Regular colonoscopies were linked to a lower risk of dying from colorectal cancer or any cause in patients with Lynch syndrome, according to a large national study in England. But they were not associated with a lower risk of developing colorectal cancer.
Regular colonoscopy with polypectomy is generally assumed to reduce colorectal cancer incidence, the researchers wrote, though evidence in Lynch syndrome has been mixed. In this analysis, patients who had colonoscopies an average of every two years or less were diagnosed with colorectal cancer more often, with higher rates of early-stage disease and no corresponding decrease in late-stage cancers.
“High-quality and timely colonoscopy remains a cornerstone of good quality care for people with Lynch syndrome,” one of the study authors, Kevin Monahan, PhD, FRCP, a consultant gastroenterologist who co-directs the St. Mark’s Centre for Familial Intestinal Cancer, London, told GI & Hepatology News. “However, more frequent colonoscopy may be less important than ensuring the colonoscopy is of high quality, as well as other holistic elements of lifelong care are managed effectively, including the prescription of aspirin to reduce long-term cancer risk and gynecological risk management.”
The study, published in Gut, included 4,732 patients with Lynch syndrome who had disease-causing changes in mismatch repair (MMR) genes and were enrolled in the English National Lynch Syndrome Registry. Researchers linked registry information with National Health Service records on hospital care, cancer, and deaths from 2010 through 2022. Patients who had been diagnosed with colorectal cancer before their Lynch syndrome diagnosis were excluded.
Researchers grouped patients based on how often they had surveillance colonoscopies. They calculated the average time between colonoscopies using each patient’s follow-up time and number of procedures. The main analysis considered colonoscopies every three years or less to be regular surveillance, while a separate analysis used a stricter cutoff of every two years or less. Patients with no recorded colonoscopies were included in the groups receiving less frequent surveillance.
Overall, 3,788 patients (80%) had at least one colonoscopy. About 64% had colonoscopies an average of every three years or less, and 33% had them every two years or less. The study included 1,856 patients with MSH2 variants (including 25 with EPCAM deletions), 1,315 with MLH1 variants, 1,011 with MSH6 variants, and 550 with PMS2 variants.
During follow-up, 334 patients were diagnosed with colorectal cancer. Cancer developed in 7.7% of patients who had colonoscopies an average of every three years or less, compared with 5.9% of those who had them less often. Overall, however, colorectal cancer risk did not differ significantly between the groups after accounting for age. More frequent surveillance also did not clearly reduce late-stage cancers in favor of earlier-stage disease. Among MLH1 carriers, patients under regular surveillance had a higher proportion of stage 1 cancers (P = .02), although stage-specific incidence rates did not differ significantly for any gene.
The picture changed under the stricter definition. Colorectal cancer was diagnosed in 10.0% of patients who had colonoscopies every two years or less, compared with 5.6% of those who had them less often. By age 50, the estimated risk of developing colorectal cancer was 33% with more frequent surveillance vs. 14% with less frequent surveillance. By age 74, the estimated risks were 63% and 37%, respectively. Similar patterns were seen across the individual MMR genes.
More frequent colonoscopies were linked to higher rates of stage 1 and 2 cancers in some age groups, but not to lower rates of stage 3 and 4 cancers. The researchers said several factors could help explain these findings. Surveillance may be failing to prevent cancer in some patients, they wrote, through a rapid adenoma-carcinoma sequence that produces interval cancers, through poor-quality colonoscopy, or through pathways of tumor development in Lynch syndrome that endoscopy cannot interrupt. Alternatively, more frequent colonoscopy may be identifying more early-stage cancers — through genuine stage shift, through overdiagnosis of cancers that would not otherwise have progressed, or as an artifact of limited follow-up time and differences in the reasons patients had colonoscopies. The biology of Lynch syndrome-related tumors, which varies by gene, may also contribute.
Two sensitivity analyses — one restricted to patients with at least one year of follow-up, the other excluding cancers diagnosed within 90 days of a first colonoscopy — produced similar but weaker results, with fewer statistically significant differences.
Regular surveillance was linked to a lower risk of death. Five of the 3,028 patients who had colonoscopies every three years or less died from colorectal cancer, compared with 17 of the 1,704 patients who had them less often. After accounting for sex, age, and socioeconomic deprivation, colonoscopies every three years or less were associated with an 86% lower risk of dying from colorectal cancer, although the estimate was based on a small number of deaths.
Deaths from any cause were also lower among patients who had more frequent colonoscopies. There were 68 deaths in the more frequently monitored group, compared with 80 in the less frequently monitored group, representing a 56% lower adjusted risk of death. However, deaths from causes other than colorectal cancer were also lower with regular surveillance, at 2.1% vs. 3.7%. The researchers attributed this to a possible healthy screenee effect, in which patients who followed surveillance recommendations were generally healthier or had other social, economic, or lifestyle advantages that contributed to better survival.
How closely patients followed surveillance recommendations also varied. Older patients were less likely to have colonoscopies at least every three years, while those living in less disadvantaged areas were more likely to do so. In the fully adjusted model, white patients were more likely than Asian patients to have regular surveillance.
The findings do not show that having colonoscopies more often prevents more cancers. Dr. Monahan said colonoscopy remains a cornerstone of care in Lynch syndrome, with quality and timeliness mattering more than interval alone.
Because this was an observational study, the researchers could not prove that surveillance caused the differences they found. Uncaptured differences between patients who did and did not attend surveillance could not be accounted for and are likely to have had a substantial effect on the results, they wrote. Because the surveillance interval was calculated as an average, it did not capture variation in the timing of individual colonoscopies, and patients who died or developed cancer early were classified as non-adherent. The researchers also could not always tell whether colonoscopies were done routinely or because patients had symptoms. Information on colonoscopy quality, test findings, family cancer history, and procedures performed privately or outside England was unavailable, and no adjustment was made for disruptions to colonoscopy activity during the COVID-19 pandemic.
Catherine Huntley reported Wellcome Trust fellowship support. Other authors reported funding from Cancer Research UK, 40tude Curing Colon Cancer, and the NIHR Cambridge Biomedical Research Centre. Dr. Monahan also reported serving as a medical advisor to Bowel Cancer UK, Lynch Syndrome UK, and NHS England.
Expert Insight
Dr. Monahan discussed the clinical implications of the study with GI & Hepatology News.
How should these findings influence the surveillance interval you recommend for patients with Lynch syndrome, particularly given the higher colorectal cancer (CRC) incidence observed with colonoscopy at intervals of two years or less?
Dr. Monahan: This is an important and clinically relevant question. Here we have demonstrated a higher CRC incidence, however also improved mortality in people undergoing more frequent surveillance. Although our data was retrospective and observational, it is a comprehensive national dataset, and it is consistent with other data from the Prospective Lynch Syndrome Database and others which suggests CRC incidence is not reduced with more frequent surveillance. We expect that a prospective evaluation of outcomes from NHS England’s Lynch Syndrome Bowel Cancer Screening Program launched in 2023 will provide more clarity about the role of interval and colonoscopic quality factors.
Surveillance every three years or less was associated with substantially lower CRC-specific mortality, despite no clear reduction in CRC incidence. What do you think explains this apparent disconnect, and how confident are you that the mortality benefit is causal?
Dr. Monahan: It is likely that attendance and adherence to colonoscopy appointments is higher in people who are more health vigilant, and these appointments provide an opportunity to address other health issues not directly related to CRC risk. This is why we think we observe reduced overall mortality. Although there may well be impact of colonoscopic surveillance on CRC mortality, our study was not designed to evaluate this question directly.
Is there anything else you’d like to say about this work?
Dr. Monahan: This project was only possible because of the support of patient advocates, and crucially the English National Lynch Syndrome Registry, to which over 15,000 people have now been ascertained. This important resource will provide many opportunities to understand the natural history of Lynch syndrome in a whole population setting.