Efficacy and Safety of Oral Janus Kinase Inhibitors in Adults and Adolescents with Alopecia Areata: A Systematic Review and Meta-Analysis of Randomised Controlled Trials

Abstract

Background

Alopecia areata is an autoimmune disorder characterised by T-cell-mediated damage of hair follicles, resulting in non-scarring hair loss that can progress to complete scalp (alopecia totalis) or body hair loss (alopecia universalis). While long-term systemic options were historically limited, the emergence of Janus kinase (JAK) inhibitors has revolutionised the treatment of alopecia areata, even in patients with moderate-to-severe disease.

Objective

We aimed to evaluate the efficacy and safety of oral JAK inhibitors compared with placebo in adults and adolescents with moderate-to-severe alopecia areata.

Methods

MEDLINE, Embase and the Cochrane Central Register of Controlled Trials databases were searched from inception to 28 November, 2025. Randomised controlled trials evaluating oral JAK inhibitors in adolescents and adults (aged ≥12 years) with alopecia areata were included. Data extraction and risk of bias assessment were performed independently by multiple reviewers. Primary outcomes were the proportion of patients who achieved Severity of Alopecia Tool (SALT) scores of ≤10 and ≤20, and 50%, 75% and 90% reductions in SALT scores from baseline, as well as treatment-related adverse events (AEs). Data were synthesised using random-effects models to calculate odds ratios (ORs) and mean differences with 95% confidence intervals (CIs).

Results

Twelve randomised controlled trials involving 4141 participants (mean age 35.8 years, mean baseline SALT score 86.0) met the inclusion criteria. All trials enrolled adult participants, with the exception of one randomised controlled trial in which adolescents comprised 15% of the study population. Six trials were rated as having a low risk of bias, while six had “some concerns” primarily because of missing outcome data. Oral JAK inhibitors were significantly more effective than placebo across all primary efficacy endpoints. At week 24, the odds of achieving SALT ≤10 were 5.69 times higher for JAK inhibitors than placebo (95% CI 3.58–9.03); by week 36, this effect increased to an OR of 9.40 (95% CI 4.58–19.29). For SALT ≤20, the OR was 7.99 (95% CI 5.13–12.44) at week 24. Patients treated with oral JAK inhibitors were significantly more likely to achieve SALT50, SALT75 and SALT90 across all timepoints (weeks 12–36). Regarding safety, JAK inhibitors were associated with a higher risk of total AEs (OR 1.60; 95% CI 1.33–1.94), but no statistically significant difference was found for serious AEs (OR 1.04; 95% CI 0.69–1.57) or treatment discontinuation because of AEs (OR 1.22; 95% CI 0.84–1.76).

Conclusions

Oral JAK inhibitors are effective in promoting scalp, eyebrow and eyelash hair regrowth in adults and adolescents with moderate-to-severe alopecia areata, with a trend of higher response rates at later timepoints across studies. The favourable safety profile of these agents is reassuring. Future research should prioritise head-to-head randomised controlled trials and long-term pharmacovigilance to define comparative efficacy and safety.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s40257-026-01060-z.

Key Points

Oral Janus kinase inhibitors are significantly more effective than placebo in promoting scalp, eyebrow and eyelash hair regrowth in adults and adolescents with moderate-to-severe alopecia areata, with a trend of higher response rates at later timepoints (up to 36 weeks) across trials.

Despite being associated with a higher risk of total adverse events, these agents demonstrate a reassuring safety profile with no significant differences in rates of serious adverse events or treatment discontinuation compared with placebo.

These findings strengthen the evidence base for oral Janus kinase inhibitors as an effective treatment option, although head-to-head trials are needed to further define comparative efficacy and safety.

Introduction

Alopecia areata (AA) is a common autoimmune disorder characterised by a T-cell-mediated attack on the hair follicle, leading to patchy non-scarring hair loss that can affect the scalp, face and other parts of the body. Bald patches can be solitary or multiple, and can even lead to complete scalp hair loss (alopecia totalis, AT), or complete scalp, facial and body hair loss (alopecia universalis, AU) [1]. Alopecia areata can affect all age groups, with 31–48% of all patients presenting before the age of 20 years [2]. For a long time, there was a lack of effective therapeutic options for moderate-to-severe disease. It is therefore hardly surprising that AA has a profoundly negative impact on health-related quality of life and social functioning for both patients and their family members, and is associated with high rates of depression and anxiety [3, 4].

Over the past decade, a mechanistic revolution has transformed therapeutic prospects through Janus kinase (JAK) inhibition. In AA, collapse of immune privilege of the hair follicle bulb leads to activation of CD8+ T cells, which produce interferon-gamma. This cytokine signals through the JAK-signal transducer and activator of transcription pathway in follicular epithelial cells to promote interleukin-15 production, creating a positive feedback loop that amplifies the autoimmune attack on the hair follicle [5, 6]. Janus kinase inhibitors target this signalling cascade, effectively disrupting the inflammatory loop and restoring hair regrowth in patients with moderate-to-severe AA.

Initial clinical translation came from open-label studies of tofacitinib (a JAK1/JAK3 inhibitor) and ruxolitinib (a JAK1/JAK2 inhibitor), which showed dramatic hair regrowth in patients with >50% scalp hair loss but lacked controls [5, 6]. Subsequent randomised controlled trials (RCTs) confirmed efficacy and safety in larger populations. Baricitinib (a JAK1/JAK2 inhibitor) was the first systemic therapy to be approved for the treatment of severe AA by the US Food and Drug Administration (FDA) and European Medicines Agency in 2022, and the Chinese National Drug Administration and Australian Therapeutic Goods Administration in 2023 [7, 8]. This followed the pivotal BRAVE-AA1 and BRAVE-AA2 trials, which demonstrated significantly higher proportions of patients achieving a Severity of Alopecia Tool (SALT) score ≤20 compared with placebo at 36 weeks [9]. An integrated safety analysis has since confirmed durable benefit and tolerability following long-term exposure [10].

A key insight emerging from these extension studies is that the longer patients remain on JAK inhibitor therapy, the greater the cumulative hair regrowth achieved. Data from 52- and 104-week follow-ups of baricitinib, and from longer-term studies of other JAK inhibitors, demonstrate continued improvement beyond initial response plateaus [11–14]. This time-dependent effect highlights why combining widely different endpoints, such as 24-week and 52-week data, in a single pooled meta-analysis may obscure true efficacy trajectories. Distinct time windows therefore warrant separate analyses to capture the progressive nature of hair regrowth.

Beyond baricitinib, several next-generation oral JAK inhibitors have completed randomised trials in AA: ritlecitinib (a JAK3/tyrosine kinase expressed in hepatocellular carcinoma inhibitor), brepocitinib (a tyrosine kinase 2/JAK1 inhibitor), deuruxolitinib (a JAK1/JAK2 inhibitor) and ivarmacitinib (a selective JAK1 inhibitor). Each has demonstrated significant scalp hair regrowth and acceptable tolerability [15–20]. Ritlecitinib received FDA approval for adolescents and adults with severe AA in June 2023, and was subsequently approved by the Chinese National Drug Administration in October 2023, the UK National Health Service in February 2024 and Therapeutic Goods Administration in June 2024. Furthermore, deuruxolitinib received FDA approval for adults with severe AA in July 2024 [7, 8]. Other JAK inhibitors such as upadacitinib (a selective JAK1 inhibitor) remain under active investigation [21, 22].

Given this rapidly expanding landscape, a rigorous synthesis of RCT evidence is now essential to benchmark efficacy and safety across agents, clarify optimal timepoints for assessment and define durability of response. The present systematic review and meta-analysis therefore aimed to (1) identify all RCTs evaluating oral JAK inhibitors in adults and adolescents with moderate-to-severe AA, (2) compare efficacy and safety outcomes stratified by pre-defined time windows, and (3) explore heterogeneity in treatment response and long-term durability.

Methods

This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines and methodological recommendations of the Cochrane Handbook for Systematic Reviews of Interventions [23]. The review protocol was prospectively registered on PROSPERO (Registration No. CRD420251181738).

Inclusion and Exclusion Criteria

Types of Studies

We included RCTs of any design (parallel, factorial or cross-over) that evaluated the efficacy and safety of JAK inhibitors in AA. Case reports, case series, non-randomised studies and secondary or subgroup analyses from parent RCTs were excluded. Specifically, subgroup analyses limited to specific populations (e.g. paediatric cohorts, AT/AU subsets) were excluded to avoid data duplication and maintain trial-level integrity.

Types of Participants

Eligible studies enrolled adolescent and adult participants (aged ≥12 years) with a clinical diagnosis of AA as defined by study-specific criteria. There were no restrictions on disease duration or prior treatments. Studies exclusively involving other forms of alopecia (e.g. androgenetic alopecia, traction alopecia or scarring alopecia) were excluded.

Types of Interventions and Comparators

We included trials evaluating any oral JAK inhibitor in the treatment of AA. Trials investigating the efficacy of topical JAK inhibitors were excluded. All but one of the studies compared oral JAK inhibitors with placebo.

Types of Outcomes

The primary outcomes of interest were changes in hair loss severity scores according to the SALT, including (1) the proportion of patients achieving SALT ≤10 and SALT ≤20, (2) the proportion of patients achieving 50%, 75% and 90% improvement in SALT score from baseline (SALT50, SALT75 and SALT90), and (3) treatment-related adverse events (AEs). Secondary outcomes included (1) percentage change in SALT score from baseline, (2) the proportion of patients demonstrating improvement in eyebrow and eyelash hair loss severity scores, and (3) the proportion of patients demonstrating improvement in quality of life or satisfaction with treatment.

Timing of Outcome Assessment

Efficacy and safety outcomes were extracted at the end of each study’s primary treatment phase. Long-term or extension data were not included unless explicitly part of the randomised treatment period.

Literature Search

A comprehensive electronic search was conducted in MEDLINE (Ovid), Embase and the Cochrane Central Register of Controlled Trials (CENTRAL) from database inception to 28 November, 2025. Search terms combined controlled vocabulary and keywords relating to JAK inhibitors and AA. These included Janus kinase inhibitors, JAK inhibitors, baricitinib, deuruxolitinib, ritlecitinib, ivarmacitinib, tofacitinib, brepocitinib, ruxolitinib, abrocitinib, upadacitinib, filgotinib, deucravacitinib, alopecia areata, alopecia totalis, alopecia universalis, alopecia circumscripta and randomised controlled trials. Details of the full search strategy is provided in Appendix 1 of the Electronic Supplementary Material (ESM).

Additional records were identified through reference list screening of relevant reviews and included studies, and through searches of ClinicalTrials.gov and the World Health Organisation International Clinical Trials Registry Platform.

Study Selection

Two reviewers (SM and BB) independently screened all titles and abstracts using predefined eligibility criteria. Full texts of potentially relevant studies were retrieved and independently assessed for inclusion. Disagreements were resolved by discussion or by a third reviewer. Duplicate publications were collated and data were extracted once per unique trial.

Data Extraction and Management

Five reviewers (SM, FD, HF, DK and BB) independently extracted data using a standardised extraction form. Extracted information included:

• Study characteristics such as author and year;

• Participant characteristics: sample size, age, sex, disease duration and baseline SALT score;

• Intervention details: JAK inhibitor type and dose(s), administration route and treatment duration;

• Outcome data: efficacy and safety endpoints, and follow-up duration.

Any discrepancies were resolved through discussion or third-party adjudication. For multi-arm trials, only data from relevant arms meeting inclusion criteria were extracted.

Risk of Bias Assessment

The risk of bias for each included study was assessed independently by five reviewers using the Cochrane Risk of Bias 2 tool [24]. Domains assessed included bias arising from the randomisation process, bias due to deviations from intended interventions, bias due to missing outcome data, bias in measurement of the outcome and bias in selection of the reported result. Each domain was rated as high, having some concerns or low. Disagreements were resolved by consensus.

Statistical Analysis

Statistical analyses were performed using R (Version 4.5.2) with the meta and metafor packages. Random-effects models were used throughout to account for between-study heterogeneity. Outcomes were stratified by timepoint and outcome type (efficacy vs safety). Odds ratios (ORs) were used for dichotomous outcomes (e.g. achieving SALT thresholds, AEs), and mean differences were used for continuous outcomes (e.g. change in SALT score), each reported with corresponding 95% confidence intervals (CIs). A continuity correction of 0.5 was applied to trials with zero-event counts in either or both arms.

For efficacy outcomes, we included all trials that reported baseline and post-treatment SALT scores at weeks 12–16, 24 or 36. Endpoints included SALT ≤10, SALT ≤20, SALT50, SALT75 and SALT90. Trials reporting only one arm, or with unmatchable treatment–control comparisons, were excluded from the pooled analysis. Endpoints such as SALT30 and SALT100, reported by a single study, were not included in pooled analyses because of insufficient data.

For safety outcomes, AEs, severe AEs, serious AEs and treatment discontinuation because of AEs were analysed separately. Severe AEs are highly intense or incapacitating, but are not necessarily dangerous or life threatening [25]. However, serious AEs result in death, are life-threatening, require hospitalisation, result in disability or permanent damage, cause a congenital anomaly or birth defect, or require intervention to prevent permanent harm [26]. Double-zero studies (i.e. 0 events in both arms) were retained and included using continuity correction. This approach follows Cochrane guidelines, which discourage exclusion of such studies owing to their importance in preserving the full evidence base and preventing inflated pooled ORs. For severe AEs, comparisons were excluded if reporting was incomplete or unmatched across arms.

Statistical significance was determined by examining whether the 95% CI of the pooled OR excluded 1. Heterogeneity was assessed using I² and τ² statistics, and Cochran’s Q test. When considerable heterogeneity was detected (I² > 50%), findings were interpreted with caution, and subgroup or sensitivity analyses were considered to explore potential sources of variability, such as differences in drug class, dose or study duration. In these cases, pooled results were interpreted with caution. Forest plots were generated from eligible comparisons with estimable treatment effects.

Results

Study Selection and Search Results

A literature search of electronic databases and clinical trial registries initially identified 731 records. After the removal of 286 duplicate entries, 445 unique records underwent title and abstract screening. Following a detailed eligibility assessment of 94 full-text reports, including three identified through alternative search methods, 12 RCTs met the inclusion criteria and were incorporated into the final meta-analysis (Fig. 1).

Fig. 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram. JAK Janus kinase, RCT randomised controlled trial

Trial and Participant Characteristics

Characteristics of the 12 included studies are summarised in Table 1. These trials, published between 2021 and 2025, represented a global cohort of 4141 participants, including 1583 men and 2558 women, with a mean (standard deviation) age of 35.8 (13.1) years. All studies enrolled adult participants, with the exception of one RCT in which adolescents comprised 15% of the study population [16]. At study entry, participants demonstrated significant disease severity, with a mean (standard deviation) baseline SALT score of 86.0 (17.8). The mean (standard deviation) duration of current episode of hair loss was 4.8 (8.5) years. One RCT included an active comparator (azathioprine) [26], with the remaining 11 studies utilising a placebo control.

Table 1.

Characteristics of 12 randomised controlled trials evaluating the efficacy and safety of oral Janus kinase inhibitors in patients with moderate-to-severe alopecia areata

Study	Drug and dose	Randomised in arm (n)	Analysed in arm (n)	Age, mean (SD), years	Age, median (IQR; range), years	Male, n (%)	Female, n (%)	Baseline SALT score, mean (SD)	Total disease duration, mean (SD), years	Total disease duration, median (IQR; range), years	Duration of current episode of hair loss, mean (SD), years	Duration of current episode of hair loss, median (IQR; range), years
King et al. [28]	Baricitinib 1 mg	28	28	38.6 (11.3)		10 (35.7%)	18 (64.3%)	89.3 (17.7)	12.5 (12.5)		4.2 (4.2)
	Baricitinib 2 mg	27	27	42.5 (13.8)		4 (14.8%)	23 (85.2%)	86.1 (19.3)	16.9 (12.8)		4.4 (5.6)
	Baricitinib 4 mg	27	27	42.4 (14.9)		2 (7.4%)	25 (92.6%)	83.4 (17.5)	12.3 (10.1)		4.0 (3.5)
	Placebo	28	28	40.5 (14.2)		12 (42.9%)	16 (57.1%)	90.0 (15.7)	16.8 (13.0)		3.5 (2.7)
King et al. [18]a	Baricitinib 2 mg	184	184	38.0 (12.8)		75 (40.8%)	109 (59.2%)	86.8 (18.0)	12.1 (9.8)		3.9 (4.7)
	Baricitinib 4 mg	281	281	36.3 (13.3)		116 (41.3%)	165 (58.7%)	85.3 (18.2)	11.8 (11.1)		3.5 (3.4)
	Placebo	189	189	37.4 (12.9)		80 (42.3%)	109 (57.7%)	84.7 (17.8)	12.6 (11.2)		3.5 (3.7)
King et al. [18]b	Baricitinib 2 mg	156	156	39.0 (13.0)		53 (34.0%)	103 (66.0%)	85.6 (18.1)	13.1 (11.8)		4.4 (6.1)
	Baricitinib 4 mg	234	234	38.0 (12.7)		90 (38.5%)	144 (61.5%)	84.8 (18.1)	11.9 (11.1)		3.9 (3.4)
	Placebo	156	156	37.1 (12.4)		58 (37.2%)	98 (62.8%)	85.0 (17.8)	11.8 (10.2)		4.7 (5.5)
King et al. [29]	Deuruxolitinib 4 mg BID	30	28 (efficacy)/29 (safety)	35.7 (11.0)		8 (26.7%)	29 (65.9%)	88.8 (16.2)			6.0 (2.78)
	Deuruxolitinib 8 mg BID	38	38	37.3 (14.2)		12 (31.6%)	22 (73.3%)	89.1 (16.4)			3.8 (2.72)
	Deuruxolitinib 12 mg BID	37	36	35.8 (12.4)		9 (24.3%)	26 (68.4%)	87.3 (18.7)			3.5 (2.3)
	Placebo	44	43 (efficacy)/44 (safety)	37.8 (13.5)		15 (34.1%)	28 (75.7%)	86.8 (18.4)			4.1 (3.34)
King et al. [17]	Deuruxolitinib 8 mg BID	351	351		37 (18-65)	134 (38.2%)	217 (61.8%)	85.5 (18.35)			2.9 (1-11)
	Deuruxolitinib 12 mg BID	215	215		36 (18-65)	84 (39.1%)	131 (60.9%)	85.2 (18.41)			2.5 (0-11)
	Placebo	140	140		38.5 (18-65)	51 (36.4%)	89 (63.6%)	88.1 (15.1)			3.3 (1-10)
Tsianakas et al. [31]	Deuruxolitinib 8 mg BID	258	233	38.4 (12.30)		81 (31.4%)	177 (68.6%)	88.1 (17.40)			3.8 (2.76)
	Deuruxolitinib 12 mg BID	129	120	39.7 (12.90)		45 (34.9%)	84 (65.1%)	86.7 (18.18)			4.0 (2.93)
	Placebo	130	119	39.7 (12.49)		42 (32.3%)	88 (67.7%)	88.9 (16.20)			3.8 (3.10)
King et al. [15]	Ritlecitinib 200/50 mgd	48	48	37 (13)		11 (23)	37 (77)	89.4 (15.8)		6.7 (0.6–52.3)		2.6 (0.3–7.5)
	Brepocitinib 60/30 mgc	47	47	34 (11)		15 (32)	32 (68)	86.4 (18.1)		8.4 (0.3–48.5)		1.9 (0.2–7.0)
	Placebo	47	47	38 (14)		18 (38)	29 (62)	88.4 (18.1)		4.8 (0.2–53.4)		2.4 (0.2–29.5)
King et al. [16]	Ritlecitinib 10 mg	63	59	34.3 (13.9)	34.0 (21.0–48.0; 13.0–58.0)	20 (32)	43 (68)	88.3 (16.9)	10.8 (10.7)	7.0 (3.1–17.2; 0.2–50.9)	3.3 (2.7)	2.2 (1.0–4.8; 0.3–9.7)
	Ritlecitinib 30 mg	132	119	33.7 (14.8)	32.5 (21.0–41.0; 12.0–73.0)	52 (39)	80 (61)	90 (15.1)	8.8 (8.9)	5.8 (2.8–10.5; 0.0–53.4)	3.6 (2.8)	2.6 (1.1–5.5; 0.3–10.0)
	Ritlecitinib 50 mg	130	124	32.4 (13.4)	30.0 (22.0–42.0; 13.0–70.0)	59 (45)	71 (55)	90.3 (14.7)	8.7 (8.7)	6.3 (2.6–10.9; 0.3–44.7)	3.2 (2.7)	2.2 (1.0–5.1; 0.2–9.9)
	Ritlecitinib 200/30 mge	130	121	33.7 (13.8)	31.5 (21.0–44.0; 12.0–65.0)	45 (35)	85 (65)	90.5 (14.2)	11.6 (11.7)	7.3 (2.9–15.1; 0.0–57.1)	3.4 (2.9)	2.7 (1.1–5.1; 0.0–13.0)
	Ritlecitinib 200/50 mgd	132	124	34.5 (15.0)	34.0 (22.0–46.5; 12.0–71.0)	51 (39)	81 (61)	90.3 (15.1)	9.9 (10.8)	6.7 (2.5–11.3; 0.3–60.1)	3.4 (2.9)	2.3 (1.0–5.5; 0.0–10.0)
	Placebo	131	130	34 (15)	32.0 (22.0–44.0; 12.0–73.0)	45 (34)	86 (66)	93 (11.5)	11.0 (11.8)	7.4 (2.6–14.4; 0.1–58.2)	3.2 (2.7)	2.5 (1.1–4.9; 0.0–10.0)
Zhou et al. [19]	Ivarmacitinib 2 mg	23	23	36.0 (10.4)		12 (52.2)	11 (47.8)	65.2 (22.7)			46.8 (27.6)
	Ivarmacitinib 4 mg	23	23	33.3 (12.2)		9 (39.1)	14 (60.9)	62.1 (20.7)			37.6 (28.1)
	Ivarmacitinib 8 mg	24	24	37.9 (13.7)		9 (37.5)	15 (62.5)	63.9 (22.5)			39.6 (22.2)
	Placebo	24	24	34.7 (9.4)		12 (50.0)	12 (50.0)	61.3 (19.7)			37.9 (26.6)
Zhou et al. [20]	Ivarmacitinib 4 mg	109	109	31.7 (8.9)	31 (18–56)	51 (46.8)	58 (53.2)	78.6 (18.3)	5.4 (5.7)	3.9 (0.7–32.4)	3.3 (2.1)	3.1 (0.7–7.7)
	Ivarmacitinib 8 mg	111	111	31.6 (10.0)	30 (18–60)	49 (44.1)	62 (55.9)	78.6 (18.7)	6.0 (5.7)	4.4 (0.5–26.8)	3.4 (2.3)	2.9 (0.5–8.0)
	Placebo	110	110	32.4 (10.3)	31 (18–63)	53 (48.2)	57 (51.8)	79.1 (18.2)	6.3 (5.5)	5.0 (0.6–24.6)	3.1 (2.1)	2.9 (0.6–8.0)
Paracha et al. [26]	Tofacitinib 5 mg BID	52	52	20.23 (7.14)		37 (71.2)	15 (28.8)	91.02 (10.21)	6.59 (4.01)
	Azathioprine 2mg/kg	52	52	22.26 (8.07)		33 (63.5)	19 (36.5)	91.02 (10.63)	7.98 (4.40)
Anderson et al. [30]	Ritlecitinib 200/50 mgd	36	36	35.1 (9.6)		11 (30.6)	25 (69.4)		8.9	3
	Placebo	35	35	34.2 (9.0)		10 (28.6%)	25 (71.4%)		8.9	3

BID twice daily, IQR interquartile range, SD standard deviation

^aRefers to the BRAVE-AA1 trial in King et al. [18]

^bRefers to the BRAVE-AA2 trial in King et al. [18]

^cBrepocitinib 60 mg once daily for 4 weeks, then 30 mg once daily for 20 weeks

^dRitlecitinib 200 mg once daily for 4 weeks, then 50 mg once daily for 20 weeks

^eRitlecitinib 200 mg once daily for 4 weeks, then 30 mg once daily for 20 weeks

Quality Assessment and Risk of Bias

Of the 12 RCTs, half (n = 6) were considered to have a low overall risk of bias. The remaining six trials were judged to have “some concerns”. These concerns were primarily localised to Domain 3 (bias due to missing outcome data) and Domain 5 (bias in the selection of the reported result). No trials were rated as having a high overall risk of bias (Appendices 2A and 2B of the ESM).

Meta-Analysis of Efficacy and Safety Outcomes

SALT ≤10

Across six RCTs involving 2847 participants, JAK inhibitors were associated with more patients achieving SALT ≤10 compared with placebo at week 24 (OR 5.69; 95% CI 3.58–9.03; I² = 50%) (Fig. 2) [17–19, 27]. In three trials extending to 36 weeks (n = 1310), this effect was amplified, with more patients treated with JAK inhibitors achieving SALT ≤10 compared with those receiving placebo (OR 9.40; 95% CI 4.58–19.29; I² = 42%) (Fig. 3) [18, 28].

Fig. 2.

Efficacy (Severity of Alopecia Tool [SALT] ≤10 and SALT ≤20) of Janus kinase (JAK) inhibitors versus placebo at week 24. BID twice daily, CI confidence interval, OR odds ratio. ^aRefers to the BRAVE-AA1 trial in King et al., 2022. ^bRefers to the BRAVE-AA2 trial in King et al., 2022. ^cRitlecitinib 200/30 mg refers to the regime of 200 mg once daily for 4 weeks, followed by 30 mg once daily for 20 weeks. ^dRitlecitinib 200/50 mg refers to the regime of 200 mg once daily for 4 weeks, followed by 50 mg once daily for 20 weeks. ^eBrepocitinib 60/30 mg refers to the regime of 60 mg once daily for 4 weeks, followed by 30 mg once daily for 20 weeks. Study references: King et al. [18]; King et al. [29]; King et al. [16]; King et al. [17]; Tsianakas et al. [27]; Zhou et al. [19]; Zhou et al. [20]

Fig. 3.

Efficacy (Severity of Alopecia Tool [SALT] ≤10 and SALT ≤20) of Janus kinase (JAK) inhibitors versus placebo at week 36. CI confidence interval, OR odds ratio. ^aRefers to the BRAVE-AA1 trial in King et al., 2022. ^bRefers to the BRAVE-AA2 trial in King et al., 2022. Study references: King et al. [28]; King et al. [18]

SALT ≤20

At weeks 12–16, two trials involving 1200 participants demonstrated that JAK inhibitors were associated with a greater proportion of patients achieving SALT ≤20 compared with placebo (OR 5.03; 95% CI 1.97–12.86; I² = 67.72%) (Fig. 4) [18]. This benefit was maintained in eight trials at week 24 (n = 3714; OR 7.99; 95% CI 5.13–12.44; I² = 62%) (Fig. 2) [16–20, 27, 29] and in three trials at week 36 (n = 1310; OR, 8.75; 95% CI 5.36–14.28; I² = 32.81%) (Fig. 3) [18, 28].

Fig. 4.

Efficacy (Severity of Alopecia Tool [SALT] ≤20) of Janus kinase (JAK) inhibitors versus placebo at weeks 12–16. CI confidence interval, OR odds ratio. ^aRefers to the BRAVE-AA1 trial in King et al., 2022. ^bRefers to the BRAVE-AA2 trial in King et al., 2022. Study references: King et al. [18]

SALT50

Three trials demonstrated that JAK inhibitors were associated with more patients achieving at least a 50% improvement in SALT score compared with placebo at weeks 12–16 (n = 1310; OR 4.96; 95% CI 3.16–7.77; I² = 12.91%) (Fig. 1 of the ESM) [18, 28]. This benefit was maintained in four trials at week 24 (n = 715; OR 4.56; 95% CI 3.13–6.64; I² = 54.0%) (Fig. 2 of the ESM) [15, 19, 20, 29].

SALT75

Across four RCTs involving 715 participants, JAK inhibitors were associated with more patients achieving at least a 75% reduction in SALT score compared with placebo at week 24 (OR 5.55; 95% CI 3.23–9.52; I² = 39%) (Fig. 2 of the ESM) [15, 19, 20, 29].

SALT90

Four trials involving 715 patients showed that treatment with JAK inhibitors was significantly more likely to result in at least 90% scalp hair regrowth compared with placebo at week 24 (OR 4.28; 95% CI 2.01–9.09; I² = 49%) (Fig. 2 of the ESM) [15, 19, 20, 29]. By week 36, three trials (n = 1310) demonstrated that the odds of achieving SALT90 were over ten times greater in patients treated with JAK inhibitors compared with placebo (OR 10.11; 95% CI 4.91–20.81; I² = 25.11%) (Fig. 3 of the ESM) [18, 28].

Adverse Events (AEs)

Safety data from 12 RCTs involving 4141 patients showed that JAK inhibitors were associated with a higher risk of total AEs compared with placebo or azathioprine (OR 1.60; 95% CI 1.33–1.94; I² = 53.43%) (Fig. 5) [15–20, 26–30]. However, there were no statistically significant differences in the rate of severe AEs (n = 4141; OR 0.86; 95% CI 0.56–1.31; I² = 0%) (Fig. 4 of the ESM) [17–20, 26, 28–30], serious AEs (n = 4141; OR 1.04; 95% CI 0.69–1.57; I² = 0%) (Fig. 5 of the ESM) [15–20, 26–30] and treatment discontinuation because of AEs (n = 4141; OR, 1.22; 95% CI 0.84–1.76; I² = 0%) between JAK inhibitors and placebo (Fig. 6 of the ESM) [15–20, 26–30].

Fig. 5.

Total adverse effects of Janus kinase (JAK) inhibitors versus placebo. AEs adverse events, BID twice daily, CI confidence interval, OR odds ratio. ^aRefers to the BRAVE-AA1 trial in King et al., 2022. ^bRefers to the BRAVE-AA2 trial in King et al., 2022. ^cRitlecitinib 200/30 mg refers to the regime of 200 mg once daily for 4 weeks, followed by 30 mg once daily for 20 weeks. ^dRitlecitinib 200/50 mg refers to the regime of 200 mg once daily for 4 weeks, followed by 50 mg once daily for 20 weeks. ^eBrepocitinib 60/30 mg refers to the regime of 60 mg once daily for 4 weeks, followed by 30 mg once daily for 20 weeks. *This study compared tofacitinib with azathioprine (2 mg/kg once daily) rather than placebo. Study references: Anderson et al. [30]; King et al. (baricitinib) [28]; King et al. (ritlecitinib and brepocitinib) [15]; King et al. [18]; King et al., [29]; King et al. [16]; King et al. [17]; Paracha et al. [26]; Tsianakas et al. [27]; Zhou et al. [19]; Zhou et al. [20]

To address potential heterogeneity introduced by the active comparator, a sensitivity analysis was performed excluding the tofacitinib versus azathioprine trial [26]. Across all four AE endpoints, removing this comparison did not materially alter the conclusions of the safety analyses. For total AEs, the corrected pooled estimate (OR 1.65; 95% CI 1.37–1.99) remained highly consistent with the original analysis (OR 1.60; 95% CI 1.33–1.94). Similarly, the pooled estimates for severe AEs (OR 0.86; 95% CI 0.56–1.31), serious AEs (OR 1.04; 95% CI 0.59–1.57) and treatment discontinuation because of AEs (OR 1.22; 95% CI 0.84–1.76) remained comparable to the original analyses and did not change the statistical interpretation of the findings. The observed differences are minor and reflect correction of the reporting error and expected instability of rare-event models, particularly for severe and serious AEs for which event counts were sparse. Overall, exclusion of the tofacitinib versus azathioprine trial [26] and correction of the pooled estimates did not meaningfully influence the direction, magnitude or interpretation of the safety outcomes. Given the sparse-event nature of these safety outcomes, the results of this sensitivity analysis are primarily descriptive and should be interpreted cautiously.

Some of the data from the included RCTs were excluded from the primary meta-analysis because of non-comparable outcome measures. These findings are summarised descriptively below.

Changes in SALT Scores

Across RCTs reporting change in SALT score, improvements at 24 weeks consistently favoured JAK inhibitors over placebo or azathioprine (36.0–59.0% for JAK inhibitors vs 2.0–29.8% for placebo or azathioprine) (Table 2) [15, 26, 31]. By 36 weeks, trials demonstrated sustained and greater reductions in SALT scores with active treatment (29.9–58.1% for JAK inhibitors vs 4.3–11.7% for placebo), reinforcing the durability of response across studies [18, 28]. The observed range of percentage change between treatment and placebo groups provides a clear descriptive overview of the magnitude and consistency of treatment response.

Table 2.

Percentage change in SALT scores with different oral JAK inhibitors and placebo

Study	JAK inhibitor and dose	Treatment duration (weeks)	% Change in SALT score from baseline	p-Value	N analysed
Tsianakas et al. [31]	Deuruxolitinib 8 mg BID	24	− 45.0%	p < 0.0001 vs placebo	258
	Deuruxolitinib 12 mg BID	24	− 53.0%	p < 0.0001 vs placebo	129
	Placebo	24	+2.0%	−	130
King et al. [15]	Brepocitinib 60/30 mgc	24	− 59.0%	p < 0.001 vs placebo	47
	Ritlecitinib 200/50 mgd	24	− 36.0%	p < 0.001 vs placebo	48
	Placebo	24	− 2.0%	−	47
Paracha et al. [26]	Tofacitinib 5 mg BID	24	− 84.50%	p < 0.0001 vs azathioprine	52
	Azathioprine 2 mg/kg daily	24	− 29.80%	−	52
King et al. [28]	Baricitinib 2 mg	36	− 48.2%	p < 0.001 vs placebo	27
	Baricitinib 4 mg	36	− 58.1%	p < 0.001 vs placebo	27
	Placebo	36	− 11.7%	−	28
King et al. [18]a	Baricitinib 2 mg	36	− 32.7%	p < 0.001 vs placebo	184
	Baricitinib 4 mg	36	− 47.1%	p < 0.001 vs placebo	281
	Placebo	36	− 9.0%	−	180
King et al. [18]b	Baricitinib 2 mg	36	− 29.9%	Not significant vs placebo	156
	Baricitinib 4 mg	36	− 48.7%	p < 0.001 vs placebo	234
	Placebo	36	− 4.3%	−	156

BID twice daily, JAK Janus kinase, SALT Severity of Alopecia Tool

^aRefers to the BRAVE-AA1 trial in King et al. (2022)

^bRefers to the BRAVE-AA2 trial in King et al. (2022)

^cBrepocitinib 60 mg once daily for 4 weeks, then 30 mg once daily for 20 weeks

^d200 mg once daily for 4 weeks, then 50 mg once daily for 20 weeks

Two trials of ivarmacitinib reported the primary outcome as the least squares mean (LSM) percentage change in SALT score from baseline. [19, 20] As LSM represents a model-adjusted mean derived from a regression analysis, it is not directly comparable to unadjusted means reported in other trials. In the phase II study, the 4-mg and 8-mg dose groups demonstrated LSM percentage changes of -56.11% (p = 0.003) and -51.01% (p = 0.010), respectively, at week 24, both significantly greater than placebo (-19.87%). The 2-mg group showed a smaller non-significant change (-30.51%). [19] These findings were corroborated by the phase III trial, in which ivarmacitinib 4 mg and 8 mg resulted in LSM percentage changes in SALT scores of -46.84% and -53.44%, respectively, at week 24 compared with -19.86% in the placebo arm (p <0.0001 for both doses).

A phase IIa ritlecitinib study reported the LSM absolute change in SALT, showing a reduction of 38.2 at week 36 with the 200 mg daily for 4 weeks followed by the 50-mg daily dosing regime compared with 6.8 in the placebo group [30].

Eyebrow and Eyelash Hair Regrowth

Across RCTs, oral JAK inhibitors demonstrated efficacy in promoting eyebrow and eyelash hair regrowth in patients with moderate-to-severe AA, despite heterogeneity in outcome measures and response definitions [15, 16, 18, 28, 32].

Although studies employed different assessment tools, including the Brigham Eyebrow Tool for Alopecia (BETA) and Brigham Eyelash Tool for Alopecia (BELA), ≥1- or ≥2-grade improvement, and Clinician-Reported Outcome (ClinRO), the benefit was remarkably consistent across agents. All evaluated JAK inhibitors showed statistically significant improvements versus placebo at primary assessment timepoints (typically 24–36 weeks), with minimal response observed in placebo groups.

A dose–response relationship was evident in several trials, with higher-dose ritlecitinib and baricitinib regimes yielding greater proportions of patients achieving significant eyebrow and eyelash hair regrowth compared with lower-dose regimes [16, 18, 28].

However, direct comparison across trials was limited by heterogeneity in outcome measures, grading scales and response thresholds. As such, a quantitative synthesis was not feasible.

Deuruxolitinib

In the THRIVE-AA2 study, deuruxolitinib produced dose-dependent improvements in BETA and BELA scores (p < 0.0001). By week 24, participants in the 12-mg twice-daily (BID) group achieved mean improvements of 1.3 (eyebrows) and 1.5 (eyelashes), while those in the placebo arm showed worsening scores [32].

Ritlecitinib and Brepocitinib

In a phase IIa trial, a ≥1 grade eyebrow improvement occurred in 51% of patients receiving ritlecitinib (200 mg daily for 4 weeks, then 50 mg daily for 20 weeks) and 72% receiving brepocitinib (60 mg daily for 4 weeks, then 30 mg daily for 20 weeks) compared with 15% of patients receiving placebo at week 24. As far as eyelashes were concerned, improvements were seen in 60% and 81% of patients respectively versus 14% in the placebo arm [15].

A larger phase IIb ritlecitinib trial used a stricter endpoint (≥2 grade improvement). At week 24, the 200 mg daily for 4 weeks followed by 30 mg daily regime resulted in this endpoint being achieved in 25% of patients for eyebrows (p = 0.0016) and 21% for eyelashes (p = 0.024), while the 200-mg/50-mg regime demonstrated higher response rates of 34% (p <0.0001) and 30% (p = 0.0001), respectively. Both dosing schedules were superior to placebo [16].

Baricitinib

A phase II trial of baricitinib also demonstrated significant improvements in eyebrow and eyelash responses. At week 36, the proportion of patients achieving a ClinRO score of 0 or 1 (indicating full coverage or minimal gaps) for eyebrows was 39.1% in the 4-mg group (p = 0.012 vs placebo) and 28.6% in the 2-mg group (p = 0.034 vs placebo) compared with 4.3% in the placebo group. For eyelashes, 60.0% and 40.0% of patients in the 4-mg and 2-mg groups, respectively, achieved this endpoint compared with 5.9% in the placebo group. However, only improvements shown in the 4-mg group were statistically significant (p = 0.041) [28].

The larger phase III BRAVE-AA1 (n = 654) and BRAVE-AA2 trials (n = 546) of baricitinib used a similar ClinRO endpoint of 0 or 1. Positive results were consistent across both trials. At week 36, the 4-mg dose of baricitinib showed statistically significant results (p <0.001 vs placebo for all analyses), with 35.2% (BRAVE-AA1) and 38.9% (BRAVE-AA2) of patients achieving this endpoint for eyebrows, and 36.2% (BRAVE-AA1) and 36.8% (BRAVE-AA2) for eyelashes. This contrasts with the placebo arms, in which the proportion of patients achieving a ClinRO score of 0 or 1 ranged from approximately 4 to 7% [18].

Patient Satisfaction

Across RCTs, oral JAK inhibitors demonstrated consistent and clinically meaningful improvements in patient satisfaction, despite considerable heterogeneity in outcome measures [15, 17, 19, 29, 32]. Although patient satisfaction was assessed using a range of instruments, including categorical satisfaction ratings (e.g. proportion of patients who were “satisfied” or “very satisfied”) and patient-reported outcome measures such as the Alopecia Areata Symptom Impact Scale (AASIS), the trend was concordant across studies, with all evaluated agents showing significant improvements compared with placebo.

A dose–response relationship was evident in multiple studies, with deuruxolitinib and ivarmacitinib trials demonstrating greater satisfaction rates with higher doses [17, 19, 20, 27]. Importantly, improvements in patient satisfaction were aligned with objective efficacy outcomes, including scalp, eyebrow and eyelash hair regrowth, supporting the clinical relevance of these findings and suggesting that substantial hair regrowth translates into meaningful patient satisfaction. However, heterogeneity in outcome measures and grading scales precluded a meta-analysis.

Deuruxolitinib

The deuruxolitinib trials demonstrated clear dose-dependent improvements in patient satisfaction. In the THRIVE-AA1 study, 42.1% of patients receiving 8 mg BID and 53.0% receiving 12 mg BID reported being “satisfied” or “very satisfied” with hair regrowth compared with 4.7% in the placebo group at week 24 (p <0.0001 for both) [17]. The THRIVE-AA2 study showed similar findings, with 46.5% and 51.7% of patients in the 8-mg BID and 12-mg BID groups being “satisfied” or “very satisfied” versus 1.7% in the placebo arm (p <0.0001 for both) [32]. A separate phase II trial demonstrated mean changes in AASIS scores of -3.8 (8-mg BID arm, p < 0.01) and -5.1 (12-mg BID arm, p <0.001) compared with -1.2 in the placebo arm [29].

Ritlecitinib and Brepocitinib

A phase IIa RCT comparing ritlecitinib and brepocitinib with placebo also measured patient-reported improvements using the AASIS. At 24 weeks, LSM changes in AASIS scores were significantly greater with ritlecitinib (-10.8) and brepocitinib (-17.4) compared with placebo (+3.0) [15].

Ivarmacitinib

Patient satisfaction outcomes for ivarmacitinib showed consistent benefit. In a phase II RCT, 47.8% (4 mg) and 50.0% (8 mg) of participants reported being “satisfied” compared with 16.7% in the placebo group at week 24. [19] A subsequent larger phase III trial (n = 330) confirmed these findings, with 56.0% (4 mg) and 63.1% (8 mg) of patients reporting satisfaction versus 23.6% of patients who received placebo (p <0.0001 for both doses). [20]

Discussion

Efficacy of Oral Janus Kinase (JAK) Inhibitors

This systematic review and meta-analysis involving 4141 patients found that oral JAK inhibitors were associated with significantly higher rates of clinically meaningful hair regrowth compared with placebo across all assessed timepoints. These findings reinforce the role of JAK-signal transducer and activator of transcription pathway inhibition in reversing disease activity in moderate-to-severe AA.

Our data highlight the significant disease burden of participants, as shown by a mean baseline SALT score of 86.0 and a mean duration of current episode of hair loss of 4.8 years. The moderate statistical heterogeneity observed in the pooled analyses likely reflects differences in baseline disease severity, JAK inhibitor selection and endpoint definitions rather than true inconsistency in therapeutic effect.

Our analysis highlights the time-dependent nature of the therapeutic response. The odds of achieving significant hair regrowth demonstrated a temporal trend, with higher response rates observed at later assessment timepoints across the pooled trials (from 12 to 16 weeks through to 36 weeks). For instance, the OR for achieving SALT≤10 increased from 5.69 at week 24 to 9.40 at week 36. Similarly, the OR for achieving SALT90 more than doubled between these two timepoints, reaching 10.11 by week 36. These results highlight the utility of sustained treatment and suggest that shorter trials may underestimate the cumulative benefit achieved with longer exposure. Furthermore, our findings for secondary outcomes, including statistically significant eyebrow and eyelash hair regrowth as well as high rates of satisfaction in patients treated with oral JAK inhibitors compared with placebo, reinforce the clinical utility of these agents beyond scalp hair regrowth alone.

Real-World Evidence

Real-world studies of oral JAK inhibitors in AA often report higher efficacy rates compared with those observed in RCTs. This discrepancy may be partly attributable to the heterogeneity of real-world populations, incorporation of adjunctive therapies and treatment durations that often extend beyond the typical timeframe of clinical trials. In clinical practice, JAK inhibitors are frequently combined with treatments such as topical or oral minoxidil, and topical, intralesional or oral corticosteroids, which are typically excluded from RCT protocols but may enhance the therapeutic effect.

In a 52-week multicentre retrospective study of 96 patients with AA treated with baricitinib, 47.4% of whom also received adjunctive therapies such as oral minoxidil, and topical or intralesional corticosteroids, 61.5% achieved SALT ≤20 [33]. Similarly, a real-world study reported that 63.2% of adults with severe AA (SALT ≥50) treated with baricitinib 4 mg achieved SALT ≤20 at week 48 [34]. In contrast, the 36-week BRAVE-AA1 and BRAVE-AA2 RCTs reported SALT ≤20 responses of 38.8% and 35.1%, respectively, among patients receiving baricitinib 4-mg monotherapy [18]. A separate real-world study showed that, among those with an inadequate response to baricitinib monotherapy, the addition of systemic corticosteroids resulted in 60% achieving SALT ≤20, with the greatest benefit observed in partial responders to baricitinib monotherapy [35].

Similarly, outcomes with tofacitinib treatment further illustrate the complementary insights gained from real-world versus RCT settings. A retrospective study of 202 patients with AA treated with tofacitinib, most of whom also received adjunctive therapies, reported that 55.9% achieved SALT50 after 18 months [36].

Likewise, a retrospective cohort study of 89 patients with AA treated with ritlecitinib (46.1% and 9.0% of whom were also treated with topical corticosteroids and topical minoxidil, respectively) reported a SALT50 response of 80.1% at 24 weeks [37]. Two recent real-world studies from Italy have shown that, among patients with severe AA (SALT ≥50) treated with ritlecitinib 50 mg, 40.2% and 50% achieved SALT ≤20 at week 24, respectively [38, 39]. These outcomes are numerically higher than those observed in the ALLEGRO phase IIb/III trial, in which 23% of patients receiving ritlecitinib 50 mg achieved SALT ≤20 at 24 weeks. However, important differences in demographics should be considered: adolescents comprised a substantially larger proportion of participants in the real-world studies (69% and 92% of patients, respectively), compared with 14% of patients in the ALLEGRO phase IIb/III trial.

The higher response rates observed in real-world cohorts may, in part, reflect differences in age distribution, patient selection, use of adjunctive therapies and variations in adherence and follow-up, which are not consistently captured or reported in observational settings.

Comparison with Previous Systematic Reviews and Novel Contributions

Several systematic reviews and meta-analyses have evaluated the efficacy and safety of JAK inhibitors in AA, consistently demonstrating favourable outcomes across a range of study designs. Early work by Phan and Sebaratnam synthesised predominantly case reports, case series and small cohort studies, highlighting promising response rates, but was limited by low-quality evidence and significant heterogeneity [40]. Subsequent reviews expanded the evidence base by incorporating prospective and randomised studies, but continued to include a considerable proportion of non-randomised data, thereby limiting the strength and interpretability of pooled estimates [41–43].

Importantly, several prior reviews included both oral and topical JAK inhibitors within pooled analyses. Given the well-established differences in efficacy between these formulations (particularly the limited efficacy of topical agents compared with oral therapies), this approach may dilute treatment effects and obscure clinically meaningful differences [40, 41, 43–45]. In addition, some analyses incorporated biologic agents used for other inflammatory conditions (e.g. dupilumab, secukinumab), which demonstrate limited or inconsistent efficacy in AA, further contributing to heterogeneity and potentially skewing pooled outcomes [48].

In contrast, the present study adopts a more rigorous and clinically focused approach. First, we restricted inclusion exclusively to RCTs, in contrast to most prior reviews that incorporated observational studies of variable quality. Although two recent meta-analyses also focused on RCTs, they included both topical and systemic therapies and did not systematically account for temporal treatment effects [44, 46]. By limiting inclusion to RCTs of oral JAK inhibitors, we reduced both methodological and therapeutic heterogeneity, providing a more robust estimate of efficacy and safety.

Second, and most importantly, this study is the first to incorporate time-stratified efficacy analyses, evaluating outcomes at clinically relevant intervals (e.g. 12–16, 24 and 36 weeks). Prior meta-analyses have generally pooled outcomes across heterogeneous or unspecified timepoints, thereby obscuring the temporal dynamics of treatment response. Our findings demonstrate a clear progression in efficacy over time, with increasing odds of achieving clinically meaningful SALT thresholds with a longer treatment duration. This temporal resolution provides clinically actionable insights, particularly in guiding expectations regarding treatment response and duration.

Finally, by standardising outcomes using validated SALT thresholds (≤10, ≤20, SALT50, SALT75, SALT90), this analysis enhances comparability across trials and improves clinical interpretability. Collectively, these methodological refinements position the present study as a more rigorous, temporally informative and clinically relevant synthesis of the current RCT evidence base for oral JAK inhibitors in AA.

Safety Profile of Oral JAK Inhibitors

The safety of JAK inhibitors is a critical consideration, particularly in the context of the FDA’s boxed warning regarding the risk of major adverse cardiovascular events, thromboembolism and malignancy [47]. Notably, this warning was based on post-marketing safety data from older patients with rheumatoid arthritis who had baseline cardiovascular risk factors, a population that differs substantially from patients with AA who are typically younger and healthier [48].

Systematic reviews and real-world data in patients with AA have consistently demonstrated that oral JAK inhibitors have a favourable safety profile. Our analysis found that, while oral JAK inhibitors were associated with a higher risk of total AEs compared with placebo (OR 1.60), there were no statistically significant differences in the rates of severe AEs, serious AEs or treatment discontinuation because of AEs. Most reported AEs were mild and included acne, upper respiratory tract infections and headache. Specifically, severe AEs were observed in 0–8.3% of patients across treatment arms compared with 0–4.5% in placebo groups. Serious AEs occurred with a comparably low frequency, affecting 0–4.2% of patients in both treatment and placebo arms. The nature of these serious events varied across trials and included infections (such as pyelonephritis, cellulitis and appendicitis), rhabdomyolysis, pulmonary embolism and isolated cases of malignancy [15–20, 26, 28–30, 32].

A 2025 network meta-analysis of 12 RCTs comprising 3840 patients further supported this overall tolerability while also identifying drug-specific differences. Baricitinib was found to be associated with higher rates of acne, urinary tract infections and hyperlipidaemia. Deuruxolitinib showed increased incidences of acne and elevated creatine phosphokinase whereas ritlecitinib was also associated with elevated creatine phosphokinase [49]. Across trials, reported AEs are generally mild and include gastrointestinal symptoms, acneiform eruptions, upper respiratory tract infections, hypercholesterolaemia and haematological abnormalities [33, 36, 49, 50].

Emerging data extending to 152 weeks have further substantiated the favourable safety profile of JAK inhibitors in patients with AA. In an analysis of patients treated continuously with baricitinib, incidence rates of AEs remained stable relative to earlier 104-week data, with no new safety concerns identified. Specifically, incidence rates per 100 patient-years remained low for serious infections (0.6), major adverse cardiovascular events (<0.1) and deep vein thrombosis or pulmonary embolism (0.1). These findings suggest that the risk of these serious AEs does not increase with cumulative exposure in patients with AA [51].

Low rates of discontinuation because of AEs observed in real-world studies align with the findings of our meta-analysis. Nevertheless, because AA often necessitates long-term therapy, ongoing pharmacovigilance and additional long-term safety data specific to the AA population remain essential to fully characterise the risk-benefit profile.

Emerging JAK Inhibitors

There is a growing body of real-world evidence supporting the efficacy of other JAK inhibitors in the treatment of AA [6, 50, 52–57]. Some of these agents are currently being investigated in clinical trials.

Upadacitinib

Upadacitinib is a JAK1 inhibitor that has shown encouraging results across several retrospective studies and case series. In a retrospective study of 17 patients who had not responded to other systemic therapies, 88.2% and 35.3% achieved SALT50 and SALT90, respectively, after 24 weeks of treatment with upadacitinib [58]. In a case series of 22 patients with severe AA, some of whom had previously not responded to baricitinib, 77% achieved SALT ≤20 after being treated with upadacitinib for a mean duration of 5.5 months, with more significant hair regrowth in those who were also treated with corticosteroids [59]. Efficacy has also been reported in paediatric populations, with a case series of three adolescents with severe AA demonstrating a median SALT score reduction of 58% after 6 months of treatment [60]. A phase III trial is underway [21], with preliminary data showing that SALT ≤20 was achieved in 42.5% and 55.0% of patients receiving doses of 15 mg and 30 mg, respectively, after 24 weeks [61].

Deucravacitinib

Deucravacitinib is a selective tyrosine kinase 2 inhibitor that has shown favourable outcomes in isolated cases of AA [62–64]. A 68-year-old man with severe AA and plaque psoriasis experienced complete hair regrowth of the scalp and eyebrows within 4 months of commencing deucravacitinib treatment [64]. Two patients with AU experienced significant hair regrowth (SALT scores of 13 and 45.2) after 7 and 6 weeks of treatment, respectively [62, 63]. Although tyrosine kinase 2 is not central to the interferon-gamma/interleukin-15 axis, these results suggest that alternative immunomodulatory pathways may be therapeutically relevant in AA. A phase II AA trial was initiated but later terminated for reasons unrelated to efficacy or safety [65].

Other JAK Inhibitors

KL130008, a JAK1/JAK2 inhibitor, is in phase II testing [66], with phase I data showing good tolerability and dose-dependent JAK-signal transducer and activator of transcription inhibition [67]. SYHX1901, a pan-JAK inhibitor, is currently being evaluated in a phase II RCT [68, 69].

Limitations

This systematic review has several limitations that should be considered when interpreting the findings. First, the included RCTs demonstrated methodological heterogeneity, including differences in trial protocols, dosing regimes and timing of outcome assessments. While our analysis aimed to stratify outcomes by follow-up duration, such variability may influence the pooled estimates.

Second, this analysis incorporated placebo-controlled studies, and the absence of direct head-to-head comparisons limits conclusions regarding the relative efficacy and safety of individual oral JAK inhibitors. Third, while recent 3-year data for baricitinib are now available, the short-to-medium follow-up periods of most other included trials limit the assessment of the long-term efficacy and safety of these agents.

Furthermore, eyebrow and eyelash hair regrowth, and patient satisfaction were assessed using heterogeneous instruments across trials. This variability precluded the synthesis of quantitative data, necessitating reliance on narrative synthesis.

In addition, publication bias was not formally assessed using funnel plot asymmetry because of the limited number of included studies (11) and the presence of multi-arm trials contributing non-independent effect sizes (shared placebo groups). Therefore, the potential for publication bias cannot be excluded.

Finally, only patients with a duration of current episode of hair loss of ≤8–10 years were included in all but one of the RCTs (Paracha et al.) [26], as it is widely accepted that long-standing alopecia (>8–10 years) may be irreversible. Consequently, our findings are not generalisable to patients with long-standing hair loss.

Future Directions

The expanding landscape of JAK inhibitors for AA presents opportunities and challenges. Although their efficacy is well established, head-to-head RCTs are needed to directly compare agents in terms of efficacy and safety, enabling more precise treatment selection based on disease subtype, comorbidities and prior therapies.

Future research should also optimise treatment regimes by exploring optimal induction and maintenance dosing, as well as combination approaches. Trials pairing JAK inhibitors with adjunctive agents (e.g. oral minoxidil and corticosteroids) may help validate the synergistic benefits observed in real-world practice.

Conclusions

In this systematic review and meta-analysis, oral JAK inhibitors were associated with significantly higher rates of clinically meaningful hair regrowth compared with placebo in adults and adolescents with moderate-to-severe AA, demonstrating a trend of higher efficacy at later timepoints (up to 36 weeks) across pooled trials. While these agents were associated with a higher risk of total AEs, there were no significant differences in the rates of serious AEs or treatment discontinuation, suggesting an acceptable safety profile in the short to medium term. These results strengthen the evidence base for oral JAK inhibitors as an effective systemic therapy for moderate-to-severe AA. However, the current lack of head-to-head data and the methodological heterogeneity across trials underscore the need for comparative trials and standardisation of outcome measures.

Supplementary Information

Below is the link to the electronic supplementary material.

Funding

Open access funding provided by Melbourne Health (RMH/RWH), Victorian Health Libraries Consortium (VHLC).

Declarations

Funding

No funding was received for the preparation of this article.

Conflicts of Interest/Competing Interests

Samuel Morriss, Chuan Hui Foo, Emily Lin, Flavia Rodrigues-Dias, Daniella Kushnir-Grinbaum, Holly FitzGerald, Shalem Leemaqz, Sharon Lee and Bevin Bhoyrul have no conflicts of interest that are directly relevant to the content of this article.

Ethics Approval

Not applicable.

Consent to Participate

Not applicable.

Consent for Publication

Not applicable.

Availability of Data and Material

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Code Availability

Not applicable.

Authors’ Contributions

SM, FD, HF, DK and BB conducted the data extraction. SM and BB conducted the study selection. Statistical analysis was performed by CHF and SL. SM and BB drafted the manuscript. BB revised the final manuscript draft. All authors reviewed, edited and approved the final version of the manuscript.