Copy number alterations in plasma cell-free DNA from metastatic gastroenteropancreatic neuroendocrine neoplasms.

Abstract

372 Background: Recent studies, including our proof-of-concept study, demonstrated the possibility to detect tumor-derived molecular alterations in cell-free DNA (cfDNA) from plasma of patients with a gastroenteropancreatic neuroendocrine neoplasm (GEP-NEN). More in-depth evaluation of the biomarker potential of cfDNA in GEP-NENs is warranted, since highly sensitive and specific blood-based biomarkers are an important unmet need for this tumor type. The aim of our study was to detect tumor-associated copy number alterations (CNAs) in cfDNA from GEP-NEN patients to molecularly characterize the tumor and to estimate tumor fraction, as a measure of tumor burden, and to evaluate changes in these parameters over time. Methods: Metastatic GEP-NEN patients were included within NETwerk, a multi-institutional network of nine hospitals in Belgium. Clinicopathological data were collected to correlate experimental and clinical findings. Plasma samples were collected from all patients and cfDNA was extracted and subjected to shallow whole-genome sequencing (WGS). Detection of CNAs and estimation of tumor fraction, based on the sequencing data, were performed using the R-based tool ichorCNA. Results: In total, 80 samples of 29 metastatic GEP-NEN patients were analyzed using shallow WGS. All patients had a well-differentiated GEP-NEN of Grade 1/2 and primary sites were pancreas (N = 15), small intestine (N = 9), colon (N = 1), caecum (N = 1), ileocaecal valve (N = 1), pylorus (N = 1) and unknown (N = 1). Median number of samples per patient was two, with a median time between first and last sampling of six months. In 25 cfDNA samples from nine patients (31%), CNAs with tumor fractions higher than 3% could be detected. The primary tumor site of all CNA-positive patients was pancreas, corresponding to 60% of included pancreatic NEN (PNEN) patients. Six of the CNA-positive patients were included at initiation of everolimus treatment. The detected CNA patterns were similar to the copy number profiles of PNENs described in literature, e.g. whole-chromosome gains of chromosomes 5, 7, 9, 12, 13, 14, 19 and 20. CNA profiles were relatively stable over time, although in two patients new alterations did arise. Tumor fractions changed over time, which could be linked to changes in tumor burden, tumor progression and treatment response according to RECIST1.1 criteria and will be further examined, including in additional samples that are being collected. Conclusions: Cell-free DNA of metastatic GEP-NEN patients contains CNAs that correspond to CNA profiles seen in tumor tissue samples. CNAs can be used to quantify the tumor fraction in cfDNA over time, which will be linked to tumor progression in our ongoing study, particularly in the promising subgroup of PNEN patients.