Comparison of Standardized and Nonstandardized Nuclear Grade of Renal Cell Carcinoma to Predict Outcome Among 2,042 Patients

Christine M. Lohse; Michael L. Blute, MD; Horst Zincke, MD, PhD; Amy L. Weaver, MS; John C. Cheville, MD

Disclosures

Am J Clin Pathol. 2002;118(6) 

In This Article

Discussion

The results of our study indicate that nuclear grades reviewed by a urologic pathologist using standardized nuclear and nucleolar features were better able to predict death due to RCC compared with nonstandardized grades taken from surgical pathology reports during the 1970-1998 period. The differences in predictive ability between the original and reviewed grades, as well as the tendency to undergrade, were apparent for clear cell, papillary, and chromophobe RCC, both univariately and after adjusting for the 1997 TNM stage. At least 35% of tumors in each of the 3 subtypes were upshifted by 1 or more grades on review. Among patients with clear cell and papillary tumors, the most common shift was from grade 2 to grade 3. There was a statistically significant difference in cancer-specific survival between patients with tumors that remained grade 2 on review and those that were upshifted to grade 3. This difference was not statistically significant among patients with chromophobe tumors (P = .08); however, comparisons of outcome among patients with this subtype were limited by the few deaths due to chromophobe RCC. Interestingly, the more common shift among patients with chromophobe tumors was from grade 1 to grade 2. Although 20% of chromophobe tumors originally were classified as grade 1, only 2% remained grade 1 on review.

Several grading systems for RCC have been proposed, all demonstrating an association with patient outcome.[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17] Yet, there is no consensus regarding which grading system should be used. The guidelines from the UICC and AJCC workshop on the diagnosis and prognosis of RCC state that a grading system should be based on standardized and reproducible criteria that recognize grade heterogeneity within a tumor. Such a system should be evaluated separately for each histologic subtype and should be assessed using a large sample of patients to examine the effect of grade on outcome after adjusting for other well-established prognostic features.[18]

Although there still is no agreement as to which grading system should be used, the most common system is the one proposed by Fuhrman et al.[3] The Fuhrman grade is based on nuclear size and shape and the prominence of nucleoli. We applied a grading system modeled after Fuhrman’s criteria, with a few differences. For example, a Fuhrman grade 2 tumor consists of larger nuclei (approximately 15 µm) that exhibit irregularities in outline and nucleoli when examined under high power (x400 magnification). In contrast, our reviewed grade 1 tumors had small, round nuclei with inconspicuous nucleoli visible only at x400 magnification, while grade 2 tumors contained round to slightly irregular nuclei with mildly enlarged nucleoli visible at x200. In the Fuhrman grading scheme, the grade assigned is the highest grade exhibited, even if only focal. However, Fuhrman et al[3] did not define the area of tumor described by "focal." The grade assigned to each tumor using the criteria in our analysis was the highest grade occupying at least 1 field at x400 magnification. We believe that pathologists at other institutions could easily apply the criteria in our study for the grading of RCC since the criteria are based on the presence of nucleoli at various magnifications. We are designing a multi-institutional study to assess the variability and reproducibility of this grading system.

We studied the association of reviewed nuclear grade with patient outcome separately for the 3 most common histologic subtypes of RCC, recognizing that these subtypes differ in their morphologic features and outcome.[20,21,22] In contrast, the study by Fuhrman et al,[3] as well as many studies that incorporate the Fuhrman grade, included a mix of subtypes of RCC.[1,2,4,5,6,7,8,9,10,11,12,14,16] We studied only patients treated with a radical nephrectomy between 1970 and 1998, while Fuhrman et al[3] included patients treated with nephrectomy, adjuvant chemotherapy, radiation, or no treatment at all, between 1961 and 1974. Also, we assessed the effect of nuclear grade on outcome using more than 2,000 patients, which permitted us to stratify our results by histologic subtype and evaluate nuclear grade in a multivariate setting. Fuhrman et al[3] studied the association of grade with outcome using 103 patients, which limited statistical comparisons to univariate assessments only. Last, we studied death due to RCC, while Fuhrman et al[3] studied death due to any cause. These issues may have contributed to the inconsistency in our results with those reported by Fuhrman et al.[3] We identified significant differences in death due to clear cell, papillary, and chromophobe RCC between patients with grade 3 tumors and patients with grade 4 tumors. In addition, we identified significant differences in outcome between patients with grade 3 and 4 tumors compared with those with grade 1 and 2 tumors. However, we found no significant difference in outcome between patients with grade 1 and patients with grade 2 tumors. Other studies[6,8] that have used the Fuhrman grading system have reported similar results, including a recent study by Ficarra et al.[17] In contrast, Fuhrman et al[3] did not identify a significant difference in outcome between patients with grade 2 and patients with grade 3 tumors.

The members attending the UICC and AJCC workshop on the diagnosis and prognosis of RCC recommended that each increase in grade correspond to a significant difference in patient outcome.[18] A limitation of the application of our grading system is that we did not identify a statistically significant difference in death due to RCC between patients with grade 1 and patients with grade 2 tumors. This potentially could be explained by differences in fixation times since an increase in the length of fixation in buffered formalin results in an increase in nuclear size and nucleolar prominence.[23] Then again, since our tumors were fixed and processed on the same day using a standardized protocol, fixation times would vary by hours, not days. We hypothesize that the nuclear changes apparent in grade 3 tumors are greater than one would expect from extended fixation. Nevertheless, we plan to evaluate the effect of tissue fixation on nuclear grade in a future study. Regardless of the fixation effects, our findings suggest that a 3-tiered grading system may be appropriate for RCC, as supported by other studies.[5,7,11,16] However, use of a 3-tiered grading system without strict adherence to standardized criteria could result in an increase in the frequency of grade 2 tumors and lessen the ability of nuclear grade to predict clinically important outcomes.

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as:

processing....