Should Fuhrman Grading in Clear Cell Renal Cell Carcinomas Be Based on Nucleoli Only?
Susanne S Chan, Linda Lee, Joseph Chin, Jose A Gomez, Madeleine Moussa, George M Yousef, Manal Y Gabril. LHSC, London, ON, Canada; St. Michael's Hospital, Toronto, ON, Canada
Background: Fuhrman grading has been shown to be second only to stage in prognostic significance for clear cell renal cell carcinomas (ccRCC). One major limitation of this system, however, is the variability in applying the grading criteria (including nuclear size, nuclear shape and nucleolar prominence) amongst pathologists. The use of nucleolar prominence as the only criteria for Fuhrman grading is under debate. In this multi-institutional study, we compare the interobserver variability and prognostic significance of using nucleolar prominence only vs. using all criteria for Furhman grading.
Design: 77 radical nephrectomies with a diagnosis of organ-confined ccRCC (1990-1998) were included in this study. All slides from all cases were re-reviewed and independently graded by three pathologists at different institutions. Each pathologist graded every ccRCC twice, once using all Fuhrman grading criteria (nuclear size, nuclear shape, nucleolar prominence) and then again based on nucleolar prominence only. Clinical data from 50 cases including length of follow-up, local recurrence, distant metastasis and death due to RCC was also collected. Data from three pathologists was used to determine the interobserver variability scores, and to determine the prognostic significance of each grading system.
Results: Interobserver Variability: The mean kappa values ranged from 0.42 to 0.66 using all Furhman grading criteria vs. 0.46 to 0.92 using nucleolar grade only between pathologists.
The mean kappa value when the two Fuhrman grading methods were applied by the same pathologist ranged from 0.24 to 0.39.
14 out of 50 cases of ccRCC developed metastasis.
|All Criteria (n=50)||Nucleoli Only (n=50)||p-Value|
|Grade 1&2||1/15 (7%)||1/5 (20%)||(p>0.999)|
|Grade 3&4||13/35 (37%)||13/45 (29%)||(p>0.999)|