Molecular Testing in Metastatic Urothelial Carcinoma

Petros Grivas, MD, PhD: There’s so much variability in the assays testing PD-L1. Usually this happens in archival paraffin tissue from the primary tumor in most patients. There are different assays and scoring systems. Each has been linked with a checkpoint inhibitor. For example, the Agilent DakoLink 22C3 assay with a CPS [combined positive score] of 10 or higher, being positive, is linked with pembrolizumab, and Ventana SP142 with 5% or higher positive immune infiltrating cells is linked with atezolizumab. It’s important to not mix and match assays. You have to use the assay for the drug.

However, both you and I say that the utility of PD-L1 testing is much lower in our practices. There’s no use because we tend to use chemotherapy for available maintenance in the frontline setting if someone is fit for cisplatin or carboplatin. We tend to keep atezolizumab and pembrolizumab for patients who are platinum unfit or cisplatin and carboplatin unfit. With that indication in the United States, they allow the use regardless of PD-L1.

The other point, which you mentioned before, is the question of the best time to do molecular testing or next-generation sequencing. I agree with you in doing it at the time of diagnosis of metastatic disease in order to have available information down the road. For example, in use of target therapies, like erdafitinib, for platinum-refractory disease or other targets for clinical trials, I tend to do genomic sequencing somatic tumor profiling with a commercial vendor. In some cases, academic centers have their own assays. Other times, when we diagnose someone with metastatic disease, I don’t wait until later, so I can have the results available.

But the use of PD-L1 testing is becoming less frequent. As a community, we try to come up with other biomarkers and try to prove clinical utility in clinical trials. The other question I have for you is in terms of efficacy for platinum-based chemotherapy. Gemcitabine-cisplatin has about a 56% response rate, and gemcitabine-carboplatin has about a 40% to 50% response rate. Some additional patients get stable disease. I would argue about 75% to 80% of patients get response to stable disease with cisplatin or carboplatin. Would you agree with that approximation?

Neeraj Agarwal, MD: Absolutely. There’s no reason not to use chemotherapy in patients with metastatic bladder cancer, as long as there’s no clear contraindication. In my view, 90% of patients with metastatic bladder cancer should receive chemotherapy in the first-line setting in our practice when they come to us.

Petros Grivas, MD, PhD: It’s a great discussion. In some patients who have lymph node–only disease and are PD-L1 positive, they have very durable benefit from a checkpoint inhibitor. We have data from longer follow-up with pembrolizumab by Peter O’Donnell at the 2021 ASCO [American Society of Clinical Oncology Annual Meeting], showing that response rate with pembrolizumab alone in PD-L1–high patients is up to 48%. Overall response rate, regardless of PD-L1, is almost 29%.

But to your point, it’s hard to know a priori who those patients are who can reliably skip chemotherapy. Because we don’t have this perfect biomarker, I agree with you and tend to use chemotherapy. A checkpoint inhibitor is still an option. The FDA is reviewing the indication of pembrolizumab and atezolizumab in the frontline setting for cisplatin-unfit patients. There have been 2 ODAC [Oncologic Drugs Advisory Committee] meetings that favored the retention of this indication. Do you have any comments? It’s hard to predict what the FDA will do, but do you expect atezolizumab and pembrolizumab to stay and keep their indications in the frontline setting for cisplatin-unfit patients?

Neeraj Agarwal, MD: I think so. There’s no reason the indication should go away, because I don’t see any other option for these patients who are not chemotherapy eligible. This is still a patient population with unmet need. There’s nothing else to fill that void if you take away the approvals of atezolizumab and pembrolizumab. I welcome the continued approval of these 2 agents in this patient population.

Petros Grivas, MD, PhD: It’s a great discussion. I agree with you. It’s good to have options and access. There are patients who are certainly PD-L1–high. But mainly, patients who are platinum unfit make up to 20% of cases in community practices, where they cannot get platinum chemotherapy, cisplatin, and carboplatin, so it’s good to have this option available.

Of course, there are trials in that setting. We have the LEAP-011 trial testing the combination of pembrolizumab plus lenvatinib vs pembrolizumab plus placebo. That trial is very interesting because pembrolizumab-lenvatinib has shown great data in advanced kidney cancer in the CLEAR trial that we saw at ASCO GU [ASCO Genitourinary Cancers Symposium] and ASCO [Annual Meeting]. There’s also some data that you and Thomas Powles presented with cabozantinib and atezolizumab suggesting that this combination of anti-VEGF and anti–PD-1 and –PD-L1 has a good rationale. I’m very excited to see what happens with the pembrolizumab-lenvatinib trial. Do you have any comments? I know you did great work with the COSMIC-021 trial.

Neeraj Agarwal, MD: There’s definitely rationale for combining VEGF-TKIs, which are multi–tyrosine kinase inhibitors and have synergy with PD-1 inhibitors. There are promising data from the combination of cabozantinib plus atezolizumab. But we need confirmatory trials, as you said, before we can even think about using them in our practice.

Transcript edited for clarity.