About 5-10% of EGFR-mutant lung ACs transform to SCLC at the time of acquired resistance. The clinical course of patients (pts) with this finding is poorly characterized.
We retrospectively reviewed the records of all 16 pts with EGFR-mutant SCLC that have been seen at our center under an IRB-approved protocol and summarized demographics, disease features, and clinical outcomes.
Among 16 pts, 10 were women. 15 had AC histology at diagnosis; one had de novo SCLC with EGFR del19. 11 were never-smokers. All but the de novo case received an EGFR tyrosine kinase inhibitor (TKI) prior to transformation (7 had > 1 prior TKI; 6 received a 3rd-gen TKI) and 14/15 were on a TKI when the SCLC was noted. Median time from diagnosis to SCLC was 29.6 mo (95% CI 10.8-38.1). 15/16 of the SCLC tumors were genotyped; all kept their founder EGFR mutation and none had T790M, including 5 pts with prior T790M-positivity. Not all samples were assessed for genotype by the same assay, though recurrent mutations observed in at least 25% of cases were TP53, PIK3CA and RB1 (full genetic data will be shown at meeting). The most common therapy given directly after SCLC diagnosis was platinum-etoposide (n = 9); all SCLC treatment lines considered, platinum-etoposide had a clinical response rate of 72% (8/11) and progression-free survival of 4.6 mo (95% CI 2.0-5.5). Seven pts also had a taxane at some point after the diagnosis of SCLC, and 4/7 (57%) responded. Median overall survival (mOS) from initial cancer diagnosis was 38.2 mo (95% CI 24.5-43.9) and mOS from time of SCLC diagnosis was 12.4 mo (95% CI 4.0-16.6).
EGFR-mutant AC that transforms to SCLC uniformly maintains its founder EGFR mutation and is mutually exclusive with T790M (though both can be observed sequentially). In our cohort, the median time from initial lung cancer diagnosis to transformation was 2.5 years. The mOS of 12.4 mo following diagnosis of SCLC is similar to that seen among pts with non-EGFR-mutant SCLC. Likewise, responses to platinum-etoposide were frequent, but transient. 4 pts also responded to a taxane. Interestingly, mOS from diagnosis was similar to expected mOS for pts that never transform to SCLC at 38.2 mo. Further investigation is needed to better elucidate optimal strategies for this group.
Clinical trial identification
Legal entity responsible for the study
Lecia V Sequist
Z. Piotrowska: Advisory board/consulting honoraria from AstraZeneca, Boehringer-Ingelheim and Ariad Pharmaceuticals. A.F. Farago: Consulting for Pharmamar, Abbvie, Takeda, Merrimack, Intervention Insights. Honorarium from Foundation Medicine. A.N. Hata: Amgen - consulting and research support Novartis - research support L. Sequist: Consulting for AZ, Ariad, BMS and Genentech – and research support from Clovis, AZ, BI, Novartis, Merrimack, Pfizer, Genentech, Merck, Johnson & Johnson. All other authors have declared no conflicts of interest.