Research Explores Novel Insights on Pediatric Hair Loss

Posters presented at the Society for Pediatric Dermatology Annual Meeting shed new light on pediatric hair loss. | Image Credit: Cienpies Design - stock.adobe.com

Posters presented at the Society for Pediatric Dermatology Annual Meeting, which took place on July 11-14, in Toronto, Ontario, Canada, focused on patients with pediatric androgenetic alopecia (AGA), the evaluation of generative artificial intelligence (AI) tools to calculate alopecia severity, and the safety analysis of ritlecitinib to treat pediatric alopecia areata (AA) to unveil promising advancements in understanding and treating pediatric hair loss.

AGA is the second most common form of hair loss among children and the most common subtype of hair loss in adults. Athough AA is still the most common type of alopecia among adolescents, AGA appears to be more prevalent than previously though.¹

The first poster was based on a retrospective cohort study across 5 academic institutions between January 2000 through June 2022.² Participants were considered eligible if they were 17 years or younger and diagnosed with AGA by a pediatric dermatologist.

A total of 203 patients were included in the study, with males comprising more than half of the population (58%). The majority of the population was classified as White (33%), followed by Asian (29.1%), and Hispanic/Latino (25.6%). Comorbidities included acne (28.1%), atopic dermatitis (7.9%), acanthosis nigricans (5.4%), polycystic ovary syndrome (4.9%), dyslipidemia (4.4%), and anxiety (4.4%).

Treatment medications ranged from topical minoxidil (76.5%), to oral minoxidil (2.9%), spironolactone (2.7%), topical corticosteroids (2.2%), 5-α reductase inhibitors (1%), other treatments (17.7%), and no treatment (8.4%). Patients reported their treatment responses as improving (24%), unchanged or worsened with treatments added to initial therapy (25%), and unchanged or worsened when treatment plans changed (12%).

Adverse events (AEs) were very mild or limited but were reported in 15 patients and included sexual dysfunction, burning/irritation, hypertrichosis, tremor, and abdominal discomfort.

The study was limited based on the retrospective review design but successfully found associations with obesity and family history of AGA. Overall, minoxidil in both topical and oral forms was the most common treatment used among patients.

Another poster focused on the evaluation of generative AI tools to calculate Severity of Alopecia Tool (SALT) scores for AA.³ The research explored the ability of generative AI tools like ChatGPT to calculate SALT scores based on de-identified 4-view images of patients’ scalps.

Methods were tested using ChatGPT and GPT-4o using prompts such as, “Tell me what you know about SALT scoring for alopecia areata." The 4-view images were then uploaded, and a tool was prompted to calculate SALT scores for each patient.

Comparisons between GPT-4 and GPT-4o showed high concordance, suggesting both versions produced scores that agreed closely with the standard rater. Based on the concordance correlation coefficient, SALT scores produced by GPT-4 agreed more closely with scores produced by clinicians, and based on intraclass correlation coefficient values, GPT-4o scores agreed more closely with clinicians’ scores.

No significant bias was observed between the raters because the Bland-Altman plots had a mean difference line close to 0 when clinician SALT scores were compared with GPT-4’s scores. Additionally, there were little to no points outside of the limits of agreement, suggesting that the raters agreed on most measurements of SALT scores. Due to the high concordance, preliminary data displayed a potential adjunct role that off-the-shelf generative AI tools like ChatGPT may play in calculating SALT scoring.

The final poster from the annual meeting showcased data from the ALLEGRO phase 2b/3 trial (NCT03732807) that tested the long-term safety of ritlecitinib for adolescents with AA.⁴ The oral JAK3/TEC family kinase inhibitor displayed efficacy and safety among patients 12 years and older with diagnosed AA.

Participants were placed into either a placebo-controlled pool (n = 105) for up to 24 weeks or an all-exposure pool (n = 181, 76 de novo) over 36 weeks. There were only 19 of the 105 participants in the placebo pool that did not receive ritlecitinib. Of those in the all-exposure pool, 172 received ritlecitinib 50 mg with or without a loading dose.

The all-exposure pool was split between adolescents who received any ritlecitinib dose in ALLEGRO-2b/3 and/or ALLEGRO-LT (10 mg, 30 mg, 50 mg, 200/30 mg, or 200/50 mg once daily) and the ritlecitinib 50-mg group of adolescents who received ritlecitinib 50 mg with or without an initial 4-week 200-mg daily loading dose in ALLEGRO 2b/3 and/or ALLEGRO-LT (NCT04006457).

More than half of the study population included participants who identified as White (89.5%) or women (63.2%) for both the placebo (67.4%) and all-exposure group (54.1%).

The median exposure in the ritlecitinib 50-mg group of the all-exposure pool was 790 days (59.2% ≥ 24 months exposure and (18% ≥ 36 months exposure).

Ritlecitinib in adolescents was administered for a median duration of 807 days for any ritlecitinib dose and 170 days for the placebo pool. Patient-years totaled 360.1 years for any ritlecitinib and 8.9 years for the placebo control.

Only a few serious AEs were identified in the any-exposure pool, including appendicitis in 2 adolescents; COVID-19 pneumonia, septic shock, delirium, and acute respiratory failure in 1 adolescent; and miscarriage, bipolar disorder and suicidal ideation, suicidal behavior, eczema, and latent tuberculosis in 1 adolescent each.

Serious infections that included appendicitis in 2 adolescents, COVID-19 and septic shock in 1 adolescent, and latent tuberculosis in 1 adolescent were found among the any-exposure group. There were no deaths, opportunistic infections, herpes zoster infections, malignancies, cardiovascula eventsr, or thrombotic events detected in the adolescent patients.

None of the AEs were related to growth disturbance, and changes in height/weight were considered age appropriate, instead of clinically significant.

The posters presented research that offers valuable insights into the management and treatment of pediatric hair loss. Although AGA appeared to be on the rise, the study on treatment responses suggested a need for further exploration of effective therapeutic options. The potential of AI in assessing alopecia severity is promising, and the long-term safety data for ritlecitinib in adolescent patients with AA is reassuring, although continued monitoring is essential, according to the study. These findings contribute to the growing body of knowledge in pediatric dermatology and may inform future clinical practices and research directions.

References:

1. Özcan D. Pediatric androgenetic alopecia: a retrospective review of clinical characteristics, hormonal assays and metabolic syndrome risk factors in 23 patients. An Bras Dermatol. 2022;97(2):166-172. doi:10.1016/j.abd.2021.06.006

2. Lee TS, O’Connor M, Castelo-Soccio L, et al. An international multicenter, retrospective cohort study of 203 patients with pediatric androgenetic alopecia. Presented at: Society for Pediatric Dermatology Annual Meeting; July 11-14, 2024; Toronto, Ontario, Canada. Poster 144.

3. Gupta R, Castelo-Soccio L, Cotsarelis G, et al. Evaluation of generative AI tools to calculate SALT scores for alopecia areata. Presented at: Society for Pediatric Dermatology Annual Meeting; July 11-14, 2024; Toronto, Ontario, Canada. Poster 78.

4. Craiglow B, Sun W, Sibbald C, et al. Updated integrated safety analysis of ritlecitinib in adolescents with alopecia areata (AA) from the allegro clinical trial program. Presented at: Society for Pediatric Dermatology Annual Meeting; July 11-14, 2024; Toronto, Ontario, Canada. Poster 33.