In february 2019, the POTOMAC trial was announced, focusing on high-risk non-muscle invasive bladder cancer (NMIBC). Now, 6 year later, the long-awaited results will be presented at ESMO 2025. Let’s have a look at the trial design and what we can expect.
Setting the scene: non-muscle-invasive bladder cancer.
The European Association of Urology (EAU) has proposed a risk classification for NMIBC. The definition of a high risk tumor in the POTOMAC trial basically corresponds to the EAU high or very high risk group.
High risk tumor according to POTOMAC, is defined as one of the following:
T1 tumor
High grade/Grade 3 tumor
Carcinoma in situ (CIS)
Multiple, recurrent & large tumors (>3 cm)
For patients in the high risk group (EAU), full-dose intravesical BCG for 1 to 3 years is indicated. Patients in the very high-risk group have an extremely high risk of tumor progression: 58,6% after 10 years according to the 2021 EAU NMIBC scoring model.
The EAU guidelines even have a strong recommendation to discuss immediate radical cystectomy in these patients. Guidelines are clear that if intravesical BCG is preferred over radical cystectomy, it should be offered for 1 to 3 years.
In other words, patients included in POTOMAC have a high risk of disease recurrence & progression to muscle-invasive bladder cancer. We also know that patients who experience disease progression from non-muscle-invasive to muscle-invasive stage have a worse prognosis than those who present with primary muscle-invasive disease.
POTOMAC Trial
Patients in the POTOMAC trial will be randomised in 1 of three arms (1:1:1 randomization)
BCG for 2 years (Standard of Care)
Durvalumab for 1 year + BCG for 2 years
Durvalumab for 1 year + BCG induction only
The primary endpoint is Disease Free Survival (DFS). Key secondary endpoints are the proportion of patients alive and disease free after 2 years, overall survival at 5 years, safety & tolerability, and health-related quality of life.
What we already know
According to a press release, durvalumab + SoC BCG for 2 years demonstrated a significant and clinically meaningful improvement in DFS compared to BCG for 2 years alone.
What are the issues?
- First and foremost, the third arm — in which patients are randomized to receive durvalumab for one year and BCG induction only — is highly problematic. These are patients at very high risk of recurrence and progression, and all major guidelines clearly state that they require at least one year of BCG maintenance following induction. This design therefore exposes high-risk patients to a potentially suboptimal regimen without addressing a clinically relevant question.
The rationale for including this arm is unclear; however, the aim may be to “spare” BCG maintenance through the addition of durvalumab. This concept itself is also problematic:
We first need to establish whether durvalumab is actually active in NMIBC, which will be determined by comparing the standard-of-care (SoC) group with the Durvalumab + SoC group. Only if a benefit is demonstrated here would it be appropriate to explore a trial comparing Durvalumab + SoC versus Durvalumab + BCG induction only.
One year of durvalumab treatment carries a substantial financial burden for healthcare systems.
Durvalumab is not without toxicity, and there are currently no comparative data regarding its tolerability relative to BCG.
- The true unmet need in NMIBC lies in patients unresponsive to BCG, where treatment options other than radical cystectomy are limited. With this trial design, including only BCG-naïve patients, a subset of patients will be overtreated (those who would respond to BCG alone).
- As I have described before, durvalumab has no proven survival benefit in the metastatic setting. It did show an OS benefit in the perioperative setting compared to neo-adjuvant chemotherapy, but we lack data on post-protocol care.
Without clear data on its efficiency in the metastatic and perioperative setting, questions can be asked if we should now expose an even broader group of patients to this toxic drug.
What answers do we hope to get?
- Does durvalumab + 2 years reduce the risk of recurrence & progression, in other words are you less likely to proceed to radical cystectomy if you add durvalumab to (actual) SoC?
- What happened to the patients in the BCG-induction only arm? I would suspect more censoring here due to patient (and doctors) dissatisfaction (drop-out)
- One of the secondary endpoints is overall survival at 5 years; will there be a difference, and will we have data on post-protocol therapy for patients developping metastases?
- Is durvalumab actually well tolerated? Since we are not in a metastatic setting, one could expect that patients will stop treatment earlier due to side effects compared to a metastatic setting in which patients are more desperate. In other words, is a systemic treatment sustainable and defendable in a subgroup of oncology patients who only has local symptoms.
Can you explain for an idiot like me why a T1 tumor already classifies as high risk? I would have guessed a T1N0M0 in the bladder being not a big problem (as detected).