Gynecologic Pathologist Agreement on Presence and Quantification of Lymphovascular Invasion in Low-stage, Low-grade Endometrial Endometrioid Adenocarcinoma.

Abstract

Lymphovascular invasion (LVI) is a prognostic and treatment-driving parameter for low-stage, low-grade endometrial endometrioid adenocarcinoma (EEAC). There is lack of organizational recommendations/guideline consensus regarding the clinically significant cutoff and pathologist agreement on cutoff thresholds has not been extensively studied. Digitized unannotated tumor-containing slides from 52 low-grade/uterine corpus-confined EEAC of no specific molecular profile (NSMP) and mismatch repair-deficient (MMRd) molecular subgroups were reviewed by 6 gynecologic pathologists. The reviewers, blinded to the originally reported LVI status, independently evaluated for the presence of LVI and the maximum number of involved vessels in a single slide or slides, per case. To quantify inter-rater reliability, intraclass correlation coefficient (ICC) estimates and their 95% confidence intervals were calculated (ICC<0.5=poor, 0.5 to 0.75=moderate, and 0.75 to 0.9=good and >0.9=excellent). A total of 280 slides were evaluated. The overall inter-rater reliability for the presence of LVI was moderate (ICC=0.53) and was slightly higher in the MMRd (ICC=0.56) compared with NSMP (ICC=0.52) group. For specific quantification of LVI, the overall inter-rater reliability was also moderate (ICC=0.71) and higher in the MMRd (ICC=0.76) compared with NSMP (ICC=0.69) group. Perfect agreement (6/6 reviewers with same assessment) only occurred for designations of no LVI or ≥5 involved vessels. There was a moderate level of inter-rater reliability that was not significantly affected by shifting the cutoff of involved vessels (from 3 to 4 to 5). Since a specific cutoff has no apparent bearing on the most reproducible cutoff for pathologists, future efforts should be directed towards obtaining clinical consensus for reporting purposes.