Circulating tumor DNA predicts neoadjuvant immunotherapy efficacy and recurrence-free survival in surgical non-small cell lung cancer patients

BACKGROUND: There is currently a lack of effective biomarkers to evaluate efficacy of neoadjuvant therapy (NAT) for resectable non-small cell lung cancer (NSCLC) patients. Circulating tumor DNA (ctDNA) has been investigated as a non-invasive tool for the assessment of tumor burden and minimal residual disease (MRD). The utility of ctDNA profiling in reflecting NAT efficacy, however, has not been confirmed. This study explored the association of ctDNA change with treatment response to NAT and recurrence-free survival (RFS) after surgery. METHODS: Eligible patients with stage IB–IIIA NSCLC were retrospectively included if they had received neoadjuvant immunotherapy combined with chemotherapy (IO+Chemo), dual immunotherapy (IO+IO), or chemotherapy alone (Chemo). We conducted ctDNA profiling before and after NAT, after surgery, and during follow-ups using an ultra-deep lung cancer-specific MRD (LC-MRD) sequencing panel. RESULTS: A total of 22 patients who received NAT followed by surgery between August 2018 and July 2019 were included in this study. The major pathological response (MPR) rates were 58.33% (7/12) in the IO+Chemo group, 25.00% (1/4) in the IO+IO group, and 16.67% (1/6) in the Chemo group. The ctDNA dynamics during NAT were highly concordant with pathologic response, demonstrating 100% sensitivity and 83.33% specificity, for an overall accuracy of 91.67%. Pre-surgery detectable ctDNA (after NAT) trended to correlate with inferior RFS [hazard ratio (HR), 7.41; 95% confidence interval (CI): 0.91–60.22, log-rank P=0.03]. At 3-8 days after surgery, ctDNA was detectable in 31.8% of patients and was an independent risk factor for recurrence (HR, 5.37; 95% CI: 1.27–22.67; log-rank P=0.01). The presence of ctDNA at 3 months after surgery showed 83% sensitivity and 90% specificity for predicting relapse (C-index, 0.79; 95% CI: 0.62–0.95). During disease monitoring after surgery, molecular recurrence by means of ctDNA preceded radiographic relapse, with a median time of 6.83 months. CONCLUSIONS: This study investigated the potential of ctDNA in evaluating NAT efficacy in NSCLC, implying the high concordance between ctDNA and pathological response. We also set out the prognostic value of perioperative ctDNA in predicting recurrence.