Impact of prior intravesical bacillus Calmette-Guerin therapy on the effectiveness of pembrolizumab for patients with metastatic urothelial carcinoma.

Abstract Objective The aim of this study was to determine whether a history of treatment for non-muscle invasive bladder cancer (NMIBC), including intravesical bacillus Calmette-Guerin (BCG) therapy, affects the treatment outcomes of pembrolizumab in patients with metastatic, chemo-resistant urothelial carcinoma (UC). Materials and methods The clinicopathological data of 755 patients with metastatic, chemo-resistant UC who received pembrolizumab were retrospectively reviewed. Best overall response and overall survival (OS) from the initiation of pembrolizumab were analyzed with regard to the history of NMIBC treatment and BCG usage using propensity score matching (PSM). Results A total of 155 (20.5%) patients had a history of NMIBC treatment, of which 97 (12.8%) had received intravesical BCG therapy. When compared to patients without a NMIBC history (median 10.0 months), the OS from the initiation of pembrolizumab for patients with a NMIBC history (13.3 months, HR [95% CI] 0.79 [0.62–1.02], P = 0.073), those with a NMIBC history and BCG (12.1 months, HR 0.87 [0.64–1.17], P = 0.356), or those with a NMIBC history but not BCG (14.5 months, HR 0.68 [0.45–1.12], P = 0.061) were not significantly different. This tendency was robust after 1:1 or 1:2 PSMs. The objective response rate (ORR, 24.5% vs. 31.0%, P = 0.222) and disease control rate (DCR, 56.1% vs. 52.1%, P = 0.501) of the 155 patients with an NMIBC history did not differ from those of 155 matched patients without an NMIBC history. Among those with an NMIBC history, the prior use of BCG did not affect OS (with vs. without BCG, 12.1 vs. 14.5 months, HR 1.29 [0.80–2.09], P = 0.295), ORR (24.5% vs. 34.0%, P = 0.298) or DCR (57.1% vs. 56.0%, P = 0.908). The ORR in BCG-treated patients was significantly lower than that in those without a NMIBC history (19.8% vs. 33.3%, P = 0.042), whereas DCR between the 2 groups did not differ significantly (55.8% vs. 54.4%, P = 0.855). Conclusions Our risk-adjusted analyses revealed that a history of prior NMIBC treatment, including intravesical BCG therapy, did not affect the treatment outcomes of pembrolizumab in metastatic UC patients.