Treatments for alopecia areata: a network meta‐analysis

Abstract

Background

Alopecia areata is an autoimmune disease leading to nonscarring hair loss on the scalp or body. There are different treatments including immunosuppressants, hair growth stimulants, and contact immunotherapy.

Objectives

To assess the benefits and harms of the treatments for alopecia areata (AA), alopecia totalis (AT), and alopecia universalis (AU) in children and adults.

Search methods

The Cochrane Skin Specialised Register, CENTRAL, MEDLINE, Embase, ClinicalTrials.gov and WHO ICTRP were searched up to July 2022.

Selection criteria

We included randomised controlled trials (RCTs) that evaluated classical immunosuppressants, biologics, small molecule inhibitors, contact immunotherapy, hair growth stimulants, and other therapies in paediatric and adult populations with AA.

Data collection and analysis

We used the standard procedures expected by Cochrane including assessment of risks of bias using RoB2 and the certainty of the evidence using GRADE. The primary outcomes were short‐term hair regrowth ≥ 75% (between 12 and 26 weeks of follow‐up), and incidence of serious adverse events. The secondary outcomes were long‐term hair regrowth ≥ 75% (greater than 26 weeks of follow‐up) and health‐related quality of life. We could not perform a network meta‐analysis as very few trials compared the same treatments. We presented direct comparisons and made a narrative description of the findings.

Main results

We included 63 studies that tested 47 different treatments in 4817 randomised participants. All trials used a parallel‐group design except one that used a cross‐over design. The mean sample size was 78 participants. All trials recruited outpatients from dermatology clinics. Participants were between 2 and 74 years old. The trials included patients with AA (n = 25), AT (n = 1), AU (n = 1), mixed cases (n = 31), and unclear types of alopecia (n = 4).

Thirty‐three out of 63 studies (52.3%) reported the proportion of participants achieving short‐term hair regrowth ≥ 75% (between 12 and 26 weeks). Forty‐seven studies (74.6%) reported serious adverse events and only one study (1.5%) reported health‐related quality of life. Five studies (7.9%) reported the proportion of participants with long‐term hair regrowth ≥ 75% (greater than 26 weeks).

Amongst the variety of interventions found, we prioritised some groups of interventions for their relevance to clinical practice: systemic therapies (classical immunosuppressants, biologics, and small molecule inhibitors), and local therapies (intralesional corticosteroids, topical small molecule inhibitors, contact immunotherapy, hair growth stimulants and cryotherapy).

Considering only the prioritised interventions, 14 studies from 12 comparisons reported short‐term hair regrowth ≥ 75% and 22 studies from 10 comparisons reported serious adverse events (18 reported zero events and 4 reported at least one). One study (1 comparison) reported quality of life, and two studies (1 comparison) reported long‐term hair regrowth ≥ 75%.

For the main outcome of short‐term hair regrowth ≥ 75%, the evidence is very uncertain about the effect of oral prednisolone or cyclosporine versus placebo (RR 4.68, 95% CI 0.57 to 38.27; 79 participants; 2 studies; very low‐certainty evidence), intralesional betamethasone or triamcinolone versus placebo (RR 13.84, 95% CI 0.87 to 219.76; 231 participants; 1 study; very low‐certainty evidence), oral ruxolitinib versus oral tofacitinib (RR 1.08, 95% CI 0.77 to 1.52; 80 participants; 1 study; very low‐certainty evidence), diphencyprone or squaric acid dibutil ester versus placebo (RR 1.16, 95% CI 0.79 to 1.71; 99 participants; 1 study; very‐low‐certainty evidence), diphencyprone or squaric acid dibutyl ester versus topical minoxidil (RR 1.16, 95% CI 0.79 to 1.71; 99 participants; 1 study; very low‐certainty evidence), diphencyprone plus topical minoxidil versus diphencyprone (RR 0.67, 95% CI 0.13 to 3.44; 30 participants; 1 study; very low‐certainty evidence), topical minoxidil 1% and 2% versus placebo (RR 2.31, 95% CI 1.34 to 3.96; 202 participants; 2 studies; very low‐certainty evidence) and cryotherapy versus fractional CO2 laser (RR 0.31, 95% CI 0.11 to 0.86; 80 participants; 1 study; very low‐certainty evidence). The evidence suggests oral betamethasone may increase short‐term hair regrowth ≥ 75% compared to prednisolone or azathioprine (RR 1.67, 95% CI 0.96 to 2.88; 80 participants; 2 studies; low‐certainty evidence). There may be little to no difference between subcutaneous dupilumab and placebo in short‐term hair regrowth ≥ 75% (RR 3.59, 95% CI 0.19 to 66.22; 60 participants; 1 study; low‐certainty evidence) as well as between topical ruxolitinib and placebo (RR 5.00, 95% CI 0.25 to 100.89; 78 participants; 1 study; low‐certainty evidence). However, baricitinib results in an increase in short‐term hair regrowth ≥ 75% when compared to placebo (RR 7.54, 95% CI 3.90 to 14.58; 1200 participants; 2 studies; high‐certainty evidence).

For the incidence of serious adverse events, the evidence is very uncertain about the effect of topical ruxolitinib versus placebo (RR 0.33, 95% CI 0.01 to 7.94; 78 participants; 1 study; very low‐certainty evidence). Baricitinib and apremilast may result in little to no difference in the incidence of serious adverse events versus placebo (RR 1.47, 95% CI 0.60 to 3.60; 1224 participants; 3 studies; low‐certainty evidence). The same result is observed for subcutaneous dupilumab compared to placebo (RR 1.54, 95% CI 0.07 to 36.11; 60 participants; 1 study; low‐certainty evidence).

For health‐related quality of life, the evidence is very uncertain about the effect of oral cyclosporine compared to placebo (MD 0.01, 95% CI ‐0.04 to 0.07; very low‐certainty evidence).

Baricitinib results in an increase in long‐term hair regrowth ≥ 75% compared to placebo (RR 8.49, 95% CI 4.70 to 15.34; 1200 participants; 2 studies; high‐certainty evidence).

Regarding the risk of bias, the most relevant issues were the lack of details about randomisation and allocation concealment, the limited efforts to keep patients and assessors unaware of the assigned intervention, and losses to follow‐up.

Authors' conclusions

We found that treatment with baricitinib results in an increase in short‐ and long‐term hair regrowth compared to placebo. Although we found inconclusive results for the risk of serious adverse effects with baricitinib, the reported small incidence of serious adverse events in the baricitinib arm should be balanced with the expected benefits. We also found that the impact of other treatments on hair regrowth is very uncertain. Evidence for health‐related quality of life is still scant.

Plain language summary

What are the benefits and risks of different treatments for alopecia areata (hair loss)?

Key messages

‐There are multiple treatment options for alopecia areata as systemic therapies such as immunosuppressants and local treatments such as hair growth stimulants, but it is unclear how helpful they are in producing new hair growth in the affected area.

‐Only one oral immunosuppressant treatment, baricitinib, showed an increment in hair regrowth.

‐Therapies appear to be safe and serious side effects are rare.

‐More good‐quality studies are required to evaluate the therapies that appear to be useful.

What is alopecia areata?

Alopecia areata is a common condition characterised by localised or diffuse hair loss on the scalp or around the body. Almost half of the patients have new hair growth without treatment, but a considerable number of them still require specific treatment.

How is alopecia areata treated?

Most patients get better spontaneously and, in some cases, the doctor and patient may choose to wait for new hair to grow. However, spontaneous growth is rare in severe cases. For patients who need medication, there are multiple treatments, including topical therapies, oral treatments, and localised corticosteroid injections.

Why did we do this Cochrane Review?

Due to the broad landscape of treatments for alopecia areata, we wanted to know the potential benefits and harms of the available treatments and to see if some of them work better than others.

What did we do?

We searched for studies that tested therapies to treat alopecia areata (AA), alopecia totalis (AT) and alopecia universalis (AU). The therapies we searched for included medicines administered orally or subcutaneously with systemic effects (that affect the body as a whole) such as immunosuppressants, and local treatments (that affect the skin surface) such as hair growth stimulants or cryotherapy. These treatments were compared with a placebo (a 'dummy' treatment that does not contain any drug) or with another medicine. We assessed the effect of the treatments on significant hair regrowth (equal or greater than 75% and measured from 12 to 26 weeks after starting the treatment or beyond 26 weeks), on well‐being (quality of life), and if they are likely to cause serious adverse events (unwanted or harmful effects).

We also summarised the results of the studies and rated our confidence in the evidence, based on factors such as study methods and number of participants.

How up‐to‐date is this review?

We included evidence up to July 2022.

What did we find?

We found 63 studies that tested 47 different therapies in 4817 people with AA, AT or AU (participants were between 2 and 74 years old).

Within the 12 comparisons (one comparison means one treatment compared to another) considered more relevant to clinical practice, short‐term hair regrowth ≥ 75% (from 12 to 26 weeks of follow‐up) was evaluated in 14 studies and long‐term regrowth ≥ 75% (after 26 weeks of follow‐up) was evaluated in two studies. Only one study assessed well‐being (quality of life) and 22 studies assessed serious adverse events (although only four of them reported at least one serious adverse event).

What are the main results of our review?

Baricitinib (an immunosuppressant medicine taken by mouth) compared to placebo increases hair regrowth ≥ 75% (both short and long‐term) and we are confident with these results. The evidence is very uncertain about the effect of topical hair growth stimulants (minoxidil) on short‐term hair regrowth ≥ 75% compared to placebo as the results largely varied from study to study. It is unclear if other treatments compared to placebo or to other medicines have an effect on hair regrowth, and we are not confident of the evidence because the results from the studies varied widely, and involved small numbers of people. Some studies did not clearly report how they were conducted, or whether the people taking part knew what they had received which could have affected the study's results. Further research is likely to change our results.

Four studies that assessed the immunosuppressant drugs dupilumab (administered subcutaneously), baricitinib, and topical ruxolitinib reported 30 serious adverse events (23 in the treatment group and 7 in the placebo group) in a total of 1332 participants. In all cases, the authors did not establish a direct relationship between serious adverse events and therapies. There were no deaths, serious infections, or cancer cases. Based on the available evidence, those treatments may have little to no effect on serious adverse events.

It is unclear if oral cyclosporine (an immunosuppressant drug) has an effect on the quality of life.

Limitations of the evidence

Our confidence in the evidence is only high for one comparison (baricitinib compared to placebo) and, for the remaining identified evidence, our confidence is generally low since most of the therapies have been evaluated in studies with weaknesses in their design, with few patients, and that have not been replicated to evaluate the consistency of the results.

Summary of findings

Summary of findings 1. Summary of findings table ‐ Oral classical immunosuppressants (prednisolone or cyclosporine) compared to placebo for adults with alopecia areata.

Oral classical immunosuppressants (prednisolone or cyclosporine) compared to placebo for adults with alopecia areata
Patient or population: adults with alopecia areata Setting: outpatients Intervention: oral classical immunosuppressants (prednisolone or cyclosporine) Comparison: placebo
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with placebo	Risk with oral classical immunosuppressants (prednisolone or cyclosporine)
Short‐term hair regrowth ≥ 75% (between 12 and 26 weeks of follow‐up)	Low		RR 4.68 (0.57 to 38.27)	79 (2 RCTs)	⊕⊝⊝⊝ Very lowa,b	No events were observed in the control group (placebo). Thus, two scenarios for baseline risk were simulated to calculate absolute effect of the intervention.
	1 per 1000	5 per 1000 (1 to 38)
	Moderate
	10 per 1000	47 per 1000 (6 to 383)
Incidence of serious adverse events	0 per 100	0 per 100 (0 to 0)	Not estimable	79 (2 RCTs)	⊕⊝⊝⊝ Very lowa,c	The evidence is very uncertain about the effect of oral classical immunosuppressants (prednisolone or cyclosporine) on incidence of serious adverse events. Zero events have been observed in all trialsc
Health‐related quality of life (AQoL‐8D) assessed with: AQoL‐8D; scale from 0 (death) to 1 (full health) follow‐up: mean 12 weeks	The mean health‐related quality of life was 0.05	MD 0.01 higher (0.04 lower to 0.07 higher)	‐	36 (1 RCT)	⊕⊝⊝⊝ Very lowa,b
Long‐term hair regrowth ≥ 75% (greater than 26 weeks of follow‐up) ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
See interactive version of this table: https://gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_437439792522728604.

^a Downgraded by one level due to serious risk of bias: lack of information on randomisation, outcome assessors and patients could be aware of the intervention received. Reasons for patients who dropped the study were not reported. In addition, there was no pre‐registered protocol.
^b Downgraded by two levels for very serious imprecision: small sample size below the OIS, and null value of the effect included in the confidence interval.
^c Downgraded by two levels for serious imprecision: zero events have been observed in the trials.

Summary of findings 2. Summary of findings table ‐ Oral classical immunosuppressants (betamethasone) compared to oral classical immunosuppressants (azathioprine, methotrexate, prednisolone) for adults with alopecia areata.

Oral classical immunosuppressants (betamethasone) compared to oral classical immunosuppressants (azathioprine, methotrexate, prednisolone) for adults with alopecia areata
Patient or population: adults with alopecia areata Setting: outpatients Intervention: oral classical immunosuppressants (betamethasone) Comparison: oral classical immunosuppressants (azathioprine, methotrexate or prednisolone)
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with oral classical immunosuppressants (azathioprine, methotrexate or prednisolone)	Risk with oral classical immunosuppressants (betamethasone)
Short‐term hair regrowth ≥ 75% (between 12 and 26 weeks of follow‐up)	250 per 1000	418 per 1000 (240 to 720)	RR 1.67 (0.96 to 2.88)	80 (2 RCTs)	⊕⊕⊝⊝ Lowa,b
Incidence of serious adverse events	0 per 100	0 per 100 (0 to 0)	Not estimable	154 (4 RCTs)	⊕⊝⊝⊝ Very lowa,c	The evidence is very uncertain about the effect of oral classical immunosuppressants (betamethasone compared to azathioprine, methotrexate or prednisolone) on the incidence of serious adverse events. Zero events have been observed in all trials.c
Health‐related quality of life ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
Long‐term hair regrowth ≥ 75% (greater than 26 weeks of follow‐up) ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
See interactive version of this table: https://gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_440406708983187840.

^a No protocol is available. Lack of information on the allocation concealment, blinding of participants or study personnel. The percentage of losses could influence the result. 
^b Downgraded by two levels for very serious imprecision. Null value of the effect included in the confidence interval.
^c Downgraded by two levels for very serious imprecision. Zero events have been observed in all trials.

Summary of findings 3. Summary of findings table ‐ Intralesional classical immunosuppresants (betamethasone or triamcinolone) compared to placebo for adults with alopecia areata.

Intralesional classical immunosuppresants (betamethasone or triamcinolone) compared to placebo for adults with alopecia areata
Patient or population: adults with alopecia areata Setting: outpatients Intervention: intralesional classical immunosuppresants (betamethasone or triamcinolone) Comparison: placebo
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with placebo	Risk with intralesional classical immunosuppresants (betamethasone or triamcinolone)
Short‐term hair regrowth ≥ 75% (between 12 and 26 weeks of follow‐up)	Low		RR 13.84 (0.87 to 219.76)	231 (1 RCT)	⊕⊝⊝⊝ Very lowa,b	No events were observed in the control group (placebo). Thus, two scenarios for baseline risk were simulated to calculate the absolute effect of the intervention.
	1 per 1000	14 per 1000 (1 to 220)
	Moderate
	10 per 1000	138 per 1000 (9 to 1000)
Incidence of serious adverse events	0 per 100	0 per 100 (0 to 0)	Not estimable	543 (4 RCTs)	⊕⊝⊝⊝ Very lowa,b,c	The evidence is very uncertain about the effect of oral classical immunosuppressants (betamethasone or triamcinolone) on the incidence of serious adverse events. Zero events have been observed in all trials.
Health‐related quality of life ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
Long‐term hair regrowth ≥ 75% (greater than 26 weeks of follow‐up) ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
See interactive version of this table: https://gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_437441152971971919.

^a Downgraded by one level due to no information of allocation concealment, no information to determine if participants and people delivering the intervention were aware of the assigned intervention, and no pre‐registered protocol or statistical plan.
^b Downgraded by two levels for very serious imprecision: small sample size below the OIS and the confidence interval includes the null value with appreciable benefits and harms.
^c Downgraded by two levels for very serious imprecision. Zero events have been observed in all trials.

Summary of findings 4. Summary of findings table ‐ Subcutaneous biologics (dupilumab) compared to placebo for adults with alopecia areata.

Subcutaneous biologics (dupilumab) compared to placebo for adults with alopecia areata
Patient or population: adults with alopecia areata Setting: outpatients Intervention: subcutaneous biologics (dupilumab) Comparison: placebo
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with placebo	Risk with subcutaneous biologics (dupilumab)
Short‐term hair regrowth ≥ 75%	Low		RR 3.59 (0.19 to 66.22)	60 (1 RCT)	⊕⊕⊝⊝ Lowa	No events were observed in the control group (placebo). Thus, two scenarios for baseline risk were simulated to calculate the absolute effect of the intervention.
	1 per 1000	4 per 1000 (0 to 66)
	Moderate
	10 per 1000	36 per 1000 (2 to 662)
Incidence of serious adverse events	0 per 1000	0 per 1000 (0 to 0)	RR 1.54 (0.07 to 36.11)	60 (1 RCT)	⊕⊕⊝⊝ Lowa
Health‐related quality of life ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
Long‐term hair regrowth ≥ 75% (greater than 26 weeks of follow‐up) ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
See interactive version of this table: https://gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_437441343526543105.

^a Downgraded by two levels for very serious imprecision: small sample size below the OIS and there were very few events, so the confidence interval includes appreciable benefits and harms.

Summary of findings 5. Summary of findings table ‐ Oral small molecule inhibitors (apremilast or baricitinib) compared to placebo for adults with alopecia areata.

Oral small molecule inhibitors (apremilast or baricitinib) compared to placebo for adults with alopecia areata
Patient or population: adults with alopecia areata Setting: outpatients Intervention: oral small molecule inhibitors (apremilast or baricitinib) Comparison: placebo
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with placebo	Risk with oral small molecule inhibitors (apremilast or baricitinib)
Short‐term hair regrowth ≥ 75% (between 12 and 26 weeks of follow‐up)	26 per 1000	197 per 1000 (102 to 380)	RR 7.54 (3.90 to 14.58)	1200 (2 RCTs)	⊕⊕⊕⊕ High
Incidence of serious adverse events	17 per 1000	25 per 1000 (10 to 61)	RR 1.47 (0.60 to 3.60)	1224 (3 RCTs)	⊕⊕⊝⊝ Lowa
Health‐related quality of life ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
Long‐term hair regrowth ≥ 75% (greater than 26 weeks of follow‐up)	32 per 1000	271 per 1000 (150 to 489)	RR 8.49 (4.70 to 15.34)	1200 (2 RCTs)	⊕⊕⊕⊕ High
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
See interactive version of this table: https://gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_437441426952298244.

^a Downgraded by two levels for very serious imprecision: null value of the effect included in the confidence interval.

Summary of findings 6. Summary of findings table ‐ Oral small molecule inhibitors (ruxolitinib) compared to oral small molecule inhibitors (tofacitinib) for adults with alopecia areata.

Oral small molecule inhibitors (ruxolitinib) compared to oral small molecule inhibitors (tofacitinib) for adults with alopecia areata
Patient or population: adults with alopecia areata Setting: outpatients Intervention: oral small molecule inhibitors (ruxolitinib) Comparison: oral small molecule inhibitors (tofacitinib)
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with oral small molecule inhibitors (tofacitinib)	Risk with oral small molecule inhibitors (ruxolitinib)
Short‐term hair regrowth ≥ 75% (between 12 and 26 weeks of follow‐up)	600 per 1000	648 per 1000 (462 to 912)	RR 1.08 (0.77 to 1.52)	80 (1 RCT)	⊕⊝⊝⊝ Very lowa,b
Incidence of serious adverse events	0 per 100	0 per 100 (0 to 0)	Not estimable	80 (1 RCT)	⊕⊝⊝⊝ Very lowc,d,e	The evidence is very uncertain about the effect of oral small molecule inhibitors (ruxolitinib compared to tofacitinib) on the incidence of serious adverse events. Zero events have been observed in all trials.
Health‐related quality of life ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
Long‐term hair regrowth ≥ 75% (greater than 26 weeks of follow‐up) ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
See interactive version of this table: https://gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_440407039166139262.

^a No study protocol is available. Lack of information on the method of randomisation. Participants and study personnel were aware of the intended interventions. Non‐adherent participants were withdrawn from the analysis.
^b Downgraded by one level for serious imprecision: small sample size below the OIS.
^c No study protocol is available. Lack of information on the method of randomisation. Participants and study personnel were aware of the intended interventions. Non‐adherent participants were withdrawn from the analysis. However, the authors reported any side effects and serious AE could be recorded independently of the knowledge of the participant allocation.
^d Downgraded by two levels for very serious imprecision: small sample size below the OIS, and null value of the effect included in the confidence interval.
^e Downgraded by two levels for very serious imprecision. Zero events have been observed in all trials.

Summary of findings 7. Summary of findings table ‐ Topical small molecule inhibitors (ruxolitinib) compared to placebo for adults with alopecia areata.

Topical small molecule inhibitors (ruxolitinib) compared to placebo for adults with alopecia areata
Patient or population: adults with alopecia areata Setting: outpatients Intervention: topical small molecule inhibitors (ruxolitinib) Comparison: placebo
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with placebo	Risk with topical small molecule inhibitors (ruxolitinib)
Short‐term hair regrowth ≥ 75% (between 12 and 26 weeks of follow‐up)	0 per 100	0 per 100 (0 to 0)	RR 5.00 (0.25 to 100.89)	78 (1 RCT)	⊕⊕⊝⊝ Lowa
Incidence of serious adverse events	0 per 100	0 per 100 (0 to 0)	RR 0.33 (0.01 to 7.94)	78 (1 RCT)	⊕⊝⊝⊝ Very lowa,b
Health‐related quality of life ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
Long‐term hair regrowth ≥ 75% (greater than 26 weeks of follow‐up) ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
See interactive version of this table: https://gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_440407113330347352.

^a Downgraded by two levels for very serious imprecision: small sample size below the OIS, and null value of the effect included in the confidence interval.
^b Some concerns in 2 domains due to the lack of detail on randomisation and allocation concealment, and lack of information on the method to assess adverse events in the treatment arms.

Summary of findings 8. Summary of findings table ‐ Contact immunotherapy (diphencyprone or squaric acid dibutil ester) compared to placebo for adults with alopecia areata.

Contact immunotherapy (diphencyprone or squaric acid dibutil ester) compared to placebo for adults with alopecia areata
Patient or population: adults with alopecia areata Setting: outpatients Intervention: contact immunotherapy (diphencyprone or squaric acid dibutil ester) Comparison: placebo
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with placebo	Risk with contact immunotherapy (diphencyprone or squaric acid dibutil ester)
Short‐term hair regrowth ≥ 75% (between 12 and 26 weeks of follow‐up)	600 per 1000	696 per 1000 (474 to 1000)	RR 1.16 (0.79 to 1.71)	99 (1 RCT)	⊕⊝⊝⊝ Very lowa,b
Incidence of serious adverse events ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
Health‐related quality of life ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
Long‐term hair regrowth ≥ 75% (greater than 26 weeks of follow‐up) ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
See interactive version of this table: https://gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_437441787556798444.

^a Downgraded by two levels due to a very serious risk of bias due to a lack of information about the randomisation process, deviations from intended interventions, missing outcome data and blinding of the outcome assessor.
^b Downgraded by one level of imprecision because of the null value of the effect was included in the confidence interval.

Summary of findings 9. Summary of findings table ‐ Contact immunotherapy (diphencyprone or squaric acid dibutil ester) compared to hair growth stimulants (minoxidil) for adults with alopecia areata.

Contact immunotherapy (diphencyprone or squaric acid dibutil ester) compared to hair growth stimulants (minoxidil) for adults with alopecia areata
Patient or population: adults with alopecia areata Setting: outpatients Intervention: contact immunotherapy (diphencyprone or squaric acid dibutyl ester) Comparison: hair growth stimulants (minoxidil)
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with hair growth stimulants (minoxidil)	Risk with contact immunotherapy (diphencyprone or squaric acid dibutyl ester)
Short‐term hair regrowth ≥ 75% (between 12 and 26 weeks of follow‐up)	600 per 1000	696 per 1000 (474 to 1000)	RR 1.16 (0.79 to 1.71)	99 (1 RCT)	⊕⊝⊝⊝ Very lowa,b
Incidence of serious adverse events ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
Health‐related quality of life ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
Long‐term hair regrowth ≥ 75% (greater than 26 weeks of follow‐up) ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
See interactive version of this table: https://gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_437685345154162408.

^a Downgraded by two levels due to a very serious risk of bias due to a lack of information about the randomisation process, deviations from intended interventions, missing outcome data and blinding of the outcome assessor.
^b Downgraded by one level of imprecision because the null value of the effect was included in the confidence interval.

Summary of findings 10. Summary of findings table ‐ Contact immunotherapy (diphencyprone) + hair growth stimulants (minoxidil) compared to contact immunotherapy (diphencyprone) for adults with alopecia areata.

Contact immunotherapy (diphencyprone) + hair growth stimulants (minoxidil) compared to contact immunotherapy (diphencyprone) for adults with alopecia areata
Patient or population: adults with alopecia areata Setting: outpatients Intervention: contact immunotherapy (diphencyprone) + hair growth stimulants (minoxidil) Comparison: contact immunotherapy (diphencyprone)
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with contact immunotherapy (diphencyprone)	Risk with contact immunotherapy (diphencyprone) + hair growth stimulants (minoxidil)
Short‐term hair regrowth ≥ 75% (between 12 and 26 weeks of follow‐up)	200 per 1000	134 per 1000 (26 to 688)	RR 0.67 (0.13 to 3.44)	30 (1 RCT)	⊕⊝⊝⊝ Very lowa,b
Incidence of serious adverse events	0 per 100	0 per 100 (0 to 0)	Not estimable	30 (1 RCT)	⊕⊝⊝⊝ Very lowb,c,d	The evidence is very uncertain about the effect of contact immunotherapy (diphencyprone) plus hair growth stimulants (minoxidil) compared to contact immunotherapy (diphencyprone) on the incidence of serious adverse events. Zero events have been observed in all trials.
Health‐related quality of life ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
Long‐term hair regrowth ≥ 75% (greater than 26 weeks of follow‐up) ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
See interactive version of this table: https://gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_437441982448018086.

^a Downgraded by one level due to the lack of information on randomisation, baseline imbalances, lack of information on whether outcome assessors were aware of the intervention received, and missing outcome data without appropriate analysis. In addition, there was no pre‐registered protocol available.
^b Downgraded by two levels for very serious imprecision: small sample size below the OIS, and the confidence interval includes the null effect and ranged from appreciable benefits to great harms.
^c There were no details on the method of measuring the outcome, but according to the information provided, it is understood that the same outcome was evaluated in both intervention groups applying the same methodology. There was no information on whether the outcome assessors were aware of the intervention received, but knowledge of the intervention received could have affected the measurement of the results.
^d Downgraded by two levels for very serious imprecision. Zero events have been observed in all trials.

Summary of findings 11. Summary of findings table ‐ Topical hair growth stimulants (minoxidil) compared to placebo for adults with alopecia areata.

Topical hair growth stimulants (minoxidil) compared to placebo for adults with alopecia areata
Patient or population: adults with alopecia areata Setting: outpatients Intervention: topical hair growth stimulants (minoxidil) Comparison: placebo
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with placebo	Risk with topical hair growth stimulants (minoxidil)
Short‐term hair regrowth ≥ 75% (between 12 and 26 weeks of follow‐up)	175 per 1000	404 per 1000 (235 to 693)	RR 2.31 (1.34 to 3.96)	202 (2 RCTs)	⊕⊝⊝⊝ Very lowa,b
Incidence of serious adverse events	0 per 100	0 per 100 (0 to 0)	Not estimable	304 (4 RCTs)	⊕⊝⊝⊝ Very lowc,d,e	The evidence is very uncertain about the effect of topical hair growth stimulants (minoxidil) on the incidence of serious adverse events. Zero events have been observed in all trials.
Health‐related quality of life ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
Long‐term hair regrowth ≥ 75% (greater than 26 weeks of follow‐up) ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
See interactive version of this table: https://gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_437441625584312094.

^a Downgraded by two levels due to lack of information about the randomisation process, missing outcome data and blinding of the outcome assessor.
^b The trials showed highly heterogeneous results (I‐squared = 93%).
^c Downgraded by two levels due to lack of information about the randomisation process, blinding of outcome assessors, reasons for missing outcome data and outcome measurement and plan analysis. 
^d Downgraded by two levels for very serious imprecision: small sample size below the OIS, and null value of the effect included in the confidence interval.
^e Downgraded by two levels for very serious imprecision. Zero events have been observed in all trials

Summary of findings 12. Summary of findings table ‐ Other therapies (cryotherapy) compared to other therapies (fractional CO2 laser) for adults with alopecia areata.

Other therapies (cryotherapy) compared to other therapies (fractional CO2 laser) for adults with alopecia areata
Patient or population: adults with alopecia areata Setting: outpatients Intervention: other therapies (cryotherapy) Comparison: other therapies (fractional CO2 laser)
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with other therapies (fractional CO2 laser)	Risk with other therapies (cryotherapy)
Short‐term hair regrowth ≥ 75%	325 per 1000	101 per 1000 (36 to 280)	RR 0.31 (0.11 to 0.86)	80 (1 RCT)	⊕⊝⊝⊝ Very lowa,b
Incidence of serious adverse events	0 per 100	0 per 100 (0 to 0)	Not estimable	80 (1 RCT)	⊕⊝⊝⊝ Very lowa,c	The evidence is very uncertain about the effect of other therapies (cryotherapy compared to fractional C02 laser) on the incidence of serious adverse events. Zero events have been observed in all trials.
Health‐related quality of life ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
Long‐term hair regrowth ≥ 75% ‐ not reported	‐	‐	‐	‐	‐	This outcome was not reported.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; OR: odds ratio; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
See interactive version of this table: https://gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_440742870117343721.

^a Insufficient information about the randomisation process. Participants and carers were aware of the assigned interventions and protocol deviations were not reported. Insufficient information for measuring of the outcome. Reporting could have been influenced by the knowledge of the intended intervention.
^b Downgraded by one level for serious imprecision: small sample size below the OIS.
^c Downgraded by two levels for very serious imprecision. Zero events have been observed in all trials.

Background

A glossary of terms is provided in Table 13.

1. Glossary of terms.

Term	Definition
Adjuvant therapy	Additional treatment given to the primary or initial therapy to maximise its effectiveness
Alopecia areata (AA)	A common, inflammatory, nonscarring type of hair loss, characterised by small patches of baldness on the scalp or around the body, as well as periods of relapse or remission
Alopecia totalis (AT)	Alopecia that affects all scalp hairs
Alopecia universalis (AU)	Alopecia that involves all scalp and body hairs
Antigen	A substance that invokes an immune response
Biologic agents	Compounds produced by living cells using recombinant DNA technology
Blistering	Circumscribed elevations of the epidermis, fluid‐filled due to separation of two layers of tissue and the leakage of plasma into the space
Catagen phase	A phase of 2 to 3 weeks of duration when growth stops and the follicle shrinks
Chemokines	Molecular messengers through which epithelial cells communicate with key cells in the immune system
Coudability hairs	Normal‐looking hairs tapered at the proximal end (i.e. the hair root)
Cryotherapy	Tissue destruction techniques produced by the application of cold on the skin
Cutaneous atrophy	Skin thinning
Dendritic Cells	The major antigen‐presenting cells; these cells capture, process, and present antigens to T cells in order to induce adaptive immunity or tolerance to self‐antigens
Downregulation	The process by which a cell reduces or suppresses the quantity of a cellular component
Exclamation mark hairs	Short hairs, 3 mm long, with irregular thickening and terminal dilation
Exudate	A fluid rich in protein and cellular elements that results from a continuous inflammatory response
First choice	Treatment regimen accepted for the primary or initial therapy
Hair follicle	A small cavity in the epidermis, from which a hair develops
Hepatotoxicity	The result of chemical‐driven liver damage
Immune privilege	Multiple mechanisms that prevent autologous attack by immune cells in certain locations, such as the hair follicle
Immunosuppressant therapy	Treatment that reduces the activity of the body's immune system
Incidence	The rate of new (or newly diagnosed) cases of the disease
Interleukins	Potent cytokines (proteins) produced by some leukocytes (white blood cells), which function as mediators of cell growth, inflammation, immunity, differentiation and repair
Keratins	A fibrous protein that occurs in the outer layer of the skin and in the hair and nails
Langerhans cells	Dendritic cells (which regulate the cellular immune response) present in all layers of the epidermis as a dense network of immune system indicators
Laser therapy	A medical treatment where a light source is used; radiant energy in the form of photons and waves is capable of producing special biological effects
Lymphocytes	Leukocyte whose main function is the regulation of the adaptive immune response
Macrophages	Cells derived from monocytes residing in various tissues; their function is to present the antigens to the lymphocytes to initiate the immune response and destroy, by phagocytosis (the process by which a cell binds to and engulfs something) the antigens and the cells that transport them
Monocytes	Circulating blood cells whose main function is phagocytosis
Patchy alopecia (PA)	Round or oval patches of alopecia located on the head or in different regions of the body
Phototherapy	Therapeutic use of ultraviolet radiation
Prevalence	Proportion of individuals from a population that present with the event in a given period of time
Pruritus	Feeling that leads to scratching
Psoralens	Photoactive medication
PUVA	Administration of a psoralen and subsequent exposure to ultraviolet radiation A
Regrowth	Reappearance or return
T cells	Cells that are programmed to recognise, respond to, and remember antigens
Tumour Necrosis Factor alpha (TNF alpha)	An inflammatory cytokine produced by macrophages/monocytes responsible for different signalling events within cells, leading to necrosis (cell death) or apoptosis (programmed cell death)
Upregulation	The process by which a cell increases the quantity of a cellular component
Vehicle	A treatment without active ingredients

DNA: deoxyribonucleic acid
TNF: tumour necrosis factor

Description of the condition

Alopecia areata is an autoimmune disease leading to nonscarring hair loss on the scalp or body, as well as periods of relapse or remission, affecting the quality of life of both patients and their caregivers (Hunt 2005; Villasante 2015). AA primarily affects the scalp; however, it also can affect nails, eyelashes, eyebrows, and other hair follicles in the patient's body (Strazzulla 2018b).

There are three types of alopecia areata, defined according to the affected area: alopecia areata (AA), which involves a partial loss of scalp hair and comprises 90% of clinical diagnoses; alopecia totalis (AT), with a total loss of scalp hair; and alopecia universalis (AU), involving all scalp and body hair (Islam 2015). Please, note that the term alopecia areata (AA) includes all types of alopecia areata (patchy), alopecia totalis (AT), and alopecia universalis (AU). In addition, the term AA can refer to only patchy alopecia areata as this condition is the most prevalent.

The cause of AA is still being researched but, currently, genetics and immune function seem to be the most relevant contributors to the disease. Recently, some studies have suggested that the destruction of the hair follicle generated by an immune disorder could play a role in the appearance of AA (Juarez‐Rendon 2017; Paus 2018; Pratt 2017; Strazzulla 2018b).

The lifetime incidence reported for the US population ranges between 1.7% and 2.1%, and the prevalence varies between populations and studies, ranging from 0.1% to 0.2% in the general population (Mirzoyev 2014; Safavi 1992; Safavi 1995). The pooled prevalence estimates are 0.08% for AT and 0.03% for AU, and the overall incidence proportion is 3.37% for AT and 0.02% for AU (Lee 2019; Lee 2020). There are no clear differences by race or gender, however, almost 20% of patients have a family history of AA. Of all cases with AA, 20% occur in childhood. The prevalence of AA in children and adolescents is 1.83%; between 10% to 51.6% of those with AA have a family history of the condition and 40.2% present their first episode of AA by 20 years old (Lee 2019; Korta 2018; Pratt 2017). The extent of hair loss is considered the most important prognostic factor; those with greater hair loss respond less to treatment and have a greater likelihood of progressing to chronic disease. Other factors related to a less favourable prognosis are early age, atopic dermatitis, autoimmune diseases, and nail changes (Lee 2017). Involvement of the nails occurs in 10% to 66% of all cases of AA and is present in 15.4% of AU cases (more frequent in severe forms of AA) (Ferreira 2016).

There is a strong association between AA and other autoimmune diseases; thyroid disorders are the most common accompanying conditions, with a prevalence of 19% (Islam 2015). Other diseases commonly associated with AA include lupus erythematous, atopic dermatitis, and psychiatric diseases (Conic 2017; Lee 2019).

The psychological and social effects of AA impact substantially on patients' health‐related quality of life. Compared with the general population, people with AA have an increased risk in three out of the four mental health domains of the SF‐36 physical and mental health summary scales ("role‐emotional, mental health, vitality") (Rencz 2016). Psychiatric disorders can trigger the onset of AA and the role of psychological stress in people with AA may be related to psychiatric comorbidities including anxiety, depression, social phobia, and personality disorders. Young male patients are at greater risk for psychological distress and suicide and require careful monitoring (Rencz 2016).

The diagnosis of AA is based on typical clinical presentation (acute alopecia in well‐circumscribed patches of normal‐appearing skin) (Gilhar 2012).

In the natural history of the disease, hair regrowth occurs in 34% to 50% of patients within one year, and 15% to 25% will progress to AT. The long‐term prognosis is directly associated with the severity of AA (Bernardis 2018; Hammerschmidt 2014). It has been a challenge for researchers and clinicians to quantify AA in real‐time. The Severity of Alopecia Tool Scoring (SALT) is a quantitative, reproducible, standardised and simple system that allows a clinical assessment of the amount of terminal hair loss in four views, and can be used to track treatment response (Olsen 2004; Strazzulla 2018b).

Description of the intervention

Most cases of AA remit spontaneously, and it may be appropriate not to medically treat if this is consistent with the patient’s wishes; however, remission is not common in severe AA. A considerable number of patients require medical management to improve the growth of their hair; such management is mainly focused on stopping the local immune response against the hair follicle. There are different types of interventions currently in use, including local and systemic corticosteroids, contact immunotherapy, topical immunosuppressants, biologic agents, laser treatment, psychological support, and cosmetic strategies (Lee 2017; Murad 2018; Pratt 2017).

Classical immunosuppressants

Normally, in the paediatric population, the treatment used as the first option is topical corticosteroids of class I and II, since they have minimal adverse effects and are easily applied in their different formulations (lotion, foam, or shampoo). Adverse events that may occur due to chronic use must be taken into account (Peloquin 2017). Intralesional corticosteroids are amongst the first‐line alternatives in patients older than 12 years old with AA (Lee 2017; Pratt 2017). Usually, triamcinolone acetonide is administered by injection with a fine needle into the superior subcutaneous tissue, with the aim of stimulating hair growth at the site of injection (Kassim 2014). Most patients need multiple injections (Pratt 2017). Oral corticosteroids (e.g. oral prednisolone, prednisone, and dexamethasone) have been used successfully for extensive and rapidly progressive AA (e.g. an extension larger than 50% of the scalp). As some patients need prolonged treatment to maintain hair growth, the benefits of this management should be balanced against the adverse effects (e.g. cutaneous atrophy, Cushing's syndrome) (Pratt 2017).

Systemic cyclosporine, methotrexate, and azathioprine could be used as an isolated treatment or as an adjuvant therapy to reduce the use of systemic corticosteroids. They have potential adverse effects (e.g. hepatotoxicity, renal failure, higher risk of infections), so a careful study of the benefit‐risk balance should be performed in each case (Pratt 2017).

There are other therapeutic options considered to be immunosuppressants, mostly used in combination with first‐line agents: topical calcineurin inhibitors (Price 2005), and phototherapy with psoralens‐(PUVA) (Whitmont 2003).

Biologic agents

Patients with AA present with hair follicle dystrophy and acceleration of hair follicles into the catagen phase, due to an over‐expression of a variety of proinflammatory cytokines in the hair follicle, along with a tissue upregulation of several cytokines such as interleukin 2, 7, 15, and 21 and tumour necrosis factor (TNF) alpha. Anti‐TNF alpha biologic agents (etanercept, adalimumab, and infliximab), and other anticytokine agents (dupilumab, ustekinumab, amongst others) are drugs that inhibit the physiological response to this pro‐inflammatory cytokine and help stop inflammation. These drugs are not the first choice for the treatment of AA, as there is scarce evidence of their efficacy (Alsantali 2011, Schwartz 2017, Waśkiel‐Burnat 2021).

Small molecule inhibitors

Other new drugs (e.g. apremilast, tofacitinib, ruxolitinib, baricitinib), have recently been proposed as potentially useful for severe forms of AA. Amongst them, the most promising alternatives are the Janus kinases (JAK) inhibitors that suppress the T‐cell mediated inflammatory responses, which promote hair growth by stimulating the activation and proliferation of hair follicle stem cells and other related mechanisms (Hosking 2018; Kostovic 2017; Liu 2018; Strazzulla 2018a).

Contact immunotherapy

Contact immunotherapy is the first line of management for patients with more than 50% hair loss on the scalp (Jang 2017; Sutherland 2015; Yoshimasu 2016). This intervention starts with an induced contact dermatitis through topical administration of a chemical sensitiser (e.g. dinitrochlorobenzene (DNCB), diphencyprone (DPCP), squaric acid dibutyl ester (SADBE), or anthralin‐dithranol) (Perera 2015). The mechanism of action of contact immunotherapy is not fully understood. Related adverse events include local (e.g. pruritus, blistering, exudate) and distant reactions (e.g. disseminated eczema, urticaria). Contact immunotherapy is free of systemic adverse effects, and it can be maintained over a long‐term period.

Hair growth stimulants

Minoxidil is a topical treatment that has been evaluated in several studies of AA in both adults and children, and its efficacy has been related to a sustained hair growth effect. In general, it is used in combination with other treatments (i.e. topical or intralesional steroids) (Fiedler‐Weiss 1987; Maitland 1984; Price 1987a).

Other therapies

Additional therapeutic options (mostly used in combination with first‐line agents) include laser therapy, cryotherapy, and vitamins (Gupta 2017; Strazzulla 2018a).

How the intervention might work

Pathogenesis of AA is related to several factors, including autoimmunity in combination with a genetic contribution (Spano 2015). In selected cases with limited patchy hair loss, spontaneous resolution occurs within the first year in 34% to 50% of cases. However, extensive areas of AA have a poor prognosis (Tosti 2006). Currently, there is no cure for AA, although there are many possible treatments focusing on the degree of hair loss and the patient's preferences.

Classical immunosuppressants

The mechanism of action of corticosteroids involves the reduction of CD3+ T cells, CD8+ T cells, CD11c+ dendritic cells and CD1a+ Langerhans cells. In addition, it has been reported that corticosteroid treatment causes a downregulation effect over genes that encode several interleukins and chemokines (proteins and molecular messengers) (IL12B, CC‐chemokine ligand 18, and IL32), as well as upregulation of genes encoding several keratins (KRT35, KRT75, and KRT86) (Fuentes‐Duculan 2016; Kurosawa 2006).

Although the mechanism of action of methotrexate in the management of AA is not completely known, the evidence suggests that it produces an inhibition of the enzyme dihydrofolate reductase, then causes an increase in adenosine and release into the extracellular space, which inhibits the accumulation of white blood cells, as well as a variety of activities of monocytes, macrophages and T cells, but it also leads to a reduction in synthesis of TNF alpha and interferon‐gamma (Hammerschmidt 2014).

Cyclosporine has an immunosuppressive effect, which allows it to block gene transcription in activated T cells that codify for different cytokines. It also regulates nuclear translocation and activation of NFAT (nuclear factor of activated T‐cells) due to the inhibition of the phosphatase activity of calcineurin (Matsuda 2000).

Azathioprine inhibits the synthesis of DNA and decreases the proliferation of T and B lymphocytes. In addition, it decreases the number of Langerhans cells and other antigen‐presenting cells in the skin (Farshi 2010).

Biologic agents

Recent studies on AA development suggest the involvement of TNF alpha; hence, some biologics such as adalimumab, infliximab and etanercept, amongst others, could have benefits in the management of this condition by blocking this factor (Alsantali 2011).

Small molecule inhibitors

Patients with AA present with hair follicle dystrophy and acceleration of hair follicles into the catagen phase, due to an over‐expression of a variety of pro‐inflammatory cytokines in the hair follicle, along with a tissue upregulation of several γ‐chain cytokines (such as interleukin 2, 7, 15, and 21) and IFN‐γ elements, which are signalled through JAK1 and JAK2. The JAK‐inhibitors are able to block JAK signals and suppress the T‐cell‐mediated inflammatory responses, which promote hair growth by stimulating the activation and proliferation of hair follicle stem cells and other related mechanisms (Schwartz 2017).

Contact immunotherapy

The mechanism of action of contact immunotherapy is not fully understood and can be both allergic (allergens such as DPCP or SADBE), and also irritative such as anthralin. Some studies suggest that the induced allergic contact dermatitis attracts CD4+ T cells away from the perifollicular region, changing the milieu of immune cells surrounding hair follicles. Other potential mechanisms include the nonspecific stimulation of T suppressor cells in the skin, the increase of local expression of transforming growth factor beta, and the activation of myeloid suppressor cells contributing to autoreactive T cell silencing (Pratt 2017).

Hair growth stimulants

Minoxidil is a vasodilator that causes hypertrichosis (excessive hair growth anywhere on the body) as a secondary effect. Some authors have proposed a role for it as a topical treatment for AA. This effect could be the result of vasodilation, which facilitates the supply of oxygen and nutrients to the hair follicles, which in turn induces the formation of new vessels (Choi 2018). On the other hand, it is important to mention that the vascular endothelial growth factor, produced by endothelial cells and expressed in hair follicles, is associated with vasodilation processes and seems to be involved in hair growth (Wu 2018).

Other therapies

There are other treatments for which mechanisms of action are not fully understood. Laser therapy has been used in children with AA, with variable results. The most common wavelengths used are 308 nanometres (nm) (excimer laser), 904 nm (diode laser), and 1540 nm (Er:Glass laser) (Al‐Mutairi 2007; Waiz 2006; Yoo 2010).

Cryotherapy has also been used in patches of AA, however, its mechanism of action is unclear. It could be related to the effect produced by exposure for a short period to liquid nitrogen, which increases blood flow and improves microcirculation through reactive vasodilatation (Jun 2017).

Why it is important to do this review

In 2017, Cochrane Skin undertook a large‐scale exercise with stakeholders for prioritisation of systematic reviews to be developed in the next two years. Specifically, one of the target conditions prioritised was AA, which is a common autoimmune disease affecting all types of people around the world, with reports of cumulative life incidence of up to 2% (Mirzoyev 2014). As mentioned above, AA affects the quality of life of patients and caregivers and interferes especially with their daily activities. In addition, AA is an important cause of absence from both school and work, and it generates consequences in the global economy and an excessive burden for the healthcare system.

In addition, it is important to highlight the contribution of Macbeth and colleagues, who presented the top 10 research priorities for AA in the UK, which include at least three issues related to the objectives of this review, as follows (Macbeth 2017).

• Are immunosuppressant therapies better than placebo in the treatment of AA?

• In AA, are biologic therapies (including JAK inhibitors and anti‐cytokine therapies) more effective than placebo in causing hair regrowth?

• Do any treatments have long‐term therapeutic benefits in AA?

Cochrane Skin published a systematic review of AA in 2008 (Delamere 2008); however, new evidence about the effectiveness and safety of potential treatments has been published in the last decade. The update of this systematic review is essential. In addition, due to the numerous interventions proposed for the management of AA, the methodology of a systematic review with network meta‐analysis will be an important tool to assess the different alternatives and to guide clinical practice through a complete comparison of proposed treatments for AA. Due to this, the scope of the original 2008 review (Delamere 2008) was expanded to undertake a network meta‐analysis. The plans for this new review were published as a protocol 'Treatments for alopecia areata: a network meta‐analysis' (Novoa‐Candia 2020).

Finally, this systematic review will identify the existing gaps in the evidence related to the management of this condition and can be used to inform new lines of AA research.

Objectives

To assess the benefits and harms of treatments for alopecia areata (AA), alopecia totalis (AT), and alopecia universalis (AU) in children and adults.

Methods

Criteria for considering studies for this review

Types of studies

We included randomised controlled trials (RCTs) that evaluated efficacy and safety or both. We considered RCTs with parallel groups, cluster‐randomised trials, cross‐over trials, experiments with repeated measures on participants, multiple intervention groups, half‐head studies, and multiple body parts.

Types of participants

We considered individuals who had been diagnosed by a medical practitioner with AA, AT, or AU, including both paediatric and adult populations. We only included studies in which there was a subset of relevant participants if it was possible to obtain specific and separate information. We imposed no restrictions on the age, sex, or ethnicity of the participants. Diagnosis of AA should be achieved by clinical examination, trichoscopy/dermoscopy, or biopsy. We excluded trials with participants suffering from androgenetic alopecia and cicatricial alopecia.

Types of interventions

Decision Set

We considered for interventions of interest those treatments that are usually available and commonly used in clinical practice by dermatologists. Interventions scarcely used and with little relevance in clinical practice, such as carboxytherapy, topical rosemary, topical photodynamic therapy with 5‐aminolaevulinic acid, ayurvedic drugs, the curative effect observation of colligation therapy, or not globally used, such as Chinese medicine were not considered relevant for this review. We excluded studies assessing interventions not included in the list below.

Included studies assessed one or more of the following interventions, delivered at any dose, duration, and follow‐up time.

Classical immunosuppressants

• Topical corticosteroids

• Topical calcineurin inhibitors

• Intralesional corticosteroids

• Systemic corticosteroids (e.g. prednisolone)

• Systemic cyclosporine

• Methotrexate

• Azathioprine

• Hydroxychloroquine

• Sulfasalazine

• Psoralens taken by mouth + exposure to ultraviolet light A (PUVA)

Subcutaneous biologics

• Etanercept

• Adalimumab

• Infliximab

• Abatacept

• Alefacept

• Dupilumab

• Ustekinumab

Small molecule inhibitors

• Apremilast

• Tofacitinib

• Ruxolitinib

• Baricitinib

Contact immunotherapy

• Dinitrochlorobenzene (DNCB)

• Diphencyprone (DPCP)

• Squaric acid dibutyl ester (SADBE)

• Anthralin (dithranol)

Hair growth stimulants

• Oral minoxidil

• Topical minoxidil

• Topical bimatoprost

Other therapies

• Cryotherapy

• Laser

• Vitamin supplementation

• Aromatherapy

• Mesotherapy (zinc, selenium, biotin, platelet‐rich plasma)

Interventions could have been administered either as a single therapy or in combination.

Reference set

The comparators were placebo, no treatment, vehicle only, or another active compound.

Types of outcome measures

We considered for this section the recommendations of the consensus of Olsen and colleagues to obtain objective outcome measures on AA. We used the Severity of Alopecia Tool (SALT), which evaluates the percentage of scalp involvement, in preference to other reported measures (Olsen 2018).

This score has now been widely adapted to rate the extent of scalp hair loss in AA (range; 0 [no hair loss]–100 [complete hair loss]). The SALT score is obtained by the summation of the percentage of hair loss in four predefined areas of the scalp (vertex – 40%, posterior – 24%, right side – 18%, and left side –18%).

Primary outcomes

1. Short‐term hair regrowth ≥ 75%: the proportion of participants with clinically significant hair regrowth between 12 and 26 weeks of follow‐up, as rated by the participant or medical practitioner (where both types of rating were available, we used medical assessment). We deemed ≥ 75% regrowth of the affected area to constitute significant hair regrowth.

2. Incidence of serious adverse events (SAE): mortality, hospitalisation, surgical intervention, temporary or permanent sequelae, and serious infections. We considered the number of participants with at least one serious adverse event.

Secondary outcomes

1. Health‐related quality of life: measured with validated generic and specific instruments such as the Alopecia Areata Quality of Life Index (AA‐QLI) (Fabbrocini 2013), Dermatology Life Quality Index (DLQI) (Finlay 1994), and Skindex (versions 29 and 16) (Chren 2012).

2. Long‐term hair regrowth ≥ 75%: the proportion of participants with long‐term sustainability of hair regrowth (greater than 26 weeks of follow‐up).

The outcomes were described as presented by the study authors. In the case of studies with extremely vague outcomes, or if the outcome was not clearly described but seemed most likely to map to one of our predefined outcomes, then we described the outcome narratively and took into consideration narrative information when interpreting the results of the meta‐analyses.

Timing of outcomes

Regarding time points and follow‐up, Olsen and colleagues recommended that at least a 12‐week observation period should be measured (Olsen 2018). Therefore, we described the results obtained in the longest follow‐up time reported by adequate numbers of studies for meaningful and representative meta‐analysis. We classified outcomes as short‐term (between 12 and 26 weeks of follow‐up) and long‐term (greater than 26 weeks of follow‐up). In the case of multiple time points of measurement, we used the result closest to 26 weeks for short‐term outcomes and closest to one year for long‐term outcomes. We assessed whether the time point assessment reported was biologically reasonable. We also considered the quantity and quality of data available for each time point to be pooled in the meta‐analysis. We took these considerations into account when assessing the quality of the study. We performed the analysis of studies with outcome data at similar time points.

Search methods for identification of studies

We identified all relevant RCTs without restrictions on language or publication status (published, unpublished, in press, or in progress).

Electronic searches

Searches of the following databases were first run on 13 October 2021 by Liz Doney, the Cochrane Skin information specialist, and repeated on 22 July 2022 by MMH and TBD (see search strategies in Appendix 1):

• The Cochrane Skin Specialised Register 2021 via the Cochrane Register of Studies (CRS‐Web).

• The Cochrane Central Register of Controlled Trials (CENTRAL); 2022, Issue 7, in the Cochrane Library.

• MEDLINE via Ovid (from 1946 onwards).

• Embase via Ovid (from 1974 onwards).

Trial registers

GSV searched the following trial registers up to 13 October 2021 and two authors (MMH, TBD) up to 22 July 2022 using the strategies described in Appendix 1.

• ClinicalTrials.gov (www.clinicaltrials.gov)

• The World Health Organisation International Clinical Trials Registry Platform (ICTRP) (trialsearch.who.int/)

Retractions and errata

A search to identify retraction statements or errata related to our included studies was undertaken in MEDLINE and Embase on 11 November 2021.

Searching other resources

Searching reference lists

We checked the bibliographies of included studies and relevant systematic reviews identified for further references to relevant trials, including the reference list of the previous Cochrane review of AA (Delamere 2008).

Correspondence with trialists/experts/organisations

When it was possible, we contacted the authors of the study reports to determine the current status of the clinical trial and the availability and publication status of the results of these studies (see Table 14 'Investigators contacted' for more information).

2. Investigators contacted.

Study‐date	Person contacted	Requested information	Contact date	Reply
IRCT20141209020250N5 2019	Alaa Al Bazzal	Current status of the trial, available and published results	21 September 2021	No response
CTRI/2019/02/017483 2019	Amit Bahuguna	Current status of the trial, available and published results	22 September 2021	No response
CTRI/2018/07/014701 2018	Astha Arora	Current status of the trial, available and published results	22 September 2021	"The trial has been completed but awaiting publication in an international journal"
IRCT20200130046311N1 2021	Azadeh Rezayat	Current status of the trial, available and published results	18 October 2021	No response
NCT04793945	Eman Riad Mohamed Hofny	Current status of the trial, available and published results	18 November 2021	No response
NCT02684123 2016 (Mikhaylov 2019)	Emma Guttman‐Yassky	Current status of the trial, available and published results	21 September 2021	"This is included now"
IRCT2013031112563N3 2014	Fardin Gharibi	Current status of the trial, available and published results	21 September 2021	No response
IRCT20100314003566N10 2020	Hamideh Azimi	Current status of the trial, available and published results	22 November 2021	No response
IRCT201410113566N5 2014 (Azimi 2018)	Hamideh Azimi	Current status of the trial, available and published results	1 October 2021	No response
NCT03473600 2018	Hanan Morsy	Current status of the trial, available and published results	21 September 2021	"The researcher didn't finish the study yet"
CTRI/2018/02/012046 2018	Marikanti Jayakumar	Current status of the trial, available and published results	22 September 2021	No response
NCT04660786 2020	Marwa Eldeeb	Current status of the trial, available and published results	18 November 2021	No response
NCT03535233 2018	Mona El‐Kaliobi	Current status of the trial, available and published results	22 September 2021	"The trial is completed but it is not yet published"
CTRI/2013/08/003880 2013	Nanda Kishore	Current status of the trial, available and published results	21 September 2021	No response
IRCT20150923024147N1 2019	Nika Kianfar	Current status of the trial, available and published results	21 September 2021	"The trial has ended some months ago and is currently under review in a journal!"
NCT02037191 2014	Pascal Joly	Current status of the trial, available and published results	21 September 2021	"Study completed. Manuscript sent to coauthors for approval"
CTRI/2018/10/015870 2018	Rahul Mahajan	Current status of the trial, available and published results	21 September 2021	"The trial is completed and the manuscript has been submitted to Pediatric Dermatology Journal for peer review"
NCT04412148 2020	Rana Hilal	Current status of the trial, available and published results	18 November 2021	"We are still working on this clinical trial"
IRCT20141209020250N4 2018 (Ghandi 2021)	Romina Daneshmand	Current status of the trial, available and published results	21 September 2021	No response
Ustuner 2017	Pelin Ustuner	Measurement of the treatment concentrations, final concentrations	22 September 2021	No response
TCTR20210521008 2021	Supisara Wongdama	Current status of the trial, available and published results	19 November 2021	No response
CTRI/2021/03/031963 2021	Surabhi Dayal	Current status of the trial, available and published results	18 November 2021	No response
CTRI/2017/02/007932 2017 (Gupta 2019)	Vishal Gupta	Current status of the trial, available and published results	22 September 2021	"We have completed the trial and the results are now published"
IRCT20181226042136N1 2019 (Asilian 2020)	Zayike Ganjei	Current status of the trial, available and published results	21 September 2021	"The study has been completed and its results have been published as an article"
IRCT2013010612031N1 2019	Zohreh Teranchinia	Current status of the trial, available and published results	22 September 2021	No response
RBR‐5kyg2r 2020	Vando Sousa	Current status of the trial, available and published results	22 November 2021	No response
NCT05414266	Adelaide A Hebert	Current status of the trial, available and published results	26 October 2022	"Not yet ready to publish data"
ISRCTN14007390	David Fleet	Current status of the trial, available and published results	26 October 2022	No response
CTRI/2022/04/041728	Farheem Begum	Current status of the trial, available and published results	26 October 2022	No response
CTRI/2022/03/040829	Hemanta Kumar Kar	Current status of the trial, available and published results	26 October 2022	"We have completed our study, but it is in analysis stage. It will take some time for submission in any good indexed journal for consideration to publish"
IRCT20211109053013N1	Nafise Yazdanian	Current status of the trial, available and published results	26 October 2022	"Fortunately, Our trial is published recently and the results are available"
KCT0006802	Ji Won Lee	Current status of the trial, available and published results	14 December 2022	"Our clinical trial is currently ongoing"
NCT05251831	Ahmed Nouh	Current status of the trial, available and published results	14 December 2022	"This paper is already published"

Adverse effects

We did not perform a separate search for adverse effects of interventions used for the treatment of AA. We only considered the adverse events described in the included studies.

Data collection and analysis

The results of the literature search were collected in an EndNote library, where duplicate studies were removed. The file was exported to Rayyan for the screening process (Ouzzani 2016).

Selection of studies

Seven authors (ACP, GSV, MGUR, DOQ, MMH, SFG, TBD) independently assessed the studies for eligibility. We reviewed the titles and abstracts of all identified studies to determine whether they fulfilled the inclusion criteria. We assessed the full texts of selected studies to confirm their relevance for inclusion. We solved any disagreements by consulting a third author (SVG). We were not blinded to the study authors’ names and institutions, journal of publication, or study results at any stage of the review. We recorded the reasons for the exclusion of potential studies in the Characteristics of excluded studies tables.

Data extraction and management

Ten authors (ACP, GSV, SFG, MMH, MGUR, DOQ, DSC, MNC, LGM, TBD) independently used a data extraction tool tailored for this systematic review, to extract information on the description of interventions, participants, outcome measures, methods, and methodological quality. This data extraction form was evaluated in a pilot test with a set of included studies.

Regarding outcomes, we extracted the number of participants allocated to each intervention group and the proportion of participants that reached more than 75% hair regrowth in the short‐term period. We extracted the proportion or incidence of participants in each group with at least one serious adverse event (mortality, hospitalisation, surgical intervention, temporary or permanent sequelae, and serious infections). We extracted from each trial the event rates and descriptions of serious adverse events. We also extracted the results related to our secondary outcomes: health‐related quality of life and the proportion of participants with long‐term hair regrowth ≥ 75%.

We resolved any disagreement by discussing it with a third review author (JZ).

Assessment of risk of bias in included studies

We used the August 2019 version of the Cochrane Risk of bias tool (RoB2) for RCTs (Sterne 2019) to evaluate the risk of bias of each included trial. The assessment was done independently by five authors (GSV, ACP, MMH, MGUR, DOQ), following the recommendations in Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2019b). Disagreements were resolved by discussion with a third author (JZ). The focus of our review was to assess the effect of assignment to the interventions at baseline. As the Handbook recommends, we evaluated each outcome using the RoB2 tool. The domains of this tool for RCTs included the following:

• 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;

• Bias in selection of the reported result.

Each domain had a group of signalling questions to retrieve relevant information for an assessment of risk of bias. The options were: 'yes', 'probably yes', 'probably no', 'no', and 'no information'. According to the answers, the risk of bias judgement for each domain would be either: 'low risk of bias', 'some concerns', or 'high risk of bias'.

The overall judgement about risk of bias depended on the result of each domain, as presented below.

• Low risk of bias: the study was judged to be at low risk of bias for all domains for this result.

• Some concerns: the study was judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain.

• High risk of bias: the study was judged to be at high risk of bias in at least one domain for this result, or the study was judged to have some concerns about multiple domains in a way that substantially lowers confidence in the result.

The risk of bias assessments informed our GRADE evaluations of the certainty of evidence for those outcomes presented in the Summary of findings tables.

For each bias judgement, we used the Excel template, which is available at https://www.riskofbias.info/welcome/rob-2-0-tool/current-version-of-rob-2.

Risk of bias in cross‐over randomised controlled trials

For this type of study, we started with RoB2 as it is and, for Domain 2 ('bias due to deviations from the intended interventions') and Domain 3 ('bias due to missing outcome data'), we used the respective signalling questions from the RoB2 tool guidance for cross‐over RCTs (Higgins 2019c).

Measures of treatment effect

We estimated the relative treatment effects using risk ratios (RRs) for each dichotomous outcome with 95% confidence intervals (CIs). In the case of a continuous outcome (i.e. quality of life scores), we analysed data using mean differences (MDs) with their respective 95% CI.

We had planned to perform an NMA, assess the consistency, and estimate the effect of the treatments using rankings based on the probability of being the best treatment and the surface under the cumulative ranking curve‐SUCRA (Chaimani 2013; Salanti 2011; Shim 2017). Since the assumptions preserving the validity of the NMA were not met, we opted to present a narrative description of the findings and a set of graphs illustrating the geometry of the networks using Stata Statistical Software, release 15 (Stata 2017).

Unit of analysis issues

Following the guidance from the Handbook (Higgins 2019d), we considered the potential impact of different designs on the analysis, including special considerations for cross‐over trials and experiments with repeated measures on participants.

Cross‐over trials

In accordance with the guidance in the Handbook (Higgins 2019c), we followed the conservative option of including data from the first trial period only.

Repeated measures on participants

We analysed the longest follow‐up time reported and classified it as short‐term (between 12 and 26 weeks of follow‐up) or long‐term (greater than 26 weeks of follow‐up) (see 'Timing of outcomes', above). During the analysis, we grouped trials that reported similar timings, to make comparisons between studies.

Other study designs considered in the protocol but not included in this version of the review (i.e. cluster‐randomised trials) are in Appendix 2.

Dealing with missing data

In studies with missing data, we contacted trial authors or sponsors, and we made requests for missing data. We created a table (Table 14) with information about the authors contacted, and the information requested and received. For the primary and secondary efficacy binary outcomes (i.e. short‐term and long‐term hair regrowth ≥ 75%), we used the intention‐to‐treat (ITT) principle for the analysis. We assumed a non‐responder imputation for lost‐to‐follow‐up participants. This meant that participants who dropped out from the trial were considered non‐responders (i.e. they did not achieve the regrowth outcome). However, the denominators used for calculating the incidence of these outcomes were based on all participants who were initially randomised.

We undertook a sensitivity analysis to assess how sensitive results were to changes to this assumption. For this analysis, we assumed a worst‐case scenario in which all dropouts in the active intervention group were considered non‐responders (i.e. as in the ITT primary analysis explained above) and all dropouts in the comparator group were considered responders (i.e. they achieved the regrowth outcome). The impact of missing data on the review's findings was discussed in the Discussion section.

Assessment of heterogeneity

We explored the variability between studies by comparing trial population characteristics across the included studies and assessed the presence of clinical or methodological heterogeneity by analysing the different sources of clinical and methodological heterogeneity (Higgins 2019a). For our pairwise analyses, we used the statistical test included in RevMan Web (RevMan Web 2019). The Chi² test performs an evaluation to explain the chance of variability, with heterogeneity occurring if a P value is less than 0.1. In addition, we used the I² value to quantify the heterogeneity, through a description of the variability in effect estimates that is due to heterogeneity. We planned to explore sources of heterogeneity by performing subgroup analysis and meta‐regressions. Meta‐regressions were not performed as there were very few studies included in the meta‐analyses ( see Subgroup analysis and investigation of heterogeneity).

Assessment of reporting biases

We had planned to assess the presence of reporting bias using funnel plots, contour‐enhanced funnel plots, and regression asymmetry tests if we identified sufficient studies (more than 10) for inclusion in the meta‐analysis (Begg 1994; Higgins 2021b; Sterne 2011). However, the number of included studies was not sufficient to do this.

Data synthesis

For estimates of RR, when two or fewer studies were available to test the same comparison for the same outcome, we performed a meta‐analysis using a fixed‐effect model based on the Mantel‐Haenszel method (Mantel 1959). Otherwise, we applied a random‐effects model using the DerSimonian‐Laird method (DerSimonian 1986).

Subgroup analysis and investigation of heterogeneity

We had planned to perform a subgroup analysis if we had identified sufficient available studies, especially for factors that could explain the differences between treatments.

The subgroups of main interest were:

• Age group (less than 12 years versus 12 years or more);

• The extent of hair loss (less than 75% versus 75% or more);

• The three types of alopecia according to the affected area: a) alopecia areata (AA); b) alopecia totalis (AT); and c) alopecia universalis (AU).

Meta‐regression was not performed as there were very few studies included in the meta‐analyses.

Sensitivity analysis

We planned to perform a sensitivity analysis using the risk of bias as a variable to explore the robustness of the findings. We planned to verify the behaviour of our estimators, including and excluding studies with a high risk of bias overall (see Assessment of risk of bias in included studies).

Summary of findings and assessment of the certainty of the evidence

We evaluated the certainty of the evidence through the GRADE approach (Schünemann 2013) which includes the assessment of the risk of bias, inconsistency, indirectness, imprecision, and publication bias for each of the outcomes of interest:

• Short‐term hair regrowth ≥ 75%;

• Incidence of serious adverse events;

• Health‐related quality of life;

• Long‐term hair regrowth ≥ 75%.

We completed the summary of findings tables using GRADEpro GDT software.

Since all the studies included were RCTs, the certainty of the evidence for an outcome was at first rated as high. However, based on the assessments of the domains that raised serious or very serious concerns, it was downgraded by one or two levels, respectively. The certainty of the evidence could be high, moderate, low, or very low.

Results

Description of studies

Results of the search

The searches of the six databases (see Electronic searches) retrieved 1314 records (see Figure 1). Our searches using other resources, such as the reference lists of the included studies (see Searching other resources), did not identify additional studies that met the inclusion criteria. We, therefore, had a total of 1314 records for the title and abstract selection.

1.

Study flow diagram

Once duplicates had been removed, we had a total of 1116 records. We excluded 892 records based on titles and abstracts. We obtained the full text of the remaining 224 records. Then, we excluded 121 studies (122 records) (see Characteristics of excluded studies). We classified seven studies as Studies awaiting classification and 16 as ongoing trials (see Characteristics of ongoing studies).

Of these, we included 63 studies (79 records). For a further description of our screening process, see the study flow diagram (Figure 1). Details of the included studies are described in the Characteristics of included studies table.

Included studies

Trial design

All 63 included studies were randomised controlled clinical trials. These studies evaluated 47 different therapies in a total of 4817 participants (see Characteristics of included studies). All studies used a parallel‐group design except one that used a cross‐over design (White 1985). The mean sample size was 78 (range: 4 to 654). Of the 63 studies, 53 (84.1%) were conducted in a single centre, six (9.5%) were multicentre (from 2 to 169 centres) (Asilian 2020; Guttman‐Yassky 2021; King 2022a; King 2022b; Olsen 2019; Strober 2009), and in four studies (6.3%), this information was not clear (Agrawal 2020; Narahari 1996; Rashad 2022; Ustuner 2017). All studies recruited outpatient participants from dermatology clinics. The studies took place worldwide, in Asia (n = 24; 38.0%), in Africa (n = 11; 17.4%), in North America (n =12; 19.0%), in Europe (n = 10; 15.8%), in South America (n = 3; 4.7%), in Oceania (n = 2; 3.1%), or in multiple continents (n = 1; 1.5%).

In total, 45 studies out of 63 (71.4%) were two‐arm and 18 (28.5%) were multi‐arm. Six of these 18 multi‐arm studies (33.3%) assessed the same experimental drug at multiple dose levels (Abdel 2020; Chu 2015; King 2022a; King 2022b; Rajan 2021; Saif 2012), 10 (55.5%) assessed at least two different drugs (Abdallah 2020; Asilian 2020; Bokhari 2018; El‐Taieb 2017; Gupta 2019; Kuldeep 2011; Lohrasb 2015; Tawfik 2022; Tosti 1986; Trink 2013), and two (11.1%) assessed both the same experimental drug at multiple dose levels and different drugs (Sousa 2020; Ustuner 2017).

Characteristics of the participants

The participants were between two and 74 years old; there were more men (n = 2379; 56%) than women (n = 2076; 44%). Age and gender were not reported for 577 and 362 participants from 14 and eight studies, respectively. Two studies of 63 (3.1%) included participants under 12 years of age (Amaral 2003; Lenane 2014). Most of the studies evaluated only the scalp, but 11 studies evaluated both scalp and body hair. The studies included participants with AA (n = 26), AT (n = 1, White 1985), AU (n = 1, Bokhari 2018), and mixed cases (including AA, AT, and AU) (n = 31). Four studies did not report the type of alopecia areata (Agrawal 2020; Kapoor 2020; Sayed 2020; Tawfik 2022).

Characteristics of the interventions

We found a wide variety of interventions and comparisons from the 63 included studies. To show this variety, we split the comparisons into two groups: active treatment versus placebo and active treatment versus another active treatment. Studies with two or more parallel arms comparing different doses of the same drug against placebo were grouped into one single 'arm' (King 2022a; King 2022b; Rajan 2021; Sousa 2020; Ustuner 2017).

Active treatment versus placebo: 33 studies

• Aromatherapy (n = 2) (Hay 1998; Ozmen 2015);

• Oral cyclosporine (n = 1) (Lai 2019);

• Diphencyprone (n = 1) (Tosti 1986);

• Intramuscular alefacept (n = 1) (Strober 2009);

• Intralesional betamethasone (n = 2) (Sousa 2020; Ustuner 2017);

• Intralesional minoxidil (n = 1) (Abdallah 2020);

• Intralesional minoxidil + triamcinolone (n = 1) (Abdallah 2020);

• Intralesional triamcinolone (n = 6) (Abdallah 2020; Chu 2015; Rajan 2021; Sousa 2020; Trink 2013; Ustuner 2017);

• Intralesional platelet‐rich plasma (n = 3) (El‐Taieb 2017; Tawfik 2022; Trink 2013);

• Low‐level laser therapy (n = 1) (Tawfik 2022);

• Mesotherapy (n = 1) (Rashad 2022);

• Oral apremilast (n = 1) (Mikhaylov 2019);

• Oral baricitinib (n = 2) (King 2022a; King 2022b);

• Oral prednisolone versus placebo (n = 1) (Kar 2005);

• Squaric acid dibutyl ester (n = 1) (Tosti 1986);

• Subcutaneous dupilumab (n = 1) (Guttman‐Yassky 2021);

• Topical betamethasone (n = 1) (Amaral 2003);

• Topical clobetasol (n = 5) (Bokhari 2018; Tosti 1986; Tosti 2006; Ucak 2012; Ucak 2014);

• Topical desoximetasone (n = 1) (Charuwichitratana 2000);

• Topical minoxidil (n = 9) (El‐Taieb 2017; Fransway 1988; Khoury 1992; Luo 1991; Price 1987a; Price 1987b; Tosti 1986; White 1985; Zaib 2017);

• Topical ruxolitinib (n = 2) (Bokhari 2018; Olsen 2019);

• Topical tofacitinib (n = 1) (Bokhari 2018).

Active treatment versus active treatment: 39 studies

• Cryotherapy (three different durations) (n = 1) (Abdel 2020);

• Cryotherapy versus fractional CO₂ laser (n = 1) (Nouh 2022);

• Cryotherapy versus PUVA (n = 1) (Sayed 2020);

• Diphencyprone versus anthralin (n = 2) (Azimi 2018; Barreto‐Rocha 2021);

• Diphencyprone versus diphencyprone + anthralin (n = 2) (Agrawal 2020; Ghandi 2021);

• Diphencyprone versus diphencyprone + topical minoxidil (n = 1) (Shapiro 1993);

• Diphencyprone versus squaric acid dibutyl ester (n = 2) (Tiwary 2016; Tosti 1986);

• Intralesional betamethasone (three different doses) (n = 2) (Sousa 2020; Ustuner 2017);

• Intralesional triamcinolone (three different doses) (n = 2) (Rajan 2021; Ustuner 2017);

• Intralesional triamcinolone versus anthralin (n = 1) (Narahari 1996);

• Intralesional triamcinolone versus excimer laser (n = 1) (Kianfar 2021);

• Intralesional triamcinolone versus intralesional betamethasone (n = 2) (Sousa 2020; Ustuner 2017);

• Intralesional triamcinolone versus intralesional methotrexate (n = 1) (Hamdino 2021);

• Intralesional triamcinolone versus intralesional minoxidil (n = 1) (Abdallah 2020);

• Intralesional triamcinolone versus intralesional minoxidil + triamcinolone (n = 1) (Abdallah 2020);

• Intralesional triamcinolone versus intralesional platelet‐rich plasma (n = 4) (Albalat 2019; Hedge 2020; Kapoor 2020; Trink 2013);

• Intralesional triamcinolone versus topical betamethasone (n = 2) (Devi 2015; Kuldeep 2011);

• Intralesional triamcinolone versus topical tacrolimus (n = 1) (Kuldeep 2011);

• Intralesional platelet‐rich plasma versus low‐level laser therapy (n = 1) (Tawfik 2022);

• Intralesional platelet‐rich plasma versus topical minoxidil (n = 1) (El‐Taieb 2017);

• Oral betamethasone versus oral azathioprine (n = 2) (Gupta 2019; Verma 2015);

• Oral betamethasone versus oral methotrexate (n = 1) (Asilian 2020);

• Oral betamethasone versus oral prednisolone (n = 1) (Goyal 2000);

• Oral methylprednisolone comparing three dosage regimens (n = 1) (Saif 2012);

• Oral ruxolitinib versus oral tofacitinib (n = 1) (Almutairi 2019);

• Topical betamethasone lotion versus betamethasone foam (n = 1) (Mancuso 2003);

• Topical betamethasone versus topical tacrolimus (n = 1) (Kuldeep 2011);

• Topical bimatoprost versus topical mometasone (n = 1) (Zaher 2014);

• Topical clobetasol versus topical hydrocortisone (n = 1) (Lenane 2014);

• Topical clobetasol versus topical ruxolitinib (n = 1) (Bokhari 2018);

• Topical clobetasol versus topical tacrolimus (n = 1) (Ullah 2022);

• Topical clobetasol versus topical tofacitinib (n = 1) (Bokhari 2018);

• Topical minoxidil versus aromatherapy (n = 1) (Lohrasb 2015);

• Topical minoxidil versus diphencyprone (n = 1) (Tosti 1986);

• Topical minoxidil versus topical methotrexate (n = 1) (Toma 2022);

• Topical minoxidil versus topical minoxidil + tacrolimus (n = 1) (Iraji 2014);

• Topical minoxidil versus squaric acid dibutyl ester (n = 1) (Tosti 1986);

• Topical minoxidil + aromatherapy versus aromatherapy (n = 1) (Lohrasb 2015);

• Topical minoxidil + aromatherapy versus topical minoxidil (n = 1) (Lohrasb 2015);

• Topical ruxolitinib versus topical tofacitinib (n = 1) (Bokhari 2018).

For each study, we provided details of the dosage regimen in the Characteristics of included studies tables.

Due to the large amount of information collected and to simplify the presentation of data in the main text, we prioritised some of the interventions based on their relevance in clinical practice. The prioritised interventions included systemic therapies (classical immunosuppressants, biologics, and small molecule inhibitors), local therapies (intralesional corticosteroids, topical small molecule inhibitors, contact immunotherapy, hair growth stimulants) and cryotherapy. We grouped these interventions, considering the type of therapy and administration route, into 12 main comparisons. These interventions were compared versus placebo or versus another different treatment. The 12 comparisons are detailed in the Effects of interventions section. In total, 23 studies were included in these comparisons.

We considered that non‐prioritised comparisons were those that included topical classical immunosuppressants, oral or intralesional minoxidil, and other therapies such as vitamin supplementation, aromatherapy, mesotherapy (zinc, selenium, biotin, platelet‐rich plasma). The results of these non‐prioritised comparisons for the outcomes of interest are reported in a series of online appendices (https://osf.io/rd7cg/?view_only=9017f5a8fe284b1dab83b5d7d194bf51).

Characteristics of the outcomes

For the efficacy outcomes, 33 of 63 studies (52.3%) reported the proportion of participants with clinically significant hair regrowth (≥ 75% regrowth of the affected area) between 12 and 26 weeks (Abdel 2020; Albalat 2019; Almutairi 2019; Azimi 2018; Barreto‐Rocha 2021; Charuwichitratana 2000; Ghandi 2021; Goyal 2000; Gupta 2019; Guttman‐Yassky 2021; Hamdino 2021; Hedge 2020; Kar 2005; King 2022a; King 2022b; Kuldeep 2011; Lai 2019; Luo 1991; Mancuso 2003; Narahari 1996; Nouh 2022; Olsen 2019; Ozmen 2015; Rashad 2022; Saif 2012; Sayed 2020; Shapiro 1993; Tawfik 2022; Tiwary 2016; Toma 2022; Tosti 1986; Tosti 2006; Ustuner 2017). Five studies of 63 (7.9%) reported the proportion of participants with clinically significant long‐term hair regrowth ≥ 75% greater than 26 weeks (Hay 1998; King 2022a; King 2022b; Saif 2012; Trink 2013). One study of 63 (1.5%) reported the health‐related quality of life outcome with the Assessment of Quality of Life‐8D (AQoL‐8D) instrument through a change from baseline (positive values reflect an improvement in quality of life) on a scale from 0 (death) to 1 (full health) (Lai 2019). Seven of 29 studies (24.1%) without the primary efficacy outcome (≥ 75% regrowth) reported a proportion of participants with more than 50% of hair regrowth (Kianfar 2021; Lohrasb 2015; Mikhaylov 2019; Rajan 2021; Strober 2009; Ullah 2022; Zaib 2017) and 21 did not report the percentage of hair regrowth as an outcome (Abdallah 2020; Agrawal 2020; Amaral 2003; Asilian 2020; Bokhari 2018; Chu 2015; Devi 2015; El‐Taieb 2017; Fransway 1988; Hay 1998; Iraji 2014; Khoury 1992; Lenane 2014; Price 1987b; Price 1987a; Trink 2013; Ucak 2012; Ucak 2014; Verma 2015; White 1985; Zaher 2014). One trial didn't provide the proportion of participants that achieved the primary outcome of regrowth (Sousa 2020).

For the safety outcome, 47 of 63 (74.6%) studies reported serious adverse events with five studies having at least one event (Guttman‐Yassky 2021; King 2022a; King 2022b; Olsen 2019; Strober 2009).

The outcomes were evaluated between 12 and 64 weeks, as follows: 12 weeks (n = 12), 16 weeks (n = 1), 18 weeks (n = 1), 20 weeks (n = 4), 21 weeks (n = 1), 24 weeks (n = 21), 28 weeks (n = 3), 32 weeks (n = 1), 36 weeks (n = 6), 48 weeks (n = 2), 52 weeks (n = 4) and 64 weeks (n = 1). The week of assessment was not clearly defined in six studies (Abdel 2020; Azimi 2018; Chu 2015; Ghandi 2021; Gupta 2019; Sayed 2020).

Considering only the prioritised interventions, 14 studies from 12 comparisons reported short‐term hair regrowth ≥ 75% (Almutairi 2019; Goyal 2000; Gupta 2019; Guttman‐Yassky 2021; Kar 2005; King 2022a; King 2022b; Lai 2019; Luo 1991; Nouh 2022; Olsen 2019; Shapiro 1993; Tosti 1986; Ustuner 2017), 22 studies from 10 comparisons reported serious adverse events (18 reported zero events (Almutairi 2019; Asilian 2020; Fransway 1988; Goyal 2000; Gupta 2019; Kar 2005; Lai 2019; Luo 1991; Mikhaylov 2019; Nouh 2022; Price 1987a; Price 1987b; Rajan 2021; Shapiro 1993; Sousa 2020; Trink 2013; Ustuner 2017; Verma 2015) and only four reported at least one (Guttman‐Yassky 2021; King 2022a; King 2022b; Olsen 2019)). Moreover, one study (1 comparison) reported quality of life (Lai 2019), and two studies (1 comparison) reported long‐term hair regrowth ≥ 75% (King 2022a; King 2022b).

Contacts with authors of included studies

We contacted the author of one included study (Ustuner 2017) to query issues regarding the unit of measurement of the drug concentrations and the final concentrations of the drugs after the dilutions, but no response was received. The list of all other contacts with investigators is shown in Table 14.

Funding

Sixteen studies declared a source of funding: seven studies declared pharmaceutical company funding (Guttman‐Yassky 2021; Khoury 1992; King 2022a; King 2022b; Mikhaylov 2019; Olsen 2019; Tosti 2006) and nine studies declared non‐profit/institutional funding (Asilian 2020; Barreto‐Rocha 2021; Bokhari 2018; Kianfar 2021; Lai 2019; Lenane 2014; Saif 2012; Shapiro 1993; Strober 2009). Thirteen studies had no funding source (Abdallah 2020; Abdel 2020; Hedge 2020; Kapoor 2020; Kar 2005; Kuldeep 2011; Sayed 2020; Tawfik 2022; Tiwary 2016; Trink 2013; Ullah 2022; Ustuner 2017; Verma 2015) and 34 did not report the source of funding.

Excluded studies

We excluded 121 studies (122 records) that didn't meet the inclusion criteria: different study design (n = 58), different intervention (n = 49), and different population (n = 10). Other reasons for exclusion were study withdrawal (n = 3) or early termination of the trial (n = 1).

The whole selection process is described in Figure 1 and the reasons for exclusion are presented in Characteristics of excluded studies.

Studies awaiting classification

We categorised seven clinical trials that had been completed but had not yet been published as awaiting classification studies (Figure 1). More details are available in the Characteristics of studies awaiting classification.

Ongoing studies

We included 16 clinical trials as ongoing studies (Figure 1). More details are described in the Characteristics of ongoing studies.

Risk of bias in included studies

We report here the risk of bias (ROB) for the prioritised comparisons. The ROB assessments for each outcome are located at the side of all forest plots (see Analysis section and Risk of bias (tables)).

For the main efficacy outcome (short‐term hair regrowth ≥ 75%), we analysed the ROB of 14 studies from 12 comparisons (Table 90; Table 93; Table 95; Table 97; Table 99; Table 102; Table 104; Table 106; Table 107; Table 108; Table 110; Table 112) (Almutairi 2019; Goyal 2000; Gupta 2019; Guttman‐Yassky 2021; Kar 2005; King 2022a; King 2022b; Lai 2019; Luo 1991; Nouh 2022; Olsen 2019; Shapiro 1993; Tosti 1986; Ustuner 2017). The overall ROB was judged as high in 8 studies, some concerns in two, and low in four. In general, the authors did not provide details of the randomisation process, or how the measurement of the outcome was planned. The study protocol was not available in most of the included studies.

Risk of bias for analysis 1.1 Short‐term hair regrowth ≥ 75%.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Kar 2005

Some concerns

Adequate method of randomisation. Baseline characteristics were well‐balanced between groups.

High risk of bias

People delivering the intervention were probably not aware of the assignment of intervention, but patients could be aware of it due to the adverse effects of treatment. Analysis was probably not appropriate, existing some potential for an impact of a failure on the results

Low risk of bias

7 out of 43 patients (16%) were lost to follow‐up. Sensitivity analysis was not performed. However, participants were lost to follow‐up because of reasons unrelated to the study.

High risk of bias

Lack of information about the method of measuring the outcome and if assessors were blind to the intervention.

High risk of bias

Protocol or registry was not found, so information on the analysis plan was not available. A simple analysis was performed

High risk of bias

Participants and outcome assessors could have been aware of the assigned interventions. Data analysis was probably not appropriate and sensitivity analysis was not performed. The study protocol was not available

Lai 2019

Low risk of bias

Blinding of participants, study researchers and outcome assessors were blinded to the allocation sequence. Baseline characteristics data were similar across both groups.

High risk of bias

Adequate blinding method. The analysis to estimate the effect of the assignment was not accurate (per‐protocol analysis). The probable substantial impact of missing data on the result

High risk of bias

4 out of 32 (12.5%) patients were lost in the follow‐up. Lack of information about protocol deviations and reasons of patients who did not complete the study

Low risk of bias

The outcome measurement was correct. The outcome assessors were also blind to their assigned intervention

High risk of bias

Protocol or registry was not found, so information on the analysis plan was not available. A simple analysis was performed

High risk of bias

A per‐protocol analysis and simple analysis was conducted. There was probably an impact of missing data on the result. Protocol deviations and reasons of patients who dropped the study were not reported, and the protocol or registry was not found.

Risk of bias for analysis 2.1 Short‐term hair regrowth ≥ 75%.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Goyal 2000

Some concerns

No information on the method of randomisation. The authors took care in balancing groups.

Some concerns

Probably participants were aware of the intended intervention. The authors excluded from the analysis those participants that did not complete the follow‐up (around 1%). No reasons for participant losses were provided.

Low risk of bias

Three out of 15 participants with missing outcome data were excluded from the analysis: one and two from each group. This would not influence the results.

High risk of bias

Although there is no information about the method for the outcome assessment, the knowledge of the intervention may influence the result.

High risk of bias

No protocol or analysis plan was available. The outcome result was collected at different time points (from week 8 to 29) till reach the complete response.

High risk of bias

No protocol is available. Lack of information on the method of randomisation, blinding of participants and personnel, and the outcome assessment was not well established.

Gupta 2019

Some concerns

Randomisation was done using a random number table but concealment allocation was not reported. Groups were comparable at baseline.

High risk of bias

No information about blinding of participants or people delivering the intervention. No deviations from the intended interventions were reported. A per‐protocol analysis was performed. Nine participants (18%) were lost at follow‐up.

High risk of bias

The percentage of losses were 20% and 16% for each group. Losses were related to logistic issues (changes in residence, contact tracing) but this information is not given by groups.

High risk of bias

Although the outcome assessment method is fine, there is no enough information to know if outcome assessors were aware of the intervention. The kwnowledge of the intervention could influence the outcome assessment.

High risk of bias

No trial protocol available or analysis plan. Complete hair regrowth was no listed as an outcome in the methods section. The outcome was measured at the end of treatment (16 weeks).

High risk of bias

No protocol is available. Lack of information on the allocation concealment, blinding of participants or study personnel. The percentage of losses could influence the result.

Risk of bias for analysis 3.1 Short‐term hair regrowth ≥ 75%.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Ustuner 2017

Some concerns

There was no information on allocation concealment but there were no baseline differences beyond that expected by chance.

Some concerns

There was no information to determine if participants and people delivering the intervention were aware of the assigned intervention There was no information to support deviations from the intended intervention. An appropriate analysis was done

Low risk of bias

All data for this outcome were available to all participants

Low risk of bias

A standardized method of evaluation was used in the different treatment groups. The outcome assessors were not aware of the intervention received by study participants

Some concerns

There is no pre‐registered protocol available and there is no information on whether there were multiple eligible outcomes measurements or multiple analyzes of the data

Some concerns

Overall due to lack of information of allocation concealment, no information to determine if participants and people delivering the intervention were aware of the assigned intervention, and no pre‐registered protocol or statistical plan.

Risk of bias for analysis 4.1 Short‐term hair regrowth ≥ 75%.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Guttman‐Yassky 2021

Low risk of bias

There was no issues on randomisation process and there was no baseline imbalance

Low risk of bias

There were no deviations from intended interventions

Low risk of bias

All randomised patients were included in the analysis

Low risk of bias

Outcome measurement was appropriate

Low risk of bias

No biases were detected in the selection of the reported results

Low risk of bias

There is a low risk of bias from the randomisation process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of reported results.

Risk of bias for analysis 5.1 Short‐term hair regrowth ≥ 75%.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

King 2022a

Low risk of bias

The randomisation process was adequate

Low risk of bias

Patients and carers were probably not aware of the assigned interventions and an appropriate analysis was made to estimate the effect of assignment.

Low risk of bias

Authors presented data for nearly all randomised patients

Low risk of bias

Methods were adequate and the outcome assessors were probably not aware of the interventions.

Low risk of bias

Results for the outcomes of interest were presented according to the protocol, without selection.

Low risk of bias

There is a low risk of bias from the randomisation process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of reported results.

King 2022b

Low risk of bias

The randomisation process was adequate.

Low risk of bias

Patients and carers were probably not aware of the assigned interventions and an appropriate analysis was made to estimate the effect of assignment.

Low risk of bias

Authors presented data for all randomised patients.

Low risk of bias

Methods were adequate and the outcome assessors were probably not aware of the interventions.

Low risk of bias

Results for the outcomes of interest were presented according to the protocol, without selection.

Low risk of bias

There is a low risk of bias from the randomisation process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of reported results

Risk of bias for analysis 6.1 Short‐term hair regrowth ≥ 75%.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Almutairi 2019

Some concerns

No information on the method of randomisation. There were no differences between baseline participant characteristics.

Some concerns

Participants and study personnel were aware of the intended interventions. The analysis plan was not appropriate, as participants who were non‐adherents with the intervention were withdrawn.

Low risk of bias

Around 6% of randomised participants withdrew from the study due to non‐compliance. No details about lack of compliance for excluded participants. Missing outcome data were balanced between groups.

High risk of bias

The study is described as an "open‐label comparative study" and outcome measuring involves visual observation and therefore a judgment.

Some concerns

No study protocol is available, but in the method section, authors reported the use of one validated tool to measure the outcome and presented data for two trial endpoints (24 and 36 weeks).

High risk of bias

No study protocol is available. Lack of information on the method of randomisation. Participants and study personnel were aware of the intended interventions. Non‐adherent participants were withdrawn from the analysis.

Risk of bias for analysis 7.1 Short‐term hair regrowth ≥ 75%.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Olsen 2019

Low risk of bias

Interactive Response Technology was used for patient study number assignment and randomisation. There was no baseline imbalance that would suggest a problem with randomisation

Low risk of bias

It was a placebo placebo‐controlled study, and the authors declare a double‐blind study. The analysis was appropriate.

Low risk of bias

All randomised patients were included in the analysis .

Low risk of bias

The method of outcome measurement was appropriate, and the outcome assessors were blind to the intervention.

Low risk of bias

The outcome was analysed according to the analysis plan reported in the study protocol.

Low risk of bias

The study protocol and analysis plan were available. Randomisation, outcome measurement and analysis were appropriate, and the participants and study team were blind to the intervention.

Risk of bias for analysis 8.1 Short‐term hair regrowth ≥ 75%.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Tosti 1986

High risk of bias

The study sample and analyses correspond to 119 patients, but according to the article only 40 patients were randomised ("The forty patients with patchy alopecia that involved less than 25% of the scalp were randomised to all the therapies used"). This implies that an adequate procedure for allocating the interventions was not carried out. No allocation concealment procedure was performed. There was no information on the baseline characteristics of the intervention groups

High risk of bias

There was no information to determine if participants and those delivering the intervention were aware of the intervention received, but doses and therapeutic scheme were differents. There were no details on deviations from the intended interventions due to trial context. The randomisation process was performed in 40 patients, but the analysis includes 119 patients, which has an impact on the final analysis of the results.

High risk of bias

The study sample was 119 patients and only 40 patients were randomised. The analysis was performed on 119, and details of the 40 randomised patients are not presented.

High risk of bias

There were no details on the method of measuring the outcome. According to the information provided, it is understood that the same outcome was evaluated in every intervention group applying the same methodology. Outcome assessors could infer the intervention group due to adverse events Knowledge of the intervention received could have affected outcome measurement

Some concerns

There is no pre‐registered protocol available and there is no information on whether there were multiple eligible outcomes measurements or multiple analyses of the data.

High risk of bias

High risk across randomisation, missing outcome data and blinding of the outcome assessor

Risk of bias for analysis 9.1 Short‐term hair regrowth ≥ 75%.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Tosti 1986

High risk of bias

The study sample and analyses correspond to 119 patients, but according to the article only 40 patients were randomised ("The forty patients with patchy alopecia that involved less than 25% of the scalp were randomised to all the therapies used"). This implies that an adequate procedure for allocating the interventions was not carried out. No allocation concealment procedure was performed. There was no information on the baseline characteristics of the intervention groups

High risk of bias

There was no information to determine if participants and those delivering the intervention were aware of the intervention received, but doses and therapeutic scheme were differents. There were no details on deviations from the intended interventions due to trial context. The randomisation process was performed in 40 patients, but the analysis includes 119 patients, which has an impact on the final analysis of the results.

High risk of bias

The study sample was 119 patients and only 40 patients were randomised. The analysis was performed on 119, and details of the 40 randomised patients are not presented.

High risk of bias

There were no details on the method of measuring the outcome. According to the information provided, it is understood that the same outcome was evaluated in every intervention group applying the same methodology. Outcome assessors could infer the intervention group due to adverse events Knowledge of the intervention received could have affected outcome measurement

Some concerns

There is no pre‐registered protocol available and there is no information on whether there were multiple eligible outcomes measurements or multiple analyses of the data.

High risk of bias

High risk across randomisation, missing outcome data and blinding of the outcome assessor.

Risk of bias for analysis 10.1 Short‐term hair regrowth ≥ 75%.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Shapiro 1993

High risk of bias

There was no information on the method of randomisation and allocation concealment. There was baseline imbalance in some characteristics that would suggest a problem with randomisation.

High risk of bias

The information includes the statement "double‐blind manner" but without specific details about it. There were no deviations from intended interventions. The analysis was per‐protocol, and due to the small sample size, two patients who were not included in the analysis could affect the final estimate.

Some concerns

Two patients dropped out of the study due to an adverse event (generalized dermatitis). The efficacy could not be established in two patients who were not included in the analysis, and due to the small sample size, it could affect the final estimate

Some concerns

There were no details on the method of measuring the outcome, but according to the information provided, it is understood that the same outcome was evaluated in both intervention groups applying the same methodology. There was no information on whether the outcome assessors were aware of the intervention received, but knowledge of the intervention received could have affected the measurement of the results.

Some concerns

There is no pre‐registered protocol available and there is no information on whether there were multiple eligible outcomes measurements or multiple analyses of the data.

High risk of bias

Overall high due to the lack of information on randomisation, baseline imbalances, lack of information on whether outcome assessors were aware of the intervention received, and missing outcome data without appropriate analysis. In addition, there was no pre‐registered protocol available.

Risk of bias for analysis 11.1 Short‐term hair regrowth ≥ 75%.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Luo 1991

Some concerns

There is no information about the randomisation process. However, there are no significant differences in baseline characteristics.

Low risk of bias

Participants and study personnel were blinded to the intervention. All participants were analysed in the group as they have been randomised.

Low risk of bias

There are no missing outcome data.

Low risk of bias

Although the outcome assessment involved visual observation, the outcome assessors were blind to the intervention.

Some concerns

No protocol available or analysis plan. Outcome results were presumptively reported for the end of the study (12 weeks), and from a single analysis.

Some concerns

No protocol is available. Lack of information about the randomisation process and outcome results were presumptively reported for the end of the study (12 weeks) and from a single analysis.

Tosti 1986

High risk of bias

The study sample and analyses correspond to 119 patients, but according to the article only 40 patients were randomised ("The forty patients with patchy alopecia that involved less than 25% of the scalp were randomised to all the therapies used"). This implies that an adequate procedure for allocating the interventions was not carried out. No allocation concealment procedure was performed. There was no information on the baseline characteristics of the intervention groups

High risk of bias

There was no information to determine if participants and those delivering the intervention were aware of the intervention received, but doses and therapeutic scheme were differents. There were no details on deviations from the intended interventions due to trial context. The randomisation process was performed in 40 patients, but the analysis includes 119 patients, which has an impact on the final analysis of the results.

High risk of bias

The study sample was 119 patients and only 40 patients were randomised. The analysis was performed on 119, and details of the 40 randomised patients are not presented.

High risk of bias

There were no details on the method of measuring the outcome. According to the information provided, it is understood that the same outcome was evaluated in every intervention group applying the same methodology. Outcome assessors could infer the intervention group due to adverse events Knowledge of the intervention received could have affected outcome measurement

Some concerns

There is no pre‐registered protocol available and there is no information on whether there were multiple eligible outcomes measurements or multiple analyses of the data

High risk of bias

High risk across randomisation, missing outcome data and blinding of the outcome assessor.

Risk of bias for analysis 12.1 Short‐term hair regrowth ≥ 75%.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Nouh 2022

Some concerns

Insufficient information about the randomisation process.

Some concerns

Participants and carers were aware of the assigned interventions and protocol deviations were not reported.

Low risk of bias

There were no missing outcome data.

High risk of bias

Although there is no information about the method for the outcome assessment, the knowledge of the intervention may had some influence in the result.

Some concerns

No registry or protocol is available and no information about the analysis plan.

High risk of bias

Insufficient information about the randomisation process. Participants and carers were aware of the assigned interventions and protocol deviations were not reported. Although there is no information about the method for the outcome assessment, the knowledge of the intervention may had some influence in the result.

The overall ROB for the main safety outcome (incidence of serious adverse events) was judged in 22 studies from 10 comparisons (Table 91; Table 94; Table 96; Table 98; Table 100; Table 103; Table 105; Table 109; Table 111; Table 113) (Almutairi 2019; Asilian 2020; Fransway 1988; Goyal 2000; Gupta 2019; Guttman‐Yassky 2021; Kar 2005; King 2022a; King 2022b; Lai 2019; Luo 1991; Mikhaylov 2019; Nouh 2022; Olsen 2019; Price 1987a; Price 1987b; Rajan 2021; Shapiro 1993; Sousa 2020; Trink 2013; Ustuner 2017; Verma 2015). Amongst them, 18 studies reported zero serious events. From the remaining four studies which reported some events, the judgement was low in two studies (King 2022a; King 2022b) and some concerns in the remaining two (Guttman‐Yassky 2021; Olsen 2019).

Risk of bias for analysis 1.2 Incidence of serious adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Kar 2005

Some concerns

Adequate method of randomisation. Baseline characteristics were well‐balanced between groups.

High risk of bias

People delivering the intervention were probably not aware of the assignment of intervention, but patients could be aware of it due to the adverse effects of treatment. Analysis was probably not appropriate, existing some potential for an impact of a failure on the results.

Low risk of bias

7 out of 43 patients (16%) were lost to follow‐up. Sensitivity analysis was not performed. However, missingness in the outcome could be considered independent of their true value.

Low risk of bias

Outcome measurement was adequate. Moreover, the outcome assessors were not aware of the assigned intervention.

High risk of bias

Protocol or registry was not found, so information on the analysis plan was not available. A simple analysis was performed.

High risk of bias

Participants could have been aware of the assigned interventions. Data analysis was probably not appropriate and sensitivity analysis was not performed. The protocol was not available.

Lai 2019

Low risk of bias

Blinding of participants, study researchers and outcome assessors were blinded to the allocation sequence. Baseline characteristics data were similar across both groups.

High risk of bias

Adequate blinding method. The analysis to estimate the effect of the assignment was not accurate (per‐protocol analysis). The probable substantial impact of missing data on the result.

Low risk of bias

Outcome data available for all/nearly all patients randomised.

Low risk of bias

The outcome measurement was correct. The outcome assessors were also blind to their assigned intervention.

High risk of bias

Protocol or registry was not found, so information of the analysis plan was not available. A simple analysis was performed.

High risk of bias

A per‐protocol analysis and simple analysis was conducted. There was also a probable substantial impact of missing data on the result. Protocol deviations and reasons of patients who dropped the study were not reported, and the protocol or registry was not found.

Risk of bias for analysis 2.2 Incidence of serious adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Asilian 2020

Some concerns

There was no information on method of allocation concealment, but there was no baseline imbalance that would suggest a problem with randomisation.

Low risk of bias

The team and patients were probably blinded to the allocation group. The analysis was appropriate.

Low risk of bias

All randomised patients were included in the analysis.

Low risk of bias

There are no details on the method of measuring the outcomes. There is no evidence that the outcome assessment was done in a different way between groups. tthe outcome assessors were probably blinded to the allocation group.

Some concerns

The protocol was registered but no information was provided on how serious adverse events were measured. Due to the nature of the outcome, it is unlikely that there will be multiple measurement methods or multiple possibilities for analysis.

Some concerns

Lack of information about the allocation concealment process and how serious adverse events were measured.

Goyal 2000

Some concerns

No information on the method of randomisation. The authors took care in balancing groups.

Some concerns

Probably participants were aware of the intended intervention. The authors excluded from the analysis those participants that did not complete the follow‐up (around 1%). No reasons for participant losses were provided.

Low risk of bias

Three out of 15 participants with missing outcome data were excluded from the analysis: one and two from each group. This would not influence in the results.

Low risk of bias

Although there is no information about the method for the outcome assessment or the knowledge of the intervention by the assessors, the result may not be influenced by this.

Some concerns

No study protocol or analysis plan were available. However, the result would not be influenced by multiple eligible outcome measurements or analyses of the data.

Some concerns

No protocol is available. Lack of information on the method of randomisation, blinding of participants or study personnel.

Gupta 2019

Some concerns

Randomisation was done using a random number table but concealment allocation was not reported. Groups were comparable at baseline.

High risk of bias

No information about blinding of participants or people delivering the intervention. No deviations from the intended interventions were reported. A per‐protocol analysis was performed. Nine participants (18%) were lost at follow‐up.

High risk of bias

The percentage of losses were 20% and 16% for each group. Losses were related to logistic issues (changes in residence, contact tracing) but this information is not given by groups.

Low risk of bias

Although there is no information about the knowledge of the intervention by the assessors, the measurement of the outcome may not be influenced by this.

Some concerns

No protocol is available or analysis plan, but the measurement of side effects were well described.

High risk of bias

No protocol is available. Lack of information on the allocation concealment, blinding of participants or study personnel. The percentage of losses could influence the result.

Verma 2015

High risk of bias

The allocation was not concealed, because each patient was instructed to apply bimatoprost to one AA patch twice daily and to apply topical corticosteroid cream once daily on the other patch for a period of 3 months.

Some concerns

Each patient were aware of their assigned intervention. All assessments were performed by three blinded physicians. There was no information to support deviations from the intended interventions.

Low risk of bias

All randomised patients were analyzed.

Low risk of bias

A standardized method of evaluation was used by three blinded physicians.

Some concerns

There is no pre‐registered protocol available and there is no information on whether there were multiple eligible outcomes measurements or multiple analyzes of the data.

High risk of bias

Overall high due to the allocation sequence was not concealed and each patient were aware of their assigned intervention.

Risk of bias for analysis 3.2 Incidence of serious adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Rajan 2021

Low risk of bias

Randomization and allocation concealment were appropriate.

Some concerns

The care provider was not blinded.

Some concerns

Missing outcome data could bias the outcome results as 30 participants did not end the study. However, missingness in the outcome could be considered independent of their true value (serious adverse events as mortality).

Low risk of bias

The response assessor was blinded to the nature of intervention. Serious adverse events were probably not influenced by the knowledge of the intervention.

Some concerns

No protocol was available. It was reported that there were no serious adverse events

Some concerns

Participants and the outcome assessor were blinded to the intervention although the care provider was not. Serious adverse events (e.g. death) were probably not influenced by the knowledge of the intervention. Missing outcome data could bias the outcome results as 30 participants did not end the study. However, missingness in the outcome is not related to its true value as were consider serious adverse events as mortality.

Sousa 2020

Some concerns

There was no information on randomisation or allocation concealment. There were no imbalances between the groups.

High risk of bias

It was a double‐blind study, but there are no details of how the participants and people delivering the interventions were blinded. Details of the analysis of adverse events were not presented in the published protocol, study methods or results.

Low risk of bias

Data for this outcome were available for all randomised participants.

Low risk of bias

There were no information on how adverse events were evaluated in these patients The results were measured by an independent committee.

Some concerns

There is a registered protocol (REBEC: RBR‐5kyg2r), which did not include safety results. In the published article, the authors mentioned safety without specific details on methods or results.

High risk of bias

Overall due to discrepancy between the protocol and the final report on adverse events. The methods for measuring adverse events were not specified and their results were not presented in the published article.

Trink 2013

Some concerns

A randomised allocation table was used, but there was no information on allocation concealment. There was no baseline imbalance that would suggest a problem with randomisation.

Low risk of bias

Participants were not aware of the intervention received. The physicians injecting the test material were not blinded to the treatment modalities. The patients were analyzed in the group to which they were randomly assigned.

Low risk of bias

The data for this outcome were available for all participants randomised.

Low risk of bias

The SALT score was used in every intervention group. The personnel involved in the evaluation of the efficacy of treatment were not aware of the intervention received by study participants.

Some concerns

There is no pre‐registered protocol available and there is no information on whether there were multiple eligible outcomes measurements or multiple analyses of the data.

Some concerns

Some concerns in 2 domains due to the lack of detail on allocation concealment, and lack of information of pre‐ registered protocol.

Ustuner 2017

Some concerns

There was no information on allocation concealment but there were no baseline differences beyond that expected by chance.

High risk of bias

Participants were not aware of the intervention received, but the physicians injecting the test material were not blinded to the treatment modalities. There was no complete information about the analysis (placebo group).

High risk of bias

The results at T3 (12 months) were informed just for Platelet‐rich plasma (PRP) but not for placebo.

Low risk of bias

The SALT score was used in every intervention group The personnel involved in the evaluation of efficacy of treatment were not aware of the intervention received by study participants.

Some concerns

There is no pre‐registered protocol available and there is no information on whether there were multiple eligible outcomes measurements or multiple analyses of the data.

High risk of bias

High risk of bias due to not reported placebo group results plus lack of information about allocation concealment.

Risk of bias for analysis 4.2 Incidence of serious adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Guttman‐Yassky 2021

Low risk of bias

There was no issues on randomisation process and there was no baseline imbalance

Low risk of bias

There were no deviations from intended interventions

Low risk of bias

All randomised patients were included in the analysis

Low risk of bias

Outcome measurement was appropriate

Low risk of bias

No biases were detected in the selection of the reported results

Low risk of bias

There is a low risk of bias from the randomisation process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of reported results.

Risk of bias for analysis 5.2 Incidence of serious adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

King 2022a

Low risk of bias

The randomisation process was adequate

Low risk of bias

Patients and carers were probably not aware of the assigned interventions and an appropriate analysis was made to estimate the effect of assignment

Low risk of bias

Authors presented data for nearly all randomised patients.

Low risk of bias

Methods were adequate and the outcome assessors were probably not aware of the interventions.

Low risk of bias

There was no evidence of selection of the reported results.

Low risk of bias

There is a low risk of bias from the randomisation process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of reported results.

King 2022b

Low risk of bias

The randomisation process was adequate.

Low risk of bias

Patients and carers were probably not aware of the assigned interventions and an appropriate analysis was made to estimate the effect of assignment.

Low risk of bias

Authors presented data for nearly all randomised patients

Low risk of bias

Methods were adequate and the outcome assessors were probably not aware of the interventions

Low risk of bias

There was no evidence of selection of the reported results.

Low risk of bias

There is a low risk of bias from the randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of reported results.

Mikhaylov 2019

Low risk of bias

Randomisation was performed using a computer‐generated randomisation list with a block of eight patients in a 1:1 ratio. Treatment allocation concealment was obtained using a phone‐call centralized procedure for each eligible patient. No baseline imbalance would suggest a problem with randomisation.

Low risk of bias

It was a double‐blind, placebo‐controlled study, and the team and patients were blinded to the allocation group. The analysis was appropriate.

Low risk of bias

All randomised patients were included in the analysis.

Low risk of bias

It was a double‐blind, placebo‐controlled study, and the team and patients were blinded to the allocation group. Adverse events were evaluated at each visit by a professional who did not know the intervention group.

Some concerns

There is no pre‐registered protocol available and there is no information on whether there were multiple eligible outcomes measurements or multiple analyses of the data.

Some concerns

Some concerns in 2 domains are due to the lack of detail on randomisation and allocation concealment, and lack of information on pre‐registered protocol.

Risk of bias for analysis 6.2 Incidence of serious adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Almutairi 2019

Some concerns

No information on the method of randomisation. There were no differences between baseline participant characteristics.

Some concerns

Participants and study personnel were aware of the intended interventions. The analysis plan was not appropriate as participants who were non‐adherents with the intervention were withdrawn.

Low risk of bias

Around 6% of randomised participants withdrew from the study due to non‐compliance. No details about lack of compliance for excluded participants. Missing outcome data were balanced between groups.

Low risk of bias

Details about monitoring of side effects were not reported in methods. Outcome assessors were aware of the delivered intervention, but serious AE could be recorded independently of the knowledge of the participant allocation.

Some concerns

No information about the study protocol. However, the authors reported any side effects and presented the AE frequency at the end of the trial.

Some concerns

No study protocol is available. Lack of information on the method of randomisation. Participants and study personnel were aware of the intended interventions. Non‐adherent participants were withdrawn from the analysis. However, the authors reported any side effects and serious AE could be recorded independently of the knowledge of the participant allocation.

Risk of bias for analysis 7.2 Incidence of serious adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Olsen 2019

Low risk of bias

Interactive Response Technology was used for patient study number assignment and randomisation. There was no baseline imbalance that would suggest a problem with randomisation

Low risk of bias

It was a placebo‐controlled study. The authors declare a double‐blind study. The team and patients were probably blinded to the allocation group. The analysis was appropriate.

Low risk of bias

All randomised patients were included in the analysis.

Some concerns

There are no details on the method to assess adverse events in the treatment arms. The authors declare a double‐blind study. The team and patients were blinded to the allocation group.

Low risk of bias

The registered protocol includes the reporting of adverse events as published in the reviewed article.

Some concerns

Some concerns in 2 domains due to the lack of detail on randomisation and allocation concealment, and lack of information on the method to assess adverse events in the treatment arms.

Risk of bias for analysis 10.2 Incidence of serious adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Shapiro 1993

High risk of bias

There was no information on the method of randomisation and allocation concealment. There was baseline imbalance in some characteristics that would suggest a problem with randomisation.

High risk of bias

The information includes the statement "double‐blind manner" but without specific details about it. There were no deviations from intended interventions. The analysis was per‐protocole, and due to the small sample size, two patients who were not included in the analysis could affect the final estimate.

Some concerns

Two patients dropped out of the study due to an adverse event (generalized dermatitis). The occurrence of SAE could not be established in two patients who were not included in the analysis, and due to the small sample size it could affect the final estimate.

Some concerns

There were no details on the method of measuring the outcome, but according to the information provided, it is understood that the same outcome was evaluated in both intervention groups applying the same methodology. There was no information on whether the outcome assessors were aware of the intervention received, but knowledge of the intervention received could have affected the measurement of the results.

Some concerns

There is no pre‐registered protocol available and there is no information on whether there were multiple eligible outcomes measurements or multiple analyzes of the data.

High risk of bias

Overall high due to the lack of information on randomisation, baseline imbalances, lack of information on whether outcome assessors were aware of the intervention received, and missing outcome data without an appropriate analysis. In addition there was no pre‐registered protocol available.

Risk of bias for analysis 11.2 Incidence of serious adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Fransway 1988

Some concerns

Insufficient information about randomisation process and balance between baseline characteristics for participants.

Some concerns

Although participants were unaware of the intervention there is no information about if cares were too. Only participants who completed the study were included in the analysis.

High risk of bias

Lack of information about the reasons for missing outcome data.

Some concerns

Insufficient information about measuring side effects. However, assessment of serious side effects probably was not affected by information regarding assigned intervention.

Some concerns

No protocol was available. No information in the method section about the AE measurements or analysis.

High risk of bias

No study protocol is available. Lack of information about the randomisation process, blinding of outcome assessors, reasons for missing outcome data and outcome measurement and plan analysis.

Luo 1991

Some concerns

There is no information about the randomisation process. However, there are no significant differences in baseline characteristics.

Low risk of bias

Participants and study personnel were blinded to the intervention. All participants were analysed in the group as they have been randomised.

Low risk of bias

There are no missing outcome data.

Low risk of bias

Although the outcome assessment involved visual observation, the outcome assessors were blind to the intervention.

Some concerns

No protocol available or analysis plan. Outcome results were presumptively reported for the end of the study (12 weeks), and from a single analysis.

Some concerns

No protocol is available. Lack of information about the randomisation process and outcome results were presumptively reported for the end of the study (12 weeks) and from a single analysis.

Price 1987a

Some concerns

There was no information on randomisation or allocation concealment, but there was no baseline imbalance that would suggest a problem with randomisation.

High risk of bias

There was not performed an appropriate data analysis and losses to follow up in the minoxidil group could affect the final estimate.

High risk of bias

Losses were greater in the minoxidil group (4/15 = 26%) compared to placebo (1/15 = 6%). The reasons of participant's losses were not given.

Low risk of bias

It was used an standard scale of hair regrowth in both treatment arms. All the follow‐up evaluations were done by a dermatologist who was blinded to the active and control groups.

Some concerns

There is no pre‐registered protocol available but in methods section there are a full description of adverse events evaluation and in results the adverse event findings were appropriately reported.

High risk of bias

Overall high due to missing outcome data without explanation and proper analysis.

Price 1987b

Some concerns

There was no information on randomisation or allocation concealment, but there was no baseline imbalance that would suggest a problem with randomisation.

Some concerns

There was no information on randomisation or allocation concealment, but there was no baseline imbalance that would suggest a problem with randomisation.

Low risk of bias

Data for this outcome were available for all randomised participants.

Low risk of bias

A complete clinical examination and a full panel of studies were performed to evaluate SAE in both groups. It was a double blind study and probably the outcome assessors were not aware of the intervention received.

Some concerns

There is no pre‐registered protocol available but in methods section there are a full description of adverse events evaluation and in results the adverse event findings were appropriately reported.

Some concerns

Overall due to a lack of detail on randomisation, and no pre‐registered protocol.

Risk of bias for analysis 12.2 Incidence of serious adverse events.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Nouh 2022

Some concerns

Insufficient information about the randomisation process.

Some concerns

Participants and carers were aware of the assigned interventions and protocol deviations were not reported.

Low risk of bias

There were no missing outcome data.

High risk of bias

Insufficient information for measuring of the outcome. Reporting could have been influenced by the knowledge of the intended intervention.

Some concerns

There is not enough information to know if the serious adverse events rate was biased by any selection of the results as there is no analysis plan for it.

High risk of bias

Insufficient information about the randomisation process. Participants and carers were aware of the assigned interventions and protocol deviations were not reported. Insufficient information for measuring the outcome. Reporting could have been influenced by the knowledge of the intended interventions.

For the outcome of health‐related quality of life, the ROB assessment was judged as high risk in one study from one comparison (Lai 2019) due to the substantial impact of missing data on the outcome result (Table 92).

Risk of bias for analysis 1.3 Health‐related quality of life.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Lai 2019

Low risk of bias

Blinding of participants, study researchers and outcome assessors were blinded to the allocation sequence. Baseline characteristics data were similar across both groups.

High risk of bias

Adequate blinding method. The analysis to estimate the effect of the assignment was not accurate (per‐protocol analysis). The probable substantial impact of missing data on the result.

Low risk of bias

Outcome data available for all/nearly all patients randomised.

Low risk of bias

The outcome measurement was correct. The outcome assessors were also blind to their assigned intervention.

High risk of bias

Protocol or registry was not found, so information of the analysis plan was not available. A simple analysis was performed.

High risk of bias

A per‐protocol analysis and simple analysis was conducted. There was also a probable substantial impact of missing data on the result. Protocol deviations and reasons of patients who dropped the study were not reported, and the protocol or registry was not found.

Two studies (King 2022a; King 2022b) from one comparison assessed long‐term hair regrowth ≥ 75% and the ROB assessment was judged as low (Table 101).

Risk of bias for analysis 5.3 Long‐term hair regrowth ≥ 75%.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

King 2022a

Low risk of bias

The randomisation process was adequate

Low risk of bias

Patients and carers were probably not aware of the assigned interventions and an appropriate analysis was made to estimate the effect of assignment

Low risk of bias

Authors presented data for nearly all randomised patients.

Low risk of bias

Methods were adequate and the outcome assessors were probably not aware of the interventions.

Low risk of bias

Results for the outcomes of interest were presented according to the protocol, without selection.

Low risk of bias

There is a low risk of bias from the randomisation process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of reported results.

King 2022b

Low risk of bias

The randomisation process was adequate.

Low risk of bias

Patients and carers were probably not aware of the assigned interventions and an appropriate analysis was made to estimate the effect of assignment.

Low risk of bias

Authors presented data for all randomised patients.

Low risk of bias

Methods were adequate and the outcome assessors were probably not aware of the interventions.

Low risk of bias

Results for the outcomes of interest were presented according to the protocol, without selection

Low risk of bias

There is a low risk of bias from the randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of reported results

The ROB assessments and judgement support for the non‐prioritised comparisons and their outcomes are available here: https://osf.io/rd7cg/?view_only=9017f5a8fe284b1dab83b5d7d194bf51.

Effects of interventions

See: Table 1; Table 2; Table 3; Table 4; Table 5; Table 6; Table 7; Table 8; Table 9; Table 10; Table 11; Table 12

As mentioned in the methods section, the NMA was not performed. The geometry of the treatment comparisons is depicted in a network plot (Figure 2).

2.

Prioritised comparisons

Considering the outcomes of interest for this review and the prioritised interventions, we pooled the data of those interventions belonging to the same pharmacological group and administration route. We obtained 12 prioritised comparisons: seven interventions compared to placebo (including 16 studies) and five interventions compared to another different intervention (including 8 studies). One study assessed had four different interventions, so it was included in three different comparisons (Tosti 1986).

Comparison 1: Oral classical immunosuppressants versus placebo

• Cyclosporine versus placebo (n = 1) (Lai 2019)

• Prednisolone versus placebo (n = 1) (Kar 2005)

Comparison 2: Oral classical immunosuppressants versus oral classical immunosuppressants

• Betamethasone versus azathioprine (n = 2) (Gupta 2019, Verma 2015)

• Betamethasone versus methotrexate (n = 1) (Asilian 2020)

• Betamethasone versus prednisolone (n = 1) (Goyal 2000)

Comparison 3: Intralesional classical immunosuppressants versus placebo

• Triamcinolone versus placebo (n = 4) (Rajan 2021; Sousa 2020; Trink 2013; Ustuner 2017)

• Betamethasone versus placebo (n = 2) (Sousa 2020; Ustuner 2017)

Comparison 4: Subcutaneous biologics versus placebo

• Dupilumab versus placebo (n = 1) (Guttman‐Yassky 2021)

Comparison 5: Oral small molecule inhibitors versus placebo

• Baricitinib versus placebo (n = 2) (King 2022a; King 2022b)

• Apremilast versus placebo (n = 1) (Mikhaylov 2019)

Comparison 6: Oral small molecule inhibitors versus oral small molecule inhibitors

• Ruxolitinib versus tofacitinib (n=1) (Almutairi 2019)

Comparison 7: Topical small molecule inhibitors versus placebo

• Ruxolitinib versus placebo (n = 1) (Olsen 2019)

Comparison 8: Contact immunotherapy versus placebo

• Squaric acid dibutyl ester or diphencyprone versus placebo (n = 1) (Tosti 1986)

Comparison 9: Contact immunotherapy versus topical hair growth stimulants

• Squaric acid dibutyl ester or diphencyprone versus topical minoxidil (n = 1) (Tosti 1986)

Comparison 10: Topical hair growth stimulants + contact immunotherapy versus contact immunotherapy

• Diphencyprone + topical minoxidil versus diphencyprone (n = 1) (Shapiro 1993)

Comparison 11: Topical hair growth stimulants versus placebo

• Minoxidil versus placebo (n = 5) (Fransway 1988; Luo 1991; Price 1987a; Price 1987b; Tosti 1986)

Comparison 12: Other therapies

• Cryotherapy versus fractional CO₂ laser (n = 1) (Nouh 2022)

The results for each prioritised comparison are presented below.

1. Oral classical immunosuppressants versus placebo

1.1 Short‐term hair regrowth ≥ 75%

Two studies compared oral prednisolone (Kar 2005) and cyclosporine (Lai 2019) versus placebo. The evidence is very uncertain about the effect of classical immunosuppressants when compared to placebo on short‐term hair regrowth ≥ 75% (risk ratio [RR] 4.68, 95% confidence interval [CI] 0.57 to 38.27; 79 participants; 2 studies; very low‐certainty evidence) (Analysis 1.1; Table 1).

1.1. Analysis.

Comparison 1: Oral classical immunosuppressants vs placebo, Outcome 1: Short‐term hair regrowth ≥ 75%

1.2 Incidence of serious adverse events

Two studies evaluated the occurrence of serious adverse events and reported zero events (Kar 2005; Lai 2019). The evidence is very uncertain about the effect of classical immunosuppressants (prednisolone or azathioprine) when compared to placebo on the incidence of serious adverse events (RR not estimable, no events observed; 79 participants; 2 studies; very low‐certainty evidence) (Analysis 1.2; Table 1).

1.2. Analysis.

Comparison 1: Oral classical immunosuppressants vs placebo, Outcome 2: Incidence of serious adverse events

1.3 Health‐related quality of life

Only one study evaluated the effect on health‐related quality of life (HRQoL), measured with the AQoL‐8D score (Lai 2019). The evidence is very uncertain about the effects of classical immunosuppressants (cyclosporine) when compared to placebo on health‐related quality of life (mean difference [MD] 0.01, 95% CI ‐0.04 to 0.07; 36 participants; 1 study; very low‐certainty evidence) (Analysis 1.3; Table 1).

1.3. Analysis.

Comparison 1: Oral classical immunosuppressants vs placebo, Outcome 3: Health‐related quality of life

1.4 Long‐term hair regrowth ≥ 75%

There were no studies reporting this outcome.

2. Oral classical immunosuppressants versus oral classical immunosuppressants

2.1 Short‐term hair regrowth ≥ 75%

Two studies compared oral betamethasone versus oral prednisolone (Goyal 2000) or azathioprine (Gupta 2019). The evidence suggests oral classical immunosuppressants may increase short‐term hair regrowth by ≥ 75% when compared to oral classical immunosuppressants (RR 1.67, 95% CI 0.96 to 2.88; 80 participants; 2 studies; low‐certainty evidence) (Analysis 2.1; Table 2).

2.1. Analysis.

Comparison 2: Oral classical immunosuppressants vs oral classical immunosuppressants, Outcome 1: Short‐term hair regrowth ≥ 75%

2.2 Incidence of serious adverse events

Four studies compared betamethasone versus methotrexate (Asilian 2020), prednisolone (Goyal 2000), or azathioprine (Gupta 2019; Verma 2015) and reported zero serious adverse events. The evidence is very uncertain about the effect of oral classical immunosuppressants when compared to oral classical immunosuppressants on the incidence of serious adverse events (RR not estimable, no events observed; 154 participants; 4 studies; very low‐certainty evidence) (Analysis 2.2; Table 2).

2.2. Analysis.

Comparison 2: Oral classical immunosuppressants vs oral classical immunosuppressants, Outcome 2: Incidence of serious adverse events

2.3 Health‐related quality of life

There were no studies reporting this outcome.

2.4 Long‐term hair regrowth ≥ 75%

There were no studies reporting this outcome.

3. Intralesional classical immunosuppressants versus placebo

3.1 Short‐term hair regrowth ≥ 75%

One study compared intralesional betamethasone or triamcinolone versus placebo (Ustuner 2017). The evidence is very uncertain about the effect of intralesional classical immunosuppressants when compared to placebo on short‐term hair regrowth ≥ 75% (RR 13.84, 95% CI 0.87 to 219.76; 231 participants; 1 study; very low‐certainty evidence) (Analysis 3.1; Table 3).

3.1. Analysis.

Comparison 3: Intralesional classical immunosuppressants vs placebo, Outcome 1: Short‐term hair regrowth ≥ 75%

3.2 Incidence of serious adverse events

Four studies compared intralesional betamethasone or triamcinolone versus placebo (Rajan 2021; Sousa 2020; Trink 2013; Ustuner 2017). The evidence is very uncertain about the effect of intralesional classical immunosuppressants when compared to placebo on the incidence of serious adverse events (RR not estimable, no events observed; 543 participants; 4 studies; very‐low certainty) (Analysis 3.2; Table 3).

3.2. Analysis.

Comparison 3: Intralesional classical immunosuppressants vs placebo, Outcome 2: Incidence of serious adverse events

3.3 Health‐related quality of life

There were no studies reporting this outcome.

3.4 Long‐term hair regrowth ≥ 75%

One study compared intralesional triamcinolone versus placebo in a total sample of 30 participants and reported that four of 15 (26.7%) participants achieved the outcome (Trink 2013). However, the authors did not provide the number of events in the placebo group and no further analysis could be performed.

4. Subcutaneous biologics versus placebo

4.1 Short‐term hair regrowth ≥ 75%

One study compared subcutaneous dupilumab versus placebo (Guttman‐Yassky 2021). The evidence suggests subcutaneous biologics result in little to no difference in short‐term hair regrowth ≥ 75% when compared to placebo (RR 3.59, 95% CI 0.19 to 66.22; 60 participants; 1 study; low‐certainty evidence) (Analysis 4.1; Table 4).

4.1. Analysis.

Comparison 4: Subcutaneous biologics vs placebo, Outcome 1: Short‐term hair regrowth ≥ 75%

4.2 Incidence of serious adverse events

One study evaluated the occurrence of serious adverse events and reported one event in the dupilumab arm (Guttman‐Yassky 2021). The evidence suggests that subcutaneous biologics may not increase the incidence of serious adverse events when compared to placebo (RR 1.54, 95% CI 0.07 to 36.11; 60 participants; 1 study; low‐certainty evidence) (Analysis 4.2; Table 4).

4.2. Analysis.

Comparison 4: Subcutaneous biologics vs placebo, Outcome 2: Incidence of serious adverse events

4.3 Health‐related quality of life

There were no studies reporting this outcome.

4.4 Long‐term hair regrowth ≥ 75%

There were no studies reporting this outcome.

5. Oral small molecule inhibitors versus placebo

5.1 Short‐term hair regrowth ≥ 75%

Two identically designed trials (BRAVE‐AA1 and BRAVE‐AA2) compared oral baricitinib versus placebo (King 2022a; King 2022b). The populations for the two studies had identical eligibility criteria and identical primary and secondary outcomes. The reported risk ratio is the pooled result estimated from the two studies. Oral small molecule inhibitors result in an increase in short‐term hair regrowth ≥ 75% when compared to placebo (RR 7.54, 95% CI 3.90 to 14.58; 1200 participants; 2 studies; high‐certainty evidence) (Analysis 5.1; Table 5).

5.1. Analysis.

Comparison 5: Oral small molecule inhibitors vs placebo, Outcome 1: Short‐term hair regrowth ≥ 75%

5.2 Incidence of serious adverse events

Two studies compared baricitinib versus placebo (King 2022a; King 2022b), and one study compared apremilast versus placebo and reported zero serious adverse events (Mikhaylov 2019). The authors mentioned that there were no deaths, opportunistic infections, thromboembolic events, or gastrointestinal perforations. The evidence suggests oral small molecule inhibitors do not increase the incidence of serious adverse events when compared to placebo (RR 1.47, 95% CI 0.60 to 3.60; 1224 participants; 3 studies; low‐certainty evidence) (Analysis 5.2; Table 5).

5.2. Analysis.

Comparison 5: Oral small molecule inhibitors vs placebo, Outcome 2: Incidence of serious adverse events

5.3 Long‐term hair regrowth ≥ 75%

Two studies assessed long‐term hair regrowth ≥ 75% (King 2022a; King 2022b). Oral small molecule inhibitors result in an increase in long‐term hair regrowth ≥ 75% when compared to placebo (RR 8.49, 95% CI 4.70 to 15.34; 1200 participants; 2 studies; high‐certainty evidence) (Analysis 5.3; Table 5).

5.3. Analysis.

Comparison 5: Oral small molecule inhibitors vs placebo, Outcome 3: Long‐term hair regrowth ≥ 75%

5.4 Health‐related quality of life

There were no studies reporting this outcome.

6. Oral small molecule inhibitors versus oral small molecule inhibitors

6.1 Short‐term hair regrowth ≥ 75%

One study compared oral ruxolitinib versus tofacitinib in a sample of 80 participants (Almutairi 2019). The evidence is very uncertain about the effect of ruxolitinib versus tofacitinib on short‐term hair regrowth ≥ 75% (RR 1.08, 95% CI 0.77 to 1.52; 80 participants; 1 study; very‐low certainty evidence) (Analysis 6.1; Table 6).

6.1. Analysis.

Comparison 6: Oral small molecule inhibitors vs oral small molecule inhibitors, Outcome 1: Short‐term hair regrowth ≥ 75%

6.2 Incidence of serious adverse events

One study compared ruxolitinib versus tofacitinib (Almutairi 2019) and reported zero serious adverse events. The evidence is very uncertain about the effect of ruxolitinib versus tofacitinib on serious adverse events (RR not estimable, no events observed; 80 participants; 1 study; very low‐certainty evidence) (Analysis 6.2; Table 6).

6.2. Analysis.

Comparison 6: Oral small molecule inhibitors vs oral small molecule inhibitors, Outcome 2: Incidence of serious adverse events

6.3 Health‐related quality of life

There were no studies reporting this outcome.

6.4 Long‐term hair regrowth ≥ 75%

There were no studies reporting this outcome.

7. Topical small molecule inhibitors versus placebo

7.1 Short‐term hair regrowth ≥ 75%

One study compared topical ruxolitinib versus placebo (Olsen 2019). The evidence suggests topical small molecule inhibitors may result in little to no difference in short‐term hair regrowth ≥ 75% when compared to placebo (RR 5.00, 95% CI 0.25 to 100.89; 78 participants; 1 study; low‐certainty evidence) (Analysis 7.1; Table 7).

7.1. Analysis.

Comparison 7: Topical small molecule inhibitors vs placebo, Outcome 1: Short‐term hair regrowth ≥ 75%

7.2 Incidence of serious adverse events

One study evaluated the occurrence of serious adverse events and reported one event in the placebo arm (Olsen 2019). The evidence is very uncertain about the effect of topical small molecule inhibitors when compared to placebo on the incidence of serious adverse events (RR 0.33, 95% CI 0.01 to 7.94; 78 participants; 1 study; very low‐certainty evidence) (Analysis 7.2; Table 7).

7.2. Analysis.

Comparison 7: Topical small molecule inhibitors vs placebo, Outcome 2: Incidence of serious adverse events

7.3 Health‐related quality of life

There were no studies reporting this outcome.

7.4 Long‐term hair regrowth ≥ 75%

There were no studies reporting this outcome.

8. Contact immunotherapy versus placebo

8.1 Short‐term hair regrowth ≥ 75%

One study compared diphencyprone and squaric acid dibutyl ester versus placebo (Tosti 1986). The evidence is very uncertain about the effect of contact immunotherapy when compared to placebo on short‐term hair regrowth ≥ 75% (RR 1.16, 95% CI 0.79 to 1.71; 99 participants; 1 study; very‐low certainty evidence) (Analysis 8.1; Table 8).

8.1. Analysis.

Comparison 8: Contact immunotherapy vs placebo, Outcome 1: Short‐term hair regrowth ≥ 75%

8.2 Incidence of serious adverse events

There were no studies reporting this outcome.

8.3 Health‐related quality of life

There were no studies reporting this outcome.

8.4 Long‐term hair regrowth ≥ 75%

There were no studies reporting this outcome.

9. Contact immunotherapy versus topical hair growth stimulants

9.1 Short‐term hair regrowth ≥ 75%

One study compared diphencyprone and squaric acid dibutyl ester versus topical minoxidil (Tosti 1986). The evidence is very uncertain about the effect of contact immunotherapy when compared to topical hair growth stimulants on short‐term hair regrowth ≥ 75% (RR 1.16, 95% CI 0.79 to 1.71; 99 participants; 1 study; very low‐certainty evidence) (Analysis 9.1; Table 9).

9.1. Analysis.

Comparison 9: Contact immunotherapy vs topical hair growth stimulants, Outcome 1: Short‐term hair regrowth ≥ 75%

9.2 Incidence of serious adverse events

There were no studies reporting this outcome.

9.3 Health‐related quality of life

There were no studies reporting this outcome.

9.4 Long‐term hair regrowth ≥ 75%

There were no studies reporting this outcome.

10. Contact immunotherapy + topical hair growth stimulants versus contact immunotherapy

10.1 Short‐term hair regrowth ≥ 75%

One study compared diphencyprone plus topical minoxidil versus diphencyprone (Shapiro 1993). The evidence is very uncertain about the effect of contact immunotherapy plus topical hair growth stimulants when compared to contact immunotherapy on short‐term hair regrowth ≥ 75% (RR 0.67, 95% CI 0.13 to 3.44; 30 participants; 1 study; very low‐certainty evidence) (Analysis 10.1; Table 10).

10.1. Analysis.

Comparison 10: Contact immunotherapy + hair growth stimulants vs contact immunotherapy, Outcome 1: Short‐term hair regrowth ≥ 75%

10.2 Incidence of serious adverse events

One study evaluated the occurrence of serious adverse events and reported zero events (Shapiro 1993). The evidence is very uncertain about the effect of contact immunotherapy plus topical hair growth stimulants when compared to contact immunotherapy on the incidence of serious adverse events (RR not estimable, no events observed; 30 participants; 1 study; very low‐certainty evidence) (Analysis 10.2; Table 10).

10.2. Analysis.

Comparison 10: Contact immunotherapy + hair growth stimulants vs contact immunotherapy, Outcome 2: Incidence of serious adverse events

10.3 Health‐related quality of life

There were no studies reporting this outcome.

10.4 Long‐term hair regrowth ≥ 75%

There were no studies reporting this outcome.

11. Topical hair growth stimulants versus placebo

11.1 Short‐term hair regrowth ≥ 75%

Two studies compared minoxidil 1% (Tosti 1986) and minoxidil 2% (Luo 1991) versus placebo. The evidence is very uncertain about the effect of topical hair growth stimulants when compared to placebo on short‐term hair regrowth ≥ 75% (RR 2.31, 95% CI 1.34 to 3.96; 202 participants; 2 studies; very low‐certainty evidence) (Analysis 11.1; Table 11). However, it is worth noting that the pooled analysis indicated high statistical heterogeneity (I‐squared = 93%). The estimation of between‐study variance was based on only two studies, making this estimation very unreliable. Of the two studies, only minoxidil 2% (Luo 1991) showed a significant effect on this outcome. When using a random‐effects model, the estimation of the effect of the intervention was similar to the one obtained with a fixed‐effect model but now very imprecise, with a very wide confidence interval which included the null (random‐effects model, RR 2.75, 95% CI 0.17 to 43.46).

11.1. Analysis.

Comparison 11: Topical hair growth stimulants vs placebo, Outcome 1: Short‐term hair regrowth ≥ 75%

11.2 Incidence of serious adverse events

Four studies compared minoxidil 3% (Fransway 1988; Price 1987a; Price 1987b) and minoxidil 2% (Luo 1991) versus placebo and reported zero serious adverse events. The evidence is very uncertain about the effect of topical hair growth stimulants when compared to placebo on the incidence of serious adverse events (RR not estimable, no events observed; 304 participants; 4 studies; very low‐certainty evidence) (Analysis 11.2; Table 11).

11.2. Analysis.

Comparison 11: Topical hair growth stimulants vs placebo, Outcome 2: Incidence of serious adverse events

11.3 Health‐related quality of life

There were no studies reporting this outcome.

11.4 Long‐term hair regrowth ≥ 75%

There were no studies reporting this outcome.

12. Other therapies

12.1 Short‐term hair regrowth ≥ 75%

One study compared cryotherapy versus fractional CO₂ laser (Nouh 2022). The evidence is very uncertain about the effect of cryotherapy when compared to fractional CO₂ laser on short‐term hair regrowth ≥ 75% (RR 0.31, 95% CI 0.11 to 0.86; 80 participants; 1 study; very low‐certainty evidence) (Analysis 12.1; Table 12).

12.1. Analysis.

Comparison 12: Other therapies, Outcome 1: Short‐term hair regrowth ≥ 75%

12.2 Incidence of serious adverse events

One study evaluated the occurrence of serious adverse events and reported zero serious adverse events (Nouh 2022). The evidence is very uncertain about the effect of cryotherapy when compared to fractional CO₂ laser on serious adverse events (RR not estimable, no events observed; 80 participants; 1 study; very low‐certainty evidence) (Analysis 12.2; Table 12).

12.2. Analysis.

Comparison 12: Other therapies, Outcome 2: Incidence of serious adverse events

12.3 Health‐related quality of life

There were no studies reporting this outcome.

12.4 Long‐term hair regrowth ≥ 75%

There were no studies reporting this outcome.

Non‐prioritised comparisons

The following link provides access to the results of the effects of the non‐prioritised comparisons and the corresponding summary of findings tables:

https://osf.io/rd7cg/?view_only=9017f5a8fe284b1dab83b5d7d194bf51

13. Sensitivity analysis

Since the number of studies per comparison was insufficient, the sensitivity analysis that we planned to evaluate the impact of risk of bias assessments on the effects of the interventions was not performed.

We have run a sensitivity analysis to assess the impact of imputing the result of short‐ and long‐term hair regrowth ≥ 75% as responders to the participants lost to follow‐up in the control group of each trial (Analysis 13.1; Analysis 13.2; Analysis 13.3; Analysis 13.4; Analysis 13.5; Analysis 13.6; Analysis 13.7; Analysis 13.8; Analysis 13.9; Analysis 13.10; Analysis 13.11; Analysis 13.12). This change in the assumption made for imputation of lost‐to‐follow‐up participants did not alter the conclusions of the effects of the interventions previously estimated in the main analysis.

13.1. Analysis.

Comparison 13: Sensitivity analysis, Outcome 1: Sensitivity analysis. Oral classical immunosuppressants vs placebo: short‐term hair regrowth

13.2. Analysis.

Comparison 13: Sensitivity analysis, Outcome 2: Sensitivity analysis. Betamethasone vs azathioprine or prednisolone: short‐term hair regrowth

13.3. Analysis.

Comparison 13: Sensitivity analysis, Outcome 3: Sensitivity analysis. Intralesional classical immunosuppressants vs placebo: short‐term hair regrowth

13.4. Analysis.

Comparison 13: Sensitivity analysis, Outcome 4: Sensitivity analysis. Subcutaneous biologics vs placebo: short‐term hair regrowth

13.5. Analysis.

Comparison 13: Sensitivity analysis, Outcome 5: Sensitivity analysis. Oral small molecule inhibitors: baricitinib vs placebo

13.6. Analysis.

Comparison 13: Sensitivity analysis, Outcome 6: Sensitivity analysis. Oral small molecule inhibitors: ruxolitinib vs tofacitinib

13.7. Analysis.

Comparison 13: Sensitivity analysis, Outcome 7: Sensitivity analysis. Topical small molecule inhibitors: short‐term hair regrowth

13.8. Analysis.

Comparison 13: Sensitivity analysis, Outcome 8: Sensitivity analysis. Contact immunotherapy vs placebo: short‐term hair regrowth

13.9. Analysis.

Comparison 13: Sensitivity analysis, Outcome 9: Sensitivity analysis. Contact immunotherapy vs minoxidil: short‐term hair regrowth

13.10. Analysis.

Comparison 13: Sensitivity analysis, Outcome 10: Sensitivity analysis. Contact immunotherapy + hair growth stimulants vs contact immunotherapy: short term hair regrowth

13.11. Analysis.

Comparison 13: Sensitivity analysis, Outcome 11: Sensitivity analysis. Topical hair growth stimulants vs placebo: short‐term hair‐regrowth

13.12. Analysis.

Comparison 13: Sensitivity analysis, Outcome 12: Sensitivity analysis. Cryotherapy vs fractional CO₂ laser: short‐term hair regrowth

14. Subgroup analysis

We did not undertake any subgroup analysis as we did not have enough studies to perform it. We could not analyse the heterogeneity induced by the age of the participants nor the extent of hair loss or the type of alopecia.

Discussion

Summary of main results

Amongst the interventions evaluated, only baricitinib compared to placebo resulted in a beneficial impact on short and long‐term hair regrowth with high certainty of evidence. However, we found inconclusive results for the risk of serious adverse effects with baricitinib compared to placebo with low certainty of evidence. No additional studies assessed long‐term hair regrowth.

Minoxidil compared to placebo may increase short‐term hair regrowth, although the certainty of the evidence is very low. The remaining interventions showed little to no impact on short‐term hair regrowth with low‐ or very‐low certainty evidence. In general, the studies had small sample sizes, which is reflected in the low precision of the effect estimates. The evidence is uncertain about the effect on the incidence of serious adverse events of the remaining interventions.

Only one study assessed health‐related quality of life comparing cyclosporine against placebo, reporting inconclusive results with very low certainty of evidence.

Overall completeness and applicability of evidence

Delamere 2008 included 17 clinical trials, while we have included 63 studies in this new review. We have identified studies on local treatments including hair growth stimulants, intralesional therapies, contact immunotherapy, cryotherapy, and a range of systemic treatments from corticosteroids to small molecule inhibitors. All this body of evidence allows professionals to know the existing landscape of treatment alternatives in AA. Our review makes it possible to elucidate the existence of treatments that, compared to others, appear to be superior in terms of efficacy with hair regrowth results greater than 75% and with adequate safety profiles. It is worth noting that most of these comparisons showed great imprecision in estimating the effects. This is the case in oral corticosteroids (Kar 2005; Lai 2019) and intralesional corticosteroids (Ustuner 2017) compared to placebo. These therapies are available in the clinical context and are routinely used by professionals involved in the management of patients with AA.

Despite this important advance in the available evidence, we have found great heterogeneity in the endpoints used to evaluate the effectiveness of the treatments, making it extremely difficult to compare the therapeutical alternatives. This review has raised the need for a consensus regarding the appropriate core outcome measures of efficacy for clinical trials in AA, and the minimum follow‐up time for these participants. Notwithstanding the fact that AA is a disease that affects the quality of life of patients who suffer from it, studies that have assessed this outcome are scarce. So, it is important to also agree on the inclusion of QoL outcomes into the standardised core outcomes set for AA. In our opinion, it could be a matter of criticism that there is a lack of replication of the many studies that evaluated the efficacy of interventions for AA. Only a few of these studies were designed to replicate previous findings. It is necessary to design clinical studies with adequate quality standards, which include a sufficient number of participants that allow valid and accurate estimates of the efficacy of the therapies that have previously shown promising results.

Quality of the evidence

In general, the certainty of the evidence for short‐term hair regrowth was rated as low or very low, mainly due to the high risk of bias and imprecision, except for the assessment of baricitinib versus placebo, where the certainty of the evidence for short and long‐term hair regrowth was high. The certainty of the evidence for the outcome of serious adverse events was rated as low due to very serious imprecision. Only one study assessed quality of life (the comparison of cyclosporine versus placebo), and the certainty of the evidence was rated as very low.

Regarding the risk of bias, the most relevant aspects were related to the lack of details about the randomisation procedure and the allocation concealment, which in some cases produced imbalances in the comparison groups. In many studies, no effort was made to prevent participants and assessors from being aware of the assigned intervention. Another relevant aspect was the presence of losses to follow‐up, without any related sensitivity analyses being considered. Finally, very few studies were registered in the clinical trial databases, which makes it difficult to verify the analysis plan and its follow‐up. Concerning imprecision, it should be noted that all the studies had sample sizes below the optimal information size and, in many of them, despite the fact that a potential effect level of the therapy was appreciated, the null value (of no effect) was contained in the confidence interval. On the other hand, in the studies in which the difference was statistically significant in favour of some treatments, wide confidence intervals were always found, which makes it difficult to assess the clinical impact of the result.

Potential biases in the review process

This systematic review was designed to perform a network meta‐analysis, which was not possible due to the network structure and the number of studies available for each comparison. Consequently, the network geometry presents a significant number of open loops and only one closed loop, in which the effect could be estimated from mixed comparisons, but very imprecise results would be obtained.

It is important to mention that the included studies, in addition to their small sample size, high risk of bias, and heterogeneous populations, applied different methods to assess the main outcome (short‐term hair regrowth ≥ 75%), which increases the risk of presenting effect pooled estimates using network meta‐analysis, due to violations of similarity and transitivity assumptions.

Consequently, the review group decided not to present the network meta‐analysis and the ranking of treatments, based on a network, mainly, of indirect comparisons, most informed by a single study for each comparison, with a small sample size, which could produce biassed and imprecise estimates that could lead to inappropriate clinical conclusions. Instead of network meta‐analysis, individual analyses and forest plots are provided for each comparison.

In our review, we found many interventions, including different drugs, with different administration routes and dosages, even for the same drug (e.g. triamcinolone). We considered including in the main text of the review those comparisons assessing the interventions that were considered more commonly used in clinical practice. We comprehensively reported all other interventions in an online appendix. Therefore, we presented the results after prioritising some interventions and grouping them whenever they belonged to the same pharmacological group. The prioritised interventions were selected by the field experts involved in the systematic review (clinical dermatologists) following the suggestions made by the clinical reviewers of previous versions of the review and by the editorial team members at the Cochrane Skin Group. Surely, using a survey amongst clinicians or running a Delphi consensus panel would have been a more appropriate method to prioritise the interventions, and would have provided greater transparency in the process. However, it is worth noting that, to avoid a selective reporting bias, all comparisons have been comprehensibly reported, either in the main text or in the online supplement.

We chose the primary outcome of short‐term hair regrowth ≥ 75% as we considered it a measure of clinical efficacy and a significant cosmetic improvement. We believe that reducing at least 75% (rather than 50%) of the affected area would reduce the psychological impact caused by the condition and would eliminate the need for wigs, improving patients' quality of life (Rencz 2016). There is a lack of consensus in the scientific community to set this threshold. For example, in one recently published study of alopecia areata, the authors chose 90%, 75%, and 50% of hair regrowth as secondary outcomes of treatment efficacy (King 2022a; King 2022b).

Agreements and disagreements with other studies or reviews

The review by Delamere 2008 evaluated the effects of the interventions used in the treatment of AA. Seventeen trials were included in that review, and none of the interventions showed significant treatment benefit in terms of hair regrowth when compared with placebo, therefore, the clinical benefits were inconclusive. New systematic reviews addressing the treatment of AA have recently appeared in the literature, but the conclusions are not consistent between them, and they present differences in design and results compared to our systematic review.

Below are the results of a set of recent systematic reviews, which have addressed research questions similar to ours.

Freire 2019 conducted a systematic review to assess the effects of topical minoxidil 5% versus placebo for AA. In this work, 59 studies were included and the outcome of interest was hair regrowth ≥ 50%. This work concluded that topical minoxidil 5% is superior to placebo in adults and children with AA (RR 8.37, 95% CI 3.16 to 22.14). Our results are not entirely comparable, since we assessed hair regrowth ≥ 75% and, for this outcome, we did not find studies that evaluated this concentration of minoxidil. However, for minoxidil 2%, our results showed superiority against placebo (Luo 1991).

In order to compare the efficacy of the different treatment options available for both mild and moderate‐to‐severe scalp AA, Gupta 2019 performed a systematic review with NMA, in which it analysed 38 studies, including randomised clinical trials, non‐randomised studies and observational studies. This work concludes, for the result of hair regrowth ≥ 75% based on a network mainly of indirect comparisons and against placebo, that there is evidence for the use of intralesional and topical corticosteroids for the treatment of mild AA; and also for DPCP, laser, SADBE, topical minoxidil and topical corticosteroids in moderate‐to‐severe disease. These results are not comparable with the results we found, given that we only included randomised clinical trials. Moreover, for reasons previously described, we chose not to perform the NMA (violation of assumptions of similarity, transitivity, and absence of evidence to support valid and accurate estimates in the network).

Yee 2020 evaluated the efficacy and tolerability of different concentrations of intralesional triamcinolone acetonide for AA in a systematic review that included seven studies using hair regrowth ≥ 50% as the primary outcome. According to their results, the hair regrowth rates were comparable at the concentrations of 5 mg/mL and 10 mg/mL (80.9% versus 76.4%), while the lower hair growth rates (62.3%) occurred when concentrations below 5 mg/ml were used. The results of this systematic review are not entirely comparable with ours, since observational studies have been included and a different cut‐off point was defined for the primary outcome (hair regrowth ≥ 50%).

Husein‐ElAhmed 2021 performed a systematic review to evaluate the efficacy of cyclosporine A in the treatment of AA, using the outcome of hair regrowth ≥ 50%. Fifteen studies were included, and a response rate was found in 73% of participants (57%‐85%). Cyclosporine A in monotherapy showed a proportion of hair regrowth in 66% of cases (50% to 79%), while cyclosporine A combined with systemic corticosteroids produced an effect in 78% of participants (48% to 93%). In our systematic review, the comparison of cyclosporine versus placebo in hair regrowth ≥ 75% was included, without being able to detect differences between groups (Lai 2019).

Oral small molecule inhibitors (JAK inhibitors) have been used to treat AA. Yu 2021 performed a systematic review to examine the efficacy in patients with AA treated with oral tofacitinib or ruxolitinib. Twelve studies were included and, for the outcome of hair growth ≥ 50%, they found a response rate for ruxolitinib of 79%, (95% CI, 66% to 87%) and for tofacitinib of 62% (95% CI, 49% to 74%); the difference was not significant (P = 0.06). In our systematic review, we found only one study comparing these two interventions showing no differences between them on hair regrowth ≥ 75% (Almutairi 2019).

Fukumoto 2021 published a systematic review with NMA, with the objective of identifying the best treatment for AA and establishing a rank of treatments using network meta‐analysis. This systematic review included a total of 54 studies, using hair regrowth ≥ 50% as the main outcome. The authors concluded that pentoxifylline plus topical corticosteroids had the highest treatment success rate compared with “no treatment,” followed by pentoxifylline alone, topical calcipotriol plus narrowband ultraviolet radiation B phototherapy, topical calcipotriol, intralesional corticosteroids, systemic corticosteroids, minoxidil plus topical corticosteroids, topical bimatoprost, psoralen ultraviolet radiation A phototherapy, and tofacitinib. Our results are not consistent with Fukumoto's conclusions, probably due to our cut‐off point for the main outcome (hair regrowth ≥ 75%), and several interventions described by Fukumoto were not included in our review (e.g. pentoxifylline). In this NMA, the authors could not provide a ranking of treatments for AA because of independent loops and wide confidence intervals. We agree with Fukumoto's decision not to propose a ranking of treatments as we find ourselves in the same situation when attempting to perform a valid NMA.

In order to evaluate therapies in AA, it is important to define a rigorous standardised outcome and a sufficiently long follow‐up period. In general, we can affirm that the inconsistency of the results described here could largely be related to these factors.

Authors' conclusions

Implications for practice.

In general, the evidence found shows inconclusive results of the effect of interventions (both systemic and local therapies) on clinically relevant hair regrowth. Although no serious adverse events were reported, the benefit‐harm balance of using therapies without clinically relevant effects on hair growth should be considered.

Only oral baricitinib, compared to placebo, shows beneficial effects on hair regrowth. Although we found inconclusive results for the risk of serious adverse effects with baricitinib, the reported small incidence of serious adverse events in the baricitinib arm should be taken into account to balance the benefits and harms of this treatment as serious adverse events could lead to medical assistance.

Dermatologists and people with AA should not make decisions about interventions for AA based on expectations of improvement of health‐related quality of life as this outcome has been barely assessed.

Implications for research.

Some limitations of the present review are clearly related to future research in alopecia areata. First, the need for a valid, reliable and widely accepted standardised outcome measure of the effectiveness of the treatments stands out, so that clinical trials in AA achieve greater homogeneity and comparability in their results. Second, it is important that AA clinical trials are designed with high standards of methodological quality, ensuring adequate control of biases and with a sample size calculation that guarantees sufficient statistical power to detect differences between treatments, when these really exist. Finally, in the context of therapeutic diversity in AA as a result of the number of existing options, clinical studies are required to corroborate or discard previous findings, improve precision and contrast the reproducibility of the effects of the treatments already evaluated.

What's new

Date	Event	Description
23 October 2023	New citation required and conclusions have changed	New interventions have been included in the review. Treatment with baricitinib resulted in a large increase in short (between 12‐26 weeks) and long‐term (greater than 26 weeks) hair regrowth ≥ 75% as compared to placebo although the results for the risk of serious adverse effects with baricitinib are inconclusive. The impact of other treatments on hair regrowth is very uncertain.
23 October 2023	New search has been performed	A new team of reviewers underwent the update of this review. No author of the previous review continued to engage in the update. New methods were proposed, specifically a new search with new selection criteria broadening the set of therapies for alopecia areata included in the review. New statistical methods were proposed (i.e. a network meta‐analysis).

History

Protocol first published: Issue 9, 2020

Date	Event	Description
16 September 2021	New search has been performed	Update methods, new analysis, and discussion.

Appendices

Appendix 1. Search strategies

Cochrane Skin Specialised Register (CRSW) search strategy

1. autoimmune hair loss* AND INREGISTER
2. (alopecia*):TI,ab AND INREGISTER
3. MESH DESCRIPTOR Alopecia Areata EXPLODE ALL AND INREGISTER
4. (non scarring hair loss*):ti,ab AND INREGISTER
5. (nonscarring hair loss*):ti,ab AND INREGISTER
6. (ophiasi* ):ti,ab AND INREGISTER
7. (sisaipho* ):ti,ab AND INREGISTER
8. (atrichia* ):ti,ab AND INREGISTER
9. (pseudopelade* ):ti,ab AND INREGISTER
10. (porrigo decalvan* ):ti,ab AND INREGISTER
11. (spot* and bald* ):ti,ab AND INREGISTER
12. (patch* and bald*):ti,ab AND INREGISTER
13. #1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8 OR #9 OR #10 OR #11 OR #12

CENTRAL (Cochrane Library) search strategy

#1 MeSH descriptor: [Alopecia Areata] explode all trees
#2 nonscarring hair loss*:ti,ab,kw
#3 non scarring hair loss*:ti,ab,kw
#4 ophiasi*:ti,ab,kw
#5 sisaipho*:ti,ab,kw
#6 atrichia*:ti,ab,kw
#7 pseudopelade*:ti,ab,kw
#8 porrigo decalvans:ti,ab,kw
#9 (bald* next spot*):ti,ab,kw
#10 (bald* next patch*):ti,ab,kw
#11 (autoimmune next hair next loss*):ti,ab,kw
#12 (alopecia near/3 (spot or spots or nonscarring or non‐scarring or areata or areatae or barbae or patch or patches or semiuniversalis or autoimmune or totalis or circumscripta or liminaris or universalis or circumscribed or diffuse or Jonston or marginal or snake‐shaped)):ti,ab,kw
#13 {or #1‐#12}

MEDLINE (Ovid) search strategy

1. exp Alopecia Areata/

2. nonscarring hair loss$.ti,ab.

3. non‐scarring hair loss$.ti,ab.

4. ophiasi$.ti,ab.

5. sisaipho.ti,ab.

6. atrichia$.ti,ab.

7. pseudopelade$.ti,ab.

8. porrigo decalvans.ti,ab.

9. ((spot$ or patch$) and baldness).ti,ab.

10. autoimmune hair loss$.ti,ab.

11. (alopecia adj3 (spot$ or nonscarring or non‐scarring or areat$ or barbae or patch$ or semiuniversalis or autoimmune or totalis or circumscripta or liminaris or universalis or circumscribed or diffuse or Jonston$ or marginal or snake‐shaped)).ti,ab.

12. or/1‐11

13. randomized controlled trial.pt.

14. controlled clinical trial.pt.

15. randomized.ab.

16. placebo.ab.

17. clinical trials as topic.sh.

18. randomly.ab.

19. trial.ti.

20. 13 or 14 or 15 or 16 or 17 or 18 or 19

21. exp animals/ not humans.sh.

22. 20 not 21

23. 12 and 22

[Lines 13‐22: Cochrane Highly Sensitive Search Strategy for identifying randomised trials in MEDLINE: sensitivity‐ and precision‐maximising version (2008 revision); Ovid format, from section 3.6.1 in Lefebvre C, Glanville J, Briscoe S, Littlewood A, Marshall C, Metzendorf M‐I, et al. Technical Supplement to Chapter 4: Searching for and selecting studies. In: Higgins JPT, Thomas J, Chandler J, Cumpston MS, Li T, Page MJ, Welch VA (eds). Cochrane Handbook for Systematic Reviews of Interventions Version 6. Cochrane, 2019. Available from: www.training.cochrane.org/handbook]

Embase (Ovid) search strategy

1. exp alopecia areata/

2. nonscarring hair loss$.ti,ab.

3. non‐scarring hair loss$.ti,ab.

4. ophiasi$.ti,ab.

5. sisaipho.ti,ab.

6. atrichia$.ti,ab.

7. pseudopelade$.ti,ab.

8. porrigo decalvans.ti,ab.

9. ((spot$ or patch$) and baldness).ti,ab.

10. autoimmune hair loss$.ti,ab.

11. (alopecia adj3 (spot$ or nonscarring or non‐scarring or areat$ or barbae or patch$ or semiuniversalis or autoimmune or totalis or circumscripta or liminaris or universalis or circumscribed or diffuse or Jonston$ or marginal or snake‐shaped)).ti,ab.

12. or/1‐11

13. crossover procedure.sh.

14. double‐blind procedure.sh.

15. single‐blind procedure.sh.

16. (crossover$ or cross over$).tw.

17. placebo$.tw.

18. (doubl$ adj blind$).tw.

19. allocat$.tw.

20. trial.ti.

21. randomized controlled trial.sh.

22. random$.tw.

23. or/13‐22

24. exp animal/ or exp invertebrate/ or animal experiment/ or animal model/ or animal tissue/ or animal cell/ or nonhuman/

25. human/ or normal human/

26. 24 and 25

27. 24 not 26

28. 23 not 27

29. 12 and 28

[Lines 13‐28: Based on terms suggested for identifying RCTs in Embase (section 3.6.2) in Lefebvre C, Glanville J, Briscoe S, Littlewood A, Marshall C, Metzendorf M‐I, et al. Technical Supplement to Chapter 4: Searching for and selecting studies. In: Higgins JPT, Thomas J, Chandler J, Cumpston MS, Li T, Page MJ, Welch VA (eds). Cochrane Handbook for Systematic Reviews of Interventions Version 6. Cochrane, 2019. Available from: www.training.cochrane.org/handbook][Enter text here]

US National Institutes of Health Ongoing Trials Register (ClinicalTrials.gov)

alopecia areata OR patchy alopecia areata OR alopecia universalis OR alopecia areata universalis OR alopecia totalis OR alopecia areata totalis

World Health Organization International Clinical Trials Registry Platform (www.who.int/clinical-trials-registry-platform)

alopecia areata OR patchy alopecia areata OR alopecia universalis OR alopecia areata universalis OR alopecia totalis OR alopecia areata totalis

Appendix 2. Methods: Unit of analysis issues

For future updates, the following unit of analysis issues will be taken into account.

Cluster‐randomised trials

In case of cluster‐randomised trials, we will perform the analysis at the same level as the allocation, using a summary measurement from each cluster, following the recommendations in the Handbook (Higgins 2019c). If a cluster‐randomised trial provides results unadjusted for its cluster design, we will reduce the effective sample size accounting for the design effect using the reported intra‐cluster correlation coefficient (ICC) or, in case the study does not report the ICC, we will use an ICC obtained from a similar trial from the literature. We will assess the effect of the unit of randomisation in a sensitivity analysis.

Multiple intervention groups

If we include trials which have relevant comparisons between more than two groups in a pairwise meta‐analysis, we will split participants in the control group into two or more groups; this is to avoid double‐counting participants. In the network meta‐analysis, multi‐arm trials are split into separate nodes of the network, so double‐counting of patients is not a concern as long as within‐trial correlation is adequately considered in the analysis (Franchini 2012).

Multiple body parts

In the context of AA, different parts of the body could be randomised to different interventions (i.e. split‐head designs). If outcomes are evaluated at the level of body parts despite being randomised at participant level, and the data were not correctly analysed, we will assess whether it is possible to approximate the correct analysis following guidance provided in the Handbook (Higgins 2019c). If studies do not provide sufficient information to approximate the correct analysis, we will exclude the data from the analysis.

Data and analyses

Comparison 1. Oral classical immunosuppressants vs placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 Short‐term hair regrowth ≥ 75%	2	79	Risk Ratio (M‐H, Fixed, 95% CI)	4.68 [0.57, 38.27]
1.2 Incidence of serious adverse events	2	79	Risk Ratio (M‐H, Fixed, 95% CI)	Not estimable
1.3 Health‐related quality of life	1	36	Mean Difference (IV, Fixed, 95% CI)	0.01 [‐0.04, 0.07]

Comparison 2. Oral classical immunosuppressants vs oral classical immunosuppressants.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
2.1 Short‐term hair regrowth ≥ 75%	2	80	Risk Ratio (M‐H, Fixed, 95% CI)	1.67 [0.96, 2.88]
2.2 Incidence of serious adverse events	4	154	Risk Ratio (M‐H, Fixed, 95% CI)	Not estimable

Comparison 3. Intralesional classical immunosuppressants vs placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
3.1 Short‐term hair regrowth ≥ 75%	1	231	Risk Ratio (M‐H, Fixed, 95% CI)	13.84 [0.87, 219.76]
3.2 Incidence of serious adverse events	4	543	Risk Ratio (M‐H, Fixed, 95% CI)	Not estimable

Comparison 4. Subcutaneous biologics vs placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
4.1 Short‐term hair regrowth ≥ 75%	1	60	Risk Ratio (M‐H, Fixed, 95% CI)	3.59 [0.19, 66.22]
4.2 Incidence of serious adverse events	1	60	Risk Ratio (M‐H, Fixed, 95% CI)	1.54 [0.07, 36.11]

Comparison 5. Oral small molecule inhibitors vs placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
5.1 Short‐term hair regrowth ≥ 75%	2	1200	Risk Ratio (M‐H, Fixed, 95% CI)	7.54 [3.90, 14.58]
5.2 Incidence of serious adverse events	3	1224	Risk Ratio (M‐H, Random, 95% CI)	1.47 [0.60, 3.60]
5.3 Long‐term hair regrowth ≥ 75%	2	1200	Risk Ratio (M‐H, Fixed, 95% CI)	8.49 [4.70, 15.34]

Comparison 6. Oral small molecule inhibitors vs oral small molecule inhibitors.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
6.1 Short‐term hair regrowth ≥ 75%	1	80	Risk Ratio (M‐H, Fixed, 95% CI)	1.08 [0.77, 1.52]
6.2 Incidence of serious adverse events	1	80	Risk Ratio (M‐H, Fixed, 95% CI)	Not estimable

Comparison 7. Topical small molecule inhibitors vs placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
7.1 Short‐term hair regrowth ≥ 75%	1	78	Risk Ratio (M‐H, Fixed, 95% CI)	5.00 [0.25, 100.89]
7.2 Incidence of serious adverse events	1	78	Risk Ratio (M‐H, Fixed, 95% CI)	0.33 [0.01, 7.94]

Comparison 8. Contact immunotherapy vs placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
8.1 Short‐term hair regrowth ≥ 75%	1	99	Risk Ratio (M‐H, Fixed, 95% CI)	1.16 [0.79, 1.71]

Comparison 9. Contact immunotherapy vs topical hair growth stimulants.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
9.1 Short‐term hair regrowth ≥ 75%	1	99	Risk Ratio (M‐H, Fixed, 95% CI)	1.16 [0.79, 1.71]

Comparison 10. Contact immunotherapy + hair growth stimulants vs contact immunotherapy.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
10.1 Short‐term hair regrowth ≥ 75%	1	30	Risk Ratio (M‐H, Fixed, 95% CI)	0.67 [0.13, 3.44]
10.2 Incidence of serious adverse events	1	30	Risk Ratio (M‐H, Fixed, 95% CI)	Not estimable

Comparison 11. Topical hair growth stimulants vs placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
11.1 Short‐term hair regrowth ≥ 75%	2	202	Risk Ratio (M‐H, Fixed, 95% CI)	2.31 [1.34, 3.96]
11.2 Incidence of serious adverse events	4	304	Risk Ratio (M‐H, Fixed, 95% CI)	Not estimable

Comparison 12. Other therapies.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
12.1 Short‐term hair regrowth ≥ 75%	1	80	Risk Ratio (M‐H, Fixed, 95% CI)	0.31 [0.11, 0.86]
12.2 Incidence of serious adverse events	1	80	Risk Ratio (M‐H, Fixed, 95% CI)	Not estimable

Comparison 13. Sensitivity analysis.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
13.1 Sensitivity analysis. Oral classical immunosuppressants vs placebo: short‐term hair regrowth	2	79	Risk Ratio (M‐H, Fixed, 95% CI)	0.61 [0.19, 2.00]
13.2 Sensitivity analysis. Betamethasone vs azathioprine or prednisolone: short‐term hair regrowth	2	80	Risk Ratio (M‐H, Fixed, 95% CI)	1.00 [0.64, 1.57]
13.3 Sensitivity analysis. Intralesional classical immunosuppressants vs placebo: short‐term hair regrowth	1	231	Risk Ratio (M‐H, Fixed, 95% CI)	13.84 [0.87, 219.76]
13.4 Sensitivity analysis. Subcutaneous biologics vs placebo: short‐term hair regrowth	1	60	Risk Ratio (M‐H, Fixed, 95% CI)	0.50 [0.11, 2.26]
13.5 Sensitivity analysis. Oral small molecule inhibitors: baricitinib vs placebo	2		Risk Ratio (M‐H, Fixed, 95% CI)	Subtotals only
13.5.1 Short‐term hair regrowth	2	1200	Risk Ratio (M‐H, Fixed, 95% CI)	7.54 [3.90, 14.58]
13.5.2 Long‐term hair regrowth	2	1200	Risk Ratio (M‐H, Fixed, 95% CI)	1.79 [1.36, 2.36]
13.6 Sensitivity analysis. Oral small molecule inhibitors: ruxolitinib vs tofacitinib	1	80	Risk Ratio (M‐H, Fixed, 95% CI)	0.96 [0.70, 1.32]
13.7 Sensitivity analysis. Topical small molecule inhibitors: short‐term hair regrowth	1	78	Risk Ratio (M‐H, Fixed, 95% CI)	0.33 [0.07, 1.55]
13.8 Sensitivity analysis. Contact immunotherapy vs placebo: short‐term hair regrowth	1	99	Risk Ratio (IV, Fixed, 95% CI)	1.16 [0.79, 1.71]
13.9 Sensitivity analysis. Contact immunotherapy vs minoxidil: short‐term hair regrowth	1	99	Risk Ratio (IV, Fixed, 95% CI)	1.16 [0.79, 1.71]
13.10 Sensitivity analysis. Contact immunotherapy + hair growth stimulants vs contact immunotherapy: short term hair regrowth	1	30	Risk Ratio (M‐H, Fixed, 95% CI)	0.67 [0.13, 3.44]
13.11 Sensitivity analysis. Topical hair growth stimulants vs placebo: short‐term hair‐regrowth	2	202	Risk Ratio (M‐H, Fixed, 95% CI)	2.31 [1.34, 3.96]
13.12 Sensitivity analysis. Cryotherapy vs fractional CO2 laser: short‐term hair regrowth	1	80	Risk Ratio (M‐H, Fixed, 95% CI)	0.31 [0.11, 0.86]

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Abdallah 2020.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Egypt
Participants	Randomised: 100 patches in 20 patients (30.15 (8.9) years, 1 female) Inclusion criteria: Patients with non‐severe AA A SALT score up to S2 with less than 50% of scalp involvement 5 scalp patches of minimally 3 cm2 Exclusion criteria: Patients with severe AA Other forms of patchy hair loss like trichotillomania, tractional alopecia, cases with known hypersensitivity to corticosteroids or minoxidil Patients with severe systemic diseases including diabetes and hypertension Patients who received systemic treatment for AA in the last 3 months or had applied any topical or intradermal treatments in the last 4 weeks
Interventions	For each participant, four out of five patches were randomised to receive four different interventions. The fifth patch was selected as control without receiving any intervention. Intervention 1: Intradermal triamcinolone acetonide 5 mg/mL (4 mL max/session) Intervention 2: Intradermal minoxidil 50 mg/mL (4 mL max/session) Intervention 3: Intradermal mix (1:1 proportion) of triamcinolone 5 mg/mL and minoxidil 50 mg/mL (4 mL max/session) Intervention 4: Micro‐needling intradermal injections with no drugs All injections were done intradermally delivering 0.1 mL/point at 1 cm interval using a 1 mL insulin syringe with a 30‐gauge needle for a maximum of 4 mL in each session. All treatments were delivered in 4 sessions, 4 weeks apart.
Outcomes	Assessment at 20 weeks LAD score Average percentage change of hair regrowth Time of initial response Adverse events
Funding source	No funding
Declarations of interest	No conflicts
Notes

Abdel 2020.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Egypt
Participants	Randomised: 75 (25.51 (9.63) years, 37 females) Inclusion criteria: Patients with recalcitrant patchy scalp alopecia areata who experienced failure of different topical and systemic treatment for at least 6 months Aged 12 years or more Patients who discontinued any other treatment modality at least one month before the study Patchy alopecia areata not exceeding 10 cm2 for each lesion Exclusion criteria: Patients with cold intolerance and those with previously diagnosed systemic or autoimmune diseases
Interventions	Once every 2 weeks for maximum of 6 sessions or earlier if complete hair regrowth occurs Intervention 1: Liquid nitrogen cryotherapy (3‐5 sec. spray technique) Intervention 2: Liquid nitrogen cryotherapy (8‐10 sec spray technique) Intervention 3: Liquid nitrogen cryotherapy (13‐15 sec spray technique)
Outcomes	Assessment at 1 month after last session, maximum of 16 weeks Clinical response: % regrowth of terminal hair Overall improvement assessed by average percentage Maintained response or recurrence of the disease Adverse effects
Funding source	No funding
Declarations of interest	No conflicts
Notes

Agrawal 2020.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: Not stated Country: India
Participants	Randomised: 70 patients Inclusion criteria: Not stated Exclusion criteria: Not stated
Interventions	Intervention 1: Diphencyprone (DPCP) from 0.01% to 2% weekly after week 2 for 24 weeks Intervention 2: Diphencyprone (DPCP) from 0.01% to 2% weekly after week 2 + anthralin for 1‐10 min 5 days a week for 24 weeks
Outcomes	Assessment at week 24 Percentage change in SALT score Side effects
Funding source	Not stated
Declarations of interest	Not stated
Notes

Albalat 2019.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Egypt
Participants	Randomised: 80 (34.29 (9.226) years, 12 females) Inclusion criteria: Healthy persons of both sexes, aged from 17 to 52 years with patchy alopecia Exclusion criteria: Patients with a history of or existing skin diseases affecting the scalp Evidence of infection or skin cancer in the treated areas Medical diseases such as diabetes and hypertension, malignancies, HIV, HCV or HBV patients in whom re‐growth is evident at baseline in the areas to be treated Under anticoagulant therapy, with hemoglobin < 10 g/dL or platelets < 150,000/mm Women who are pregnant or nursing Patients with evidence of adrenal cortex abnormality or previous significant adverse reaction to intralesional steroids Alopecia totalis or universalis Patients who have been treated with intralesional steroids, systemic steroids, anthralin, minoxidil, or other medication which may affect hair re‐growth within 1 month of the baseline visit
Interventions	From 3 to 5 sessions, once every 2 weeks for 12 weeks Intervention 1: Intralesional triamcinolone 5 mg/mL (3 mL max) Intervention 2: Intralesional platelet‐rich plasma injection (3 mL max)
Outcomes	Assessment at 12 weeks Improvement in SALT score Hair regrowth Dermoscopic evaluations (determine the number of dystrophic hairs and its markers including yellow, black dots, exclamation mark, short villous, short broken, and white hair) Side effects
Funding source	Not stated
Declarations of interest	No conflicts
Notes

Almutairi 2019.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Kuwait
Participants	Randomised: 75 (32.1 (5.16) years, 32 females) Inclusion criteria: Patients > 18 years of age with a clinical diagnosis of severe AA, for at least 6 months duration, and resistant to other treatment modalities Exclusion criteria: Patients with less than 30% scalp involvement Co‐existent androgenetic alopecia Active scalp inflammation Active malignancy or history of malignancy History of systemic or infectious disease Pregnant women
Interventions	Intervention 1: Oral ruxolitinib, 20 mg twice daily for 24 weeks Intervention 2: Oral tofacitinib, 5 mg twice daily for 24 weeks
Outcomes	Duration of initial hair growth (assessed at 24 weeks) Percentage change in SALT score (assessed at 24 weeks) Treatment response (% hair regrowth) (assessed at 24 weeks) Relapse rate (assessed at 36 weeks) Adverse effects
Funding source	Not stated
Declarations of interest	No conflicts
Notes

Amaral 2003.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Brazil
Participants	Randomised: 35, (4.9 (2.8) years, 22 females) Inclusion criteria: Diagnosis of partial alopecia (of hairy areas) Alopecia totalis (AT) (of hairy areas) Alopecia universalis (AU) Exclusion criteria: Not stated
Interventions	Intervention 1: Betamethasone dipropionate cream 0.5 mg/mL. Thin layer twice a day for 6 months Intervention 2: Placebo cream, thin layer twice a day for 6 months
Outcomes	Assessment at 24 weeks Repilation response Influence of atopy in the therapeutic response
Funding source	Not stated
Declarations of interest	Not stated
Notes

Asilian 2020.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 2 Country: Iran
Participants	Randomised: 36 (21 females) Inclusion criteria: Over 50% of scalp involvement Age range 16‐60 years Refractory disease with local and systemic therapies, which was defined as unresponsiveness to systemic therapies Patients' consent to participate in the study Exclusion criteria: Pregnancy or lactating Haemoglobin levels less than 9 mg/dL White blood cell count under 4000 mm Platelet level under 100,000 mm Over two‐fold increase in hepatic enzyme levels more than the normal limit Increased erythrocyte sedimentation rate A positive tool exam for stringyloides stercoralis Positive test for viral hepatitis The presence of contraindications for oral betamethasone use and hypersensitivity to methotrexate
Interventions	Intervention 1: Oral betamethasone 1 mg three times on Saturdays for 24 weeks Intervention 2: Oral methotrexate 5 mg three times on Fridays for 24 weeks Intervention 3: Oral betamethasone 1 mg + oral methotrexate 5 mg three times on Saturdays + three times on Fridays for 24 weeks
Outcomes	Assessment at 28 weeks Hair regrowth assessed by decrease in SALT score
Funding source	Isfahan University of Medical Sciences
Declarations of interest	Not stated
Notes

Azimi 2018.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Iran
Participants	Randomised: 50 (26.58 (13.1) years, 27 females) Inclusion criteria: Patients with a clinical diagnosis of AA who have not responded to corticosteroid treatment in the past 3 months Patients who have expressed their full consent to participate in the study Patients over 2 years of age Exclusion criteria: Existence of skin or other accompanying systemic diseases causing severe contact dermatitis or severe urticaria relative to diphencyprone (DPCP) or anthralin Having undergone another treatment over the past month Patients with malignancy or dyscrasia Pregnant and lactating women
Interventions	Intervention 1: Diphencyprone (DPCP) from 0.01 to 20 mg/mL once a week Intervention 2: Anthralin 5 mg/mL from 20 to 60 min daily
Outcomes	Assessment at up to 24 weeks after treatment Recovery rate Duration of treatment Complications Relapse rate
Funding source	Not stated
Declarations of interest	Not stated
Notes

Barreto‐Rocha 2021.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Brazil
Participants	Randomised: 24 (21 females) Inclusion criteria: AA for more than one year SALT score >= 50% more than 30 days without treatment Exclusion criteria: Not stated
Interventions	Intervention 1: Diphencyprone once weekly for 24 weeks Intervention 2: Anthralin 20 mg/mL once weekly for 24 weeks
Outcomes	Assessment at 24 weeks Percentage of hair regrowth Median SALT score Side effects
Funding source	Institutional funding: FUNADERM ‐ Fundo de Apoio à Dermatologia
Declarations of interest	No conflicts
Notes

Bokhari 2018.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Australia
Participants	Randomised: 16 (females) Inclusion criteria: Patients with diagnosis of alopecia areata Exclusion criteria: Not stated
Interventions	Intervention 1: Ruxolitinib ointment 10 mg/mL, twice daily for 12 weeks Intervention 2: Tofacitinib ointment 20 mg/mL, twice daily for 12 weeks Intervention 3: Clobetasol dipropionate ointment 0.5 mg/mL, twice daily for 12 weeks Intervention 4: Placebo ointment, twice daily for 12 weeks
Outcomes	Assessment at 24 weeks Changes in global photographies 4‐point scale for invertigator's and patient's global assessment (hair regrowth scales)
Funding source	Institutional funding: a research project grant from the Australia Alopecia Areata Foundation
Declarations of interest	No conflicts
Notes

Charuwichitratana 2000.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Thailand
Participants	Randomised: 70 (34.24 (11.62) years) Inclusion criteria: Patients with diagnosis of patchy alopecia areata who did not receive any treatment for at least 1 month prior to the study Exclusion criteria: Patients younger than 15 years Pregnant women
Interventions	Intervention 1: Desoximetasone cream 25 mg/mL, twice a day for 12 weeks Intervention 2: Topical placebo (posology not detailed)
Outcomes	Assessment at 12 weeks Hair regrowth Adverse effects
Funding source	Not stated
Declarations of interest	Not stated
Notes

Chu 2015.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: USA
Participants	Randomised: 4 (median age 34 years, 2 females) Inclusion criteria: Patients with untreated patchy alopecia areata Exclusion criteria: Not stated
Interventions	Intervention 1: Intralesional triamcinolone 2.5 mg/mL (1.5 mg), once every 6 weeks for 36 weeks Intervention 2: Intralesional triamcinolone 5 mg/mL (3 mg), once every 6 weeks for 36 weeks Intervention 3: Intralesional triamcinolone 10 mg/mL (6 mg), once every 6 weeks for 36 weeks
Outcomes	Assessement at 36 weeks: Hair regrowth Hair density Hair caliber increase Side effects
Funding source	Not stated
Declarations of interest	No conflicts
Notes

Devi 2015.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Pakistan
Participants	Randomised: 226 (34.36 (8.7) years, 62 females) Inclusion criteria: Patients aged 18 ‐ 50 years of either gender, who had never received treatment for hair loss and having non‐scarring patchy hair loss (fewer than 3 patches) Exclusion criteria: Patients with extensive alopecia areata (more than 3 patches), alopecia universalis, alopecia totalis, and ophiasis Patients taking systemic glucocorticoids or immunosuppressives
Interventions	Intervention 1: Intralesional triamcinolone 10 mg/mL, once every 3 weeks for 12 weeks Intervention 2: Betamethasone valerate cream 1 mg/mL, twice daily for 12 weeks
Outcomes	Assessment at 12 weeks Hair regrowth
Funding source	Not stated
Declarations of interest	Not stated
Notes

El‐Taieb 2017.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Egypt
Participants	Randomised: 90 (51 females) Inclusion criteria: Patients included between 10 and 40 years, with no therapy for at least 3 months before study Exclusion criteria: Children aged less than 10 years Pregnant and lactating women Immunocompromised patients Patients having active scalp inflammation
Interventions	Intervention 1: Minoxidil topical solution 50 mg/mL, twice daily for 12 weeks Intervention 2: Intralesional platelet‐rich plasma, once every 4 weeks for 12 weeks Intervention 3: Panthenol cream (placebo), twice daily for 12 weeks
Outcomes	Assessment at 12 weeks Percentage of hair growth Percentage of short vellus hair Percentage of yellow dots
Funding source	Not stated
Declarations of interest	No conflicts
Notes

Fransway 1988.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: USA
Participants	Randomised: 21 (mean 35.8 years, 13 females) Inclusion criteria: More than 65 per cent loss of scalp hair from alopecia areata without scarring or evidence of regrowth Between the ages of 18 and 65 and in good general health without significant systemic disease Not have been treated for at least 4 weeks before entering the study and not be receiving significant systemic therapy for this or other major illnesses All women of childbearing age were to use an effective means of contraception Exclusion criteria: Not stated
Interventions	Intervention 1: Minoxidil topical solution 30 mg/mL twice a day for 1 year Intervention 2: Placebo topical solution twice a day
Outcomes	Assessment at 52 weeks Regrowth of sparse vellus hair Regrowth of terminal hair Adverse effects
Funding source	Not stated
Declarations of interest	Not stated
Notes

Ghandi 2021.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Iran
Participants	Randomised: 50 (mean 31 (8.8) years, 23 females) Inclusion criteria: Being treatment‐naïve for 3 months prior to enrolment Being 18 years of age or older Having scalp hair loss > 25% based on the Severity of Alopecia Tool (SALT) score Agreeing to co‐operate for regular follow‐up Exclusion criteria: Patients with alopecia affecting < 25% of scalp area Patients that were lactating, pregnant or intending to become pregnant Immunosuppressed patients Patients with a history of receiving systemic or local therapy within 3 months or a history of severe reaction or intolerance of DPCP or anthralin Patients with recent progressive and severe hair loss that required systemic immunosuppressant therapy Patients with a significant intercurrent medical illness or localised scalp inflammation or infection
Interventions	Intervention 1: Diphencyprone (DPCP) from 0.01 to 10 mg/mL once a week for 24 weeks Intervention 2: Diphencyprone (DPCP) from 0.01 to 10 mg/mL + anthralin from 5 to 10 mg/mL on day three once a week for 24 weeks
Outcomes	Assessment at 24 weeks Percentage of scalp hair regrowth (rates) Mean percentage of scalp hair regrowth Side effects
Funding source	Not stated
Declarations of interest	Not stated
Notes

Goyal 2000.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: India
Participants	Randomised: 30 Inclusion criteria: Patients with AA with acutely evolving and/or extensive disease, with no other medical illness and no taking systemic treatment prior presenting at department Exclusion criteria: Not stated
Interventions	Intervention 1: Oral betamethasone 5 mg, 2 consecutive days every week Intervention 2: Oral prednisolone 40 mg, once a day
Outcomes	Assessment at 12 weeks (resistant cases were followed up to 24 weeks) Pull test of hair loss Growth of hair in the patch Appearance of adverse effects Time taken for onset of response since therapy Time of appearance of adverse effects
Funding source	Not stated
Declarations of interest	Not stated
Notes

Gupta 2019.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: India
Participants	Randomised: 50 (26.6 (7.38) years, 14 females) Inclusion criteria: Patients of ≥ 18 years of age, with AA affecting ≥ 10% scalp area for at least 3 months Exclusion criteria: Patients having significant spontaneous regrowth of terminal hair History of use of topical or intralesional therapy in the past 1 month and/or phototherapy and/or systemic therapy in the last 2 months Uncontrolled renal, hepatic, haematological or heart disease, or absolute contraindication to corticosteroids or azathioprine
Interventions	Intervention 1: Oral azathioprine 300 mg once weekly for 16 weeks or until complete hair regrowth Intervention 2: Oral betamethasone 5 mg on 2 consecutive days every week for 16 weeks or until complete hair regrowth
Outcomes	Assessment at 16 weeks Primary outcomes: Average percentage scalp hair regrowth Average change in SALT score Side effects Secondary outcomes: Comparison of percentage of patients who achieved SALT‐50 Median global assessment score Patient global assessment based on visual analog scoring from 0 to 100 Percentage of patients showing cosmetically acceptable hair regrowth Investigator Global Assessment by evaluation of photographs on a VAS of 0–100 Proportion of patients not responding or worsening Percentage of patients showing relapse
Funding source	Not stated
Declarations of interest	No conflicts
Notes

Guttman‐Yassky 2021.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: multicentre (number not stated) Country: USA
Participants	Randomised: 60 participants (43 females, 13 in the placebo arm and 30 in the dupilumab arm). Participants were randomised in a 2:1 ratio to receive dupilumab (n = 40) or placebo (n = 20). The mean age was 44 years, 71.7% were female (46.5 ± SD 14.4 in the placebo arm and 41.6 ± SD 13.8 in the dupilumab arm). Inclusion criteria: Subjects who are able to understand and voluntarily sign an informed consent document, and adhere to the study visit schedule and other protocol requirements Approximately one‐third of subjects will have active AD skin or a concomitant history of AD. Negative pregnancy test; use of an approved contraceptive method If subject is a female of non‐childbearing potential, proper documentation should be provided. Negative tuberculosis test prior to baseline visit (subjects with a positive or indeterminable test result must have a documented negative workup for tuberculosis and/or completed standard tuberculosis therapy). Specific laboratory criteria, as detailed in the full protocol, should be met (including complete blood count parameters, serum creatinine, and liver function tests) Subjects who are judged to be in otherwise good overall health Exclusion criteria: Subjects who are pregnant or breastfeeding The cause of hair loss is indeterminable and/or other concomitant causes of alopecia History of AA with no evidence of hair regrowth for ≥ 10 years An active bacterial, viral, or helminth parasitic infection or a history of ongoing, recurrent severe infections requiring systemic antibiotics Subject with a known or suspected underlying immunodeficiency or immunocompromised state A concurrent or recent history of severe, progressive, or uncontrolled renal, hepatic, haematological, intestinal, metabolic, endocrine, pulmonary, cardiovascular, or neurological disease Active hepatitis B, hepatitis C, human immunodeficiency virus (HIV), or positive HIV serology at the time of screening A suspected or active lymphoproliferative disorder or malignancy or a history of malignancy within 5 years before the baseline assessment Subjects who received a live attenuated vaccine ≤ 30 days prior to study randomisation; any uncertain or clinically significant laboratory abnormalities that may affect the interpretation of study data or endpoints Any other medical or psychological condition that, in the opinion of the investigator, may present additional unreasonable risks as a result of study participation and/or interfere with clinic visits and necessary study assessments History of adverse systemic or allergic reactions to any component of the study drug Severe, untreated asthma or a history of life‐threatening asthma exacerbations while on appropriate anti‐asthmatic medications Use of systemic immunosuppressive medications, including, but not limited to, cyclosporine, systemic or intralesional corticosteroids, mycophenolate mofetil, azathioprine, methotrexate, tacrolimus, or ultraviolet (UV) within 4 weeks prior to randomisation Use of an oral JAK inhibitor within 12 weeks prior to the baseline visit Use of topical corticosteroids, and/or tacrolimus, and/or pimecrolimus within 1 week prior to the baseline visit Previous treatment with dupilumab, current use or a plan to use antiretroviral therapy at any time during the study
Interventions	Intervention 1: Subcutaneous dupilumab 300 mg once weekly for 24 weeks Intervention 2: Placebo 300 mg once weekly for 24 weeks
Outcomes	Primary outcome: Change in SALT score assessed at week 24 Secondary outcomes: Change in SALT from week 48 to week 24 and to baseline The proportion of patients achieving 30%/50%/75% SALT score improvement (SALT30/50/75) at week 24 and 48 Change in the Alopecia Areata Symptom Impact Scale (AASIS) and Alopecia Areata Quality of Life Index questionnaires (AA‐QLI), with Alopecia Areata Physician's Global Assessment (AA‐PGA) improvement Change in eyelash and eyebrow scores and in Eczema Area and Severity Index (EASI) Safety assessments Safety labs and blood markers potentially related with AA/atopy (e.g. percentage of eosinophils, total serum IgE) were evaluated at baseline and weeks 12, 24, 36 and 48
Funding source	Pharmaceutical company: Regeneron/Sanofi
Declarations of interest	There were conflicts. EGY has served as a consultant for AbbVie, Amgen, Allergan, Asana Bioscience, Celgene, Concert, Dermira, DS Biopharma, Escalier, Galderma, Glenmark, Kyowa Kirin, LEO Pharmaceuticals, Lilly, Mitsubishi Tanabe, Novartis, Pfizer, Regeneron, Sanofi, and Union Therapeutics; a member of advisory boards of Allergan, Asana Bioscience, Celgene, DBV, Dermavant, Dermira, Escalier, Galderma, Glenmark, Kyowa Kirin, LEO Pharma, Lilly, Novartis, Pfizer, Regeneron, and Sanofi; and a recipient of research grants from AbbVie, AnaptysBio, AntibioTx, Asana Bioscience, Boehringer‐ Ingelheim, Celgene, DBV, Dermavant, DS Biopharma, Galderma, Glenmark, Innovaderm, Janssen Biotech, Kiniska Pharma, LEO Pharmaceuticals, Lilly, Medimmune, Sienna Biopharmaceuticals, Novan, Novartis, Ralexar, Regeneron, Pfizer, UCB, and Union Therapeutics. MGL has received grant support from Ortho Dermatologics, UCB, AbbVie, Amgen, Eli Lilly, Incyte, and Janssen Research and Development, grant support and consulting fees from Pfizer, Dermavant Sciences, LEO Pharma, Boehringer Ingelheim, and Arcutis Biotherapeutics, and consulting fees from Allergan, Almirall, Avotres Therapeutics, BirchBioMed, Bristol‐Myers Squibb, Cara Therapeutics, Castle Biosciences, Corrona, EMD Serono, Evelo Biosciences, Inozyme Pharma, Meiji Seika Pharma, Menlo Therapeutics, Mitsubishi Pharma, NeuroDerm, Promius Pharma–Dr. Reddy's Laboratories, Theravance Biopharma, Verrica Pharmaceuticals, Aditum Bio, BMD Skincare, and Kyowa Kirin. JGK has received grants paid to The Rockefeller University from Amgen, Boehringer Ingelheim, Bristol‐Myers Squibb, Dermira, Innovaderm, Janssen, Kadmon, Kineta, Kyowa, LEO Pharma, Lilly, Novartis, Paraxel, Pfizer, Provectus, Regeneron, and Vitae and personal fees from AbbVie, Baxter, Biogen Idec, Boehringer Ingelheim, Bristol‐Myers Squibb, Delenex, Dermira, Janssen, Kadmon, Kineta, Lilly, Merck, Novartis, Pfizer, Sanofi, Serono, and XenoPort. All other authors declared no conflict of interest.
Notes

Hamdino 2021.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Egypt
Participants	Randomised: 40 (0 females) Inclusion criteria: Adult patients with patchy AA on the scalp Less than 50% of scalp surface area Did not receive any treatment for AA during the last 3 months Exclusion criteria: Children and patients less than 18 years old Pregnant and lactating women Patients with chronic hepatic or haematological disorders Immunocompromised patients Patients with extensive types (alopecia totalis, universalis and surface area > 50%) Patients who received any treatment for AA within 3 months prior to the study also, those with recent AA less than 3 months
Interventions	Intervention 1: Methotrexate 25 mg/mL once every 3 weeks for a maximum of 12 weeks Intervention 2: Triamcinolone acetonide 10 mg/mL once every 3 weeks for a maximum of 12 weeks
Outcomes	Assessment at 24 weeks SALT score Hair regrowth (regrowth scale) Trichoscopic findings Patients satisfaction Side effects
Funding source	Not stated
Declarations of interest	No conflicts
Notes

Hay 1998.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: UK
Participants	Randomised: 84 Inclusion criteria: Patients with diagnosis of alopecia areata Exclusion criteria: Patients with a medical history of hypertension, epilepsy or pregnancy, concomitant androgenetic alopecia
Interventions	Intervention 1: Aromatherapy (essential oils mixed in a carrier oil) once a day for 28 weeks Intervention 2: Carrier oil (placebo), once a day
Outcomes	Assessment at 28 weeks 4‐point scale to measure changes in hair growth Improvement in alopecia Reduction in affected area, cm2
Funding source	Not stated
Declarations of interest	Not stated
Notes

Hedge 2020.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: India
Participants	Randomised: 50 Inclusion criteria: Not stated Exclusion criteria: Patients who received any treatment (systemic/topical) for AA in past 8 weeks, coexisting androgenetic alopecia, bleeding diathesis, pregnant and lactating females
Interventions	Intervention 1: Triamcinolone acetonide 10 mg/mL, once every 4 weeks for 12 weeks Intervention 2: Intralesional platelet‐rich plasma, once every 4 weeks for 12 weeks Intervention 3: Intralesional saline 9 mg/mL (placebo), once every 4 weeks for 12 weeks
Outcomes	Assessment at week 20: SALT score Absolute regrowth Percentage regrowth Adverse effects
Funding source	No funding
Declarations of interest	No conflicts
Notes

Iraji 2014.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Iran
Participants	Randomised: 96 (26 females) Inclusion criteria: Less than 50% involvement of the scalp No prior history of autoimmune disease No recorded sensitivity to tacrolimus Age above 2 years Not pregnant and not planning to become pregnant during the course of study No history of corticosteroids in the last three months, and no history of topical corticosteroids, anthralin or PUVA treatment Having had the diagnosis for less than 6 months Exclusion criteria: Not stated
Interventions	Intervention 1: Minoxidil topical solution 50 mg/mL + tacrolimus cream 1 mg/mL, twice a day for 8 weeks Intervention 2: Minoxidil topical solution 50 mg/mL + placebo cream, twice a day for 8 weeks
Outcomes	Assessment at 24 weeks Mean difference in surface areas
Funding source	Not stated
Declarations of interest	Not stated
Notes

Kapoor 2020.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: India
Participants	Randomised: 44 (27.1 (7.08) years, 22 females) Inclusion criteria: Subjects aged 18–50 years with AA AA on the scalp (< 25% involvement) Patients who had not taken any form of treatment in the last 6 months Subjects who were willing to sign the written consent form before participating in the study Exclusion criteria: Subjects aged < 18 or > 50 years Patients with a history of diabetes, hypertension, thromboembolism, bleeding disorders and abnormal coagulogram Patients who were already on steroids Patients with active infection at the local site and with keloidal tendency.
Interventions	Intervention 1: Intradermal triamcinolone acetonide 10 mg/mL, once every 3 weeks for 12 weeks Intervention 2: Intralesional platelet‐rich plasma, once every 3 weeks for 12 weeks
Outcomes	Assessment at week 24: Trends in SALT score and mean SALT score difference with respect to baseline Percentage reduction in baseline SALT score used to assess the grade of improvement Side effects
Funding source	No funding
Declarations of interest	No conflicts
Notes

Kar 2005.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: India
Participants	Randomised: 36 participants (10 females) Intervention 1: 23 participants, 6 females (30%). Mean age of onset 23.0 ± SD 7.2 years Intervention 2: 20 participants, 4 females (25%). Mean age of onset 27.5 ± SD 9.2 years Intervention 1: average duration of disease 3.07 ± SD 1.3 years intervention; 2: average duration of disease 2.8 ± SD 2.0 years Inclusion criteria: Patients with severe AA (40% loss of scalp hair or > 10 patches scattered over the scalp and body) Exclusion criteria: Patients with diabetes mellitus, thyroid disorders, peptic ulcer, severe hypertension, cardiac failure, active infection, nephropathy, alopecia universalis, or pregnancy, and those who had received oral/injectable steroid therapy during the past 3 months
Interventions	Intervention 1: Oral prednisolone 200 mg once weekly for 12 weeks Intervention 2: Oral placebo once weekly for 12 weeks
Outcomes	Hair regrowth at 12 weeks Relapse at 24 weeks Side effects at 24 weeks
Funding source	No funding
Declarations of interest	No conflicts
Notes

Khoury 1992.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: USA
Participants	Randomised: 20 (11 females) Inclusion criteria: Patients with alopecia areata (AA) involving 26‐99% of the scalp Exclusion criteria: Patients who had ever received minoxidil before, or any other topical or systemic treatment during the 3 months prior to their enrolment
Interventions	Intervention 1: Minoxidil topical solution 5 mg/mL twice daily for 12 weeks Intervention 2: Placebo topical solution twice daily for 12 weeks
Outcomes	Assessment at 24 weeks Hair regrowth Number and appearance of hair follicles Presence of perifollicular mononuclear leukocyte infiltrates
Funding source	Pharmaceutical company funding: a grant from The Upjohn Company
Declarations of interest	Not stated
Notes

Kianfar 2021.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Iran
Participants	Randomised: 60 (24.26 (10.7) years, 25 females) Inclusion criteria: Definite diagnosis of AA Aged 18–60 Having at least two patches or more on the face or scalp Patch size of 2 × 2 to 4 × 4 cm2 No use of any hair‐related treatment in the past 3 months Exclusion criteria: Other types of AA such as totalis, universalis Pregnancy or breastfeeding Underlying diseases, such as infection, immunological disorders, and endocrine diseases, which cause hair loss Photosensitive skin Patients' unwillingness to continue cooperation in the trial Progressive AA during the trial
Interventions	Intervention 1: 308 nm excimer laser. Once a week for 3 months Intervention 2: triamcinolone 5 mg/mL injections, every 4 weeks for 3 months
Outcomes	Assessment at 1, 12 and 16 weeks: Hair regrowth and the occurrence of adverse events with photographs, trichoscopy and dermoscopy Quality of life with the Dermatology Life Quality Index (DLQI)
Funding source	Not stated
Declarations of interest	No conflicts
Notes

King 2022a.

Study characteristics
Methods	Study design: BRAVE‐AA1: double‐blind RCT adaptative phase 2/3 although only results of the phase 3 were considered here Number of centres: 169 Country: South Korea, Mexico, Japan, Taiwan, China, Israel, Australia, Brazil, Argentina, USA
Participants	Randomised: BRAVE‐AA1: 654 participants Placebo: n = 189, mean age 37.4 ± SD 12.9, 109 females (57.7 %) Baricitinib 2 mg: n = 184, mean age 38.0 ± SD 12.8, 109 females (59.2%) Baricitinib 4 mg: n = 281, mean age 36.3 ± SD 13.3, 165 females (58.7%) Inclusion criteria: same for BRAVE‐AA1 and BRAVE‐AA2 (King 2022b) Severity of Alopecia Tool (SALT) score of 50 or higher and a current episode of alopecia areata lasting more than 6 months to less than 8 years, without spontaneous improvement (≤ 10‐point reduction in the SALT score) during the previous 6 months Patients with episodes longer than 8 years if episodes of hair regrowth, spontaneous or while receiving treatment, were observed on the affected areas of the scalp during the past 8 years At least 18 years and less than or equal to 60 years for males (less than or equal to 70 years of age for females) at the time of informed consent Exclusion criteria: same for BRAVE‐AA1 and BRAVE‐AA2 (King 2022b) “Diffuse” pattern alopecia areata or other forms of alopecia Patients who had been treated with the following agents were excluded: topical glucocorticoids on the scalp or eyebrows within 1 week before randomisation, systemic or intralesional glucocorticoids (including intra‐articular injections) within 8 weeks before randomisation, and topical or oral JAK inhibitors within 4 or 8 weeks, respectively, before randomisation Patients with an inadequate response to oral JAK inhibitors after at least 12 weeks of treatment
Interventions	Same for BRAVE‐AA1 and BRAVE‐AA2 (King 2022b) Intervention 1: Oral baricitinib 2 mg, once a day for 36 weeks Intervention 2: Oral baricitinib 4 mg, once a day for 36 weeks Intervention 3: Placebo, once a day for 36 weeks For the analysis, active intervention 1 (oral baricitinib 2 mg) and active intervention 2 (oral baricitinib 4 mg) were grouped into a single arm which was compared to placebo.
Outcomes	Same for BRAVE‐AA1 and BRAVE‐AA2 (King 2022b) Primary outcomes Assessment at week 36: Percentage of participants achieving SALT ≤ 20 Secondary outcomes Assessment at week 36: Percentage of participants with PRO for scalp hair assessment score of 0 or 1 with a ≥ 2‐point improvement from baseline amongst participants with a score of ≥ 3 at baseline Percentage of participants achieving ClinRO measure for EB hair loss 0 or 1 with ≥ 2‐point improvement from baseline Percentage of participants achieving ClinRO measure for Eye Lash (EL) hair loss 0 or 1 with ≥ 2‐point improvement from baseline Percent change from baseline in SALT score SALT score of 10 or less Decrease of at least 90% in SALT score Assessment at week 24: SALT score of 20 or less SALT score of 10 or less Assessment of week 16: SALT score of 20 or less Assessment of week 12: Decrease of at least 50% in SALT score
Funding source	Funded by Eli Lilly under licence from Incyte
Declarations of interest	There were conflicts of interest. Most of the authors were employees or had a relationship with the company through a grant or via an advisory role.
Notes	King 2022a and King 2022b refer to two identical randomised controlled trials published simultaneously in the same manuscript at NEJM (King B, Ohyama M, Kwon O, Zlotogorski A, Ko J, Mesinkovska NA, et al for the BRAVE‐AA Investigators. Two phase 3 trials of baricitinib for alopecia areata. New England Journal of Medicine; p2022;386(18):1687‐99. [DOI: 10.1056/NEJMoa2110343])

King 2022b.

Study characteristics
Methods	Study design: BRAVE‐AA2: double‐blind RCT phase 3 Number of centres: 169 Country: South Korea, Mexico, Japan, Taiwan, China, Israel, Australia, Brazil, Argentina, USA
Participants	Randomised: BRAVE‐AA2: 546 participants Placebo: n = 156, mean age 37.1 ± SD 12.4, 98 females (62.8%) Baricitinib 2 mg: n = 156, mean age 39.0 ± SD 13.0, 103 females (66.0%) Baricitinib 4 mg: n = 234, mean age 38.0 ± SD 12.7, 144 females (61.5%) Inclusion criteria: same for BRAVE‐AA1 (King 2022a) and BRAVE‐AA2 Severity of Alopecia Tool (SALT) score of 50 or higher and a current episode of alopecia areata lasting more than 6 months to less than 8 years, without spontaneous improvement (≤ 10‐point reduction in the SALT score) during the previous 6 months Patients with episodes longer than 8 years if episodes of hair regrowth, spontaneous or while receiving treatment, were observed on the affected areas of the scalp during the past 8 years At least 18 years and less than or equal to 60 years for males (less than or equal to 70 years of age for females) at the time of informed consent Exclusion criteria: same for BRAVE‐AA1 (King 2022a) and BRAVE‐AA2 “Diffuse” pattern alopecia areata or other forms of alopecia Patients who had been treated with the following agents were excluded: topical glucocorticoids on the scalp or eyebrows within 1 week before randomisation, systemic or intralesional glucocorticoids (including intra‐articular injections) within 8 weeks before randomisation, and topical or oral JAK inhibitors within 4 or 8 weeks, respectively, before randomisation Patients with an inadequate response to oral JAK inhibitors after at least 12 weeks of treatment
Interventions	Same for BRAVE‐AA1 (King 2022a) and BRAVE‐AA2 Intervention 1: Oral baricitinib 2 mg, once a day for 36 weeks Intervention 2: Oral baricitinib 4 mg, once a day for 36 weeks Intervention 3: Placebo, once a day for 36 weeks For the analysis, active intervention 1 (oral baricitinib 2 mg) and active intervention 2 (oral baricitinib 4 mg) were grouped in a single arm which was compared to placebo.
Outcomes	Same for BRAVE‐AA1 (King 2022a) and BRAVE‐AA2 Primary outcomes Assessment at week 36: Percentage of participants achieving SALT ≤ 20 Secondary outcomes Assessment at week 36: Percentage of participants with PRO for scalp hair assessment score of 0 or 1 with a ≥ 2‐point improvement from baseline amongst participants with a score of ≥ 3 at baseline Percentage of participants achieving ClinRO measure for EB hair loss 0 or 1 with ≥ 2‐point improvement from baseline Percentage of participants achieving ClinRO measure for Eye Lash (EL) hair loss 0 or 1 with ≥ 2‐point improvement from baseline Percent change from baseline in SALT score SALT score of 10 or less Decrease of at least 90% in SALT score Assessment at week 24: SALT score of 20 or less SALT score of 10 or less Assessment of week 16: SALT score of 20 or less Assessment of week 12: Decrease of at least 50% in SALT score
Funding source	Funded by Eli Lilly under licence from Incyte
Declarations of interest	There were conflicts of interest. Most of the authors were employees or had a relationship with the company through a grant or via an advisory role.
Notes

Kuldeep 2011.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: India
Participants	Randomised: 78 (26 females) Inclusion criteria: Patients with asymptomatic, non‐scarring, patchy hair loss over scalp and elsewhere, patients with age range of 11 to 50 years in both genders who never received treatment for hair loss Exclusion criteria: Patients with extreme ages (< 10 or > 50 years), pregnancy or lactation, extensive (more than 3 patches) or atypical alopecia areata (alopecia totalis, alopecia universalis, ophiasis, diffuse) and history of allergy to glucocorticoids or tacrolimus
Interventions	Intervention 1: Betamethasone valerate foam 1 mg/mL, twice daily for 12 weeks Intervention 2: Intralesional triamcinolone 10 mg/mL, once every 3 weeks for 12 weeks Intervention 3: Tacrolimus ointment 1 mg/mL, twice daily for 12 weeks
Outcomes	Assessment at 24 weeks Hair regrowth analysed by hair regrowth grade Side effects
Funding source	No funding
Declarations of interest	No conflicts
Notes

Lai 2019.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Australia
Participants	Randomised: 36 participants (mean 41 ± 14.5 years, 29 females) Intervention 1: Cyclosporine, mean age 36.4 ± 11.3, 13 females (72.2%) Intervention 2: Placebo, mean age 45.7 ± 16.2, 16 females (88.9%) Inclusion criteria: Adults aged 18 to 65 years of age with moderate‐to‐severe AA Exclusion criteria: Pregnancy and lactation History of any lymphoproliferative disorder, HIV, hepatitis B, or hepatitis C Hypersensitivity to any ingredient of the medication Use of any hair regrowth treatments before the study without an adequate washout period (generally 5 half‐lives) Inability to adhere to study procedures and visits Any acute or chronic medical or laboratory abnormality that may increase the risk of study participation (i.e. clinically significant, severe, progressive, and uncontrolled diseases)
Interventions	Intervention 1: Oral cyclosporine (4 mg/kg/d), twice daily for 12 weeks Intervention 2: Oral placebo twice daily for 12 weeks
Outcomes	Assessment at 12 weeks Primary outcomes: Proportion of participants achieving 30, 50, 75 and 100% reduction in SALT Change from baseline in SALT score Change from baseline in nonvellus hair count Proportion of participants achieving at least a 1 grade improvement in eyelashes and eyebrows assessment scale scores Secondary outcomes: Change from baseline in Assessment Quality of Life‐8D and Alopecia Areata Symptom scale scores
Funding source	Institutional funding: Supported by the Australia Alopecia Areata Foundation
Declarations of interest	No conflicts
Notes

Lenane 2014.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Canada
Participants	Randomised: 41 (mean 7.3 years, 23 females) Inclusion criteria: Children 2 to 16 years of age with a clinical confirmation of alopecia areata and at least 10% of their scalp surface area affected, whose parents provided written informed consent Exclusion criteria: Any skin or medical condition requiring oral corticosteroids, immunosuppressants, or light therapy within 4 weeks prior to the study; use of inhaled or intranasal steroids during the 14 days prior to enrolment; and/or any topical medications during the 7 days prior to enrolment
Interventions	Intervention 1: Clobetasol propionate cream 0.5 mg/mL, twice daily for 24 weeks Intervention 2: Hydrocortisone cream 10 mg/mL, twice daily for 24 weeks
Outcomes	Assessment at 24 weeks Primary outcomes: Change in scalp surface area with hair loss over time Secondary outcomes: Percentage of change in scalp surface area with hair loss between baseline and 24 weeks At least 50% reduction in scalp surface area with hair loss from baseline to 24 weeks The frequency of adverse events such as skin atrophy, striae, or local symptoms
Funding source	Institutional funding: This study was supported by a RESTRACOMP (Research Training Competition, Research Institute, The Hospital for Sick Children) grant and by Physician Services Inc (PSI) grant 3227216019.
Declarations of interest	No conflicts
Notes

Lohrasb 2015.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Iran
Participants	Randomised: 189 (92 females) Inclusion criteria: Diagnosis of alopecia areata Duration of the disease less than 6 months Exclusion criteria: A person suffering from another type of regional hair loss. Alopecia areata caused by fungal diseases Hair loss due to certain diseases such as cancer (chemotherapy) A person suffering from another type of regional hair loss
Interventions	Intervention 1: Minoxidil topical solution 20 mg/mL, twice daily for 24 weeks Intervention 2: Aromatherapy (rosemary topical solution), twice daily for 24 weeks Intervention 3: Minoxidil topical solution 20 mg/mL with rosemary topical solution, twice daily for 24 weeks
Outcomes	Assessment at 36 weeks Recovery process Hair regrowth and whether the hair grown villous or terminal, complete or single Side effects Nail changes
Funding source	Not stated
Declarations of interest	Not stated
Notes

Luo 1991.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: China
Participants	Randomised: 162 participants (63 females, 38.8%) Intervention 1: 102 participants (42 females, 41.2%), age ranged from 11‐58 years old Intervention 2: control group with 60 participants (21 females, 35%), age ranged from 12‐57 years old Inclusion criteria: Not stated Exclusion criteria: Not stated
Interventions	Intervention 1: minoxidil topical solution 20 mg/mL, twice a day for 12 weeks Intervention 2: placebo topical solution twice daily for 12 weeks
Outcomes	Assessment at 12 weeks Effective rate Curative rate Side effects
Funding source	Not stated
Declarations of interest	Not stated
Notes

Mancuso 2003.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Italy
Participants	Randomised: 61 (41 (13) years, 35 females) Inclusion criteria: Male and female patients Aged > 18 years With mild‐to‐moderate (S1 or S2 grade, i.e. hair loss < 26%) patchy AA. Exclusion criteria: Severe AA (= S3) Positive history of steroid allergy Previous (< 4 weeks) topical or systemic treatments for AA Pregnant or lactating women
Interventions	Intervention 1: Betamethasone valerate foam 1 mg/mL, twice daily for 12 weeks Intervention 2: Betamethasone dipropionate lotion 0.5 mg/mL, twice daily for 12 weeks
Outcomes	Assessment at 20 weeks Primary outcomes: Hair regrowth score between groups Secondary outcomes: Safety and tolerability
Funding source	Not stated
Declarations of interest	Not stated
Notes

Mikhaylov 2019.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: USA
Participants	Randomised: 30 (mean 41.8 years, 21 females) Inclusion criteria: Males or females, 18 years or older at the time of signing the informed consent document Subject with a diagnosis of patchy scalp alopecia areata present for at least 6 months, and up to a maximum of 10 years Patients with ≥ 50% and < 95% total scalp hair loss at baseline as measured using the SALT score to qualify as moderate‐to‐severe AA; and 95%‐100% scalp hair loss to qualify as AA totalis/universalis Must meet the following laboratory criteria: White blood cell count ≥ 3000/mm3 (≥ 3.0 x 109/L) and < 14,000/mm3 (< 14 x 109/L); Platelet count ≥ 100,000/μL (≥ 100 x 109/L); Serum creatinine ≤ 1.5 mg/dL (≤ 132.6 μmol/L); AST (SGOT) and ALT (SGPT) ≤ 2 x upper limit of normal (ULN). If the initial test shows ALT or AST > 2 times the ULN, one repeat test is allowed during the screening phase. Total bilirubin ≤ 2 mg/dL (34 μmol/L). If the initial test shows total bilirubin > 2 mg/dL (34 μmol/L), one repeat test is allowed during the screening phase. Haemoglobin ≥ 10 g/dL (≥ 6.2 mmol/L) Females of childbearing potential (FCBP) must have a negative pregnancy test at screening and baseline. Male subjects (including those who have had a vasectomy) who engage in activity in which conception is possible must use barrier contraception. No evidence of hair regrowth present at baseline Exclusion criteria: Clinically significant (as determined by the investigator) cardiac, endocrine, pulmonary, neurologic, psychiatric, hepatic, renal, haematologic, or immunologic disease, or other major uncontrolled diseases that will affect the health of the subject during the study or interfere with the interpretation of study results Hepatitis B surface antigen positive at screening (visit 1) Hepatitis C antibody‐positive at screening (visit 1) History of positive human immunodeficiency virus (HIV), or congenital or acquired immunodeficiency Active TB or a history of inadequately treated TB Active substance abuse or a history of substance abuse within six months prior to screening Pregnant or breastfeeding History of allergy to any component of the PI Major surgery within eight weeks prior to screening (visit 1) and/or planned surgery during the length of the study Malignancy or history of malignancy, except for: treated (i.e. cured) basal cell or squamous cell in situ skin carcinomas; treated (i.e. cured) cervical intraepithelial neoplasia (CIN) or carcinoma in situ of the cervix with no evidence of recurrence within 5 years prior to screening (visit 1). Unstable asthma; prior episode(s) of life‐threatening asthma; or asthma that requires inhaled budesonide or equivalent at >1200 μg/day or fluticasone propionate at > 880 μg/day along with another anti‐asthmatic drug such as a long‐acting beta‐agonist A history of and/or concurrent condition of serious hypersensitivity (e.g. anaphylaxis) to drugs, foods, or other allergens without access to emergency rescue medication such as epinephrine Persistent or recurring bacterial infection requiring systemic antibiotics, or clinically significant viral or fungal infections, within two weeks of screening Active skin infection requiring systemic antimicrobials at baseline/randomisation Skin lesion(s) due to conditions other than AA that would interfere with the study specified assessments Prior treatment with apremilast, or participation in a clinical study involving apremilast Use of phototherapy (i.e. UVB, UVA) or systemic immunosuppressive drugs (including, but not limited to, cyclosporine, corticosteroids, mycophenolate mofetil, azathioprine, methotrexate, or tacrolimus), or oral preparations of herbal immunomodulatory medications within four weeks prior to baseline/randomisation Use of interferon‐γ within 12 weeks prior to baseline/randomisation (visit 2). Use of abatacept, adalimumab, certolizumab pegol, etanercept, golimumab, infliximab, or tocilizumab within 12 weeks prior to baseline/randomisation Use of oral janus kinase (JAK) inhibitors (e.g. tofacitinib, ruxolitinib) within 12 weeks prior to baseline/randomisation Use of omalizumab, rituximab, ustekinumab, alefacept, briakinumab, or other therapeutic antibody products within 24 weeks prior to baseline/randomisation (visit 2) Use of any investigational drug within four weeks or five PK or PD half lives (whichever is longer) prior to baseline/randomisation (visit 2) Use of topical corticosteroid preparations, topical calcineurin inhibitors, or other topical preparations with immunomodulatory properties within 2 weeks prior to baseline/randomisation Prior history of suicide attempt at any time in the subject’s lifetime prior to baseline (visit 2) or major psychiatric illness requiring hospitalisation within 3 years prior to baseline History of male or female pattern hair loss Ludwig stage III or Hamilton > stage V
Interventions	Intervention 1: Oral apremilast 30 mg, twice daily for 24 weeks Intervention 2: Oral placebo, twice daily for 24 weeks
Outcomes	Primary outcomes Proportion of patients achieving SALT50 or greater at 24 weeks Secondary outcomes Percent change in SALT score at weeks 24 and 48 Proportion of subjects achieving an alopecia areata Physician’s Global Assessment (AAPGA) score of 3 or above at weeks 24 and 48 Percentage change from baseline in the Alopecia Areata Symptom Impact Scale (AASIS) at weeks 24 and 48 Percentage change from baseline in the Alopecia Areata Quality of Life questionnaire (AA‐QoL) at weeks 24 and 48 Semiquantitative score using SALT subclasses at weeks 24 and 48 Adverse events
Funding source	Pharmaceutical company funding: this study was funded by Celgene (Grant number AP‐CL‐ALOP‐PI‐005776).
Declarations of interest	Authors declared conflicts "Daniela Mikhaylov, Ana Pavel, Christopher Yao, Giselle Singer, Mark Taliercio, Rachel Karalekas, John Nia, Peter Hashim, Grace Kimmel, Danielle Baum, Yasaman Mansouri, and An‐jali S. Vekaria have no conflicts of interest to disclose. Dr. Emma Gutt‐man‐Yassky is a board member of Sanofi Aventis, Regeneron, Stie‐fel/GlaxoSmithKline, MedImmune, Celgene, Anacor, AnaptysBio, Celsus, Dermira, Galderma, Glenmark, Novartis, Pfizer, Vitae and Leo Pharma; has received consultancy fees from Regeneron, Sanofi, Med‐Immune, Celgene, Stiefel/GlaxoSmithKline, Celsus, BMS, Amgen, Drais, AbbVie, Anacor, AnaptysBio, Dermira, Galderma, Glenmark, LEO Pharma, Novartis, Pfizer, Vitae, Mitsubishi Tanabe and Eli Lilly; and has received research support from Janssen, Regeneron, Celgene, BMS, Novartis, Merck, LEO Pharma and Dermira. Dr. Mark Lebwohl is an employee of Mount Sinai, which receives research funds from Amgen, Anacor Pharmaceuticals Inc, Boehringer Ingelheim, Celgene, Eli Lilly and Company, Janssen Biotech, Kadmon Corporation, LEO Pharmaceuticals, MedImmune, Novartis Pharmaceuticals Corporation, Pfizer, Sun Pharmaceuticals Industries Ltd, and Valeant Pharmaceuticals. The authors did not receive any form of compensation, either directly or indirectly, from any company or agency related to the development, authorship, or publication of this article."
Notes

Narahari 1996.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: unclear Country: India
Participants	Randomised: 100 (23 females) Inclusion criteria: Patients with single lesion on the scalp, less than 6 months of duration, first episode, without ophiasis, age group between 20‐50, belonging to the type I of Ikeda's classification and not having tried any other forms of medication Exclusion criteria: Not stated
Interventions	Intervention 1: Anthralin cream 11.5 mg/mL, once every two weeks for 8 weeks Intervention 2: Intralesional triamcinolone acetonide 10 mg/mL once every two weeks for 8 weeks
Outcomes	Assessment at 24 weeks Complete regrowth of hair Adequate response Adverse effects Relapse
Funding source	Not stated
Declarations of interest	Not stated
Notes

Nouh 2022.

Study characteristics
Methods	Methods Study design: RCT, parallel groups Number of centres: 1 Country: Egypt
Participants	Randomised: 80 (34.17 years) Inclusion criteria: Age above 5 years Cases with less than 25% scalp involvement Presence of at least two alopecic patches over the scalp Exclusion criteria: Intolerance to cold Active secondary infection Patients were not willing to follow the protocol
Interventions	Intervention 1: Fractional CO2 laser Intervention 2: Cryotherapy method. Every 2 weeks for a maximum of five sittings (at 0, 2, 4, 6, and 8 weeks)
Outcomes	The efficacy of treatment will be evaluated with: Digital photography before and after treatment Regrowth scales every 2 weeks Trichoscopy every 2 weeks SALT score
Funding source	Not stated
Declarations of interest	No conflict of interest
Notes

Olsen 2019.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 17 Country: USA
Participants	Randomised: 78 (51 females) Inclusion criteria: Patients aged ≥ 18 to ≤ 70 years with a Severity of Alopecia Tool (SALT) score at baseline of 25% to 99% scalp hair loss and either a positive hair pull test or presence of exclamation mark hairs. More than or equal to 10% of the randomised population could be patients with AT or AU of < 2 years duration. Exclusion criteria: Key exclusion criteria included AT, AU or exclusively ophiasis; < 12 months of 25% to < 50% hair loss and < 6 months of 50% to 99% hair loss; evidence of diffuse, spontaneous terminal hair regrowth; history of spontaneous remission in previous 2 years; and cytopenias at screening
Interventions	Intervention 1: Ruxolitinib cream 15 mg/mL, twice daily for 24 weeks Intervention 2: Placebo cream twice daily for 24 weeks
Outcomes	Assessment at 24 weeks Primary outcomes Percentage of all patients achieving > 50% improvement in SALT score Secondary outcomes Percentage of patients with 50% to 100% scalp involvement at baseline achieving a SALT50 Percentage of patients achieving a ≥ 90% improvement in SALT score
Funding source	Pharmaceutical company funding: Incyte Corporation
Declarations of interest	Authors declared conflicts "EAO has served as a consultant and investigator for Aclaris and as a consultant for Bioniz, Incyte, Lilly, and Pfizer. She did not receive any compensation from Incyte in writing this article. DK and KS are employees and stockholders of Incyte Corporation. MKH has served as a consultant for Incyte, Concert, and Bioniz, and as an investigator for Incyte, Aclaris, Lilly, and Pfizer".
Notes

Ozmen 2015.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Turkey
Participants	Randomised: 20 (21.5 (11.2) years, 14 females) Inclusion criteria: At least five years of age with localised AA patches Exclusion criteria: AU, AT, difuse AA or any other dermatoses involving the scalp, known prior allergic reactions to the medication, pregnancy and women who were lactating Patients who had used any topical treatment in the previous two months
Interventions	Intervention 1: Aromatherapy (essential oils mixed in a carrier oil) once daily for 12 weeks Intervention 2: Carrier oil (placebo) once daily for 12 weeks
Outcomes	Assessment at 20 weeks Mean hair growth rate Mean clinical assessment Extent of affected area Adverse effects
Funding source	Not stated
Declarations of interest	Not stated
Notes

Price 1987a.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: USA
Participants	Randomised: 30 (23 females) Inclusion criteria: Patients with extensive alopecia areata affecting more than 50% of the scalp area or with the more extensive conditions of alopecia totalis and alopecia universalis Exclusion criteria: Subjects with cardiovascular disease, hypertension, or severe systemic illness
Interventions	Intervention 1: Minoxidil topical solution 30 mg/mL twice daily for 52 weeks Intervention 2: Placebo topical solution twice daily for 52 weeks
Outcomes	Assessment at 52 weeks Hair growth Adverse effects
Funding source	Not stated
Declarations of interest	Not stated
Notes

Price 1987b.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: USA
Participants	Randomised: 30 (females) Inclusion criteria: Patients with extensive alopecia areata affecting 25% to 100% of the scalp Exclusion criteria: Those with hypertension, cardiovascular disease, or severe systemic illness
Interventions	Intervention 1: Minoxidil topical solution 30 mg/mL twice daily for 12 weeks Intervention 2: Placebo topical solution for 12 weeks
Outcomes	Assessment at 64 weeks Hair growth Adverse effects
Funding source	Not stated
Declarations of interest	Not stated
Notes

Rajan 2021.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: India
Participants	Randomised: 105 (38 females). One hundred and sixty‐eight patches were divided into 4 quadrants. Each quadrant was randomised to each intervention. Inclusion criteria: age more than 12 years, having at least one scalp AA patch, AA patch size more than 2.5 cm x 2.5 cm, less than 50% scalp involvement, AA patches of less than one year duration and at least 1 month treatment wash‐off period Exclusion criteria: patches showing hair regrowth, local infection, pregnant or lactating women and patients with immune‐compromised state, uncontrolled diabetes mellitus, tuberculosis, hypertension, and bleeding diathesis
Interventions	Intervention 1: Intralesional triamcinolone acetonide 2.5 mg/mL, once every 4 weeks for 12 weeks Intervention 2: Intralesional triamcinolone acetonide 5 mg/mL, once every 4 weeks for 12 weeks Intervention 3: Intralesional triamcinolone acetonide 10 mg/mL, once every 4 weeks for 12 weeks Intervention 4: Placebo, once every 4 weeks for 12 weeks
Outcomes	Assessment at 12 weeks Hair density assessment using RGS Presence or absence of dermoscopic disease activity signs Local adverse effects
Funding source	Not stated
Declarations of interest	No conflicts
Notes

Rashad 2022.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Egypt
Participants	Randomised: 60 (11 females) Inclusion criteria: Adult patients with localised patchy alopecia areata Exclusion criteria: Patients who received systemic or topical treatment for alopecia areata in the last 3 months prior to the start of the study Patients who were pregnant or lactating Patients who had bleeding or coagulation disorders or were immunocompromised Patients with known history of hypersensitivity to vitamin D, patients taking vitamin D supplements in the last 6 months, patients treated with topical vitamin D analogues, and patients with diseases known to alter vitamin D level as autoimmune diseases, liver, or renal diseases
Interventions	Intervention 1: Intralesional injection of vitamin D3 2.5 mg/ mL every 4 weeks for a maximum of 3 sessions Intervention 2: Intralesional injection of normal saline 0.9% every 4 weeks for 3 sessions
Outcomes	Assessment at every month for 3 months: Percentage of hair regrowth with the 5‐point semiquantitative regrowth score (RGS) Dermoscopy for yellow dots, broken hairs black dots, and tapering hairs, and signs of improvement of the disease including short vellus hairs and terminal hairs
Funding source	Not stated
Declarations of interest	No conflicts
Notes

Saif 2012.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Saudi Arabia
Participants	Randomised: 42 (19.5 (7.4) years, 20 females) Inclusion criteria: Patients diagnosed with either AU, AT, or OA Exclusion criteria: Patients with contraindications to steroid therapy
Interventions	Intervention 1: Oral methylprednisolone 15 mg/kg/day, 3 consecutive days once every 2 weeks for 24 weeks Intervention 2: Oral methylprednisolone 15 mg/kg/day, 2 consecutive daily pulses every 3 weeks Intervention 3: Oral methylprednisolone 15 mg/kg/day, 3 consecutive daily pulses every 3 weeks
Outcomes	Assessment at 36 weeks Percentage of hair regrowth Adverse events
Funding source	Institutional funding: College of Medicine Research Center of King Saud University, and the Saudi Society for Dermatology and Dermatologic Surgery, Riyadh, Kingdom of Saudi Arabia
Declarations of interest	No conflicts
Notes

Sayed 2020.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Egypt
Participants	Randomised: 52 (16 females) Inclusion criteria: Aged more than 12 with mild‐moderate cases of AA, covering less than 25% the total scalp area Exclusion criteria: Patients on simultaneous immune suppressive therapy or having any severe medical illness or associated autoimmune disease, patients with cataract, pregnant and lactating women and any contraindication or documented hypersensitivity reaction to each of the procedures Patients on corticosteroids, minoxidil, or other treatment for AA were instructed to stop treatment 1 month before the study.
Interventions	Intervention 1: Superficial liquid nitrogen cryotherapy spray (2‐3 seconds), every 2 weeks for a maximum of six sessions (12 weeks max). Intervention 2: Topical PUVA. Starting dose was 0.25‐0.5J/cm2 weekly increments of 0.12‐0.25J/cm2 twice‐weekly for a maximum of 6 weeks.
Outcomes	Assessment 12 weeks after the end of treatment Percentage of terminal hair growth Recurrence (considered when there was loss of more than 50% of hair or increase of more than 50% in the size and extent of the alopecia patch after good terminal hair regrowth) Adverse effects recorded after each session
Funding source	No funding
Declarations of interest	No conflicts
Notes

Shapiro 1993.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Canada
Participants	Randomised: 15 participants (9 females, 60%) Mean age: ranged from 25 to 70 years Duration of the disease: ranged from 2 to 55 years (mean 12 years) Inclusion criteria: Alopecia areata affecting more than 50% of the scalp Exclusion criteria: Subjects with cardiovascular disease, hypertension, or serious medical illness
Interventions	Intervention 1: Diphencyprone (DPCP) from 0.001 to 10 mg/mL once a week + minoxidil topical solution 50 mg/mL twice a day for 24 weeks Intervention 2: Diphencyprone (DPCP) once a week + placebo topical solution twice a day for 24 weeks
Outcomes	Assessment at 24 weeks Proportion of responders Side effects
Funding source	Institutional funding: grant from National Alopecia Areata Foundation
Declarations of interest	Not stated
Notes

Sousa 2020.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Brazil
Participants	Randomised: 12 Inclusion criteria: Patchy alopecia areata Exclusion criteria: Not stated
Interventions	Intervention 1: Intralesional triamcinolone 2.5 mg/mL once every 4 weeks for 12 weeks Intervention 2: Intralesional betamethasone 0.375 mg/mL once every 4 weeks for 12 weeks Intervention 3: Intralesional betamethasone 1.75 mg/mL once every 4 weeks for 12 weeks Intervention 4: Intralesional saline 9 mg/mL (placebo) once every 4 weeks for 12 weeks
Outcomes	Assessment at 12 weeks Repilation Number of dystrophic hairs Adverse effects
Funding source	Not stated
Declarations of interest	Not stated
Notes

Strober 2009.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 5 Country: USA
Participants	Randomised: 45 (36.3 (11.1) years, 32 females) Inclusion criteria: 18 to 65 years of age, with a diagnosis of chronic, severe, scalp AA defined as at least a 50% to 95% patchy scalp hair loss of at least 6 months’ duration Exclusion criteria: Patients with alopecia totalis, alopecia universalis, or coexisting significant androgenetic alopecia (Norwood‐Hamilton stage IV or greater in males, Ludwig stage III in females) Individuals with CD4 T‐lymphocyte counts below the lower limit of the reference range at screening (as determined by the local laboratory); those with either abnormal hepatic function or haematologic test results Known history of unstable cardiovascular or pulmonary disease, or poorly controlled diabetes mellitus; were seropositive for human immunodeficiency, hepatitis C, or hepatitis B viruses Patients who had a history of recurrent bacterial, fungal, atypical mycobacterial, viral or opportunistic infections; a history of lymphoproliferative or malignant disease; a history of active tuberculosis (TB) or were currently undergoing treatment for active TB; or were pregnant or breastfeeding Individuals with a history of treatment with alefacept or treatment with another investigational medication
Interventions	Intervention 1: Intramuscular alefacept 15 mg once weekly for 12 weeks Intervention 2: Intramuscular placebo
Outcomes	Assessment at 24 weeks Percentage of participants with at least a 50% improvement from the baseline in their SALT score Extent of hair loss Adverse events
Funding source	Institutional funding: a grant from the National Alopecia Areata Foundation, San Rafael, California, and by Biogen Idec and Astellas Pharma US
Declarations of interest	Authors declared conflicts "Dr Strober serves as a speaker for Astellas Pharma US. Dr McMichael serves on the Scientific Advisory Council to the National Alopecia Areata Foundation."
Notes	The drug alefacept was discontinued by its manufacturers, Astellas Pharma, in 2011.

Tawfik 2022.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Egypt
Participants	Randomised: 30 (mean 28.80 ± 6.8 years, 4 females) Inclusion criteria: Patients who suffered from chronic, recurrent disease of minimally 2‐year period, which was resistant to other lines of treatment At least three patches of AA. Doctors demanded patients stop any intralesional, topical, or systemic treatments for AA 90 days before their enclosure. Exclusion criteria: Pregnancy, lactation, presence of any general disease (such as thyroid diseases and blood clotting disorders), other local scalp or hair disorder and the consumption of medicines that impede growth of hair as chemotherapeutics
Interventions	Intervention 1: Platelet‐rich plasma (PRP) injection once a week Intervention 2: Low‐level laser (or) light therapy (LLLT) three times per week Intervention 3: Topical placebo three times per week
Outcomes	Assessment at 1 month and 3 months next to the termination of sessions: Effect on the growth of hairs, density or hair number/cm² and breadth (diameter) measured by: Photography Standardised phototrichograms Assessment at 12 weeks after the last session: Participant’s satisfaction scale
Funding source	None
Declarations of interest	No conflict of interests
Notes

Tiwary 2016.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: India
Participants	Randomised: 24 (10 females) Inclusion criteria: Immunocompetent patients Patients willing for regular follow‐up after consenting for treatment Patients > 18 years of age AA affecting > 20% of scalp area No local/systemic therapy since at least 6 months Exclusion criteria: Patients not willing to undergo treatment and/or regular follow‐up Pregnancy Serious intercurrent medical illness Significant cardiovascular disease AA affecting < 20% of scalp Localised scalp infections/ inflammation Immunosuppressed patients History of local/systemic therapy within 6 months
Interventions	Intervention 1: Squaric acid dibutyl ester (SADBE) from 0.1 to 10 mg/mL once a week for 24 weeks Intervention 2: Diphencyprone (DPCP) from 0.1 to 10 mg/mL once a week for 24 weeks
Outcomes	Assessment at 24 weeks Percentage change in baseline SALT score Percentage change
Funding source	No funding
Declarations of interest	No conflicts
Notes

Toma 2022.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Egypt
Participants	Randomised: 50 patients Inclusion criteria: Patients with localised AA 18 years old and up, of both sexes Exclusion criteria: Patients with alopecia totalis or alopecia universalis Allergy or hypersensitivity to any component of the treatment products Pregnant and lactating women Patients with any underlying systemic disorders Patients who had received any other modalities of treatment in the last 2 months
Interventions	Intervention 1: Topical methotrexate 1% gel twice a day until complete improvement or a maximum period of 12 weeks Intervention 2: Topical minoxidil 5% gel twice a day until complete improvement or a maximum period of 12 weeks
Outcomes	Assessment at 6 and 12 weeks Hair regrowth score using a 5‐point semi‐quantitative scale
Funding source	Not stated
Declarations of interest	No conflicts of interest
Notes

Tosti 1986.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Italy
Participants	Randomised: 119 participants (mean 30 years, 65 females, 54.6%) Intervention 1: 44 (30 adults, 14 children) Intervention 2: 35 (34 adults, 1 child) Intervention 3: 20 (17 adults, 3 children) Intervention 4: 20 (all adults) Age: ranged from 5 to 74 years (average, 30 years) Duration of the disease: ranged from 3 months to 30 years (median, 2 years) Inclusion criteria: Patchy alopecia involving less than 40% of the scalp Exclusion criteria: Not stated
Interventions	Intervention 1: Squaric acid dibutyl ester (SADBE) from 0.0001 to 10 mg/mL once a week for 24 weeks Intervention 2: Diphencyprone (DPCP) from 0.0001 to 10 mg/mL once a week for 24 weeks Intervention 3: Minoxidil topical solution 10 mg/mL twice a day for 24 weeks Intervention 4: Placebo topical solution twice a day for 24 weeks The four arms of the study were grouped according to the type of intervention: topical contact immunosuppressants (i.e. squaric acid dibutyl ester and diphencyprone) were grouped together and compared to placebo or to minoxidil.
Outcomes	Assessment at 24 weeks Proportion of patients with 100% hair regrowth
Funding source	Not stated
Declarations of interest	Not stated
Notes

Tosti 2006.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Italy
Participants	Randomised: 34 (40 (13) years, 26 females) Inclusion criteria: Patients with moderate‐to‐severe AA Exclusion criteria: Patients with known hypersensitivity to topical steroids and concomitant bacterial skin infections
Interventions	Intervention 1: Clobetasol propionate foam 0.5 mg/mL, twice daily, 5 days/week for 12 weeks Intervention 2: Placebo foam 0.5 mg/mL, twice daily 5 days/week for 12 weeks
Outcomes	Assessment at 24 weeks Hair regrowth rate Evolution of Alopecia Grading Score Evolution of regrowth score Adverse events
Funding source	Pharmaceutical company funding: Mipharm Spa sponsored the study and was responsible for preparation of study drugs, protocol, case report form design and the collection and statistical analysis of the data.
Declarations of interest	Not stated
Notes

Trink 2013.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Italy
Participants	Randomised: 45 (25 females) Inclusion criteria: Male and female healthy AA patients with a chronic, recurring disease with at least 2 years duration, and consisting of between 4‐6 symmetrically distributed patches of hair loss for 1 year without treatment Exclusion criteria: Any other medical condition or other scalp or hair diseases
Interventions	A total of three treatments were given for each patient, with an interval of one month from each other. Intervention 1: Intralesional platelet‐rich plasma once every 4 weeks Intervention 2: Intralesional triamcinolone acetonide 2.5 mg/mL once every 4 weeks Intervention 3: Intralesional sterile water (placebo) once every 4 weeks
Outcomes	Assessment at 52 weeks Hair regrowth measured by SALT score Percentage of dystrophic hairs Burning/itching sensation Cell proliferation measured by Ki‐67 levels Adverse effects
Funding source	No funding
Declarations of interest	No conflicts
Notes	This study did not report the number of events for the secondary outcomes (long‐term hair regrowth > 75%) in the placebo group.

Ucak 2012.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Turkey
Participants	Randomised: 103 (24.11 (12.5) years, 41 females) Inclusion criteria: Patients with alopecic plaques in the scalp, beard and eyebrows Exclusion criteria: Patients who were being followed with the diagnosis of AT and AU and those who used systemic drugs for the treatment of AA
Interventions	Intervention 1: Clobetasol propionate cream 0.5 mg/mL twice a day for 12 weeks Intervention 2: Vaseline cream (placebo) twice daily for 12 weeks
Outcomes	Assessment at 12 weeks Hair regrowth score Recovery score Percentage recovery Side effects
Funding source	Not stated
Declarations of interest	No conflicts
Notes

Ucak 2014.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Turkey
Participants	Randomised: 60 (24.26 (10.7) years, 25 females) Inclusion criteria: Patients with alopecic plaques affecting the scalp, eyebrows and ⁄ or beard Exclusion criteria: AT and AU Use of systemic drugs to treat AA Immunosuppression Use of immunosuppresive treatments Pregnancy
Interventions	Intervention 1: Clobetasol propionate cream 0.5 mg/mL twice a day for 12 weeks Intervention 2: Vaseline cream (placebo) twice daily for 12 weeks
Outcomes	Assessment at 12 weeks Hair regrowth score Recovery score Percentage recovery Side effects Relapse (assessed at 24 weeks)
Funding source	Not stated
Declarations of interest	No conflicts
Notes

Ullah 2022.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Pakistan
Participants	Randomised: 116 (20 females) Inclusion criteria: Patients with an age range of 11‐50 years Number of patches (≤ 5) No previous treatment for hair loss Duration of alopecia areata (≤ 6 months) Exclusion criteria: Pregnancy, lactating mothers Atypical alopecia areata Patients under immunosuppressive drugs and systemic glucocorticoids
Interventions	Intervention 1: Clobetasol propionate 0.05% ointment twice a day for 3 months Intervention 2: Topical tacrolimus 0.1% twice a day for 3 months
Outcomes	Assessment at 3 months: Efficacy assessed by percentage of hair regrowth with SALT score
Funding source	Not stated
Declarations of interest	No conflicts
Notes

Ustuner 2017.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Turkey
Participants	Randomised: 231 patches in 83 participants (23 females, 27.7%). Mean age 29.05 ± 9.08 years Saline group: mean age 31.11 ± 5.21 years Betamethasone group: mean age 27.83 ± 8.67 years Triamcinolone group: mean age 29.93 ± 10.46 years Inclusion criteria: 16‐55 years of age, 1‐6 patches of AA, duration of present patches being less than 3 years, localised AA (alopecia subtotalis; less than 50% scalp or beard area involved), and patches with a diameter of less than 3 cm Exclusion criteria: Being pregnant or lactating, history of steroid allergy, immunocompromised diseases, bleeding diathesis, active scalp inflammation, and taking systemic steroids or any other topical medications for AA Patients that had taken any modality of therapy within the last three months
Interventions	Intervention 1: Intralesional betamethasone 4 mg/mL (3 mL max.) once every 4 weeks for a maximum of 6 sessions Intervention 2: Intralesional betamethasone 2 mg/mL (3 mL max.) once every 4 weeks for a maximum of 6 sessions Intervention 3: Intralesional betamethasone 1 mg/mL (3 mL max.) once every 4 weeks for a maximum of 6 sessions Intervention 4: Intralesional triamcinolone 10 mg/mL (3 mL max.) once every 4 weeks for a maximum of 6 sessions Intervention 5: Intralesional triamcinolone 5 mg/mL (3 mL max.) once every 4 weeks for a maximum of 6 sessions Intervention 6: Intralesional triamcinolone 3 mg/mL (3 mL max.) once every 4 weeks for a maximum of 6 sessions Intervention 7: Intralesional saline 9 mg/mL (placebo) (3 mL max.) once every 4 weeks
Outcomes	Assessment at 24 weeks Number of required sessions Hair regrowth score Treatment success rates Percentage of terminal hairs Adverse effects
Funding source	No funding
Declarations of interest	No conflicts
Notes

Verma 2015.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: India
Participants	Randomised: 52 (26.6 (7.38) years, 16 females) Inclusion criteria: Between 18‐46 years of age with alopecia areata involving 10% scalp area Exclusion criteria: Not stated
Interventions	Intervention 1: Oral azathioprine 300 mg once weekly for 16 weeks Intervention 2: Oral betamethasone 5 mg on two consecutive days weekly for 16 weeks
Outcomes	Assessment at 16 weeks Average percentage hair regrowth Change in median percentage scalp hair loss using severity of alopecia tool (SALT) Adverse effects Relapse
Funding source	No funding
Declarations of interest	Not stated
Notes

White 1985.

Study characteristics
Methods	Study design: RCT, cross‐over Number of centres: 1 Country: UK
Participants	Randomised: 15 (median 39.9 years, 10 females) Inclusion criteria: Not stated Exclusion criteria: Not stated
Interventions	Intervention 1: Minoxidil topical solution 30 mg/mL twice a day for 16 weeks Intervention 2: Placebo topical solution twice daily for 16 weeks
Outcomes	Assessment at 16 and 32 weeks Hair regrowth Side effects
Funding source	Not stated
Declarations of interest	Not stated
Notes

Zaher 2014.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Egypt
Participants	Randomised: 30 (35.4 (8.89) years, 12 females) Inclusion criteria: Adults with at least 2 non‐adjacent scalp AA patches. All recruited patients had a Severity of Alopecia Tool (SALT) score of S1 (<= 25% hair loss) with either stationary or progressive course. Exclusion criteria: Patients with alopecia totalis, alopecia universalis, ophiasis and/or those who had received treatment for AA within 3 months prior to the study Patients with reported spontaneous hair regrowth Patients with associated systemic and/or other dermatological diseases
Interventions	Intervention 1: Bimatoprost topical solution 0.3 mg/mL twice daily for 12 weeks Intervention 2: Mometasone furoate foam 1 mg/mL once daily for 12 weeks
Outcomes	Assessment at 12 weeks Average SALT score Average percentage of hair regrowth Time of initial response Side effects
Funding source	Not stated
Declarations of interest	No conflicts
Notes

Zaib 2017.

Study characteristics
Methods	Study design: RCT, parallel groups Number of centres: 1 Country: Pakistan
Participants	Randomised: 80 (26.4 (7.1) years, 41 females) Inclusion criteria: Patients of alopecia areata, clinically diagnosed Exclusion criteria: Not stated
Interventions	Intervention 1: Minoxidil topical solution 50 mg/mL twice a day for 12 weeks Intervention 2: Liquid paraffin (placebo) twice daily for 12 weeks
Outcomes	Assessment at 12 weeks Hair regrowth
Funding source	Not stated
Declarations of interest	Not stated
Notes

AA: Alopecia areata
AA‐PGA: Alopecia Areata Physician's Global Assessment
AA‐QLI: Alopecia Areata Quality of Life Index questionnaire
AA‐QoL: Alopecia Areata Quality of Life questionnaire
AASIS: Alopecia Areata Symptom Impact Scale
ALT: Alanine aminotransferase
AST: Aspartate transaminase
AT: Alopecia totalis
AU: Alopecia universalis
AD: Atopic dermatitis
CIN: Cervical intraepithelial neoplasia
ClinRO: Clinician‐reported outcome
CO₂: Carbon dioxide
DLQI: Dermatology Life Quality Index
DPCP: Diphencyprone
EASI: Eczema Area and Severity Index
EB: Eyebrow
EL: Eye Lash
FCBP: Females of childbearing potential
HIV: Human immunodeficiency virus
HCV: Hepatitis C virus
HBV: Hepatitis B virus
IgE: Immunoglobulin E
IP: Investigational product
JAK: Janus kinases
Ki‐67: Ki‐67 protein
LAD: Lesional area and density
LLLT: Low‐level laser (or) light therapy
min: Minutes
max: Maximum
OA: Ophiasis alopecia
OP: Operational
PD: Pharmacodynamics
PI: Pharmaceutical ingredient
PK: Pharmacokinetics
PRO: Patient‐reported outcome
PRP: Platelet‐rich plasma
PUVA: Psoralen and ultraviolet A
RCT: Randomised controlled trial
RGS: Regrowth score
SADBE: Squaric acid dibutyl ester
SALT: Severity of alopecia areata tool
SD: Standard deviation
sec: Seconds
SGOT: Serum glutamic oxaloacetic transaminase
SGPT: Serum glutamate pyruvate transaminase
TB: tuberculosis
ULN: Upper limit of normal
UVA: Ultraviolet A
UVB: Ultraviolet B
VAS: Visual analogue scale

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Avgerinou 2008	Wrong study design
Bautista 2010	Wrong intervention
Bernardo 2003	Wrong study design
Choe 2017	Wrong study design
Cipriani 2001	Wrong study design
Cotellessa 2001	Wrong study design
Denli 1995	Wrong study design
Ehsani 2009	Wrong intervention
El Taieb 2019	Wrong intervention
El‐Ashmawy 2018	Wrong intervention
El‐Mofty 2018	Wrong intervention
Farah 2016	Wrong intervention
Giorgio 2020	Wrong intervention
Guttman‐Yassky 2018	Wrong intervention
Guttman‐Yassky 2018a	Wrong intervention
Guttman‐Yassky 2019	Wrong study design
Jaiswal 2018	Wrong intervention
Jung 2017	Wrong study design
Kurosawa 2006	Wrong study design
Lattouf 2015	Wrong population
Moosavi 2019	Wrong intervention
Olsen 1992	Wrong study design
Ross 2005	Wrong study design

Characteristics of studies awaiting classification [ordered by study ID]

CTRI/2018/07/014701.

Methods	Randomised, parallel ‐ group trial Location: India
Participants	60 participants. Inclusion criteria: Patients with a clinical diagnosis of alopecia areata restricted to the scalp. Patients with disease of limited area, not requiring systemic treatment.‐Patients above the age of 18 years. Exclusion criteria: Patients suffering from any significant systemic illness. History of taking intralesional injections or any topical or systemic treatment for alopecia areata in the past twelve weeks.
Interventions	Microneedling with the injection of 10mg/mL intralesional triamcinolone acetonide as multiple 0.1 mL injections per cm2. Microneedling with the injection of 10 mg/mL triamcinolone acetonide with a dermaroller as 0.1 mL per cm2.
Outcomes	Primary outcome: Percentage regrowth and uniformity of regrowth and to compare the efficacy of intralesional triamcinolone acetonide to triamcinolone acetonide through micro‐needling. Secondary outcomes: Atrophy over the patch. Pigmentation changes. Telangiectasia.
Notes	Awaiting. We contacted Dr Astha Arora and the response was "The trial has been completed but awaiting publication in an international journal".

CTRI/2018/10/015870.

Methods	Randomised, parallel‐group trial Location: India
Participants	Inclusion criteria: Patients with severe non‐progressive alopecia areata with a SALT score of more than 40 and age less than 18 years.
Interventions	DPCP topical once weekly in half head after sensitisation with 2% DPCP starting with 0.001% DPCP, increasing by 10% every week till desired pruritus or erythema is reached. Oral mini pulse corticosteroids (dexamethasone) 2.5 mg twice a week.
Outcomes	Primary outcomes: Change in SALT score from baseline in both the groups. Cosmetically acceptable hair regrowth or cosmetically unacceptable hair regrowth in both the groups. Secondary outcome: The number of adverse effects in group A versus group B.
Notes	Awaiting. We contacted Dr Rahul Mahajan and the response was "The trial is completed and the manuscript has been submitted to pediatric Dermatology journal for peer review".

IRCT20211109053013N1.

Methods	Randomised, parallel‐group trial Location: Iran
Participants	108 participants Inclusion criteria: Patients with alopecia areata limited (less than 3 alopecia regions) in beard or scalp Failure to receive treatment up to one month before the beginning of the study Patients with informed consent to enter the study Exclusion criteria: Patients with extensive alopecia (more than 3 alopecia regions Patients with systemic diseases and other skin diseases, eye disorders, atrophy or skin infection in the alopecia area Patients sensitive to each of the therapeutic components used in the study Pregnant and nursing women
Interventions	Latanoprost eye drop 0.005% solution Minoxidil 5% solution Latanoprost eye drop 0.005% solution and Minoxidil 5% solution Betamethasone 0.1% solution and minoxidil 5% solution Betamethasone 0.1% solution and latanoprost eye drop 0.005% solution Betamethasone 0.1% solution
Outcomes	Assessment at the beginning and 2 and 6 weeks after treatment: Primary outcome: Percentage of SALT (Severity of Alopecia Tool) Secondary outcome: Patients' satisfaction with treatment
Notes	Awaiting. We contact Dr. Yazdanian and the response was "The trial was published recently and the results are available". The study was published on 18/10/2022, after our search date.

NCT02037191.

Methods	Randomised, parallel, double‐blind trial Location: France
Participants	90 participants. Inclusion criteria: Age 18 to 70 years old. Informed consent. Severe type of AA defined as: 3.1 ‐ AA totalis, universalis 3.2 ‐ evolving for more than 6 months despite previous treatments including photothérapy (PUVA or UVB), applications of super potent topical corticosteroid ( clobetasol propionate), applications of minoxidil 5%, or IV pulse corticosteroids 3.3 ‐ DLQI score superior or equal to 10. Exclusion criteria: Pregnant or breastfeeding women. VIH + patients. Active hepatitis B or C, treatment with immunosuppressants (cyclosporine, mycophenolate mofetil, cyclophosphamide, azathioprine, methotrexate) within 2 months before inclusion. Severe cardiac arrhythmia or severe cardiac insufficiency or severe coronary disease. Liver disorder. Alcohol consumption. Renal failure. Severe diabetes mellitus. Past history of severe infection. Past history of neoplasia ( excluding BCC). Karnofsky index < 50 %. Severe lung disorder. Mental impairment. Symptomatic osteoporosis. Blood cytopenia (haemoglobin < 10 g/l, leucocytes< 3000/mm3, platelets< 100 000/mm3). Albuminemia < 25 g/l.
Interventions	Methotrexate 20 to 25 mg once a week for 6 months. Patients who experience at least a 25% hair regrowth after the month 5 evaluation will continue methotrexate or placebo from month 6 to the end of the study. Placebo patients who will experience at least a 25% hair regrowth after the month 5 evaluation will continue methotrexate or placebo from month 6 to the end of the study. Non‐responder patients in both arms A and B will be re‐randomized to receive from month 6 to the end of the study: methotrexate alone or with prednisone 0.3 mg/kg/day.
Outcomes	Primary outcome Complete or almost complete hair regrowth at the month 12 evaluation in patients treated with methotrexate or placebo who have not been re‐randomized after the month 5 evaluation. Secondary outcome Rate of Global Regrowth Assessment (GRA) at months 6 and 12. Delay of hair regrowth. Rate of relapse. Quality of life. Safety.
Notes	Awaiting. We contact Dr Pascal Joly and the response was " The study is completed and the manuscript sent to coauthors for approval".

NCT03535233.

Methods	Randomised, parallel ‐group trial Location: Egypt
Participants	40 participants. Inclusion criteria: Patients presenting with scalp alopecia areata, patchy type, of at least 2 months duration. Exclusion criteria: Alopecia totalis and alopecia universalis. Alopecia areata solely affecting the beard. Pregnant and lactating. Patients known to have autoimmune diseases e.g. autoimmune thyroid disease, vitiligo or SLE. Patients receiving systemic treatment relevant to alopecia areata within 3 months before enrollment into the study or topical treatment relevant to alopecia areata within 2 months before. Patients with a dermatological condition affecting the scalp other than AA: e.g. psoriasis, eczema. Patients with psychiatric illness or psychological state interfering with compliance or influencing the expectation of the patient.
Interventions	Intralesional triamcinolone acetonide 5 mg/mL monthly Minoxidil 5% topical solution applied twice daily and topical clobetasol propionate 0.05% cream applied once daily every night.
Outcomes	Primary outcome: Clinical assessment using SALT score. Secondary outcome: Dermoscopic evaluation. Quantitative measurement of Transforming growth factor‐beta 1 (TGF‐β1). Quantitative measurement of Interleukin‐23 (IL‐23).
Notes	Awaiting. We contacted Dr Mona El‐Kalioby and the response was: "This trial was completed but it is not yet published".

NCT05251831.

Methods	Randomised, double‐blinded, parallel‐group trial Location: Egypt
Participants	40 participants Inclusion criteria: Patients diagnosed clinically and dermatoscopically as alopecia areata of the scalp (uni‐ or 50 multilocularis). Age ranged between 18‐50 years in both sex. Exclusion criteria: Previously treated patients. Pregnant women. Patients with hematological disorders.
Interventions	Intradermal injections of PRP activated with calcium chloride 10% solution. Intradermal injections of PRP without activation.
Outcomes	Primary outcome Hair length in group A Secondary outcomes Hair length in Group B
Notes	Awaiting. We contacted Dr Ahmed Nouh and the response was "This paper is already published in Skin Appendage Disorders". The study was published on 12/10/2022, after our search date.

TCTR20210521008.

Methods	Randomised, double‐blinded, parallel‐group trial Location: Thailand
Participants	10 participants Inclusion criteria: Patients aged equal to or more than 18 years.‐Patients with chronic severe alopecia areata, alopecia totalis, or alopecia universalis. Exclusion criteria: Patients who had other active skin diseases at the scalp, who received systemic immunosuppressive or corticosteroid treatment within the past 6 months, and/or who had a history of resistance to DCP
Interventions	One side of the scalp was treated with combined DCP solution and 0.5% anthralin for 6 months. The other side of the scalp was treated with DCP alone for 6 months.
Outcomes	Primary outcome ‐ The efficacy of combination therapy of topical DCP and topical 0.5% anthralin versus topical DCP alone. Measurement of regrowth from baseline to 6 months Severity of Alopecia Tool (SALT) scores
Notes	Awaiting. We contacted Dr Supisara Wongdara and the response was "This research is finished and in the process of submitting the manuscript".

Characteristics of ongoing studies [ordered by study ID]

CTRI/2013/08/003880.

Study name	Comparing three different modalities in patchy hair loss
Methods	Randomised, parallel‐group trial Location: India
Participants	60 participants Inclusion criteria: Patient willing for the study. Patients with a circumscribed patch of hair loss without scarring. Patients who have not taken any treatment for alopecia areata during the last six months. Patients age more than 10 years. Exclusion criteria: Patients not willing to participate. Patients having scarry alopecia. Patients with secondary infection. Patients with pregnancy and lactation. Extensive lesions.
Interventions	Intralesional injection of triamcinolone acetonide 10mg/mL diluted 1:1 with distilled water using insulin syringe, once every three weeks for 12 weeks. Minoxidil 5% over the lesion twice a day for 12 weeks. Tacrolimus ointment 0.1% applied twice a day for 12 weeks.
Outcomes	Safety and efficacy of three modalities of treatment in Alopecia Areata Safety and efficacy of Intralesional steroid, Tacrolimus(0.1%) ointment and Minoxidil (5%) in the treatment of Alopecia areata
Starting date	October 12, 2012
Contact information	Mahajabeen Madarkar, Dr Nanda Kishore B drnandakishoreb@rediffmail.com, mahajabeenmadarkar@gmail.com Department of dermatology Father muller medical college Father muller Road Kankanady Mangalore‐575002 Dakshina Kannada KARNATAKA 575002 India
Notes	Ongoing. The status of the trial is not available. We have contacted the author on 21/09/2021, without a response.

CTRI/2018/02/012046.

Study name	A study to compare the efficacy of two different procedural treatments of patchy hair loss
Methods	Randomised, parallel‐group, placebo controlled trial Location: India
Participants	75 participants Inclusion criteria: Clinically diagnosed cases of alopecia areata in people more than 18 years old. Moderate to severe AA with a SALT score of more than 25. Not on treatment in the past 3 months for AA. Exclusion criteria: Patients with bleeding disorders. Patients who are on anticoagulant medication. Patients who had a keloidal tendency. Patients with thrombocytopenia. Patients suffering from HIV and other infectious diseases.
Interventions	Autologous platelet‐rich plasma therapy into the lesional area over the scalp by insulin syringe once in 4 weeks, for 16 weeks. Intralesional saline intradermally with an insulin syringe, 0.1‐mL injections per square cm, once in 4 weeks for 16 weeks. Triamcinolone acetonide intradermally with an insulin syringe, 0.1‐mL injections per square cm, once in 4 weeks for 16 weeks.
Outcomes	Primary outcome Improvement in SALT and burning itching score after 16 weeks (4th sitting) of treatment. Secondary outcome Maintenance of improvement in SALT and burning itching score six months after the 4th sitting of treatment.
Starting date	February 21, 2018
Contact information	kvtgopal77@gmail.com, drjk.marikanti@gmail.com MAHARAJAHS INSTITUTE OF MEDICAL SCIENCES, Department of Dermatology, Op room No.1 NELLIMARLA, VIZIANAGARAM(DT), ANDHRAPRADESH, INDIA Vizianagaram ANDHRA PRADESH 535217 India
Notes	Ongoing. The status of the trial is not available. We have contacted the author on 21/09/2021, without a response.

CTRI/2019/02/017483.

Study name	A study to compare effect of three common locally applied medicines in alopecia areata (a disease causing patchy hair loss) of scalp
Methods	Randomized, parallel‐group, multiple arm trial Location: India
Participants	120 participants Inclusion criteria: Alopecia Areata (AA) patients aged 12‐45 years. AA of scalp with a Severity of Alopecia Tool (SALT) score of S1. Patients having 03 or fewer patches. Consent to be part of the study, photograph & regular follow up. Exclusion criteria: History of side effects to bimatoprost, mometasone or tacrolimus. Patients were treated earlier within 3 months of the study. Patients with any evidence of atrophy in the lesions. Patients with alopecia totalis, alopecia universalis or ophiasis pattern of AA. Patients with a history of spontaneous hair re‐growth. Pregnant or lactating patients.
Interventions	0.03% Bimatoprost twice daily for 24 weeks. Mometasone 0.1% once daily for 24 weeks. 0.1% Tacrolimus once daily for 24 weeks.
Outcomes	Primary outcome Hair regrowth‐ 5 points semi‐quantitative score, regrowth scale (RGS)and Global assessment of 100%. Secondary outcomes Relapse. Atrophy.
Starting date	May 2, 2019
Contact information	Amit Bahuguna Skin Department Command Hospital(WC) Chandimandir 134107 Panchkula, HARYANA India afmcrocks@gmail.com Army Medical Corps
Notes	Ongoing. The status of the trial is not stated. We have contacted the author on 21/09/2021, without a response. On WHO clinical trials webpage: "Recruitment status: open to recruitment".

CTRI/2021/03/031963.

Study name	Microneedling with topical autologous Platelet rich plasma versus intralesional Triamcinolone acetonide in Alopecia Areata
Methods	Randomised, parallel‐group clinical trial Location: India
Participants	40 participants Inclusion criteria: Clinically diagnosed as alopecia areata of the scalp less than or equal to 5 patches and less than or equal to 25 % scalp involvement and who have not taken any treatment for hair loss in the past 3 months and are willing to consent for enrollment in the study. Exclusion criteria: Patients with history of bleeding disorders,on anticoagulant medications, with an active infection at local site, having keloidal tendency, having psoriasis or lichen planus because of risk of Koebner phenomenon or any other dermatological disease, having alopecia totalis or alopecia universalis, who had taken any treatment for hair‐loss in the past 3 months, with history of hypersensitivity to formulation used in the study, severe systemic illness or malignancy, pregnant and lactating mothers.
Interventions	Combination of micro needling with topical autologous platelet rich plasma every 4 weeks for 16 weeks. Intralesional triamcinolone acetonide 10 mg/mL every 4 weeks for 16 weeks.
Outcomes	Primary outcome Decrease in the mean SALT (Severity of Alopecia Areata Tool)scores from the baseline and in 4, 8, 12 and 16 weeks.
Starting date	March 15, 2021
Contact information	Dr Surabhi Dayal Address: Deptt. of Dermatology, Venereology and Leprology Pt B.D. Sharma, PGIMS, Rohtak 124001 Rohtak, HARYANA India Telephone: 9896205319 Email: surabhidayal7@gmail.com Affiliation: Pt B.D. Sharma, PGIMS, Rohtak Dr Mani Mehra Address: Skin OPD,Room no. 238 Deptt. of Dermatology, Venereology and Leprology Pt B.D. Sharma, PGIMS, Rohtak Deptt. of Dermatology, Venereolog 124001 Rohtak, HARYANA India Telephone: 9896205319 Email: surabhidayal7@gmail.com Affiliation: Pt B.D. Sharma, PGIMS, Rohtak
Notes	Ongoing. The status of the trial is not available. We have contacted the author on 18/10/2021, without a response.

CTRI/2022/03/040829.

Study name
Methods	Methods: Randomised, parallel‐group clinical trial Location: India
Participants	Participants: 60 participants Inclusion criteria: Alopecia areata patients ≥ 18 years Alopecia areata patients who have not received any other systemic or topical treatment for the disease in the past 1 month Exclusion criteria: Pregnant or lactating females Patients with history of or existing scalp skin diseases, infections Severe systemic illness (as uncontrolled DM or hypertension, liver or renal diseases) Immune‐compromised patients History of keloids or hypertrophic scars
Interventions	Fractional carbondioxide laser with topical corticosteroid Intralesional corticosteroid injection
Outcomes	Primary outcome: Hair regrowth Secondary outcome: Patient satisfaction with 7 point global‐assesment and 10 point VAS scales Adverse effects associated with treatment SALT score
Starting date	March 10, 2022
Contact information	Dr Raveena: drraveena96@gmail.com Dr Hemantakumarkar: hkkar_2000@yahoo.com Kalinga Institute of Medical Sciences Dept of Dermatology Patia Bhubhaneswar Khordha ORISSA 751024 India
Notes	Ongoing. We contacted Dr Raveena. Email response: "We have completed our study, but it is in analysis stage".

CTRI/2022/04/041728.

Study name	CTRI/2022/04/041728
Methods	Randomised, parallel‐group clinical trial Location: India
Participants	90 participants Inclusion criteria: Patients >18 years of age All patients with localized alopecia ( < 5 patches) on the scalp Exclusion criteria: Early age of onset of alopecia areata Nail changes Duration > 1 year Family history of Alopecia Patterns‐ophiasis, sisaipho, reticular Active infection at lesional site Associated autoimmune disease Pregnancy Received any prior treatment within a month Hormonal association Endocrine abnormality
Interventions	Calcipotriol 0.005% ointment Mometasone 0.1% cream Tacrolimus 0.1% ointment
Outcomes	Primary outcome: Assessment at 4, 8 and 16 weeks: Efficacy with pre and post pictures kept at baseline and follow up visits, SALT score and dermoscopy Secondary outcome: Assessment at 4, 8 and 16 weeks: Safety profile in long term use
Starting date	(First received 04 August 2022)
Contact information	Dr Farheen Begum: farheem.begum@gmail.com Dr Ajaya Jena: ajayajenabbsr@gmail.com Department of Dermatology Venerology and Leprosy IMS and SUM Hospital K8 Kalinga Nagar Bhubaneswar K8 Kalinga Nagar Bhubaneswar Khordha ORISSA 751003 India
Notes	The status of the trial is not available. We have contacted the author on 26/10/2022, without a response. Overall trial status: "Not yet recruiting"

IRCT20100314003566N10.

Study name	Comparative efficacy of different concentrations of triamcinolone acetonid in alopecia areata
Methods	Randomised, single‐blind, parallel‐group clinical trial Location: Iran
Participants	45 participants Inclusion criteria: Alopecia areata patients with one or more patch lesions. Patients 15 years and older. Patients who declare their full consent to participate in the study. The duration of the disease is 3 to 12 months and there is no sign of new hair growth at the site of the lesion. Exclusion criteria: Children Presence of skin diseases such as psoriasis, seborrheic dermatitis, and malignancy and infection at the site. Patients with malignancy and blood dyscrasia. Pregnant and lactating women. Patients with mental retardation and physical disability. Patients who have received topical or systemic corticosteroids and topical treatments in alopecia areata or immunosuppressive drugs over the past month. People with autoimmune, type I diabetes, etc. No seizure.
Interventions	Intralesional triamcinolone acetonide 2.5 mg/mL, 5 mg/mL and 10 mg/mL.
Outcomes	Primary outcome Clinical examination of number and length of new hair at lesion site Secondary outcome Clinical examination of new hair growth rate on injection site
Starting date	1 July 2020
Contact information	Hamideh Azimi Tabriz University of Medical Sciences Iran (the Islamic Republic of) Phone: +98 41 1540 6612 Email: addressazimih@tbzmed.ac.i
Notes	Ongoing. The status of the trial is not available. We have contacted the author on 22/11/2021, without a response.

IRCT20141209020250N5.

Study name	I RCT 20141209020250N5
Methods	Randomised, parallel‐group trial Location: Iran
Participants	Inclusion criteria: Patients more than 18 years old with AA. Hair loss grater than 50%.. Possibility to regular visit. Exclusion criteria: Pregnancy or lactation. Systemic immunosuppressive therapy during recent 3 months. Hematologic dycrasia. No response to diphencyprone test. Having androgenetic alopecia. Persistent totalis or universalis areata alopecia from infancy.
Interventions	Topical diphencyprone with new regimen. Topical diphencyprone with old regimen.
Outcomes	Primary outcome Terminal hair regrowth rate with topical diphencyprone with the old regimen in comparison to a new regimen based on SALT score. Secondary outcome Rate of change in SALT score than at base value, after intervention and at the end of course of treatment.
Starting date	November 22, 2019
Contact information	Doctor Alaa Al Bazzal Razi Dermatology Hospital, Razi Blind Alley, Vahdate‐Islami Square, Vahdate‐Islami Street, Tehran 1199663911 +98 21 5560 9951 alabazal9@gmail.com
Notes	Ongoing. The status of the trial is not available. We have contacted the author on 21/09/2021, without a response.

IRCT20200130046311N1.

Study name	Comparison of the effect of diphenciprone alone and diphenciprone with intralesional injection of platelet‐rich plasma in alopecia areata
Methods	Randomised, double‐blinded, parallel‐group clinical trial Location: Iran
Participants	20 participants Inclusion criteria: A patient with alopecia areata whose definitive diagnosis was made on clinical or histological basis and their involvement was severe (> 50% scalp area and less than 100%). Positive response to test with 2% diphenciprone solution. Exclusion criteria: Patients with alopecia totalis or universalis. Receive any systemic treatment during the last 6 months. Children under 18 years. Pregnant and lactating women.‐Immunosuppressed patient. Patient with platelet dysfunction.‐Patient with a history of injecting steroids on the site over the past month. Fever, systemic infection and local infection at the injection site.
Interventions	Combination of diphenciprone and platelet‐rich plasma injection on one side of the head. Topically diphencyprone and placebo injection on the other side.
Outcomes	Primary: Severity alopecia tool score (SALT score). Trichoscopic evaluation at baseline, 3 and 6 months later. Secondary: SALT score evaluation at baseline, 3 months and 6 months later.
Starting date	April 13, 2021
Contact information	Azadeh Rezayat Razi Dermatology Specialty Hospital, Razi Stand, Islamic Unity Square, Islamic Unity St., Tehran, Iran 1199663911 Tehran Iran (Islamic Republic of) Telephone: +98 2155630553 Email:azadehrezayat@gmail.com Affiliation: Tehran University of Medical Sciences
Notes	Ongoing. The status of the trial is not available. We have contacted the author on 18/10/2021, without a response.

ISRCTN14007390.

Study name	ISRCTN14007390
Methods	Randomised, parallel‐group clinical trial Location: UK
Participants	150 participants Inclusion criteria: Alopecia areata (<50% SALT score) Not currently receiving treatment for hair loss Exclusion criteria: Severe alopecia areata (totalis/universalis)
Interventions	0.25%, 0.5%, 1.0%, 2% topical STS‐01 once a day for 6 months Topical placebo once a day for 6 months
Outcomes	Primary outcome: Assessment at 0, 2, 4 and 6 months with a 2‐month follow up: Severity measured using the Severity of Alopecia Tool (SALT) Secondary outcome: Assessment at 0, 2, 4 and 6 months with a 2‐month follow up: Regrowth area measured by graphical measuring of the size of the patch AA quality of life measured using Subject Alopecia Areata Symptom Impact Scale (AASIS) and Alopecia Areata Quality of Life Index (AAQLI) Global assessment using the Clinical Global Impression (CGI) score Cytokines measured using assay of interleukins
Starting date	January 01, 2019
Contact information	Mr Tony Brown: tonybrown@manentia.co.uk Mr David Fleet: davidfleet@manentia.co.uk Laburnum House East Grimstead Salisbury SP5 3RT United Kingdom +44 (0)7533002238
Notes	The status of the trial is not available. We have contacted the author on 26/10/2022, without a response. Overall trial status: "Ongoing"

KCT0006802.

Study name
Methods	Randomised, open, parallel‐group trial Location: Korea
Participants	15 participants Inclusion Criteria: Alopecia areata volunteers with a lesion of 3 cm or more in the long diameter within 1 year on the scalp. A person who voluntarily decides to participate after hearing the detailed explanation of this clinical trial and agrees in writing to abide by the precautions. Men and women over the age of 20. Volunteers available for follow‐up during the trial period. Exclusion Criteria: Those who have not observed hair regrowth even after treatment (oral, topical, injection) for more than 6 months after the occurrence of bald spots. Those with clear hair regrowth in the alopecia lesion. Those who show ophiasis pattern. Those who have received drug treatment (oral, topical, injection) for alopecia areata within one month. Those with other diseases of the scalp or hair. Women who are pregnant or lactating. Those with cold intolerance or immune diseases other than alopecia areata. Those with psychiatric illness. Those judged by the principal investigator to be ineligible as a study subject.
Interventions	On 1st and 3rd quadrants, cryotherapy is performed for a total of 2 cycles of 10 seconds per cycle or 20 seconds per cycle, respectively, every 2 weeks and receive up to 6 sessions. On 2nd and 4th quadrants, triamcinolone intralesional injection is done on one side and any treatment is not done on the other side, every 2 weeks and receive up to 6 sessions.
Outcomes	Principal outcome Improvement rate of alopecia lesion Secondary outcome Lesion area Change rate of hair density and thickness
Starting date	February 2021
Contact information	Ohsang Kwon Seoul National University College of Medicine, 101, Daehak‐ro Jongno‐gu, Seoul 03080, Rep. of Korea +82‐2‐2072‐2418 oskwon@snu.ac.kr Seoul National University Hospital Ji Won Lee Seoul National University College of Medicine, 101, Daehak‐ro Jongno‐gu, Seoul 03080, Rep. of Korea +82‐2‐2072‐3708 fondant92@gmail.com Seoul National University Hospital
Notes	Ongoing. We have contacted the authors in 14/12/2022, without a response.

NCT03473600.

Study name	Cryotherapy versus steroids In alopecia areata: trichoscopic evaluation.
Methods	Randomised, parallel‐group trial Location: Egypt
Participants	40 participants Inclusion Criteria: Age >12 years. Both sexes will be included. Newly diagnosed cases. Exclusion Criteria: Children < 12 years. Pregnancy and lactation. Patients with active scalp infection. Patients with cold sensitivity (regarding the first group). Any newly onset medical systemic illness.
Interventions	Cryotherapy with liquid‐nitrogen spray, two cycles each one 3‐5 seconds, one session every two weeks for three months. Intralesional injection of triamcinolone‐acetonide once every three weeks for three months, using insulin syringes.
Outcomes	Principal outcome The incidence of patients with hair regrowth measuring the severity of alopecia tool score and dermoscopic examination for three months.
Starting date	November 2018
Contact information	Faculty of Medicine. Assiut, Egypt, 71111. Hanan Ahmed Morsy, assistant professor. Hanan_morsy2003@yahoo.com Aya Youssef Mohamed, Lecturer. Aya_badran@yahoo.com
Notes	Ongoing. We contacted Hanan Ahmed Morsy. Email response: "The researcher didn’t finish the study yet".

NCT04412148.

Study name	Modified SALT score for alopecia areata
Methods	Randomised, single‐group, double‐blinded clinical trial Location: Egypt
Participants	20 participants Inclusion criteria: Localized scalp AA ( 2 patches and ≤50% scalp involvement, duration more than 6 months) patients who did not receive any medication for at least 2 months before starting the study and those who agreed to join the study and signed written consent. Exclusion criteria: Pregnant, and lactating females Alopecia totalis or universalis or ophiasis or cicatritial alopecia Usage of systemic treatment of alopecia areata 2 months prior to the study. Any scalp lesion within the treated area Bleeding diathesis, severe anaemia or platelet disorders Medical conditions such as autoimmune diseases.
Interventions	Active arm: intralesional injection of triamcinolone acetonide every 2 weeks for 6 sessions. Placebo arm: intralesional injection of saline.
Outcomes	Primary outcomes Percentage involvement. Time frame: 6 months to 1 year. Modified Severity of Alopecia Tool (SALT) score. Time frame: 6 months to 1 year.
Starting date	January 1, 2020
Contact information	Rana Hilal, MD1001488869 ext +20 rana.hilal@kasralainy.edu.eg Kasr El Aini Hospital Egypt
Notes	Ongoing. We contacted Dr Hilal and the response was " We are still working on this clinical trial”.

NCT04660786.

Study name	Intralesional vitamin D in alopecia areata
Methods	40 participants Randomised, parallel‐group clinical trial Location: Egypt
Participants	Inclusion Criteria: Site: scalp. Extent: localized alopecia areata (less than 50% scalp involvement). Exclusion Criteria: Other skin diseases affecting the scalp. Pregnant or lactating women. Bleeding or coagulation disorders Immunocompromised patients. Known hypersensitivity to vitamin D3. Patients who received systemic or topical treatment for alopecia areata in the last month. Patients taking vitamin D supplements in the last 6 months. Patients treated with topical vitamin D analogues. Patients with diseases known to alter vitamin D level as autoimmune diseases, liver or renal diseases. Obesity (defined as body mass index ≥ 25).
Interventions	Intralesional triamcinolone acetonide 5mg/mL every 4 weeks for 3 sessions Intralesional vitamin D 2.5 mg/mL every 4 weeks for 3 sessions
Outcomes	Primary outcome Regrowth Score (RGS). Time frame: 3 months.
Starting date	November 1, 2020
Contact information	Marwa Eldeeb 01200029774 marwa.eldeeb16@alexmed.edu.eg
Notes	Ongoing. The status of the trial is not available. We have contacted the author on 18/11/2021, without a response.

NCT04793945.

Study name	Excimer Light and Topical Steroid in Treatment of Alopecia Areata
Methods	Randomised, parallel‐group clinical trial Location: Egypt
Participants	30 participants Inclusion criteria: Aged 16‐40 years. Good general health. A clinical diagnosis of multiple Alopecia Areata of the scalp. Hair loss <25 percent of the total scalp. No use of drugs or hair care products influencing hair growth in the last 2 months. Exclusion criteria: Other causes of hair loss (like endocrine or immunological diseases). Skin disease in the treatment area. Pregnant or lactating patients. Hair transplantation history. Photosensitivity history. Malignant tumor history.
Interventions	Topical steroid and 308nm‐Excimer light therapy twice weekly. Topical steroid and 308nm‐Excimer light therapy once weekly.
Outcomes	Primary outcomes Severity of Alopecia Tool score. Time frame: 9 months Digital photographs. Time frame: 9 months. The hair regrowth in the patches of alopecia. Time frame: 9 months. Dermoscopic evaluation. Time frame: 9 months. A patient satisfaction score. Time frame: 9 months. Adverse effects. Time frame: 9 months. Dermoscopic evaluation. Time frame: 9 months.
Starting date	April 1, 2021
Contact information	Eman R Mohamed Hofny, professor 01005298992 ext 002e_riad@yahoo.com Yasmeen M Tawfik Mhany PHD01006033331 ext 002 dr.yasminmostawfik@yahoo.com
Notes	Ongoing. The status of the trial is not available. We have contacted the author on 18/11/2021, without a response.

NCT05414266.

Study name	Study Evaluating Efficacy of Topical Squaric Acid Dibutyl Ester in Children and Adolescents With Alopecia Areata
Methods	Randomised, parallel‐group trial Location: USA
Participants	36 participants Inclusion criteria: Evidence of a personally signed and dated informed consent document indicating that the participant or a legally acceptable representative/parent(s)/legal guardian has been informed of all pertinent aspects of the study. Willing and able to comply with scheduled visits, treatment plan, laboratory tests, and other study procedures. Not pregnant or breast feeding Participants who are women of child bearing potential(WOCBP) must agree to use one highly effective method of contraception(with a failure rate of <1% per year during the intervention period and for at least 28 days after the last dose of study intervention. Negative highly sensitive pregnancy (urine) at the Day 1 visit before the first dose of study intervention for WOCBP Have a clinical diagnosis of AA with no other etiology of hair loss (e.g., telogen effluvium,androgenetic alopecia, etc.) greater than 10% hair loss of the scalp, including without evidence of terminal hair regrowth within 6 months at both screening and baseline visits Current episode of hair loss less than 5 years. must be on a stable regimen of permitted concomitant medication Must agree to avoid prolonged exposure to the sun and not to use tanning booths, sun lamps or other ultraviolet light sources in the treatment area during the study. Exclusion criteria: Other scalp disease that may impact AA assessment (e.g., scalp psoriasis, dermatitis, etc). Active systemic diseases that may cause hair loss (e.g., lupus erythematosus, thyroiditis, systemic sclerosis, lichen planus, etc). Any psychiatric condition including recent or active suicidal ideation or behavior Ongoing or recent history of any other uncontrolled and/or clinically significant medical or psychiatric disease or condition which, in the PI's medical opinion, should exclude participation in the study. Any present malignancies or history of malignancies with the exception of adequately treated or excised non metastatic basal cell or squamous cell cancer of the skin or cervical carcinoma in situ. History (single episode) of disseminated herpes zoster or disseminated herpes simplex, or a recurrent (more than one episode of) localized, dermatomal herpes zoster History of systemic infection requiring hospitalization, parenteral antimicrobial therapy, or as otherwise judged clinically significant by the investigator within 6 months prior to Day 1 Known primary or secondary immunodeficiency disorder or a first‐degree relative with a hereditary immunodeficiency Significant trauma or major surgery within 1 month of the first dose of study intervention. Considered in imminent need for surgery. Participants with elective surgery scheduled can only be enrolled in Study SADBE‐high dose (HD) with the approval of the investigator. Active acute or chronic infection requiring treatment with oral antibiotics, antivirals, anti‐parasitics, anti‐protozoals, or antifungals within 4 weeks prior to Day 1 or any active systemic or local infection not meeting other exclusion criteria within 1 week prior to Day 1. Participant has a known hypersensitivity or previous allergic reaction to any of the active or inactive ingredients in the study intervention Anticipated treatment with prohibited concomitant medication(s) during the course of the study Received Herbal medications with either unknown properties or pharmaceutical properties that impact AA within 1 week of first dose of study intervention Received topical steroids (e.g., steroid cream, steroid ointment) on areas under assessment (i.e., scalp, eyebrows). within 2 week of first dose of study intervention Received Ultraviolet B (UVB) phototherapy, Psoralen Ultraviolet A (PUVA) therapy, other phototherapy, contact immunotherapy [e.g.diphenylcyclopropenone (DPCP), and 1‐chloro‐2,4‐dinitrobenzene (DNCB)], topical irritants(e.g., anthralin), and liquid nitrogen cryotherapy within 4 weeks of first dose of study intervention Received Other topical or systemic treatments that could affect AA including: Immune suppressants (e.g., cyclosporine A, azathioprine, methotrexate (MTX), sulfasalazine, mycophenolate mofetil (MMF), everolimus, ibrutinib, Intralesional, topical, oral, or injectable (intramuscular or intraarticular)steroids or Oral or topical minoxidil within 8 weeks of first dose of study intervention or within 5 half‐lives (if known), whichever is longer Received Any Janus kinase (JAK) inhibitor for use in any disease indication or other immunomodulatory biologic agents within 12 weeks of first dose of study intervention or 5 half‐lives (if known), whichever is longer Investigator site staff members directly involved in the conduct of the study and their family members, site staff members otherwise supervised by the investigator, or participants who are University of Texas employees, including their family members, directly involved in the conduct of the study. Participants with shaved heads must not enter the study until hair has grown back to a reasonable level and is considered stable, in the opinion of the investigator. Have an active history of alcohol or substance abuse within 1 year prior to Day 1. Participant is unable to communicate or cooperate with the investigator due to language problems, poor mental development, or impaired cerebral function.
Interventions	Placebo solution applied every month for 4 months, then Squaric Acid Dibutyl Ester (SADBE) solution 5% applied every month for 8 months Squaric Acid Dibutyl Ester (SADBE) solution 5% every month for 12 months
Outcomes	Primary outcomes: Hair loss as assessed by the number of subjects that score <30 on the Severity of alopecia tool (SALT) Hair loss as assessed by the number of subjects that score <50 on the Severity of alopecia tool (SALT) Overall hair loss as assessed by the SALT score AA hair loss as assessed by the SALT score Change in overall hair loss as assessed by the Severity of alopecia tool (SALT) Change in AA hair loss as assessed by the Severity of alopecia tool (SALT) Overall hair loss as assessed by the number of subjects with at least 50% improvement in the SALT score as assessed by the Severity of alopecia tool (SALT) AA hair loss as assessed by the number of subjects with at least 50% improvement in the SALT score as assessed by the Severity of alopecia tool (SALT) Number of subjects that achieved a 1‐point increase in eyebrow assessment as assessed by the eyebrow assessment (EBA) score Number of subjects that achieved a 2‐point increase in eyebrow assessment as assessed by the eyebrow assessment (EBA) score Number of participants with hair loss as described as a 1‐point decrease in the Clinician Global Impression ‐ Alopecia Areata (CGI‐AA) Number of participants with hair loss as described as a 2‐point decrease in the Clinician Global Impression ‐ Alopecia Areata (CGI‐AA) Secondary outcomes: AA as assessed by the Patient's Global Impression of Change (PGI‐C) scale Number of subjects with moderately improved AA as assessed by the Patient's Global Impression of Change (PGI‐C) scale Number of subjects with greatly improved AA as assessed by the Patient's Global Impression of Change (PGI‐C) scale Quality of life as assessed by the Dermatology Life Quality Index( DLQI) Quality of life as assessed by the Children's Dermatology Life Quality Index (CDLQI) Quality of life as assessed by the Alopecia Areata Quality of Life Index (AA‐QLI)
Starting date	August 2022
Contact information	Adelaide A Hebert, MD 713‐500‐8276 Adelaide.A.Hebert@uth.tmc.edu Eugenio G Galindo, MD 713‐500‐8278 Eugenio.G.Galindo@uth.tmc.edu
Notes	Ongoing. We contacted Dr Adelaide Hebert. Email response: "Not yet ready to publish data. Enrollment just starting".

Differences between protocol and review

• A network meta‐analysis was planned, but it was not performed as we found too few trials comparing the same treatment alternatives to produce a connected network. There were several multi‐arm trials, and most direct comparisons were based on only one study. This makes it virtually impossible to evaluate inconsistency and model heterogeneity. For the above reasons, we were also unable to achieve the objective of establishing a ranking of treatments for alopecia areata.

• No time point was specified for the primary outcome of hair regrowth (> 75%) in the protocol version of this review. However, in the review, we set the timing of the main outcome as short‐term, defined as between 12 and 26 weeks.

• We considered other validated generic quality‐of‐life measures.

• We planned to use Covidence as a screening tool. However, Rayyan software was used instead.

• As per the suggestion of the Managing Editor (Central Editorial Service Editorial Team) and the Editorial Assistant, we have included a sensitivity analysis to see the impact of changing the assumption about the imputed results on participants lost to follow‐up.

• Initially, we planned to report in the summary of findings tables the primary outcomes only. However, for completeness of reporting, we also included the secondary outcomes.

• Due to the large amount of information collected and to simplify the presentation of data in the main text, we prioritised some of the interventions, grouping them into 12 main comparisons. The prioritisation was suggested by the clinical reviewers of the previous version, consensuated with the editorial team members at the Cochrane Skin Group and finally, their relevance to current clinical practice was confirmed by the field experts involved in the review (dermatologists). The interventions have been split between the main text of the review (named “prioritised” comparisons) or have been reported in the online supplement hosted at OSF platform.

Contributions of authors

MNC, GSV, SVG, MGUR, LG, DSC, JT, DOQ, JZ conceptualized the study, framed the review question and wrote the protocol. ACP, GSV, MGUR, DOQ, DSC, MMH, SFG, TBD screened potential studies resulting from the searches. ACP, GSV, SFG, MMH, MGUR, DOQ, DSC, MNC, LGM, TBD extracted data. GSV, ACP, MMH, MGUR, DOQ assessed risk of bias of all studies. AGG and GSV verified extracted data and performed data analysis. GSV, ACP, AGG, SFG, MMH wrote the drafts of the review. JZ critically reviewed the draft. All authors approved the final version of the review.

Disclaimer

This project was supported by the National Institute for Health Research, via Cochrane Infrastructure funding to Cochrane Skin. This systematic review was supported in peer review by the Complex Reviews Support Unit, also funded by the National Institute for Health Research (project number 14/178/29). The views and opinions expressed therein are those of the authors and do not necessarily reflect those of the Systematic Reviews Programme, NIHR, NHS or the Department of Health.

Sources of support

Internal sources

• Fundación Universitaria de Ciencias de la Salud‐FUCS, Colombia

  FUCS has supported the participation of GSV, MNC, and MGUR

• Hospital Universitario Ramón y Cajal (IRYCIS), Spain

  IRYCIS has supported the participation of JZ

• CIBER of epidemiology and public health (CIBERESP), Spain

  CIBERESP has supported the participation of JZ

External sources

• The National Institute for Health Research (NIHR), UK

  The NIHR, UK, is the largest single funder of Cochrane Skin.

• NIHR Complex Reviews Support Unit, UK

  This protocol was supported (peer review) by the Complex Reviews Support Unit, also funded by the National Institute for Health Research (project number 14/178/29)

Declarations of interest

Miriam Mateos‐Haro has no conflicts of interest.

Monica Novoa‐Candia has no conflicts of interest.

Guillermo Sánchez Vanegas has no conflicts of interest.

Andrea Correa‐Pérez has no conflicts of interest.

Andrea Gaetano‐Gil has no conflicts of interest.

Silvia Fernández‐García has no conflicts of interest.

Daniel Ortega‐Quijano has no conflicts of interest.

Mayra Gizeth Urueña Rodriguez has no conflicts of interest.

David Saceda‐Corralo has no conflicts of interest.

Tayeb Bennouna‐Dalero has no conflicts of interest.

Lucia Giraldo has no conflicts of interest.

Jaqueline Tomlinson has no conflicts of interest.

Sergio Vaño‐Galván declares receiving consulting fees from Eli Lilly and Company and Pfizer Canada INC.

Javier Zamora has no conflicts of interest.

These authors should be considered joint first author

New