Real‐World Efficacy and Safety of Tofacitinib for Patients With Alopecia Areata: A Retrospective Multi‐Center Cohort Study

ABSTRACT

The approval of baricitinib and ritlecitinib for alopecia areata (AA) marks the beginning of the Janus kinase (JAK) inhibitor era in AA treatment. Tofacitinib, a first‐generation JAK1/3 inhibitor, has theoretical potential to treat AA; however, real‐world evidence, particularly on comparative outcomes across patient subgroups and patient‐reported outcomes, remains limited. In this multi‐center, retrospective cohort study, we evaluated the efficacy and safety of tofacitinib in refractory AA using objective clinical assessments, including SALT scores and four‐point eyebrow and eyelash scales, alongside patient‐reported outcomes, namely DLQI/CDLQI and VAS scores. After 12 weeks of treatment, tofacitinib substantially reduced SALT scores and promoted eyebrow and eyelash regrowth, accompanied by meaningful improvements in patient‐reported outcomes. Efficacy was greater in patients with severe AA than in those with non‐severe AA and was consistent across pediatric and adult populations. Some patients experienced transient disease worsening, which was manageable without treatment escalation. Despite the study's small sample size and retrospective design, tofacitinib demonstrated significant efficacy in both objective and patient‐reported measures, supporting its use as a viable therapeutic option for severe AA and underscoring the value of incorporating patient‐reported outcomes in real‐world studies.

Study Highlights

• What is the current knowledge on the topic?

  • ○
    FDA‐approved JAK inhibitors have opened a new era in alopecia areata (AA) treatment. Tofacitinib, a first‐generation JAK1/3 inhibitor, has shown theoretical potential but limited real‐world data.
• What question did this study address?

  • ○
    This study evaluated the real‐world efficacy and safety of tofacitinib in patients with refractory AA, incorporating both objective clinical assessments and patient‐reported outcomes, and investigated how treatment responses vary across disease severity and age subgroups.
• What does this study add to our knowledge?

  • ○
    This study found that, beyond reducing SALT scores and promoting eyebrow/eyelash regrowth with no severe AEs observed, tofacitinib also improved quality of life and psychological well‐being. Treatment responses were more pronounced in patients with severe AA, with efficacy evident across pediatric and adult populations.
• How might this change clinical pharmacology or translational science?

  • ○
    Tofacitinib may serve as a safe, effective option for refractory or severe AA in both adults and pediatric patients, warranting broader clinical use and further study.

1. Introduction

Alopecia areata (AA) is a common non‐scarring hair disorder. As is reported, nearly 2% of the population has been disturbed by AA, and the lifetime risk of it is up to 1.7% [1]. Clinically, the traditional therapies of AA include corticosteroids, minoxidil, and immunosuppressants. However, the treatment cycle is always long, and the efficacy is sometimes unsatisfactory, and these therapies are associated with some side effects, which impose substantial personal and societal healthcare burdens.

Abundant evidence has shown that the Janus kinase (JAK)‐signal transducer and activator of transcription pathway plays a crucial role in the pathogenesis of AA, and blocking the activity of the signaling pathway has achieved gratifying outcomes in numerous cases [2, 3]. In 2023, the FDA officially approved baricitinib and ritlecitinib for AA's clinical applications, which means the treatment of AA has stepped into the era of JAK inhibitors [4, 5, 6].

Tofacitinib, the first‐generation JAK inhibitor, has been applied to the treatment of rheumatoid arthritis with a potent curative effect for years [7, 8]. It specifically targets JAK 1 and JAK 3 and can inhibit cytotoxic CD8⁺ NKG2D⁺ T cell‐related inflammation, which theoretically hints at its potential to retrieve the destruction of hair follicles by autoreactive CD8⁺ T cells in AA [3, 9, 10]. To date, relevant real‐world evidence on the clinical response and safety of tofacitinib in the AA population remains limited. Existing research predominantly focuses on holistic AA populations or single subgroups with limited comparisons across distinct patient categories. Furthermore, empirical data on patient‐reported outcomes are still absent.

To address the evidence gaps, this study evaluated tofacitinib's therapeutic outcomes in AA through objective clinical assessments and subjective measures, while conducting comparative analyses of treatment responses and safety profiles across diverse patient subgroups.

2. Methods

2.1. Setting

Having been approved by the Institutional Review Board at Medical Ethics Committee of Dermatology Hospital, Southern Medical University (number IIT‐2024‐122, dated January 3, 2025.), a 3‐year observation from January 2020 to September 2023 was conducted at Union Hospital, Tongji Medical College, Huazhong University of Science and Technology and Dermatology Hospital of Southern Medical University. This research is a multi‐center, retrospective study based on real‐world data.

2.2. Patients

Sixty‐one AA patients were enrolled in this work. All patients were diagnosed as AA based on the evidence of clinical manifestations and dermoscopic features. Moreover, they also meet at least one of the following inclusive criteria: (1) were classified as one of the refractory subtypes (alopecia totalis, alopecia universalis [AU], ophiasis, and sisaipho); (2) responded poorly to the traditional medications; (3) had some unbearable side effects after the traditional medications. Exclusion criteria included other causes of alopecia, previous use of any JAK inhibitors, or suffering from contraindications of JAK inhibitors.

2.3. Study Design

A baseline assessment using the severity of alopecia tool (SALT) was conducted before therapy. The SALT was used to quantify scalp hair loss, with the scalp divided into four weighted regions (vertex 40%, right side 18%, left side 18%, posterior 24%): SALT = (Hair loss % in vertex × 0.4) + (Hair loss % in right side × 0.18) + (Hair loss % in left side × 0.18) + (Hair loss % in posterior × 0.24) [11]. Scores range from 0 to 100, with higher values indicating more extensive hair loss. SALT scoring was performed by three trained dermatologists and averaged for analysis. Afterward, patients undertook the treatment with Tofacitinib tablets (initial dose: adults, 5 mg, bid; juveniles, 5 mg, qd) accompanied by conventional therapies or not. The observation spanned at least one therapeutic course, each lasting for 12 weeks. In every return visit, SALT scoring was proceeded to reassess the condition. For AU patients, an extra clinical assessment of eyebrows and eyelashes hair loss (eyelash assessment or eyebrow assessment by the investigator with four‐point scales from 0 [none] to 3 [normal]) was performed [12].

Additionally, since AA is a chronic disfiguring condition exerting a considerable disease burden in both physical and mental aspects and undermining the patient's life quality, the patients were required to recollect and compare their relevant physical and mental conditions before and after the entire treatment, then complete the visual analog scale (VAS) and dermatologic life quality index (DLQI) or Children's dermatology life quality index (CDLQI) to evaluate Tofacitinib's impacts on potential emotional distress and impairment in quality of life bred by AA [13, 14, 15]. The assessment tools were defined as follows: DLQI/CDLQI: Used for patients to evaluate quality of life impact, with scores interpreted as follows: 0–1 (no impact), 2–5 (minor impact), 6–10 (moderate impact), 11–20 (severe impact), and 21–30 (extreme impact). VAS: A 100‐point scale (0 = maximum distress, 100 = no distress) to assess the intensity of patients' subjective distress caused by AA.

Other extracted data were as follows: demographic variables, family history, comorbidities, prior and concomitant medication for AA, and adverse events (AEs) during the treatment.

2.4. Statistical Analysis

SPSS (SPSS Inc., Chicago, IL, USA) and GraphPad Prism (GraphPad Software Inc., San Diego, CA, USA) were used for further analysis of the data. Means with standard deviation (SD) and medians with interquartile ranges (IQRs) were calculated for the quantitative data. Data were evaluated using descriptive and analytic statistics, including paired T‐test, ordinary one‐way ANOVA analysis, and two‐way ANOVA analysis, all of which were two‐tailed. p < 0.05 was considered statistically significant.

2.5. Definitions

The patients were compared by two different dimensions including age and severity: pediatric patients were defined as those < 18 years old, and adult patients as those ≥ 18 years old; severe patients were defined as those with ≥ 50% scalp hair‐loss measured by SALT at baseline, and correspondingly, non‐severe patients denoted the patients with < 50% scalp alopecia measured by SALT at baseline.

Regarding treatment response criteria: SALT score decreasing > 50% from the baseline was identified as having a significant response to the treatment [16]. A SALT score reducing to 20 or less signified a meaningful treatment outcome for AA [17].

3. Results

3.1. Baseline Demographic Characteristics

Sixty‐one patients with AA were included, including 25 males and 36 females. Ages ranged from 4 to 57 years, with disease duration ranging from 1 to 240 months. The mean age at disease onset was 21 years. Pediatric patients accounted for 36% of the cohort (n = 22; mean age, 13 years), while 39 patients (64%) were adults. A family history of AA was reported in only one patient. As for the comorbidity, sleep disorder was referred by 8 patients, and 10 patients had allergic rhinitis. Additional comorbidities include atopic dermatitis, eczema, hyperthyroidism, etc.

The detailed baseline demographic characteristics are listed in Table 1.

TABLE 1.

Demographic characteristics of the patients at baseline.

Characteristic	Value
Sex
Male, n (%)	25 (41.0%)
Female, n (%)	36 (59.0%)
Age of starting tofacitinib treatment, mean (SD)	26.6 (14.1)
< 12, n (%)	6 (9.8%)
12–17, n (%)	16 (26.2%)
≥ 18, n (%)	39 (63.9%)
Duration of disease, month, median (IQR)	24 (12–67.5)
Clinical subtypes
AU, n (%)	23 (37.7%)
Others, n (%)	38 (62.3%)
Received any treatments for AA before
Yes, n (%)	51 (83.6%)
No, n (%)	10 (16.4%)
Other current treatments for AA
Compound betamethasone injection, n (%)	36 (59.0%)
Minoxidil tincture, n (%)	51 (83.6%)
Halometasone cream, n (%)	28 (45.9%)
Desonide cream, n (%)	3 (4.9%)
Oral immunomodulators (Thymopeptide or Spleen aminopeptide or Pidotimod), n (%)	32 (52.5%)
Nutritional supplements, n (%)	10 (16.4%)
Chinese patent drugs, n (%)	18 (29.5%)
Microneedle therapy with red light	5 (8.2%)
Comorbidities
Sleep disorders, n (%)	12 (19.7%)
Allergic rhinitis, n (%)	10 (16.4%)
Atopic dermatitis, n (%)	5 (8.2%)
Neurosis, n (%)	4 (6.6%)
Eczema, n (%)	3 (4.9%)
Urticaria, n (%)	2 (3.3%)
Asthma, n (%)	1 (1.6%)
Irregular menstruation, n (%)	1 (1.6%)

Abbreviations: AA, alopecia areata; AU, alopecia universalis.

3.2. Efficacy of Tofacitinib Treatment for AA Measured by SALT

All patients completed a minimum of 12‐week treatment, with follow‐up available for 37, 25, 21, and 8 patients at Weeks 24, 36, 48, and 72, respectively (Figure 1A–G and Table S1).

FIGURE 1.

Patient hair response after using tofacitinib. (A) Response evaluated by mean SALT scores. (B) Response evaluated by median SALT scores. (C) Response evaluated by mean SALT score improvement. (D) Response evaluated by median SALT score improvement. (E) Response based on SALT score 20 or less. (F) Response based on SALT score improvement increasing ≥ 50%. (G) Representative photos recording hair regrowth of a single responder. Vertical bars represent the standard error of the mean (SEM). Owing to the small populations which were less than 3 in some groups at the later stage, we only exhibited the changes from Week 0 to Week 36.

3.2.1. Overall Treatment Response

The mean baseline SALT score for the overall cohort was 70.9. After the first treatment course, the mean SALT score declined to 41.3, corresponding to an average improvement rate of 41.6% from baseline (calculated for each patient as [(Baseline SALT score—Follow‐up SALT score)/Baseline SALT score] × 100%, with the mean taken across all patients). At Week 24, the mean SALT score further decreased to 28.1, and most subsequent follow‐ups showed only mild fluctuations, indicating a plateau of treatment efficacy. During the entire observation period, 40 patients (67.8%) achieved > 50% improvement in SALT scores, 32 patients (54.2%) reached a clinically meaningful response, and 12 patients (20.3%) attained a SALT score of ≤ 10. Full descriptive statistics are provided in Tables S2–S5.

3.2.2. Severe Versus Non‐Severe AA

Among severe patients (n = 46), the mean baseline SALT score was 82.5, decreasing to 47.4 after the first course (improvement 42.9%) and further to 29.7 at Week 24. For non‐severe patients (n = 15), the mean SALT score decreased from 35.4 to 22.7 after the first course (improvement 37.5%) and to 17.9 at Week 24. At Week 24, 21 patients achieved > 50% improvement (17 severe, 4 non‐severe), and 19 patients reached a SALT score ≤ 20 (16 severe, 3 non‐severe), including 9 patients with SALT ≤ 10. Full descriptive statistics are provided in Tables S2–S5.

3.2.3. Adult Versus Pediatric

In pediatric patients (n = 22), the mean baseline SALT score was 77.7, declining to 46.6 after the first course (improvement 41.2%) and 25.7 at Week 24. Twelve pediatric patients achieved a significant response, and six attained a SALT score ≤ 20. Among adult patients (n = 39), the mean baseline SALT score was 67.2, decreasing to 38.3 after the first 12 weeks (improvement 41.8%) and 29.6 at Week 24; 65.2% achieved a significant response, and 20.5% reached a SALT score ≤ 20. Full descriptive statistics are provided in Tables S2–S5.

3.2.4. Recrudescence During Follow‐Up

During the study, 18 patients experienced transient disease worsening: eight during dose maintenance, four during dose reduction, and 6 after treatment discontinuation. Except for seven new‐onset cases relapsing within 1 month and four cases lost to follow‐up after recurrence, all patients had an alleviation again measured by SALT after prolonging the medication of Tofacitinib at the same dose.

3.3. Efficacy of Tofacitinib Treatment in Eyebrow and Eyelash for AA

In 25 patients with AU, the mean baseline eyebrow assessment score was 0.7 (median, 1.0), and the mean eyelash score was 1.2 (median, 1.0) (Figure 2a–c). Following tofacitinib treatment, regrowth of eyebrows and eyelashes was observed in all patients over time. At Week 12, mean scores increased to 1.6 (median, 2.0) for eyebrows and 2.0 (median, 2.0) for eyelashes, rising further to 1.9 (median, 2.0) and 2.2 (median, 2.0) at Week 24. Seven patients achieved complete eyebrow regrowth, while 13 patients attained full scores for eyelash regrowth. Detailed changes are provided in Table S6.

FIGURE 2.

Patient eyebrows and eyelashes response after using tofacitinib. (a) Response evaluated by mean eyebrow/eyelash assessment scores. (b) Response evaluated by median eyebrow/eyelash assessment scores. (c) Representative photos recording eyebrows regrowth of a single responder (eyelashes regrowth was shown in Figure S1 for privacy consideration). Vertical bars represent the standard error of the mean (SEM). We did not include the data of Week 96 in Figure 2a,b, because only one AU patient re‐visited at this time point.

3.4. Efficacy of Tofacitinib Treatment for AA Measured by DLQI/CDLQI and VAS

Among 39 collected questionnaires, 37 were deemed valid. The overall mean DLQI/CDLQI score decreased from 14.7 (median, 16.0) at baseline to 9.6 (median, 7.0) after treatment. In severe patients, the score declined from 14.8 (median, 16.0) to 9.9 (median, 7.0), while in non‐severe patients, it decreased from 14.2 (median, 15.5) to 9.6 (median, 7.0) (Figure 3a–d and Table S7).

FIGURE 3.

Patient‐oriented response after using tofacitinib. (a) Response evaluated by mean DLQI/CDLQI scores. (b) Response evaluated by median DLQI/CDLQI scores. (c) Response evaluated by mean VAS score improvement. (d) Response evaluated by median VAS score improvement. Definition of significant difference: P < 0.05. ****, symbolized p < 0.0001, in comparison to baseline by paired‐samples T test.

Similarly, VAS scores improved markedly after treatment. The mean score increased from 17.3 (median, 0.0) to 51.3 (median, 50.0) in the overall cohort, from 12.8 (median, 0.0) to 49.2 (median, 42.0) in severe patients, and from 40.5 (median, 40.5) to 62.2 (median, 70.5) in non‐severe patients.

3.5. Safety of Tofacitinib Treatment for AA

During follow‐up, AEs were reported in 17 patients (27.9%), including four pediatric patients. No severe AEs occurred, and none of the participants discontinued treatment (Table 2).

TABLE 2.

Adverse events in overall study period.

Patients with AEs	Amount
Total, n (%) *	17 (27.9%)
Pediatric, n (%)	4 (6.6%)
Adult, n (%)	13 (21.3%)
Severe AEs, n (%)	0 (0.0%)
Infections, n (%)	5 (8.2%)
Influenza, n (%)	2 (3.3%)
Upper respiratory tract infection (Pharyngitis, consillitis, etc.), n (%)	2 (3.3%)
Conjunctivitis, n (%)	2 (3.3%)
Acne, n (%)	4 (6.6%)
Weight gain, n (%)	4 (6.6%)
Stomach upset, n (%)	1 (1.6%)
Palpitation, n (%)	1 (1.6%)
Chest distress, n (%)	1 (1.6%)
Hematological parameters changing, n (%)	4 (6.6%)
Dyslipidemia, n (%)	4 (6.6%)
Coagulation abnormality, n (%)	3 (4.9%)

Abbreviation: AEs, adverse events.

^* This proportion refers to the number of patients who experienced AEs relative to the total number of patients.

The most common AEs were infections (five patients, 8.2%; influenza, upper respiratory tract infection, conjunctivitis), weight gain (> 5 kg, 4 patients, 6.6%), acne (4 patients, 6.6%), and gastrointestinal discomfort (1 patient, 1.6%). Notably, one patient experienced palpitations and another experienced chest discomfort; both events resolved completely without additional interventions.

Laboratory abnormalities were observed in some patients. Dyslipidemia occurred in four patients (6.6%), including changes in serum apolipoprotein A and elevations in total cholesterol, low‐density lipoprotein cholesterol (LDL‐C), apolipoprotein B, and apolipoprotein E. Coagulation parameters, including prothrombin time, international normalized ratio (INR), thrombin time, or serum fibrinogen levels, fluctuated in four patients (6.6%).

4. Discussion

This real‐world, multi‐center study evaluates the efficacy and tolerability of tofacitinib in Chinese patients with AA, uniquely integrating objective clinical assessments, patient‐reported outcomes, and subgroup analyses by disease severity and age.

As early as the end of the first medication course (Week 12), tofacitinib demonstrated significant efficacy compared to baseline, with both mean and median SALT scores showing notable reductions and a mean improvement rate of 41.6%. Visible hair regrowth was observed in several patients within the first month, indicating a rapid onset of effect. The mean SALT score continued to decline by Week 24 and subsequently reached a plateau with only minor fluctuations in later follow‐ups, with the mean improvement rate increasing to 64.3% before stabilization. In contrast, the median SALT score showed a sustained decrease through Week 96. Over the entire treatment period, 40 patients (67.8%) achieved more than 50% improvement in SALT scores, 32 patients (54.2%) reached clinically meaningful treatment success, and 12 patients (20.3%) attained a stringent response with a SALT score of 10 or less. Briefly, the above data not only implied the rapid curative effect of Tofacitinib in most AA patients but also manifested that it was of notability and generality. Representative hair regrowth in one patient is shown in Figure 1G.

In the subgroup analysis, a clear time‐dependent effect was observed in the severe AA group, consistent with the overall cohort, supporting the potential of tofacitinib for treating severe AA. Unlike most previous studies that included only severe patients [8, 16, 18, 19, 20], we also evaluated 15 non‐severe patients who had experienced poor responses to long‐term conventional therapy and were willing to try tofacitinib. Although SALT scores decreased and improvement rates increased in this group, the changes did not reach statistical significance, suggesting that tofacitinib may offer limited benefits in non‐severe AA and highlighting the need for severity‐targeted treatment strategies. Comparisons between adults and pediatric patients showed similar efficacy and tolerability, indicating that tofacitinib can be safely and effectively applied across age groups in clinical practice.

In patients with AU in our cohort, tofacitinib treatment also promoted substantial regrowth of eyebrows and eyelashes (Figure 2a,b), with most patients showing benefits at an early stage. Figure 2c illustrates the eyebrow regrowth process in one patient over the course of treatment (for privacy reasons, eyelash changes are presented in Figure S1), demonstrating that eyebrow and eyelash regrowth occurred earlier than scalp hair.

Patient‐reported outcomes, including DLQI/CDLQI and VAS, were used to evaluate subjective improvement. As shown, tofacitinib effectively enhanced quality of life and alleviated patients' subjective distress, reflecting positive effects on scalp discomfort, emotional well‐being, daily activities, and overall psychosocial burden.

In our cohort, transient exacerbation of alopecia during treatment was observed in 18 patients. Available follow‐up data indicated that most episodes could be controlled with unchanged tofacitinib dosage. Similar transient worsening during treatment has also been reported in patients receiving other JAK inhibitors, including baricitinib [21], and ivarmacitinib [22]. These exacerbations are unlikely to represent sustained disease progression or treatment failure, as disease control could be re‐established without treatment escalation. This phenomenon may reflect interindividual variability in response to JAK inhibitor therapy, as well as potential lifestyle‐ or adherence‐related factors during treatment, such as increased stress, sleep disturbance, or irregular medication use. Maintaining an adequate period of dose stabilization and implementing gradual tapering before treatment withdrawal may help reduce the risk of disease recurrence [23]. Future prospective studies are warranted to better characterize the mechanisms underlying these transient fluctuations in disease activity during JAK inhibitor therapy.

The safety profile of tofacitinib observed in our study aligns with global evidence, addressing key considerations raised regarding its clinical applicability. The most common AEs and overall AE rate (27.9%) are consistent with published real‐world data and meta‐analyses focusing on AA (overall AE rate: 15%–43.2%). Notably, no severe AEs occurred, which is in line with the established safety profile of tofacitinib in AA [24, 25, 26, 27, 28]. The transient palpitations and chest discomfort in 2 patients resolved spontaneously without intervention, further supporting its tolerability.

Ethnically, real‐world studies show consistent AE profiles of tofacitinib in Chinese and non‐Chinese AA patients (Chinese: infections 5%–15.9%, acne 12.7%–20%, dyslipidemia 7.9%–10.0% [29, 30]; non‐Chinese: infections 9.2%–21.3%, acne 3.1%–5.9%, dyslipidemia 12.3%–33.2%) [25, 26], with no ethnic‐specific safety signals identified. Although acne is more prevalent and dyslipidemia is less common in Chinese patients, these findings are based on small‐to‐moderate sample sizes (20–202 patients) without matching for acne and dyslipidemia‐related confounders (e.g., age), requiring validation in larger‐scale studies.

Coagulation abnormality and dyslipidemia warrant longer‐term observation. Nevertheless, with a maximum follow‐up exceeding 2 years in this study, tofacitinib demonstrated general safety and good tolerability across all AA subgroups, including pediatric patients. The study did not systematically assess laboratory parameters after discontinuation, precluding conclusions on their reversibility. In our experience, timely pharmacologic intervention—when indicated—is a prudent strategy to ensure long‐term treatment safety in patients with moderate‐to‐high cardiovascular risk.

5. Limitations

Limitations of our study include its small sample size, retrospective design, and the unavoidable recall bias existing in DLQI/CDLQI and VAS scores.

This study suggests that:

1. Tofacitinib demonstrates significant efficacy and safety in AA patients, improving clinical outcomes, quality of life, and reducing psychological burden across age groups.

2. Compared with patients with severe AA, the benefits of tofacitinib are limited in non‐severe AA, supporting a severity‐stratified treatment approach, in which JAK inhibitors are primarily indicated for severe cases.

3. The duration of dose maintenance and gradual tapering during tofacitinib therapy may influence disease recurrence, while the plateau phase of SALT scores may guide individualized dose adjustment and discontinuation.

Overall, tofacitinib represents a promising therapeutic option among JAK inhibitors for severe AA, and further large‐scale studies are warranted to evaluate its long‐term efficacy and safety.

Author Contributions

Y.L., Y.J., J.C, and S.Y. wrote the manuscript. Y.L. and C.H. designed the research. J.C., S.Y., and Y.Z. performed the research; R.T. and Y.H.L. analyzed the data.

Funding

This work was supported partially by a grant from the General Program of the National Natural Science Foundation of China (No. 81972565).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Representative photos recording eyebrows and eyelashes regrowth of a single responder.

Table S1: The total number of patients in different subgroups at different time points.

Table S2: Mean and median SALT scores in different subgroups at different time points.

Table S3: Mean and median SALT score improvement rate^# compared to the baseline in different subgroups at different time points.

Table S4: The total number of the patients with a SALT score ≤ 20 in different subgroups at different time points.

Table S5: The total number of the patients with SALT score improvement rate > 50% compared to the baseline in different subgroups at different time points.

Table S6: Mean and median Eyebrow/eyelash assessment scores of AU patients at different time points.

Table S7: Mean and median DLQI/CDLQI and VAS scores in different subgroups at different time points.