Abstract 41P
Background
ROS1+ is a distinct molecular subset of NSCLC with a therapeutically druggable target. Information regarding the potential effect of ROS1+ fusion partners and resistance mechanisms on clinical characteristics and outcomes of TKI treatment in the real-world and in vitro remain limited.
Methods
In this multicenter study, we identified 55 ROS1+ patients in the past 10 years. Fusion partners were identified for 25 patients using Archer analysis. Clinical data were retrieved retrospectively for 48 patients. In vitro experiments with Ba/F3 cells expressing SLC34A2-ROS1 fusion were performed to investigate on-target resistance mechanisms.
Results
The median age was 62 years, 54% was female, and 33% were never smokers. ROS1 fusion partners were EZR (n=7), CD74 (n=9), SDC4 (n=6), SLC34A2 (n=1), LDLR (n=1) and HLA (n=1). Nine patients were not treated with TKI due to localized disease (n=6) or ECOG PS >3 (n=3). Thirty-nine patients received TKI (crizotinib; n= 36, non-crizotinib; n=3), of which 12 received it as 2nd line treatment after 1 - 4 lines of chemotherapy. The response rate of crizotinib was 56% with a median progression-free survival (mPFS) of 5 months (95% CI 2.8–7.3). Second line TKI treatment with lorlatinib in 15 patients resulted in an ORR of 40% and a mPFS 3.7 months [95% CI 2.2–5.2]. The median overall survival in the total cohort (n=48) was 24 months (95% CI 20.2–28.2). Univariate analysis showed no statistical correlation between fusion partners and survival on TKI treatment. 7/15 patients underwent a tumour biopsy before 2nd line TKI lorlatinib treatment, which revealed on-target resistance mutations in 3 patients (L2026M (n=1) and G2032R (n=2)). These patients showed no response to lorlatinib. Additionally, in vitro studies showed that SLC34A2-ROS1 BaF3 cells with L2026M or G2032R are more resistant to lorlatinib compared to unmutated SLC34A2-ROS1 BaF3 cells.
Conclusions
No associations were observed in this real-world ROS1+ population of the fusion partners in relation to TKI outcome. Second line treatment with lorlatinib was ineffective in patients with on-target resistance mutations. Consistent with these findings, lorlatinib was unable to suppress downstream signaling of SLC34A2-ROS1 G2032R or L2026M in vitro.
Legal entity responsible for the study
The authors.
Funding
Has not received any funding.
Disclosure
W. Timens: Financial Interests, Institutional, Other, Consultancy payments: Merck Sharp & Dohme, Bristol Myers Squibb. J.T.J.N. Hiltermann: Financial Interests, Institutional, Advisory Board: BMS, AZD, Roche, Boehringer, Pfizer; Financial Interests, Institutional, Research Grant: BMS, AstraZeneca, Roche; Non-Financial Interests, Personal, Principal Investigator: BMS, AstraZeneca, Roche, Novartis, Merck, GSK, Amgen. A.J. van der Wekken: Financial Interests, Institutional, Principal Investigator: AstraZeneca; Financial Interests, Institutional, Research Grant: Boehringer Ingelheim, Pfizer, Roche, Takeda, Janssen Cilag, Lilly, Amgen, Merck. All other authors have declared no conflicts of interest.