Residual ctDNA Detection Predicts Benefit From Definitive Chemoradiotherapy and Immune Checkpoint Inhibitors in Locally Advanced NSCLC

Abstract

Therapeutic efficacy differs across unresectable locally advanced non-small cell lung cancer (LA-NSCLC) patients receiving definitive chemoradiotherapy (CRT), and the dynamic change in circulating tumor DNA (ctDNA) is closely associated with clinical outcomes. However, evidence for the predictive effect of ctDNA in LA-NSCLC patients with CRT and immunotherapy, the current standard of care, remains limited. It is unclear whether residual ctDNA after definitive CRT can guide clinical decisions on further use of immune checkpoint inhibitors (ICIs) consolidation. We prospectively included 73 patients with unresectable stage II-III NSCLC. All patients received definitive concurrent or sequential CRT, with the prescribed dose of 60 Gy and ≥2 cycles of platinum-based chemotherapy. Thirty-seven (50.7%) patients further underwent ICIs therapy (18 durvalumab, 9 pembrolizumab, 10 others). Peripheral blood samples were collected from all patients before any treatment (baseline), 1 month after CRT (post-CRT), and at the time of progression. All plasma specimens were analyzed with next-generation sequencing panel of 474 cancer-related genes. Plasma samples with ≥1 variant detected were defined as detectable ctDNA. The primary endpoint was progression-free survival (PFS). After the median follow-up of 25.4 months, median overall survival (OS) was not reached (NR), and median PFS was 16.7 months (95% CI, 12.4-26.6) for all patients. Compared with baseline, ctDNA abundance significantly decreased after definitive CRT (P<0.001) but relatively increased at progression (P = 0.051). Patients treated with CRT plus ICIs exhibited significantly longer OS (median, NR vs 22.1 months [95% CI, 16.7-NR]; P = 0.002) and PFS (median, 24.8 months [95% CI, 16.1-NR] vs 11.4 months [95% CI, 6.5-NR]; P = 0.016) than those with CRT alone. Post-CRT residual ctDNA was associated with significantly poorer OS (median, 18.3 months [95% CI, 14.8-NR] vs NR; P = 0.002) and PFS (median, 6.5 months [95% CI, 5.6-NR] vs 24.8 months [95% CI, 18.8-NR]; P<0.001), whereas baseline ctDNA predicted neither OS (P = 0.310) nor PFS (P = 0.570). For patients with post-CRT detectable ctDNA, further ICIs therapy brought significant benefit in both OS (median, NR vs 14.8 months [95% CI, 12.0-NR]; P = 0.012) and PFS (median, 16.1 months [95% CI, 5.8-NR] vs 6.2 months [95% CI, 4.2-NR]; P = 0.043). However, in patients with post-CRT ctDNA clearance, there was no significant difference in OS (P = 0.080) or PFS (P = 0.151) between patient with and without ICIs, which suggested less clinical benefit. Post-CRT residual ctDNA predicted worse survival in LA-NSCLC, but indicated more benefit from further ICIs therapy and thereby could facilitate personalization of consolidation immunotherapy. Further prospected studies with large sample size are warranted to validated these findings.