Ghassan K. Abou-Alfa, MD: Riad brought up a very important point: endpoints. We are all more inclined to use overall survival because it’s really the ultimate reference. People are alive and are doing well. But, of course, there’s the in-between of many other things. To simplify the question and make sure we put it in context, why would you say that there are certain arguments for pushing for progression-free survival, for example, in clinical trials in HCC [hepatocellular carcinoma]?
Amit Singal, MD: Before I get to that point, I just want to say one other thing. We have more and more tools that we’re comparing, right? You already have all of the embolization therapies. We talked about bland embolization, chemoembolization, radioembolization. Now you also have SBRT [stereotactic body radiation therapy] as a local therapy. I think this is a theme that you’re going to hear throughout this discussion today. We have to be more nuanced, in terms of how we discuss these things, particularly as we start to say these are relatively the same in terms of overall survival. To start to push the field forward, we need to look at either radiologic markers of response or biomarkers so we can actually start making smarter decisions. We won’t end up depending on local expertise but will try to make the right decision for the individual patient in front of us. I know that we’re not there right now, but I really think that’s where we have to push the field to go in the future.
Now, to get back to your question, in terms of these intermediate outcomes and why these are important. Riad talked a bit about this. Particularly for local therapies, when you talk about BCLC [Barcelona Clinic Liver Cancer]—A and B patients, it could be tough to compare therapies when you have the postprogression therapies that confound any kind of interpretation of these comparative studies. And so, when you take a look at this, you can use some of these more intermediate outcomes, like time to progression. Or if you’re looking at the radiation literature, you can look at these local controls as a way to compare these therapies in terms of their actual efficacy for what they’re trying to achieve.
Ghassan K. Abou-Alfa, MD: Yes, this is a bit debatable. If anything, as Riad said, there’s quite a bit of interest in regard to what those other endpoints mean. Recently, we heard about TTUP. Riad, tell us a bit more about TTUP. It came right from the left field and is rather intriguing.
Riad Salem, MD: So TTUP means time to untreatable progression.
Ghassan K. Abou-Alfa, MD: Untreatable progression. I need time to even understand that.
Riad Salem, MD: Yes. Let me take some time to try to explain it as best I can. The idea here is that when you’re performing an embolotherapy, there has to be a right balance as to how many embolotherapies you implement to maximize response, overall survival, purported overall survival, and permit the patient to be well enough and to be treated early enough with systemic therapy. That’s where the balance is. And so, the concept of TTUP comes in saying, “Hey, I want to make sure that I treat but don’t have a patient decompensating to Child-Pugh B or Child-Pugh C. I don’t want to embolize until, for example, I’ve occluded all of the vessels.” It means that you don’t want to be in a situation where you want to do a locoregional therapy like embolization and the vessels are thrombosed or the patient is now Child-Pugh C.
The idea of TTUP is, “Let me try to maximize the benefit for the patient. When they’re still well enough, let’s initiate systemic therapy.” That’s the concept. It’s intriguing, and it’s interesting. Unfortunately, one of the issues that arises is the variability in the Child-Pugh score and the variability in some of these definitions that people have used. TTUP, by my review, has already been defined in 3 or 4 different ways. I think leaders and scientists have to get together to figure out, if we’re going to use that, what the definition is. And then we can all follow that same sort of pattern and proceed forward.
Ghassan K. Abou-Alfa, MD: Peter, do you have any favored endpoints? Do you think this is the right one to use, in regard to those studies?
Peter Galle, MD, PhD: It was very nicely pointed out by Riad that this is a very individualized, ill-defined sort of situation. However, this is clinical reality. We know that with progression, if you have it in a sharply defined situation—something where you stop a clinical trial because the endpoint was reached—you might still continue with the same drug. We have been doing that. We have seen patients reach the clinical endpoint of progression, and we have kept them on a drug for 2 years after the trial because we saw a drop in alpha-fetoprotein and the patient was doing fine. So progression is very complex in hepatocellular carcinoma. Several trials have tried to capture that. They allowed for clinically relevant progression, or something like that, so that you can continue because you consider that, although it’s objectively speaking “radiographic progression,” to be nothing that would prevent you from continuing.
Coming to endpoints, we had that discussion. It’s driven by oncologists that we need something, particularly in a situation where we go from first-line to second-line to third-line therapy. There’s so much along those lines that is impacting therapy, and you need to have something in addition to overall survival. Progression-free survival is probably what most people then consider to be relevant, but the trials tell us that those surrogate markers are not always translating into the real endpoint—overall survival. Honestly, I’m stuck. I’m a bit confused on what to recommend. It’s the big question that you are asking during any discussion with Big Pharma: What endpoint would you recommend? To come back to the initial discussion, it’s very relevant to keep patients on a drug although they might have progressed. That requires a dose reduction, or a break, and time and a second assessment. Determining how to bring that into a clinical trial is a good question.
Riad Salem, MD: I just want to add one other thing, if I may, about endpoints. For me, this is one of the most important issues that leaders need to address in the field: to define endpoints that are going to be practical so the field can continue to move forward. It’s long been my contention as an interventional radiologist working mostly with BCLC-A and B patients, where the survival is very long, that I’m sort of stuck, as Peter was saying. I’m statistically paralyzed because I can’t do 5000 patient trials. The BCLC-C field is so dynamic, with 5 or 6 or 7 trials. My postprogression argument that I used 10 years ago with sorafenib [Nexavar] is now complicated by 5 other post progressions. It’s impossible to perform an overall survival study in those fields. The objective endpoint is the best one, if we can use it, but we’re statistically stuck.
We need to find endpoints that are relevant. We need to find endpoints that people will act on, that are relevant for patients, to move the field forward. There’s a reason why there are only 2 TACE [transcatheter arterial chemoembolization] studies from 20 years ago that we still quote. They were controlled to no treatment, and nothing else has happened. Those were in small, single centers. There were 40 patients. Nothing has happened since then. There’s a reason. It’s the statistics, and it’s the need for this overall survival endpoint.
Transcript Edited for Clarity