Track

Clinical Studies

Abstract

Cutaneous melanoma is classified into various subtypes, with varying prognoses. The integrated 31-gene expression profile (i31-GEP) combines the 31-GEP score with clinicopathological factors to predict recurrence risk. We analyzed i31-GEP risk stratification among various cutaneous melanoma tumor subtypes. Patients with stage I-III CM (n=1,743) were included. Subtypes included superficial spreading (n=826), nodular (n=201), acral (n=14), desmoplastic (n=24), or other (n=678). Kaplan-Meier analysis with log-rank test was used to analyze recurrence-free survival (RFS) for superficial spreading, nodular, and other. Due to lower numbers, only accuracy metrics were calculated for acral and desmoplastic. Risk thresholds were set at ≤10%, >10-20%, or >20% i31-GEP predicted risk for low, intermediate, or high risk of recurrence, respectively. The i31-GEP significantly stratified patients into low-, intermediate-, and high-risk groups for superficial spreading (5-year RFS: 97.3%, 83.1%, and 57.0%, p<0.001), nodular (5-year RFS: 95.0%, 80.3%, and 62.5%, p=0.001), and other subtypes (5-year RFS: 95.8%, 85.8%, and 54.9%, p<0.001). For acral, the i31-GEP sensitivity was 100% (4 recurrences in 4 high/intermediate); negative predictive value (NPV) was 100% (no recurrences in 9 low risk). For desmoplastic, sensitivity was 80% (4 recurrences in 5 high/intermediate); NPV was 83.3% (1 recurrence in 6 low risk). Within cutaneous melanoma subtypes, the i31-GEP test identified patients at significantly higher risk of recurrence (p<0.01), for whom increased surveillance or treatment may be warranted.  These findings demonstrate the i31-GEP test provides personalized risk prediction within patient groups defined by a given pathological factor, refining risk prediction to help guide risk-aligned surveillance and treatment management decisions.