Driver coexistence characteristics of ALK-fusion in Chinese patients with lung cancer.

Abstract

8530 Background: ALK fusions and other driver mutations are usually mutually exclusive. With the widespread application of genetic testing techniques, the coexistence of ALK fusions and other driver mutations could be detected, such as ALK fusion and EGFR driver mutations. However, there were no systematical studies about the coexistence of ALK fusions and other driver mutations. Here, we retrospectively investigated the coexistence of ALK fusions and other driver mutations. Methods: Samples with ALK fusions were extracted from a Chinese lung cancer cohort, which from OncoPanscan (Genetron Health) based sequencing of tissue. Driver mutations of EGFR, ROS1, RET, NTRK1/2/3, BRAF, MET, KRAS and ERBB2 could be detected. Results: In the cohort, 692 samples with ALK fusions contained the intact kinase domain. These samples could be classified into three forms: canonical fusions (only EML4 partner, n = 601), single non-canonical fusions (only non- EML4 partner, n = 51) and complex non-canonical fusions (two or more partners, n = 40). Among the 692 samples, only 20 samples (20/692, 2.89%) coexisted with other driver mutations, which indicated that the driver coexistence were rare. 70% (14/20) of driver coexistence were happened on canonical fusions ( EML4-ALK) samples. They were coexisting with EGFR L858R (3), EGFR L858R plus ROS1 fusion (1), EGFR L858R plus MET amplification (1), KRAS G12C/D/V (3), MET amplification (3), MET Exon14 skipping (1), ERBB2 amplification (1), and MET amplification plus ERBB2 amplification (1), respectively. 20% (4/20) of driver coexistence samples were single non-canonical ALK fusions coexisting with EGFR 19del (1) or 20ins (1) or L858R (1) or MET amplification (1). Another 10% (2/20) samples were complex non-canonical ALK fusions coexisting with EGFR G719S (1) or RET fusion (1). Most co-mutations have corresponding targeted inhibitors, maybe these patients can be treated by combined or sequential therapies. Among canonical fusions, single non-canonical fusions and complex non-canonical fusions of ALK, the frequency of the samples without coexistence of driver mutations was respectively 97.67% (587/601), 92.16% (47/51), 95.00% (38/40), without significant difference (P = 0.056). Maybe single/complex non-canonical fusions are also strong drivers as canonical fusions. Conclusions: In this cohort, very few of ALK fusion patients coexisted with other driver mutations. Among the co-existence samples, ALK fusion were mainly coexisting with the site mutations of EGFR and KRAS, amplifications of MET and ERBB2, fusions of ROS1 and RET. These samples maybe obtain more effective outcomes by combined or sequential therapies.