No abstract

Gaps in colonoscopy skills among endoscopists, primarily due to experience, have been identified, and solutions are critically needed. Hence, the development of a real-time robust detection system for colorectal neoplasms is considered to significantly reduce the risk of missed lesions during colonoscopy. Here, we develop an artificial intelligence (AI) system that automatically detects early signs of colorectal cancer during colonoscopy; the AI system shows the sensitivity and specificity are 97.3% (95% confidence interval [CI] = 95.9%–98.4%) and 99.0% (95% CI = 98.6%–99.2%), respectively, and the area under the curve is 0.975 (95% CI = 0.964–0.986) in the validation set. Moreover, the sensitivities are 98.0% (95% CI = 96.6%–98.8%) in the polypoid subgroup and 93.7% (95% CI = 87.6%–96.9%) in the non-polypoid subgroup; To accelerate the detection, tensor metrics in the trained model was decomposed, and the system can predict cancerous regions 21.9 ms/image on average. These findings suggest that the system is sufficient to support endoscopists in the high detection against non-polypoid lesions, which are frequently missed by optical colonoscopy. This AI system can alert endoscopists in real-time to avoid missing abnormalities such as non-polypoid polyps during colonoscopy, improving the early detection of this disease.

No abstract

Endoscopic submucosal dissection (ESD) is known as a curative treatment for colorectal superficial neoplasms. There is however a need for more long-term clinical data to establish the full advantages of colorectal ESD regarding very low recurrence rates. The aim of this retrospective study was to determine long-term clinical outcomes of colorectal ESD. A total of 423 lesions treated by ESD for colorectal adenoma/dysplasia or adenocarcinoma between 1998 and 2008 at a single high volume referral center were included. We conducted a retrospective survey on patients with follow-up and obtained complete 1-, 3-, and 5-year outcome data for 358 (85 %), 292 (69 %), and 209 (49 %) lesions, respectively. Curative resection was defined when the pathological specimen had carcinoma-free resection margins, irrespective of piecemeal or en bloc resection, without submucosal deep invasion (≥ 1000 µm), lymphovascular involvement, or a poorly differentiated adenocarcinoma component.After a median 4.9 years of follow-up, the 3-year overall cumulative endoscopic recurrence rate and cancerous recurrence rate were 2.9 % (95 % confidence interval [95 %CI] 1.2 - 4.7) and 1.1 % (0 - 2.1), respectively. The 5-year overall cumulative endoscopic recurrence and cancerous recurrence rates were 3.8 % (1.7 - 5.9) and 1.6 % (0.1 - 3.0), respectively. In 361 lesions eligible for endoscopic follow-up, the 3-year endoscopic recurrence and cancerous recurrence rates were 2.4 % (0.8 - 4.1) and 0.4 % (0 - 1.4), respectively. Multivariate analysis revealed that piecemeal resection and submucosal deep tumor invasion were associated with recurrence. The current study demonstrated favorable long-term clinical outcomes of colorectal ESD when en bloc curative resection is achieved.

To elucidate the role of GNAS mutations in colorectal tumourigenesis, we performed a mutation analysis in a total of 234 colorectal tumours, including adenomas, serrated lesions and adenocarcinomas. Activating GNAS mutations were found in 20 of the 24 villous adenomas (83%) but were absent in all the other tumours, except for one tubulovillous adenoma (3%) and two adenocarcinomas (3%). KRAS and BRAF mutations were always mutually exclusive. KRAS mutations were frequent in villous (67%) and tubulovillous (60%) adenomas but were rare or absent in tubular adenomas (6%) and serrated lesions, including hyperplastic polyps, sessile serrated polyps/sessile serrated lesions and traditional serrated adenomas (0-9%). BRAF mutations were found in four villous adenomas (17%) and in the large majority of serrated lesions (81-92%), but were absent in tubular and tubulovillous adenomas. Seventeen villous adenomas (71%) harboured GNAS mutations concomitantly with KRAS or BRAF mutations. Immunohistochemically, all the villous adenomas retained mismatch repair protein expression, suggesting that they are microsatellite-stable. The current study showed that the presence of activating GNAS mutations, in association with KRAS or BRAF mutations, is a characteristic genetic feature of colorectal villous adenoma.