Background: PD1/PDL1 pathway inhibition has activity in BC. NAC effect on PD1, PDL1, PDL2expression is unknown. We assessed PD1, PDL1, PDL2 TT expression in BC patients (pts) before (pre) & after (post) NAC, and its potential impact on NAC response and progression-free survival (PFS). Methods: 40 pts with surgery after NAC (diagnosed 2006-2013) were identified. Paraffin-embedded TT from pre-NAC (TURBT) and/or post-NAC (cystectomy) was analyzed (IHC: low 1-3 +; high 4/5 +) for PD1, PDL1, PDL2 expression. NanoString (nCounter, custom 800-gene set) was used for TT gene expression profiling. Fisher’s exact test, Cochran-Armitage trend test, and proportional hazards models were used for analysis of IHC data & clinical factors. Correlation between mRNA & protein expression was explored. Results: Median age at diagnosis was 65 (39-82), 92% men, 28% never-smoker; 38 pts had NAC (63% Gem-Cis, 18% dd-MVAC, 8% Gem-Carbo, 8% MVAC, 3% Cis-Etop); 6/40 had pT0, 16/40 downstaged vs TURBT; 7/39 pN+, 12/37 LVI+. Primary histology at cystectomy: 25 pure urothelial carcinomas (UC), 7 UC/squamous features, 1 small cell carcinoma. PD1, PDL1, PDL2 protein was expressed pre- & post-NAC. There was significant correlation between PDL1 & PDL2 pre-NAC (p< .0001), PDL1 & PDL2 post-NAC (p< .0001), PDL1 & PD1 post-NAC (p=.0008), PDL2 & PD1 post-NAC (p=.007), and post vs pre- NAC change in both PDL1 & PDL2 protein expression (p< .0001). 76% of paired tissues had reduction in PD1, 44% in PDL1, and 41% in PDL2 protein post vs pre- NAC. High post-NAC PDL1 correlated with higher pT stage & no downstaging (Table). Post-NAC high PDL1 & PDL2 protein expression correlated with shorter PFS (p= .04, HR=4.26; p= .03, HR=4.65, respectively). There was significant correlation between mRNA & protein expression for PDL1 & PDL2 (not for PD1) but no significant difference in PD1, PDL1, PDL2 mRNA expression post vs pre- NAC. Conclusions: NAC affects PD1, PDL1, PDL2 TT IHC expression; this may impact clinical trial designs; validation is needed.