A comparative analysis of grading systems in renal adenocarcinoma.

While the histological grade of a renal cell carcinoma is of prognostic significance there is poor concordance amongst pathologists in the use of these grading systems. Many grading systems have been described, but none has achieved widespread acceptance. The objective of this study was to assess the degree of interobserver variation amongst four experienced pathologists in their use of four commonly applied grading systems. The pathologists reviewed a series of 88 cases of renal cell carcinoma. Grades were detailed on a proforma which consisted of a breakdown of each grading system. Cohen's kappa was calculated for each pair of observers for each system. The mean kappa scores for each system were compared using the Tukey honestly significant differences method. Mean kappa was highest for the grading system of Syrjanen and Hjelt and this grading system also had a higher mean kappa than two of the other systems tested. The most striking feature of the results was the degree to which the pathologists differed in their assessments. The grading system of Syrjanen and Hjelt was shown to be subject to less interobserver variability than other commonly used classifications and we are of the opinion that it should become the standard method.