Abstract
ABSTRACT
Melanoma in skin of color (SOC), defined as Fitzpatrick skin tones IV‐VI, presents unique epidemiologic, diagnostic, and therapeutic challenges compared to non‐Hispanic White (NHW). Although SOC patients have a lower incidence of melanoma, they experience higher melanoma‐specific fatality rates, which may be attributable to later‐stage diagnoses, differences in the distribution patterns of melanoma subtypes, and systemic barriers to care. SOC patients, particularly non‐Hispanic Black individuals, have a higher proportion of acral lentiginous and mucosal melanomas, which arise in non‐sun‐exposed areas, carry lower mutational burdens, and respond less favorably to immunotherapy and BRAF/MEK‐targeted therapies. These anatomic and biological differences contribute to delayed recognition and reduced treatment efficacy. Concurrently, the rising incidence of melanoma in situ across racial and ethnic groups without corresponding decreases in invasive disease or similarly increased mortality rates raises concern for overdiagnosis. Artificial intelligence‐based diagnostic tools show promise in NHW populations but are limited by the underrepresentation of SOC in training datasets and diminished positive predictive value in low‐prevalence populations. The estimated number needed to screen to prevent one melanoma death in SOC populations ranges from approximately 46,000 to 201,000, underscoring the challenge of melanoma screening in this group. Suggested solutions and improvements include expanding healthcare access through patient navigators, developing culturally tailored education programs focused on identifying early invasive disease, diversifying AI training datasets to include acral and mucosal melanoma, biobanking rare melanoma subtypes, and decentralizing clinical trials into community settings. A multifaceted approach is essential to addressing overdiagnosis and improving melanoma outcomes in the SOC populations.
image