Abstract 1315: Identification of actionable targets in advanced cancer patients from circulating tumor DNA using clinical decision support platforms

Abstract

Abstract BACKGROUND: We are entering exciting times in precision oncology, with next-generation sequencing enabling the individualized screening of multiple patient- and tumor-specific genomic and transcriptomic changes. As new molecular information flows into the clinic, the pursuit of innovative targeted therapies and defining the actionable genome are at the forefront. The amount of molecular information, much of it of unknown clinical relevance, has rendered the interpretation of detected aberrations a daunting task for clinical geneticists and oncologists alike. In this regard, several clinical decision support platforms have emerged on the market and need to be validated before implementation in routine clinical care of advanced cancer patients. METHODS: Here, we present a retrospective, descriptive study in adult patients with solid tumors using plasma DNA for the identification of actionable targets. We generated genomic profiles of 48 selected patients with advanced breast (BC), colorectal (CRC) and non-small cell lung cancer (NSCLC; samples selected had tumor fraction under 5%) using shallow whole-genome sequencing in parallel to the AVENIO ctDNA Expanded Panel, which covers 77 genes spanning 192kb. Somatic copy number alterations (SCNAs) and somatic variants were annotated for pathogenicity and subsequently matched to therapies using the following clinical decision support platforms: NAVIFY Mutation Profiler (NMP, Roche), CureMatch BIONOV (CureMatch) and QCI Interpret (QCI-I, Qiagen). We assessed the various features of each tool, correlated genomic features with actionable targets, compared genomic characteristics between tumor entities and evaluated the alignment of treatment recommendations. RESULTS: Tumor fraction in plasma DNA was highest in patients with BC (24.95% [range 10.52-48.93]) compared to CRC (14.94% [range 4.07-54.69]) and NSCLC (4.46% [range 1.63-45.42]) patients, whereas fraction genome altered was highest in CRC (median FGA = 0.289). Of the total reported aberrations per patient, both CureMatch and NMP identified the highest median percent actionability in CRC patients, whereas NSCLC had the highest actionability according to QCI-I. The total number of matching clinical trials was highest in CRC patients across platforms. CureMatch, a pathway-based analysis, suggested ranked combination therapies to target the entire molecular profile. CONCLUSIONS: This study seeks to provide insight into alternative treatment options and available clinical trials which were possibly previously overlooked due to incomplete molecular knowledge of the patient's tumor. These analyses may prove extremely valuable for increasing oncologists' confidence in and efficiency of employing molecular cancer diagnostics in a routine setting, in turn increasing clinicians' ability to use genomics-guided personalized cancer therapy when deciding on treatments in the future. Citation Format: Samantha O. Perakis, Sabrina Weber, Ricarda Graf, Qing Zhou, Jakob M. Riedl, Nadia Dandachi, Marija Balic, Armin Gerger, Ed Schuuring, Harry J. Groen, Jochen B. Geigl, Michael R. Speicher, Ellen Heitzer. Identification of actionable targets in advanced cancer patients from circulating tumor DNA using clinical decision support platforms [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 1315.