Real-world application of liquid biopsy in gastrointestinal tumors: Experience from a comprehensive cancer center.

Abstract

e14546 Background: Determination of circulating tumoral DNA (ctDNA) in plasma has recently emerged as an alternative to genomic testing in solid biopsy. Although different clínical trials have suggested that ctDNA alterations can tailor patient's treatment, its routinary utility in clinical practice is still controversial. Methods: A total of 74 consecutive plasma samples from 63 patients (pts) have been collected from October 2017 to December 2018 in our center and analyzed using two different next-generation sequencing (NGS) platforms: OncoTRACE (OT) and Guardant360 (G360). Main epidemiological, pathological and clinical data have been retrospectively extracted from medical records and correlation with ctDNA alterations have been performed using Chi-Square test. Application of liquid biopsy results has been decided after discussion in molecular tumor board and physician’s opinion. Results: OT and G360 were performed in 63 and 11 plasma samples respectively. Most part of the samples were from pancreatic (34; 45.9%) and colorrectal (CRC) (28; 25.7%) adenocarcinoma pts although biliary tract, gastric and oesophageal carcinoma pts were also included (4 each; 5.4%). Plasma samples were analyzed after progression to ≥two systemic therapies in 48.6%. At least one pathogenic alteration of ctDNA was detected in 44.6% of the samples. No correlation has been observed between presence of alteration in ctDNA and tumor origin, surgery of the primary, number and location of metastases and CEA and CA19.9. ctDNA alteration could be targetable with approved or experimental therapies in 14 pts (20.9%; 6 pts OT and 8 pts G360): 9 pts with CRC ( BRAF, EGFR, RAS, PI3KCA, PTEN), 2pts with pancreatic ( IDH1, BRCA1) and 1 pt with gastric, oesophageal and biliary tract carcinoma ( PTEN, EZH2, IDH1). However, these results have been applied in 4 pts (6% of all pts, 28.6% with targeteable ctDNA alterations): 3 pts were treatment-naïve and 1 pt had progressive disease to 2 systemic therapies. Most part of the patients with targetable alterations who did not receive treatment was due to clinical deterioration. Conclusions: Detection of alterations in ctDNA using NGS platforms is feasible in gastrointestinal tumor patients and can help physicians to decide treatment, especially in CRC. However, many patients could not be treated according to ctDNA as plasma samples were collected in late stages of the disease.