Colorectal cancer screening with flexible sigmoidoscopy can reduce mortality: A discussion with Dr. Andrew Swartz

A re-analysis of data on colorectal cancer screening (CRC) has found that flexible sigmoidoscopy reduces all-cause mortality with an absolute risk reduction of 3 deaths per 1,000 persons. This is clinically meaningful when compared with the results of other screening methods, according to the authors of the re-analysis. Their findings were published in the journal Annals of Internal Medicine.

In 2016, the US Preventive Services Task Force (USPSTF) released an updated evidence report on colorectal cancer screening. That report concluded that no CRC modalities had been shown to reduce all-cause mortality.

But there was a discrepancy in the USPSTF report, according to Andrew W. Swartz, MD, and colleagues, who authored the re-analysis article. This observation led them to double-check the data. When they did, they found that CRC screening with flexible sigmoidoscopy “reduces all-cause mortality with an absolute risk reduction that is clinically important relative to other preventive interventions.”

They calculated that the absolute risk reduction for flexible sigmoidoscopy was 3 deaths per 1,000 persons (age 50 to 74 years) invited to screening after 11.5 years of follow-up.

“If the primary goal of screening is to reduce the risk for death, then the evidence supporting flexible sigmoidoscopy is substantially stronger than that of other screening methods,” Dr. Swartz and colleagues concluded. “We believe that colorectal cancer screening guidelines warrant reassessment to incorporate this evidence.”

In this interview, lead author Dr. Swartz of Yukon-Kuskokwim Delta Regional Hospital in Bethel, AK, describes how he recognized the problem in the USPSTF data, what it could mean for patients, and whether flexible sigmoidoscopy should become the preferred CRC screening method.

MDLinx: Why did you suspect that there was an anomaly in the USPSTF evidence report for colorectal cancer screening?

Dr. Swartz: The 2016 USPSTF meta-analysis used results from two age cohorts of the Norwegian Colorectal Cancer Prevention (NORCCAP) trial. When the NORCCAP study was originally published in JAMA in 2014, I had noticed a discrepancy in the reported statistics, which prompted me to write a letter to the editor. In response, the authors explained that their reported number of deaths did not correlate with their summary statistic because NORCCAP had really been two separate trials and, despite the fact that they reported only the total number of deaths, two cohorts could not be analyzed as one group.

When the 2016 USPSTF evidence report for CRC was published, I naturally looked to see how they had handled the two cohorts of NORCCAP, and I immediately noticed that they had used the aggregated data. I recalculated the meta-analysis, and I was quite surprised to see that when the NORCCAP cohorts are included as two separate trials, flexible sigmoidoscopy reduces all-cause mortality.

MDLinx: What is the implication of your finding in the context of CRC screening?

Dr. Swartz: The importance of the all-cause mortality reduction is not that we expect colorectal cancer screening to reduce deaths from causes other than colorectal cancer, but rather that screening should not increase deaths from other causes.

Cancer screening trials have long shown reductions in their target-cancer deaths, but this is frequently offset by increased deaths from other causes, leading to no change in overall deaths. This has led some evidence-based medicine adherents to question the validity of the disease-specific death reductions. It remains unclear if screening is causing extra off-target deaths (eg, surgical complications, infection, etc.) or if the cause-of-death determinations are biased, or both.

But just changing the attributed cause-of-death is not the goal; we are trying to reduce death. Flexible sigmoidoscopy is now the first cancer screening method (for any type of cancer) to achieve an overall death reduction compared with usual care in randomized trials. Across five trials, flexible sigmoidoscopy shows a consistent reduction in the risk of actually dying; the effect is not statistically significant in any individual trial, but it is statistically significant with a properly performed meta-analysis. This is a landmark outcome, but the USPSTF missed it because they used the aggregated NORCCAP results that confounded their meta-analysis.

MDLinx: Although this finding may be statistically significant, how is it clinically significant? That is, flexible sigmoidoscopy is uncommon these days, and the overwhelming majority of colorectal cancer screening is done by colonoscopy.

Dr. Swartz: First, the absolute risk reduction is far better than that of other cancer screening methods. Three deaths per 1,000 invited persons means 1 death prevented per every 333 persons invited. If you look at other methods of cancer screening, you must invite 1,000 to 5,000 persons to prevent a single death attributed to the target cancer, but with no change in the risk of actually dying.

Second, flexible sigmoidoscopy is much less expensive, can be performed by primary care physicians (thus, increasing access to screening), and is somewhat safer than colonoscopy. Also, we have no clinical trial results to support either colonoscopy or fecal immunochemical testing (FIT). Given that we now have very strong evidence that flexible sigmoidoscopy is the most effective method of cancer screening to date (for any type of cancer), we should strongly reconsider it to be our preferred screening method for colorectal cancer.

While it is true that flexible sigmoidoscopy has become uncommon in the US, it has not been discredited or become outdated, but rather just fallen out of favor. There is no reason why we could not return to it as the primary method of colorectal cancer screening.

MDLinx: Have you heard yet whether any medical organizations are reconsidering their colorectal cancer screening guidelines in light of your finding?

Dr. Swartz: We have not heard directly from any guideline authorities. However, the Vice Chairman of the USPSTF, Doug Owens, MD, commented on it in a Medscape Medical News article.

"The Task Force last looked at screening for colorectal cancer in 2016, so it is not yet in the process of updating this recommendation,” Dr. Owens told Medscape. “The Task Force looks forward to reviewing all relevant evidence when it updates its review in the future.”

Meanwhile, the US Multi-Society Task Force on Colorectal Cancer (which represents the American College of Gastroenterology, the American Gastroenterological Association, and the American Society for Gastrointestinal Endoscopy) just released its latest guideline in July 2017, which ranked flexible sigmoidoscopy as a “Tier 2” test while giving colonoscopy and FIT testing “Tier 1” recommendations.

Our findings definitely challenge those conclusions.

About Dr. Swartz: Andrew W. Swartz, MD, is trained in rural family medicine and provides comprehensive care (including colorectal cancer screening) at Yukon-Kuskokwim Delta Regional Hospital in Bethel, AK.