Advanced non-small cell lung cancer (NSCLC) treatment has significantly improved with targeted therapies. Comprehensive biomarker testing, including for ROS1 rearrangements, is recommended before first-line treatment. However, tissue availability often limits testing. Liquid biopsies, analyzing circulating tumor DNA (ctDNA), offer a less invasive alternative. This study, part of the Blood First Assay Screening Trial (BFAST), focuses on Cohort D, evaluating entrectinib's efficacy and safety in ROS1-positive NSCLC patients identified solely by liquid biopsies. Entrectinib, a potent ROS1, TRK, and ALK inhibitor, exhibits CNS penetration, a crucial factor given the high incidence of brain metastases in ROS1-positive NSCLC. Prior studies using tissue-based testing demonstrated entrectinib's effectiveness, motivating this investigation to confirm these findings using liquid biopsies. The aim was to demonstrate consistency between liquid biopsy-based and tissue-based selection for entrectinib treatment in ROS1-positive NSCLC patients.

Numerous studies highlight the improved survival outcomes associated with targeted therapies in oncogene-driven advanced NSCLC. Despite recommendations for comprehensive biomarker testing, real-world data shows low rates of multi-biomarker testing. Limitations in tissue availability and the invasiveness of biopsies contribute to this gap. Liquid biopsies, utilizing ctDNA analysis, offer a faster, less invasive approach with high concordance to tissue biopsies. While less sensitive, their faster turnaround times and equivalent time-to-treatment offer a significant clinical advantage. Previous studies have demonstrated entrectinib's efficacy in ROS1-positive NSCLC identified via tissue-based testing, showing durable responses and intracranial efficacy. The BFAST trial, with its multi-cohort design, aims to validate these findings using liquid biopsies as the sole basis for patient selection.

BFAST is a global, open-label, multi-cohort trial. Cohort D enrolled treatment-naïve patients ≥18 years with stage IIIB/IV ROS1-positive NSCLC, confirmed solely by liquid biopsy NGS testing (FoundationOne Liquid CDx or FoundationACT). Patients received entrectinib 600 mg daily. Efficacy endpoints included investigator-assessed confirmed ORR (primary endpoint), CBR, DoR, PFS, and OS. Safety was assessed by monitoring adverse events (AEs). Biomarker analyses, including ROS1 fusion partners, co-mutations (e.g., TP53), ctDNA fraction (CTF), and ctDNA clearance, were performed as post-hoc exploratory analyses. Statistical analysis included Kaplan-Meier curves for survival endpoints and descriptive statistics for biomarker data. The sample size calculation aimed to demonstrate consistency with the ORR observed in previous tissue-based entrectinib studies.

From January 11, 2018, to December 9, 2020, 5220 patients were screened; 92 (1.8%) had ROS1-positive NSCLC detected by liquid biopsy. Fifty-five treatment-naïve patients were enrolled, with 54 having measurable disease. The primary endpoint was met: the investigator-assessed confirmed ORR was 81.5% (95% CI: 68.6–90.8), consistent with the 73.4% (95% CI: 63.3–82.0) ORR from the integrated analysis of entrectinib (tissue-based testing). Median DoR was 13.0 months (investigator) and 16.7 months (IRF); median PFS was 12.9 months (investigator) and 14.8 months (IRF). OS data were immature, but the 12-month OS rate was 79%. The safety profile was consistent with previous reports; most TRAEs were non-serious. Biomarker analysis revealed CD74 as the most common ROS1 fusion partner; no difference in response was observed based on fusion partner. TP53 mutations were associated with numerically shorter DoR and PFS. There was no association between baseline CTF and clinical outcomes, although CTF weakly correlated with tumor burden. CtDNA clearance by cycle 3 day 1 was associated with longer DoR and PFS. Analysis of acquired resistance mechanisms identified ROS1 G2032R in two patients.

The BFAST Cohort D results confirm the efficacy and safety of entrectinib in ROS1-positive NSCLC, using liquid biopsies for patient selection. The consistent ORR compared to tissue-based studies validates liquid biopsies as a valuable tool for identifying suitable candidates for entrectinib. While numerical differences in DoR and PFS were observed, they could be attributed to differences in patient populations (e.g., detectable ctDNA at baseline correlating with higher tumor burden, TP53 mutation prevalence). The study highlights entrectinib's potential for delaying or preventing CNS metastases, although further investigation is needed due to the low number of patients with baseline CNS metastases. Post-hoc biomarker analyses suggest that ctDNA clearance might predict clinical outcomes, and TP53 mutations may be associated with poorer prognosis. However, the limited sample size necessitates cautious interpretation of these findings. The study's strength lies in demonstrating the clinical utility of liquid biopsies for selecting patients for targeted therapy, improving access and reducing invasiveness.

This study demonstrates the clinical applicability of liquid biopsies in identifying ROS1-positive NSCLC patients suitable for entrectinib treatment. The consistent efficacy and safety observed in this trial, compared to previous tissue-based studies, reinforces the value of liquid biopsies in streamlining treatment decisions. Further research should focus on validating the prognostic potential of ctDNA clearance and TP53 mutations, and explore the relationship between ctDNA dynamics and clinical response. Larger, comparative studies are needed to definitively assess the clinical benefits of entrectinib compared to other ROS1 inhibitors.

The limitations of this study include the relatively small sample size, lack of a comparator arm, and short follow-up time at data cutoff. The use of two different clinical trial assays (FoundationOneLiquid CDx and FoundationACT) could have introduced variability, although high concordance was demonstrated. The inherent limitation of ctDNA-based testing is its dependence on tumor shedding into the blood, potentially excluding some patients with low tumor burden. The post-hoc exploratory biomarker analyses, due to small sample size, should be interpreted with caution, and further studies are warranted to validate these findings.