European Urology

International Bladder Cancer Nomogram Consortium
J Clin Oncol 2006;24:3967–72
Experts’ summary:
This article describes a nomogram for patients with bladder cancer who have undergone radical cystectomy and pelvic lymph node dissection. The data were comprised of 9064 patients from 12 institutions worldwide. Institutional data spanned many years; however, pathologic staging was reformatted to the 1997 American Joint Committee on Cancer (AJCC) staging and histologic grading information was converted to a consensus system consisting of a binary system of low versus high grade to provide the most reliable information for input into the nomogram. The goal of the study was to develop a prognostic tool to predict 5-yr bladder cancer recurrence risk after radical cystectomy that demonstrated higher predictive accuracy than standard AJCC TNM staging and pathologic subgroupings. Cox proportional hazards regression analyses were used to examine multiple variables including sex, age, histology (adenocarcinoma, transitional cell, and squamous cell carcinomas), pelvic lymph node status (positive or negative), pathologic stage, and grade (high or low). The nomogram was validated by performing testing on each institution using a reconstructed nomogram that excluded that institution's data. Patients who had received preoperative or adjuvant systemic chemotherapy or pelvic radiation therapy were excluded. The final nomogram predicted 5-yr disease-free survival with higher accuracy than risk stratification based on AJCC staging or standard pathologic subgroupings. The authors concluded that the nomogram was an improved prognostic instrument that can be used for improved patient counseling with regard to the need of adjuvant therapy.
Experts’ comments:
Molecular and histopathologic studies indicate that urothelial carcinomas present as a heterogeneous group of tumors that evolve from two phenotypic variants with distinct biologic behavior. Consequently, accurate risk stratification models are important to guide decision-making with regard to adjuvant chemotherapy and design clinical trials to evaluate new treatment strategies. The AJCC stages, which are based on pathologic tumor invasion (T) and lymph node involvement (N), represent the gold standard predictor of recurrence after radical cystectomy. However, the separate use of two variables, pT and pN, appears inadequate to accurately predict reoccurrence [1]. The power of nomograms is embedded in their ability to allow individualized prediction of outcome based on a combination of variables. The International Bladder Cancer Nomogram Consortium (IBCNC) nomogram represents a significant achievement in developing predictive tools that allow risk estimates to be made for recurrence in patients with bladder cancer undergoing radical cystectomy and lymphadenectomy. Nevertheless, this study also has inherent limitations that raise important issues for future research.

Nomograms are comprised of data from multiple institutions, which provide improved statistical power. The data are limited to variations in patient selection based on management and treatment protocols specific to each institution. Although the exclusion criteria in this patient population included patients that received neoadjuvant or adjuvant chemotherapy, many other more subtle and less quantifiable criteria may have further sub-selected the population leading to potential bias. Nevertheless, the decision to exclude patients who had undergone neoadjuvant or adjuvant chemotherapy is a wise one, given the variations in the regimens used because of the number of institutions and the time span of patient inclusion. Inclusion would have most likely reduced the predictive power of the tool.

The multi-institutional nature of this study and variations of surgical techniques contributed to difficulties in properly assessing lymph node status. The extent of pelvic lymph node dissection and number of lymph nodes removed has been shown to influence bladder cancer survival independent of node disease status [2]. Although the final nomogram incorporates node status, the extent of the surgery is not factored in. Another limitation is the lack of central pathologic review and the use of a limited grading system. Furthermore, because the majority of patients with squamous bladder cancers were from Egypt, where most cases are secondary to schistosomiasis, it is unclear wether this histologic distinction would offer accurate prediction to patients with tumors of squamous histology in patients in other areas without endemic Schistosoma hematobium.

The IBCNC is comprised of 12 urologic institutes of excellence and the data represent outcomes specific to this high standard of care. Multivariate analysis of data from the randomized cooperative Southwest Oncology Group trial comparing neoadjuvant chemotherapy plus cystectomy with cystectomy alone for locally advanced bladder cancer demonstrated that surgical margins were independently associated with survival after cystectomy [3]. Therefore, the quality of surgery is an important prognostic factor and thus it remains to be determined if output of this nomogram can be translated to radical cystectomy carried out in community practice. Future validation studies of this nomogram should be performed on patients from institutions with varying surgical volume to provide meaningful information to the community urologist.

A second generation of prognostic nomograms may be further improved by incorporating validated molecular (gene expression, protein-tumor or urine, and DNA) biomarkers. A recent study by Shariat [4] evaluated a nomogram for superficial papillary tumors that estimated risk of disease recurrence and progression in a cohort of 2542 patients and demonstrated statistically significant improvement in accuracy with the inclusion of nuclear matrix protein 22 (NMP22) urine level. Given the multiplicity and complexity of the biochemical pathways involved in the urothelial cell tumorigenesis, the incorporation of multiple molecular markers will undoubtably be most useful.

Finally, we would like to echo the authors’ observed deficiencies in the construction of the database and their recommendation of standardizing data exchange formats to accelerate future collaborative research projects. They developed Caisis (www.caisis.org), a freely available integrated clinic and research management system designed to standardize documentation of patient data for this purpose.

In conclusion, the IBCNC nomogram is an important step forward in the development of accurate prognostic tools for patients diagnosed with muscle-invasive bladder cancer and highlights the power and benefits of international collaboration.