Abstract 1144: Dynamic changes in circulating tumor DNA (ctDNA) in patients treated with sotorasib for KRAS G12C mutant non-small cell lung cancer

Abstract Background: Sotorasib was recently approved for KRAS G12C mutant advanced non-small cell lung cancer (NSCLC). We explored whether the presence of specific mutations and dynamic changes in circulating tumor DNA (ctDNA) would be of prognostic/predictive value for patients (pts) with KRAS G12C mutations on sotorasib. Methods: Serial plasma was collected from pts with advanced KRAS G12C mutant NSCLC treated on the registrational CodeBreaK 100 trial, including at baseline (cycle 1, day 1, C1D1), cycle 2, day 1 (C2D1), and at end of treatment (EOT). At one site (Memorial Sloan Kettering), plasma was collected at additional timepoints, including C1D2 and C1D8. Plasma was analyzed using the Resolution Bioscience ctDX LungTM assay, a 23-gene panel. Results: Plasma was collected from 179 pts, for a total of 624 evaluable samples. At baseline, 77% of pts (129/168) had detectable plasma KRAS G12C mutations. Lower G12C allele frequency (AF) at baseline correlated with better % change in radiographic Sum of Diameters (SoD) (Spearman’s rho=0. 22, p 0. 001). Pts without baseline detectable plasma KRAS G12C had a median PFS of 11. 5 months as compared to 4. 1 months in those with detectable G12C (HR 2. 8, p0. 001), and were more likely to respond (HR = 2. 76, p 0. 001). Pts with no baseline detectable ctDNA also had longer PFS (11. 1 versus 5. 1 months, HR 2. 07, p=0. 01). While responses were seen across a variety of genomic alterations, complete responders had fewer co-mutations overall. 75% (3) of complete responders (CRs) had no mutations of any kind detected vs 12% (19) of non-CRs (p=0. 008, Fisher exact test). The number of altered genes was lower in responders (p0. 001, Poisson regression), and CRs alone tended to have fewer altered genes (p=0. 06), though the number of cases was very small (4). EGFR amplification at baseline was observed in patients who achieved a partial response (PR), but not a CR. The early kinetics of ctDNA comparing C1D2 to C1D1 and C2D1 to C1D8 did not differ between pts who had a decrease in SoD by RECIST on first scans (completed around C3D1) and those who had an increase (p0. 1, tested by within and between group medians) ; both increases and decreases in G12C AF were seen at these time points regardless of radiographic changes. A subset of both responders (26% 13/50) and non-responders (54% 42/78), had a transient drop in ctDNA between C1D1 and C2D1 with a subsequent increase after C2D1. Those with no detectable ctDNA at C2D1 had longer median PFS (8. 3 months) compared to those with detectable ctDNA (4. 4 months, p=0. 007). Conclusion: At baseline, lower G12C AF and lower number of altered genes were associated with improved outcomes, suggesting there may be benefit to early initiation of sotorasib when co-mutational burden is low. Changes in ctDNA early in treatment are dynamic. Absence of detectable ctDNA at C2D1 may be associated with longer PFS. Citation Format: Yonina R. Murciano-Goroff, Ferdinandos Skoulidis, Vamsidhar Velcheti, Gerald S. Falchook, Grace Dy, Suresh Ramalingam, Martin J. Edelman, Jürgen Wolf, Antoine Italiano, Jorge S. Reis-Filho, Pedram Razavi, Alexander Drilon, Sandra Misale, Rona Yaeger, Mark Li, Antreas Hindoyan, Agnes Ang, Abraham A. Anderson, David Hong, Piro Lito, Ramaswamy Govindan, Bob T. Li. Dynamic changes in circulating tumor DNA (ctDNA) in patients treated with sotorasib for KRAS G12C mutant non-small cell lung cancer abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts) ; 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83 (7Suppl): Abstract nr 1144.