Risk Stratification for FIT-Based CRC Screening in Elderly

WASHINGTON — A risk-stratified upper age limit may be beneficial for colorectal cancer (CRC) screening among patients who are ages 75 and older, according to a study presented at the annual Digestive Disease Week^® (DDW).

In particular, interval CRC risk can vary substantially based on the fecal hemoglobin (f-Hb) concentration in the patient’s last fecal immunochemical test (FIT), as well as the number of prior screening rounds.

“Less is known about what happens after the upper age limit has been reached and individuals are not invited to participate in more screening rounds. This is important as life expectancy is increasing, and it is increasingly important to consider the most efficient way of screening the elderly,” said lead author Brenda van Stigt, a PhD candidate focused on cancer screening at Erasmus University Medical Center in Rotterdam, the Netherlands.

In the Netherlands, adults between ages 55 and 75 are invited to participate in stool-based CRC screening every 2 years. Based on a fecal immunochemical testing (FIT) threshold of 47 μg Hb/g, those who test positive are referred to colonoscopy, and those who test negative are invited to participate again after a 2-year period.

FIT can play a major role in risk stratification, Ms van Stigt noted, along with other factors that influence CRC risk, such as age, sex, and CRC screening history. Although this is documented for ages 55-75, she and colleagues wanted to know more about what happens after age 75.

Ms van Stigt and colleagues conducted a population-based study by analyzing Dutch national cancer registry data and FIT results around the final screening at age 75, looking at those who were diagnosed with CRC within 24 months of their last negative FIT. The researchers assessed interval CRC risk and cancer stage, accounting for sex, last f-Hb concentration, and the number of screening rounds.

Among 305,761 people with a complete 24-month follow-up after a negative FIT, 661 patients were diagnosed with interval CRC, indicating an overall interval CRC risk of 21.6 per 10,000 individuals with a negative FIT. There were no significant differences by sex.

However, there were differences by screening rounds, with those who had participated in three or four screening rounds having a lower risk than those who participated only once (HR, 0.49).

In addition, those with detectable f-Hb (> 0 μg Hb/g) in their last screening round had a much higher interval CRC risk (HR, 4.87), at 65.8 per 10,000 negative FITs, compared with 13.8 per 10,000 among those without detectable f-Hb. Interval CRC risk also increased over time for those with detectable f-Hb.

About 15% of the total population had detectable f-Hb, whereas 46% of those with interval CRC had detectable f-Hb, Ms van Stigt said, meaning that nearly half of patients who were diagnosed with interval CRC already had detectable f-Hb in their prior FIT.

In a survival analysis, there was no association between interval CRC risk and sex. However, those who participated in three or four screening rounds were half as likely to be diagnosed than those who participated once or twice, and those with detectable f-Hb were five times as likely to be diagnosed.

For late-stage CRC, there was no association with sex or the number of screening rounds. Detectable f-Hb was associated with not only a higher risk of interval CRC but also a late-stage diagnosis.

“These findings indicate that one uniform age to stop screening is suboptimal,” Ms van Stigt said. “Personalized screening strategies should, therefore, also ideally incorporate a risk-stratified age to stop screening.”

The US Preventive Services Task Force recommends that clinicians personalize screening for ages 76-85, accounting for overall health, prior screening history, and patient preferences.

“But we have no clear guidance on how to quantify or weigh these factors. This interesting study highlights how one of these factors (prior screening history) and fecal hemoglobin level (an emerging factor) are powerful stratifiers of subsequent colorectal cancer risk,” said Sameer D. Saini, MD, AGAF, director and research investigator at the VA Ann Arbor Healthcare System’s Center for Clinical Management Research. Dr Saini wasn’t involved with the study.

At the clinical level, Dr Saini said, sophisticated modeling is needed to understand the interaction with competing risks and identify the optimal screening strategies for patients at varying levels of cancer risk and life expectancy. Models could also help to quantify the population benefits and cost-effectiveness of personalized screening.

“Finally, it is important to note that in many health systems, access to quantitative FIT may be limited,” he said. “These data may be less informative if colonoscopy is the primary mode of screening.”

Ms van Stigt and Dr Saini reported no relevant disclosures.

This article originally appeared on MDedge.com, part of the Medscape Professional Network.