AI-Enabled SKOUT Device Improves Colorectal Cancer Screening

Iterative Scopes is touting the study results for its automated polyp detection algorithm for colorectal cancer (CRC) screening, SKOUT.

Trial data show statistical superiority of overall adenoma detection compared to routine colonoscopy. The 27% relative improvement in adenomas per colonoscopy (APC) detection in a relatively average-risk patient population reflects one additional adenoma detected among every 4.5 patients screened and occurred without decreasing safety or workflow efficiency, important considerations for gastroenterologists performing the procedures. The results from the registration trial were also accepted for publication in Gastroenterology.
 
SKOUT is a computer-aided device (CADe) that uses artificial intelligence and advanced computer vision technology to recognize suspicious tissue and provide real-time feedback to gastroenterologists. The goal is to detect suspicious lesions early, which has been shown to decrease incidence of colorectal cancer (CRC).¹ CRC is the second leading cause of cancer deaths in the United States and Europe.^2,3
 
Iterative Scopes submitted data from the clinical trial to the U.S. Food and Drug Administration in a 510(k) filing for review in November 2021. The 510(k) clearance for SKOUT is pending, it is not currently for sale in the United States.
 
“The results from this study indicate that CADe polyp-detection devices can enhance the quality of colonoscopy and are an important addition in the endoscopy suite,” said Aasma Shaukat, M.D., MPH, the Robert M. and Mary H. Glickman professor of Medicine and Gastroenterology at NYU Grossman School of Medicine. “Improving detection of adenomas per colonoscopy without increasing the rate of removal of healthy tissue has the potential of making our colonoscopy practice even more effective in the prevention of colorectal neoplasia.”
 
Colonoscopies are routinely performed to screen for CRC with the goal of detecting precancerous lesions, known as adenomas.¹ Studies show that colonoscopies are highly operator dependent, and endoscopists miss up to 26% of adenomas.⁴
 
In the SKOUT registration study, 1359 randomized subjects were included in the primary analysis in two separate cohorts. One cohort underwent a colonoscopy procedure with the implementation of the SKOUT AI software, and the other group underwent a colonoscopy procedure without implementation of the SKOUT AI software. Twenty-two board-certified, experienced gastroenterologists from five academic medical and community centers in the United States participated in the trial.
 
Both primary study endpoints were met. Adenomas per colonoscopy increased significantly with the use of SKOUT compared to the control, and true histology rate (THR) did not significantly change between cohorts. APC is a quality metric of colonoscopy procedure efficacy and is defined as the number of detected adenomas in a procedure divided by the total number of screening colonoscopies.⁴ A recently published analysis of state cancer registries found that incidence of post-colonoscopy CRC within three years of baseline examination decreased among endoscopists with higher APC rates.⁵
 
THR is a measure of resection of non-neoplastic lesions. A decrease in THR may indicate a rise in the resection of histologically unimportant (healthy) tissue. In this study THR did not change significantly with the use of SKOUT compared to the control. These results were consistent in both the screening and surveillance populations.
 
The study resulted in additional findings demonstrating improved detection of clinically important lesions. The study met secondary endpoints showing that the use of SKOUT did not affect procedure withdrawal time nor total procedure time. This indicates that this technology may be able to be integrated smoothly into routine clinical practice. Full study results have been accepted by Gastroenterology and will be released in the near term.
 
Iterative Scopes applies artificial intelligence-based precision medicine for gastroenterology to help optimize clinical trials investigating inflammatory bowel disease (IBD) treatment. The technology is also designed to potentially enhance colorectal cancer screenings. Its proprietary artificial intelligence and computer vision technologies can potentially improve the accuracy and consistency of endoscopy readings. Iterative Scopes is initially applying these advances to impact polyp detection for colorectal cancer screenings and working to standardize disease severity characterization for inflammatory bowel disease. Longer term, the company plans to establish more meaningful endpoints for GI diseases, which may be better predictors of therapeutic response and disease outcomes. Spun out of MIT in 2017, the company is based in Cambridge, Mass.

References 
¹ Kahi CJ, Imperiale TF, Juliar BE, et al. Effect of screening colonoscopy on colorectal cancer incidence and mortality. Clin Gastroenterol Hepatol 2009;7:770-775.
² Dyba T, Randi G, Bray F, et al. The European cancer burden in 2020: Incidence and mortality estimates for 40 countries and 25 major cancers. European Journal of Cancer 2021; 157: 308-347.
³ Rebeca S, Kimberly M, Ann Godin S, Colorectal Cancer Statistic, 2020. CA Cancer J Clin 2020; 70:145-164.
⁴ Zhao S, Wang S, Pan P, et al. Magnitude, risk factors, and factors associated with adenoma miss rate of tandem colonoscopy: a systematic review and meta-analysis. Gastroenterology 2019;156:1661-1674.e11.
⁵ Anderson J, Hisey W, Robinson C, et al. Mean adenomas per colonoscopy is inversely associated with post colonoscopy colorectal cancers: data from the New Hampshire Colonoscopy Registry validating this quality measure. Paper presented at: American College of Gastroenterology 2021 Annual Scientific Meeting; October 22-27, 2021; Las Vegas, Nevada.