Diagnosing Barrett’s esophagus and associated dysplasia using the WATS3D three dimensional computer-assisted biopsy system resulted in significantly higher interobserver agreement compared with standard histopathology, according to recent study data.
“This study addresses a major clinical challenge in gastroenterology,” Prateek Sharma, MD, associate professor of medicine in the division of gastroenterology and hepatology at University of Kansas School of Medicine, said in a press release. “The high rates of interobserver variability among pathologists evaluating Barrett’s esophagus complicate diagnosis and risk stratification and can confound the gastroenterologist’s ability to make effective treatment decisions for the patient.”
The wide-area transepithelial sampling (WATS) procedure uses a minimally invasive brush biopsy technique to acquire wide-area tissue samples, which are then analyzed by a high-speed computer scan, the researchers wrote.
“These data suggest the WATS3D biopsy can help overcome the limitations of traditional histopathological assessment of esophageal precancer,” Prashanth R. Vennalaganti, MD, GI fellow, University of Kansas School of Medicine, said in the press release. “WATS3D uses a computer neural network that chooses the 200 most suspicious areas from the biopsy and projects the same cells on a computer monitor to every pathologist evaluating the sample. This may potentially explain the high interobserver agreement observed in the study.”
Aiming to determine interobserver agreement in the diagnosis of Barrett’s esophagus and associated dysplasia using the WATS3D (CDx Diagnostics) technique, Vennalaganti, Sharma and colleagues randomly selected 149 WATS slides with varying degrees of Barrett’s esophagus-associated dysplasia, which were obtained from the CDx Diagnostics repository. All samples were then graded by four masked pathologists (who were trained in analysis of WATS samples) as nondysplastic, low-grade dysplasia or high-grade dysplasia/esophageal adenocarcinoma.
The mean kappa value was 0.86 (95% CI, 0.75-0.97) for all three diagnoses combined, “nearly perfect” at 0.95 (95% CI, 0.88-0.99) for high-grade dysplasia/esophageal adenocarcinoma, substantial at 0.74 (95% CI, 0.61-0.85) for indeterminate for dysplasia/low-grade dysplasia, and 0.88 (95% CI, 0.81-0.94) for nondysplastic Barrett’s esophagus.
“One of the most meaningful findings in this study is the high interobserver agreement for the diagnosis of low-grade dysplasia, which has historically had the lowest rate of agreement even among expert pathologists,” Lauren Gerson, MD, MSc, associate clinical professor of medicine from University of California San Francisco, said in the press release. “Improved detection of low-grade dysplasia will allow more gastroenterologists to change their surveillance protocols or perform endoscopic interventions for this patient cohort.”
“In summary, the diagnosis of [Barrett’s esophagus] and associated dysplasia using WATS aided by computer analysis has high interobserver agreement and appears to be significantly higher compared with previously published data using histopathology specimens,” the researchers concluded. – by Adam Leitenberger
Disclosures: Gerson, Sharma and Vennalaganti report no relevant financial disclosures. Please see the full study for a list of all other researchers’ relevant financial disclosures.