FormalPara Key Summary Points

There has been no universal agreement on how to best manage alopecia areata (AA), possibly because of the lack of options approved for this indication and the heterogeneous reporting of data across drug classes, with little consistency of data collection or clinical endpoints used.

A systematic literature review identified that the most robust available evidence is for the Janus kinase (JAK) inhibitor class, in particular baricitinib, which was consistently shown to be more beneficial than placebo across various clinical outcomes in adults with at least 50% scalp hair loss.

Overall, hair regrowth varied widely for the other classes of drugs (systemic corticosteroids, topical immunotherapy, cyclosporine A, methotrexate, and azathioprine), was generally low for patients with moderate-to-severe AA, and relapses were common.

The current management of patients with moderate-to-severe AA often relies on the use of treatments that have not been well evaluated in clinical trials, but the most robust available evidence supports the use of baricitinib, and other oral JAK inhibitors, in patients with severe AA.

Introduction

Alopecia areata (AA) is an autoimmune disease associated with non-scarring hair loss that typically affects the scalp but can also affect other parts of the body; its severity can range from small patches of hair loss to total hair loss [1,2,3,4]. Hair regrowth is common in the early stages of the disease (spontaneous or on treatment), but tends to be rare for patients with chronic and extensive hair loss [1, 5]. AA can be associated with a substantial quality-of-life impairment and psychological burden [6,7,8].

Until 2022, physicians had traditionally relied on treatments prescribed off-label [9], as there were no systemic therapies for AA approved by major regulatory bodies such as the US Food and Drug Administration (FDA) or European Medicine Agency (EMA). Several guidelines and an expert consensus have been published, but no universal agreement exists on how to best manage AA [1, 4, 10, 11].

With the recent approval of baricitinib [12, 13], an oral selective Janus kinase (JAK)1/JAK2 inhibitor, for the treatment of AA, there is a need to better understand the evidence supporting the management of AA in clinical practice. To this end, we conducted a systematic literature review (SLR) to describe the evidence pertaining to the efficacy and tolerability of treatments currently recommended for moderate-to-severe forms of AA.

Methods

SLR

We conducted an SLR in accordance with guidelines from the National Institute for Health and Care Excellence [14]. The eligibility criteria that defined the scope of studies to be synthetized in the SLR were defined according to Population, Intervention, Comparators, Outcomes, and Study (PICOS) design criteria and are listed in Table 1. The list of interventions included in the search strategies was identified from treatment guidelines and consensus statements [1, 10, 11, 15].

Table 1 Summary of PICOS eligibility criteria for the SLR
Full size table

Search Strategies and Information Sources

Searches for peer-reviewed publications were conducted in Embase, MEDLINE, and using the Cochrane website. Conference proceedings (as summarized in Table 2) for the years 2019–2022 were searched.

Table 2 Conference proceedings searched for articles of interest
Full size table

Clinical trial registries (ClinicalTrials.gov, the World Health Organization International Clinical Trials Registry Platform, and the European Union Clinical Trials Register), key Health Technology Assessment (HTA) databases, and other relevant websites were also searched to identify ongoing trials and drug assessments. Additionally, the bibliographies of several published SLRs were hand-searched for eligible records.

Literature searches were originally performed on July 2, 2021, and they were updated on February 4, 2022, and again on July 11, 2022. The full search strategies used for each search are presented in Table S1.

Selection Procedure

Titles and abstracts of identified publications were independently checked by two reviewers in parallel to select those that would proceed to full-text review; discrepancies in decisions were referred to a third reviewer. The same process was applied to publications selected for full-text review to ascertain final eligibility. Searches of conference proceedings, clinical trial registries, HTA databases, and other websites were performed by a single reviewer and checked by a second reviewer. Conference abstracts that presented data also available in a peer-reviewed publication were excluded. If duplicate abstracts with identical information were identified, only the most recent abstract was selected. A Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram was created.

Data Extraction

Data were extracted by one reviewer and independently checked by another. To facilitate the synthesis of evidence, selected records were allocated to one of three categories based on population, intervention, and study design. The first category (category 1) included randomized controlled trials (RCTs) versus placebo, with interventions recommended for the treatment of adults with moderate-to-severe forms of AA (systemic corticosteroids, topical immunotherapy, systemic immunosuppressants [i.e., cyclosporine A, methotrexate, azathioprine], and oral JAK inhibitors) [1, 10, 11, 15]. A suitable number of studies in this category may allow meta-analysis or combined analysis across treatments. The second category (category 2) included non-randomized trials and observational studies as well as RCTs without a placebo arm, with interventions recommended for the treatment of adults with moderate-to-severe forms of AA. The final category (category 3) included records with data from adults with mostly mild-to-moderate forms of AA or severity of hair loss at baseline that was not documented; records of studies assessing interventions that did not fall into category 1 or 2 (i.e., that were not recommended or were recommended in mild forms of AA or as adjunctive therapy only) were also included in category 3. Category 1 describes studies providing the highest level of evidence, while category 2 provided data for a narrative summary beyond placebo-controlled RCTs. Records in all three categories met the inclusion criteria, but only those included in categories 1 and 2 met the objective of our SLR. As a result, full data extraction was carried out for records falling into categories 1 and 2, whereas records falling into category 3 were summarized only. There was no formal data extraction from relevant entries identified in clinical trial registries, HTA databases, and other websites; rather, this search was conducted for informative purposes to better understand the treatment landscape.

Narrative Summary

The narrative summary focused on category 1 and 2 records, which reported data related to interventions recommended for the treatment of adults with moderate-to-severe forms of AA: systemic corticosteroids, topical immunotherapy, systemic immunosuppressants (cyclosporine A, methotrexate, azathioprine), and oral JAK inhibitors.

Quality Assessment

Quality assessment was performed by one reviewer and then checked by a second reviewer for all fully extracted records, except for conference proceedings and category 3 records. Assessment of the quality of RCTs was conducted using guidance provided by the Cochrane Collaboration [16]. The Centre for Reviews and Dissemination (CRD) checklist [17] was used for non-randomized trials and observational studies.

Ethical Approval

This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.

Results

Of 2557 identified records, a total of 53 were retained for data extraction. Nine records reported data from seven RCTs versus placebo (category 1), and 44 reported data from unique non-randomized trials, observational studies, or RCTs with no placebo arm (category 2) (Fig. 1). Data related to category 1 and 2 records are presented in Tables 3, 4, 5, 6, 7, and 8 [18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70]. Information on records allocated to category 3 and excluded from the current summary is provided in Table S2.

Fig. 1
figure 1

PRISMA flow diagram showing the evolution of studies included in the current systematic narrative review. Category 1: RCTs versus placebo, with interventions recommended for moderate-to-severe forms of AA (systemic corticosteroids, systemic immunosuppressants, JAK inhibitors, contact immunotherapy). Category 2: uncontrolled and observational studies and RCTs without placebo, with interventions recommended for moderate-to-severe forms of AA. Category 3: mixed populations of adults and children; only or mostly adults with mild-to-moderate forms of AA or severity of hair loss at baseline that was not documented; interventions that did not fall into category 1 or 2 (i.e., they were not recommended in guidelines or were recommended in mild forms of AA or as adjunctive therapy only). aTwo of these records were abstracts that reported preliminary data from studies with full publications. These records were subsequently excluded to give a total of 53 references with extracted data (9 category 1 and 44 category 2). AA alopecia areata, HTA Health Technology Assessment, JAK Janus kinase, PRISMA Preferred Reporting Items for Systematic Reviews and Meta-Analyses, RCT randomized controlled trial

Full size image
Table 3 Summary of category 1 and 2 systemic corticosteroid studies included in the SLR
Full size table
Table 4 Summary of category 2 topical immunotherapy studies included in the SLR
Full size table
Table 5 Summary of category 1 and 2 cyclosporine A studies included in the SLR
Full size table
Table 6 Summary of category 2 methotrexate studies included in the SLR
Full size table
Table 7 Summary of category 2 azathioprine studies included in the SLR
Full size table
Table 8 Summary of category 1 and 2 JAK inhibitor studies included in the SLR
Full size table

Systemic Corticosteroids

Efficacy

One category 1 study was identified (Table 3) comparing 3 months of weekly oral prednisolone pulse therapy and placebo in 43 patients with severe AA [18]. Severe AA was defined as ≥ 40% scalp hair loss or > 10 patches scattered over the scalp and body. Results were reported only for the 36 study completers. After 3 months, 8 of 20 (40%) patients on corticosteroids achieved > 30% hair regrowth, compared with none of those receiving placebo (p < 0.03). Only two patients in the corticosteroid group achieved > 60% regrowth. At the end of a 3-month post-treatment follow-up, two of the eight responders had relapsed.

Four category 2 studies were identified. Treatment protocols varied across studies, and sample sizes ranged from 31 to 89 patients. In studies that included patients with AAs of varying severity, response rates to corticosteroid therapy were lower in those with extensive hair loss [20, 22]. In one study that reported data from 31 patients with alopecia totalis (AT) or universalis (AU), 22 (71%) achieved ≥ 75% scalp hair regrowth after 4 months of oral dexamethasone [21]; persistent response (defined as the maintenance of > 75% regrowth for at least 3 months after therapy withdrawal) was observed for 10 patients. Across studies, relapse rates after discontinuation of corticosteroids ranged from 25 to 100% [18, 20].

Tolerability

Corticosteroid therapy was associated with a higher rate of adverse events (AEs) versus placebo (55% vs. 13%) [18]. AEs reported with corticosteroids across the five studies most commonly included acne, weight gain, gastrointestinal symptoms, dysmenorrhea, striae, myalgia, and edema [18,19,20,21,22].

Topical Immunotherapy

Efficacy

No placebo-controlled studies were identified. Data were extracted from 13 category 2 studies mostly involving diphenylcyclopropenone (12 studies), with initial sample sizes varying from 10 to 64 [23,24,25,26,27,28,29,30,31,32,33,34,35]. Overall, response varied widely across the studies, with rates of complete response (where reported) ranging from 0 to 52% in study completers (Table 4). Baseline severity of disease and definition of response were not consistent across studies. Efficacy was lower in patients with AT/AU than in those with less severe scalp hair loss in one study [25], but this was not seen in another study [28]. Rates of relapse after treatment discontinuation ranged from 22 to 69% [24,25,26, 29, 32].

Tolerability

Notable AEs across studies included lymphadenopathy, contact dermatitis of the face or neck, generalized dermatitis, eczema, pruritus, hyperpigmentation, and urticaria (Table 4). The rate of AE-related discontinuations ranged from 6 to 25% in studies that reported these data [24, 26,27,28, 31, 33, 34].

Cyclosporine A

Efficacy

Two records reported the interim data [71] and final data [36] from a single RCT conducted in 36 adults with moderate-to-severe AA who received oral cyclosporine A 4 mg/kg/day or placebo for 3 months (Table 5). The final analysis performed on data from 32 patients found no statistically significant difference for the primary endpoint (50% improvement in Severity of Alopecia Tool score; SALT50) and most secondary endpoints after 3 months of treatment. Data from four category 2 studies were also extracted (Table 5). Sample sizes ranged from 6 to 88. Cyclosporine A was mostly used in monotherapy.

In a prospective study conducted in 15 patients with severe AA (> 50% scalp hair loss), five (33%) patients achieved ≥ 75% scalp hair regrowth over treatment durations ranging from 1 to 12 months [37]. In a retrospective study including 51 patients with mild to severe AA treated with cyclosporine A, 55% of patients achieved > 50% hair regrowth [39]. A reported relapse after treatment discontinuation occurred in 2 patients, with an additional 20 showing improved but sustained hair loss, from the total treated population of 51 patients (initial responses of individual patients were not described) in one study [39] and both responding patients in another [40].

Tolerability

In the category 1 study, 83% of patients in both the cyclosporine A and placebo groups reported at least one AE [36]. AEs reported at a higher rate with cyclosporine A than with placebo included urinary tract infection, musculoskeletal and respiratory disorders, gastrointestinal symptoms, hypertrichosis, hypertension, and hirsutism [36]. Similar AE profiles were reported across the category 2 studies (Table 5) [37,38,39,40].

Methotrexate

Efficacy

No placebo-controlled RCT was identified. One RCT was conducted in 36 patients with severe AA (SALT score 100%) randomized to receive oral methotrexate, oral betamethasone pulse therapy, or both combined [42]. At month 6 (end of treatment), the mean SALT score was 78% for methotrexate, 63% for betamethasone, and 62% for the combination; mean SALT score was unchanged 3 months after methotrexate treatment completion (Table 6). Across-group comparisons were not significant, and no patient achieved complete scalp hair regrowth. In the four remaining studies, methotrexate was mostly used in combination with systemic corticosteroids. Sample sizes ranged from 10 to 31, and, where provided, baseline severity ranged from > 50% to complete hair loss. Relapse rates ranged from 37 to 88% [41, 43,44,45]. Relapses were sometimes observed on treatment [41, 43], but mostly occurred when methotrexate and/or concomitant systemic corticosteroids were either tapered or discontinued [41, 43, 45].

Tolerability

Across studies, reported AEs included gastrointestinal complaints, leucopenia, and increases in transaminase activity (Table 6) [41,42,43,44,45].

Azathioprine

Efficacy

No placebo-controlled RCT was identified (Table 7). A prospective study provided data on 14 patients with AU treated with azathioprine. Six patients achieved ≥ 75% scalp hair regrowth and two relapsed after the withdrawal of azathioprine [46].

Tolerability

AEs were reported in five patients and included diarrhea, elevation of liver enzymes, pancreatitis, and bone marrow suppression (Table 7). Treatment was discontinued in four patients.

JAK Inhibitors

Efficacy

Seven category 1 records reporting data from five placebo-controlled RCTs were identified (Table 8).

King et al. [47] reported the outcomes of two phase 3 double-blind, parallel-group RCTs (BRAVE-AA1, N = 654 and BRAVE-AA2, N = 546). In each study, adults with severe AA (≥ 50% scalp hair loss) were randomized to receive baricitinib 4 mg or 2 mg or placebo. The primary outcome (proportion of patients achieving a SALT score ≤ 20 at week 36) was achieved by each dose in both trials (p < 0.001 for each dose versus placebo) (Table 8). Significant results were observed for most secondary outcomes for baricitinib 4 mg, including the achievement of eyebrow and eyelash hair regrowth. When baricitinib 2 mg was considered, response rates were generally lower than with baricitinib 4 mg.

Subgroup analyses of these two trials were provided in three congress reports. These showed that the efficacy of baricitinib was not notably affected by an atopic background [48], was achieved in patients with various degrees of scalp hair loss [49], and that benefit was obtained with continued treatment in patients without an initial response (SALT score > 20 at week 36) [50] (Table 8). Findings of the phase 2 portion of BRAVE-AA1, conducted in 110 adult patients with severe AA (≥ 50% scalp hair loss), support the phase 3 findings [51].

Another phase 2 randomized, placebo-controlled trial, ALLEGRO [52], assessed two unlicensed JAK inhibitors, ritlecitinib (a JAK3/tyrosine kinase expressed in hepatocellular carcinoma [TEC] family kinase irreversible inhibitor) and brepocitinib (an inhibitor of tyrosine kinase 2 and JAK1), for the treatment of patients with AA (≥ 50% scalp hair loss). Significantly better results were observed for both JAK inhibitors compared with placebo (both p < 0.0001) for the primary endpoint (change from baseline in SALT score at week 24). Numerical improvements versus placebo were also seen with both drugs across several additional outcome measures related to scalp, eyebrow, and eyelash regrowth.

King et al. [53] reported a phase 2 randomized, placebo-controlled trial of deuruxolitinib, an oral JAK1/2 inhibitor, in adults with moderate-to-severe AA (≥ 50% scalp hair loss). Patients were randomized to receive deuruxolitinib (4 mg, 8 mg, or 12 mg twice daily) or placebo for 24 weeks. There was a dose-dependent increase in the percentage of patients achieving SALT50 at week 24 (primary endpoint), with statistical significance versus placebo (p < 0.001) for the 8 mg and 12 mg groups (Table 8).

A total of 17 category 2 studies were identified. Sample size ranged from 6 to 90 patients and baseline severity of hair loss ranged from > 30% scalp hair loss to AT/AU. In these studies, JAK inhibitors were mostly used in monotherapy, with tofacitinib (a JAK1/3 inhibitor) the most studied.

Most patients experienced a relapse of hair loss after tofacitinib discontinuation in category 2 studies; data concerning relapse have not been reported for the other JAK inhibitors (Table 8).

Tolerability

AEs occurring at a higher frequency with JAK inhibitors than with placebo included upper respiratory tract infections, urinary tract infections, nasopharyngitis, acne, blood creatine phosphokinase and transaminase increases, headache, and nausea. Across studies, AEs were mostly mild or moderate in severity (Table 8).

Quality Assessment

Major concerns for potential sources of bias were identified, especially related to blinding, the distribution of patients, and intention to treat. Only the category 1 BRAVE-AA1 phase 2 [51]  and BRAVE-AA1/AA2 phase 3 studies of baricitinib [47], the phase 2 study with ritlecitinib and brepocitinib [52], and the Lai study [36] investigating the efficacy of cyclosporine A mentioned both the methods used for randomization and allocation concealment. Overall, except for these placebo-controlled trials, the studies from extracted records were considered to be of moderate-to-poor quality. Table 9 includes details of the quality assessment of the RCTs (placebo controlled and active controlled) identified and included in the SLR (see Table S3 for the quality assessment of the non-randomized included trials).

Table 9 Critical appraisal of the included randomized trials
Full size table

Discussion

Here, we summarize the available evidence pertaining to the clinical effects of treatments used in AA, with a focus on interventions recommended for use in adult patients with moderate-to-severe forms of AA. Among the 53 records retained for data extraction, only 9 reported data from 7 placebo-controlled RCTs (category 1). Five of these studies assessed one or several JAK inhibitors in adult patients with at least 50% scalp hair loss and demonstrated a consistent benefit over placebo across various clinical outcomes [47, 51,52,53]. The other two studies assessed oral prednisolone and cyclosporine A. Oral prednisolone pulse therapy was compared to placebo in a small monocentric trial, but the findings were not conclusive [18]. In the other trial, which was also small and monocentric, oral cyclosporine A was not superior to placebo after 3 months of therapy [36, 70]. Overall, the results in terms of hair regrowth varied widely across the remaining identified studies and were generally low for patients with moderate-to-severe AA. Relapses were commonly observed during treatment and upon discontinuation. AEs were generally consistent with the known safety profile of each intervention.

Our results align with those of a Cochrane review performed in 2008, which focused on RCTs evaluating the effectiveness of both topical and systemic interventions for AA, AT, and AU that were published prior to the introduction of JAK inhibitors [9]. The authors’ conclusions included that none of the interventions identified showed a significant treatment benefit in terms of hair growth when compared to placebo, that only a few treatments used for AA had been well evaluated in RCTs, and that most of these RCTs were small and had been reported poorly [9]. A more recent systematic review examining the effectiveness of systemic treatments for people with AA, AT, or AU also concluded that there was no specific systemic therapy that is supported by a robust body of evidence from RCTs [72].

A network meta-analysis (NMA) of RCTs comparing two or more AA treatments conducted in patients with AA of any severity and published prior to August 2019 included only three of the RCTs identified in the current SLR: the placebo-controlled RCTs by Kar et al. [18] for corticosteroids and Lai et al. [36] for cyclosporine A and the RCT comparing the JAK inhibitors ruxolitinib and tofacitinib (no placebo control) by Almutairi et al. [55]. In that NMA, the authors concluded that the highest treatment success rate was associated with pentoxifylline plus topical corticosteroids, followed by pentoxifylline alone [73]. Pentoxifylline is not currently recommended for the treatment of patients with AA; hence, only category 3 data were identified for this agent in the current SLR (see Table S2).

As part of the current SLR, we investigated whether the conduct of an NMA or indirect treatment comparison of the treatments used to manage adults with severe AA was possible using a feasibility assessment [74]. For the placebo-controlled trials, it would be statistically possible to conduct separate pairwise anchored indirect comparisons to compare the JAK inhibitor baricitinib with the oral immunosuppressant cyclosporine A and with the oral steroid prednisolone due to the use of a common placebo in the trials. However, such a comparison was not recommended due to several major limitations. The sample sizes in the cyclosporine and prednisolone trials were very small compared to the available data for baricitinib; hence, the results would be highly uncertain. In addition, complete data for potential prognostic factors and effect modifiers were not available. The remaining studies were uncontrolled, with small sample sizes and high heterogeneity between patient populations. Baseline characteristics and the timing of follow-up were not consistently reported, and outcome measures varied across studies. Unanchored indirect comparisons were not recommended due to the limited data and the strong assumptions that would be required for such an analysis [75]. An indirect comparison of individual JAK inhibitors was not performed as most are not available in clinical practice and the tofacitinib trials were predominantly non-randomized and/or retrospective.

Limitations/Strengths

The conclusions to be drawn regarding the efficacy and safety of treatments recommended for patients with moderate-to-severe AA are primarily limited by the excessive heterogeneity across the included records with respect to study design and population (severity, disease duration) and by the overall moderate-to-poor quality of the conduct and reporting of the studies.

The differing definitions for the severity of AA at baseline and the use of a wide range of response outcomes are also problematic. The SALT score is an objective, validated, and standardized tool recommended by experts for grading the extent and density of scalp hair loss in patients with AA in clinical trials [76, 77]. While SALT-based parameters were reported in the placebo-controlled RCTs identified for JAK inhibitors and cyclosporine A, most of the other studies used another method to assess percentage hair regrowth, often without providing any details on how the assessment was performed. This heterogeneity between studies prevented the conduct of a NMA of the available data.

Against these limitations, the strengths of our study should be taken into consideration. The broad scope of the SLR aimed to ensure the identification of the relevant evidence supporting the treatments for moderate-to-severe AA recommended in current guidelines. Additionally, the assessment of the quality of the included randomized and non-randomized studies using Cochrane and CRD checklists, respectively, means that the conclusions drawn can be considered robust.

Conclusion

The current management of patients with moderate-to-severe AA still relies on the use of treatments that have not been well evaluated in clinical trials. The most robust evidence identified supported the use of the EMA- and FDA-approved baricitinib, and other currently unapproved oral JAK inhibitors, in patients with severe AA.