Pattern-based classification of invasive endocervical adenocarcinoma, depth of invasion measurement and distinction from adenocarcinoma in situ: interobserver variation among gynecologic pathologists

• DOI: 10.1038/modpathol.2016.86
• Publication Date: 2016-05-13T13:51:32Z
• Authors: Carlos Parra-Herran; Monica Taljaard; Bojana Djordjevic; M Carolina Reyes; Lauren Schwartz; John K Schoolmeester; Ricardo R Lastra; Charles M Quick; Anna Laury; Golnar Rasty; Marisa R Nucci; Brooke E Howitt
• Keywords: Concordance; Carcinoma in situ; Kappa

A pattern-based classification for invasive endocervical adenocarcinoma has been proposed as predictive of the risk nodal metastases. We aimed to determine reproducibility such in context common diagnostic challenges: distinction between situ and depth invasion measurement. Nine gynecologic pathologists independently reviewed 96 cases (two slides per case). They diagnosed each case or carcinoma classifying latter following pattern (non-destructive), B (focally destructive) C (diffusely destructive). Depth invasion, when applicable, was measured (mm). Overall, diagnosis good (κ=0.65). Perfect agreement (9/9 reviewers) seen only 11 (11%), all destructively (10 1 B). In all, ≥5/9 reviewer concordance achieved 82/96 (85%). Distinction carcinoma, regardless pattern, showed slight (κ=0.37). Likewise, restricted versus poor (κ=0.23). non-destructive (in situ+pattern A) destructive (patterns B+C) significantly higher (κ=0.62). Estimation excellent (ICC=0.82). However, different measurements resulting differing FIGO stages were (from at least 79% cases). On basis interobserver agreement, is best diagnosing which carries a Agreement including A, poor. Given nil spread lesions, term 'endocervical with growth' can be considered difficult, after excluding invasion. measurement highly reproducible among pathologists; thus, approach complement, but should not replace, metric.