Presence of frequent underlying RAS mutations in cutaneous squamous cell carcinomas and keratoacanthomas (cuSCC/KA) that develop in patients during vemurafenib therapy.

Abstract

8520 Background: Vemurafenib (V; RG7204, PLX4032), an oral inhibitor of oncogenic V600E mutant BRAF, was evaluated for safety, efficacy, and pharmacokinetics in dose escalation and melanoma extension cohorts from the Phase I study (n=87; Flaherty, et al. NEJM, 2010). Development of cuSCC/KA was observed in melanoma patients (pts). Here we describe further molecular characterization of confirmed cases of cuSCC/KA. Methods: Pts receiving V therapy were monitored for toxicity including development of skin lesions. Suspicious lesions were excised and sent to central laboratories for confirmation of cuSCC/KA histology and for molecular analysis. Analyses included genetic analysis of H/K/N-RAS, CDKN2A and TP53, as well as P53 protein expression. Results: 35 cuSCC/KA lesions developed in 15 pts during therapy—34/35 (97%) were diagnosed as SCC, KA type or SCC, with features of KA and 1/35 (3%) were diagnosed as SCC, without features of KA. All lesions were histologically well differentiated and were managed with complete local excision. No lesions recurred and therapy was not interrupted. The majority of lesions developed within 8–21 wks after initiation of therapy. Genetic analyses were performed on 21 cuSCC/KA from 10 melanoma pts. 15/21 (71%) had mutations at codons 12, 13 or 61 of H-, K- or NRAS (with 14/15 having an HRAS mutation). No specimens had BRAF mutations in exon 15. A CDKN2A mutation was observed in 1/21 (5%) lesions (A73V) and 4/18 (22%) had exonic TP53 mutations. P53 nuclear expression varied, but was highest in the samples with P53 mutations (C176W and Y236C). Conclusions: CuSCC/KA lesions that developed during V therapy had a high incidence of RAS mutations in codons commonly mutated in sporadic cuSCC/KAs in the general population (HRAS mutation rate 12–46%, Nouri K., Skin Cancer, 2008). Given the short latency between exposure to V and the development of cuSCC/KA, these preliminary findings suggest that skin cells with pre-existing RAS mutations may develop into these lesions upon exposure to V. Further investigation and larger sample size is needed to determine if RAS alone or additional factors are required to induce cuSCC/KA.