Presentation Authors: Madeleine L. Burg*, Thomas G. Clifford, Michael Lin-Brande, Gus Miranda, Jie Cai, Sumeet Bhanvadia, Hooman Djaladat, Anne K. Schuckman, Siamak Daneshmand, Los Angeles, CA
Introduction: Frailty has been associated with worse postoperative outcomes after radical cystectomy (RC) in studies using a prospective frailty assessment or a frailty index. Herein we aimed to determine which method is most predictive of complications after RC.
Methods: Patients >65 years undergoing open or robotic RC were prospectively assessed at a tertiary-care referral center using Fried Frailty Criteria (FFC): grip strength, gait speed, physical activity, feelings of exhaustion, and shrinking (>10 pounds of weight loss in past year). Patients were scored using the 11-item modified frailty index (mFI), 15-item urologic oncology modified frailty index1 (oncFI), and 5-item simplified frailty index (sFI).2 Thirty-day complications were recorded per the Clavien-Dindo classification system (overall and >Grade IIIa). Univariate and multivariable logistic regression analyses were performed.
Results: 123 patients were assessed preoperatively, with 109 patients receiving full FFC assessment. 59 patients (48%) had >1 30-day complication with 16 patients (13%) having a high-grade complication. On univariate analysis, oncFI (p=0.04) and shrinking (p=0.005) were predictive of overall complications, while FFC (p=0.02) and sFI (p=0.02) were predictive of high-grade complications. On multivariable logistic regression, shrinking (OR 3.79, 95% CI 1.64-9.26, c-index 0.70) and oncFI >3 (OR 3.26, 95% CI 1.50-7.33, c-index 0.69) were predictive of overall complications, after adjusting for patient age, pathologic stage, and type of urinary diversion. For high-grade complications, being intermediately frail or frail on FFC (OR 4.87, 95% CI 1.39-22.77, c-index 0.68) and sFI >3 (OR 68.00, 95% CI 3.83-1,000, c-index 0.68) were predictive of high-grade complications.
Conclusions: In this study shrinking was as predictive as oncFI of increased risk for any complication after RC. FFC was the best predictor for high-grade complications, although this analysis is limited by few high-grade events. mFI may not be predictive of complications after RC, contradicting prior registry studies. Further prospective studies are warranted on larger institutional cohorts.