Biomarkers Emerge for Pathologic Overall Response With Gem-iDRS Plus Cetrelimab in MIBC

A biomarker analysis of the phase 2 SunRISe-4 trial showed that elevated tumor mutational burden, high genomic disease burden, and PD-L1 expression were each significantly associated with pathologic overall response (pOR) to neoadjuvant gemcitabine intravesical drug release system (Gem-iDRS; Inlexzo) plus cetrelimab or cetrelimab alone in patients with muscle-invasive bladder cancer (MIBC); however, none of the markers evaluated predicted pathologic complete response or recurrence-free survival.¹

Results presented during the 2026 American Urological Association Annual Meeting showed that in the overall cohort (n = 88), pOR rates were significantly higher in patients with high baseline TMB (n = 46), high GDB (n = 43), or high PD-L1 (n = 28) compared with those with low levels of each marker, respectively. More specifically:

• TMB high: pOR 41% (n = 32 biomarker-positive responders/79 patients evaluable); odds ratio (OR), 3.92 (95% CI, 1.41-11.52); P = .01
• GDB high: pOR 50% (n = 29/58); OR, 4.03 (95% CI, 1.02–18.13); P = .03
• PD-L1 high: pOR 32% (n = 20/62); OR, 3.10 (95% CI, 0.97–10.64); P = .04

MMRd status was not significantly associated with pOR in the overall cohort (n = 18). For MMRd-evaluable patients, the pOR was 18% (n = 8/44, [OR, 0.69; 95% CI, 0.17-2.69; P = .76]).

Regarding pCR, ORs for high TMB, high GDB, PD-L1 expression, and MMRd were 1.75 (95% CI, 0.62-5.13; P = .25), 2.10 (95% CI, 0.52-10.76; P = .36), 1.55 (95% CI, 0.49-5.02; P = .44), and 0.73 (95% CI, 0.15-3.17; P = .74), respectively

“High genomic disease burden markers (TMB, GDB) and PD-L1 expression at baseline were associated with high pathologic overall response to neoadjuvant gem-iDRS plus cetrelimab or cetrelimab alone in patients with MIBC; further investigation in a larger study is warranted to further evaluate the predictive utility of these biomarkers,” according to Felix Guerrero-Ramos, MD, et al. Guerrero-Romas is an attending urologist in the Department of Urology at Hospital Universitario 12 de Octubre in Madrid, Spain.

Clinical Context

The standard of care for MIBC is radical cystectomy (RC) with neoadjuvant cisplatin-based chemotherapy; however, up to 50% of patients may be ineligible for or refuse cisplatin-based treatment. Effective and better-tolerated neoadjuvant options are therefore needed for patients with MIBC who are not candidates for the latter regimen.

SunRISe-4 evaluated neoadjuvant gem-iDRS—a drug delivery system enabling sustained intravesical gemcitabine exposure—combined with cetrelimab, a PD-1 checkpoint inhibitor, in this patient population.² In the primary analysis of SunRISe-4, neoadjuvant gem-iDRS plus cetrelimab demonstrated a pCR rate of 37.5% (95% CI, 27.4%-48.5%), a pOR rate of 53.4% (95% CI, 42.5%-64.1%), and a 1-year RFS rate of 77.4% (95% CI, 66.6%-85.1%) in patients with MIBC.

The current biomarker analysis was conducted to determine which genomic and immune characteristics at baseline may be associated with these clinical outcomes, with the goal of identifying patients most likely to benefit from the combination or from cetrelimab monotherapy.¹

Design of SunRISe-4 and Current Analysis

SunRISe-4 is a phase 2, open-label, randomized controlled trial enrolling patients aged 18 years or older with histologically confirmed cT2–T4a N0M0 MIBC of predominant urothelial carcinoma histology and an ECOG performance status of 0 or 1. Patients were also required to be ineligible for or have refused cisplatin-based chemotherapy and be scheduled for RC.

Upon enrollment, patients were stratified by visible residual disease at transurethral resection of bladder tumor (complete vs incomplete [≤3 cm]) and tumor stage at MIBC diagnosis (cT2 vs cT3-T4a). Those in cohort 1 (n = 101) received 225 mg of gem-iDRS plus 360 mg of cetrelimab intravenously every 3 weeks for 12 weeks; cohort 2 (n = 58) received the same dose of cetrelimab as monotherapy. RC was performed at approximately week 12.

The primary endpoint was pCR rate, with secondary end points including RFS and safety and exploratory end points including pOR rate, overall survival, and biomarker analysis.

The current biomarker analysis was designed to evaluate associations between baseline TMB, mismatch repair deficiency (MMRd), GDB, and PD-L1 status and clinical outcomes in this patient population. It included data from the overall cohort as well as cohort 1 alone; cohort 2 was not analyzed separately due to its small sample size. At the data cutoff of May 9, 2025, 73 tissue samples and 67 urine samples were collected. Baseline patient samples were analyzed via whole exome sequencing (WES) for tissue using Signatera (Natera) and for urine using UroAMP (Convergent Genomics).

Baseline Biomarker Frequencies

Among biomarker-evaluable patients in the overall cohort, comprising 61 patients from cohort 1 and 27 patients from cohort 2, the frequency of each biomarker was as follows:

• TMB high: 58% (n = 51/88)
• MMRd: 41% (n = 19/46)
• GDB high: 75% (n = 50/67)
• PD-L1 high: 42% (n = 29/69)

In the overall cohort, pairwise analyses revealed no statistically significant associations between TMB, GDB, PD-L1, or MMRd status when compared to one another, indicating these markers capture independent, non-overlapping dimensions of tumor biology.

In an analysis of cohort 1 alone, pOR rates were found to be significantly higher in patients with high baseline TMB (n = 35) or high PD-L1 (n = 22) compared with those with low levels:

• TMB high: pOR 44% (n = 25 biomarker-positive patients/57 patients evaluable); OR, 3.52 (95% CI, 1.02-12.89); P = .02
• PD-L1 high: pOR 37% (n = 17/46); OR, 3.89 (95% CI, 0.96-18.18); P = .04
• GDB high (n = 29): pOR 49% (n = 20/41); OR, 3.02 (95% CI, 0.62-15.86); P = .16
• MMRd (n = 15): pOR 25% (n = 8/32); OR, 0.63 (95% CI, 0.11-3.21); P = .72

As seen in the overall cohort, no biomarker was significantly associated with pCR in cohort 1:

• TMB high: OR, 1.78 (95% CI, 0.52-6.55; P = .40)
• PD-L1 high: OR, 1.38 (95% CI, 0.37–5.29; P = .76)
• GDB high: OR, 1.40 (95% CI, 0.28–7.87; P = .73)
• MMRd: OR, 0.57 (95% CI, 0.10–2.90; P = .49).

Continuous TMB and RFS Analyses

Using baseline TMB on a continuous scale derived from WES data, TMB was significantly higher in patients achieving a pCR or pOR compared with those who did not achieve these end points in cohort 1 (Wilcoxon P = .0003 for pCR; Wilcoxon P = .009 for pOR). A similar trend was observed in the overall cohort.

Additionally, patients with a high baseline TMB trended toward longer RFS compared with those with a low baseline TMB, though this did not reach statistical significance (HR, 1.75; 95% CI, 0.74-4.15; log-rank P = .2). No trends were observed between RFS and baseline PD-L1, MMRd, or GDB.

Gene Alterations and TMB Status

Mutational data from WES and urinary tumor DNA (utDNA) were assessed to determine which gene alterations predicted TMB-high status at baseline in the overall cohort. Among the genes examined, TP53 and ATM alterations were significantly associated with high TMB. Specific data are as follows:

• TP53: altered in 75% (n = 38/51) of TMB-high vs 43% (n = 16/37) of TMB-low patients (OR, 3.8; 95% CI, 1.4-10.5; P = .004)
• ATM: altered in 24% (n = 12/51) of TMB-high vs 5% (n = 2/37) of TMB-low patients (OR, 5.3; 95% CI, 1.1-52.0; P = .036)

Other genes examined, such as RB1, PIK3CA, TSC1, ERCC2, FGFR3, and POLE, did not demonstrate statistically significant associations with TMB-high status at baseline.

Based on the overall data from this analysis, Guerrero-Ramos and colleagues concluded that further investigation in a larger study is warranted to evaluate the predictive utility of TMB, GDB, and PD-L1 in this setting and to determine whether these markers can prospectively identify patients most likely to benefit from neoadjuvant gem-iDRS plus cetrelimab in MIBC.

Disclosures: Guerrero-Ramos F reported stock and other ownership interests in Pfizer, CG Oncology, and Johnson & Johnson; honoraria from AstraZeneca, Johnson & Johnson, BMS GmbH & Co. KG, Combat Medical, Merck, Pfizer, Nucleix, Roche, Palex, Johnson & Johnson, and MSD Oncology; consulting or advisory roles with AstraZeneca, Pfizer, Johnson & Johnson Oncology, BMS GmbH & Co. KG, Johnson & Johnson, Roche, Nucleix, Merck, Menarini Group, and Photocure; speakers' bureau participation with Pfizer, Johnson & Johnson Oncology, AstraZeneca, Combat Medical, Nucleix, and BMS GmbH & Co. KG; research funding from Roche and Combat Medical; patents, royalties, or other intellectual property related to BlaDimiR, a miRNA-based diagnostic urine test; expert testimony for Nucleix, Johnson & Johnson, and Menarini Group; and travel, accommodations, or expenses from Johnson & Johnson, AstraZeneca, Johnson & Johnson Oncology, Ipsen, Combat Medical, Nucleix, Recordati, Pfizer, SALVAT, Italfarmaco, and Fidia Farmaceutici S.p.A.

References

1. Guerrero-Ramos F, Necchi A, Shore N, et al. Association of genomic disease burden markers and programmed death-ligand 1 with clinical outcomes in patients with muscle-invasive bladder cancer treated with gemcitabine intravesical system + cetrelimab or cetrelimab alone in the phase 2 SunRISe-4 trial. Presented at: 2026 AUA Annual Meeting; May 15–18, 2026; Washington, DC. Abstract IP36-06.

2. Necchi A, Guerrero-Ramos F, Crispen PL, et al. Gemcitabine intravesical system plus cetrelimab or cetrelimab alone as neoadjuvant therapy in muscle-invasive bladder cancer: SunRISe-4 primary analysis and biomarker results. J Clin Oncol. 2026;44(7):586-597. doi:10.1200/JCO-25-02382