Abstract
Tumor mutational burden (TMB) has recently been identified as a biomarker of response to immune checkpoint inhibitors in many cancers, including melanoma. Co-assessment of TMB with inflammatory markers and genetic mutations may better predict disease outcomes. The goal of this study was to evaluate the potential for TMB and somatic mutations in combination to predict the recurrence of disease in advanced melanoma. A retrospective review of 85 patients with stage III or IV melanoma whose tumors were analyzed by next-generation sequencing was conducted. Fisher's exact test was used to assess differences in TMB category by somatic mutation status as well as recurrence locations. Kaplan-Meier estimates and Cox-proportional regression model were used for survival analyses. The most frequently detected mutations were TERT (32.9%), CDKN2A (28.2%), KMT2 (25.9%), BRAF V600E (24.7%), and NRAS (24.7%). Patients with TMB-L + BRAFWT status were more likely to have a recurrence [hazard ratio (HR), 3.43; confidence interval (CI), 1.29-9.15; P = 0.01] compared to TMB-H + BRAF WT. Patients with TMB-L + NRASmut were more likely to have a recurrence (HR, 5.29; 95% CI, 1.44-19.45; P = 0.01) compared to TMB-H + NRAS WT. TMB-L tumors were associated with local (P = 0.029) and in-transit (P = 0.004) recurrences. Analysis of TMB alone may be insufficient in understanding the relationship between melanoma's molecular profile and the body's immune system. Classification into BRAFmut, NRASmut, and tumor mutational load groups may aid in identifying patients who are more likely to have disease recurrence in advanced melanoma.