Off-label Use of Azathioprine in Dermatology: Title and subTitle BreakA Systematic ReviewOff-Label Azathioprine Use

Azathioprine is an immune-modulating drug widely used in medicine today. Originally, azathioprine was developed and used to prevent graft rejection in transplant surgery, but soon after its introduction, application expanded to a range of autoimmune and immune-mediated diseases. The therapeutic effect is thought to derive from its purine antimetabolism, and azathioprine is often seen as a corticosteroid-sparing agent. In the United States, azathioprine use in dermatology is off-label. Although in most European countries the use of azathioprine is licensed for the treatment of pemphigus vulgaris and dermatomyositis (http://www.eudrapharm.eu), it is extensively used off-label.

In an off-label setting, patients are treated with drugs that are not registered for that specific indication.¹ Only pharmaceutical companies can apply for registration of drugs for specific indications. However, the interests of pharmaceutical companies and those of physicians and patients often differ regarding the registration of drugs in difficult-to-treat and refractory dermatologic conditions or in rare conditions with subsequently scarce therapeutic options. As a result, in daily practice, many drugs are prescribed off-label, approaching 50% of prescriptions in dermatology.¹

International and national regulatory agencies are applying new rules and regulations concerning off-label medication use (http://www.fda.gov, http://www.mhra.gov.uk, and http://www.rivm.nl). Moreover, dermatologists can be held accountable for any unfavorable effects that result from off-label prescription.

To ensure evidence-based health care decision making, guidelines should be developed for drugs that are often prescribed off-label. Because azathioprine is one such drug, we aimed to summarize the available evidence in the literature regarding the efficacy, effectiveness, and safety of off-label treatment with azathioprine in dermatologic patients. On the basis of this evidence and taking clinical experience into account, we formulated clinical recommendations for the application of azathioprine in specific conditions.

Included were randomized controlled trials (RCTs), cohort studies, and case series with more than 5 patients in which dermatologic patients treated with off-label azathioprine were assessed to determine the efficacy, effectiveness, and/or safety of the drug with or without concomitant therapy. Excluded were reviews, case series with fewer than 5 patients, animal and in vitro studies, abstracts, unpublished data, articles concerning topical treatment, articles concerning diseases primarily treated by other specialists, and duplicate publications. No restrictions were imposed regarding age, sex, skin type, or date of publication. Only articles in English, French, German, or Dutch were included.

On October 9, 2009, a literature search of the MEDLINE (January 1, 1950, to October 1, 2009), EMBASE (January 1, 1980, to October 9, 2009), and CENTRAL databases (1996-2009) was performed. The keyword azathioprine and its synonyms were used in combination with the names of skin diseases (Table 1). Reference lists of included articles and reviews regarding azathioprine were hand-searched to find additional eligible articles.

Of the articles located, the title and/or abstract was screened for off-label azathioprine treatment in dermatologic patients and selected if considered potentially relevant. To determine eligibility, the full text of the selected articles was assessed according to the predefined inclusion and exclusion criteria. Data concerning methodologic quality, study characteristics, efficacy, and safety were extracted using a data extraction form. Study characteristics included study population (number, age range, sex of included patients, dermatologic disease concerned, previous treatments, and inclusion criteria for participation), design of the study (study type, duration of treatment, dosage of azathioprine, and duration of follow-up period), publication date, safety and severity outcome measurements used, and subsequent outcomes, including onset of action and duration of remission. Preferably, changes in mean objective outcome measurements from baseline to end of active treatment were reported. If this was not possible, percentages of patients with a specific outcome were presented.

All stages of study selection and data extraction were performed by 2 reviewers (M.E.S. and C.M.J.M.B.) independently. Disagreements regarding study selection and data extraction were solved by discussion.

We assessed RCTs according to the criterion grading system described in the Cochrane Handbook for Systematic Reviews of Interventions 5.0.2 (updated in February 2008).² To assess the risk of bias, the following measurements were assessed: sequence generation, concealment of allocation, masking (of participants, researchers, and outcome assessment), handling of withdrawals and losses, and other potential threats to validity. Each item was scored as adequate, inadequate, or unclear. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) checklist for observational studies was used to identify methodologic errors in cohort studies.³ Case reports were not assessed for methodologic quality.

Each reviewer graded the total body of evidence found for each disease according to the grading system proposed by Robinson et al.⁴ Because it is inappropriate to formulate safety conclusions based on information derived from studies with relatively small case numbers and short follow-up periods, we compared the adverse events (AEs) and serious adverse events (SAEs) reported in the included studies with the safety data from the Summary of Product Characteristics dated May 27, 2009 (http://www.medicines.org.uk/emc/medicine/2881/SPC/Imuran+Tablets+25mg/). Particular attention was given to the identification of AEs that seemed to be specific for an indication.

The clinical recommendations are based on the available evidence derived from the literature in combination with considerations such as known AEs, patient preferences, availability of other treatment options, organizational aspects, social consequences, and costs. Two reviewers (M.E.S. and C.M.J.M.B.) independently assessed the methodologic quality, strength of evidence, and magnitude of effect and generated clinical recommendations balancing the benefits, burdens, risks, and costs. All details were discussed with the members of the working group.

The Figure summarizes the selection process. An initial search retrieved 3870 articles. After screening the title and/or abstract for eligibility, 148 articles were selected. The main reason for exclusion was lack of relevance (eg, use of azathioprine in registered indications). After screening the full texts of the articles, 43 articles matched our inclusion and exclusion criteria, of which 11 were RCTs, 2 were cohort studies, and 30 were case series.^{5 - 47} In total, 1128 patients with 12 different dermatologic diseases were described. Study demographics are listed in Table 2.

Concomitant treatment was allowed in most studies, not reported in 9 studies, and not allowed in 3. Because azathioprine is often regarded and used as a corticosteroid-supporting and corticosteroid-sparing agent, oral corticosteroids were most frequently used as concomitant medication (n = 20). Oral corticosteroids were also used to support patients between initiation and the onset of action of azathioprine, after which the corticosteroid dosage was tapered. Topical corticosteroids were allowed in 9 studies. Diagnostic criteria and disease definitions were often not specified. Also, the definition of the onset of action was mostly absent.

The severity outcome measurements used were heterogeneous. Objective and validated severity outcome measurements were used in only 11 (25.6%) of the studies and included the Six Area, Six Sign Atopic Dermatitis (SASSAD) severity score; Severity Scoring of Atopic Dermatitis index; Psoriasis Area Severity Index; modified Rodnan skin score; scores on visual analog scales; and percentage of body surface area affected by disease.^{48 - 50} Nonvalidated outcome measurement included the modified Psoriasis Area Severity Index and the Clinical Severity Score. In most of the studies, only descriptive means were given to indicate changes in disease severity.

In 4 studies, thiopurine methyltransferase (TPMT) activity was measured before the initiation of azathioprine therapy. In 2 of these studies, the dosage of azathioprine was adjusted to the degree of TPMT activity, and efficacy was shown not to be altered by this.^{32 - 33} Onset of action varied from 1 to 28 weeks and occurred on average 4 to 6 weeks after the initial dosage. Duration of remission varied from 0 to 93 months.

Study results are summarized in Table 3. Studies reporting azathioprine efficacy in patients with 2 different dermatologic diseases were discussed separately for each disease.

As seen in the risk of bias table, the methodologic quality of the included RCTs varied substantially (Table 4). Adequate randomization was performed in 7 (63.6%) of the studies and was unclear or not performed in 4 (36.4%). Concealment of allocation was judged unclear in 3 studies (27.3%) and inadequate in 2 (18.2%). Masking was judged adequate in 3 studies (27.3%) and was inadequate or unclear in 8 (72.7%). Summary of the quality of evidence and the evidence-based recommendations for clinical practice are given in Table 5.

All 10 studies^{8 - 9 ,22 - 23 ,26 ,28 ,30 ,32 - 33 ,35} concerning azathioprine in patients (n = 319) with moderate to severe atopic dermatitis, whose conditions were often refractory to conventional treatment, showed an overall decrease in disease severity after active treatment with azathioprine. Two included RCTs that showed that azathioprine treatment had results significantly more favorable than placebo treatment.^{8 ,33} A mean SASSAD improvement of 26% when receiving azathioprine treatment was found after 3 months of treatment with 2.5 mg/kg of azathioprine in the study by Berth-Jones et al.⁸ Noticeably, 12 of the 37 patients withdrew during azathioprine treatment, 4 of whom did so because of AEs, compared with 4 of the 37 patients who withdrew during placebo treatment. Because an intention-to-treat analysis was performed, the risk of overestimation of the effect is considered minimal. Meggitt et al³³ showed a mean SASSAD improvement of 37% in the azathioprine group vs 20.6% in the placebo group on intention-to-treat analysis (P < .01). Six patients in the azathioprine group withdrew, compared with 1 in the placebo group. A meta-analysis of the RCTs could not be performed because of lack of data.

Eight case series, of which 3 were prospective, were included concerning 221 patients with atopic dermatitis. One study³⁵ was performed solely with children and showed a good to excellent response in 85% of the patients. In the study by Hon et al,²² a mean Severity Scoring of Atopic Dermatitis improvement of 27% was seen during a 6-month period. The cohort study by Meggitt and Reynolds³² found a mean SASSAD improvement of 26%, which was statistically significant. In all the other studies, descriptive means were used to report changes in disease severity. In those studies, at least 60% of the patients had a favorable response to azathioprine, although no clear definition was given. In 3 studies, complete remissions were reported in 40.5% to 58.3% of the patients. Onset of action of azathioprine ranged from 1 week to 7 months.

Four SAEs were described when using azathioprine to treat atopic dermatitis: severe pancytopenia (TPMT activity not measured), pancreatitis, non-Hodgkin lymphoma, and a ruptured cerebral aneurysm. The last 2 SAEs occurred in the follow-up period.

Seven published studies^{5 ,7 ,10 - 11 ,18 - 19 ,41} were found in the literature in which patients with bullous pemphigoid were treated with azathioprine: 3 RCTs, 1 cohort study, and 3 case series. In total, 252 patients were enrolled, with a mean age of approximately 76 years. In all patients, the diagnosis of bullous pemphigoid was established by histologic analysis of skin biopsy specimens. Concerning the RCTs, Beissert et al⁷ compared methylprednisolone plus azathioprine with methylprednisolone plus mycophenolate mofetil. Results show that in both groups, 100% remission was achieved. Time to remission was more prompt in the azathioprine group, although this was not statistically significant. Burton et al¹¹ found a significant reduction (45%) of the cumulative prednisone dosage during a 3-year period when adding azathioprine. Although the percentage of remissions and number of deaths were lower in the prednisone plus azathioprine group, it was not significant. In the study by Guillaume et al,¹⁹ prednisolone monotherapy was compared with prednisolone plus azathioprine and prednisolone plus plasma exchange. This trial was interrupted after the interim analysis showed no appreciable benefit resulting from the addition of azathioprine or from adding plasma exchange to prednisolone. A meta-analysis could not be performed because of clinical and methodologic heterogeneity.

In the cohort study by Ahmed et al,⁵ 2 groups were compared: prednisone and prednisone plus azathioprine. A 50% reduction in the maintenance dosage of prednisolone was found in the azathioprine group and a 30% reduction was found in the duration of drug therapy. Regarding the case series, Greaves et al¹⁸ found that the dosage of prednisolone at which the patients relapsed was lower when patients were using concomitant azathioprine. They found that by giving azathioprine, it was possible to withdraw prednisolone maintenance treatment in 9 of 11 patients without relapses. In the follow-up study by Burton and Greaves,¹⁰ 44% of these patients remained in remission during a 4-year period. One study⁴¹ used azathioprine as monotherapy and showed a good to very good response in 4 of the 5 patients.

Because of the mean age of the study participants, the relatively long follow-up period, and the severity of the disease, SAEs were not uncommon. Fourteen patient deaths occurred, of which 7 were unspecified, 3 were due to cerebral vascular accidents, 2 were due to heart diseases, 1 was due to an adenocarcinoma, and 1 was due to a preexistent mamma carcinoma.

In total, 3 studies^{29 ,34 ,47} were included in which evidence for the efficacy of azathioprine in patients with chronic actinic dermatitis was found. Murphy et al³⁴ performed a double-masked, placebo-controlled RCT. At treatment months 1, 3, and 6, the mean reduction in rash score and extent of rash and itch on a visual analog scale was statistically different between the azathioprine and placebo groups. The trial was prematurely terminated because the clinical status of actively treated patients showed marked improvement compared with placebo. The included RCT has serious limitations concerning randomization, concealment of allocation, and masking. Thereby, the 2 treatment groups were not comparable concerning severity of the disease at baseline and UV exposure during the trial.

In the 2 case series, 14 patients with chronic actinic dermatitis were included. In total, the conditions of 57% to 92% of the patients cleared or improved markedly. Three patients died, of a cerebral vascular accident, airway disease, or heart disease, respectively, after 12 to 15 months of azathioprine treatment.

One nonmasked RCT and 2 case series were eligible for this study.^{12 - 13 ,20} Regarding the RCT, prednisolone with azathioprine (2.0 mg/kg/d) was compared with different conventional antirheumatic treatments (including nonsteroidal anti-inflammatory drugs, hydroxychloroquine, aurothioglucose, penicillamine, and sulfasalazine) in 19 patients with cutaneous vasculitis characterized by palpable purpura, ulcers, nail fold infarcts, peripheral gangrene, and histologic features of small vessel infiltration.²⁰ In the azathioprine group, a significant reduction of skin vasculitis occurred after 18 months; 7 patients (88%) achieved complete remission. This finding was not statistically significant compared with the conventional treatment group. One patient in the azathioprine group withdrew because of gastrointestinal adverse effects.

Concerning the methodologic quality, details regarding randomization and concealment of allocation were not given, and no masking was performed. Furthermore, the causes of vasculitis were not specified, and it was unclear how the severity of vasculitis was measured and by whom.

Of the 19 patients included in the case series, 83% to 100% were responsive to azathioprine treatment and were able to lower or discontinue their dosage of prednisone.^{12 - 13} One patient experienced pancreatitis during the trial. Two other SAEs of infectious origin occurred: septic arthritis and epidural abscess. One patient with severe and rapid-onset vasculitis died of renal failure during treatment with azathioprine, which could be attributed to the natural course of the underlying disease.

One nonmasked RCT in which patients with leprosy type 1 reaction were treated with azathioprine (3.0 mg/kg/d) and concomitant prednisolone (40 mg/d to 5 mg/d) or prednisolone monotherapy was found to be eligible.³¹ In total, 40 patients were enrolled. At week 12, skin symptoms improved in 52% of the patients in the azathioprine plus prednisolone group vs 65% of the patients in the prednisolone-only group. At 12 and 24 weeks, the nerve function (sensory and voluntary muscle testing) showed no improvement in either group. Nerve pain and tenderness improved in both groups. No statistically significant difference was observed between the groups concerning all these outcomes. The RCT has serious limitations because the study was not masked, it was inadequate in randomization and concealment of allocation, and the outcome tool used was not validated.

Four studies^{40 ,42 ,44 - 45} concerning azathioprine treatment in parthenium dermatitis were found: 1 single-masked, parallel group RCT; 1 cohort comparing dosage regimens; and 2 case series. Concerning the RCT, 73% in the azathioprine group (100 mg/d) vs 72% in the betamethasone group (2 mg/d) had an excellent response to treatment.⁴⁴ During follow-up, 45% had a relapse in the azathioprine group compared with 67% in the betamethasone group (P > .05). It was unclear in the RCT who measured the severity of parthenium dermatitis and whether this was performed in a masked fashion. Also, a substantial number of patients were lost to follow-up without specifications given.

In the cohort study, 3 dosage regimens were compared: 100 mg daily, 50 mg daily plus 300 mg every 28 days, and 100 mg/d plus 300 mg every 28 days.⁴² Complete remission was achieved by 11 patients (50%) in group 1, as well as 2 patients (18%) in group 2 and 3 patients (30%) in group 3. These results were not statistically significant. Patients who were using azathioprine for less than 6 months were excluded from analysis. Furthermore, the groups were not clearly described and the outcome measurement not defined. Of the 32 patients with parthenium dermatitis included in the case series, approximately 60% had a greater than 80% reduction in severity scores.

In total, 6 eligible case series^{6 ,15 ,17 ,21 ,27 ,46} were found in the literature, in which 147 patients with moderate to severe psoriasis, erythrodermic psoriasis, and/or arthritis psoriatica, whose conditions were often refractory to conventional therapy, were treated. Overall, the case studies gave similar results for effectiveness, with approximately 80% of the patients experiencing marked improvement. Three SAEs were described during treatment: a pulmonary embolus, severe anemia, and myocardial infarction.

Four studies^{14 ,25 ,36 - 37} were eligible in which patients with scleroderma or cutaneous involvement of systemic sclerosis were treated: a nonmasked RCT and 3 case series. In the RCT, cyclophosphamide (2 mg/kg/d) combined with prednisolone was more effective than azathioprine (2.5 mg/kg/d) combined with prednisolone after 30 months.³⁶ The RCT had serious limitations in randomization (according to the dates of birth of the individuals studied), concealment of allocation, and masking. The case series demonstrated some beneficial effects of azathioprine in combination with prednisone.

A nonmasked RCT was performed comparing oral psoralen–UV-A (PUVA) photochemotherapy twice per week and azathioprine (0.6-0.75 mg/d) with PUVA monotherapy (twice/wk) in 92 patients with vitiligo.³⁹ Thirty-two patients were lost to follow-up during the study (reasons not specified). After 4 months, the mean total repigmentation rate was 58.4% in the PUVA plus azathioprine group vs 24.8% in the PUVA-only group (P < .001). Combining phototherapy with systemic immunomodulating drugs is not advised by the working group because of the risk of developing nonmelanoma skin cancer.

For these diseases, the evidence was limited to 1 case series with fewer than 10 patients per disease.^{13 ,16 ,24 ,38 ,43} Therefore, the quality of evidence is considered low (level C) for each disease and the estimate of effect uncertain. Efficacy and safety details are given in Table 3.

A strong clinical recommendation was given for the use of azathioprine as an alternative treatment in atopic dermatitis. High quality of evidence was found for a moderate therapeutic response. Although the efficacy is not as pronounced as for cyclosporine (SASSAD improvement of 39% to 57%),⁵¹ azathioprine can be considered an alternative treatment option for severe atopic dermatitis.

A level 2A recommendation was given for the use of azathioprine in bullous pemphigoid in combination with oral corticosteroids. Moderate strength of evidence (level B) was found for a moderate therapeutic response of azathioprine with concomitant corticosteroids, although no significant benefit was found for combined treatment with azathioprine and oral corticosteroids compared with monotherapy with oral corticosteroids. Azathioprine is considered to be a corticosteroid-sparing agent that can reduce corticosteroid-related adverse events. A level 2A recommendation was given for azathioprine in combination with topical or oral corticosteroids as an alternative treatment option for parthenium dermatitis. Whether these results can be extrapolated to all forms of airborne contact dermatitis remains uncertain. Also, a level 2A clinical recommendation was given for chronic actinic dermatitis, leprosy type 1 reactions (in combination with oral corticosteroids), and cutaneous vasculitis (in combination with oral corticosteroids). Evidence of a moderate therapeutic effect existed, although the strength of evidence was moderate (level B or B/C). Azathioprine should only be considered if other treatments (registered or known alternatives) have failed. A moderate to good treatment response was found for erythema multiforme, hand dermatitis with lymphedema, lichen planus, pityriasis rubra pilaris, and psoriasis. Because this was mostly based on only 1 case series per disease, the strength of evidence was considered very low (level C), and a weak recommendation (level 2B) was given. Evidence generated from these studies should be used as a proof of concept and thus should be followed by well-designed further research with a higher level of evidence. Because no beneficial effect of azathioprine on scleroderma seemed to exist, a level 2B recommendation against the use in scleroderma (in combination with oral corticosteroids) was given. Although a moderate treatment response for azathioprine in combination with PUVA was found for vitiligo, a level 2A recommendation against its use in this context was given because of an unfavorable risk-benefit ratio.

Overall, the strength of evidence concerning safety issues was low to very low (level C). When comparing the adverse events stated in the Summary of Product Characteristics with the reported AEs and SAEs, we conclude that the reported AEs and SAEs corresponded with the known adverse effects or that they might be explained by the age of the patients and/or the severity of their disease (eg, bullous pemphigoid). Gastrointestinal concerns were common (n = 120), as were increased liver enzyme levels (n = 131) and mild infections (n = 36). Severe infections were reported in 3 events. Azathioprine-induced pancreatitis was observed twice. Malignant neoplasms occurred 3 times: an adenocarcinoma; a mamma carcinoma, which was preexistent; and a non-Hodgkin lymphoma 8 months after discontinuation of azathioprine therapy. Concerns exist that prolonged treatment with azathioprine predisposes patients to tumors, predominantly squamous cell carcinomas and non-Hodgkin lymphomas, because carcinogenicity is well recognized in organ transplant recipients treated with azathioprine and other immunosuppressive drugs.^{52 - 53} Squamous cell carcinomas did not occur in our study, which could be explained by the relatively short duration of treatment and follow-up. Patients treated for short to medium periods, as in inflammatory bowel disease, appear not to be at increased risk for tumors.⁵⁴ The myelosuppressive effect of azathioprine was commonly reported: leukocytopenia (n = 127), mainly lymphocytopenia (n = 71), thrombocytopenia (n = 9), anemia (n = 9), and pancytopenia (n = 2). Most of these cases were mild and transient. However, severe myelotoxic reactions were seen. Because the risk for severe myelotoxic effects of azathioprine is increased in individuals with a TPMT deficiency, measuring TPMT activity before the initiation of azathioprine therapy is advised and might prevent life-threatening myelosupression.^{55 - 56}

The studies in this review were mostly of a short-term nature, with limited populations and often incomplete or missing reports of AEs, probably because of the emphasis on efficacy in the study design. Therefore, conclusions regarding (long-term) safety could not be drawn. Nonetheless, the reported AEs and SAEs did not reveal any particular new safety concern.

The main difficulty in assessing the off-label efficacy and safety of azathioprine was the generally poor quality of study design (inadequate randomization, concealment of allocation, and masking procedures) of RCTs and other studies in general, the heterogeneity of the data, the high dropout rates, and a lack in comprehensive reporting. A major source of heterogeneity was the use of different and unvalidated diagnostic and outcome measurements and the many studies in which only descriptive means were used. Therefore, the numerous case series were prone to selection and publication bias, and the retrospective design raised concerns regarding recruitment bias and confounding.

In this review, we excluded case series with fewer than 5 patients because we considered the additional value of including approximately 300 case series with fewer than 5 patients to be minimal and because doing so would lead to an unclear overview. To overcome the fact that potentially important safety issues could be missed, we selected a random sample of 20 case reports and screened those articles for relevant safety issues. Because we did not encounter any, we concluded it was appropriate to exclude those case reports. In 2 studies that compared their results against an undefined control group, the control groups were excluded from analysis and the studies were considered case series. In studies concerning systemic conditions (eg, systemic sclerosus or vasculitis), patients without dermatologic involvement were excluded.

This review clearly reveals the need to standardize the manner in which studies regarding off-label treatment are being conducted. It is vital to perform methodologically sound studies using standardized and validated diagnostic and outcome measures to reduce heterogeneity and thus to enable meta-analysis.^{57 - 58} As shown, the available evidence for off-label treatment is scarce in some indications, and promising case series are often not followed by well-designed RCTs. If we do not move forward and start generating a high level of evidence for (rare) diseases, for which we do not have (many) effective or safe treatments, improvement of treatment options will stagnate. Initiatives to generate high levels of evidence, such as the United Kingdom Dermatology Clinical Trials Network, are necessary. Therefore, long-term registries, such as the Psonet initiative for psoriasis, should be created to generate long-term safety data for off-label drug use within dermatology.⁵⁹ Comparing our results with the data provided in standard textbooks,^{60 - 62} which are used in many medical faculties and are widely supported by professionals, showed that this review can be of extra value to new versions of textbooks.

Results from this review can be used to update or formulate clinical practice guidelines for off-label azathioprine prescription and thus to ensure evidence-based health care decision making. Future studies should be performed to confirm the role of azathioprine in diseases that currently have a weak recommendation because of low quality of evidence but in which azathioprine has shown promising results.

Correspondence: Mandy E. Schram, MD, Department of Dermatology, Academic Medical Center, Room A0-252, University of Amsterdam, PO Box 22700, 1100 DE Amsterdam, the Netherlands (m.e.schram@amc.uva.nl).

Accepted for Publication: November 2, 2010.

Author Contributions: All authors had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Schram, Borgonjen, Bik, van der Schroeff, van Everdingen, and Spuls. Acquisition of data: Schram, Borgonjen, and Bik. Analysis and interpretation of data: Schram, Borgonjen, Bik, van Everdingen, and Spuls. Drafting of the manuscript: Schram, Borgonjen, Bik, and Spuls. Critical revision of the manuscript for important intellectual content: Schram, Borgonjen, Bik, van der Schroeff, van Everdingen, and Spuls. Statistical analysis: Schram, Borgonjen, and Bik. Obtained funding: Spuls. Administrative, technical, and material support: Schram, Borgonjen, and Bik. Study supervision: van der Schroeff, van Everdingen, and Spuls.

Financial Disclosure: None reported.

Funding/Support: This study was supported in part by the Dutch Society of Dermatology and Venereology.

Role of the Sponsors: The sponsors had no role in the design and conduct of the study; in the collection, analysis, and interpretation of data; or in the preparation, review, or approval of the manuscript.

Additional Information: Drs Borgonjen and Bik contributed equally to this article

Additional Contributions: We are indebted to Monique E. Wessels, MSc, of the Dutch Order of Medical Specialists. We thank Carla A. F. M. Bruijnzeel, MD, PhD, Sjan (A. P. M.) Lavrijsen, MD, PhD, Eugène P. van Puijenbroek, MD, PhD, Lode Wigersma, MD, PhD, H. Bing Thio, MD, PhD, Kees (A. C.) van Grootheest, MD, PhD, and Elke M. G. J. de Jong, MD, PhD, of the Off-Label Project Group of the Dutch Society of Dermatology and Venereology. We thank Pieter van der Valk, MD, PhD, William R. Faber, MD, PhD, Amber Y. Goedkoop, MD, PhD, Annemieke Horikx, PharmD, PhD, Rutger I. F. van der Waal, MD, PhD, Annemieke Floor-Schreudering, PharmD, Wouter Goldschmidt, MD, PhD, and Noortje (E. L.) Swart, PharmD, PhD, of the Off-Label Working Group of the Dutch Society of Dermatology and Venereology.