Risk of post colonoscopy colorectal cancer following screening colonoscopy with low-risk or no adenomas: A population-based study

Post-colonoscopy colorectal cancer: A population-based cohort study of fecal occult blood test-positive colonoscopies

BACKGROUND

Data on post-colonoscopy colorectal cancers (PCCRCs) after fecal occult blood test (FOBT)-positive colonoscopies is scarce (guaiac-based (gFOBT) or fecal immunochemical test (FIT)).

AIMS

Evaluate the prevalence and characteristics of PCCRCs in the French gFOBT CRC screening program.

METHODS

Retrospective population-based cohort study of all gFOBT-positive colonoscopies performed among individuals aged 50-74 between 2003 and 2014 within the CRC screening program organized in the Haut-Rhin (Alsace, France). The main outcome was PCCRC-3y rate. Adenoma detection rates (ADRs) calculated on gFOBT-positive colonoscopies were compared to those calculated on FIT-positive colonoscopies performed by the same gastroenterologists.

RESULTS

Overall, 9106 gFOBT-positive colonoscopies performed by 36 gastroenterologists were included. Sixteen PCCRC-3y and 31 PCCRC-5y were diagnosed (68.8 % and 58.1 % were true interval PCCRCs respectively). The unadjusted PCCRC-3y rate was 2.4 % [95 % confidence interval (CI) 1.4 %-3.9 %]. The risk for PCCRC-5y was significantly higher when the gastroenterologist's ADR was <35 % compared to ≥35 % (HR 2.17 [95 %CI 1.19-3.93]). The mean absolute difference for ADR between gFOBT- and FIT-positive colonoscopies was 16.3 % in favor of FIT-positive colonoscopies.

CONCLUSION

PCCRC-3y prevalence was low, estimated at 2.4 %. We suggest that the minimum standard for ADR in gFOBT- and FIT-positive colonoscopies should be set at 35 % and 50 % to 55 % respectively, in the French screening program.

Faecal haemoglobin concentration predicts all-cause mortality

BACKGROUND

Population-based screening for colorectal cancer by a faecal immunochemical test (FIT) is recommended by the European Union. Detectable faecal haemoglobin can indicate colorectal neoplasia as well as other conditions. A positive FIT predicts an increased risk of death from colorectal cancer but might also predict an increased risk of all-cause mortality.

METHODS

A cohort of screening participants was followed using the Danish National Register of Causes of Death. Data were retrieved from the Danish Colorectal Cancer Screening Database supplemented with FIT concentrations. Colorectal cancer specific and all-cause mortality were compared between FIT concentration groups using multivariate cox proportional hazards regression models.

FINDINGS

In 444,910 Danes invited for the screening program, 25,234 (5·7%) died during a mean follow-up of 56·5 months. Colorectal cancer caused 1120 deaths. The risk of colorectal cancer death increased with the increasing FIT concentration. The hazard ratios ranged from 2·6 to 25·9 compared to individuals with FIT concentrations <4 μg hb/g faeces. Causes other than colorectal cancer caused 24,114 deaths. The risk of all-cause death increased with the increasing FIT concentration, with the hazard ratios ranging from 1·6 to 5·3 compared to individuals with FIT concentrations <4 μg hb/g faeces.

INTERPRETATION

The risk of colorectal cancer mortality increased with the increasing FIT concentrations even for FIT concentrations considered negative in all European screening programs. The risk of all-cause mortality was also increased for individuals with detectable faecal blood. For colorectal cancer specific mortality and all-cause mortality, the risk was increased at the FIT concentrations as low as 4-9 μg hb/g faeces.

FUNDING

The study was funded by the Odense University Hospital grants A3610 and A2359.