Clinical Predictors of Early Cutaneous Melanoma Recurrence

Early-stage (IA, IB, IIA) cutaneous melanoma (CM) is typically managed with wide local excision (WLE). Despite this, some patients experience relapse, suggesting the presence of preoperative micrometastases. Identifying patients at high risk for relapse could inform the use of adjuvant therapies. To date, no studies have compared coding mutations among patients who have stage IA-IIA CM with and without relapse.

Researchers assess clinical and pathologic factors to support the use of whole-exome sequencing (WES) in identifying mutations associated with relapse. The results will be presented at American Society of Clinical Oncology 2025 Annual Meeting from May 30 to June 3, 2025.

The researchers conducted a retrospective study of 180 patients with stage IA-IIA CM diagnosed between 2003 and 2022. Data were obtained from the University of Pittsburgh Medical Center Hillman Melanoma Program biospecimen registry (1,511 of 7,750 total patients). Of 1,511 patients, 414 experienced relapse (case patients), 808 did not, and the remainder had incomplete data. Inclusion criteria required for were available follow-up data, pathology reports, primary tumor specimens WLE with negative margins, and a negative finding on sentinel lymph node biopsy.

Control patients were defined as those with no relapse (NR) for at least 3 years after WLE and documented follow-up within 18 months. Case and control patients were matched 1:1 based on age (50 years and older), stage, and sex. Variables analyzed included Breslow thickness, Clark level, ulceration, mitotic rate (MR), lymphovascular invasion, tumor-infiltrating lymphocytes, regression, age, sex, stage, and histologic subtype. The researchers evaluated overall survival (OS), relapse-free survival (RFS), relapse rates, and timing and site(s) of relapse.

Among both the relapse and NR groups, stage distribution was 19% for IA, 38% for IB, and 43% for IIA. Women comprised 43% of the NR group and 41% of the relapse group. Median age was 61 years for patients in the relapse group and 59 years for those in the NR group. Superficial spreading melanoma was detected in 43% of the relapse group and 46% of the NR group, and nodular melanoma was more frequent in the relapse group (26%) than in the NR group (20%).

OS was significantly lower in the relapse group (49%) compared with the NR group (99%) (hazard ratio [HR], 51.2; P<0.001). Timing of relapses was as follows: 18.9% occurred within 1 year, 44.4% between 1 and 3 years, 35.6% between 3 and 10 years, and one case of relapse occurred after 10 years. Among relapses, 34% were local, 21% were nodal, and 44% were distant (including 14% with brain metastases).

Older age was associated with a reduced risk for distant relapse (odds ratio [OR], 0.96; P=0.04) and brain metastases (OR, 0.95; P=0.03). Relapse was more common in patients with MR of 1/mm² or higher and residual tumor in the WLE specimen. An MR of 1/mm² or higher was associated with worse RFS overall (HR, 1.99; P=0.03), within the first year (HR, 1.99; P=0.03), and at 1 to 3 years (HR, 2.25; P=0.02). Relapse-free survival was also worse with residual tumor in the WLE specimen (HR, 1.83; P=0.005) and lymphovascular invasion (HR, 3.22; P=0.02).

Worse OS was seen with stage 1B compared with IA (HR, 3.33; P=0.01), whereas stage IIA trended worse than IA (HR, 2.02; P=0.2). Of note, brisk tumor-infiltrating lymphocytes were associated with poorer OS (HR, 2.95; P=0.003). Breslow thickness predicted worse OS in stage IIA (HR, 1.86; P=0.04) and across all stages when adjusted for increasing MR (HR, 1.34; P=0.04) or MR of 2/mm² or higher (HR, 1.35; P=0.03). However, MR of 1/mm² or higher alone did not significantly affect OS.