Abstract
Background/Aim: In 2012, the International Society of Urological Pathology (ISUP) recommended replacing Fuhrman with ISUP for grading renal cell carcinoma (RCC). Our aim was to report recurrence-free survival (RFS) and assess prognostic value of ISUP and Fuhrman for predicting recurrence using original pathology assessment and routine follow-up data. Patients and Methods: In this single-institution retrospective cohort study, 686 patients underwent a single session total or partial nephrectomy due to nonmetastatic RCC (nmRCC). Of those, 564 had tumors prospectively graded according to either ISUP or Fuhrman, which defined the cohorts. RFS was defined as the interval from surgery to local recurrence and/or metastasis. Differences in RFS were calculated with log rank test. Cox models adjusted for risk factors were used for predicting recurrence. Results: During a median follow-up of 36 months in the ISUP group (n=152), 11% developed recurrent disease. RFS was significantly lower for grade 4 compared to 1-3 (p<0.001), but non-significant between 1-3. Grade was the only significant predictor in multivariate analyses. During a median follow-up time of 50 months in the Fuhrman group (n=412), 16% developed recurrent disease. There was a significant difference in RFS between grades 2 and 3 (p=0.003) and between 3 and 4 (p<0.001), but non-significant between 1 and 2 (p=0.063). Grade, positive surgical margin, tumor size ≥4 cm, and pT were significant predictors of recurrence in multivariate analyses. Conclusion: ISUP grading alone is an accurate tool for predicting recurrence in patients with nmRCC.
Renal cell carcinomas (RCC) represent a highly heterogeneous group of solid cancers (1, 2). The European Association of Urology (EAU) recommends total nephrectomy (TN) or partial nephrectomy (PN) for nonmetastatic RCC (nmRCC) (3). After surgery, the rate of recurrence ranges from 5 to 30% depending on several factors, such as the tumor type, histological characteristics, and disease stage (1, 4-7). The wide range of recurrence rates illustrates that risk profiling is essential for individualized follow-up and personalized counseling.
Fuhrman grading of RCCs was the gold standard for assessing tumor grade for almost 30 years and is implemented in most clinical nomograms for predicting recurrence (8-11). However, at a consensus meeting in 2012, the International Society of Urological Pathology (ISUP) recommended replacing the Fuhrman grading system with a new grading system, ISUP 1-4, for grading clear cell carcinoma and papillary carcinoma. The consensus meeting recommended not to include chromophobe RCC (chrRCC) in the new system (2, 12-17).
It is suggested that ISUP grading is superior to Fuhrman grading in predicting recurrence (18-21). However, all studies have been retrospective and were based on reviewed specimens originally graded according to the Fuhrman system. This may have caused an over or underestimation of the predictive value of the ISUP system in clinical practice. The current EAU guidelines emphasize the need for external validation with prospectively applied ISUP grading and clinical follow-up data.
Thus, the aim of this study was to report the recurrence-free survival (RFS) and assess the prognostic value of the ISUP system and the Fuhrman system in two independent cohorts for predicting recurrence using the original pathological assessment and routine clinical follow-up data.
Patients and Methods
This retrospective quality control study was based on prospectively registered data. We included all patients who underwent open or laparoscopic TN and PN due to nmRCC between January 2006 and November 2018 at Oslo University Hospital. A waiver for informed consent was issued, and the study was approved by the Local Data Protection Officer (18/20654).
Inclusion and exclusion criteria. All patients who underwent a single session surgery with open or laparoscopic TN and PN for nmRCC between 2006 and 2018 with at least one follow-up radiological examination were included.
Exclusion criteria were benign tumors, non-RCC malignancies, multiple resections at different sessions and/or hereditary malignant renal tumors (i.e., von Hippel Lindau, Tuberous Sclerosis), previous history of RCC, metastasis at the time of surgery or absent/unavailable follow-up data for any reason.
Patient data. Sex, age at surgery and type of surgery were retrieved from the patient records. The histological examination defined the tumor type, tumor size, pathological tumor (pT) stage, pathological lymph node (pN) stage and nuclear grade (Fuhrman grade before 2014 and ISUP grade after 2014). Tumor necrosis was reported for specimens assessed according to the ISUP system. Cancer extending into the cut surface in the surgical specimen was defined as a positive surgical margin (PSM). In cases of bilateral tumors, we reported the tumor with the highest grade. If the tumor grade was equal on both sides, we reported the largest tumor. All specimens were prospectively examined by two staff consulting pathologists (double reading) who were subspecialized in uropathology. No specimens were reviewed.
Follow-up. All patients were regularly followed up with routine radiology (i.e., thoracic and abdominal computed tomography (CT), chest X-ray, abdominal ultrasound) and clinical examination based on their risk profile according to Leibovich score/model (10, 22).
Endpoints. The rates of recurrent disease and 5-year RFS for the patients with tumor graded according to either ISUP or Fuhrman. The areas under the curves (AUC) of ISUP grade and Fuhrman grade for predicting recurrent disease. The risk of recurrent disease according to sex, age, tumor grade, surgical margin status, tumor necrosis, tumor size, pN and pT for the patients with tumor graded according to either ISUP or Fuhrman.
Statistical analysis. Recurrent disease was defined as local recurrence and/or metastasis based on original radiology reports and/or histological reports from salvage surgery and/or clinical data from the patient records. RFS was defined as the time (months) from surgery to the time to local recurrence and/or metastasis (whichever came first). Data were obtained at the time of diagnosis of local recurrence and/or metastases or the latest clinical or radiological examination. We calculated the median time to recurrent disease with the interquartile range (IQR). Results were stratified according to the ISUP and Fuhrman grades.
The 5-year overall RFS and 95% confidence interval (CI) are reported. We used Kaplan–Meier curves to illustrate differences in RFS curves at ISUP grades 1-4 and Fuhrman grades 1-4. The log-rank test was used to assess any differences in survival curves. Cox regression analyses were used to analyze the risk of recurrence (univariate and multivariate analysis). Results are expressed as hazard ratios (HR) with 95% CIs. We used logistic regression with a binary outcome to calculate the probability of recurrence and the corresponding AUC using all significant covariates. A p-value <0.05 was considered statistically significant, and IBM SPSS Statistics for Windows, Version 28.0. (Armonk, NY, USA) and MedCalc for Windows, version 20.009 (MedCalc Software, Ostend, Belgium) were used for the statistical analyses.
Results
From 2006 to 2018, 686 patients with nmRCC underwent a single session PN or TN. Of these, we included 642 (94%) patients with complete surgical data and at least one radiological follow-up (Figure 1). The ISUP and Fuhrman groups included 152 patients and 412 patients, respectively. In 78 patients, the tumor was ungradable. The patient and tumor characteristics of all 642 patients are shown in Table I.
Flowchart illustrating included and excluded patients.
Patient and tumor characteristics for all patients.
During a median follow-up time of 45 months for all patients (IQR=20-60 months), 14% (89/642, 95%CI=11-17) developed recurrent disease. Of the 89 patients with recurrent disease, 67 (75%) had metastasis only, 9 (10%) had local recurrence only, and 13 (15%) had both local recurrence and metastasis. The 5-year RFS across all groups was 85% (95% CI=82-89), and the median time to recurrence was 17 months (IQR=4-37 months).
Recurrence-free survival in the ISUP group. During a median follow-up time of 36 months (IQR=21-54 months) in the ISUP group, 11% (16/152, 95%CI=6-17) developed recurrent disease (Table II). The median time to recurrence was 8 months (IQR=3-22 months), and the 5-year overall RFS was 90% (95%CI=85-95). The ISUP grade was the only significant predictor for recurrence in multivariate analyses (HR=8.4) (Table III). The AUC of ISUP for predicting recurrent disease was 0.895 (95%CI=0.83-0.94) (Figure 2). There were no cases of recurrent disease among those with ISUP grade 1, and those with ISUP grade 4 had a 50 times higher risk of recurrence than those with ISUP grades 2-3 (HR=51, 95%CI=16-169). Those with ISUP grade 4 demonstrated significantly lower RFS than those with ISUP grades 1,2, and 3 (p<0.001) (Figure 3). There was no significant difference in RFS between ISUP grades 1,2, and 3.
Surgical procedures and outcome for all patients.
Cox regression analyses for assessment of risk of recurrent disease in the two cohorts.
Receiver operating characteristic curves illustrating the diagnostic accuracy of the International Society of Urological Pathology (ISUP) and Fuhrman.
Kaplan–Meier curves showing the differences in recurrence-free survival according to the International Society of Urological Pathology (ISUP) grades 1-4.
Recurrence-free survival in the Fuhrman group. During a median follow-up time of 50 months (IQR=18-61 months) in the Fuhrman group, 16% (67/412, 95%CI=13-21) developed recurrent disease (Table II). The median time to recurrence was 20 months (IQR=5-39 months) and the 5-year RFS was 93% (95%CI=91-96). In multivariate analyses, the Fuhrman grade, PSM, tumor size ≥4cm, and pT were all significant predictors of recurrence (Table III). The AUC of Fuhrman and the AUC of Fuhrman+PSM+size≥4cm+pT was 0.720 (95%CI=0.67-0.76) and 0.870 (95%CI=0.83-0.90), respectively (Figure 2). There was a significant difference in RFS between those with Fuhrman grades 2 and 3 (p=0.003) and between grades 3 and 4 (p<0.001), but there was non-significant difference between grades 1 and 2 (p=0.063) (Figure 4).
Kaplan–Meier curves showing the differences in recurrence-free survival according to Fuhrman grades 1-4.
Discussion
To the best of our knowledge, this is the first study reporting on recurrence rates according to prospectively applied ISUP grading and real-life clinical data. Previous studies were based on retrospectively reviewed and reclassified pathological specimens, which may have skewed the results. We believe it is essential to publish real-life clinical data using original reports and follow-up data. In this way, the unbiased performance of routine tumor grading is reported.
Our study demonstrated that the ISUP grade was the only significant predictor for recurrence in multivariate analysis in patients with tumors graded according to the ISUP system. Haas et al. showed that there is a strong association between disease-free survival and OS (23). Therefore, identifying risk factors for recurrence is essential to provide better treatment and optimal follow-up.
In our study, the 5-year RFS across all groups was 85%, which is similar to other studies showing 5-year RFS rates of 80-88% (4, 19, 24). We did not see any cases of recurrence in patients with ISUP grade 1, while most patients with ISUP grade 4 developed recurrence. We are not aware of other studies specifically reporting on RFS according to prospectively applied ISUP grading, but two retrospective studies demonstrate excellent cancer-specific survival in cases of ISUP grades 1-2, intermediate survival in cases of ISUP grade 3 and poor survival in cases of ISUP grade 4 (18, 25).
We investigated several suspected risk factors for recurrence, such as tumor grade, tumor necrosis, tumor size, PSM, pN, and pT. In the ISUP group, only the tumor grade remained significant in multivariate analyses, and the AUC of ISUP was 0.895. In contrast, Kim et al. failed to show that the ISUP system was an independent predictor of recurrence when ISUP grades 1-2 were compared to ISUP grades 3-4 (19). Dagher et al. demonstrated that both the ISUP grade and pT3 were significant risk factors for recurrent disease, but they did not include other risk factors for analysis (18). To the best of our knowledge, the predictive value of PSM and tumor size has not been described in studies using the ISUP grading system.
We did not detect an impact of tumor necrosis on RFS in our study, which contrasts with other studies (19, 20). Delahunt et al. recommended applying the presence or absence of microscopic tumor necrosis in the ISUP grading system due to an added value for predicting cancer-specific survival, especially for clear cell RCC (ccRCC) (25). Other studies recommend quantifying tumor necrosis using different percentage cut-off values (26, 27). For ccRCC, there is a consensus to report tumor necrosis as absent or present in percentage, but there is currently no consensus on how to implement these findings in the ISUP system (16). For papillary RCC (pRCC) and chromophobe RCC (chrRCC), there is no consensus on how to report tumor necrosis. In our study, tumor necrosis was not systematically quantified, and a more structured approach might improve its prognostic value.
In contrast to the ISUP group, recurrence occurred with all Fuhrman grades, which indicates that the ISUP grade provides better discrimination between patients at very low risk, and those at very high risk. In the Fuhrman group, tumor grade, size, PSM, and pT stage were all significant predictors of recurrence (Table III). This is in accordance with previous studies, and the established SSIGN prognostic tool (stage, size, grade, and necrosis) (28). In our study, a PSM was the strongest predictor of recurrence in the Fuhrman group, although some studies report that a PSM has a little or no impact on survival, whereas others report a predictive value in high-risk patients (7, 29-32). The overall rate of PSM in our study was 6%, while other studies report a PSM rates between 0.1 and 13.6%, depending on the patient selection and type of surgery (7, 32-34).
In our study, the AUC of ISUP was higher than the AUC of Fuhrman for predicting recurrence. Direct comparison cannot be done as the tests were not performed on the same cohort. Nevertheless, the two independent cohorts were similar and represent clinical practice from two different time periods at the same hospital (Table I).
The concordance indexes (c-indexes) used in other studies correspond to the AUCs used in this study. Other studies report c-indexes for the ISUP system between 0.634 and 0.863 and for the Fuhrman system between 0.700 and 0.864 (19, 21). Correa et al. reported c-indexes for eight different clinical risk nomograms used in clinical practice, of which the Krakiewicz system obtained the highest score of 0.86 (8). The median observation time of that study was 12 years, making it difficult to compare their results with our study.
Nomograms tend to be complicated and depend on several variables. Compared to significantly more complex models based on several postoperative variables, our study indicates that the ISUP grade alone may provide equally high predictive value. Whether updated risk nomograms may add to the prognostic value of the ISUP grading system remains to be determined.
Strengths and weaknesses. The main limitation of this study is the relatively short follow-up period and sample size in the ISUP group. However, this is to be expected as ISUP grading is a much newer system. All patients were followed according to their risk profile defined by the Leibovich score, which is based on the Fuhrman system. In our hospital, Fuhrman grades 1-4 have been substituted with ISUP grades 1-4 and all patients are followed accordingly. This may have influenced the frequency of follow-up and possibly the time to recurrence in the ISUP group, but it is uncertain how this may have affected our results.
A significant proportion of the patients had a tumor that was not graded according to either the ISUP system or the Fuhrman system. This group of patients was highly heterogeneous and detailed group reporting is not feasible due to the vastly different tumor biology and prognosis. A revision of the WHO classification of urogenital tumors was published in 2016 (35). Some of the ungradable tumors in our study could fit some of the new entities defined in this updated version and could possibly have been provided with an ISUP grade.
In this study, we can only provide RFS data rather than cancer-specific survival, as we do not have complete data on oncological treatment and cause of death.
Conclusion
In conclusion, this study shows that routine ISUP grading provides a simple and accurate tool for risk stratification in patients with nmRCC.
Footnotes
Authors’ Contributions
Kristina Flor Galtung: Conceptualization, methodology, data collection, data curation, statistical analysis and interpretation of results, writing, project managing. Peter Mæhre Lauritzen: Conceptualization, methodology, data collection, writing, reviewing, and editing, supervision. Eduard Baco: supervision, writing, reviewing, and editing. Rolf Eigil Berg: supervision, writing, reviewing, and editing. Anca Mihaela Naas: writing, reviewing, and editing. Erik Rud: Conceptualization, methodology, data collection, statistical analysis, and interpretation of results, writing and editing, supervision.
Conflicts of Interest
The Authors declare no conflicts of interest in relation to this study.
- Received April 7, 2022.
- Revision received April 25, 2022.
- Accepted April 26, 2022.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.