Fecal Recovery of Ingested Cellular DNA: Implications for Noninvasive Detection of Upper Gastrointestinal Neoplasms

Background: Stool DNA testing represents a potential noninvasive approach to detect upper gastrointestinal (UGI) neoplasms. However, little is known about fecal recovery efficiency of DNA exfoliated from UGI tumors. Aims: The purpose of this study was to establish a human ingestion model that quantitatively approximates daily cellular shedding from UGI neoplasms and to estimate fecal DNA marker recovery rates. Methods: Healthy volunteers (n = 10) ingested two scheduled doses of raw salmon, 0.3 and 30 g, simulating the mass exfoliated daily from 1 to 4.5 cm lesions. To approach a steady-state, each dose was ingested over three consecutive days in randomized order. Following defecation of an indicator dye ingested with test meals, stools were collected over 48 h. Ingested salmon DNA was captured from stools using probes targeting pathognomonic Salmonidae sequences (SlmII). Captured DNA was quantified using PCR primers to generate 178, 138, 88 and 55 bp amplicons. Results: SlmII sequences were recovered from all stools following salmon ingestion; recovery was proportional to amount ingested (p = 0.004). Fecal recovery of ingested salmon varied inversely with amplicon size targeted; mean recovery rates of SlmII were 0.49, 0.91, 3.63, and 7.31 copies per 100,000 copies ingested for 178, 134, 88, and 55 bp amplicons, respectively (p < 0.0001). Longer oro-anal transit was associated with reduced recovery. Conclusions: While recovery efficiencies are low, ingested cellular DNA simulating daily amounts shed from UGI tumors can readily be detected in stool. Assay of shorter-fragment analyte increases recovery. This ingestion model has potential value in studying the effects of perturbations relevant to the fecal recovery of DNA exfoliated from UGI tumors.