Presentation Authors: Yu-Kuang Chen*, Tzu-Ping Lin, Taipei City, Taiwan, Shih-Yen Lu, Taoyuan City, Taiwan, Yen-Hwa Chang, Junne Yih Kuo, Hsiao-Jen Chung, Howard H.H. Wu, Shing-Hwa Lu, Eric Yi-Hsiu Huang, Chi-Cheh Lin, Yu-Hua Fan, I-shen Huang, William J. Huang, Alex T.L. Lin, Kuang-Kuo Chen, Taipei City, Taiwan
Introduction: Risk stratification with different models in bladder cancer have been introduced for prognosis prediction, including European Organization for Research and Treatment of Cancer (EORTC) risk tables, and Spanish Urological Club for Oncological Treatment (CUETO) scoring model. However, they use 1973 WHO classification, having different treatment such as BCG therapy, and were developed based on western population. We aim to evaluate the accuracy of these models with our single medical center database in Taiwan.
Methods: From 2007 to 2015, 1613 patients with pathology report of non-muscle invasive bladder urothelial carcinoma (UC) were enrolled. Patients with history of primary (pure) CIS, muscle invasive or upper tract UC were excluded. The definition of recurrence was reappearance of UC in the bladder, and progression was advancing in stage, metastasis or death caused by UC. EORTC and CUETO score have been calculated for each patient, then assigned into 4 groups. Regression model were created and obtained the concordance indexes.
Results: A total of 958 patients were in the cohort, with 55-month follow-up in average. The mean age was 73 (Â±12.7) years old, 835 (81.9%) male patients, 103 (10.1%) concurrent CIS, 659 (64.6%) primary occurrence and 274 (26.9%) patients received adjuvant intravesical BCG instillation. Four hundred twenty-four (44.3%) recurrence and 133 (13.9%) progression were noted. After stratification, the 5-year recurrence rate of 4 groups were 0%, 37%, 55% and 82% (EORTC, figure), then 40%, 42%, 53% and 68% (CUETO). The 5-year progression rate were 0%, 9%, 17% and 35% (EORTC, figure), then 4%, 7%, 14% & 26% (CUETO). All showing statistical significance. The concordance index of disease recurrence was 0.592 and 0.586 respectively for EORTC and CUETO model, showing suboptimal discrimination, while better in respect of disease progression (0.658 and 0.643, respectively).
Conclusions: In our single institute cohort and Asian population, both EORTC and CUETO model reflected the trend of outcomes. EORTC score is a better predictor but still overestimated the risk. Due to guideline migration and more intensive adjuvant treatment, a modified risk calculator for NMIBC might be warranted.