Native and rearranged ALK copy number and rearranged ALK cell count in NSCLC: Implications for ALK inhibitor therapy.

Abstract

7534 Background: ALK rearranged NSCLC responds well to ALK inhibitors. Clinically, >15% cells showing rearrangements by break-apart FISH classifies tumors as ALK(+). Native ALK copy number gain has also been reported. We explored the significance of native and rearranged ALK copy number on crizotinib outcomes and whether >15% reflected a clear biological distinction in the frequency of ALK(+) cells. Methods: Copy number and genomic status of ALK assessed by FISH. A total of 1426 NSCLC clinical specimens, 174 ALK(+) and 1252 ALK(-) by standard criteria, and 26 NSCLC cell lines (2 ALK(+) and 24 ALK(-)) were investigated. Results: Native ALK gene mean copy number was significantly higher in ALK(-) than in ALK(+) cases (2.8, SD 0.93, range 1.2-11.4 vs. 1.8, SD 0.79; range 0.6 to 5.2; p<0.01). Frequency of native ALK copy number gain (≥3 copies/cell in ≥40% cells) was 19% in ALK(+) and 62% in ALK(-) tumors (p<0.001). In ALK(-) tumors, abundant focal amplification of native ALK was rare (0.8%); scanty amplification (<10% tumor cells) occurred in 1.1%, and duplication of the entire native ALK or of the ALK 3’ end occurred in 3.5%. Among ALK(-) cell lines, mean native ALK copy number ranged 2.1-6.9 and was not correlated with in vitro crizotinib sensitivity (IC50s 0.34-2.8 uM). In (+) patients, neither native or rearranged ALK copy number, nor percentage cell count correlated with maximal tumor shrinkage or PFS with crizotinib. Clinical specimens with 0-9%, 10-15%, 16-30%, 31-50% and >50% of ALK+ cells were found in 79.3%, 8.5%, 1.4%, 2.7% and 8.1% of cases, respectively. Conclusions: Lower native ALK copy number in ALK(+) NSCLC suggests ALK fusion occurs early, preceding chromosomal instability. Elevated native ALK copy number rarely reflects focal amplification and native ALK copy number increases alone are not associated with sensitivity to ALK inhibition in vitro. Neither native or rearranged copy number, nor positive cell count within ALK(+) tumors influences clinical benefit from ALK inhibition. As 8.5% of ALK(-) cases fall within 5% of the established >15% cell positive threshold, further investigation of ALK status by other diagnostic techniques in this subset may be warranted.