Pathology should be questioned as the gold standard for diagnosis of colorectal lesions ≤3mm, especially when high-confidence optical evaluation identifies an adenoma, according to a study published in Gastroenterology.
Discordance between endoscopic and pathologic diagnoses for colorectal lesions is common. To evaluate the discrepancy between endoscopic and pathologic diagnoses, researchers evaluated consecutive colorectal lesions ≤3 mm that were diagnosed during optical evaluation as adenatomous (N=644) with high-confidence by a single experienced endoscopist with expertise in optical evaluation (DKR). Lesions of this size were chosen because they were perceived to have a substantial risk for failed pathologic identification. An image dataset was evaluated using an established real-time artificial intelligence (AI) clinical decision support solution (CDSS), which was blinded to the endoscopic and pathologic diagnoses.
Concordant pathologic diagnosis was found in 71.1% of cases (n=458). Discrepancies between endoscopic and pathologic diagnoses occurred in 28.9% of lesions (n=186). Pathologic diagnoses included were hyperplastic polyp (n=85), sessile serrated polyp (n=2), and normal mucosa (n=99). CDSS agreed with the endoscopic diagnosis in 89.6% of lesions (n=577) overall; in discordant endoscopic and pathologic diagnoses, CDSS agreed with the endoscopic diagnosis in 90.3% of lesions (n=168). CDSS agreed with the endoscopic diagnosis in 90.9% of lesions identified by pathology to be normal mucosa (n=90).
Limitations of this study include possible optical misclassification and erroneous resecting of adjacent normal tissue, though this is unlikely for these to have been key players because all exams were performed by an experienced endoscopist with expertise in optical evaluation. Further research is needed to confirm these findings and assess the impact on surveillance intervals, including photo-documentation of optical evaluation and appropriate tissue resection, assessment of specimen quality and tissue fragmentation, specimen resectioning, and assessing endoscopic and pathologic inter-rater reliability.
“This study is the first description of a potential future application of AI; the arbitration between endoscopist and pathologist when discordant diagnoses occur … It is therefore imperative that all endoscopists incorporate optical evaluation, coupled with high-quality photo documentation, into clinical practice,” the researchers concluded. “In the interim, endoscopists should consider a more conservative approach for deciding on the appropriate surveillance colonoscopy interval, when endoscopic and pathologic discrepancy is encountered.”
Disclosure: Several study authors declared affiliations with the pharmaceutical industry. Please see the original reference for a full list of authors’ disclosures.
Shahidi N, Rex DK, Kaltenbach T, Rastogi A, Ghalehjegh SH, Byrne MF. Use of endoscopic impression, artificial intelligence, and pathologist interpretation to resolve discrepancies from endoscopy and pathology analyses of diminutive colorectal polyps [published online December 18, 2019]. Gastroenterology. doi: 10.1053/j.gastro.2019.10.024