International study group on rectal cancer regression grading: interobserver variability with commonly used regression grading systems.

The aim of this study was to ascertain the level of concordance among gastrointestinal pathologists for regression grading in rectal cancers treated with neoadjuvant chemoradiation. Seventeen gastrointestinal pathologists participated using the Mandard, Dworak, and modified rectal cancer regression grading systems to grade 10 representative slides that were selected from 10 cases of rectal cancer treated with long-course neoadjuvant chemoradiation. The slides were scanned with a whole-slide scanner generating dynamic digitized images. The results showed very little concordance across the 3 grading systems, with κ values of 0.28, 0.35, and 0.38 for the Mandard, Dworak, and modified rectal cancer regression grading systems, respectively. In only 1 of 10 study cases was there unanimous grading concordance using the modified rectal cancer regression grading system. It was felt that these systems lacked precision and clarity for reproducible, accurate regression grading. The study concluded that there was a need for a simple, reproducible regression grading system with clear criteria, a cumulative or composite score taking into account all sections of the tumor bed that is sampled rather than the worst section (highest grade), and there should be a uniform method of sampling of these specimens.