Abstract
Introduction
Skin cancers are the most common malignancy worldwide, and identifying populations with altered risk is essential for informing prevention and surveillance strategies. Emerging evidence suggests that alopecia areata (AA) may be associated with reduced skin cancer risk, potentially reflecting enhanced cytotoxic immune activity.
Methods
We conducted a systematic review and meta-analysis to evaluate the incidence of skin cancers in patients with AA. PubMed, Embase, Scopus, and ClinicalTrials.gov were searched through July 2025. Of 1, 039 records identified, eight studies met inclusion criteria, with six included in the quantitative synthesis.
Results
AA was associated with a statistically significant reduction in melanoma incidence (OR 0.58; 95% CI, 0.36-0.94; p = 0.028). Overall skin cancer risk was reduced but not statistically significant (OR 0.58, 95% CI, 0.27-1.22). Pooled estimates for basal cell carcinoma (OR 0.43; 95% CI, 0.11-1.75) and squamous cell carcinoma (OR 0.66; 95% CI, 0.28-1.57) suggested directionally reduced associations, but did not reach statistical significance. Between-study heterogeneity was high (I2 >80% for BCC and SCC), however sensitivity analyses, including leave-one-out tests, confirmed the stability of the protective trend.
Discussion
These findings suggest that AA is not associated with increased skin cancer risk and demonstrates an inverse association with melanoma incidence, although substantial heterogeneity and sensitivity analyses warrant cautious interpretation. These results provide important baseline context for patient counseling and for interpreting longterm safety data as systemic therapies, including JAK inhibitors, are increasingly used.
Systematic Review Registration
https://www.crd.york.ac.uk/prospero/, identifier CRD420251123793.
Keywords: alopecia areata (AA), basal cell carcinoma (BCC), melanoma, nonmelanoma skin cancer (NMSC), skin cancer, squamous cell carcinoma (SCC)
1. Introduction
Skin cancers, including melanoma, basal cell carcinoma (BCC), and squamous cell carcinoma (SCC), are amongst the most common malignancies worldwide and represent a substantial burden for patients and health systems (1–4). Identifying populations with either heightened or reduced risk has important implications for prevention strategies, screening recommendations, and mechanistic insights into tumor biology. Importantly, the scalp is a sun-exposed site, and patients with alopecia may be particularly vulnerable to UV-induced carcinogenesis. In scarring alopecias, chronic inflammation and permanent follicular destruction may create a microenvironment that facilitates keratinocyte carcinogenesis, contributing to an increased incidence of BCC and SCC on the scalp (5). By contrast, alopecia areata (AA) is a non-scarring autoimmune alopecia, and emerging evidence suggests that its immunologic profile may paradoxically be inversely associated with melanoma and other skin cancers. As systemic immunomodulatory therapies for AA expand, understanding disease-specific cancer risk has become increasingly important for patient counseling and long-term safety assessment.
AA is a chronic, autoimmune hair loss disorder with a lifetime prevalence of approximately 2% (6). It is characterized by perifollicular T cell infiltration and disruption of hair follicle immune privilege (7). Unlike many autoimmune diseases, which are typically associated with an increased cancer risk (8, 9), AA may be associated with a reduced incidence of malignancies. This paradox mirrors observations in vitiligo, another autoimmune skin disorder marked by cytotoxic T cell activity, in which multiple studies have demonstrated a lower incidence of melanoma and nonmelanoma skin cancers (NMSC) (10).
Despite these observations, the relationship between AA and skin cancer remains uncertain. Several large population-based studies report significantly decreased incidence of melanoma and NMSC in AA, while others show neutral or even increased risks. This inconsistency leaves clinicians uncertain about the true relationship between AA and skin cancer risks, complicating patient counseling and long-term surveillance.
The recent approval of Janus kinase (JAK) inhibitors for AA underscores the urgency of clarifying this association, as long-term immunosuppression could alter cancer risk profiles. These highly effective therapies modulate immune pathways that may influence cancer susceptibility, yet long-term safety data remain limited (11–13). Understanding the baseline risk of skin cancer in patients with AA is therefore essential for both clinical practice and long-term therapeutic safety.
To address this gap, we conducted a systematic review and meta-analysis of observational studies evaluating the incidence of skin cancers in AA, quantifying risk for melanoma, BCC, and SCC. We aim to clarify the direction and magnitude of risk and contextualize these findings within dermatology, oncology, and immunology.
2. Methods
2.1. Search strategy and study selection
A systematic review was conducted in accordance with PRISMA guidelines, with the protocol registered in PROSPERO (#CRD420251123793). PubMed, Embase, Scopus, and ClinicalTrials.gov databases from inception through July 2025, using the terms “alopecia areata” AND (“skin cancer” OR “basal cell carcinoma” OR “squamous cell carcinoma” OR “melanoma”). Eligible studies were observational, non-interventional human research reporting incidence of skin cancers in patients with AA compared to non-AA controls. Review articles, conference abstracts without quantitative data, animal studies, interventional trials, and non-English studies were excluded.
A total of 1, 039 records were identified through database searches, of which 244 were duplicates (Figure 1). After screening 795 titles and abstracts, 18 articles underwent full-text review. Ten studies were excluded, no extractable outcome data (n = 4), review articles (n = 4), reporting of non-skin cancers (n = 1), or non-specific alopecia (n = 1). Eight studies met inclusion criteria and were included in the systematic review.
Figure 1.
PRISMA flow diagram of study selection. Flow diagram depicting literature search and study selection process. Of 1, 039 records identified across Embase, PubMed, Scopus, and ClinicalTrials.gov, eight studies met inclusion criteria for the systematic review.
2.2. Data extraction and measures
Three reviewers (SIG, IK & PVW) independently screened studies, extracted data, and assessed quality. For each study, we recorded cohort sizes, numbers of cancer cases, and reported effect estimates (odds ratios [ORs], hazard ratios [HRs], or standardized incidence ratios [SIRs]) with corresponding 95% confidence intervals (CIs). Outcomes were classified as overall skin cancer or by subtype (melanoma, basal cell carcinoma (BCC), squamous cell carcinoma (SCC)). Discrepancies were resolved by consensus.
2.3. Risk of bias assessment
Risk of bias was assessed with the ROBINS-E tool, appropriate for exposure-outcome studies (Supplementary Table 1) (14). Two reviewers (SIG & PVW) independently rated seven domains, with consensus on overall judgements (low, moderate, serious, critical).
2.4. Statistical analysis
For the meta-analysis, odds ratios (ORs) were calculated directly from raw incidence data whenever possible. Six studies contributed to the pooled estimate for overall skin cancer risk (Miller et al., Mostaghimi et al., Conic et al., George et al., Delcoigne et al., and Lee et al.). For these studies, number of skin cancer cases and cohort sizes were extracted from the published text, tables, or Supplementary Materials, and study-specific ORs were derived from these data. Studies reporting hazard ratios (HR), incidence ratios (IR), or standardized incidence ratios (SIR) without sufficient raw incidence data were included in the systematic review but were not pooled in the meta-analysis.
Log-transformed ORs were pooled using random-effects models with restricted maximum likelihood (REML) and Hartung-Knapp adjustment to account for between-study variance. Pooled effect sizes were expressed as ORs with 95% CIs. Heterogeneity was assessed using Cochran’s Q-test and quantified with the I² statistic (15). Subgroup analyses examined melanoma, BCC, and SCC separately. Sensitivity analyses included leave-one-out re-estimation of pooled effects and influence diagnostics. Publication bias was evaluated with funnel plots and Egger’s regression test (16). All analyses were performed in R (version 4.5.1) using the metafor package (version 4.8) (17), and were conducted by AA.
3. Results
3.1. Systematic review
Eight retrospective studies were included, spanning the United States (18–21), Sweden (22), Denmark (23), Taiwan (24), and Korea (25) (Table 1). Sample sizes ranged from just over 500 patients with AA (18, 20) to more than 650, 000 (25), with a mean age of approximately 40 years and a nearly equal distribution of males and females (51.4% female). Five studies included patients with alopecia totalis or universalis (AT/AU) in addition to patchy AA (18, 20, 21, 23, 25). Two studies did not include matched controls (23, 24).
Table 1.
Characteristics of observational studies included in the systematic review.
| Author (year) | Country | Study Design | AA Patients (n) | Controls (n) | AA Subtype(s) | Mean/Median age (% female) |
Melanoma outcome (n) |
SCC outcome (n) |
BCC outcome (n) |
Overall skin cancer (%) |
||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| AA | Control | AA | Control | AA | Control | AA | Control | |||||||
| Miller (2015) | USA | Retrospective study | 584 | 172 | AA, AT, AU | 35.53 (68.5%) | 3 | 3 | 3 | 1 | 9 | 6 | 2.6% | 5.8% |
| Mostaghimi (2016) | USA | Retrospective study | 1, 414 | 4, 242 | AA | 46.5 (65%) |
17 | 78 | 18 | 87 | 33 | 194 | 4.8% | 8.5% |
| Conic (2018) | USA | Retrospective study | 563 | 563 | AA, AT, AU | 42.47 (74.6%) | 2 | 19 | 12 | 50 | 22 | 129 | 6.4% | 35.2% |
| George (2023) | USA | Retrospective study | 8, 784 | 26, 352 | AA, AT, AU | 45.6 (55.6%) | 16 | 47 | 52 | 119 | 87 | 238 | 1.8% | 1.5% |
| Delcoigne (2025) | Sweden | Retrospective study | 18, 541 | 244, 573 | AA | 35 median (63%) | 52 | 825 | 65 | 776 | N/A | 0.63% | 0.65% | |
| Sørensen (2025) | Denmark | Retrospective study | 2, 778 | 0 | AA, AT, AU | 39 median (63%) |
< 5 | N/A | < 5 | N/A | 28 | N/A | < 1.3% | N/A |
| Chen (2018) | Taiwan | Retrospective study | 162, 499 | 0 | AA | 32.39 (52%) |
8 | N/A | NMSC | 0.023% | N/A | |||
| 30 | N/A | |||||||||||||
| Lee (2018) | Korea | Retrospective study | 668, 604 | 3, 343, 020 | AA, AT, AU | 40.3 (49.1%) | N/A | N/A | N/A | 0.0063% | 0.0063% | |||
AA, alopecia areata; AT, alopecia totalis; AU, alopecia universalis; BCC, basal cell carcinoma; SCC, squamous cell carcinoma; NMSC, non-melanoma skin cancer. Sample sizes reflect AA patients and controls included in each study. For Sørensen (2025), melanoma and SCC counts <5 were masked per Danish registry policy; overall incidence was estimated as <1.3%. Chen (2018) reported NMSC as a combined category without separate SCC or BCC outcomes.
Several studies reported a lower incidence of skin cancer in AA patients compared to controls. In US cohorts, Conic et al. found significantly reduced incidence of melanoma (OR 0.10; 95% CI, 0.02-0.44), BCC (OR 0.14; 95% CI, 0.09-0.22), and SCC (OR 0.22; 95% CI, 0.12-0.42) (all p < 0.001) (20). Miller et al. also reported reduced odds of melanoma (OR 0.29; 95% CI, 0.06-1.45), BCC (OR 0.43; 95% CI, 0.15-1.23), SCC (OR 0.88; 95% CI, 0.09-8.44), and NMSC overall (OR 0.47; 95% CI, 0.19-1.15), though only SCC reached statistical significance (p = 0.012) (18). Similarly, Mostaghimi et al. observed a 35% reduction in melanoma risk (OR 0.65; 95% CI, 0.39-1.09) and a comparable reduction in NMSC (OR 0.63; 95% CI, 0.48-0.81) (19). In Sweden, Delcoigne et al. demonstrated lower hazards for invasive melanoma (HR 0.53; 95% CI, 0.34-0.83), melanoma in situ (HR 0.81; 95% CI, 0.56-1.16), SCC (HR 0.65; 95% CI, 0.43-0.98), and SCC in situ (HR 0.89; 95% CI, 0.64-1.23) (22). Notably, a nationwide Danish registry reported a markedly decreased incidence of SCC (IR 0.16; 95% CI, 0.04-0.42) and malignant melanoma (IR 0.25; 95% CI, 0.09-0.54), while BCC incidence was neutral but appeared lower in patients with AT/AU (IR 0.89; 95% CI, 0.24-2.29) (23). In Taiwan, Chen et al. reported a significantly lower standardized incidence ratio for NMSC (SIR 0.59; 95% CI, 0.38-0.80) (24).
Although the scalp is sun-exposed and theoretically vulnerable in alopecia, few studies reported cancer location in detail. Delcoigne et al. noted that, aside from a possibly lower proportion of facial/scalp melanomas, cancer distribution in AA was largely similar to controls (22). Mostaghimi et al. also found no significant difference in anatomic location (19). Conic et al. provided the most detailed analysis, showing that in AA patients, SCC and BCC were most common on the head/neck and upper extremities, whereas melanomas occurred predominantly on the trunk and extremities (20). Compared with controls, AA patients had a relatively greater proportion of skin cancers on the upper extremities and fewer melanomas localized to the scalp. Collectively, these findings suggest that AA is not associated with a disproportionate increase in scalp-specific cancers, despite loss of hair coverage.
Other findings were more neutral or contradictory. George et al. reported no significant differences in melanoma, SCC, and BCC incidence between AA patients and matched controls (21). In Korea, Lee et al. found a slightly increased risk of skin cancer in AA overall (HR 1.09; 95% CI, 0.79-1.52), but a decreased risk in AT/AU (HR 0.42; 95% CI, 0.10-1.68) (25). Chen et al. also reported an increased risk of melanoma (SIR 1.54; 95% CI, 0.47-2.61) in AA compared to the general population (24).
Risk of bias assessment with ROBINS-E revealed variation across study designs. Single-center cohorts (Miller, Conic) were judged at serious risk due to referral populations, unmeasured confounding, and outcome measurement concerns. Large US claims studies (Mostaghimi, George) were rated moderate risk, reflecting diagnostic coding limitations and incomplete confounder adjustment. National registries (Delcoigne, Sørensen, Lee, Chen) were also rated moderate risk overall: although case ascertainment and representativeness were strong, residual confounding from UV exposure, skin phototype, and treatment history could not be excluded. No study was rated at low overall risk. Overall, most studies were categorized as moderate risk (Supplementary Table 1).
3.2. Meta-analysis
3.2.1. Overall skin cancer
Six studies contributed to the pooled analysis of overall skin cancer. Patients with AA demonstrated lower odds compared with controls, though the result did not reach statistical significance (OR 0.58; 95% CI, 0.27-1.22; p = 0.12) (Figure 2). Between-study heterogeneity was very high (I2 = 95%), so we applied REML with Hartung–Knapp adjustment to generate more conservative CIs.
Figure 2.
Overall skin cancer in patients with alopecia areata compared with controls. Four-panel figure comprising: (1) Funnel plot of log(OR) versus standard error with pseudo-95% confidence limits to assess small-study effects (2) Forest plot showing study-specific odds ratios and the random-effects (REML, Hartung-Knapp) pooled estimate; squares are proportional to study weight, diamonds show the pooled OR with 95% CI; heterogeneity statistics (I2, τ2, and Q) are reported. (3) Leave-one-out analysis showing the re-estimated pooled OR after repeatedly excluding each study (vertical dashed line marks the all-studies pooled OR). (4) Egger’s regression assessing small-study effects, displaying the regression line, intercept, and two-sided p-value. OR < 1 indicates lower overall skin cancer risk in AA versus controls.
3.2.2. Basal cell carcinoma
Four studies reported BCC outcomes. The pooled analysis suggested a lower risk of BCC among AA patients, but the estimate was not statistically significant (OR 0.43; 95% CI, 0.11-1.75; p = 0.15) (Figure 3). Between-study heterogeneity was very high (I2 = 95%).
Figure 3.
Basal Cell Carcinoma (BCC) in patients with alopecia areata compared with controls. Four-panel figure comprising: (1) Funnel plot of log(OR) versus standard error with pseudo-95% confidence limits to assess small-study effects (2) Forest plot showing study-specific odds ratios and the random-effects (REML, Hartung-Knapp) pooled estimate; squares are proportional to study weight, diamonds show the pooled OR with 95% CI; heterogeneity statistics (I2, τ2, and Q) are reported. (3) Leave-one-out analysis showing the re-estimated pooled OR after repeatedly excluding each study (vertical dashed line marks the all-studies pooled OR). (4) Egger’s regression assessing small-study effects, displaying the regression line, intercept, and two-sided p-value. OR < 1 indicates lower BCC risk in AA versus controls.
3.2.3. Squamous cell carcinoma
Five studies reported SCC incidence. The pooled estimate showed a directionally consistent reduction in risk that did not reach statistical significance (OR 0.66; 95% CI, 0.28-1.57; p = 0.26) (Figure 4). Between-study heterogeneity was high (I2 = 84%).
Figure 4.
Squamous Cell Carcinoma (SCC) in Patients with Alopecia Areata Compared with Controls. Four-panel figure comprising: (1) Funnel plot of log(OR) versus standard error with pseudo-95% confidence limits to assess small-study effects (2) Forest plot showing study-specific odds ratios and the random-effects (REML, Hartung-Knapp) pooled estimate; squares are proportional to study weight, diamonds show the pooled OR with 95% CI; heterogeneity statistics (I2, τ2, and Q) are reported. (3) Leave-one-out analysis showing the re-estimated pooled OR after repeatedly excluding each study (vertical dashed line marks the all-studies pooled OR). (4) Egger’s regression assessing small-study effects, displaying the regression line, intercept, and two-sided p-value. Note: OR < 1 indicates lower SCC risk in AA versus controls..
3.2.4. Melanoma
Five studies contributed melanoma outcomes. In contrast to BCC and SCC, the pooled analysis showed a statistically significant inverse association with melanoma among AA patients (OR 0.58; 95% CI, 0.36-0.94; p = 0.028) (Figure 5). Between-study heterogeneity was moderate (I2 = 58%). Although statistically significant, this inverse association was sensitive to exclusion of large registry-based cohorts, with attenuation of significance in leave-one-out analyses.
Figure 5.
Melanoma in patients with alopecia areata compared with controls. Four-panel figure comprising: (1) Funnel plot of log(OR) versus standard error with pseudo-95% confidence limits to assess small-study effects (2) Forest plot showing study-specific odds ratios and the random-effects (REML, Hartung-Knapp) pooled estimate; squares are proportional to study weight, diamonds show the pooled OR with 95% CI; heterogeneity statistics (I2, τ2, and Q) are reported. (3) Leave-one-out analysis showing the re-estimated pooled OR after repeatedly excluding each study (vertical dashed line marks the all-studies pooled OR). (4) Egger’s regression assessing small-study effects, displaying the regression line, intercept, and two-sided p-value. OR < 1 indicates lower melanoma risk in AA versus controls.
3.2.5. Sensitivity analysis
Across all outcomes, heterogeneity was a prominent feature. Overall skin cancer, BCC, and SCC each demonstrated very high heterogeneity (I2 = 84-95%), with little reduction on leave-one-out testing, suggesting that inconsistency was not driven by any single study. In contrast, melanoma showed only moderate heterogeneity (I2 = 58%) and more stable estimates, although statistical significance attenuated when certain large cohorts were excluded. These findings indicate that while the direction of effect consistently suggested lower odds of skin cancer in AA, variability between studies warrants cautious interpretation of pooled estimates. No evidence of publication bias was detected on funnel plot inspection or Egger’s regression.
4. Discussion
This systematic review and meta-analysis synthesizes available observational evidence examining skin cancer risk in patients with AA and healthy controls. Across multiple cohorts, AA was associated with a statistically significant inverse association with melanoma incidence, with consistent directionally reduced associations for BCC and SCC despite non-significant pooled estimates. Substantial between-study heterogeneity was observed, particularly for keratinocyte carcinomas, reflecting variability in study design, populations, and outcome ascertainment. Given the observational nature of the included studies, these findings should be interpreted as associations rather than evidence of causality.
The immunologic mechanisms discussed below are intended to provide biologic plausibility for the observed epidemiologic patterns and should be regarded as hypothetical frameworks rather than causal inferences derived from this meta-analysis or from the underlying observational studies.
AA is characterized by dense infiltration of autoreactive CD8+ T cells with collapse of hair follicle immune privilege (7). These effector cells produce interferon (IFN)-γ, a cytokine central to tumor immunosurveillance through enhanced antigen presentation and cytotoxic lymphocyte activation (26). Immune pathways implicated in AA therefore overlap with mechanisms involved in antitumor immunity, providing a biologically plausible mechanistic context for the observed inverse association with melanoma (27, 28).
Interleukin (IL)-15 further contributes to this immune milieu by sustaining tissue-resident memory CD8+ T cells and natural killer (NK) cells, both of which play complementary roles in tumor immunosurveillance (29). Unlike IL-2, IL-15 preferentially maintains cytotoxic effector populations without promoting regulatory T cell expansion, favoring persistent immune activation (30, 31). Experimental oncology studies demonstrate that IL-15-driven immune circuits enhance cytotoxic lymphocyte recruitment and tumor control, offering a biologically coherent, though indirect, framework through which AA-associated immune activity may intersect with cancer susceptibility (32–36). NK cells may be particularly relevant in melanoma, given their capacity to target malignant cells that evade immune detection via MHC class I downregulation (37–42).
Collectively, these observations suggest that AA may serve as a human disease model characterized by sustained cytotoxic immune activation, in which IFN-γ- and IL-15-dependent pathways simultaneously drive autoimmune hair follicle destruction and may intersect with pathways relevant to antitumor immunity. While these mechanistic interpretations cannot establish causality from observational data, they provide a coherent biological context for the epidemiologic patterns observed in this meta-analysis and help explain why melanoma, in particular, demonstrates the most consistent protective signal.
From a clinical perspective, these findings provide important context for patient counseling. While chronic autoimmune diseases are often associated with elevated malignancy risk, AA does not appear to share this pattern for skin cancer. Importantly, the data do not support intensified skin cancer surveillance beyond standard population-based recommendations in patients with AA in the absence of additional risk factors. Individualized risk assessment should continue to consider ultraviolet exposure, skin phototype, family history, and prior immunosuppressive therapies.
These observations are particularly relevant in the era of Janus kinase (JAK) inhibitor therapy for moderate-to-severe AA. Regulatory concerns regarding malignancy risk with systemic JAK inhibition in other disease contexts necessitate careful interpretation of long-term safety data in AA (43, 44). Establishing a disease-specific baseline risk is essential for contextualizing future pharmacovigilance findings and distinguishing treatment-related effects from disease-intrinsic immune biology.
Several limitations merit consideration. All included studies were observational, precluding causal inference. The small number of eligible studies limited statistical power and restricted subgroup analyses. Adjustment for key confounders, including cumulative ultraviolet exposure, skin phototype, geographic variation, and treatment history, was inconsistent across studies. The included cohorts spanned diverse geographic regions with differing ultraviolet exposure levels, population skin phototypes, and healthcare systems, which likely contributed to the substantial heterogeneity observed. Several studies combined patchy AA with alopecia totalis or universalis, which may represent biologically distinct phenotypes with differing immune intensity and cancer risk. Moreover, outcome ascertainment relied largely on administrative coding or registry data, introducing potential misclassification. Despite these limitations, sensitivity analyses supported consistent directionality of effect estimates, and no strong evidence of publication bias was detected.
5. Conclusion
This systematic review and meta-analysis demonstrates that AA is not associated with an increased risk of skin cancer and rather establishes an inverse association with melanoma incidence. These findings support the concept that AA represents a distinct autoimmune phenotype characterized by heightened cytotoxic immune surveillance, with potential implications for skin cancer risk. By establishing a disease-specific baseline risk, this work provides essential context for patient counseling and for interpreting long-term safety data as systemic therapies, including JAK inhibitors, become increasingly integrated into AA management. Continued prospective and mechanistic studies will be critical to further refine risk stratification and guide evidence-based clinical decision-making.
Funding Statement
The author(s) declared that financial support was not received for this work and/or its publication.
Footnotes
Edited by: Piotr K. Krajewski, Wroclaw Medical University, Poland
Reviewed by: Beata Jastrząb-Miśkiewicz, Wrocław Medical University, Poland
Justyna Putek, Regional Specialist Hospital, Poland
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Author contributions
SG: Supervision, Conceptualization, Project administration, Validation, Methodology, Writing – review & editing, Writing – original draft, Data curation, Software, Formal analysis, Investigation, Visualization. AA: Data curation, Software, Methodology, Validation, Investigation, Formal analysis, Writing – original draft. IK: Visualization, Investigation, Validation, Formal analysis, Methodology, Writing – original draft, Data curation, Writing – review & editing. PW: Investigation, Writing – original draft, Visualization, Formal analysis, Data curation, Methodology. KN: Supervision, Writing – review & editing, Validation. AT: Supervision, Validation, Writing – review & editing. JJ: Investigation, Validation, Writing – review & editing, Conceptualization, Methodology, Supervision, Resources, Data curation, Project administration, Visualization, Writing – original draft.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fonc.2026.1787992/full#supplementary-material
Risk of Bias Assessment Using ROBINS-E Framework. Risk of bias ratings across seven domains for each included study, classified as low, moderate, serious, or critical. Overall risk of bias was determined according to ROBINS-E guidance: the overall judgment was anchored to the most severe domain-level rating. Clinic-based cohorts were generally rated at higher risk due to potential confounding and outcome measurement bias, whereas registry-based cohorts demonstrated lower overall bias.
References
- 1. Wang R, Chen Y, Shao X, Chen T, Zhong J, Ou Y, et al. Burden of skin cancer in older adults from 1990 to 2021 and modelled projection to 2050. JAMA Dermatol. (2025) 161:715–22. doi: 10.1001/jamadermatol.2025.1276. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Roky AH, Islam MM, Ahasan AMF, Mostaq MS, Mahmud MZ, Amin MN, et al. Overview of skin cancer types and prevalence rates across continents. Cancer Pathog Ther. (2025) 3:89–100. doi: 10.1016/j.cpt.2024.08.002. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Jones SMW. Financial worry in people with cancer: Relationship to employment and outcomes. Psychooncology. (2022) 31:1835–42. doi: 10.1002/pon.6034. PMID: [DOI] [PubMed] [Google Scholar]
- 4. Aggarwal P, Knabel P, Fleischer AB. United States burden of melanoma and non-melanoma skin cancer from 1990 to 2019. J Am Acad Dermatol. (2021) 85:388–95. doi: 10.1016/j.jaad.2021.03.109. PMID: [DOI] [PubMed] [Google Scholar]
- 5. Lee J, de Gregorio L, Tosti A. Scarring alopecia and risk of skin cancer: A literature review. Skin Appendage Disord. (2025) 11(5):424–31. doi: 10.1159/000545345. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Dainichi T, Kabashima K. Alopecia areata: What’s new in epidemiology, pathogenesis, diagnosis, and therapeutic options? J Dermatol Sci. (2017) 86:3–12. doi: 10.1016/j.jdermsci.2016.10.004. PMID: [DOI] [PubMed] [Google Scholar]
- 7. Bertolini M, McElwee K, Gilhar A, Bulfone-Paus S, Paus R. Hair follicle immune privilege and its collapse in alopecia areata. Exp Dermatol. (2020) 29:703–25. doi: 10.1111/exd.14155. PMID: [DOI] [PubMed] [Google Scholar]
- 8. He M, Lo C-H, Wang K, Polychronidis G, Wang L, Zhong R, et al. Immune-mediated diseases associated with cancer risks. JAMA Oncol. (2022) 8:209–19. doi: 10.1001/jamaoncol.2021.5680. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Ehrenfeld M. Autoimmune diseases and cancer. In: Autoimmunity: from bench to bedside. Bogota, Colombia: El Rosario University Press; (2013). Available online at: https://www.ncbi.nlm.nih.gov/books/NBK459441/. [PubMed] [Google Scholar]
- 10. Mohammed AA, Lengyel AS, Meznerics FA, Szondy I, Walter A, Nagy B, et al. Cancer risk in vitiligo: No evidence of increased prevalence-a systematic review and meta-analysis. Dermatol Ther (Heidelb). (2025) 15:3009–3024. doi: 10.1007/s13555-025-01520-0. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Dillon K-A. A comprehensive literature review of JAK inhibitors in treatment of alopecia areata. Clin Cosmet Investig Dermatol. (2021) 14:691–714. doi: 10.2147/CCID.S309215. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Sanchez K, Englander H, Salloum L, Gregoire S, Biba U, Ershadi S, et al. Evaluating current and emergent JAK inhibitors for alopecia areata: A narrative review. Dermatol Ther (Heidelb). (2025). doi: 10.1007/s13555-025-01517-9. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Egeberg A, Linsell L, Johansson E, Durand F, Yu G, Vañó-Galván S. Treatments for moderate-to-severe alopecia areata: A systematic narrative review. Dermatol Ther (Heidelb). (2023) 13:2951–91. doi: 10.1007/s13555-023-01044-5. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Higgins JPT, Morgan RL, Rooney AA, Taylor KW, Thayer KA, Silva RA, et al. A tool to assess risk of bias in non-randomized follow-up studies of exposure effects (ROBINS-E). Environ Int. (2024) 186:108602. doi: 10.1016/j.envint.2024.108602. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. (2003) 327:557–60. doi: 10.1136/bmj.327.7414.557. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ. (1997) 315:629–34. doi: 10.1136/bmj.315.7109.629. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Viechtbauer W. Conducting meta-analyses in R with the metafor package. J Stat Softw. (2010) 36:1–48. doi: 10.18637/jss.v036.i03 [DOI] [Google Scholar]
- 18. Miller R, Conic RZ, Bergfeld W, Mesinkovska NA. Prevalence of comorbid conditions and sun-induced skin cancers in patients with alopecia areata. J Investig Dermatol Symp Proc. (2015) 17:61–2. doi: 10.1038/jidsymp.2015.44. PMID: [DOI] [PubMed] [Google Scholar]
- 19. Mostaghimi A, Qureshi S, Joyce C, Guo Y, Huang KP. Reduced incidence of skin cancer in patients with alopecia areata: A retrospective cohort study. Cancer Epidemiol. (2016) 41:129–31. doi: 10.1016/j.canep.2016.02.009. PMID: [DOI] [PubMed] [Google Scholar]
- 20. Conic RRZ, Miller R, Piliang M, Bergfeld W, Atanaskova Mesinkovska N. Risk of sun-induced skin cancers in patients with alopecia areata, alopecia totalis and alopecia universalis. J Eur Acad Dermatol Venereol. (2018) 32:e409–11. doi: 10.1111/jdv.15002. PMID: [DOI] [PubMed] [Google Scholar]
- 21. George P, Jagun O, Liu Q, Wentworth C, Napatalung L, Wolk R, et al. Incidence rates of infections, Malignancies, thromboembolism, and cardiovascular events in an alopecia areata cohort from a US claims database. Dermatol Ther (Heidelb). (2023) 13:1733–46. doi: 10.1007/s13555-023-00937-9. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Delcoigne B, Lysell J, Askling J. Incidence, stage and outcome of Malignant melanoma, keratinocyte and other cancers in individuals with vitiligo or alopecia: Intraindividual or familial risks? Br J Dermatol. (2025) 193:66–73. doi: 10.1093/bjd/ljaf074. PMID: [DOI] [PubMed] [Google Scholar]
- 23. Sørensen SBT, George P, Jagun O, Wolk R, Napatalung L, Zwillich SH, et al. Rates of infections, Malignancies, cardiovascular outcomes, and death in individuals with hospital-treated alopecia areata: A registry-based cohort study in Denmark. Acta Derm Venereol. (2025) 105:adv42646. doi: 10.2340/actadv.v105.42646. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Chen C-C, Chang Y-T, Liu H-N, Chen Y-J. Cancer risk in patients with alopecia areata: A nationwide population-based matched cohort study. Cancer Med. (2018) 7:2153–9. doi: 10.1002/cam4.1448. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Lee JH, Song Y, Do Han K, Park YM, Lee JY, Park Y-G, et al. Cancer risk by the subtype of alopecia. Sci Rep. (2018) 8:9748. doi: 10.1038/s41598-018-28142-1. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Castro F, Cardoso AP, Gonçalves RM, Serre K, Oliveira MJ. Interferon-gamma at the crossroads of tumor immune surveillance or evasion. Front Immunol. (2018) 9:847. doi: 10.3389/fimmu.2018.00847. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Nielsen M, Presti M, Sztupinszki Z, Jensen AWP, Draghi A, Chamberlain CA, et al. Coexisting alterations of MHC class I antigen presentation and IFNγ signaling mediate acquired resistance of melanoma to post-PD-1 immunotherapy. Cancer Immunol Res. (2022) 10:1254–62. doi: 10.1158/2326-6066.CIR-22-0326. PMID: [DOI] [PubMed] [Google Scholar]
- 28. Buferne M, Chasson L, Grange M, Mas A, Arnoux F, Bertuzzi M, et al. IFNγ producing CD8+ T cells modified to resist major immune checkpoints induce regression of MHC class I-deficient melanomas. Oncoimmunology. (2015) 4:e974959. doi: 10.4161/2162402X.2014.974959. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Lee H, Park S-H, Shin E-C. IL-15 in T-cell responses and immunopathogenesis. Immune Netw. (2024) 24:e11. doi: 10.4110/in.2024.24.e11. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Arneja A, Johnson H, Gabrovsek L, Lauffenburger DA, White FM. Qualitatively different T cell phenotypic responses to IL-2 versus IL-15 are unified by identical dependences on receptor signal strength and duration. J Immunol. (2014) 192:123–35. doi: 10.4049/jimmunol.1302291. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Waldmann TA. The biology of interleukin-2 and interleukin-15: Implications for cancer therapy and vaccine design. Nat Rev Immunol. (2006) 6:595–601. doi: 10.1038/nri1901. PMID: [DOI] [PubMed] [Google Scholar]
- 32. Cai M, Huang X, Huang X, Ju D, Zhu YZ, Ye L. Research progress of interleukin-15 in cancer immunotherapy. Front Pharmacol. (2023) 14:1184703. doi: 10.3389/fphar.2023.1184703. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Bergamaschi C, Pandit H, Nagy BA, Stellas D, Jensen SM, Bear J, et al. Heterodimeric IL-15 delays tumor growth and promotes intratumoral CTL and dendritic cell accumulation by a cytokine network involving XCL1, IFN-γ, CXCL9 and CXCL10. J Immunother Cancer. (2020) 8:e000599. doi: 10.1136/jitc-2020-000599. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Heger L, Hatscher L, Liang C, Lehmann CHK, Amon L, Lühr JJ, et al. XCR1 expression distinguishes human conventional dendritic cell type 1 with full effector functions from their immediate precursors. Proc Natl Acad Sci. (2023) 120:e2300343120. doi: 10.1073/pnas.2300343120. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Bödder J, Zahan T, van Slooten R, Schreibelt G, de Vries IJM, Flórez-Grau G. Harnessing the cDC1-NK cross-talk in the tumor microenvironment to battle cancer (2021). [DOI] [PMC free article] [PubMed]
- 36. Robinson TO, Schluns KS. The potential and promise of IL-15 in immuno-oncogenic therapies. Immunol Lett. (2017) 190:159–68. doi: 10.1016/j.imlet.2017.08.010. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. DhatChinamoorthy K, Colbert JD, Rock KL. Cancer immune evasion through loss of MHC class I antigen presentation (2021). [DOI] [PMC free article] [PubMed]
- 38. Bessoles S, Grandclément C, Alari-Pahissa E, Gehrig J, Jeevan-Raj B, Held W. Adaptations of natural killer cells to self-MHC class I (2014). [DOI] [PMC free article] [PubMed]
- 39. de Souza Fernandez Pereira M, Carr DR, Gatti-Mays ME, Olsen MR, Setty BA, Shahwan KT, et al. Natural killer cell recognition and control of epithelial cancers. Cancer J. (2022) 28:263–9. doi: 10.1097/PPO.0000000000000610. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Lee JH, Shklovskaya E, Lim SY, Carlino MS, Menzies AM, Stewart A, et al. Transcriptional downregulation of MHC class I and melanoma de-differentiation in resistance to PD-1 inhibition. Nat Commun. (2020) 11:1897. doi: 10.1038/s41467-020-15726-7. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Cornel AM, Mimpen IL, Nierkens S. MHC class I downregulation in cancer: Underlying mechanisms and potential targets for cancer immunotherapy. Cancers (Basel). (2020) 12:1760. doi: 10.3390/cancers12071760. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Taylor BC, Balko JM. Mechanisms of MHC-I downregulation and role in immunotherapy response (2022). [DOI] [PMC free article] [PubMed]
- 43. Ytterberg SR, Bhatt DL, Mikuls TR, Koch GG, Fleischmann R, Rivas JL, et al. Cardiovascular and cancer risk with tofacitinib in rheumatoid arthritis. N Engl J Med. (2022) 386:316–26. doi: 10.1056/NEJMoa2109927. PMID: [DOI] [PubMed] [Google Scholar]
- 44. Ahmed S, Pfeiffer RM, Jackson SS, Song M, Mariette X, Seror R, et al. Cancer risk of Janus kinase inhibitors and biological disease-modifying antirheumatic drugs among older Americans with rheumatoid arthritis. Rheumatol (Oxford). (2025) 64:4172–80. doi: 10.1093/rheumatology/keaf092. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Risk of Bias Assessment Using ROBINS-E Framework. Risk of bias ratings across seven domains for each included study, classified as low, moderate, serious, or critical. Overall risk of bias was determined according to ROBINS-E guidance: the overall judgment was anchored to the most severe domain-level rating. Clinic-based cohorts were generally rated at higher risk due to potential confounding and outcome measurement bias, whereas registry-based cohorts demonstrated lower overall bias.
Data Availability Statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.