Novel histological scoring system predicts remission in UC
Click Here to Manage Email Alerts
A histology-based artificial intelligence system exhibited the highest correlation with endoscopic activity and predicted histological remission in patients with ulcerative colitis, according to research.
“Histological remission is an emerging treatment target and is an important outcome in UC clinical trials due to its association with favorable outcomes. However, challenges remain on how to incorporate histology into clinical practice,” Xianyong Gui, MD, of the University of Washington School of Medicine, and colleagues wrote. “Recently, we conducted a prospective international multicenter study to develop the Paddington International virtual ChromoendoScopy ScOre (PICaSSO) endoscopic score. ... The PICaSSO endoscopic score had better correlation than Mayo Endoscopic Score and UC Endoscopic Index of Severity with multiple histological scores.”
Taking this research a step further, Gui and colleagues aimed to develop a simplified histological score that could reflect microscopic mucosal inflammation and healing, predict clinical outcomes, respond to therapy and be implemented into AI systems. Using the PICaSSO Histological Remission Index (PHRI), 614 biopsies from 307 patients with UC underwent endoscopic and histological evaluation.
According to study results, PHRI correlated “strongly” with endoscopic scores (P < .05). Assessing the correlation between various histopathological components and endoscopic scores, researchers noted the neutrophil infiltration in the lamina propria and epithelium showed the strongest correlation compared with other histological features (P < .05). Further, patients with a PHRI score greater than 0 vs. equal to 0 had more negative clinical outcomes at 12 months (48.65% vs. 13.91%).
In addition, the preliminary AI algorithm discerned active vs. quiescent UC with a sensitivity, specificity and accuracy of 78%, 91.7% and 86%, respectively.
“PHRI is a simple and reproducible histological index that correlates strongly with endoscopic activity and predicts clinical outcomes in UC. It is therefore ideally suited for adoption in clinical practice, as well as for consideration in clinical trials and central readouts, if further validated to fulfill requirements of U.S. FDA or European Medicines Agency requirements,” Gui and colleagues concluded. “Further studies are ongoing to validate the deep learning-based computer-aided classifier before it can be adopted in clinical practice.”
Collapse
Read more about
Click Here to Manage Email Alerts