How long does it take until the effects of endoscopic screening on colorectal cancer mortality are fully disclosed?: a Markov model study

Abstract

A recent randomized trial has suggested persisting protection from colorectal cancer (CRC) incidence and mortality of a single flexible sigmoidoscopy for up to 17 years and possibly beyond. We performed a simulation study to explore the time course and magnitude of protection provided by screening colonoscopy against CRC death over 25 years. Using data from the German national screening colonoscopy registry, a multistate Markov model was set up based on the adenoma-carcinoma pathway to estimate cumulative CRC mortality when different proportions of the population have a single screening colonoscopy at age 55, or two screening colonoscopies at ages 55 and 65. Cumulative CRC mortality continuously increased with age and reached 2.6 and 1.7% at age 80 in the absence of screening for men and women, respectively. A single colonoscopy at age 55, even with limited uptake, would lead to much lower cumulative mortality (0.7% for men and 0.5% for women at age 80 under 100% uptake). Relative mortality reduction continued to increase over more than 10 years and reached the maximum around 12-13 years after screening. Absolute risk reduction steadily increased throughout follow-up and more than half of the total risk reduction would occur between 15-25 years. A repeat colonoscopy 10 years later further enhanced the effects and cumulative mortality remained at 0.1-0.2% under 100% uptake. Even a single (once-only) screening colonoscopy has the potential to prevent most of CRC mortalities. Protective effects are expected to be long-lasting and to become fully manifest after more than two decades from screening.

Keywords: colonoscopy; colorectal cancer; mortality; screening; simulation.