Is Disease-Free Survival the Best Endpoint for Adjuvant Nivolumab in High-Risk, Muscle-Invasive Urothelial Carcinoma?

Get Permission

Derek Raghavan, MD, PhD, FACP, FRACP, FASCO

Derek Raghavan, MD, PhD, FACP, FRACP, FASCO

The role of adjuvant treatment for invasive, high-grade bladder cancer remains controversial and challenging. Sternberg et al reported a statistically significant progression-free survival benefit from adjuvant combination gemcitabine/cisplatin (GC) or MVAC (methotrexate, vinblastine, doxorubicin, cisplatin) chemotherapy after radical cystectomy in an underpowered EORTC study several years ago.1 Of note, this was a well-designed study, which was simply not supported by the clinicians of the investigating group (a consortium of EORTC, U.S. cooperative groups, and others); as a result, there was no statistically significant difference in overall survival when crossover patients were analyzed (ie, those who received GC or MVAC at the time of relapse). What was puzzling about the EORTC study was that the major clinical impact was among patients with advanced T-stage tumors who had negative nodes, which cast into doubt the quality of the surgery prior to adjuvant chemotherapy. Several other randomized trials and meta-analyses have failed to prove the role of adjuvant chemotherapy in this setting.2

The CheckMate 274 trial, recently reported by Bajorin et al3 and summarized in this issue of The ASCO Post, is a well-designed, randomized trial with appropriate power and statistics, as one would expect from its leadership. Among more than 700 randomly assigned patients with high-risk, resected invasive bladder cancer, the investigators showed impressive, clinically important, and statistically significant differences in disease-free survival between those allocated to adjuvant nivolumab (240 mg every 2 weeks) vs placebo. Patterns of toxicity were reported, and adverse events were significant but manageable.

Overall Concerns About CheckMate 274

I do think this was a well-run study. However, I do have some concerns:

The duration of follow-up is much too short to be definitive; this is consistent with the absence of overall survival data. However, the publication of the study in The New England Journal of Medicine, which usually implies a definitive answer, will give encouragement to some readers to believe the study has proven the utility and efficacy of adjuvant nivolumab for bladder cancer after resection.

The pattern of recorded toxicity suggests there may well have been underreporting of side effects experienced over a potential time frame of 6 months on treatment. This is not surprising, given the large number of investigators and sites involved internationally in the trial.

Only 4 of more than 700 patients were Black (no data regarding -Hispanic patients and only 1 Native American patient). Notwithstanding, this was an international study, and so that is very disappointing in an era focused on improving equity of care!

There are previous, provocative early data suggesting that adjuvant systemic therapy may be of benefit for upper tract tumors.4 Although just 20% of randomized cases had upper tract tumors, this had the potential to confound the overall analysis; this is more puzzling since a forest plot of the data suggests possibly worse outcomes in those with upper tract tumors.

The quality of surgery was variable, with more than 25% of N0 cases having fewer than 10 nodes resected.

It appears that PD-L1 assessment was made on any available urothelial cancer tissue. We have previously reported that there may be significant discrepancies between expression of PD-L1 and other targeted therapy markers in primary tumors and concomitant metastatic deposits in urothelial malignancy.5,6 It may be that this aspect of the study underreports the significance of this expression pattern.

Disease-Free Survival: Words of Caution

So, what about the price of gaining significant disease-free survival? The investigators chose the maintenance dose of 240 mg every 2 weeks, based on prior published maintenance regimens from other tumors, and this seems perfectly reasonable. For patients who remained free of disease, there was the potential for 26 doses per protocol. At a cost of around $7,000 for 240 mg of nivolumab, this comes to $182,000 per patient. That is pretty reasonable in the current high-tech/high-expense era, if there is a major increase in survival, notwithstanding that this figure represents drug cost alone.

What surprised me is that the journal published this paper without robust overall survival data, and that is my major concern. Everyone who designs adjuvant studies recognizes that effective treatments, of any type, will produce a disease-free survival benefit. Some years ago, Dr. Cora Sternberg gave an eloquent presentation of the previously mentioned EORTC adjuvant GC/MVAC randomized study at the plenary session of the ASCO Annual Meeting. This was followed by a somewhat controversial discussion and Q & A session that emphasized several problems with the analysis of that study. The absence of a statistically significant survival benefit due to the crossover was noted—that is, patients who relapsed on the control arm were treated with systemic chemotherapy, which brought the survival curves closer together, vitiating a statistically significant overall survival benefit. This demonstrates why I question disease-free or progression-free survival as valid final endpoints for the present study.

Disease-free and progression-free survival have become favored tools of investigators and pharmaceutical companies, and this may be appropriate when considering metastatic disease. When one reviews the extensive literature on adjuvant therapy, the innovations that have been incorporated into standard practice and have stood the test of time have all demonstrated clinically relevant and statistically significant increases in overall survival. This has not yet been demonstrated for adjuvant nivolumab for resected invasive bladder cancer. The treatment is expensive (a total nivolumab price for this study of around $64 million, which presumably was covered by the study sponsor); there are significant associated toxicities; and I believe the presentation of these trial results without overall survival data is premature, especially as randomized studies of similar targeted agents have been negative.

There is no doubt that immune-oncology agents such as nivolumab are important new treatments for advanced bladder cancer, but it remains unclear whether adjuvant use will improve cure rates. Although one understands that fiscal cost should not preclude optimal care for patients, the expenditure of $64 million (at least) for the next group of 353 patients receiving such treatment should be justified by a proven overall survival benefit.

Closing Thoughts

Finally, again, this study was well designed. My key concerns relate to the critical need for those who recruit patients internationally to such studies to make greater efforts to ensure representation of all races and ethnicities, and this is also a responsibility of the study sponsors. Also, it appears that the U.S. Food and Drug Administration has lowered the bar for acceptable marketing data to focus on disease-free survival in the adjuvant context, and one could easily question whether that is a data-driven or politically driven stance. 

Dr. Raghavan is President of the Levine Cancer Institute at Atrium Health.

DISCLOSURE: Dr. Raghavan has served as a consultant or advisor to Gerson Lehrman Group and has served in as an institutional consultant or advisor to Caris Life Sciences.


1. Sternberg CN, et al: Immediate versus deferred chemotherapy after radical cystectomy in patients with pT3-pT4 or N+ M0 urothelial carcinoma of the bladder (EORTC 30994). Lancet Oncol 16:76-86, 2015.

2. Raghavan D: Chemotherapy for invasive bladder cancer: Five simple rules learned over 30 years. Bladder Cancer 1:3-13, 2015.

3. Bajorin DF, et al: Adjuvant nivolumab versus placebo in muscle-invasive urothelial carcinoma. N Engl J Med 384:2102-2114, 2021.

4. Birtle A, et al: Adjuvant chemotherapy in upper tract urothelial carcinoma (the POUT trial). Lancet 395:1268-1277, 2020.

5. Burgess EF, et al: Discordance of high PD-L1 expression in primary and metastatic urothelial carcinoma lesions. Urol Oncol 37:299.e19-299.e25, 2019.

6. Grigg CM, et al: Human epidermal growth factor receptor 2 overexpression is frequently discordant between primary and metastatic urothelial carcinoma and is associated with intratumoral human epidermal growth factor receptor 2 heterogeneity. Hum Pathol 107:96-103, 2021.

Related Articles

Adjuvant Nivolumab Improves Disease-Free Survival in High-Risk Muscle-Invasive Urothelial Carcinoma

Dean F. Bajorin, MD

Dean F. Bajorin, MD

As reported inThe New England Journal of Medicine by Dean F. Bajorin, MD, of the Department of Medicine, Memorial Sloan Kettering Cancer Center, and colleagues, an interim analysis of the phase III CheckMate 274 trial has shown improved disease-free survival with adjuvant...