Colorectal cancer (CRC) screening is effective at reducing deaths from CRC, but a new analysis shows that one form of screening — flexible sigmoidoscopy — also reduces the risk for all-cause mortality.
This is at odds with the latest statement on CRC screening issued by the US Preventive Services Task Force (USPSTF).
In its most recent guidelines, published in June 2016, the Task Force found convincing evidence that CRC screening substantially reduces disease-related mortality, although it did not recommend any one screening approach over another, as reported by Medscape Medical News at the time.
But while screening was effective at reducing death from CRC, the same did not hold true for mortality from other causes. In its guidelines, the USPSTF stated that “to date, no method of screening for colorectal cancer has been shown to reduce all-cause mortality in any age group.”
However, the new analysis has reached the opposite conclusion.
A team of researchers, led by Andrew W. Swartz, MD, from the Yukon-Kuskokwim Delta Regional Hospital, Bethel, Alaska, reanalyzed the USPSTF evidence and found that screening with flexible sigmoidoscopy “reduces all-cause mortality with an absolute risk reduction that is clinically important relative to other preventive interventions.”
This analysis was published online August 22 in Annals of Internal Medicine.
“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,” the authors conclude. “We believe that colorectal cancer screening guidelines warrant reassessment to incorporate this evidence.”
However, the USPSTF told Medscape Medical News that it has no immediate plans for revisiting its recommendations.
“The Task Force last looked at screening for colorectal cancer in 2016, so it is not yet in the process of updating this recommendation,” said Doug Owens, MD, Task Force vice chair and general internist and associate director of the Center for Innovation to Implementation at the Veterans Affairs Palo Alto Health Care System, California.
“The Task Force looks forward to reviewing all relevant evidence when it updates its review in the future,” he told Medscape Medical News.
Analysis Confounded?
The conclusion reached by the USPSTF regarding death from all causes was based partly on a meta-analysis of four randomized trials that compared flexible sigmoidoscopy screening with no screening, note Dr Swartz and colleagues.
That meta-analysis aggregated results from the two age cohorts taken from the NORCCAP (Norwegian Colorectal Cancer Prevention) study, “as if these cohorts were a single trial,” they write.
The NORCCAP study consisted of two distinct trial cohorts because of a postscreening decision to expand the inclusion age to younger individuals. Thus, the two groups were randomly assigned separately.
They point out that the younger cohort, aged 50 to 54 years, had a lower rate of events and were randomly assigned to a different ratio of screened individuals vs controls. The younger group was randomly assigned with a screen–control ratio of 1:5.4, rather than the 1:3 ratio used in the older group.
“Therefore, the meta-analysis in the USPSTF evidence report may be confounded because the aggregated NORCCAP results were used,” they write.
All-Cause Death Reduced
In their own study, Dr Owens and colleagues created a Simpson paradox repeated meta-analysis of all-cause mortality outcomes for screening flexible sigmoidoscopy using the two NORCCAP age cohorts as individual trials.
When these cohorts were considered as two separate groups (instead of being aggregated together), the relative risk (RR) for all-cause mortality favored screening with flexible sigmoidoscopy.
The RR for all-cause mortality favored screening among the younger group of participants (RR, 0.96; 95% confidence interval [CI], 0.87 – 1.06) compared with the older cohort (RR, 0.98; 95% CI, 0.94 – 1.03).
For the combined summary estimate of both groups, the RR was 0.98 (95% CI, 0.94 – 1.02), but when they were put into a single group rather than combined meta-analytically as two separate groups, the RR for all-cause mortality was 1.07 (95% CI, 1.02 – 1.12), which does not favor screening.
When all flexible sigmoidoscopy trials were put in a meta-analysis, using the individual NORCCAP study cohorts, results showed that all-cause mortality was reduced (RR, 0.975 [95% CI, 0.959 – 0.992]; P = .004; I2 = 0%) at 11 to 12 years.
With use of the assumed risk for death among the overall US population aged 50 to 74 years, the absolute risk reduction extrapolated to 3.0 deaths per 1000 persons invited to screening (95% CI, 1.0 – 4.9) after a follow-up of 11.5 years.
One author is supported in part by a grant from the American Cancer Society, and another by grant from the National Institutes of Health, National Institute of General Medical Sciences.
Ann Intern Med. Published online August 22, 2017. Abstract
For more from Medscape Oncology, follow us on Twitter: @MedscapeOnc
Tidak ada komentar:
Posting Komentar