Purpose: American Joint Committee on Cancer staging represents the gold standard for prediction of recurrence after radical cystectomy in patients with invasive bladder cancer. We tested the hypothesis that American Joint Committee on Cancer stage based predictions may be improved when pathological tumor and node stage information is combined with additional clinical and pathological variables within a prognostic nomogram.
Materials and methods: We used Cox proportional hazards regression analysis to model variables of 728 patients with transitional cell carcinoma of the bladder treated with radical cystectomy and bilateral pelvic lymphadenectomy at 1 of 3 participating institutions. Standard predictors, pT and pN, were complemented by age, gender, tumor grade at cystectomy, presence of lymphovascular invasion, presence of carcinoma in situ in the cystectomy specimen, neoadjuvant chemotherapy, adjuvant chemotherapy and adjuvant radiotherapy. The concordance index was used to quantify the accuracy of regression coefficient based nomograms. A total of 200 bootstrap resamples were used to reduce overfit bias and for internal validation. Calibration plots were used to graphically explore the performance characteristics of the multivariate nomogram.
Results: Followup ranged from 0.1 to 183.4 months (median 24.9, mean 36.4). Recurrence was recorded in 249 (34.2%) patients with a median time to recurrence of 108 months (range 0.8 to 131.9). Actuarial recurrence-free probabilities were 69.6% (95% CI 65.8%-73.0%), 60.2% (55.8%-64.3%) and 52.9% (47.3%-58.1%) at 2, 5 and 8 years after cystectomy, respectively. Two-hundred bootstrap corrected predictive accuracy of American Joint Committee on Cancer stage based predictions was 0.748. Accuracy increased by 3.2% (0.780) when age, lymphovascular invasion, carcinoma in situ, neoadjuvant chemotherapy, adjuvant chemotherapy and adjuvant radiotherapy were added to pathological stage information and used within a nomogram.
Conclusions: A nomogram predicting bladder cancer recurrence after cystectomy is 3.2% more accurate than American Joint Committee on Cancer stage based predictions. Moreover, a nomogram approach combines several advantages such as easy and precise estimation of individual recurrence probability at key points after cystectomy, which all patients deserve to know and all treating physicians need to know.