Entrectinib in ROS1 fusion-positive non-small-cell lung cancer: integrated analysis of three phase 1-2 trials

Abstract

Background: Recurrent gene fusions, such as ROS1 fusions, are oncogenic drivers of various cancers, including non-small-cell lung cancer (NSCLC). Up to 36% of patients with ROS1 fusion-positive NSCLC have brain metastases at the diagnosis of advanced disease. Entrectinib is a ROS1 inhibitor that has been designed to effectively penetrate and remain in the CNS. We explored the use of entrectinib in patients with locally advanced or metastatic ROS1 fusion-positive NSCLC.

Methods: We did an integrated analysis of three ongoing phase 1 or 2 trials of entrectinib (ALKA-372-001, STARTRK-1, and STARTRK-2). The efficacy-evaluable population included adult patients (aged ≥18 years) with locally advanced or metastatic ROS1 fusion-positive NSCLC who received entrectinib at a dose of at least 600 mg orally once per day, with at least 12 months' follow-up. All patients had an Eastern Cooperative Oncology Group performance status of 0-2, and previous cancer treatment (except for ROS1 inhibitors) was allowed. The primary endpoints were the proportion of patients with an objective response (complete or partial response according to Response Evaluation Criteria in Solid Tumors version 1.1) and duration of response, and were evaluated by blinded independent central review. The safety-evaluable population for the safety analysis included all patients with ROS1 fusion-positive NSCLC in the three trials who received at least one dose of entrectinib (irrespective of dose or duration of follow-up). These ongoing studies are registered with ClinicalTrials.gov, NCT02097810 (STARTRK-1) and NCT02568267 (STARTRK-2), and EudraCT, 2012-000148-88 (ALKA-372-001).

Findings: Patients were enrolled in ALKA-372-001 from Oct 26, 2012, to March 27, 2018; in STARTRK-1 from Aug 7, 2014, to May 10, 2018; and in STARTRK-2 from Nov 19, 2015 (enrolment is ongoing). At the data cutoff date for this analysis (May 31, 2018), 41 (77%; 95% CI 64-88) of 53 patients in the efficacy-evaluable population had an objective response. Median follow-up was 15·5 monhts (IQR 13·4-20·2). Median duration of response was 24·6 months (95% CI 11·4-34·8). In the safety-evaluable population, 79 (59%) of 134 patients had grade 1 or 2 treatment-related adverse events. 46 (34%) of 134 patients had grade 3 or 4 treatment-related adverse events, with the most common being weight increase (ten [8%]) and neutropenia (five [4%]). 15 (11%) patients had serious treatment-related adverse events, the most common of which were nervous system disorders (four [3%]) and cardiac disorders (three [2%]). No treatment-related deaths occurred.

Interpretation: Entrectinib is active with durable disease control in patients with ROS1 fusion-positive NSCLC, and is well tolerated with a manageable safety profile, making it amenable to long-term dosing in these patients. These data highlight the need to routinely test for ROS1 fusions to broaden therapeutic options for patients with ROS1 fusion-positive NSCLC.

Funding: Ignyta/F Hoffmann-La Roche.

Figure 1:. Responses to entrectinib

(A) Best responses to entrectinib in the efficacy-evaluable population. (B) Best responses for patients with and without baseline CNS disease. (C) Best intracranial responses in patients with measurable CNS disease at baseline. According to blinded independent central review, 11 patients were responders. The best response to entrectinib in ROS1 inhibitor-naive patients with ROS1 fusion-positive lung cancers is shown as the maximum percentage improvement in the sum of longest diameters of identified target lesions compared with baseline. At baseline eight patients had lesions of less than 1 cm, which were considered unmeasurable. These eight patients were evaluated as having only non-target lesions and could only have complete responses or progressive disease, therefore results for 45 patients are shown in A and B. All assessments shown were based on blinded independent central review. SLD=sum of longest diameters.

Figure 2:. Time-to-event analyses

Kaplan–Meier curves of (A) duration of response, (B) progression-free survival, (C) overall survival, and (D) time to CNS progression. All assessments shown were based on blinded independent central review. Tick marks indicate censored patients.