Background:
Current methods to detect spread of gastro-esophageal cancer into the peritoneum are inadequate. Accurate staging of the peritoneum is critical as it guides patient management. We aim to improve staging by testing if tumour DNA can be detected in peritoneal lavage fluid collected during routine staging for gastro-esophageal cancer. First, we need to optimise the extraction of cell free peritoneal tumour DNA (ptDNA) from dilute large volume peritoneal lavage (PL) collected from patients.
Methods:
DNA was extracted from 10mL of cytology-positive PL using four DNA extraction kits, two column-based: QIAmp Circulating Nucleic Acid Kit (QCNA) and QIAamp DNA Blood Maxi Kit; and two magnetic bead-based: BioChain cfPure® Cell Free DNA Kit (cfPure) and MagMAX™ Cell-Free DNA Kit. The top two performing kits based on a criteria of yield, quality, efficiency and cost were then used to extract DNA from 50mL of PL from the same cytology-positive sample and 10mL PL from three cytology-positive and three cytology-negative samples.
Results:
The top two performing kits were QCNA and cfPure according to the criteria, specifically total DNA yield. In both kits, while 50mL yields a greater DNA amount, 10mL was sufficient for intended downstream studies, specifically methylation profiling. Overall, the QCNA kit had a greater total DNA yield than cfPure using both 10mL or 50mL input, and in both cytology-positive and negative samples. All DNA extractions were evaluated on the Agilent TapeStation 4150 and demonstrated distinct peaks representing low molecular weight DNA fragments (corresponding to cell free DNA) and larger fragment size.
Conclusions:
QCNA kit results in optimal total DNA yield independent of volume or cytology status, with 10mL being sufficient for our downstream purposes. Next, we will investigate which DNA size fraction is enriched for cell free ptDNA. This will be done by performing whole exome sequencing on DNA extracted from matched tumour sections and buffy coats to identify tumour specific somatic mutations suitable for detection by digital droplet PCR in PL DNA separated based on molecular weight. This will inform decision making on which kit is optimal to extract a further 200 patients’ samples.