Authors: Peronne Joseph, Zhengyi Chen, Abhishek Ajay, Pingfu Fu, Afshin Dowlati
Published: 2024-06-28
DOI: 10.1200/jco.2024.42.16_suppl.8107
Source: Full article
8107 Background: Dynamic changes in circulating tumor DNA (ctDNA) have not been studied in small-cell lung cancer (SCLC). Here, we assessed the changes of ctDNA over time and clinical outcomes. Methods: We performed a 105-gene, hybrid-capture, next-generation sequencing (NGS) liquid biopsy assay that detects SNVs, in/dels, CNVs, and chromosomal rearrangements. Patients had ctDNA assessed at 3 time points: 1) before initiation of front-line therapy; 2) at remission (within 3-6 weeks of 4th cycle of chemotherapy+/- immunotherapy), and 3) at the time of relapse (based on radiographic imaging). Study outcome measures include OS, PFS, DFS (duration of response from completion of 4th cycle of systemic treatment until relapse), and overall response rate (ORR). Survivor distribution was estimated using Kaplan-Meier methods. Cox model was also used to estimate the effects on OS, PFS and DFS. The effects of predictors on ORR were estimated using logistic regression. The T-test was used to examine the difference in continuous measures between the two groups, and the chi-square test was used to examine the association between two categorical factors. All tests were two-sided, and p-values < 0.05 were considered significant. Results: 62 patients were enrolled. The median follow-up was 12.3 mos. Patients were treated with standard-of-care chemotherapy +/- immunotherapy for 4 cycles. 71% had extensive stage disease. At baseline, 100% of patients had detectable variant alleles in TP53,and 70% had variant RB1. Patients with ES-SCLC had significantly higher maximum (max) Variant Allelic Fraction (VAF) (alteration with highest VAF) compared to LS-SCLC (59.1 vs 27.2%, p = 0.04). Median max VAF at baseline was 52.3%, 0.3% at remission, and 38.4% at relapse, indicating a response to therapy. However, ORR did not correlate with baseline max VAF (p = 0.2). Baseline max VAF (comparing those < or > VAF of 45%) was not associated with OS (19.4 vs 12.2 mos, p = 0.6) but highly impacted PFS (11.2 vs. 6.1 mos, p = 0.02) and DFS (12.1 vs 2.2 mos, p = 0.004). Baseline max VAF was an independent predictor of PFS and DFS. A drop in max VAF from pretreatment to remission (comparing those with a > or < 99% decrease) also did not associate with OS, PFS, or DFS. Patients with RB1 alterations at pretreatment that had residual RB1 alterations at the time of clinical remission had significantly worse PFS (6.3 vs 12.1 mos) and DFS (1 vs 8 mos, p = 0.02). Conclusions: Baseline max VAF ctDNA levels strongly correlate with PFS and DFS (duration of remission from completion of 4th cycle of treatment) but not OS and ORR. Most strikingly, those with detectable residual RB1 alterations at completion of proscribed chemotherapy have an extremely short DFS of 1 month that justifies early assessment for clinical trials. This approach has led to a near 90% success rate in enrollment on 2nd line clinical trials or SOC options.