KEAP1 mutated renal cell carcinoma (RCC): Characterization of an emerging molecularly defined RCC subtype.

Abstract

4547 Background: KEAP1 is a tumor suppressor and negative regulator of the NRF2 pathway, and inactivating KEAP1 mutations (mts) have been reported in patients (pts) with RCC with similar morphology to fumarate hydratase (FH)-deficient RCC (FH-RCC). In FH-RCCs, the NRF2 pathway is activated through fumarate-led inactivation of KEAP1, and we hypothesized that KEAP1 mts are drivers in RCC, similar to FH mts in FH-RCC. We sought to characterize RCC with KEAP1 mts as a separate RCC subtype and compare to FH-RCC and clear cell (cc)RCC. Methods: Among consecutive pts with RCC consented to tumor-normal DNA sequencing via MSK-IMPACT (NCT01775072), we identified patients with germline or somatic mutations in KEAP1 or FH and no other known driver mts (ie VHL, MET, TFE3 alterations), and categorized these as “KEAP1-RCC” or “FH-RCC.” Clinicopathologic characteristics and outcomes were analyzed and compared to pts with FH-RCC and a previously annotated subset ccRCC (n=162). Immunohistochemical (IHC) staining for NQO1, marker of NRF2 activation, was performed. Time on systemic treatment and overall survival (OS) from time of sequencing were assessed. Results: Among 928 pts with RCC, 13 (1.4%) and 26 (2.8%) had RCCs with KEAP1 and FH mts, respectively. KEAP1 and FH mts were mutually exclusive. Median age was younger in FH-RCC (47 vs 63) (Table). When compared to ccRCC, OS was significantly worse for FH-RCC (HR 2.4, 95% CI 1.4-4.1; p=0.02) but not for KEAP1-RCC (HR 1.07, 95% CI 0.29-3.0; p=0.89). All KEAP1-RCC and FH-RCC were histologically classified as non-cc except one KEAP1-RCC that had 3p loss and no VHL mt. All available KEAP1 and FH-RCC were NQO1+ on IHC; control ccRCC were all negative. In the KEAP1-RCC cohort, we identified a female with an unclassified RCC and a germline KEAP1 truncating variant; RCC tumor had a second KEAP1 somatic mutation and was NQO1+ on IHC. The germline variant cosegregated to a sister with lung cancer (IHC NQO1+) and anal cancer. Conclusions: RCC with KEAP1 mts and no other genomic drivers are primarily non-cc with papillary features, have functional evidence of NRF2 activation, and although high-grade may have better outcomes than FH-RCC. KEAP1-RCC appears to be an emerging molecularly defined RCC subtype with clinical behavior similar to FH-RCC, likely as a result of converging on NRF2 pathway activation. Clinical characteristics. KEAP1-RCC (n=13) FH-RCC (n=26) ccRCC (n=162) Age (range) 63 (26-71) 47 (20-74) 56 (24-78) Male 8 (62%) 17 (65%) 125 (77%) Tumor size (cm), median (IQR) 5.6 (5.1, 11.0) 8.0 (5.0, 14.0) 8.2 (6.0, 10.5) Histology FH-deficient 0 13 (50%) 0 Papillary features 8 (62%) 4 (15%) 0 Unclassified 1 (8%) 8 (31%) 0 ccRCC 1 (8%) 0 162 (100%) Other/Unknown 3 (23%) 1 (4%) 0 Tumor grade, high 13 (100%) 26 (100%) 134 (83%) Metastatic 9 (69%) 25 (96%) 150 (93%)