522O - Mutation tracking in circulating tumour DNA (ctDNA) detects minimal residual disease (MRD) in patients with localized colorectal cancer (CRC) and identifies those at high risk of recurrence regardless of stage, lack of CDX2 expression and CMS subtype

Abstract

BackgroundAlthough staging provides prognostic information and guides therapeutic decisions, 30-50% of patients with localized CRC will relapse despite optimal primary treatment. CEA and imaging studies are insufficient to detect micrometastases at an early stage. The identification of prognostic markers beyond TNM is crucial to define high risk of relapse and to establish potential therapeutic strategies to optimize adjuvant treatment. Methods150 patients diagnosed with localized CRC were prospectively recruited from October 2015 to October 2017 at our institution. Clinicopathological features (stage, grade, vascular/perineural invasion, sidedness, MMR status and CDX2 expression) were collected. DNA and RNA extracted from FFPE samples were assessed with a custom 29-gene panel recurrently mutated in CRC (NGS) and a validated NanoCRCA assay (NanoString). ctDNA from plasma was tracked in serial samples to detect MRD (ddPCR). Plasma Interleukin-6 levels were measured (ELISA). Log-rank test, univariate and multivariate Cox regression analysis and ROC curves were used for statistics. ResultsKnown somatic mutations in tumour were found in 120 (80%) patients for mutation tracking ctDNA analysis. After a median follow-up of 24.7 months, 18 patients recurred. Postoperative CEA was not predictive of disease-free survival (P=0.229). ctDNA in serial plasma after surgery predicted metastatic relapse with a median lead time of 10 months over radiologic recurrences [HR 11.33; P=0.0001]. CMS subtypes were significantly associated with CDX2 expression (loss in CMS1, P=0.03), IL-6 levels (high in CMS1 and CMS4, P=0.002) and perineural invasion (present in CMS1 and CMS4, P=0.001). CMS1 and CMS4 subtypes were significantly associated with relapse (P=0.016). A multivariable analysis confirmed T stage, loss of CDX2 expression and mutation tracking ctDNA were significantly associated with recurrence (P=0.041, P=0.004, P<0.001, respectively). ConclusionsIn localized CRC, plasma ctDNA detects MRD in relapsing patients earlier than clinical methods and opens an opportunity for precision treatment. Legal entity responsible for the studyThe authors. FundingInstituto de Salud Carlos III (PI15/02180 and PI18/01909 to AC; PI18/01508 to TF). NT was supported by Rio Hortega contract CM15/00246 from the Instituto de Salud Carlos III. DR was supported by Joan Rotes contract 16/00040 from the Instituto de Salud Carlos III. TF was supported by Joan Rodes contract 17/00026 from the Instituto de Salud Carlos III. VG was supported by ESMO 2014 fellowship programme, and by Rio Hortega contract CM18/00241 from the Instituto de salud Carlos III. DisclosureA. Cervantes: Advisory / Consultancy, Speaker Bureau / Expert testimony, Research grant / Funding (institution): Merck Serono; Advisory / Consultancy, Speaker Bureau / Expert testimony, Research grant / Funding (institution): Roche; Advisory / Consultancy, Research grant / Funding (institution): Beigene; Advisory / Consultancy, Speaker Bureau / Expert testimony, Research grant / Funding (institution): Bayer; Advisory / Consultancy, Speaker Bureau / Expert testimony, Research grant / Funding (institution): Servier; Advisory / Consultancy, Research grant / Funding (institution): Lilly; Advisory / Consultancy, Research grant / Funding (institution): Novartis; Advisory / Consultancy, Research grant / Funding (institution): Takeda; Advisory / Consultancy, Research grant / Funding (institution): Astellas; Advisory / Consultancy: Pierre Fabre; Speaker Bureau / Expert testimony: Amgen; Speaker Bureau / Expert testimony: Foundation Medicine; Research grant / Funding (institution): MedImmune; Research grant / Funding (institution): BMS; Research grant / Funding (institution): AstraZeneca; Research grant / Funding (institution): MSD; Research grant / Funding (institution): Sierra Oncology. All other authors have declared no conflicts of interest.