Assessment of the Utility of the Fuhrman Grading System for the Major Subtypes of Renal Cell Carcinoma

Abstract

Renal cell carcinoma (RCC) is the most prevalent form of malignant renal epithelial neoplasia and is ranked tenth in the world in terms of incidence of visceral carcinoma. Approximately one third of all RCCs will have already metastasized by the time symptoms become apparent and in addition to the unpredictable behaviour characteristic of these tumours, RCC remains expressly resistant to chemotherapy and radiation treatments. It is now recognized that there are several types of RCC, each with different histological and genetic features. Recent studies have indicated that each of the major subtypes of RCC (clear cell RCC, papillary RCC and chromophobe RCC) differ in their clinical behaviour and prognosis. Tumour stage at diagnosis is well recognized as the most important of the predictors of outcome for individual cases. Tumour grade has also been assessed as a prognostic indicator with varying success. Of the current grading systems in use, that of Fuhrman, Lasky and Limas is most widey utilized. The Fuhrman classification was first described in 1982 prior to the formal subclassification of RCC, and as such has not been validated for the major subtypes of renal cell carcinoma described in the Rochester/Heidelberg classification of 1997 and in the WHO classification of 2004. The purpose of this project is to determine the inter-relationship of each of the components of the Fuhrman grading, and to determine if the Fuhrman grading system or any of the individual components of the grading system are of prognostic significance for each of the three major subtypes of RCC. Cases of clear cell RCC and papillary RCC were retrieved from the files of the New South Wales Cancer Registry. Additionally cases of chromophobe RCC were retrieved from the files of the New South Wales Cancer Registry, the University of Verona (Italy) and Witten/Herdecke University (Germany). TNM stage, pT category, tumour diameter and patient follow-up were retrieved from file. Haematoxylin and eosin stained slides of tumour specimens of all three major renal cell carcinoma subtypes were retrieved and graded according to the Fuhrman grading system. Specimens were assigned a focal Fuhrman grade based on the area of tumour showing the highest grade of nuclear pleomorphism present in each case alongside a whole grade designated according to the most prevalent nuclear grade evident within the whole tumour. Two nucleolar grades were allocated to each specimen based on the criteria for nucleolar prominence as described in the Fuhrman system, and were assessed focally and for the whole tumour, as for Fuhrman grading described above. Nuclear size and shape were measured by nuclear morphometry. Computerized morphometric analysis was conducted using the Sigma Scan Pro5 image analysis software which examined the following nuclear features; area, compactness, perimeter, shape factor and major axis. Comparative univariate and multivariate analytical techniques were employed to investigate prognostic relevance in an attempt to improve prognostic criteria for patients diagnosed with RCC. In addition, the components of the Fuhrman grading system (nucleolar prominence, nuclear size and nuclear pleomorphism) were analysed for independent prognostic significance. Of the 90 papillary cases studied, 59 were type 1,31 were type 2 and when categorized according to TNM stage information, 33 cases were TNM stage 1, 26 TNM stage 2, 18 TNM stage 3 and 12 TNM stage 4. When ordered according to pT category, 14 cases were pT1a, 20 pT1b, 25 pT2, 15 pT3a, 4 pT3b, and 11 pT4. The morphometric range measurements were for nuclear area 27.63-116.39 μM2, major axis length 6.70-14.06 μM, nuclear perimeter 20.05-41.77 μM, shape factor 0.805-0.878 and compactness 14.33-15.66. Survival information showed focal Fuhrman grade and papillary RCC type to be significantly associated with each other, further univariate analysis of the various measures revealed that of the Fuhrman grade components, only focal nucleolar grade was associated with survival. Multivariate analysis showed that focal nucleolar grade and tumour diameter were independently associated with survival and that only TNM stage retained independent prognostic significance over other parameters. Eighty-seven chromophobe tumours were examined in this series, with significant associations detected between nucleolar focal grade, Fuhrman focal grade and the morphometric measures of size and shape. No such associations were found between nucleolar whole grade and Fuhrman whole grade, while morphometric range measurements were for nuclear area 26.14-100.74 μM2, major axis length 6.49-13.21 μM, nuclear perimeter 19.73-39.28 μM, shape factor 0.80-0.89 and compactness 14.26-15.84. Univariate analysis showed that only age and tumour size were significantly associated with survival, however neither the nucleolar or Fuhrman grades showed any significant association with survival. One hundred and twenty one cases of organ confined clear cell RCC were investigated in this study and the morphometric measurement ranges were for nuclear area 18.19-88.30 μM2, major axis length 5.36-13.66 μM, nuclear perimeter 16.26-40.68 μM. shape factor 0.73-0.88 and compactness 14.41-17.60. Significant associations were found for size parameters and focal and whole case nucleolar grade, while no associations were found between nucleolar grade and measures of nuclear shape. Univariate analysis showed that focal Fuhrman grades were significantly associated with survival. Nuclear area and nuclear major axis were found to be significantly associated with survival, while focal nucleolar grade was only weakly associated with survival. No association was found between survival and measures of nuclear shape. The overall conclusions from this study were that for papillary RCC, nucleolar grade but not Fuhrman grade is appropriate. Neither Fuhrman grade nor any of its components is appropriate for chromophobe RCC and that, while Fuhrman grading of clear cell RCC was of prognostic significance, the value of grading would he enhanced if it focused upon nuclear size and focal nucleolar grade.