Careers  |  Sign In  |  Register

Sensitivity of colorectal cancer biomarkers overestimated for screening

Reuters Health • The Doctor's Channel Daily Newscast

NEW YORK (Reuters Health) – The reported sensitivity of numerous blood-based markers proposed for the early detection of colorectal cancer (CRC) are probably higher than might be expected in screening settings, because the estimates come from clinical studies with a high proportion of advanced cases.

Those conclusions, reported in the American Journal of Gastroenterology online October 19, come from a study by Dr. Hermann Brenner and colleagues at the German Cancer Research Center, Heidelberg.

“In recent years, intensive efforts have been made to identify blood-based markers that may provide a promising alternative for non-invasive CRC screening,” the authors point out. They also note, however, that the stage distribution of CRC in the screening setting is typically shifted the earlier stages compared with that seen in clinicals setting, so sensitivity estimates derived from clinical settings have to be adjusted.

The current study aimed to do that “by combining reported stage-specific estimates of sensitivity with the expected stage distribution of CRC in the screening setting.”

The researchers identified 73 articles for this purpose, with 134 evaluations of 55 different protein, cytological, mRNA or DNA markers.

Nearly all the studies had been conducted in hospitals. “There was a strong increase in sensitivity by stage in most studies,” the team found, and there was a greater proportion of higher-stage disease than seen in a reference population based on the German screening colonoscopy program.

Adjusting for this difference, the estimated sensitivities of markers expected in the screening setting were lower than the reported sensitivities in 90% of cases. “The median absolute reduction in sensitivity after stage adjustment was 9.0% units, and the median relative reduction was 19.5%,” according to the report.

In fact, in most cases, sensitivities dropped below 50%.

“Ideally, the validity of screening tests should be evaluated in the screening setting,” Dr. Brenner and colleagues point out. “Where this approach cannot be realized, adjustment of sensitivity to the stage distribution expected in the screening setting should be considered to obtain realistic and comparable estimates of sensitivity,” they conclude.


Sensitivity Estimates of Blood-Based Tests for Colorectal Cancer Detection: Impact of Overrepresentation of Advanced Stage Disease

Am J Gastroenterol 2010.