S252 Biopsy of Non-Tumor Sites After Biopsy of a Colorectal Cancer Is Associated With Metachronous Cancers: A Case-Control Study

Abstract

Introduction: A recent proof-of-concept study demonstrated the plausibility of iatrogenic tumor seeding via colonoscopy as a cause of subsequent metachronous colorectal cancer (CRC). We aimed to evaluate the association of colonoscopy-related biopsy of non-tumor sites following tumor biopsy and subsequent risk of metachronous CRC. Methods: We performed a matched case-control study of adult patients with an initial CRC diagnosed by colonoscopy between January 2006 and June 2018 who underwent curative resection in a large, community-based integrated healthcare setting. Cases were individuals with a second primary CRC diagnosed between 6 months and 4 years after the initial CRC (metachronous CRC). Each case was matched with up to five controls by age, sex, and race/ethnicity; controls were without a second CRC diagnosis. The primary exposure was biopsy at the site of metachronous CRC (or corresponding site in controls) after biopsy of the initial CRC during index colonoscopy, which was ascertained through chart review while blinded to case status. The association was evaluated using conditional logistic regression, and sensitivity analyses performed to assess for potential confounders. Results: During the study period, 14,119 patients were diagnosed with an initial CRC by biopsy during colonoscopy and 106 patients were subsequently diagnosed with a second CRC. Of these, 49 were metachronous CRC cases, of which 47 were successfully matched to 222 controls. After exclusions for inflammatory bowel disease and hereditary CRC syndrome, 30 cases and 148 controls were included in the analytic sample (Figure 1). Biopsy of non-tumor sites after biopsy of the initial CRC was associated with a >3-fold increased risk of metachronous CRC compared to no endoscopic manipulation (Table 1, odds ratio (OR) 3.53, 95% confidence interval (CI) 1.06-11.78). In sensitivity analyses, excluding patients with incomplete polyp removal at the reference site (1 case and 1 control) and then adjusting for bowel preparation adequacy or extent of examination did not substantially change the point estimates but widened CIs (OR with 95% CI 3.54 [0.93, 13.47] and 3.62 [0.96, 13.66], respectively). Conclusion: Biopsy of non-tumor sites after CRC biopsy was associated with a >3-fold increased risk of metachronous CRC. These findings suggest biopsy of non-tumor sites following tumor biopsy may be a risk factor for iatrogenic tumor seeding, though we cannot exclude alternate etiologies such as incomplete polypectomy.Figure 1.: Cohort Development DiagramTable 1.: Odds of Having Biopsy After Tumor Biopsy at the Reference Site* in Cases and Controls | Footnote: *Site of metachronous cancer in cases, and the corresponding site in matched controls