448
Background: Neoadjuvant and adjuvant chemotherapy improve relapse-free survival (RFS) and overall survival of patients (pts) with MIBC. Limited information is available regarding risk prediction of local (R
L
) vs. distant relapse (R
D
), including role of perioperative chemotherapy (POC), added to radical cystectomy (Cy). Methods: Data from 1,559 pts, treated at 29 centers from the U.S., Europe, Israel, and Canada, were collected. Pts received Cy for MIBC from 02/90 to 12/13. Of these pts, 782 (50.2%) received POC and Cy, 777 (49.8%) Cy alone. R
L
and R
D
were defined as follows: pelvic lymph-nodes/soft tissue only and any extra-pelvic recurrences, respectively. Cumulative incidence methods were used to estimate time-to-(TT) R
L
/R
D
, which accounts for the competing risk of other types of relapse. Univariable and multivariable (MVA) Cox regression analyses were performed from the complete-case dataset (n = 1,082). Risk groups were defined according to the number of adverse factors, with corresponding nomogram-based R
L
and R
D
risk estimation. All tests were two-sided and statistical significance was defined as a p-value of 0.05 or less. Results: A total of 830 pts (55%) developed a relapse, 447 in the Cy group and 383 in the Cy+POC group. On MVA, POC administration was associated with longer TTR
L
(p < 0.001) and TTR
D
(p < 0.001). Other factors associated with R
L
: histology (non-UC, odds ratio [OR] = 1.47, 95%CI: 1.07-2.01, p = 0.022), pT-stage (p < 0.001), pN-stage (p = 0.038), surgical margins (p < 0.001). For R
D
: Charlson score (p = 0.006), pT-stage (p < 0.001), pN-stage (p < 0.001). The c-index of the model for R
L
was 0.685 (95% bootstrapped CI: 0.664-0.718), and for R
D
was 0.684 (0.655-0.721). Three risk group categories were obtained for both endpoints (0, 1-2, and > 2 risk factors). Results were confirmed after applying 90-day or 180-day landmark analyses, pending external validation. Conclusions: In the largest study that separately analyzed R
L
and R
D
risk, we were able to provide risk tools that may be used to optimize locoregional treatments and compare POC benchmark with new drugs in the perioperative setting.