Abstract A025: Tumor fraction (TF) is associated with real-world progression-free survival (rwPFS) in advanced non-small-cell lung cancer (aNSCLC) patients treated with chemotherapy (chemo)

Abstract Background: Genomic ctDNA molecular response (gMR) predicts clinical benefit from immune checkpoint inhibitors, but associations with benefit from chemo are inconsistent. gMR may predict benefit from chemo after filtering to tissue-confirmed somatic variants, suggesting interference from clonal hematopoiesis (CH) mutations. Using methylation tumor fraction (TF) for MR assessment may mitigate the risk of CH interference without tissue and with higher sensitivity. Here, we report a retrospective study assessing the feasibility of using methylation- based TF from the Guardant Infinity platform to predict clinical benefit from chemo in a real- world cohort of patients (pts) with aNSCLC. Methods: Plasma samples were analyzed from pts (n=67) with aNSCLC, undergoing treatment with chemo only in 1st line or later setting, who had serial methylation TF results from tissue-free genomic and epigenomic ctDNA NGS. The present analysis included 27 pts with a baseline sample (T1) within 105 days prior to treatment start and an on-treatment timepoint (T2) 2-15 weeks after treatment start and >10 days prior to treatment end. gMR was based on the % change in mean variant allele fraction (VAF) T2/T1. TF % change was defined as TF (T2) / TF (T1). Relationships between time-to-next- treatment (TTNT; proxy for progression-free survival) and gMR or TF % change from T1 to T2 were explored using Cox proportional hazards (CPH) analysis. Multiple cutpoints were evaluated to label pts as molecular responders or non-responders for outcomes analysis. Gender, age, and therapy line were included as covariates for all analyses. Results: A ctDNA change from T1 to T2 was quantifiable in 85% of pts (18/27) by TF change versus 67% of pts by gMR due to improved limit of quantification for TF relative to somatic variants. CPH revealed that gMR defined by any ctDNA % change threshold was not associated with TTNT (p=(0.182- 0.862)). However, pts with a ≥95% decrease in TF from T1 to T2 had a significant association with improved TTNT (HR 0.14; 95% CI (0.03, 0.80); p=0.027). Furthermore, molecular responders defined by any decrease cutpoint >70% were significantly associated with improved TTNT (p≤0.033). Conclusions: A decrease in TF >70% early during chemotherapy was significantly associated with improved TTNT, even in this small cohort. The change in ctDNA was quantifiable in more patients by TF change than by gMR and included patients with no somatic variants detected. In contrast, gMR defined by any ctDNA % change threshold was not associated with TTNT. These results demonstrate feasibility of methylation-based TF to predict benefit from chemo in aNSCLC patients without tissue and with higher sensitivity than gMR, laying groundwork for future studies to address validity of serial TF monitoring in this patient population. Citation Format: Jing Wang, Sean Gordon, Errin Lagow, Tara Dinman, Carin Espenschied, Shile Zhang, Kimberly Banks. Tumor fraction (TF) is associated with real-world progression-free survival (rwPFS) in advanced non-small-cell lung cancer (aNSCLC) patients treated with chemotherapy (chemo) [abstract]. In: Proceedings of the AACR Special Conference: Liquid Biopsy: From Discovery to Clinical Implementation; 2024 Nov 13-16; San Diego, CA. Philadelphia (PA): AACR; Clin Cancer Res 2024;30(21_Suppl):Abstract nr A025.