Added Value With CAD, Mucosal Exposure Device for Pinpointing Adenomas

Use of a mucosal exposure device alone, or in combination with computer-aided detection (CADe), led to improvements in all adenoma detection parameters during standard colonoscopy, a researcher reported.

In a study of 942 patients, Endocuff Vision-assisted colonoscopy (EAC) alone (47%), CADe alone (55%), or CADe-EAC (59%) passed the aspirational adenoma detection rate (ADR) threshold of over 40% for high ADR detectors versus standard colonoscopy (41%) control (P<0.01), according to Krittaya Mekritthikrai, MD, of the King Chulalongkorn Memorial Hospital in Bangkok.

And the proximal ADR (pADRs) of CADe, EAC, and CADe-EAC, were significantly higher than control (30%, 35%, 37%, and 24%, respectively, P<0.05), she stated in a presentation at the Digestive Disease Week virtual meeting.

"Our study demonstrated that in high ADR detectors, using either Endocuff alone or in the combination with CADe could significantly improve all adenomas parameters," Mekritthikrai said. "However, using CADe can significantly improve the number of adenomas and proximal adenomas detected during colonoscopy."

Mekritthikrai and colleagues also found that the mean number of adenomas per colonoscopy (APC) and proximal APC (pAPC) per colonoscopy were all significantly higher than control:

• CADe: 1.05 for APC and 0.68 pAPC
• EAC: 1.19 and 0.75
• CADe plus EAC: 1.28 and 0.78
• Control: 0.69 and 0.36

The advancement of endoscopy technology had led to improved rates of adenoma detection over the last decade, and the researchers have suggested a mucosal device could aid in the detection of adenomas that remain hidden "behind the colonic fold." CADe is thought to increase the detection of "subtle adenoma" and those that are unrecognized by endoscopists, Mekritthikrai explained. For screening, both CADe and EAC work to improve adenoma detection, but comparative data is currently not available to support their individual use or combined use of the two techniques.

"The efficacy of combining these two techniques has never been studied," she added.

"The bottom line is that the use of aids such as CADe or artificial intelligence, as well as devices that allow more mucosa to be seen such as EAC, should help improve polyp and adenoma detection, all of which in turn will decrease the rate of colorectal cancer," said David Greenwald, MD, of Mount Sinai Hospital in New York City, who was not involved in this study. "Using any or both of these techniques is better than using neither, and we as colonoscopists should be employing assisting devices and new technology to help us detect more polyps."

Mekritthikrai and colleagues enrolled patients who underwent a screening colonoscopy from September 2020 to November 2021. They randomized 237 patients to CADe alone, 235 to EAC alone, 233 to CADe-EAC, and 237 to standard colonoscopy. All groups received a high-definition, white-light colonoscopy by trainees or attending endoscopists. Before insertion, the Endocuff-Vision (ARV 110, Olympus) was attached to the tip of the colonoscope, while CADe (CAD EYE, Fujifilm) came with a digital box and voice arm that was activated prior to insertion.

Overall, 60% of patients were female and the mean age was 61. Mean BMI was 24. The average Boston bowel preparation score was 8 and 12.5% of patients had a family history of colorectal cancer. About 14% of the total study population were smokers and about 14% had a positive fecal occult blood test. Procedures had an average withdrawal time of 7.8 minutes and a cecal intubation time of 5.7 minutes.

Between the EAC group and control, significant ADR and pADR increases were seen even after adjusting for multiple comparisons. Also, ADR and pADR increases were also seen between the CADe-EAC group and the control, but not when comparing the CADe group to the control.

Notably, adjusted detection rates for APC and pAPC were also all greater in the CADe (1.43, 1.70), EAC (1.85, 2.16), and CADe-EAC (1.94, 2.30) groups versus control.

Comment