Multiple ctDNA- based biomarkers predict benefit from selective RET Inhibition in non-small cell lung cancer patients: exploratory analysis of a prospective study

TL;DR

This study shows that measuring ctDNA in patients with RET fusion-positive lung cancer can predict treatment response and disease progression, offering non-invasive biomarkers for early intervention.

Contribution

The study introduces multiple ctDNA-based biomarkers that predict benefit from RET inhibition in RET fusion-positive NSCLC patients.

Findings

Baseline PIK3CA co-mutations were linked to worse progression-free survival in RET fusion-positive NSCLC patients.

Lower baseline ctDNA levels were associated with better progression-free survival across multiple metrics.

Early ctDNA clearance correlated with prolonged PFS and better disease control.

Abstract

Selective RET inhibitors such as pralsetinib have become the standard of care for patients with RET fusion-positive non-small cell lung cancer (NSCLC). Serial analysis of circulating tumor DNA (ctDNA) has proven effective in monitoring disease control/progression and therapeutic response in NSCLC. In this prospective study, we analyzed longitudinal ctDNA profiles (at baseline, week 8, and at progression) in Chinese patients with advanced RET fusion-positive NSCLC treated with pralsetinib (NCT03037385), utilizing allele frequency-based, cfDNA quantity-normalized, and methylation-based metrics. Associations between ctDNA dynamics, tumor response, and genomic alterations were assessed. A total of 21 patients were enrolled. Baseline PIK3CA co-mutations were associated with inferior progression-free survival (PFS; 3.0 vs. 12.4 months, P < 0.001). Superior PFS was observed in patients with lower baseline ctDNA levels across all metrics: allele frequency-based (HR = 0.24; 95% confidence interval [CI], 0.07–0.80; P = 0.012), cfDNA quantity-normalized (HR = 0.20; 95% CI, 0.05–0.71; P = 0.006), and methylation-based (HR = 0.09; 95% CI, 0.01–0.85; P = 0.010). Early ctDNA clearance at the first radiographic assessment was also associated with prolonged PFS (median PFS not reached vs. 4.8 months; P = 0.002) and enhanced disease control (71.4% vs. 0%). Moreover, three distinct ctDNA dynamic profiles—clearance-rebound, reduction-rebound, and sustained clearance—were associated with different progression patterns (systemic progression, new extrathoracic lesions, new intracranial/intrathoracic lesions). No evidence of histologic transformation was identified at the time of progression. KRAS G12R and other non-canonical alterations emerged in ctDNA-rebound samples. Molecular progression preceded radiographic progression by a mean interval of 2.2 months. These findings suggest that ctDNA-based surveillance using multiple metrics, enables early forecasting of tumor response and progression in RET fusion-positive NSCLC. Early ctDNA clearance and dynamic profiles provide non-invasive biomarkers for early intervention, warranting further validation with ctDNA-guided strategies.

The online version contains supplementary material available at 10.1186/s40364-025-00809-8.

Figures