Abstract

Background: Older patients with cutaneous melanoma (CM) are less likely to have positive sentinel lymph node biopsy but more likely to have poor outcomes.  Therefore, it is critical to identify which patients are truly at low or high risk of poor outcomes.  The 31-GEP stratifies patients into low (Class 1A), intermediate (Class 1B/2A), and high (Class 2B) risk of recurrence, metastasis, and death and can guide risk-aligned treatment/management plans.

Methods: SEER registries’ data (diagnosis=2013-2018) were linked to 31-GEP clinically tested patient data.  Analyses were performed using a de-identified dataset. Patients 65 years with stage I–III CM were included (n=6,003).  Kaplan-Meier analysis with log-rank test was used to analyze overall survival (OS) and melanoma-specific survival (MSS).  Cox multivariable regression models were used for risk prediction.

Results: Class 1A patients had significantly higher 5-year MSS/OS than those with 1B/2A or 2B results (MSS=96.1%, 91.6%, and 81.4%, p<0.001; OS=81.5%, 78.4%, and 59.7%, p<0.001).  In multivariable analysis, Class 2B was the strongest predictor of overall mortality (HR=1.86, p<0.001) and melanoma-specific mortality (HR=4.69, p<0.001); 1B/2A (HR=1.52, p=0.007), Breslow thickness (HR=1.10, p<0.001), and age (HR=1.08, p<0.001) were significant predictors of overall mortality. Class 1B/2A (HR=2.86, p=0.002), Breslow thickness (HR=1.15, p<0.001), and age (HR=1.05, p<0.001)were significant predictors of melanoma-specific mortality. Mitotic rate and ulceration were not significant.

Conclusions: The 31-GEP stratified OS/MSS in patients 65 years. The 31-GEP provides valuable information about patient prognosis, independent of risk factors used in staging, that can help guide personalized, risk-aligned clinical management decisions for older patients with CM.