Correlation between re-biopsy site and detection of T790M mutation in NSCLC patients treated with EGFR TKI.

Abstract

e20620 Background: Re-biopsy is important to decide the treatment after EGFR-tyrosine kinase inhibitor (TKI) failure in non-small cell lung cancer (NSCLC) patients. We hypothesized that the T790M mutation in EGFR might show heterogeneity depending on the re-biopsy site. Methods: NSCLC patients who had received initial EGFR-TKI since January 2009 to December 2016, at any stage and recurrence after surgery and at any line of treatment, were included. Results: In total, 128 patients were included. Median age at EGFR-TKI therapy initiation was 73 (range, 38–97) years; 67% patients were female, all were Asian, 56% had never smoked, and 99% had adenocarcinoma. Of total 128 patients, 109 showed progressive disease. Median progression-free survival (PFS) was 10 (0.56–57) months. Median period since EGFR-TKI failure until the first re-biopsy was 197 (0–1322) days. Re-biopsy was performed 50 times in 42 patients; the number of T790M positive, negative, and pathologically negative patients was 20, 17, and 5, respectively, and the number of re-biopsies in these patients was 20, 22, and 8, respectively. Median PFS was longer in T790M positive patients than in negative patients significantly (17 [11–24] vs. 7.6 [4.3–11] months, P = 0.007). Characteristics such as gender, smoking status, proportion of stage IV, time between EGFR-TKI failure and first re-biopsy, and number of biopsies did not affect the T790M status in the biopsies. T790M positive group had more exon 19 deletions than negative group significantly (75% vs. 23%, P = 0.012). Biopsies at primary lesion, distant, and pleural effusion (PE) were 25% vs. 50%, 60% vs. 36%, and 15% vs. 14%, respectively, in the T790M positive vs. negative groups. Compared with the biopsy-site at diagnosis, the site was same as before in 35% vs. 50% cases (primary lesion [20% vs. 45%], distant [10% vs. 4.5%], and PE [5% vs. 0%]) and was new in 55% vs. 41% cases (distant lesions [45% vs. 27%] and PE [10% vs. 14%]) in the T790M positive vs. negative groups, respectively. Conclusions: In NSCLC patients treated with EGFR-TKI, re-biopsy was performed in distant lesions more frequently in the T790M positive cases than in negative cases. However, the T790M status was not correlated with the re-biopsy site.