Stratified absolute benefit with adjuvant immunotherapy (aIO) in muscle-invasive bladder cancer (MIBC): A meta-analysis.

person Syed Arsalan Ahmed Naqvi Division of Hematology and Medical Oncology, Mayo Clinic, Scottsdale, AZ info_outline Syed Arsalan Ahmed Naqvi, Akshat Saxena, Kunwer Sufyan Faisal, Arifa Bibi, Muhammad Ali Khan, Ewan Kemar Cobran, Syed A. Hussain, Irbaz Bin Riaz, Alan Haruo Bryce, Parminder Singh

Authors person Syed Arsalan Ahmed Naqvi Division of Hematology and Medical Oncology, Mayo Clinic, Scottsdale, AZ info_outline Syed Arsalan Ahmed Naqvi, Akshat Saxena, Kunwer Sufyan Faisal, Arifa Bibi, Muhammad Ali Khan, Ewan Kemar Cobran, Syed A. Hussain, Irbaz Bin Riaz, Alan Haruo Bryce, Parminder Singh Organizations Division of Hematology and Medical Oncology, Mayo Clinic, Scottsdale, AZ, Division of Hematology and Medical Oncology, Mayo Clinic, Phoenix, AZ, Ziauddin Medical University, Karachi, Pakistan, Mayo Clinic College of Medicine and Science, Scottsdale, AZ, University of Sheffield and Sheffield Teaching Hospitals, Sheffield, United Kingdom, Mayo Clinic, Scottsdale, AZ, City of Hope, Phoenix, AZ, Mayo Clinic, Phoenix, AZ Abstract Disclosures Research Funding No funding sources reported Background: Should patients (pts) with MIBC receive aIO given the DFS benefit seen in AMBASSADOR and CheckMate 274 trials is an important clinical question. Thus, we aimed to quantify the magnitude of absolute benefit across different subgroups to aid clinicians in shared decision-making with their patients. Methods: Phase III trials evaluating aIO as compared to observation or placebo (obs/plac) in pts with MIBC were considered eligible. Outcomes of interest included DFS and OS. Pre-specified subgroups included disease location, PDL1 status, neoadjuvant chemotherapy (NAC) receipt, and tumor and nodal (TN) staging. Pairwise comparisons were made using a DerSimonian-Lairds (DL) random-effects meta-analysis. Stratified baseline risks (obtained from clinical trials) and relative effect estimates were used to estimate corresponding intervention risks (CIRs). The difference in CIRs and baseline risk was calculated to present absolute risk differences (ARDs) in each subgroup category. Results: Analysis included three trials with 2220 pts (pembrolizumab, nivolumab, and atezolizumab assessed in each). No significant difference was observed in OS with aIO compared to obs/plac (hazard ratio: 0.91; 0.76-1.09). However, aIO significantly improved DFS compared to obs/plac (0.76; 0.64-0.91) in the overall population. In pts with bladder/lower tract disease, aIO improved DFS compared to obs/plac (0.72; 0.56-0.92); however, no significant difference was observed in DFS in pts with upper tract disease. There was a statistically significant effect modification by disease location (p: 0.01). Adjuvant IO significantly improved DFS in PDL1- (0.78; 0.64-0.94) but not in PDL1+ pts (0.75; 0.52-1.08). Likewise, aIO significantly improved DFS in pts who received NAC (0.69; 0.53-0.90) but not in pts who did not receive NAC (0.89; 0.75-1.05). In terms of TN staging, aIO improved DFS in T2-3N0 (0.73; 0.57-0.94), T4N0 (0.76; 0.61-0.95), and in T2-4N1 patients (0.75; 0.59-0.95). No statistically significant modifications were observed among these subgroups. Stratified ARDs for DFS by each subgroup are outlined (Table). Conclusions: When delayed disease progression is a vital consideration, a risk-adapted strategy for the optimal patient selection to receive aIO should be considered. Disease location (lower tract/bladder), node positivity, and receipt of NAC emerged as groups likely to derive greater benefit with aIO. Anticipated Absolute Effects of Adjuvant IO ARD per 1000 (95% CI) Overall 95 fewer (146 fewer to 38 fewer) Upper tract 59 more (48 fewer to 175 more) Lower tract/bladder 115 fewer (194 fewer to 30 fewer) PD-L1+ 88 fewer (153 fewer to 22 fewer) PDL1- 98 fewer (206 fewer to 28 fewer) No NAC 41 fewer (97 fewer to 17 more) NAC 130 fewer (211 fewer to 38 fewer) T2-3N0 98 fewer (165 fewer to 20 fewer) T4N0 97 fewer (168 fewer to 19 fewer) T2-4N1 105 fewer (189 fewer to 19 fewer)