Contextualizing Cretostimogene in the Evolving Treatment Landscape

The expanding portfolio of approved and emerging bladder-sparing therapies for BCG-unresponsive non–muscle-invasive bladder cancer (NMIBC) has intensified interest in cross-trial comparisons, yet the methodologic barriers to such comparisons are substantial—and, in Mark D. Tyson II, MD, MPH's view, often underappreciated. In the sixth and final segment of this series, Tyson, a professor of urology at Mayo Clinic in Phoenix, Arizona, addresses how cretostimogene grenadenorepvec fits within the broader treatment landscape, why direct numeric comparisons across single-arm trials are problematic, and what the heavily pretreated nature of the BOND-003 (NCT04452591) population says about the agent's potential.

Tyson identifies several structural reasons why cross-trial comparison in BCG-unresponsive NMIBC is unreliable. Biopsy requirements varied meaningfully across trials: BOND-003 (NCT04452591) mandated biopsies at 12 months, whereas KEYNOTE-057 (NCT02625961) did not require biopsy confirmation, meaning complete response rates in that trial reflect clinical rather than pathologic responses—a distinction with real implications for how results should be interpreted. Beyond end point definitions, Tyson notes that the FDA's BCG-unresponsive definition, while uniform, encompasses a wide spectrum of disease biology: A patient with high-volume CIS who has received 30 or more doses of BCG over many years occupies a very different biologic space than a patient with limited CIS who narrowly met the definition after 7 doses—yet both are eligible for the same trials. In small, nonrandomized, open-label studies, even a handful of patients at either end of that spectrum can materially shift aggregate results. He notes that what the BOND-003 data do show—a high complete response rate, duration of response extending beyond 24 months, and no grade 3 or higher treatment-related adverse events—constitutes a notable profile on its own terms, while acknowledging that definitive comparison to other agents would require a randomized head-to-head trial.

On the question of activity in a heavily pretreated population, Tyson describes the BOND-003 enrollment as particularly notable: As approved agents accumulated in the years following the FDA's 2018 guidance, trial populations became progressively more refractory, and many patients who enrolled in BOND-003 had already received gemcitabine-docetaxel, pembrolizumab (Keytruda), or both. The fact that cretostimogene demonstrated meaningful activity in this context suggests the agent may retain utility even after multiple prior bladder-sparing options have been exhausted—although Tyson is careful to note that he does not think in terms of lines of therapy and that each additional agent a patient moves through increases his concern about losing the curative window for cystectomy.

On sequencing, he describes an individualized approach weighing disease features, cystectomy fitness, prior therapies, agent tolerability, and logistical burden—and defers to clinical trial enrollment whenever an option is available. His closing message for urologists is direct: durable bladder preservation is a more realistic goal than it has ever been for selected patients with high-risk BCG-unresponsive NMIBC, but achieving it demands attention to durability and tolerability alongside initial response rates, and an unwavering commitment to timely cystectomy when disease trajectory warrants it.