Detection of EGFR T790M mutation in bronchial washing fluid and therapeutic efficacy of lazertinib in T790M-positive NSCLC patients after first-line EGFR-TKI failure: An open-label, multicenter, single-arm exploratory study

Background: After progression on first-line (1L) EGFR-TKI in EGFR-mutant non-small cell lung cancer(NSCLC), detecting the T790M resistance mutation is pivotal to guide third-generation EGFR-TKIs. Bronchial washing fluid (BWF) digital droplet PCR (ddPCR) may improve detection versus plasma, yet its clinical utility and outcomes require clarification. Methods: The Leclaza IIT 020 was an open-label, multicenter, single-arm, exploratory study evaluating patients with advanced NSCLC who experienced progression on 1L EGFR-TKI and underwent bronchoscopy for ddPCR-based T790M testing in BWF with paired plasma. Key eligibility included 1L EGFR-TKI failure and T790M positivity on bronchial washing by ddPCR. Patients with BWF T790M positivity received lazertinib 80 mg once daily. The primary endpoint was objective response rate (ORR); secondary endpoints included progression-free survival (PFS),disease control rate (DCR), and descriptive concordance across modalities. Results: Among 26 consecutively tested patients, BWF T790M was positive in 10 (38.5%) and negative in 16. Baseline demographics were statistically comparable between the BWF-positive and -negative groups (median age 72.5 years [IQR 66.0-77.8]; female 76.9%). Most patients had stage IV disease at diagnosis (92.3%) and prior exposure to afatinib (~65%); smoking history and comorbidity profiles were also similar between the groups. Detection rates by modality were: plasma 6/26 with a detectable ddPCR signal (23.1%), but all plasma samples remained below the clinical reporting threshold (0/26 reportable), and BWF 10/26 (38.5%). Concordance analyses showed higher agreement for BW compared with initial tumor genotyping and moderate concordance with plasma, supporting the added yield of bronchoscopic sampling when plasma is negative or below threshold. Ten BWF-positive patients received lazertinib; best overall responses were partial response in 5 (50.0%), stable disease in 4 (40.0%), and progressive disease in 1 (10.0%), yielding an ORR of 50.0% and a DCR of 90.0%. Median PFS was 12.8 months (95% CI, 6.1-not estimable), with 6- and 12-month PFS of 72.7% and61.4%, respectively. Median follow-up was 9.4 months (95% CI, 3.7-12.4). Conclusions: In patients progressing on 1L EGFR-TKI, ddPCR on BWF identifies T790M more frequently than clinically reportable plasma testing and enriches for candidates who achieve favorable outcomes on lazertinib (ORR ~50.0%, median PFS ~12.8 months). BWF testing provides a practical escalation pathway when plasma is negative or below threshold, particularly inpatients with limited extra-thoracic disease burden. Prospective validation in larger cohorts is warranted.