CREST and POTOMAC: Redefining immunotherapy’s role in NMIBC

Investigators recently commented on findings from the phase 3 CREST and POTOMAC trials in a Journal of Urology editorial.¹ In this interview, author Daniel Antonio González Padilla, MD, a urologist with Clínica Universidad de Navarra in Madrid, Spain, discusses the rationale, findings, and clinical implications of the CREST (NCT04165317) and POTOMAC (NCT03528694) trials evaluating immune checkpoint inhibitors (ICIs) in high- and very high-risk non–muscle invasive bladder cancer (NMIBC). Despite optimal transurethral resection and BCG therapy, these patients face substantial recurrence (30% to 50% at 5 years) and progression risks (up to 40%), underscoring an unmet need for more effective bladder-preserving strategies. Both trials were designed around the biologically plausible hypothesis that PD-1/PD-L1 signaling contributes to BCG resistance, and that adding ICIs—sasanlimab in CREST and durvalumab (Imfinzi) in POTOMAC—could improve outcomes beyond BCG alone.

The trials were largely similar in design, although POTOMAC used European Association of Urology risk stratification, allowing the inclusion of large, multiple and recurrent low-grade tumors, and allowed patients with remote prior BCG exposure, potentially diluting efficacy signals. Both studies used disease-free or event-free survival as primary endpoints, which Dr. González Padilla emphasizes are highly meaningful to patients focused on avoiding recurrence, repeated procedures, and radical cystectomy while preserving quality of life.

CREST demonstrated a 7% absolute improvement in freedom from high-grade recurrence with 2 years of sasanlimab plus BCG, with particularly striking—but exploratory—signals in carcinoma in situ, where 3-year recurrence-free survival approached 90%. POTOMAC showed a more modest 4.4% absolute benefit, highlighting that durvalumab’s efficacy appeared dependent on concurrent and sustained BCG maintenance, reinforcing the importance of intravesical therapy for synergy.

However, these benefits must be weighed against toxicity: approximately 25% of patients experienced grade 3 to 4 immune-related adverse events, some potentially lifelong. Although progression rates to muscle-invasive disease remained low (<5% at 3 years), neither trial reduced progression compared with BCG alone. Dr. González Padilla concludes that ICIs may earn a place in future guidelines once approved, but likely for carefully selected, very high-risk patients—potentially guided by biomarkers such as PD-L1—making shared decision-making essential.

What unmet clinical needs in high-risk NMIBC motivated the design of the CREST and POTOMAC trials, and how did each trial conceptualize the role of immune checkpoint inhibitors relative to standard BCG therapy?

High- and very high-risk non-muscle invasive bladder cancer has a very high risk of recurrence. It can be up to 30% to 50% in 5 years. That's super high, even after complete endoscopic resection and even after appropriate BCG management, so there is really a need for new treatments in this area. [NMIBC] also has a high progression risk that can range from 10% to 40% at 5 years. That means that if a patient progresses, it will be to muscle-invasive disease that will may require a cystectomy. Considering that BCG is the best treatment available, there is really a high need for new treatments here. BCG has remained the standard treatment through the 2000s, and both the POTOMAC and CREST trials are based on the bioplausible theory that PD-1 expression may have a role in BCG resistance mechanisms. Basically, these trials are trying to [assess whether] adding durvalumab or sasanlimab will improve outcomes in these patients.

Can you compare the inclusion criteria and risk stratification approaches used in CREST vs POTOMAC, and how these differences may have influenced efficacy signals and clinical applicability?

The trials were practically identical. There are probably only 2 things that are worth mentioning. One is that POTOMAC used the European Association of Urology classification, which also includes multiple large, low-grade recurrent tumors, which are not usually included in high-risk categories, or at least not in the American classification. The other one is that POTOMAC actually allowed patients who have received BCG in the past if more than 3 years since the last dose have passed. We don't know how many of these patients were included. We don't know how many of them were low grade, and we don't know how many of them had BCG in the past, so we don't know how much this could have influenced the results, but considering that POTOMAC showed a little bit less absolute benefit, it could have influenced the final outcomes

Both trials used disease-free or event-free survival as primary end points. From a clinician's perspective, do these end points adequately capture what matters most to patients with high-risk NMIBC, particularly in terms of bladder preservation and quality of life?

I think they do. I think these trials capture everything that is important to these patients. These are hard outcomes that are really relevant for these patients. They want to remain disease free. They don't want to have more tumors. They don't want to have more transurethral resections. They don't want to have more intravesical treatment. They just want to be left alone with no more tumors. They want to avoid cystectomy. As much as patients don't like to have resections or have intravesical treatments, they do prefer to keep their bladders. If you ask a patient, they will all prefer to keep their bladders if possible. Of course, they want to keep their quality of life. They want to live their life in a full way. So I think these 2 trials captured all that matters to patients, and I think they were very well done in this area.

What were the most important efficacy signals observed in the CREST trial, and how should clinicians interpret these results when considering immune checkpoint inhibition earlier in the NMIBC treatment paradigm?

The CREST trial showed that BCG induction and maintenance, plus subcutaneous sasanlimab monthly for 2 years, can [result in a] 7% [absolute benefit in] the outcomes of these patients. They can be free from high-grade recurrence up to 7% more than with BCG induction and maintenance only. It also showed improved responses with carcinoma in situ, which is particularly difficult to manage. [Investigators] reported a 90% recurrence-free survival from carcinoma in situ at 3 years, vs only 67%, although we have to consider that the CREST trial included only a very limited number of patients with carcinoma in situ, which was different from the POTOMAC trial. [CREST] included only [approximately] 50 patients per arm. This could be only due to chance that they found this, but it's something that we cannot ignore. They had some very positive results when it comes to carcinoma in situ.

POTOMAC evaluated checkpoint inhibition in combination with BCG. What did this trial teach us about potential synergy between intravesical immunotherapy and systemic immune checkpoint blockade?

This was particularly interesting. The POTOMAC trial added durvalumab for only 1 year instead of 2 years, [as was seen in CREST with sasanlimab]. The [investigators] showed an absolute benefit of 4.4%. They showed that in adding durvalumab to BCG induction and maintenance for 2 years and durvalumab for only 1 year, they had the best outcomes. Arm B, where it was just BCG induction and 1 year of durvalumab, it was this arm got the worst results of the whole trial. So this has to be discarded as an option, but it was very good to try it. In arm C, it was the standard of care. So we can see that adding durvalumab to the treatment improves the outcomes in a very limited fashion, so it has to be considered that maybe it's not worth including adding [Durvalumab], but we can definitely say that the benefit of durvalumab will be present as long as BCG is still being given. So maintenance of BCG will be key for the benefits of durvalumab.

Given that high-risk NMIBC is often treated with curative intent, how do the immune-related adverse events observed in these trials affect the risk-benefit calculus compared with radical cystectomy or continued intravesical therapy?

As of now, BCG is the best treatment available. We can say that at 3 years, about 75% of patients will remain disease free with just BCG induction and maintenance. With these 2 trials [CREST and POTOMAC], we can say that we can improve that to 82%. That is either a meaningful percentage or not that meaningful, depending on how you want to look at it. Either the glass is half full or half empty. We can say that even if patients receive full treatment with sasanlimab or durvalumab added to their BCG induction and maintenance, 25% of patients who received immunotherapy with durvalumab or sasanlimab had grade 3 or 4 adverse events. That means that those patients required hospitalization, intensive treatment, and it was a life-threatening adverse event. That's something to take into account. Twenty-five percent of the patients in these trials had a recurrence, even with these treatments. So even if you got all the adverse events from the treatment, you could still end up with recurrence. In those patients, [the only other options] are a radical cystectomy or another clinical trial. We don't have anything else effective to give them. It's a very difficult situation when you have a patient with high-risk or very high-risk disease, and you can offer these treatments that have high toxicity and high risk of adverse events but also the best chance of keeping them disease free. Shared decision-making will be key with these scenarios and patients will have to know about these options.

Do the results of CREST and POTOMAC meaningfully inform decisions about delaying or avoiding radical cystectomy, or do they instead reinforce the importance of timely surgical intervention in certain high-risk subsets of patients?

Although the results of these trials do not directly inform us about this, we can say that it is safe to try these therapies in these patients in high- and even very high-risk disease, because progression to muscle-invasive disease in these trials was below 5% at 3 years. Neither of these trials showed that they could decrease progression compared with BCG only, but they do show us that trying these treatments is safe, as only a minority of patients will progress to muscle-invasive disease.

Based on the lessons learned from these trials, do you foresee immune checkpoint inhibitors gaining a defined role in future NMIBC treatment guidelines, or are we still in a exploratory phase?

I think these trials are solid enough to be included in the guidelines. Once these drugs gain authorization from the authorities, I think they should be offered to patients. Of course, these are not treatments that can be offered to everyone freely. These are treatments that are very toxic; 25% of patients will have a grade 3 or 3 adverse event. Some of these adverse events will be lifelong, so patients can end up with hypothyroidism or adrenal insufficiency for the rest of their lives, just for a treatment that aims at decreasing the chances of having a tumor coming back to their bladder, but not necessarily prolonging their life and not necessarily improving their quality of life. All things considered, we have to say that maybe the benefit that these drugs provide is not worth the chances of getting those adverse events. For example, neoadjuvant chemotherapy in muscle-invasive bladder cancer, where you can give a patient intravenous cisplatin and gemcitabine, those patients have an absolute benefit of 5% in overall survival. So it's practically the same percentage of benefit but for surviving, not dying from the disease. In this case [POTOMAC and CREST trials], they just have a lower risk of recurrence. Should these treatments be offered to every patient with high- or very high-risk disease? I think that's a matter of debate, and I think those patients will have to be very [carefully] selected, probably only in very high-risk patients, and probably those who are PD-L1-positive. Maybe in the future, when we have more publications on biomarkers and predictors of response from these trials, we will be able to choose which patients are the best to receive these treatments.

REFERENCE

1. González-Padilla DA, Subiela JD, Villacampa-Aubá F. Immune checkpoint inhibitors for high-risk non-muscle invasive bladder cancer: Lessons learnt from the CREST and POTOMAC trials. J Urol. 2025 Nov 25:101097JU0000000000004864. doi:10.1097/JU.0000000000004864