Entrectinib Shows Real-World PFS Benefit in ROS1+ NSCLC

Robert C. Doebele, MD

Entrectinib (Rozlytrek) has demonstrated superiority in real-world progression-free survival (PFS), overall survival (OS), and time-to-treatment discontinuation (TTD) compared with crizotinib (Xalkori) in ROS1-positive non—small cell lung cancer (NSCLC) in a retrospective study, explained Robert C. Doebele, MD, PhD.

The study included 122 patients with ROS1-positive NSCLC, 53 of whom received entrectinib and 69 received crizotinib. Those who received crizotinib were from the Flatiron Health electronic health records—derived database, and entrectinib-treated patients were vetted from the phase I/II ALKA-372-001, STARTRK-1, STARTRK-2 trials.

Results showed that the median TTD was doubled in patients who received entrectinib versus crizotinib at 14.6 months versus 8.4 months, respectively, and the real-world PFS was 19.0 months with entrectinib and 8.5 months with crizotinib. The median OS could not be estimated with entrectinib compared with 19.9 months in crizotinib.

However, further research is needed to confirm these findings, Doebele explained.

“One of the questions we most often get asked is, ‘How does this compare with crizotinib?’ That answer becomes very complicated by the fact that there are multiple crizotinib trials, all with a different range of response rates, PFS, and…enrollment of patients with brain metastases,” said Doebele, an associate professor in the Division of Medical Oncology at the School of Medicine, University of Colorado.

In an interview with OncLive, Doebele discusses this retrospective trial comparing the efficacy of entrectinib with crizotinib in the treatment of patients with ROS1-positive NSCLC.

OncLive: What was the rationale for examining the TTD and real-world PFS of entrectinib compared with crizotinib?

Doebele: We designed a study to try and understand how entrectinib compares with crizotinib in ROS1-positive NSCLC. Others and I have recently presented a single-arm, phase II integrated analysis of patients with ROS1-positive NSCLC treated with entrectinib. We know that the response rate for entrectinib was 77% in ROS1-positive NSCLC patients, the PFS was in the range of 20 months, and the duration of response was in the range of greater than 2 years.

The difficulty is we saw response rates from 60% to 70% from the different crizotinib trials. More importantly, we saw PFS anywhere from 9 months to 19 months, with an average somewhere in the range of 13 to 15 months. The other problem is making sure that any comparisons that we make have trials where crizotinib was administered in patients with brain metastases. That's one of the unique features of entrectinib; it has central nervous system (CNS) penetration.

How was this retrospective study designed and what did you find?

The study is a relatively simple design in the sense that we took the entrectinib-treated patients from the integrated analysis of the ALKA-372-001, STARTRK-1, and STARTRK-2 trials. We took the data that has been previously presented on these 53 patients with ROS1-positive NSCLC and compared that with a real-world cohort of ROS1-positive patients treated with crizotinib who were derived from the Flatiron Health database.

What we really tried to do was match the patients in the crizotinib virtual arm to the patients who we enrolled in the STARTRK-1, STARTRK-2, and ALKA-372-001 trials. That means enrolling patients with a higher proportion of brain metastases, which was not observed in prior crizotinib trials. We tried to match performance data and other features of these patients because, as you can imagine, one of the criticisms may be on how we selected the crizotinib-treated patients. We tried very hard to match those as closely as possible to patients in the entrectinib trials.

What we observed in this virtual clinical trial was doubling of the TTD. Entrectinib had an average time of approximately 14 or 15 months, and crizotinib was [approximately] 8 months, which matches the clinical trial results from some of the crizotinib studies in ROS1-positive NSCLC. Also, OS was not reached in the entrectinib arm, whereas it was only 20 months in the crizotinib arm. This evidence is supportive of entrectinib being a superior therapy to crizotinib; that would coincide with data that we have seen from our ALK inhibitors. Adding CNS penetration to a TKI improves PFS and outcomes for our patients.

With these types of data, could crizotinib be removed from the paradigm altogether?

The implications of these results, even based on the single-arm phase II trial, we know the ROS1-positive patients either have brain metastasis at baseline or risk of progressing in the brain on crizotinib. That alone is reason enough to consider entrectinib as the new frontline therapy for ROS1-positive NSCLC, assuming it's FDA approved.

This evidence adds to that argument. Entrectinib is a better drug. Even with the absence of a randomized clinical trial, which we think would probably take 5 or 7 or more years to read out, we think this is the strongest evidence that entrectinib should be the new frontline therapy for patients with ROS1-positive NSCLC.

What are the next steps for entrectinib?

The next steps, in regard to entrectinib, is understanding resistance that occurs following entrectinib failure. We know that there are next-generation drugs that are being evaluated, such as repotrectinib. Understanding the types of resistance we see with entrectinib may help us better understand which drug should follow entrectinib because, of course, one of our goals is always to have new therapies for our patients, even if they're progressing on our first-line therapies.

Doebele RC, Perez L, Huong Trinh H, et al. Time-to-treatment discontinuation (TTD) and real-world progression-free survival (rwPFS) as endpoints for comparative efficacy analysis between entrectinib trial and crizotinib real-world ROS1 fusion-positive (ROS1+) NSCLC patients. J Clin Oncol. 2019;37 (suppl; abstr 9070). doi: 10.1200/JCO.2019.37.15_suppl.9070.