you advise a wait-and-see approach or perform a biopsy?

Opt for early detection

Given the patient’s likely genetic predisposition to skin cancer and his regular, lengthy exposure to sunlight (see Risk factors), you would be wise to follow your cutaneous examination with a biopsy.

The ABCDs of visual assessment. The classic clinical presentation of melanoma is well known. The most publicized means for identifying potentially atypical pigmented skin lesions is the “ABCD” mnemonic (asymmetry, border irregularity, color variegation, and lesion diameter >6 mm) (FIGURES 1 AND 2).^7,8 The ABCDs are primarily an educational tool for patients. This mnemonic was expanded to include E, representing evolution of a pigmented lesion.⁹ Any pigmented lesion observed to change significantly on serial examinations warrants biopsy to exclude melanoma.

Caveat: not all melanomas are pigmented. Amelanotic melanomas are a diagnostic challenge and may be lethal if left unattended.

Routine screening. Large-scale skin cancer screening has been performed and found to be a statistically ineffective means of detection. Moreover, the US Preventive Services Task Force found insufficient evidence to recommend for or against routine counseling by primary care physicians to prevent skin cancer.¹⁰

A large-scale educational and screening campaign was performed in Italy from 1991 to 1996. During this period, 90,000 educational leaflets were distributed to a target population of approximately 243,000. A total of 2050 individuals requested a skin examination, resulting in detection of 13 melanomas.¹¹ However, 92.3% of the melanomas were thin (<1.52 mm deep). Despite the lack of statistical significance for such screenings, many organizations do perform them and find melanomas, which can be life saving for those few individuals.

Anatomic areas to focus on. While melanoma can affect any anatomic region, it is especially common on sun-exposed areas, including the head, neck, and upper extremities. Acral lentiginous melanoma is found on palmar, plantar, and subungual regions. Include the scalp, ocular mucosa, and oral cavity in your examination.

FIGURE 1
Melanoma in situ

FIGURE 2
Melanoma in situ

Which biopsy technique to choose?

A properly performed biopsy is mandatory to accurately diagnose and microstage the tumor. No other test reliably surpasses routine histologic examination.

When possible, arrange for a full-thickness excisional biopsy with a narrow (2-mm) rim of normal skin, especially for large lesions in which sampling error may be a factor.

Tissue samples are submitted for histologic examination in formalin for permanent sections. Frozen sections are not recommended for the initial diagnosis of melanoma due to artifactual changes from the freezing process.

When lesions are too large for primary excision, or if an optimal cosmetic result is important to the patient, an incisional or punch biopsy may be taken from the area believed to be the deepest portion of the tumor.

For the largest lesions, multiple punch or incisional biopsies may help to reduce the risk of sampling error. Diagnostic and management problems arise when the initial biopsy does not sample the complete skin thickness or when large lesions are not sampled adequately.

Classification of melanoma

The traditional melanoma classification scheme includes several subtypes (see Detailed melanoma classifications). Though these terms have limited clinical utility, we include them for the sake of completeness and because the terminology may arise in consultation with colleagues.

Most important for the prognosis is a determination of the tumor’s depth of invasiveness.

In situ or invasive?

Melanoma in situ lesions are confined to the epidermis and may extend along hair appendages. It often occurs on sun-exposed areas (FIGURES 1 AND 2). Histologically, melanoma in situ is an asymmetric and poorly circumscribed proliferation of melanocytes usually larger than 6 mm in diameter (FIGURE 3).

Melanocytes form irregular nests that are not equidistant and have areas of confluence (FIGURE 3). Single melanocytes predominate over nests in some areas and may entirely replace the basal layer.

Additional histologic features include single melanocytes above the dermal-epidermal junction (Pagetoid spread) and uneven distribution of pigment. In many cases the dermis has inflammatory infiltrates, melanophages, and evidence of solar-damage. Particularly in anatomical areas rich in hair follicles, the neoplastic cells may spread into the epithelium of the hair follicles without extending to the dermis.

Melanoma in situ may be quite large in diameter (horizontal growth phase) without becoming invasive; however, always be concerned about invasion.

Invasive melanoma shares the same histologic features but invades the dermis or subcutaneous fat (vertical growth phase). Aside from the conventional criteria used for the histologic diagnosis of melanoma, acral melanomas may show an increased number of dendritic melanocytes loaded with melanin (FIGURE 4). Desmoplastic melanoma is most commonly found on the head and neck; however, it varies in clinical presentation (FIGURE 5). It is often associated with other types of melanoma, is common in older persons, and has a slight male predilection.

FIGURE 3
Melanocytes

FIGURE 4
Acral melanoma

FIGURE 5
Dermoplastic melanoma

Microstaging: The key to good management

The categories “superficial spreading” and “nodular” are based on the seminal work of Wallace Clark, who described putative growth phases of cutaneous melanoma.¹² Clark hypothesized that melanoma initially grows horizontally and only later begins an invasive vertical growth phase.

The horizontal growth phase is common in sun-exposed sites and often occurs over a long period of time. The vertical growth phase is a much more aggressive growth pattern that, if left unchecked, can be lethal. This is often seen in nodular melanoma (FIGURE 6).

Measurement of vertical growth (“microstaging”) is the most important prognostic indicator for localized cutaneous melanoma.¹³

Clark described 5 levels (“Clarks’ levels,” I–V) of invasion (FIGURE 7).

Independently of Clark, Breslow described tumor thickness as an important prognostic factor.^14,15 Breslow thickness is measured from the granular layer of the epidermis to the deepest area of invasion (FIGURE 7). Most reports indicate that, overall, Breslow tumor thickness more closely correlates with clinical outcome.¹⁶

Special circumstances that affect microstaging. Although Clark and Breslow measurements are both conceptually simple, a number of factors, including hair follicle involvement and ulceration, must be considered when measuring melanoma depth. It is often useful to have a Breslow thickness and a Clark’s level for primary cutaneous melanomas (both methods are used for staging thin primary lesions).

The location of the primary lesion affects interpretation of measurement results. For example, a Breslow thickness of 0.5 mm confers a different meaning on the eyelid (very thin skin, Clark’s level proportionately deeper than tumor thickness might indicate) than the back (thick skin, proportionately more superficial Clark’s level than tumor thickness would indicate).

Measurement of vertical involvement is used to stage the tumor with the TNM classification. This was recently updated as outlined by Balch et al¹³ in 2001 (TABLE 1).

FIGURE 6
Nodular melanoma

FIGURE 7
Classifying melanoma: Both Clark’s level and Breslow measurement are often used for staging

TABLE 1
Using the TNM staging system to determine prognosis for melanoma

Preferred management and contingencies

Primary treatment for cutaneous melanoma is wide local excision of the primary site.

How wide should excision margins be? The appropriate margin of normal skin (measured from the biopsy scar or lateral border of residual melanoma) varies with Breslow tumor thickness.¹⁷ With thin and intermediate thickness, a surgical margin encompassing 1 cm of normal skin is the consensus.¹⁷ For thicker lesions, many authorities recommend a 2-cm margin. A 1-cm margin may be inadequate due to the risk of local recurrence; however, whether a 2 cm or 3 cm margin is optimal remains unclear.^18,19TABLE 2 reviews treatment guidelines for surgical excision of primary cutaneous melanoma.^20-24

Biopsied tissue submitted as permanent sections. The planned primary excision site should be oriented and submitted to pathology in formalin for permanent section examination. Frozen sections are generally not recommended; when the pathologist assesses surgical margin status, scattered melanocytes in adjacent sun-damaged skin may lead to uncertainty. Moreover, artifacts from the freezing process make interpretation difficult.

2 qualifiers for the above advice: immunostains have increased the utility of frozen sections, and frozen sections were not used in the clinical trials from which current treatment options were derived. These considerations apply particularly to the use of frozen section analysis in Mohs micrographic surgery (see below).

The wide local excision specimen can be examined by multiple surgical pathology orientation methods. In many cases, if the primary lesion was completely excised by the initial biopsy, the pathologist will examine the area of the biopsy scar for residual melanoma and select portions of the lateral margins for examination. Alternatively, the entire lesion, all of the margins, or even the entire specimen may be submitted for examination depending on the clinical circumstances and concerns of the pathologist and surgeon, such as incomplete removal at the time of initial biopsy or close deep or lateral margins. Communication between surgeon and pathologist must be clear and unambiguous.

Mohs micrographic surgery increasingly used for melanoma.^25-31 This procedure may be performed with frozen or permanent sections, using concomitant immunohistochemical staining with either method of embedding. This has been especially useful in cosmetically sensitive areas such as the face, ear, nose, genitalia, and extremities.³⁰ Mohs micrographic surgery with permanent sections also makes possible the most definitive surgical margin analysis.

TABLE 2
Recommended incision margins

Sentinel lymph node biopsy

Lymphatic mapping and sentinel lymph node biopsy have significantly changed the initial approach to the patient with melanoma and have reinvigorated debate over the importance and management of regional lymph nodes in cutaneous malignant melanoma. The anatomic basis of, and techniques for, lymphatic mapping and sentinel lymph node biopsy are well-described elsewhere and are not reviewed here.³²

Histolopathologic data from regional lymph nodes is the most important prognostic factor for cutaneous melanoma,¹³ and sentinel lymph node biopsy is proven as a staging procedure—as reflected in the revised AJCC TNM Staging System published in 2001 (TABLES 1 AND 2).¹³ This staging system differs from the previous one in several ways:

1. Thickness and ulceration are the primary predictors of survival with localized melanoma (stages I and II)
2. Number of metastatic lymph nodes involved and tumor burden are most important predictors of survival in stage III melanoma
3. Anatomic site of metastasis (distant) and presence/absence of elevated lactate dehydrogenase (LDH) are primary predictors of survival with stage IV melanoma
4. Ulceration should prompt an overall upstaging for melanoma stages I–III
5. Satellite metastases around a primary melanoma and in-transit metastases together indicate a stage III melanoma
6. Staging decisions are based on sentinel node biopsy/lymphatic mapping.¹³

Multiple studies have reviewed the changes between the new and old TNM staging systems for melanoma.¹³

When to biopsy. Clinical experience supports reserving lymphatic mapping and sentinel lymph node mapping for patients with primary lesions 1 mm or thicker. Some authorities have recommended sentinel lymph node biopsy for lesions <1 mm thick if they are ulcerated or if a discrepancy exists between the Breslow tumor thickness and Clark’s level. For such lesions, biopsy has been recommended for Clark’s level >III.³³

Though sentinel lymph node biopsy can be performed successfully after wide local excision, doing it before excision will prevent disruption of lymphatics for node mapping.³⁴ The sentinel lymph node should be submitted to pathology for permanent section analysis. Frozen section analysis is not recommended for melanoma due to its lower sensitivity.³⁵

Though not widely adopted, some surgeons use intraoperative touch prep cytology as part of an initial intraoperative evaluation.³⁶ If the surgeon is prepared to proceed with regional node dissection and a “black” grossly involved node is encountered, intraoperative frozen section may be used to confirm metastastic melanoma. Neither touch preps nor frozen sections are likely to demonstrate the small, isolated metastatic tumor cells commonly identified in small, grossly normal lymph nodes.

What to expect from the pathologist. If blue dye is used for localization, the pathologist should document the presence of that color in the lymph nodes submitted. Outside of a clinical trial, fresh lymph node tissue is not usually preserved for polymerase chain reaction (PCR) or other molecular methods for detecting tumor cells.

Although examination techniques vary in some details among institutions, general principles are well accepted and include submission of the entire lymph node, step or serial sections, and, when initial sections do not reveal metastatic tumor, use of immunohistochemistry for melanocyte-associated antigens (see Immunohistochemical techniques and antibody testing).

Fine points regarding interpretation of sentinel lymph nodes. First, the rationale behind biopsy of sentinel lymph nodes is that the lymphatic system is a potential means of metastasis. However, it is not the only means of spread. Hematogenous and tissue spread are potential mechanisms as well.

Second, the lymphatic system acts as a drain but not a dam for lymphatic flow. Thus a negative lymph node biopsy does not guarantee that a metastasis has not already occurred.

Third, the increased sensitivity of identification of positive sentinel lymph nodes with more specific tests (ie, immunohistochemistry, PCR, rtPCR) calls a negative result into question. A sentinel lymph node that is “negative” with immunohistochemistry may in fact be positive with gene amplification techniques (PCR, rtPCR).

Fourth, surgeons performing sentinel node biopsies vary in thoroughness when identifying which node or nodes contain the dye. Lymphoscintigraphy is more accurate than using dyes; however, many surgeons use both.⁴⁴ Thus, while sentinel node biopsy has demonstrated a role in staging, its use and the interpretation of results must be performed with these variables in mind.

Regional lymph node dissection: Does it benefit patients?

Patients found to have metastatic melanoma in a sentinel lymph node are advised to undergo regional lymph node dissection. Unfortunately, few data address survival with this therapy. The role of sentinel lymph node biopsy and elective lymph node dissection in the management of cutaneous malignant melanoma generates debate, much of which focuses on alternate interpretations of the data from the era of elective lymph node dissection for intermediate-thickness melanoma.

The debate over elective lymph node dissection has been summarized by others in detail.⁴⁷

Proponents of an aggressive strategy correctly point out that data from trials are simply not comparable to what is seen in the current era of sentinel lymph node biopsy.

Opponents of an aggressive strategy, who are numerous and vocal,^48-50 correctly observe that no survival benefit has been demonstrated for any treatment strategy based on regional lymph node dissection. However, to be balanced, such long-term outcome data are simply not yet available. Certainly questions will continue to arise regarding the theoretical validity and cost-effectiveness of sentinel lymph node biopsy.^48,49,51

How good is adjuvant therapy?

Adjuvant therapy for cutaneous melanoma can include regional external beam radiation, systemic cytotoxic chemotherapy, or immunotherapy. Radiation has a limited role, being used postoperatively to treat regional lymph node dissection beds at high risk of recurrence and to treat locally recurrent and in-transit disease.⁵²

Systemic cytotoxin (often dacarbazine) has not been effective as adjuvant therapy.⁵³ When combined chemo- and immunotherapy is used for metastatic disease, it is with the hope of discovering an effective combination that might also be used in the adjuvant setting.⁵⁴ However, it has been the initial success with immunotherapy that frames the debate regarding adjuvant systemic therapy for cutaneous malignant melanoma.

High-dose interferon alpha 2b. For thick or node-positive cutaneous malignant melanoma, a statistically significant increase in overall survival with high-dose interferon alpha 2b was demonstrated in a landmark Eastern Cooperative Oncology Group trial.⁵⁵ Though the results have been confirmed in other trials, the clinical benefit is still debated. The expense and toxicity of the trial, as well as the absence of benefit in a second ECOG trial of high-dose systemic therapy, have contributed to the uncertainty.⁵⁶

In an effort to reduce toxicity, low-dose systemic interferon alpha 2b has been tested in numerous trials and found to be of no benefit.⁵⁶ High-dose alpha 2b interferon therapy continues to be a focus of active clinical investigation by national and international cooperative groups. In addition to high-dose systemic interferon, a large number of immunotherapeutic approaches for the systemic treatment of melanoma have been devised and are under active investigation, including non-specific immune adjuvants such as Bacile Calmette-Guerin, levaminsole, vaccines prepared from autologous melanoma cells and vaccines designed against chemically defined antigens (gangliocides).⁵⁷ Because of the controversies surrounding adjuvant systemic therapy, patients are best served by participating in an ongoing clinical trial if the expense and toxicity of high dose systemic interferon alpha 2b are deemed to be unacceptable.

Acknowledgments

We thank Brian C. Brockway, M.S. (Department of Medical Media, Veterans affairs Medical Center, Augusta, Georgia) for illustrations.

CORRESPONDENCE
Joshua E. Lane, MD, 308 Hospital Drive, Suite 200, Macon, GA 31217. E-mail: [email protected]

Whereas quantitative and qualitative analyses of 41 randomized controlled trials (RCTs) strongly suggests that all types of NSAIDs are more effective than placebo (>50% pain relief) against an acute episode of tension-type headache (TTH), the evidence also shows that no single nonsteroidal anti-inflammatory drug (NSAID) is more effective than another in this setting.

How, then, to choose an NSAID? Many of the 41 articles we reviewed reported on the side effects of NSAIDs. No clear differences were reported in the number of side effects between the NSAIDs and placebo. However, differences were found among the types of NSAIDs. Our results agree with those found by Henry et al,¹ who concluded from their meta-analysis that ibuprofen, compared with other NSAIDs, had the lowest relative risk of serious gastrointestinal complications. Given the lack of important differences in efficacy among NSAIDs for relieving an acute episode of TTH, using the most effective dose of a drug that is well tolerated by a patient is a reasonable basis for selection. Ibuprofen, therefore, generally may be advocated.

When acetaminophen is preferred. Our results suggest NSAIDs might be more effective than acetaminophen for TTH. However, because NSAIDs are allergenic for some people, and they must not be used in association with anticoagulants,² acetaminophen might be an alternative in these situations. When giving acetaminophen, the dose of the medication might be important due to a possible dose-response relationship.

Why this review was needed

Tension-type headache, also known as tension headache or muscle contraction headache, is the most commonly experienced type of headache (see Episodic tension-type headache). Population-based studies suggest prevalence rates of 35% to 40% in adults.^3-5

Persons experiencing an acute episode of TTH most often self-treat with mild, non-narcotic analgesics for initial pain relief. Studies have suggested that acetaminophen and NSAIDs like aspirin, ibuprofen, naproxen, and ketoprofen are effective in reducing headache symptoms. But a variety of drugs, dosages, and combinations have been described. No systematic review has, until now, described the efficacy and tolerability of analgesics for the treatment of acute episodes of TTH. Good quality-controlled trials and a systematic review form the basis for evidence-based treatment guidelines, which provide a basis for the individual patient.

We aimed to describe and assess the data from RCTs concerning the efficacy and tolerability of analgesics for the treatment of acute episodes of TTH in adult patients. Details of our Methods and Results follow.

Methods

Search strategy

Medline and EMBASE were searched from inception to January 2005 using the terms tension-type headache, tension headache, stress headache, or muscle contraction headache together with the search strategy for identifying RCTs described by Robinson and Dickerson.⁶ The Cochrane Controlled Trials Register was searched using the words tension headache or tension-type headache or muscle contraction headache. Additional strategies for identifying trials included searching the reference lists of review articles and included studies.

Study selection

Only RCTs including analgesic medicine used in the treatment or management of TTH conducted among adult patients (aged 18 years or older), with reasonable criteria designed to distinguish TTH from migraine, were selected for our review. The use of a specific set of diagnostic criteria (eg, IHS 1988 and Ad Hoc 1962)^7,8 was not required, but TTH diagnoses had to be based on at least some of the distinctive features of TTH—eg, bilateral in location, no nausea or vomiting, mild or moderate intensity, or no exacerbation by exercise.

Main outcome measures were pain relief or recovery over 2 to 6 hours.

Two authors (LD, AV) independently screened titles and abstracts of identified studies for eligibility. All potentially relevant studies were retrieved as full papers and then again independently reviewed by 2 authors (LD, AV). Disagreements were resolved through consensus where possible, or by arbitration with a third author (MB). Crossover designs often presented data from treatment groups, as if the trial was a parallel group trial. The results from these studies were excluded from data-analysis if no results from both arms were presented or a binary correlation coefficient was available.⁹

Methodological quality and data extraction

Two authors (LD with MB, BK, or AV) independently rated the methodological quality of the included trials using the Delphi list.¹⁰ The Delphi list is a generic criteria list developed by international consensus and consists of the following 9 items: 1) randomization; 2) adequate allocation concealment; 3) groups similar at baseline; 4) specification of eligibility criteria; 5) blinding of outcome assessor; 6) blinding of care provider; 7) blinding of patient; 8) presentation of point estimates and measures of variability; 9) intention-to-treat-analysis. One extra item was added: 10) withdrawal or dropout rate unlikely to cause bias. All selected methodological criteria were scored as yes (=1), no (=0) or don’t know (=0). A quality score of a trial was computed by counting the number of positive scores, with equal weights applied on all items. In case of a disagreement between the 2 authors, consensus was used to resolve disagreement. When consensus could not be reached, a third author made the final decision (MB or AV).

Extraction of data from the original reports was performed by 1 author (LD) and checked by a second (AV). Disagreements were resolved by consensus. Extracted information included (if available) demographic data, detailed description of the intervention and control (ie, dose given, study duration, rescue medication), data on pain relief or recovery, and information on adverse effects measured during a treatment period of 2 to 6 hours. When a trial protocol permitted the use of rescue medication prior to the outcome time (2 to 6 hours), then the latest outcome assessment not confounded by the use of rescue medication was extracted

Data analysis

A quantitative analysis was limited to clinically homogenous studies for which the study populations, interventions and outcomes were considered to be similar. For each study, the number of patients who were recovered (often defined as more than 50% pain relief) was used to calculate relative risk (RR) with 95% confidence interval (CI). RRs and 95% CI were presented using the random effects model. Data are presented as treatment success, indicating that an RR >1 represents a better outcome for the first mentioned medication group.

In parallel studies, when more than 1 comparison from the same study (ie, aspirin 650 mg vs placebo and ibuprofen 400 mg vs placebo) was used for the statistical pooling of NSAIDs vs placebo, the results from the placebo group were evenly spread out over the 2 comparisons and the number of patients in the placebo group was divided by 2 in order to prevent double counting (personal communication RJPM Scholten, Dutch Cochrane Centre).

Because only a subset of available trials provides sufficient data for inclusion in the quantitative analysis, also a qualitative analysis was performed. We summarized findings by strength of evidence, nature of intervention and control treatments. The evidence was judged to be strong when multiple high-quality trials produced generally consistent findings.¹¹ Results were considered consistent if over 75% of the studies reported similar results on the same outcome measure. It was judged to be moderate when multiple low-quality trials or one high-quality and 1 or more low-quality trials produced generally consistent findings. Evidence was considered to be limited when only 1 low-quality RCT existed and conflicting when the findings of existing trials were inconsistent. We arbitrarily regarded trials with methodological quality scores of 6 or more as of high quality.¹¹

Relation between funding source of the RCTs and conclusions

We extracted the sources of funding of the RCTs from the text, statements of sources of support, authors’ affiliations, and acknowledgements. Funding sources were classified as nonprofit organizations, not reported, both nonprofit and for-profit organizations, or for-profit organizations.¹² For-profit organizations were defined as companies that might acquire financial gain or loss depending on the outcome of the trial.¹² Funding included provision of grants, study material (drug, placebo), or manpower (authorship, statistical analysis, or other assistance).¹² We used the effect sizes between medication(s) and placebo to evaluate whether funding source affected outcome.

Results

Search results

A total of 1878 publications were identified by our search strategy. Finally, 41 RCTs met our inclusion criteria and 4 papers concerned double publications (FIGURE 1),^13-16 leaving a total of 41 trials which were included in this review. Thirteen of these RCTs used a crossover design.^15,7-27

FIGURE 1
How the 41 trials made our cut for the review

Description of studies

Full details of the included studies are presented in TABLE W1. The number of participants included in each trial ranged from 12 to 900 (mean=252.7 patients), with a total of 10,363 patients included. The mean percentage of participants who dropped out from the trials was 15.2% (range=0%–61.9%). Age of participants (for studies reporting this information) ranged from 18 to 87 years. Overall, the percentage of women was generally higher than men (mean=69.3%; range=36%–97%). Fifteen trials used the criteria of the International Headache Society to classify TTH,^{14,17,19-21,24,28-36} 12 trials used the Ad Hoc Committee’s criteria,^{13,23,26,37,45} while the remaining studies did not use a formal classification.

Twenty-five studies compared 1 or more types of NSAIDs with placebo,^{13-17,22-24,26-36,38,41-43,45-47} 17 studies compared 1 or more doses of acetaminophen with placebo,^{17-21,25,30-34,41,44-46,48,49} 7 studies compared different types of NSAIDs,^{15,26,28,29,35-37} 9 studies compared 1 or more types of NSAIDs with acetaminophen,^{17,30-34,41,45,46} and 13 studies compared other analgesics with placebo.^{15,18,25,27,39,40,44,49,50-53}

The quality score (with positive items in parenthesis) is presented in the “Notes” section of TABLE W1. The interobserver reliability of the methodological quality assessment was high (κ=0.85). There was disagreement between the 2 authors in 7.5% of the criteria, but after consensus no disagreement persisted. The median quality score was 5 (range 1–9). Using a cutoff point of 6 out of 10 criteria, 15 studies (36.6%) were considered to be of high quality.^{15,17,19,21,22,24,25,28-30,32-34,36}

Only 1 study reported a concealed randomization method.³⁴ Other methodological flaws, which often scored “negative” or “unclear,” were blinding of the care provider (unclear 88%) and an intention-to-treat analysis (unclear 30% and negative 60%).

Effectiveness of analgesics

TABLE 1 gives the quantitative analysis for high-quality studies, low-quality studies, and for all studies for the different comparisons of NSAIDs, acetaminophen, and placebo.

TABLE 1
Quantitative analysis for the different studies for the comparisons of NSAIDs, acetaminophen and placebo

1. NSAIDs vs placebo

Twenty-five studies compared one or more types of NSAIDs with placebo, of which 10 are of high quality.^{15,17,22,24,29,30,32-34,36,45}

Quantitative analysis. Sufficient data were available in 15 studies,^{13,14,29-38,41,45,47} of which 6 were of high quality.^{29,30,32-34,36,45} Because some trials included 3 or more treatment groups, data were available for 28 comparisons. We found a significant effect in favor of NSAIDs compared with placebo on short-term pain relief (see TABLE 1 and FIGURE W1).

Qualitative analysis. The 10 high-quality studies reported 30 comparisons, of which in 26 (86.6%) NSAIDs were significantly more effective compared with placebo for short-term pain relief (strong evidence).

Adverse events. Twenty studies reported during a 2 to 6 hours treatment period data on adverse events. For the NSAID group (n=2061) frequently mentioned side effects were nausea (4.6%), photophobia (3.1%), vomiting (2.7%), phonophobia (1.7%), aching limbs (1.2%), dizziness (1.1%), and drowsiness (1.0%). For the placebo group (n=1323), these were nausea (7.0%), photophobia (4.8%), vomiting (3.9%), phonophobia (3.4%), aching limbs (2.0%), drowsiness (1.7%), and dizziness (1.0%). The pooled RR for the number of patients reporting side effects for 14 studies with sufficient data was 0.96 (95% CI, 0.7–1.3), indicating no significant difference.

2. Acetaminophen vs placebo

Seventeen studies compared 1 or more doses of acetaminophen with placebo; 9 were high-quality studies.^{17,19,21,25,30-34,45}

Quantitative analysis. The pooled analysis of 5 high-quality trials^30,32-34,45 and 3 low-quality trials^31,41,44 showed that acetaminophen was significantly more effective compared with placebo for patients on short-term pain relief (TABLE 1 and FIGURE W2). This result was due to the studies comparing acetaminophen with placebo. The only high-quality trial³⁴ with acetaminophen 500 mg failed to show a difference in short-term pain relief compared with placebo (TABLE 1).

Qualitative analysis. The 9 high-quality studies reported 16 comparisons, of which 10 (62.5%) mentioned that acetaminophen showed significantly more pain relief than placebo (conflicting evidence). In 2 high-quality studies,^17,34 we found no significant differences between acetaminophen 500 mg and placebo (strong evidence), but in the 9 high-quality studies, in 10 out of 14 comparisons (71.4%) acetaminophen 1000 mg showed significantly more pain relief compared with placebo (conflicting evidence).

Adverse events. Twelve studies reported data on adverse events. For the acetaminophen group (n=3715), frequently mentioned side effects were stomach discomfort (3.9%), dizziness (1.6%), nervousness (0.7%), nausea (0.4%), and drowsiness (0.3%). For the placebo group (n=3700), these were stomach discomfort (3.7%), nervousness (0.7%), nausea (0.6%), dizziness (0.5%), and drowsiness (0.3%). The pooled RR for the number of patients reporting side effects was 1.3 (95% CI, 0.9–1.7), indicating no significant difference.

3. NSAIDs vs acetaminophen

Nine studies compared 1 or more types of NSAIDs with acetaminophen, of which 6 are of high-quality.^17,30-34,45

Quantitative analysis. The pooled analysis of 5 high-quality studies^30-34,45 and 2 low-quality studies^31,41 showed a significant difference in short-term pain relief in favor of NSAIDs (TABLE 1).

Qualitative analysis. Six high-quality studies showed that in 9 out of 13 comparisons (69%) NSAIDs were not significantly more effective than acetaminophen for short-term pain relief in patients with acute episodes of TTH (conflicting evidence).

Adverse events. Seven studies reported data on adverse events. The pooled RR for number of patients reporting side effects was 1.3 (95% CI, 0.97–1.6), indicating no significant difference.

4. Comparison between different NSAIDs

Seven studies compared different types of NSAIDs,^{15,26,28,29,35-37} of which 4 provided data.

Quantitative and qualitative analysis. The analysis the between different types of NSAIDs no differences in short-term pain relief can be found; RR vary between 0.9 and 1.5 (TABLE 1).

Adverse events. The adverse effects were reported involving the central nervous system (ie, dizziness, drowsiness, vertigo), gastrointestinal system (ie, nausea, vomiting, gastrointestinal upset or discomfort), and the body as a whole (ie, light-headed, fatigue, cramps, asthenia, chills).

Naproxen and zomepirac gave more adverse events involving the central nervous system than aspirin, ibuprofen, and ketoprofen. Naproxen and zomepirac were also more often associated with gastrointestinal side effects than ibuprofen and ketoprofen.

Furthermore, aspirin was more associated with gastrointestinal complaints than ibuprofen. Side effects such as fatigue and cramps (body as whole) occurred significantly more often with ketoprofen compared with aspirin and ibuprofen, naproxen compared with ketoprofen, and zomepirac compared with aspirin.

5. Other analgesics vs placebo

Qualitative analysis. There is insufficient evidence to either support or refute the effectiveness of all other analgesics compared with placebo, due to the fact that most analgesics were a unique combination of analgesics with caffeine or peppermint oil. Also, the low methodological quality of nearly all these studies and the low number of studies per comparison made drawing conclusions difficult.

Optalidon and Tonopan were compared with placebo in 3 substudies of 1 high-quality study, and we found significant more pain relief using these analgesics than placebo.¹⁵ No adverse events were stated in these studies.

The combination of acetaminophen and caffeine was compared with placebo in 2 studies of high quality^25,49 showed that the combination of acetaminophen with caffeine is more effective than placebo (moderate evidence).

The combination of acetaminophen, aspirin, and caffeine was compared with placebo in 4 substudies of the same high-quality study.²⁵ Data from these studies suggest that this combination is significantly more effective than placebo. All groups reported low numbers of side effects as stomach discomfort, nervousness, and dizziness.

Relation between funding source and effect estimates

The pooled effect estimates in placebo-controlled trials stratified by funding are shown in TABLE 2. No major differences in effect sizes were found between the different funding sources.

TABLE 2
Relation between funding source and effect estimate, intervention vs placebo only

Methodological quality of included studies

This review shows that many RCTs on the efficacy of analgesics in TTH have methodological shortcomings. Using a cut-off point of 6 out of 10 criteria, only 35% of the included studies were found to be of high quality. Most authors failed to explicitly specify the method of treatment allocation and blinding procedure. In many studies authors stated that the trial had a double-blind procedure, however, when the blinding procedure was not explicitly reported (ie, identical looking tablets) we did not score 1 or more blinding items positive. These flaws can be prevented in future trials.

We are unaware of any prior systematic reviews or meta-analyses that have assessed the efficacy and tolerability of analgesics in the treatment of acute episodes of tension-type headache in adults. We conducted the review according to the high Cochrane standard, resulting in a review of high validity. Our review succeeded in identifying a large number of only randomized trials. Also the methodological quality did not explain the possible association between funding and effect estimates.

Although systematic reviews offer the least biased method of summarizing research literature, our review should be considered with the following limitations in mind. First, we decided not to contact the authors for additional information, because most trials were published before 1995. Second, some of the medications have only been evaluated in 1 or 2 studies, which may limit the generalizability of the findings. We do not think these factors have influences our conclusions.

CORRESPONDENCE
Arianne P. Verhagen, PhD, Department of General Practice, Erasmus Medical Centre, PO Box 1738, 3000 DR Rotterdam, The Netherlands. E-mail: [email protected]

Whereas quantitative and qualitative analyses of 41 randomized controlled trials (RCTs) strongly suggests that all types of NSAIDs are more effective than placebo (>50% pain relief) against an acute episode of tension-type headache (TTH), the evidence also shows that no single nonsteroidal anti-inflammatory drug (NSAID) is more effective than another in this setting.

How, then, to choose an NSAID? Many of the 41 articles we reviewed reported on the side effects of NSAIDs. No clear differences were reported in the number of side effects between the NSAIDs and placebo. However, differences were found among the types of NSAIDs. Our results agree with those found by Henry et al,¹ who concluded from their meta-analysis that ibuprofen, compared with other NSAIDs, had the lowest relative risk of serious gastrointestinal complications. Given the lack of important differences in efficacy among NSAIDs for relieving an acute episode of TTH, using the most effective dose of a drug that is well tolerated by a patient is a reasonable basis for selection. Ibuprofen, therefore, generally may be advocated.

When acetaminophen is preferred. Our results suggest NSAIDs might be more effective than acetaminophen for TTH. However, because NSAIDs are allergenic for some people, and they must not be used in association with anticoagulants,² acetaminophen might be an alternative in these situations. When giving acetaminophen, the dose of the medication might be important due to a possible dose-response relationship.

Why this review was needed

Tension-type headache, also known as tension headache or muscle contraction headache, is the most commonly experienced type of headache (see Episodic tension-type headache). Population-based studies suggest prevalence rates of 35% to 40% in adults.^3-5

Persons experiencing an acute episode of TTH most often self-treat with mild, non-narcotic analgesics for initial pain relief. Studies have suggested that acetaminophen and NSAIDs like aspirin, ibuprofen, naproxen, and ketoprofen are effective in reducing headache symptoms. But a variety of drugs, dosages, and combinations have been described. No systematic review has, until now, described the efficacy and tolerability of analgesics for the treatment of acute episodes of TTH. Good quality-controlled trials and a systematic review form the basis for evidence-based treatment guidelines, which provide a basis for the individual patient.

We aimed to describe and assess the data from RCTs concerning the efficacy and tolerability of analgesics for the treatment of acute episodes of TTH in adult patients. Details of our Methods and Results follow.

Methods

Search strategy

Medline and EMBASE were searched from inception to January 2005 using the terms tension-type headache, tension headache, stress headache, or muscle contraction headache together with the search strategy for identifying RCTs described by Robinson and Dickerson.⁶ The Cochrane Controlled Trials Register was searched using the words tension headache or tension-type headache or muscle contraction headache. Additional strategies for identifying trials included searching the reference lists of review articles and included studies.

Study selection

Only RCTs including analgesic medicine used in the treatment or management of TTH conducted among adult patients (aged 18 years or older), with reasonable criteria designed to distinguish TTH from migraine, were selected for our review. The use of a specific set of diagnostic criteria (eg, IHS 1988 and Ad Hoc 1962)^7,8 was not required, but TTH diagnoses had to be based on at least some of the distinctive features of TTH—eg, bilateral in location, no nausea or vomiting, mild or moderate intensity, or no exacerbation by exercise.

Main outcome measures were pain relief or recovery over 2 to 6 hours.

Two authors (LD, AV) independently screened titles and abstracts of identified studies for eligibility. All potentially relevant studies were retrieved as full papers and then again independently reviewed by 2 authors (LD, AV). Disagreements were resolved through consensus where possible, or by arbitration with a third author (MB). Crossover designs often presented data from treatment groups, as if the trial was a parallel group trial. The results from these studies were excluded from data-analysis if no results from both arms were presented or a binary correlation coefficient was available.⁹

Methodological quality and data extraction

Two authors (LD with MB, BK, or AV) independently rated the methodological quality of the included trials using the Delphi list.¹⁰ The Delphi list is a generic criteria list developed by international consensus and consists of the following 9 items: 1) randomization; 2) adequate allocation concealment; 3) groups similar at baseline; 4) specification of eligibility criteria; 5) blinding of outcome assessor; 6) blinding of care provider; 7) blinding of patient; 8) presentation of point estimates and measures of variability; 9) intention-to-treat-analysis. One extra item was added: 10) withdrawal or dropout rate unlikely to cause bias. All selected methodological criteria were scored as yes (=1), no (=0) or don’t know (=0). A quality score of a trial was computed by counting the number of positive scores, with equal weights applied on all items. In case of a disagreement between the 2 authors, consensus was used to resolve disagreement. When consensus could not be reached, a third author made the final decision (MB or AV).

Extraction of data from the original reports was performed by 1 author (LD) and checked by a second (AV). Disagreements were resolved by consensus. Extracted information included (if available) demographic data, detailed description of the intervention and control (ie, dose given, study duration, rescue medication), data on pain relief or recovery, and information on adverse effects measured during a treatment period of 2 to 6 hours. When a trial protocol permitted the use of rescue medication prior to the outcome time (2 to 6 hours), then the latest outcome assessment not confounded by the use of rescue medication was extracted

Data analysis

A quantitative analysis was limited to clinically homogenous studies for which the study populations, interventions and outcomes were considered to be similar. For each study, the number of patients who were recovered (often defined as more than 50% pain relief) was used to calculate relative risk (RR) with 95% confidence interval (CI). RRs and 95% CI were presented using the random effects model. Data are presented as treatment success, indicating that an RR >1 represents a better outcome for the first mentioned medication group.

In parallel studies, when more than 1 comparison from the same study (ie, aspirin 650 mg vs placebo and ibuprofen 400 mg vs placebo) was used for the statistical pooling of NSAIDs vs placebo, the results from the placebo group were evenly spread out over the 2 comparisons and the number of patients in the placebo group was divided by 2 in order to prevent double counting (personal communication RJPM Scholten, Dutch Cochrane Centre).

Because only a subset of available trials provides sufficient data for inclusion in the quantitative analysis, also a qualitative analysis was performed. We summarized findings by strength of evidence, nature of intervention and control treatments. The evidence was judged to be strong when multiple high-quality trials produced generally consistent findings.¹¹ Results were considered consistent if over 75% of the studies reported similar results on the same outcome measure. It was judged to be moderate when multiple low-quality trials or one high-quality and 1 or more low-quality trials produced generally consistent findings. Evidence was considered to be limited when only 1 low-quality RCT existed and conflicting when the findings of existing trials were inconsistent. We arbitrarily regarded trials with methodological quality scores of 6 or more as of high quality.¹¹

Relation between funding source of the RCTs and conclusions

We extracted the sources of funding of the RCTs from the text, statements of sources of support, authors’ affiliations, and acknowledgements. Funding sources were classified as nonprofit organizations, not reported, both nonprofit and for-profit organizations, or for-profit organizations.¹² For-profit organizations were defined as companies that might acquire financial gain or loss depending on the outcome of the trial.¹² Funding included provision of grants, study material (drug, placebo), or manpower (authorship, statistical analysis, or other assistance).¹² We used the effect sizes between medication(s) and placebo to evaluate whether funding source affected outcome.

Results

Search results

A total of 1878 publications were identified by our search strategy. Finally, 41 RCTs met our inclusion criteria and 4 papers concerned double publications (FIGURE 1),^13-16 leaving a total of 41 trials which were included in this review. Thirteen of these RCTs used a crossover design.^15,7-27

FIGURE 1
How the 41 trials made our cut for the review

Description of studies

Full details of the included studies are presented in TABLE W1. The number of participants included in each trial ranged from 12 to 900 (mean=252.7 patients), with a total of 10,363 patients included. The mean percentage of participants who dropped out from the trials was 15.2% (range=0%–61.9%). Age of participants (for studies reporting this information) ranged from 18 to 87 years. Overall, the percentage of women was generally higher than men (mean=69.3%; range=36%–97%). Fifteen trials used the criteria of the International Headache Society to classify TTH,^{14,17,19-21,24,28-36} 12 trials used the Ad Hoc Committee’s criteria,^{13,23,26,37,45} while the remaining studies did not use a formal classification.

Twenty-five studies compared 1 or more types of NSAIDs with placebo,^{13-17,22-24,26-36,38,41-43,45-47} 17 studies compared 1 or more doses of acetaminophen with placebo,^{17-21,25,30-34,41,44-46,48,49} 7 studies compared different types of NSAIDs,^{15,26,28,29,35-37} 9 studies compared 1 or more types of NSAIDs with acetaminophen,^{17,30-34,41,45,46} and 13 studies compared other analgesics with placebo.^{15,18,25,27,39,40,44,49,50-53}