Efficacy and Safety of Topical Sirolimus Therapy for Facial Angiofibromas in the Tuberous Sclerosis Complex : A Randomized Clinical Trial

Key Points

Question  What is the optimal concentration of topical sirolimus for the treatment of facial angiofibromas in tuberous sclerosis complex?

Findings  In this randomized clinical trial of 18 children and 18 adults, sirolimus gel at a concentration of 0.2% demonstrated significantly more improvement at the end of treatment compared with placebo, with minimum toxic effects.

Meaning  Topical treatment with a 0.2% concentration of sirolimus gel will be a useful medication for facial angiofibromas in tuberous sclerosis complex.

Importance  Inhibitors of mammalian target of rapamycin complex 1, such as sirolimus, effectively target skin lesions in tuberous sclerosis complex (TSC). However, systemic treatment causes adverse effects, and topical sirolimus has shown promise in the treatment of facial angiofibromas.

Objective  To evaluate the efficacy, safety, and optimal concentration of a topical sirolimus gel vs placebo for treatment of facial angiofibromas in TSC.

Design, Setting, and Participants  A double-blind, placebo-controlled, parallel-group, dose-escalation, phase 2 randomized clinical trial using 3 sirolimus gel concentrations was performed at Osaka University Hospital, Osaka, Japan. Thirty-six patients with TSC and facial angiofibromas, including 18 aged 3 to 18 years (children) and 18 aged 19 to 65 years (adults), were enrolled from December 10, 2013, to July 17, 2014. Analysis was by intention to treat.

Interventions  The adult and child groups were each subdivided into 3 groups (n = 12 each) and randomized to receive sirolimus gel concentrations of 0.05%, 0.1%, or 0.2% or placebo using a web-response system in a 2:1 fashion. The medication was applied to the patient’s lesions twice per day for 12 weeks. Each patient underwent assessment at 2, 4, 8, and 12 weeks during treatment and at 4 weeks after discontinuation of the treatment (16 weeks).

Main Outcomes and Measures  The primary end point, planned before starting data collection, was an improvement factor, represented as a variable composed of tumor size reduction and a lessening of the redness of the 3 target tumors at 12 weeks relative to baseline.

Results  All 36 patients (13 male and 23 female; median age, 40 years; range, 6-47 years) completed the study analyses. The improvement factor was statistically significant in all active treatment groups receiving 0.2% sirolimus (mean [SD], 1.94 [0.68]; P < .001) and not in the adult subgroups receiving 0.1% (mean [SD], 0.88 [0.85]; P = .31) and 0.05% (mean [SD], 1.63 [1.11]; P = .09) concentrations of sirolimus. No significant adverse effects were observed. Mild skin dryness (13 patients [36%]) and irritation (11 patients [31%]) were observed. Low blood levels of sirolimus (<0.25 ng/mL) were detected in adults (1 patient [25%] in the 0.1% adult subgroup and 2 patients [50%] in the 0.2% adult subgroup) and particularly in children (1 patient [25%] in the 0.05% child subgroup, 2 patients [50%] in the 0.1% child subgroup, and 4 patients [100%] in the 0.2% child subgroup).

Conclusions and Relevance  Topical sirolimus gel is safe and effective for facial angiofibromas in TSC. The optimal concentration of sirolimus was 0.2%.

Trial Registration  umin.ac.jp Identifier: UMIN000012420

Tuberous sclerosis complex (TSC) is an autosomal dominant inherited disorder that is characterized by systemic hamartomas, epilepsy, cognitive impairment, and hypopigmented macules.¹Quiz Ref ID Tuberous sclerosis complex is a result of the constitutive activation of mammalian target of rapamycin complex 1 (mTORC1)^2,3 through the dysfunction of products encoded by hamartin (TSC1 [UniProtKB Q92574])⁴ or tuberin (TSC2 [UniProtKB P49815]).⁵ Therefore, mTORC1 inhibitors, such as sirolimus, have become therapeutic agents⁶ that are effective for systemic lesions, including skin lesions.^7-16 However, the systemic administration of mTORC1 inhibitors causes various adverse effects.¹⁷ Topical sirolimus treatment is suitable for TSC skin lesions because it is an easy, painless treatment with fewer toxic effects than the systemic treatment. Recently, many case reports and comparative tests using topical sirolimus formulations modified from internal medicine^18-33 have been reported, but no clinical trials. This study describes the efficacy, safety, and optimal concentration of topical sirolimus formulations prepared using the strict universal criteria of the good manufacturing practice (GMP) system, which permits use of bulk powder for facial angiofibromas in TSC in humans in the context of a randomized clinical trial.

This study was approved by the institutional review board of the Osaka University Hospital, Osaka, Japan. All patients voluntarily participated in this trial and provided written informed consent. A legal representative (eg, parent or guardian) provided written consent on behalf of children and patients with cognitive disability. The trial protocol complies with the Good Clinical Practice guidelines per the ministerial ordinance on clinical trials and the ethical principles of the declaration of Helsinki and is available in Supplement 1.

Concentrations of 0.5 mg, 1 mg, and 2 mg of GMP-level sirolimus bulk powder (Fujian Kerui Pharmaceutical Co, Ltd) were mixed with a 100-mg gel (main ingredients are ethanol and carboxyvinyl polymer) to form 0.05%, 0.1%, and 0.2% concentrations of sirolimus gel in a GMP-level pharmaceutical department in the Osaka University Hospital. This gel showed the best absorption of sirolimus compared with other vehicles.

This study was a double-blind, placebo-controlled, parallel-group, dose-escalation, phase 2 randomized clinical trial that aimed to evaluate the efficacy and safety of topical sirolimus formulations and to estimate the optimal concentration. To obtain statistical significance, 36 patients were required for the statistical assumption based on the incidence of the disease and the effective rate of drug treatment (eMethods in Supplement 2). We identified patients with definite TSC based on the clinical diagnostic criteria in the guidelines of the Japanese Dermatological Association,³⁴ which are similar to the criteria of Roach et al³⁵ for 18 patients aged 3 to 18 years (child group) and 18 patients aged 19 to 65 years (adult group) with facial angiofibromas who intended to participate in this trial.

Based on the Pharmaceutical and Medical Devices Agency guidance for a sample size of 36 patients (eMethods in Supplement 2), the 3 sirolimus gel concentrations (0.05%, 0.1%, and 0.2%) and the possible age range of participants from 3 to 65 years were approved. The trial started with the 0.05% concentration of sirolimus gel treatment for the adult group. After we reviewed the safety of the 0.05% concentration for all participants at 4 weeks, we performed a trial of the 0.1% concentration of sirolimus gel for the adult group. A trial of the 0.2% concentration of sirolimus gel was performed for the adult group after confirming the safety of the 0.1% concentration. For children, each trial was performed after safety confirmations of each treatment concentration in the adult group were completed (Figure 1).

The patients enrolled in this trial were aged 6 to 47 years, had a definite clinical diagnosis of TSC according to the diagnostic criteria used in Japan, and had at least 3 isolated angiofibromas that were 2 mm or larger in their longest diameters. We excluded patients who had used generalized mTOR inhibitors within the past 12 months; had received surgical treatments, including laser treatments, within the past 6 months; or had used topical tacrolimus within the past 3 months.

Eligible patients were divided into adults (aged 19-65 years) and children (aged 3-18 years). Each age group was further divided into 3 subgroups and randomized to treatments with the 0.05%, 0.1%, or 0.2% concentration of sirolimus gel using a web-response system in a 2:1 fashion to receive each sirolimus concentration or a placebo (Figure 1). Each concentration of sirolimus gel or placebo was applied to the lesions twice per day for 12 weeks. Each patient underwent assessment at 2, 4, 8, and 12 weeks during treatment and at 4 weeks after discontinuation of the treatment (16 weeks). To assess treatment efficacy, the diameter and redness of 3 target tumors were measured, and all test lesions were photographed at every visit. To assess treatment safety, blood tests were conducted at Osaka University Hospital, and blood sirolimus concentrations were analyzed by LSI Medience Ltd, a pharmaceutical analysis center, at every visit.

The primary end point was an improvement factor (range, 4.0 to −2.0, with lower scores indicating less improvement), which is represented as a variable composed of tumor volume reduction and lessening of the redness of the 3 target tumors at 12 weeks relative to baseline. Secondary end points included an improvement factor at 16 weeks, defined as a tumor volume reduction and lessening of the redness of the 3 target tumors at 12 weeks relative to baseline. Tumor volume was calculated as (long diameter × short diameter × short diameter) × 1/2.^36,37 A score of 2.0 indicated an 80% or greater reduction of angiofibroma volume relative to baseline; 1.0, a 50% to 80% reduction; 0.5, a 20% to 50% reduction; 0.0, less than a 20% reduction or an increase; and −1.0, a 20% or greater increase. To assess the redness of tumor objectively, a 6-stage redness grading scale was defined by dividing the range between the highest red color (scale 5) and normal skin color (scale 0) into 6 levels, ranging from 1 to 6 points, using a color form guide (Pantone Color Bridge Coated [C]; Pantone). A score of 1 indicated the color 489 C or a lighter color, whereas a score of 6.0 indicated a more intense red than the color 704 C. The value for redness was calculated as the difference of the summed scores of 3 angiofibroma targets from baseline to 12 weeks. For the improvement score of redness, a score of 2.0 indicates a change of 9 or more in the angiofibroma redness value relative to baseline; 1.0, a change of 6 to 8; 0.5, a change of 3 to 5; 0.0, a change of 2 to −2; and −1, a change of −3 to −12 (eFigure in Supplement 2). As the change of redness became pale, the evaluation value is shown as a minus volume. Improvement score of redness is used to calculate the improvement factor, which is represented as the sum of the improvement score of redness and the score of tumor volume reduction.

Other prespecified exploratory end points in the trial included improvement of white macules, erythema, and height of the plaques. Improvement of white macules was measured as the ΔL using a spectrophotometer at baseline and completion (12 weeks), as described elsewhere.²⁶ Reduction of the height of plaques was divided into progressive, unchanged, less than 25% shortening, 25% to 75% shortening, and more than 75% shortening. Improvement of erythema was assessed by a redness reduction that was measured as the ΔA using a spectrophotometer at baseline and completion (12 weeks).

General improvement and patient satisfaction were also examined as secondary end points. General improvement was evaluated using the following 5 stages: remarkable improvement, moderate improvement, slight improvement, unchanged, and worse. Patient satisfaction was evaluated in 5 grades as extremely satisfied, satisfied, slightly satisfied, dissatisfied, and quite dissatisfied. Statistical analysis was conducted via the same method as the improvement factor. The correlation of the improvement factor with general improvement or patient satisfaction was calculated.

We monitored adverse events throughout the study. At each patient visit, a blood test was performed, and the blood sirolimus concentration was examined. The double-blind clinical trial study was performed using a web-response system for all patients.

We used Wilcoxon rank sum tests and Shirley-Williams multiple comparison tests^38,39 to assess the improvement factor (primary end point), reduction of tumor volume, lessening of redness, degree of general improvement, and patient satisfaction before and after the topical sirolimus treatments (1 sided, 5% significance level). We used Spearman rank correlation analysis to assess correlation among the improvement factor, general improvement, and patient satisfaction. Analysis was by intention to treat, and P < .05 was considered significant.

From December 10, 2013, to July 17, 2014, 36 patients aged 6 to 47 years (median age, 40) and undergoing treatment at Osaka University Hospital were randomized to one of the study treatments. All study participants were included in the analysis. Of the 36 patients, 18 were 18 years or younger (child group), and 18 patients were 19 years or older (adult group). In the placebo group, 6 patients were male and 6 were female. In the 0.05% sirolimus group, 6 patients were male and 2 were female. In the 0.1% sirolimus group, 4 patients were male and 4 were female. In the 0.2% sirolimus group, 1 patient was male and 7 were female. The baseline characteristics are summarized in eTable 1 in Supplement 2.

The mean (SD) improvement factor for all patient groups at the end of the treatment (12 weeks) was statistically significant in the 0.2% (1.94 [0.68]; P < .001, Wilcoxon rank sum and Shirley-Williams tests), 0.1% (1.06 [0.62]; P = .06, Wilcoxon rank sum test; P = .03, Shirley-Williams test), and 0.05% (1.63 [0.95]; P = .01, Wilcoxon rank sum test; P = .01, Shirley-Williams test) sirolimus subgroups (Figure 2A). The mean (SD) improvement factor in the child group was also statistically significant in the 0.2% (2.25 [0.87]; P = .01, Wilcoxon rank sum test; P = .003, Shirley-Williams test), 0.1% (1.25 [0.29]; P = .04, Wilcoxon rank sum test; P = .04, Shirley-Williams test), and 0.05% (1.63 [0.95]; P = .03, Wilcoxon rank sum test; P = .03, Shirley-Williams test) sirolimus subgroups (Figure 2B). In the adult group, the mean (SD) improvement factor was statistically significant only in the 0.2% sirolimus subgroup (1.63 [0.25]; P = .048, Wilcoxon rank sum test; P = .03, Shirley-Williams test). We found no significant difference in the adult 0.1% (mean [SD], 0.88 [0.85]; P = .31, Wilcoxon rank sum test; P = .18, Shirley-Williams test) or 0.05% (mean [SD], 1.63 [1.11]; P = .09, Wilcoxon rank sum test) sirolimus subgroups (Figure 2C).

The tumor volume for all patients at the end of treatment was statistically significant in the 0.2% (mean [SD], −35% [9%]; P = .004, Wilcoxon rank sum test; P = .002, Shirley-Williams test), 0.1% (mean [SD], −27% [9%]; P = .06, Wilcoxon rank sum test; P = .048, Shirley-Williams test), and 0.05% (mean [SD], −38% [9%]; P = .03, Wilcoxon rank sum test; P = .003, Shirley-Williams test) sirolimus subgroups. Change in tumor volume in the child group was statistically significant only in the 0.2% sirolimus subgroup (mean [SD], −29% [8%]; P = .03, Wilcoxon rank sum test; P = .005, Shirley-Williams test). No significant difference was observed in the adult subgroups (Figure 3A).

Reduction of tumor redness for all patients at the end of treatment was statistically significant in the 0.2% (mean [SD], −5.9 [0.9]; P = .001, Wilcoxon rank sum and Shirley-Williams tests), 0.1% (mean [SD], −3.7 [0.9]; P = .10, Wilcoxon rank sum test; P = .03, Shirley-Williams test), and 0.05% (mean [SD], −4.9 [0.9]; P = .005, Wilcoxon rank sum test; P = .001, Shirley-Williams test) sirolimus subgroups. Reduction of tumor redness in the child group was also statistically significant in the 0.2% (mean [SD], −6.0 [1.2]; P = .03, Wilcoxon rank sum test; P = .007, Shirley-Williams test), 0.1% (mean [SD], −2.4 [1.4]; P = .14, Wilcoxon rank sum test; P = .04, Shirley-Williams test), and 0.05% (mean [SD], −5.6 [1.3]; P = .048, Wilcoxon rank sum test; P = .02, Shirley-Williams test) sirolimus subgroups. In the adult group, reduction of tumor redness was statistically significant only in the 0.2% sirolimus subgroup (mean [SD], −5.0 [1.2]; P = .048, Wilcoxon rank sum test; P = .04, Shirley-Williams test) (Figure 3B).

The improvement factor at 4 weeks after discontinuation of treatment (16 weeks) was statistically significant for all patients in the 0.2% (mean [SD], 1.2 [0.5]; P = .03, Wilcoxon rank sum test; P = .04, Shirley-Williams test), 0.1% (mean [SD], 1.2 [0.5]; P = .04, Wilcoxon rank sum test; P = .04, Shirley-Williams test), and 0.05% (mean [SD], 1.4 [0.9]; P = .03, Wilcoxon rank sum test; P = .03, Shirley-Williams test) sirolimus subgroups (Figure 2). Unlike the result for all patients, we observed no significant difference in the adult and child subgroups.

Representative photographs of patients in each sirolimus subgroup are shown in Figure 4. Although the facial angiofibromas in all patients improved, those in the child group (Figure 4A) improved more than those in the adult group (Figure 4B) for all sirolimus concentrations. In the child and adult groups, the 0.2% concentration of sirolimus gel was the most effective. In the child subgroup that received the 0.2% concentration of sirolimus gel, the number of angiofibromas was reduced, the tumor size diminished, and tumor redness faded. Some of the angiofibromas disappeared completely, and forehead plaques became flat and discolored (Figure 4, Patient 1). In addition, with the 0.05% concentration of sirolimus gel, the angiofibromas diminished and became discolored, demonstrating that the 0.05% treatment concentration was particularly effective for the child subgroups.

We analyzed general improvement among patients using the Wilcoxon rank sum test and Shirley-Williams multiple comparison test. All the treatment subgroups except the 0.1% and 0.05% adult subgroups showed statistically significant differences at the end of the treatment (Table and eTable 2 in Supplement 2).

We analyzed patient satisfaction using the Wilcoxon rank sum test. The child subgroups who underwent active treatment showed statistically significant differences in median score (interquartile range) compared with the child placebo subgroup (0.2% sirolimus subgroup: 2.00 [2.00-2.50]; P = .02; 0.1% sirolimus subgroup: 2.00 [1.00-3.00]; P = .02; and 0.05% sirolimus subgroup: 2.50 [1.50-3.00]; P = .04). However, only the 0.05% sirolimus subgroup among all patients showed a statistically significant difference at 12 weeks (median [interquartile range], 3.00 [1.50-3.00]; P = .02) (Table and eTable 3 in Supplement 2).

We performed a Spearman correlation analysis on the primary end point (improvement factor) and for general improvement and patient satisfaction. The Spearman correlation coefficient between the improvement factor and general improvement was −0.584 (P < .001), thus confirming a correlation of the improvement factor and general improvement. Similarly, a Spearman correlation coefficient between the improvement factor and patient satisfaction was −0.477 (P = .003), thus confirming a correlation of the improvement factor and patient satisfaction.

During the clinical trial, a status epilepticus event occurred in the adult placebo subgroup, and a case of severe pneumothorax occurred in the adult 0.2% sirolimus subgroup. Because the status epilepticus was caused by TSC and the pneumothorax was based on TSC lymphangioleiomyomatosis, these symptoms were determined to be independent of the study drug, and the trial was continued.

Quiz Ref IDWe found no serious adverse event related to the study drug. Dry skin (ie, dryness without sign of inflammation) was observed in 1 of 12 patients (8%) in the placebo subgroup, 4 of 8 patients (50%) in the 0.05% sirolimus subgroup, 3 of 8 patients (38%) in the 0.1% sirolimus subgroup, and 5 of 8 patients (62%) in the 0.2% sirolimus subgroup (eTable 4A in Supplement 2). Irritated skin (ie, skin with signs of inflammation) was also observed in 3 of 12 patients (25%) in the placebo subgroup, 2 of 8 patients (25%) each in the 0.05% and 0.1% sirolimus subgroups, and 4 of 8 patients (50%) in the 0.2% sirolimus subgroup (eTable 4B in Supplement 2). Because the degree of dry and irritated skin was mild to moderate and did not require treatment other than a topical use of petroleum jelly or moisturizer, the clinical trial was continued.

Abnormalities were not observed in results of blood biochemical or urine tests. A positive blood sirolimus concentration at the end of treatment was observed in 2 of 4 patients (50%) in the 0.2% sirolimus adult subgroup and 1 of 4 patients (25%) in the 0.1% sirolimus adult subgroup. A positive blood sirolimus concentration was also observed in 4 of 4 patients (100%) in the 0.2% sirolimus child subgroup, 2 of 4 patients (50%) in the 0.1% sirolimus child subgroup, and 1 patient (25%) in the 0.05% child subgroup. Although the ratio of positive blood sirolimus concentration increased as the trial continued (eTable 5 in Supplement 2), the highest sirolimus blood concentration of 0.25 ng/mL was seen at 12 weeks. This value was significantly lower than the values found with oral administration (trough concentration, 5-15 ng/mL).

White macules and erythema were identified as secondary end points in the trial protocol before the start of the clinical trial. However, the numbers of the participants with these lesions (4 with white macules and 9 with erythemas) were too small to be statistically analyzed.

To evaluate the efficacy, safety, and optimal concentration of topical sirolimus gel, we conducted a double-blind, placebo-controlled, parallel-group, dose-escalation, phase 2 randomized clinical trial using 3 concentrations of sirolimus gel (0.05%, 0.1%, and 0.2%). We enrolled 18 patients 18 years or younger (child group) and 18 patients 19 years or older (adult group). All participants completed the trial and underwent analysis without exclusion (Figure 1).

The primary end point, the improvement factor, was statistically significant in all treatment subgroups except the 0.1% and 0.05% sirolimus adult subgroups (Figure 2). The statistically significant difference of sirolimus treatment was retained 4 weeks after discontinuation of treatment (16 weeks) among all patients. A statistically significant difference was obtained in the tumor size reduction among all patients and in the 0.2% sirolimus child subgroup (Figure 3A). The reduction rate of tumor redness was statistically significant among all patients and all child subgroups (Figure 3B). Quiz Ref IDThese results indicated that topical sirolimus gel was effective for facial angiofibromas due to TSC and was more effective for children than for adults. Topical sirolimus gel was more effective for reducing tumor redness than for reducing tumor size. The high treatment efficacy is a likely cause of the high degree of satisfaction in the child group (Table). As described in the Results section, P values of the Spearman correlation coefficient among the improvement factor, general improvement, and degree of patient satisfaction were significant (P < .001). This result indicated correlation of the improvement factor, general improvement, and degree of patient satisfaction and that the evaluation methods were adequate. The results indicate that topical sirolimus gel was an effective treatment for facial angiofibromas for objective effects and subjective improvements. The improvement factor at 16 weeks (4 weeks after discontinuation of treatment) was statistically significant in all patient subgroups receiving active treatment. In previous studies,^20,24 the efficacy for reducing redness and tumor regression were retained for 4 weeks after treatment discontinuation of the gel or tacrolimus base but not the ointment.^20,24 Long-term observation after treatment discontinuation is necessary to confirm these findings.

In this trial, we observed no significant adverse events related to the application of topical sirolimus gel. In the clinical laboratory test, we observed no abnormalities. As shown in eTable 4 in Supplement 2, dry skin and irritated skin occurred in all treatment subgroups and the placebo subgroups. Patients in the treatment subgroups exhibited dry and irritated skin at higher rates (12 patients [50%] and 8 patients [33%], respectively) than placebo (1 patient [8%] and 3 patients [25%], respectively). Although this finding could be due to the vehicle, the observation of the highest rates in the 0.2% concentration group suggests that the drug may be the culprit. These symptoms were mild, and all patients continued the trial and recovered.

The ratio of patients with positive blood sirolimus concentrations increased in the 0.2% and 0.1% sirolimus subgroups as the trial period continued (eTable 5 in Supplement 2). Quiz Ref IDThe ratio of patients with positive blood sirolimus levels was higher in the child group than in the adult group. Skin thinness in children may affect the absorption of sirolimus, resulting in better treatment effects for the child group compared with the adult group. However, blood sirolimus concentrations for the topical treatment (<0.25 ng /ml) were lower than those for the oral sirolimus administration (trough concentration, 5-15 ng/mL). In a previous clinical trial study using a topical sirolimus formulation for angiofibromas^20,24 or hypomelanotic macules^23,26 in TSC, no blood sirolimus level was detected. One reason for this result was the detection limit difference for sirolimus concentrations in previous reports and in the present trial of 0.3 and 0.1 ng/mL, respectively. Another likely reason was the difference of cutaneous permeability for the sirolimus gel. In previous studies,^20,23,24,26 a sirolimus gel modified from an oral drug was used, whereas the topical sirolimus gel in this trial was prepared from sirolimus bulk powder. This difference improved cutaneous permeability for the topical sirolimus gel by approximately 1.2 times.

All the scores in the 0.2% sirolimus child subgroup, including improvement index, reduction of tumor size, lessening of redness, general impression, and level of satisfaction, were significantly higher than those of the placebo child subgroup. In addition, the 0.2% concentration of sirolimus gel increased the improvement factor in the adult subgroups (Figure 2C). Quiz Ref IDThe trial results indicate that the optimal concentration of sirolimus gel was 0.2%. The 0.05% concentration of sirolimus gel was valuable for the child group, resulting in a lower blood sirolimus concentration. In this trial, treatment outcomes for the 0.1% concentration of sirolimus gel were poorer than for the 0.05% concentration. In the 0.1% subgroup, 1 patient with an anomalous result was included. Owing to the small group sizes, this inclusion may account for the poor result in the 0.1% sirolimus subgroup. In this trial, patient satisfaction, especially adult patient satisfaction, was inadequate. Because patient satisfaction is a simple impression, discordance between objective response and satisfaction may depend on the patient’s high expectation. Higher expectation in the adult group may result in lower satisfaction. The results of this clinical trial indicated that topical sirolimus gel was a safe and effective treatment for facial angiofibromas caused by TSC and that a 0.2% concentration was optimal.

Topical sirolimus gel is safe and effective for facial angiofibromas caused by TSC, especially for children. However, the overall patient numbers were small (n = 36), the trial period of this study was only 12 weeks, and the initial conditions of tumor volume, redness, and severity varied between the patient subgroups. The study results indicate that a precise, long-term randomized clinical trial with a larger patient study group is necessary to confirm these findings.

Corresponding Author: Mari Wataya-Kaneda, MD, PhD, Department of Dermatology, Graduate School of Medicine, Osaka University, 2-2 Yamada-Oka, Suita, Osaka 565-0871, Japan (mkaneda@derma.med.osaka-u.ac.jp).

Accepted for Publication: August 3, 2016.

Published Online: November 12, 2016. doi:10.1001/jamadermatol.2016.3545

Author Contributions: Dr Wataya-Kaneda had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Wataya-Kaneda, Nakamura, Tanaka, Katayama.

Acquisition, analysis, or interpretation of data: Wataya-Kaneda, Nakamura, Hayashi, Matsumoto, Yamamoto, Katayama.

Drafting of the manuscript: Wataya-Kaneda, Nakamura, Matsumoto, Yamamoto.

Critical revision of the manuscript for important intellectual content: Wataya-Kaneda, Tanaka, Hayashi, Katayama.

Statistical analysis: Wataya-Kaneda, Yamamoto.

Obtained funding: Wataya-Kaneda.

Administrative, technical, or material support: Wataya-Kaneda, Nakamura, Hayashi, Matsumoto, Katayama.

Study supervision: Wataya-Kaneda, Katayama.

Conflict of Interest Disclosures: None reported.

Funding/Support: This study was supported by an H24-intractable-general-008 grant from the Ministry of Health, Labor and Welfare of Japan, grant 25461690 from the Ministry of Education, Culture, Sports, Science and Technology of Japan, and practical research project 15ek0109082h0001 for rare and intractable diseases from the Japan Agency for Medical Research and Development.

Previous Presentation: This paper was presented at the 2016 annual meeting of the Japanese Society of Tuberous Sclerosis Complex; November 12, 2016; Osaka, Japan.

Additional Contributions: We thank the patients for granting permission to publish this information. Thomas H. Darling, MD, Department of Dermatology, Uniformed Services University of the Health Sciences, Bethesda, Maryland, offered helpful suggestions and discussion in the preparation of this manuscript. Aya Tanaka, MD, and Hanako Koguchi-Yoshioka, MD, Department of Dermatology, Osaka University, assisted in the treatment of patients. They received no compensation for this work.