What is the best histopathological classification for celiac disease? Does it matter?

The rebuttal of Oberhuber’s sub-division of Marsh III type in celiac disease (1) and its subsequent correspondence in the past and present issues of GHFBB allow us to review the current thinking on the histopathological classification of celiac disease. In the Definition of Medicine, Avicenna argues: -“While some divide medicine into a theoretical and a practical [applied] science, others may assume that it is only theoretical because they see it as a pure science. But, in truth, every science has both a theoretical and a practical side”. Thus, if you are working in your area in a community hospital without a dedicated pathologist to celiac disease, you could use either the classification made by Corazza & Villanacci (2) or by Ensari (3). Both classifications are practical and have proven to be useful with good specificity and sensitivity. Both the Oberhuber and the Corazza & Villanacci classifications allow discriminating latent celiac disease from patients with normal mucosa. Their classification may help in identifying and treating patients at an early stage (4). It is crucial to state that several biopsy samples of the duodenum and one from the bulbus duodeni should be taken. The specimens should be properly oriented to be able to assess the correct histopathological features. These classifications were developed to avoid the morphometric techniques, which at present are not used in routine histological laboratories. Marsh used morphometry in his extensive and now classical studies of the small intestine in celiac disease (58). Judging by the citation index the two classifications (76 and 33 respectively) cannot compete with the 1283 citations of the Marsh article of 1992 and the by now (September 2015) 656 citations of the Oberhuber et al article of 1999 even considering the difference in years of the publications. The morphometric techniques are timeconsuming and unpractical when technical facilities are not available. Ensari wanted to avoid the subjective element of the subdivision of the Type III lesions which were introduced by both Rostami et al (9, 10) and Oberhuber et al (11). When comparing the interobserver reproducibility of the “Marsh-Oberhuber” classification versus the one used by Corazza & Villanacci it was shown that the overall mean kappa value was 0.35 (fair) for the “Marsh-Oberhuber” classification and 0.55 (moderate) for the Corazza & Villanacci classification (12). Based on a sophisticated comparison between scanning electron microscopy studies and histological assessment, Marsh et al found that the Oberhuber's revisions of Marsh III into three subcategories (IIIa, IIIb, IIIc) are misinterpretations of the histological appearances of flattened mucosae. Therefore, Marsh et al, concluded that EDITORIAL