215 Using ALK IHC test to identify potential responders to ALK inhibitors outside of NSCLC

Abstract

Background: In addition to anaplastic lymphomas and NSCLC, other solid malignancies can exhibit aberration in ALK expression potentially amenable to benefit from TKI ALK inhibitors. ALK expression by immunohistochemistry (IHC) has been extensively studied in NSCLC as a cost-effective and accurate screening tool for ALK rearrangement, and may identify patients (pts) with other mechanisms of ALK over-expression; however it has not been systematically studied outside of NSCLC. Methods: To validate ALK IHC test in a CLIA-certified lab, 86 cancer cases with documented ALK rearrangement by FISH (Abbott, Vysis break apart) results were tested by IHC using anti-ALK antibody D5F3 (Ventana) and evaluated by two pathologists (threshold: 3+ in >1% of cells). 43 NSCLC and 43 tumors of 23 other cancer types were included. Follow-up with treating physicians on ALK positive (pos) pts was attempted. Additionally, a prospective dataset of >2000 routinely profiled tumors tested for ALK IHC was analyzed. Results: In 43 NSCLC, ALK IHC was concordant with FISH in 93% (sensitivity = 90%, specificity = 95%). In non-NSCLC, all ALK-FISH pos tumors were also pos for ALK IHC (N = 4), and all IHC negative (neg) tumors were neg by FISH (N = 35) (sensitivity = 100%, specificity = 90%). Of the 4 IHC+/FISH+ tumors, 1 was myxoid leiomyosarcoma of the uterus (Pt A), 1 was inflammatory myofibroblastic tumor (IMT) (B), 1 was neuroendocrine tumor of the lung (C) and 1 was neuroblastoma (D). Of the 4 discordant (IHC +/FISH −) tumors, 1 was melanoma (E), 2 were serous ovarian cancers (F and G) and 1 was small cell cancer of the cervix (H). Follow-up was available for 5 IHC pos pts (A, B, F, G and H). A and B were treated with crizotinib: A deceased within 2 months and B has ongoing partial response to crizotinib for >2 years to date. G will start on a phase 1b crizotinib trial and outcome will be presented. F and H were not treated with ALK inhibitors. Prospective collection of ALK IHC on a large cohort of tumors (N > 2000) from >24 cancer types and additional follow up will be presented. Conclusions: Using FISH for comparison, our pilot study of 43 tumors of 23 cancer types outside of NSCLC shows the sensitivity and specificity for ALK IHC in non-NSCLC to be 100% and 90%. ALK IHC therefore carries great potential as a screening tool for ALK rearrangement and could potentially identify other ALK aberrations, thus identifying potential responders for ALK inhibitors in various cancers. No conflict of interest.