close
close

Does AI-assisted colonoscopy help find advanced neoplasia?

Does AI-assisted colonoscopy help find advanced neoplasia?

A new meta-analysis confirms that the use of computer-aided detection (CADe) systems during colonoscopy finds more polyps and adenomas than conventional colonoscopy, but the effect on the detection of advanced colorectal neoplasia (ACN) remains unclear.

The research team, who focused on advanced neoplasia due to their clinical importance, found a small increase in the ACN detection rate with CADe, but no difference in the number of ACNs detected per colonoscopy.

For ACN, “the ones we really care about, the findings were conflicting,” Dennis Shung, MD, MHS, PhD, of the Yale School of Medicine, New Haven, Connecticut, said in an interview.

There was a “small positive signal” indicating the potential of AI to detect advanced neoplasia. “However, we can’t say for sure that it will help you detect advanced colonic neoplasia,” Shung said.

Jeremy Glissen Brown, MD, MSc, a gastroenterologist at Duke Health, Durham, North Carolina, who was not involved in the study, said in an interview that it is “one of the most comprehensive systematic reviews and meta-analyses to date.” , involving both parallel and tandem randomized clinical trials (RCTs) of CADe in colonoscopy.”

“The results are generally consistent with previous RCTs and meta-analyses and demonstrate an improvement in key quality metrics, primarily an increase in the number of adenomas per colonoscopy (APC), an increase in the adenoma detection rate (ADR) and a decrease in number of missed adenomas. (AMR),” Glissen Brown noted.

The analysis was published online on October 21 Annals of Internal Medicine.

Bigger, more accurate analysis

Previous meta-analyses of AI-assisted colonoscopy included up to 33 RCTs. In their updated meta-analysis, Shung and colleagues included 44 RCTs with 36,201 cases.

“The large sample size in this study allowed us to more closely examine the efficacy of CADe in the diagnosis of clinically relevant colonic lesions, which was not feasible in previous RCTs and reviews given the smaller sample size,” they wrote.

For polyp detection, CADe-enhanced colonoscopy outperformed conventional colonoscopy in terms of mean number of polyps detected per colonoscopy (1.59 vs. 1.27; incidence rate difference (IRD), 0.35) and polyp detection rate (54% vs. 46.5% ; rate ratio (RR), 1.21).

The same applied to the detection of adenomas. CADe-enhanced colonoscopy had higher mean APC (0.98 vs. 0.78; IRD, 0.22) and ADR (44.7% vs. 36.7%; RR, 1.21), associated with lower AMR (16.1% vs. 35.3%; RR, 0.47).

The mean ACN per colonoscopy was similar with and without CADe improvement (0.16 vs. 0.15; IRD, 0.01), but there was a small increase in the ACN detection rate (12.7% vs. 11. 5%; RR, 1.16).

Results from a subgroup analysis suggest a reduced benefit of CADe in patients with positive fecal immunochemical test results, “which may indicate an attenuated benefit for using CADe systems for regular screening practices,” the research team wrote.

A sensitivity analysis of overall adenoma detection by baseline adverse event showed that the benefit of CADe systems increased among providers with lower adverse event rates.

The use of CADe systems led to resection of almost two additional non-neoplastic polyps per ten colonoscopies and a “marginal” increase in waiting time (0.53 minutes), which may have “limited clinical significance,” the authors noted.

There were no apparent differences in performance between the different CADe systems used in the included studies.

All studies were rated as “very concerning” due to common bias. Other limitations include the heterogeneity of the study, the lack of blinding between conventional and CADe-assisted colonoscopies, and unexplained confounding factors.

Ready for PrimeTime?

Is the routine adoption of AI-assisted colonoscopy ready for prime time? Glissen Brown thinks so, with some caveats.

“We are at a pivotal point in researching CADe for routine use in colonoscopy. CADe has been ready for prime time in the United States since at least 2021, and CADe’s study has made the field of gastroenterology a clinical leader in the number of high-quality randomized trials of AI interventions,” said Glissen Brown. .

However, outside the clinical trial setting, questions about successful deployment and implementation remain, he said.

“These include, but are in no way limited to, ways to optimize AI-human interaction to create a successful partnership between AI providers, issues of reimbursement and costs, and issues of ethical development and deployment of AI,” said Glissen Brown.

“We must also continue to assess methods for estimating CADe use on the downstream outcomes that matter, such as the effects of CADe on reducing colon cancer rates, post-colonoscopy colorectal cancer (CRC) rates and the effect CADe could have. have CRC-related mortality. In addition, more studies on patient voice and patient preferences regarding AI use are urgently needed,” said Glissen Brown.

The American Gastroenterological Association has developed recommendations regarding the use of CADe systems during colonoscopy.

Doctors are invited to review the draft guideline and share feedback during the public comment period, which ends Oct. 28.

The research had no specific funding. Information for study authors is available with the original article. Glissen Brown is a consultant for Medtronic, Olympus and Odin Vision. He was also the lead author of one of the studies included in the meta-analysis.