Circulating tumour DNA (ctDNA) burden quantified through low-pass whole genome sequencing as an early pharmacodynamic biomarker of therapeutic response in patients enrolled on phase 1 trials.

Abstract

23 Background: Early phase trials are conducted to determine the recommended dose and safety of an investigational therapy. Targeted therapies and immunotherapy may not always have dose-dependent toxicities. Therefore, dose finding is increasingly reliant on optimal biological dosing. Pharmacodynamic biomarkers are increasingly important to demonstrate proof of concept. Using ctDNA allows for dynamic and early tumour assessments compared to imaging. We sought to evaluate ctDNA as an early pharmacodynamic biomarker of therapeutic response. Methods: Patients (pts) enrolled on phase 1 trials at the National Cancer Centre Singapore were prospectively consented to collect serial blood samples at defined timepoints (pre-treatment, mid-cycle 1, end of cycle 1, and with subsequent imaging evaluations). ctDNA burden was quantified using low pass whole genome sequencing (lpWGS) with both established (ichorCNA) and novel (cell-free DNA degradation) bioinformatic analyses of ctDNA burden, and targeted panel sequencing. Early kinetics of ctDNA burden prior to imaging evaluation was evaluated as response (reduction in ctDNA burden) versus progression (increase in ctDNA burden). Results: Between May 2021 to Oct 2022, 68 plasma samples (median 3, range 2-5) were collected from 22 pts. Median age was 56 years (range 37-77) and 73% were male. Pts were enrolled on trials of targeted therapy (55%), immunotherapy (40%) or combinations of both (5%). Common tumour types were lung (36%), colorectal (23%) and pancreatic (9%) cancer. Best response on phase 1 trial as per imaging and RECIST v1.1 was partial response (PR) in 9%, stable disease (SD) in 45%, progressive disease (PD) in 27% and not evaluable (NE; taken off trial early due to toxicity) in 18%. Quantification of ctDNA burden using cell-free DNA fragment length pattern demonstrated enhanced limit of detection in detecting baseline ctDNA compared to cell-free DNA copy number profile based ichorCNA method (100% vs 59%). For pts on targeted therapies, early kinetics of ctDNA burden indicated ctDNA response in 9/13 pts, with subsequent imaging showing PR (n=2), SD (n=2), PD (n=2) or NE (n=3). In the remaining 4/13 pts on targeted therapies with early ctDNA progression, subsequent imaging showed SD (n=3) or NE (n=1). For pts on immunotherapy, early ctDNA progression was seen in 8/10 pts, with subsequent imaging showing SD (n=5) and PD (n=3). Patient enrolment and sample analysis is ongoing to further validate these initial findings and updated data will be presented. Conclusions: Early ctDNA evaluation via lpWGS may represent an early pharmacodynamic biomarker of therapeutic response. ctDNA has the potential to provide early proof of concept, particularly for targeted therapies, that may influence subsequent decisions for dose determination.