Radical cystectomy versus trimodality therapy for muscle-invasive bladder cancer: a multi-institutional propensity score matched and weighted analysis
Alexandre R Zlotta,Leslie K Ballas,Andrzej Niemierko,Katherine Lajkosz,Cynthia Kuk,Gus Miranda,Michael Drumm,Andrea Mari,Ethan Thio,Neil E Fleshner,Girish S Kulkarni,Michael A S Jewett,Robert G Bristow,Charles Catton,Alejandro Berlin,Srikala S Sridhar,Anne Schuckman,Adam S Feldman,Matthew Wszolek,Douglas M Dahl,Richard J Lee,Philip J Saylor,M Dror Michaelson,David T Miyamoto,Anthony Zietman,William Shipley,Peter Chung,Siamak Daneshmand,Jason A Efstathiou
DOI: https://doi.org/10.1016/S1470-2045(23)00170-5
Abstract:Background: Previous randomised controlled trials comparing bladder preservation with radical cystectomy for muscle-invasive bladder cancer closed due to insufficient accrual. Given that no further trials are foreseen, we aimed to use propensity scores to compare trimodality therapy (maximal transurethral resection of bladder tumour followed by concurrent chemoradiation) with radical cystectomy. Methods: This retrospective analysis included 722 patients with clinical stage T2-T4N0M0 muscle-invasive urothelial carcinoma of the bladder (440 underwent radical cystectomy, 282 received trimodality therapy) who would have been eligible for both approaches, treated at three university centres in the USA and Canada between Jan 1, 2005, and Dec 31, 2017. All patients had solitary tumours less than 7 cm, no or unilateral hydronephrosis, and no extensive or multifocal carcinoma in situ. The 440 cases of radical cystectomy represent 29% of all radical cystectomies performed during the study period at the contributing institutions. The primary endpoint was metastasis-free survival. Secondary endpoints included overall survival, cancer-specific survival, and disease-free survival. Differences in survival outcomes by treatment were analysed using propensity scores incorporated in propensity score matching (PSM) using logistic regression and 3:1 matching with replacement and inverse probability treatment weighting (IPTW). Findings: In the PSM analysis, the 3:1 matched cohort comprised 1119 patients (837 radical cystectomy, 282 trimodality therapy). After matching, age (71·4 years [IQR 66·0-77·1] for radical cystectomy vs 71·6 years [64·0-78·9] for trimodality therapy), sex (213 [25%] vs 68 [24%] female; 624 [75%] vs 214 [76%] male), cT2 stage (755 [90%] vs 255 [90%]), presence of hydronephrosis (97 [12%] vs 27 [10%]), and receipt of neoadjuvant or adjuvant chemotherapy (492 [59%] vs 159 [56%]) were similar between groups. Median follow-up was 4·38 years (IQR 1·6-6·7) versus 4·88 years (2·8-7·7), respectively. 5-year metastasis-free survival was 74% (95% CI 70-78) for radical cystectomy and 75% (70-80) for trimodality therapy with IPTW and 74% (70-77) and 74% (68-79) with PSM. There was no difference in metastasis-free survival either with IPTW (subdistribution hazard ratio [SHR] 0·89 [95% CI 0·67-1·20]; p=0·40) or PSM (SHR 0·93 [0·71-1·24]; p=0·64). 5-year cancer-specific survival for radical cystectomy versus trimodality therapy was 81% (95% CI 77-85) versus 84% (79-89) with IPTW and 83% (80-86) versus 85% (80-89) with PSM. 5-year disease-free survival was 73% (95% CI 69-77) versus 74% (69-79) with IPTW and 76% (72-80) versus 76% (71-81) with PSM. There were no differences in cancer-specific survival (IPTW: SHR 0·72 [95% CI 0·50-1·04]; p=0·071; PSM: SHR 0·73 [0·52-1·02]; p=0·057) and disease-free survival (IPTW: SHR 0·87 [0·65-1·16]; p=0·35; PSM: SHR 0·88 [0·67-1·16]; p=0·37) between radical cystectomy and trimodality therapy. Overall survival favoured trimodality therapy (IPTW: 66% [95% CI 61-71] vs 73% [68-78]; hazard ratio [HR] 0·70 [95% CI 0·53-0·92]; p=0·010; PSM: 72% [69-75] vs 77% [72-81]; HR 0·75 [0·58-0·97]; p=0·0078). Outcomes for radical cystectomy and trimodality therapy were not statistically different among centres for cancer-specific survival and metastasis-free survival (p=0·22-0·90). Salvage cystectomy was done in 38 (13%) trimodality therapy patients. Pathological stage in the 440 radical cystectomy patients was pT2 in 124 (28%), pT3-4 in 194 (44%), and 114 (26%) node positive. The median number of nodes removed was 39, the soft tissue positive margin rate was 1% (n=5), and the perioperative mortality rate was 2·5% (n=11). Interpretation: This multi-institutional study provides the best evidence to date showing similar oncological outcomes between radical cystectomy and trimodality therapy for select patients with muscle-invasive bladder cancer. These results support that trimodality therapy, in the setting of multidisciplinary shared decision making, should be offered to all suitable candidates with muscle-invasive bladder cancer and not only to patients with significant comorbidities for whom surgery is not an option. Funding: Sinai Health Foundation, Princess Margaret Cancer Foundation, Massachusetts General Hospital.