Real-World Use of Durvalumab + Tremelimumab in Advanced HCC

Transcript:

Josep Llovet, MD: Let’s talk about the STRIDE regimen and how we manage toxicities. The STRIDE regimen provides an opportunity in the front line to treat patients with checkpoint inhibitors if they’re at very high risk of variceal bleeding. This is an incredible plus. You were talking about gastric varices. There are also esophageal varices. Some physicians with big varices or intermediate—spots—do not want to wait. That’s 1 thing. In pathology, we check with an endoscopy to see if this is solved. You need 2 endoscopies, 1 at baseline and another to make sure that this is solved and that the wound is healing. This creates a lot of layers. Arndt, how do you handle the issue of varices in your flowchart and work-up for advanced HCC [hepatocellular carcinoma] management with checkpoint inhibitors?

Arndt Vogel, MD: It’s important to identify patients who are at risk of having varices. Your patient had platelets level of 111 per mm3. It wasn’t so severe, but he still had varices. With these indirect surrogates of portal hypertension, sometimes you need to be cautious. I strongly recommend specifically in … to look for varices. As a hepatologist, I can do it the next day. I don’t need to wait. The highest bleeding risk is after 7 to 10 days. We also need to keep in mind that after the ligation, not everything is good. The risk decreases after 2 weeks. In a perfect world, you would double check and see other varices treated completely. In many cases we do a re-treatment, which takes time. If you have to wait for the initial endoscopy, you have to do a re-ligation, it can take 4 to 8 weeks, which is a lot of time. If you have a high tumor burden or even portal infiltration, you don’t want to wait. This is clearly an argument to switch to another regimen and not wait for atezolizumab-bevacizumab.

Josep Llovet, MD: Any other comments regarding varices?

Amit Singal, MD: Arndt’s point about noninvasive markers not being applied to HCC is critical. Sometimes there’s a misconception that the Baveno VII criteria have been applied in non-HCC patients. Those have not been validated in HCC patients. This is an area of need, but the only way for us to do this is to check for it by EGD [esophagogastroduodenoscopy]. To Arndt’s point, even if you’re Child-Pugh A, sometimes there’s this assumption that he’s otherwise compensated. What’s the risk of varices? About 30% to 40% of Child-Pugh A patients have varices. Many of those can be large. For this type of thing, we need to be doing the EGD. Katie, you wrote a nice editorial that accompanied the IMbrave150 study that specifically talked about the importance of doing so. This is critical.

When we think of banding, it’s not just 1 and done. Banding is a serial process. It would be different if we didn’t have other therapies, but we do. We have other effective therapies. If you’re going through that banding protocol, then the STRIDE regimen and lenvatinib are both other possibilities in that patient population that we should be considering. [We should not be] overapplying what we regard as a highly efficacious therapy of atezolizumab-bevacizumab, but this is in select patients.

Katie Kelley, MD: You talked about Baveno VII for cirrhotic patients. But even noncirrhotic patients with rapidly progressive portal hypertension can have catastrophic upper–GI [gastrointestinal] bleeding. They’re coming in with a platelet level of 300,000 per mm3 and hepatitis B with no cirrhosis, no fibrosis, but a main portal vein tumor thrombus. With fast-progressing tumors, you don’t want to wait for an endoscopy. [That’s where] I’ve seen bleeding. The take-home message is that noncirrhotic patients should be considered for endoscopy, especially for main portal vein tumor thrombus.

Amit Singal, MD: Main portal vein invasion was 1 of the risk factors for GI bleeding on the IMbrave150 study. This is not only a risk factor for varices but also for variceal bleeding on atezolizumab-bevacizumab. This is 1 of the key factors. The other thing you brought up, which is critical, is that it’s not static. That patient who progresses and has new main portal vein invasion compared with baseline may have developed varices. You may need to reconsider an assessment of varices if somebody progresses and develops new main portal vein invasion, even if you have a “recent EGD.” These things can happen over time.

Josep Llovet, MD: Let’s talk about toxicity. When checking the toxicity, anti-CTLA4 is 1 of the agents that produces hepatotoxicity. But when checking the results, the adverse events, of the STRIDE [regimen], they were not significantly different from the durvalumab profile alone. I heard that 20% of patients in the trial were in need of high-dose corticosteroids. When are you starting that? What dose are you giving? Can you brief us on how you manage immune-related adverse events related to durvalumab treatment?

Katie Kelley, MD: I’m a believer in steroid requirement rate as a toxicity metric for all our I/O [immuno-oncology] trials. When we review papers, we should be asking for this. It’s a very pragmatic way to assess the impact of immune-related toxicity. As you said, it was just a little over 20% for the STRIDE regimen, which is much less than the high-dose ipilimumab from nivolumab-ipilimumab.

Josep Llovet, MD: Nivolumab-ipilimumab combination.

Katie Kelley, MD: This is up to 50%. In terms of the protocol for STRIDE, that guides my experience. I’ll extrapolate that into the real world should there be regulatory approval, which we expect and hope for. We usually use 1 mg/kg prednisone equivalent for sick patients with rapidly progressive liver function deterioration. I fortunately haven’t seen that, but you might admit those patients for intravenous steroids to make sure it’s delivered well. Then you wean them over time. For mild skin toxicity, I’d start with even a lower dose of prednisone or not start it at all. If it’s enough of a rash to warrant prednisone, maybe start with a lower dose. That’s been my practice, and most patients have responded well.

Transcript edited for clarity.