Manuel Leiva

PURPOSE Immunotherapy targeting PD-L1 improves outcomes in patients with unresectable stage III non–small cell lung cancer (NSCLC) and no progression after definitive, concurrent chemoradiotherapy (cCRT). Earlier administration of immunotherapy, simultaneously with cCRT, may improve outcomes further. METHODS Eligible patients were randomly assigned (2:1) to receive either durvalumab or placebo administered from the start of cCRT. Patients without progression after completing cCRT received consolidation durvalumab or placebo (per initial random assignment) until progression. The primary end point was progression-free survival (PFS) by blinded independent central review. Key secondary end points included objective response rate (ORR), overall survival (OS), the proportion of patients alive at 24 months (OS24), and safety. RESULTS In total, 328 patients were randomly assigned to receive durvalumab (n = 219) or placebo (n = 109). There was no statistically significant difference with durvalumab versus placebo in PFS (hazard ratio [HR], 0.85 [95% CI, 0.65 to 1.12]; P = .247) or OS (HR, 1.03 [95% CI, 0.78 to 1.39]; P = .823); OS24 was 58.4% versus 59.5%, respectively. Confirmed ORR was 60.7% with durvalumab versus 60.6% with placebo (difference, 0.2% [95% CI, −15.2 to 16.3%]; P = .976). With durvalumab versus placebo, respectively, maximum grade 3 or 4 adverse events (AEs) occurred in 53.4% versus 59.3% of patients, pneumonitis or radiation pneumonitis (group term) in 28.8% (grade ≥3: 4.6%) versus 28.7% (grade ≥3: 5.6%), AEs leading to discontinuation of durvalumab or placebo in 25.6% versus 12.0%, and fatal AEs in 13.7% versus 10.2%. CONCLUSION Among patients with unresectable stage III NSCLC, durvalumab administered from the start of cCRT failed to demonstrate additional benefit compared with cCRT plus placebo. Consolidation durvalumab following definitive cCRT remains the standard of care in this setting.

5502 Background: In LACC, there is an unmet need for prognostic biomarkers as about 1/3 of patients (pts) relapse after chemoradiotherapy (CRT). The global randomized CALLA trial (NCT03830866) of durvalumab (D) in combination with CRT followed by D (D+CRT arm) vs CRT (CRT arm) did not significantly improve progression-free survival (PFS) in a biomarker unselected intent-to-treat (ITT) population. We analyzed the association of ultrasensitive ctDNA detection with relapse and survival in the largest ctDNA data set in LACC to date. Methods: Adult women with Stage IB2-IIB node positive (N+) or IIIA-IVA any N LACC (ITT) were randomized 1:1 to D+CRT or CRT alone. NeXT Personal (Personalis, Fremont, CA), an ultrasensitive tumor-informed MRD assay with up to 1,800 patient-specific variants from WGS, was used for ctDNA analysis from Cycle 1 Day 1 (C1D1; baseline [BL]), C3D1, and 6 mo post treatment initiation. Correlations were analyzed between ctDNA detection and outcomes (PFS, overall survival [OS]). Results: Of 770 pts randomized, the biomarker-evaluable population (BEP) comprised 185, 186, and 130 pts at BL, C3D1, and 6 mo, respectively. BL pt characteristics, PD-L1, PFS, and OS between BEP and ITT populations were generally similar. ctDNA was detected in 99% of pts at BL and decreased after treatment, reaching 23% in the D+CRT and 36% in the CRT arm at 6 mo. The lower detection rate in the D+CRT arm was associated with the PD-L1 tumor area positivity (TAP) ≥20% subpopulation. At BL, pts with low (<BL median [5268.2 ppm]) ctDNA levels had a reduced risk of progression vs pts with high (≥median) ctDNA levels (PFS hazard ratio [HR] D+CRT 0.57 [95% CI, 0.26-1.26]; CRT 0.62 [0.31-1.23]). Pts with detectable ctDNA at C3D1 or 6 mo had a higher risk of progression independent of treatment arm (Table). No differences in risk of progression between the D+CRT vs CRT arms were observed based on ctDNA detection. Correlations between ctDNA and OS will be presented. Conclusions: This pre-planned analysis of a large, global LACC population from CALLA demonstrates the high sensitivity of a personalized ctDNA assay. High ctDNA levels at BL were associated with higher risk of progression or death. Lower ctDNA detection rates after treatment with D+CRT and CRT correlated with improved survival and highlight increased tumor control by D, especially in the PD-L1 TAP ≥20% subpopulation. This analysis supports the potential utility of ultrasensitive ctDNA analysis to guide treatment decisions in LACC. Clinical trial information: NCT03830866 . D+CRT CRT Not detected Detected Not detected Detected C3D1 n=60 n=33 n=56 n=37 Median PFS (95% CI), mo NC (NC-NC) 14.03 (7.49-NC) NC (NC-NC) 10.68 (7.39-NC) HR (95% CI) 0.23 (0.11-0.50) 0.15 (0.07-0.33) 6 mo n=49 n=15 n=42 n=24 Median PFS(95% CI), mo NC (NC-NC) 10.35 (7.49-NC) NC (NC-NC) 12.98 (10.38-NC) HR (95% CI) 0.04 (0.01-0.16) 0.04 (0.01-0.17) NC, not computed/not reached.