Introduction:

Systemic pembrolizumab is a second-line therapy for BCG-unresponsive non-muscle invasive bladder cancer (NMIBC). Intravesical administration, however, is still under investigation but has the potential to provide similar or improved antitumor activity with less toxicity. To date, few biomarkers can identify response to BCG. Our objective was to perform an analysis of immune cells in the urine of BCG-unresponsive NMIBC patients enrolled in a Phase I trial of intravesical pembrolizumab (MK-3475) with BCG and correlate immune populations with clinical response.  

Methods:

This prospective Phase I clinical trial accrued nine patients between August 2017 and November 2019 (NCT02808143). At enrollment, four patients were TaHG (two with CIS), two were T1 and three Tis. The trial was stopped prematurely due to the pandemic. Median follow-up was 33.4 months among alive patients. Urine was collected at the start of the study, after an initial dose of MK-3475, and then before and after every dose of treatment (alternating MK-3475 plus BCG or BCG alone). Post-treatment urine was collected 6 hours after instillation. Analysis of urine cells was performed using flow cytometry and normalized to volume of urine collected. Patient with early recurrence had tumor present ≤ 25 weeks after treatment initiation.

Results:

At baseline, total urine CD45+ immune cells were low with the majority being granulocytes (CD45⁺SSC^hi, 86.9% ± 6.9%) followed by lymphocytes (CD45⁺SSC^lo, 2.8% ± 4.1%). The initial dose of intravesical MK-3475 (Week -2) caused a significant decrease in total CD45+ (78.2% decrease, p=0.026) which was more prominent in patients with late (>25 weeks) compared to early recurrence (≤ 25 weeks). In patients with late recurrence, the combination MK-3475 + BCG decreased total CD45+ immune cells from baseline (87.5% average decline) compared to BCG alone (117.0% average increase). At the end of induction, patients with later recurrence had a lower neutrophil/lymphocyte ratio (NLR) (0.5 ± 0.2) compared to patients with early recurrence (493.7 ± 575.6) (Fig 1).

Conclusion:

Initial investigation of the urine immune microenvironment in patients treated with intravesical MK-3475 and BCG identified changes in immune populations that correlate with response to therapy. Intravesical MK-3475 decreased CD45+ cells. High NLR may be prognostic for poor response to MK-3475 and BCG. Future research, focusing on the role of the innate immune system, may identify the role of neutrophils in BCG therapy and as biomarker for BCG response.

Funding: Merck Sharp & Dohme, Robert H. Lurie Cancer Center