Abstract 971: Validation of a tumor-naïve circulating tumor DNA (ctDNA) response monitoring panel in advanced non-small cell lung cancer (aNSCLC)

Abstract

Abstract Introduction: Monitoring of ctDNA can be a minimally invasive complement to tumor imaging for assessing treatment effect. Technical limitations require methods to distinguish tumor signal from clonal hematopoiesis (CH), often including non-tumor sequencing. We developed a tumor-naïve panel (FoundationOne® Monitor) to quantify ctDNA tumor fraction (TF). TF analytical validation used peripheral blood mononuclear cell (PBMC) sequencing. We then leveraged plasma collected serially from patients with aNSCLC in the real-world (rw) Prospective Clinico-Genomic (PCG; NCT04180176) study to investigate the utility of TF for monitoring therapy (tx) response. Methods: TF was quantified using a combination of aneuploidy and variant allele frequencies of genomic alterations (GAs), while excluding CH mutations and aneuploidy using fragmentomic signal from cfDNA. In the PCG study, data from consenting patients were collected from electronic health records from 23 participating Flatiron Health Research Network sites. We analyzed plasma collected 6-15 weeks after start of tx in exploratory and validation cohorts per prespecified criteria. We defined molecular response (MR) as undetectable TF on tx regardless of baseline TF. Hazard ratios (HR) and 95% confidence Interval (CI) were calculated with univariate Cox proportional hazard regression. Results: For TF validation, 1135 samples separate from the PCG study with paired PBMC results were included. Overall, 24/4274 (0.56%) CH derived non-aneuploidy GAs detected in 1134 samples were falsely classified as somatic. Of these, 4 were detected among 317 samples with no tumor signal, resulting in a false positive TF value (specificity = 98.7%). CH derived aneuploidy was observed in 27 PBMCs and was appropriately filtered during TF estimation in all cases. Assessing the impact of CH aneuploidy filtering on sensitivity, we only identified 1/320 (0.31%) samples where non-CH derived aneuploidy was omitted. To assess clinical validity, 222 patients were analyzed from the PCG study. MR was assessed after a median of 10.8 weeks of tx (IQR 8.9-12.0). In a subset of 152 patients treated with physicians’ choice, MR was associated with favorable rw progression free survival (rwPFS: 9.4 v 2.8 months [mo]; HR = 0.30; 95% CI: [0.21-0.44]) and rw overall survival (rwOS: 22.0 v 7.5 mo; HR = 0.33 [0.22-0.49]). We validated this finding on 70 patients receiving immunotherapy (50 with chemo). Again, MR was associated with favorable rwPFS (9.0 v 2.8 mo; HR = 0.28 [0.16-0.51]) and rwOS (20.2 v 9.4 mo; HR = 0.42 [0.23-0.78]). Conclusions: We describe a highly specific tumor naïve algorithmic filtration of non-tumor signal to enable high confidence ctDNA quantification and MR assessment. On tx MR is associated with favorable outcomes. These findings may enable personalized tx approaches tailored to a patient’s risk of progression and downstream cancer morbidity. Citation Format: Anne C. Chiang, Russell W. Madison, Yanmei Huang, Alexander Fine, Dexter X. Jin, Geoffrey R. Oxnard, Jason Hughes, Zoe June Assaf, Yi Cao, Vladan Antic, Ole Gjoerup, Amanda Young, David Fabrizio, Shaily Lakhanpal, Richard Zuniga, Katja Schulze, Lincoln W. Pasquina. Validation of a tumor-naïve circulating tumor DNA (ctDNA) response monitoring panel in advanced non-small cell lung cancer (aNSCLC) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 971.