REDUCTION OF INCIDENT COLORECTAL CANCER AFTER NEGATIVE COLONOSCOPY BY SUBSEQUENT FECAL IMMUNOCHEMICAL TESTING

Background: Sessile serrated adenoma/polyp (SSA/P) is the recently recognized precursor lesion for colorectal cancer (CRC) and contributes to approximately 20-30% of sporadic CRC.However, due to the largely unknown natural history, the current guidelines for colonoscopy surveillance of patients after removal of SSA/Ps are inconsistent and based on very weak supporting evidence.Methods: We evaluated the long-term risk of CRC incidence and mortality after removal of SSA/Ps in Sweden, by linking the ESPRESSO, a nationwide database of GI histopathology records, to the Cancer Register, Cause of Death Register, and Patient Register.SSA/Ps were identified based on an algorithm that combines morphology and topography codes and free text search, which has been validated through manual review of original pathology reports in a random sample of 160 patients from ESPRESSO.For each patient, we selected 5 controls who had undergone lower endoscopy with a normal biopsy, matched on age, sex, and year of biopsy (within 2 years).We calculated the hazard ratio (HR) and 95% confidence interval (CI) of CRC incidence and mortality by comparing SSA/ P patients to controls via Cox regression, after adjusting for income, education, sites, and number of prior endoscopies and clinic visits.Results: We identified 7,483 patients who had undergone removal of SSA/Ps from 1993 through 2016 (mean age: 61 years; women: 55%), and matched them with 37,415 normal controls.During a median follow-up of 5.1 years (maximum, 24.0), we documented 135 and 238 cases of incident CRC in the SSA/P and control groups, respectively.Patients with SSA/P removal had a higher incidence rate of CRC than control individuals (31.0 vs. 9.6 per 10,000; multivariable HR, 3.18, 95% CI, 2.51-4.03).The risk elevation did not appear to differ by cancer sublocation (P=0.39), with an HR of 3.50 (95% CI, 2.55-4.79)for proximal colon cancer, 2.39 (95% CI, 1.49-3.84)for distal colon cancer, and 3.16 (95% CI, 2.00-4.98)for rectal cancer.When stratified by age at baseline, a stronger association of SSA/P with CRC risk was observed for younger than older individuals (<50 years: HR, 5.54, 95% CI, 1. ≥75 years: 2.14, 95% CI,; P for interaction=0.001).For the mortality analysis, we documented 44 and 125 CRC deaths in the SSA/P and control groups, respectively.The mortality rate of CRC was significantly higher in the SSA/P group than the control group (10.0 vs. 5.0 per 10,000; HR, 1.87, 95% CI, 1.27-2.74).Conclusions: Individuals with SSA/Ps have an increased risk of CRC incidence and mortality.The risk elevation for CRC incidence does not significantly vary by cancer sublocation, but appears to be stronger for younger than older individuals.Our findings support existing guidelines recommending complete removal of SSA/Ps and close follow-up evaluation of patients with this lesion.