Detection of mutations and copy number alterations in circulating DNA from pancreatic neuroendocrine tumor patients

Abstract

Background: Chromogranin A, the most frequently used circulating biomarker for diagnosis and follow-up of pancreatic neuroendocrine tumor (PNET) patients, has several limitations. Research in other tumor types has indicated the biomarker potential of circulating tumor DNA (ctDNA). However, ctDNA remains unexplored in PNETs. In this study, we aimed to detect and profile ctDNA in plasma of PNET patients. Methods: Tumor tissue, perioperative blood samples and clinicopathological data were prospectively collected from 10 PNET patients undergoing surgery for their primary tumor at the Antwerp University Hospital. An additional blood sample was collected from one case during follow-up, when patient had disease progression. Whole exome sequencing was performed on tumor and germline DNA to identify somatic variants and copy number alterations (CNAs). For every patient a somatic single nucleotide variant (SNV) was selected and a digital droplet PCR assay was developed to detect this SNV in DNA isolated from plasma. Shallow whole genome sequencing (sWGS) was performed on plasma DNA to identify CNAs. Results: In two patients, the somatic SNV could be detected in the perioperative plasma sample at variant allele fractions (VAFs) of 19% and 21%. Interestingly, both patients had metastatic disease and succumbed within two years after surgery, while the other eight patients presented with localized disease and are currently disease-free. The follow- up plasma sample of one of the positive cases showed an increase in VAF to 57%. Next, sWGS was performed on ctDNA-positive plasma to detect CNAs. A significant correlation (p<0.01) was found between CNAs in primary tumor and CNAs in perioperative plasma sample. The CNA profile of the follow-up sample showed increased genetic instability. Conclusions: We provide evidence for the presence of ctDNA in patients with a metastatic PNET. Non-metastatic cases were ctDNA-negative. An increase in VAF and genetic instability were found in the follow-up sample of one of the metastatic cases, suggesting potential for ctDNA as follow-up marker. Furthermore, CNAs in primary tumor and plasma sample were significantly correlated, proposing ctDNA as an alternative for molecular profiling of tissue in metastatic patients.