Quantcast

Diagnostic problems in the subtyping of renal tumors encountered by five pathologists.

Research paper by Intan I Kümmerlin, Fiebo F ten Kate, Frank F Smedts, Thomas T Horn, Ferran F Algaba, Isabel I Trias, Jean J de la Rosette, M Pilar MP Laguna

Indexed on: 22 Oct '08Published on: 22 Oct '08Published in: Pathology - Research and Practice



Abstract

The diagnostic problems in the subtyping of renal tumors were evaluated by a panel of five pathologists studying a set of selected tumors. Five pathologists independently assessed a single hematoxylin-and-eosin (HE)-stained slide from 28 selected renal tumors. After this independent assessment, the pathologists reevaluated and discussed all discordant cases. Additional HE-stained sections and immunohistochemically (IHC) stained slides were available. The generalized kappa for interobserver agreement was calculated. After independent assessment of the HE-stained slides, the five pathologists unanimously reached an agreement in the decision between malignant and benign in 82% of the cases. Fifty percent of the cases were correctly subclassified. The overall generalized kappa value for the five pathologists was 0.320 (CI 95% 0.090-0.551), which is considered a moderate agreement. A 100% agreement was reached for all 28 cases after examination of more slides from different tumor areas and IHC-stained sections. An accurate histologic distinction between benign and malignant renal tumors is possible on one HE-stained section. Correct assignment of the subtype is difficult on one slide alone and relies on IHC-markers and additional slides. Tumors composed of an eosinophilic cell type and tumors with a papillary growth pattern were the major causes of an incorrect diagnosis on an HE-stained section alone.

More like this: