In adults aged 75 years or older, the risk for death from causes other than colorectal cancer (CRC) greatly exceeded the 10-year risk for diagnosis or CRC-specific death, regardless of previous adenoma diagnosis, a large study of US veterans found.

Reporting results in JAMA, researchers led by Samir Gupta, MD, MSCS, gastroenterologist at University of California San Diego Health, professor of medicine at University of California San Diego School of Medicine, and staff physician at the VA San Diego Healthcare System in San Diego, observed that for veterans having colonoscopy before age 75, the subsequent cumulative 10-year incidence of non-CRC death ranged from 46.9% to 48.4%. Those rates far exceeded the cumulative incidence of CRC (1.1%) and CRC death (0.5%), both in those with prior adenoma (0.7%) and without adenoma (0.4%).

Given such low yields, “patients and their clinicians may choose to prioritize healthcare for other health conditions over going for another colonoscopy,” Gupta told Medscape Medical News. “As gastroenterologists, we’re routinely faced with the challenging question of whether older adults with a prior history of polyps, who might have an increased risk for colon cancer, will benefit from exposure to repeat colonoscopy and whether the potential benefits outweigh the potential risks.”

The study aimed to see if estimating the risk for CRC in this population and comparing this risk to people without such a history (for whom current guidelines recommend against repeat colonoscopy) as well as to their competing mortality risk from other causes could inform decisions about repeat colonoscopy. The analysis also included a patient frailty component.

“Our hypothesis was that colorectal cancer risk in older adults who survived cancer-free until age 75 would be low, and that this would be dwarfed by the competing risk for non-CRC mortality,” Gupta said.

“We were surprised by the contrasts provided by including absolute risk estimates,” he added. “Although people with a prior history of polyps had a higher risk for colon cancer relative to people who had a prior normal colonoscopy, the absolute risks were very low.” The 5-year risk for dying from CRC was only about 0.2%, while the competing 5-year risk for dying from a non-CRC cause was high at about 22%.

Study Details

The retrospective analysis looked at 91,952 veterans (76% Caucasian) from January 2006 to December 2019, with a median age of 71 at their last colonoscopy and 98% men. Of those having a colonoscopy before age 75, 25,538 (27.8%) had adenoma and 66,414 (72.2%) did not.

At the 10-year follow-up, the cumulative incidence of CRC was 1.1% (95% CI, 0.8%-1.3%) in those with adenoma vs 0.7% (95% CI, 0.5%-0.8%) in those without adenoma (P < .001). Also at the 10-year follow-up, the cumulative incidence of CRC death was 0.5% (95% CI, 0.3%-0.7%) in those with adenoma vs 0.4% (95% CI, 0.3%-0.5%) in those without adenoma (P = .005).

Even for those with adenoma, the incidence of CRC death was substantially exceeded by the cumulative incidence of non-CRC death across all frailty levels, ranging from 34.2% among nonfrail individuals to 82.0% among severely frail individuals.

The findings expand previous research by offering longitudinal estimates of CRC risk and taking frailty into account — and by comparing these estimates with competing risks for non-CRC death. “We postulate that accounting for frailty and competing risk for non-CRC death is critically relevant to clinical practice because clinicians and patients engaged in decision-making need to put CRC risk in the context of a patient’s personal health status and associated risk for non-CRC death,” the authors wrote.

“With respect to risk for cancer, our data are among the first to quantify the cumulative risk,” Gupta said. “With respect to risk for non-CRC mortality and all-cause mortality, our data are congruent with what is already known. What is novel in our study is our ability to look at risks for multiple outcomes in the same study and compare the risks directly.”

Their incidence estimates are similar to those of two recent cross-sectional studies. For example, a UK study reported a low cumulative incidence of CRC of 393 per 100,000 person-years of follow-up for adults at least 75 years of age after the removal of intermediate-risk adenomas. And a meta-analysis of individuals with prior polyp removal reported that 2.1% of adults aged 70 years or older vs 1.4% of those aged 50-70 years were diagnosed with CRC at follow-up surveillance colonoscopy.

“The importance of a study like this is that the outcomes are tracked for a large cohort of patients, providing these important insights,” Rajesh Keswani, MD, interventional gastroenterologist at Northwestern Medicine in Chicago who was not involved in the study, told Medscape Medical News. “It’s difficult for a gastroenterologist to see these in clinical practice as we take care of patients at select moments in time, and we are not their primary care physicians.”

As to the importance of including a frailty component, “All clinicians understand that no age can capture the overall health of a patient. While a screening colonoscopy might be appropriate for an extremely healthy 77-year-old, it is almost certainly not for a very ill 74-year-old. Thus, considering patient frailty is as important as age,” Keswani said.

“As we have shifted the colonoscopy screening guidelines earlier, to age 45 for average-risk patients, we must simultaneously think about when it’s time to stop,” Keswani added. “Reserving colonoscopy after age 75 to select patients will certainly ease colonoscopy demand, especially in resource-limited settings.”

The bottom line, said Gupta, is that for patients who had a colonoscopy and survived cancer-free to age 75 — regardless of whether the most recent colonoscopy showed a precancerous polyp or not — the risk is very low and far lower than the risk for dying from another cause. “With this in mind, patients and their clinicians may choose to prioritize healthcare for other health conditions over going for another colonoscopy.”

Gupta added that these findings may update clinical guidelines with the recommendation that repeat colonoscopy should not be routinely performed in this patient population: “The decision should be based on shared decision-making that takes into account the very low risk for colon cancer.”

This study was supported by the Veterans Health Administration, the Department of Veterans Affairs Health Services Research and Development, the National Cancer Institute, the National Institute on Aging, and the Kaiser Permanente Delivery Science and Applied Research Physician Researcher Program. Support for data from the Veterans Administration and the Centers for Medicare & Medicaid Services was provided by the Department of Veterans Affairs Office of Research and Development.

Gupta reported receiving personal fees from Guardant Health, Geneoscopy, Universal Diagnostics, and Freenome. Keswani disclosed consulting for Covidien PLC.