Histologic classification of thymic epithelial tumors: Comparison of established classification schemes

The object of our multicenter retrospective study was to compare the new histologic World Health Organization (WHO) classification and the classical histologic Bernatz classification in terms of interobserver agreement and prognostic importance. The influence of coexisting diseases was also analyzed...

Full description

Saved in:
Bibliographic Details
Published inInternational journal of cancer Vol. 98; no. 6; pp. 900 - 906
Main Authors Rieker, Ralf J., Hoegel, Josef, Morresi‐Hauf, Alicia, Hofmann, Walter J., Blaeker, Hendrik, Penzel, Roland, Otto, Herwart F.
Format Journal Article
LanguageEnglish
Published New York Wiley Subscription Services, Inc., A Wiley Company 20.04.2002
Wiley-Liss
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:The object of our multicenter retrospective study was to compare the new histologic World Health Organization (WHO) classification and the classical histologic Bernatz classification in terms of interobserver agreement and prognostic importance. The influence of coexisting diseases was also analyzed using the Charlson score. We evaluated 218 patients from 5 different hospitals who were treated between 1967 and 1998. The statistical methods of analysis included Kaplan‐Meier estimates of survival curves and the application of Cox proportional hazards models to identify sets of prognostic factors for survival. Interobserver agreement was assessed by kappa coefficients. For both WHO and Bernatz classifications, interobserver agreement was good (weighted kappa > 0.87). However, the subdiversification of the “bioactive” WHO subgroup (B1, B2, B3) resulted in an interobserver agreement of only 0.49 within this group. In multivariable models, both the WHO classification and the Bernatz classification including carcinomas showed similar prognostic capabilities. The B3 type in the WHO classification and the predominantly epithelial type in the Bernatz classification had an intermediate prognostic ranking in comparison with the carcinomas and with the other subgroups. For both classifications, further simplification and subclassification into 3 subgroups led to classes with good discriminative power in respect to survival. In addition, very good interobserver agreement was observed in the simplified classifications. Comorbidity, sex, age of the patient and lymphofollicular hyperplasia had no major influence on overall survival. Both classifications showed similar prognostic power. Interobserver agreement of the type B subgroups was only moderate. By simplification of the classifications, subgroups with distinct survival could be identified. © 2002 Wiley‐Liss, Inc.
Bibliography:Fax: +49‐6221‐56‐5251
ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ISSN:0020-7136
1097-0215
DOI:10.1002/ijc.10255