Comparison of Germline versus Somatic BAP1 Mutations for Risk of Metastasis in Uveal Melanoma

Abstract

Background: Germline mutations in BAP1 have been associated with BAP1-Tumor Predisposition Syndrome (BAP1-TPDS), a predisposition to multiple tumors within a family that includes uveal melanoma (UM), cutaneous melanoma, malignant mesothelioma and renal cell carcinoma. Alternatively, somatic mutations in BAP1 in UM have been associated with high risk for metastasis. In this study, we compare the risk of metastasis in UM that carry germline versus somatic BAP1 mutations and mutation-negative tumors. Methods: DNA extracted from 142 UM and matched blood samples was sequenced using Sanger or next generation sequencing to identify BAP1 gene mutations. Results: Eleven of 142 UM (8%) carried germline BAP1 mutations, 43 (30%) had somatic mutations, and 88 (62%) were mutation-negative. All BAP1 mutations identified in blood samples were also present in the matched UM. There were 52 unique mutations in 54 tumors. All were pathogenic or likely pathogenic. A comparison of tumors carrying somatic vs. germline mutations, or no mutations, showed a higher frequency of metastasis in tumors carrying somatic mutations: 74% vs. 36%, P=0.03 and 74% vs. 26% P<0.001, respectively. Tumors with a somatic mutation compared to mutation-negative had an older age of diagnosis of (61.8 vs. 52.2 years, P=0.002), and shorter time to metastasis (16 vs. 26 months, P=0.04). Kaplan-Meier analysis further showed that tumors with somatic (vs. germline) mutations demonstrated a greater metastatic risk (P=0.03). Cox multivariate analysis showed in addition to chromosome-3 monosomy and larger tumor diameter, the presence of BAP1 somatic, but not germline mutations, was significantly associated with risk of metastasis(P=0.02). Personal or family history of BAP1-TPDS was available for 79 of the cases. All eight cases with germline mutations reported a history of BAP1-TPDS, which was significantly greater than what was observed in cases with somatic mutations (10 of 23, P=0.009) or mutation-negative cases (11 of 48, P<0.001). Conclusions: Defining germline vs. somatic nature of BAP1 mutations in UM can inform the individual about both the risk of metastasis, and the time to metastasis, which are critically important outcomes for the individual. This information can also change the cascade screening and surveillance of family members.