Skip to main content


Neoadjuvant Immune Checkpoint Inhibitors Unlikely Cost-Effective for Muscle-Invasive Bladder Cancer

February 26, 2021


Ali Khaki, MD, Stanford University, CA, discusses results from a cost-effectiveness analysis comparing neoadjuvant immune checkpoint inhibitors with cisplatin-based chemotherapy for muscle-invasive bladder cancer.

These results were presented at the virtual 2021 ASCO Genitourinary Cancers Symposium.


Hi, I'm Ali Khaki. I'm a clinical assistant professor at Stanford University, and I'm pleased to share with you all some work that we presented at GU ASCO 2021. The title of our abstract was "Cost Effectiveness Analysis of Neoadjuvant Immune Checkpoint Inhibition Versus Cisplatin‑Based Chemotherapy in Muscle‑Invasive Bladder Cancer."

As a way of background, neoadjuvant cisplatin‑based chemotherapy is a standard of care for treatment of patients with muscle‑invasive bladder cancer. However, recently, multiple single‑arm phase 2 clinical trials using neoadjuvant immune checkpoint inhibitors have shown very promising pathologic complete response rates for patients with muscle‑invasive bladder cancer.

We aim to compare the cost effectiveness of neoadjuvant immune checkpoint inhibition to cisplatin‑based chemotherapy in this population. We used the decision‑analytic simulation model, and in our poster, we show our decision tree.

Our comparison groups were immune checkpoint inhibitors and cisplatin‑based chemotherapy, and for our primary analysis, we compared pembrolizumab, based on the PURE‑01 trial, to dose‑dense MVAC, a commonly used cisplatin‑based chemotherapy regimen in the neoadjuvant setting.

We also modeled separately the immune checkpoint inhibitor atezolizumab based on the ABACUS trial and the combination of nivolumab and ipilimumab based on the NAMUKO trial and also used as an alternate cisplatin‑based chemotherapy regimen gemcitabine and cisplatin.

The outputs to our model were costs, pathologic complete response rates, 2‑year recurrence‑free survival, and the incremental cost-effectiveness ratio, which we defined as the costs per 2‑year recurrence‑free survival saved.

The perspective we used the health care payer perspective and the time horizon was 2 years. The data sources that we used for our modeling was we got pathologic complete response rates and 2‑year recurrence‑free survival rates from clinical trials.

Though none of the immune checkpoint inhibitor trials have presented 2‑year RFS data so far, so we extrapolated this from the chemotherapy trials, and costs, which were used from the average sales price, published by the Center for Medicaid and Medicare Services.

We also performed sensitivity analyses, which included a 1‑way sensitivity analysis, a probabilistic sensitivity analysis, and a scenario analysis, where we evaluated the value‑based price needed for pembrolizumab to be cost‑effective.

In our poster, you will see in the middle panel our main results. The costs of a neoadjuvant course of dose‑dense MVAC was estimated to be about $22,000 and was associated with a 2‑year recurrence‑free survival of about 57.5%.

Conversely, pembrolizumab, a neoadjuvant course was estimated to cost $30,500, or about $8000 more than dose‑dense MVAC, with an improvement in 2‑year recurrence-free survival of about 1.5%. This resulted in an incremental cost‑effectiveness ratio of $552,000 per 2‑year RFS saved.

Similarly, nivolumab and ipilimumab and all 3 checkpoint inhibitors, when compared to gemcitabine and cisplatin had incremental cost‑effectiveness ratios that exceed half a million. Actually, it exceeded $1 million per 2‑year RFS saved.

The one exception was atezolizumab, when compared to dose‑dense MVAC, where atezolizumab was incrementally less costly, only $18,838 per neoadjuvant course, and incrementally more effective, with improvement in 2‑year recurrence‑free survival of 0.2%. It received a dominant designation and was considered to be more cost‑effective than dose‑dense MVAC.

We also present in our poster one-way sensitivity analyses that identify the costs of both dose‑dense MVAC and pembrolizumab, as well as the probability of no relapse without PCR as the factors in our model that the model is most sensitive to, and probabilistic sensitivity analysis, where we estimated about 11% of simulations that we performed with pembrolizumab compared to dose‑dense MVAC met this $100,000 willingness to pay threshold that we had targeted.

We also did a scenario‑based analysis, where we looked for a value‑based price for pembrolizumab to be cost‑effective and found that, if pembrolizumab was 21% cheaper, then it could be considered cost‑effective when compared to dose‑dense MVAC, and over 90% cheaper, or less costly, I should say, it could be considered cost‑effective when compared to gemcitabine and cisplatin.

The one notable finding was that atezolizumab was cost‑effective, and this was likely driven by the fewer doses of atezolizumab, only 2 doses in the ABACUS trial, whereas in the PURE‑01 trial and the NAMUKO trial, there were more doses of the other medications.

In conclusion, we show in our study neoadjuvant pembrolizumab and other immune checkpoint inhibitors in general do not seem cost‑effective, with the exception of atezolizumab. This was likely driven by fewer doses of that medication.

Pembrolizumab may approach a cost‑effective threshold with a 21% rebate when compared to dose‑dense MVAC. The limitations we should highlight in this study is that we used a projection model and not real‑world data.

We're also limited by single‑arm trials, and so this is not based off randomized trials. This is based off very early data in this space. Thank you. 

Khaki AR, Shan Y, Nelson R, et al. Cost-effectiveness analysis of neoadjuvant immune checkpoint inhibition (ICI) versus cisplatin-based chemotherapy (CBC) in muscle-invasive bladder cancer (MIBC). Presented at: the virtual 2021 ASCO Genitourinary Cancers Symposium; February 11-13, 2021. Abstract 419.

Dr Khaki reports no relevant financial relationships.

Back to Top