Pembrolizumab Plus BCG Elicits 92% CR Rate in BCG-Naive Very High-Risk T1 NMIBC

Pembrolizumab (Keytruda) combined with intravesical Bacillus Calmette-Guérin (BCG) generated a 92% clinical complete response (cCR) rate at 6 months in patients with BCG-naive “very high-risk” (VHR) high-grade T1 (HGT1) non–muscle-invasive bladder cancer (NMIBC), according to data from a phase 2 investigator-initiated trial (NCT03504163) presented at the 2026 American Urological Association Annual Meeting.¹

Thirty-four of 37 enrolled patients achieved a cCR at 6 months (95% CI, 79%-97%), exceeding the prespecified threshold of at least 70% required to reject the null hypothesis. At a median follow-up of 22 months, the median duration of response had not been reached, and no patient had progressed to muscle-invasive bladder cancer (MIBC) or metastatic disease.

“Our findings support pembrolizumab plus BCG as a potential bladder-preserving option in carefully selected patients who decline cystectomy, but are at very high risk for progression with BCG alone.”
-Eugene Pietzak, MD

Pietzak is a lead investigator of the study and a urologic oncologist in the Department of Surgery at Memorial Sloan Kettering Cancer Center (MSKCC) in New York, New York.

What was the rationale for adding pembrolizumab to BCG in very high-risk T1 NMIBC?

High-grade NMIBC is a heterogeneous disease, and the subset classified as VHR T1—which is defined by the co-occurrence of HGT1 disease, concomitant carcinoma in situ (CIS), and one or more additional adverse pathologic features—occupies a clinical position closer to muscle-invasive disease than to lower-risk NMIBC. These tumors carry an elevated risk of clinical understaging, harbor the molecular machinery for invasion, and may carry micrometastatic disease, placing them at substantially increased risk of progression with BCG alone.

Current guidelines from the National Comprehensive Cancer Network and the European Association of Urology recommend up-front radical cystectomy as the preferred approach for this population. However, Pietzak noted that cystectomy represents overtreatment for an estimated 50% to 75% of such patients, and BCG monotherapy may constitute undertreatment for 25% to 50% of patients. This treatment gap formed the rationale for evaluating immune checkpoint blockade (ICB) in combination with BCG.

Prior data have suggested that PD-1/PD-L1 signaling may be a mechanism of BCG resistance, particularly in HGT1 tumors with CIS. Preclinical studies have showcased enhanced CD4 T-cell immunity when anti-PD-L1 therapy is added to BCG. Importantly, the three large randomized controlled trials evaluating ICB plus BCG in the broader NMIBC population (CREST: NCT04165317; POTOMAC: NCT03528694; ALBAN: NCT03799835) enrolled very few patients with VHR HGT1 plus CIS, Pietzak noted, leaving the role of treatment intensification in this cohort undefined.

How was the phase 2 trial of pembrolizumab plus BCG designed?

This single-arm, single-center, phase 2 study enrolled BCG-naive adult patients with histologically confirmed HGT1 N0M0 urothelial cancer and concomitant CIS, plus at least one additional adverse risk feature including persistent or recurrent T1 disease on restaging transurethral resection of bladder tumor (TURBT), several T1 tumors, T1 tumor size of at least 3 cm, T1b disease defined as deep or extensive lamina propria invasion, variant histology, prostatic urethral involvement, or lymphovascular invasion. All enrolled patients had been recommended to undergo immediate radical cystectomy, but declined surgery.

Those who received previous intravesical BCG or any ICI or other T-cell–directed agents were excluded. Other exclusion criteria included having pure squamous carcinoma, pure adenocarcinoma, or any small cell or neuroendocrine disease. Those with a history of muscle-invasive disease or prostatic stromal invasion were not permitted.

All participants underwent restaging TURBT and examination under anesthesia at MSKCC within 8 weeks of study entry. Patients received intravenous pembrolizumab at 400 mg every 6 weeks for up to 9 doses, equating to approximately 48 weeks. BCG (TICE strain, 50 mg) induction began 3 weeks after the first pembrolizumab dose, followed by SWOG-style maintenance BCG at 3, 6, and 12 months. Additional maintenance was allowed based on availability during the global BCG shortage.

In addition to the primary end point of cCR at 6 months, secondary end points included recurrence-free survival, cystectomy-free survival, and safety. bladder-intact disease-free survival (BI-DFS) would also be available in a post hoc analysis.

Among the 43 patients assessed for eligibility, 37 were enrolled and included in the intention-to-treat cohort. The median age was 67 years (range, 34-86); 86% of patients were male, and 100% had an ECOG performance status of 0 or 1. All patients had HGT1 and CIS; 70% had diffuse or multifocal CIS. The most common additional adverse features were T1 tumor size larger than 3 cm (68%), extensive lamina propria invasion (57%), several T1 tumors (43%), and T1b disease (37%). Notably, 73% of patients harbored two or more adverse features, and 51% harbored three or more of these features.

What were the recurrence-free survival and bladder-intact outcomes?

Among 6-month complete responders, response was durable over long-term follow-up: 12-month high-grade recurrence-free survival (HG-RFS) was approximately 88%, and the median HG-RFS had not been reached. The 6-month HG-RFS rate was 97.3%; at 12 months, this rate was 88.5%. Four patients (11%) experienced high-grade recurrence; 2 patients had recurrent HGT1, 1 had bladder CIS/Tis, and 1 had prostatic urethral CIS. Two patients (5%) ultimately underwent radical cystectomy: one for recurrent HGT1 (final pathology: pTis N0) and one for a pubovesical fistula arising in the context of prior robotic-assisted laparoscopic prostatectomy and radiotherapy (final pathology: pT0 N0).

In a post hoc analysis using the definition established by Apolo et al,² the 12-month BI-DFS rate was approximately 94%, with the median BI-DFS not reached.¹

What were the safety data of pembrolizumab plus BCG in this setting?

Treatment-related adverse effects (TRAEs) of any grade occurred in 33 of 37 patients (89%). Grade 3 or higher TRAEs were observed in 32% of patients, and grade 3 or higher immune-related adverse effects (irAEs) occurred in 21%.

The most common TRAEs of any grade were fatigue (35%), rash (32%), pruritus (27%), and hypothyroidism (22%). Elevated alanine aminotransferase/aspartate aminotransferase (hepatitis) was the most frequent grade 3 or higher TRAE, occurring in 5 patients: 4 had grade 3 events and 1 had a grade 4 event. Two patients developed grade 3 pneumonitis, 1 developed grade 4 laryngeal edema, and 1 developed grade 4 sepsis.

One death occurred approximately 3 weeks after the first pembrolizumab dose, before BCG initiation; the cause was unknown, and an autopsy was declined. Given the temporal association, the event was classified as a possibly treatment-related grade 5 TRAE, although the patient had multiple comorbidities.

When benchmarked against the CREST, POTOMAC, and ALBAN randomized trials examining sasanlimab, durvalumab (Imfinzi), and atezolizumab (Tecentriq), respectively, the grade 3 or higher TRAE rate of 32% in this study was numerically higher (CREST: 29%; POTOMAC: 21%; ALBAN: 23%), as was the grade 3 or higher irAE rate of 21% (CREST: 16%; POTOMAC: 8%; ALBAN: 9%).^3-5

Pietzak noted¹ that this higher toxicity signal may reflect the VHR T1 population’s more intensive treatment course, and emphasized that radical cystectomy, the recommended alternative, carries its own rate of serious adverse effects of approximately 22%.

What were the limitations of this research?

Pietzak acknowledged several limitations. He noted that this was a relatively small, single-arm, non-randomized trial conducted at a single high-volume cancer center with multidisciplinary bladder cancer expertise, which may limit generalizability. The absence of mandated post-treatment biopsies may overestimate response rates, although he added that longer follow-up may mitigate this concern. Patient-reported outcomes were not assessed, which Pietzak flagged as important context for shared decision-making.

Data from the randomized phase 3 KEYNOTE-676 trial (NCT03711032) are pending and expected to provide further informative insights in this setting.

Disclosures: Dr Pietzak disclosed serving in a consulting or advisory role for Urogen, Janssen, and Merck & Co; serving on uncompensated clinical trial steering committees for Tyra and Pfizer; and serving on data safety monitoring committees for QED Therapeutics and CG Oncology. Research funding (funds to institution) was provided by Janssen, Merck & Co, National Cancer Institute/Alliance for Clinical Trials in Oncology, Predicine, and Veracyte. Funding for this trial and associated correlatives was provided by Merck Sharp & Dohme LLC, a subsidiary of Merck & Co, Inc.

References

1. Pietzak E, Feld E, Frydenlund N, et al. A phase II study of intravenous pembrolizumab combined with intravesical bacillus Calmette-Guérin (BCG) for patients with BCG-naïve “very high-risk” T1 non-muscle invasive bladder cancer traditionally recommended for radical cystectomy. Presented at: 2026 American Urological Association Annual Meeting; May 15-18, 2026; Washington, DC. P2.
2. Apolo AB, Atiq S, Kydd AR, et al. Top advances of the year: bladder cancer. Cancer. 2025;131(16):e70020. doi:10.1002/cncr.70020
3. Shore ND, Powles TB, Bedke J, et al. Sasanlimab plus BCG in BCG-naive, high-risk non-muscle invasive bladder cancer: the randomized phase 3 CREST trial. Nat Med. 2025;31(8):2806-2814. doi:10.1038/s41591-025-03738-z
4. De Santis M, Redorta JP, Nishiyama H, et al. Durvalumab in combination with BCG for BCG-naive, high-risk, non-muscle-invasive bladder cancer (POTOMAC): final analysis of a randomised, open-label, phase 3 trial. Lancet. 2025;406(10516):2221-2234
5. Roupret M, Bertaut A, Pignot G, et al. LBA107 ALBAN: A phase III, randomized, open-label, international study of intravenous (iv) atezolizumab and intravesical Bacillus Calmette-Guérin (BCG) vs BCG alone in BCG-naïve high-risk, non-muscle-invasive bladder cancer (NMIBC). Ann Oncol. 2025;36(suppl 2):S1649-S1650. doi:10.1016/j.annonc.2025.09.126