TACE & Immunotherapy for HCC: Treatment Selection & Timing

[Dr. Christopher Beck]
Julius, before I ask a lot of the questions about TACE and how that goes to immunotherapy, would you just take a minute and explain TACE and how you do the procedure and just lay that out for us? Just a little gearhead talk.

[Dr. Julius Chapiro]
TACE, transarterial chemoembolization, is actually a very old procedure. It's been designed in Japan and then basically tested in patients mostly with what we today would describe as intermediate stage disease over the course of more than 30 years, well over 30 years actually now. The initial protocols have been done with lipiodol, which is basically an ethiodized oil which functions as a-- some people call it a triple threat, drug carrier embolic agent as well as imaging biomarker. That mixture, that is usually mixed in a emulsion with-- historically three drugs have been used. This is usually doxorubicin, cisplatin, and mitomycin.

Cisplatin was gone for a little while in powder form from the market and a lot of people have used one or two drugs. Standardization of the chemoembolic mix was a problem, but I think it was addressed fairly well over the past 15 to 20 years. I think when we do conventional chemoembolization or chemoembolization with lipiodol, we do it in a fairly standardized fashion and then after the administration of the mix, we close the back door of the feeding artery, usually with particulate embolics or gel foam. It has been demonstrated that particles are better and produce more durable responses than gel foam, which is a temporary embolic.

Then, of course, for a while, there has been interest in more, some people say, reproducible form of TACE and that was the drug-eluting beads TACE which was tested, introduced about 15 years ago and tested predominantly to avoid systemic spread of the chemotherapy that is being introduced and also limit the side effects and be in a more targeted fashion. While it has been demonstrated that it is able to reduce the systemic side effects and confine the chemotherapy more to the liver and to the targets, the real outcome metrics, the overall survival tumor response rates have not been much different from conventional chemoembolization.

Because the majority of the world from all intra-arterial therapies is still using chemoembolization, it has been also the therapy of choice for most of the randomized controlled trials and high-level data. It is still the only high-level intra-arterial therapy with level 1A evidence in HCC and has been that for a while. Predominantly, I think, driven by Asia and the availability of the therapy across the board. That's what I will say about C-TACE as a brief introduction, or C-TACE and DEB-TACE.

[Dr. Christopher Beck]
Can I just dig in a little bit though? In your practice, are you using DEBs or lipiodol?

[Dr. Julius Chapiro]
Except for the VA where lipiodol's a non-formulary, we're using almost exclusively lipiodol in our institution.

[Dr. Christopher Beck]
Okay. What's your drug cocktail?

[Dr. Julius Chapiro]
Cisplatin has been irregularly available, so we're using mostly a double drug cocktail. Unfortunately, some of my colleagues prefer to change it from time to time. Some people would mix it with beads when they administer the lipiodol. In my personal experience and the way how I do it, particularly because we like standardization, I feel that a standardized approach, highly standardized way how you mix it, what you use, what kind of drugs you use is extremely important. Because doxorubicin is the one thing that is always available, I mostly use a single drug cocktail with lipiodol.

Other people will say, why not using two or three? It's just because the remainder of the drugs are not consistently available. I believe that unless you do something the same way all over again, probably your results are not sustainably comparable. You have to have a rules-based practice, I would say.

[Dr. Christopher Beck]
Richard, I'd actually like to bring you in and just ask, do you have any commentary on the TACE procedure or with your interventional radiology colleagues? Because I can just say from personal experience, that is sometimes a knock against interventional radiology from our medical or surgical colleagues that they, depending on which interventionalist you send it to, you may be getting a TACE, but the TACE can differ between practitioners.

[Dr. Richard Finn]
Yes, I've been around a while, I finished my training in 2003 and joined the liver program even as a fellow. I've seen the evolution. I remember the Precision 5 study, and that was a big discussion in the space. I grew up in a center that was very much C-TACE based and have a lot of respect for my colleagues. At that time, Tony Gomes was probably our main vascular person. I worked very closely with Dave Lu and Steve Raman, who are more ablation people. As the program has grown, more people have come into the space, Justin McWilliams, and more recently in the past few years, Sid Padia, who has really brought Y90 to UCLA.

To me, the most important part of the procedure was the embolic part, to be honest. I know people have looked at TAE versus TACE, and there's various data. The exact drugs to me, there's nothing magical about doxorubicin or cisplatin or whatever, kinamycin. These are all cytotoxics. They're poisons. Certainly, I think when you're injecting a poison right into the tumor, as long as it's safe, it probably doesn't matter what you put in, because it's different than putting something in the vein that goes everywhere. Even in that context, a lot of times, putting combinations together of cytotoxics only improves toxicity and not necessarily outcomes.
That is why things like response, which you can always typically improve with combining things, cytotoxics don't always lead to improvements in survival outcomes. Given it's different here, it's going in the vasculature directly, I suspect that the embolic part, because liver cancer, you see that on your angiograms, it's a very vascular tumor. It's unique in that way. A lot of our drugs target the vascular nature of liver cancer. I do think the embolic part of the TACE is important. I wouldn't stand and tell an interventional radiologist how to do a procedure or what to do during that procedure.

I think they do it very well, and they know best. I think what you're referring to also is levels of evidence and data-driven studies, and I do have an issue with the field. Some people say, "Look, you don't need a clinical trial to know a parachute works."

[Dr. Christopher Beck]
That's a good line.

[Dr. Richard Finn]
Or people say, "Look, liver transplant, the initial study by Mazzaferro in New England Journal was 40 patients or less, and liver transplant became the standard. There's never been a randomized controlled study for a liver transplant. With that being said, there have been randomized studies going back to 2000 and 2002 from the Barcelona group and in Hong Kong, studies of TACE versus placebo. These were done, and they showed that TACE was better than placebo to improve survival in a randomized controlled study. It seems after that though, a lot of these studies have fallen off, these high-level evidence studies.

To that point, Precision 5, from what I remember, that was a randomized study of DEB-TACE versus C-TACE. There were some toxicity issues that Julius referred to, but really no benefit in survival outcomes, I don't think, and response rates, I think, were pretty much the same. Since then, a lot of the radiology literature is retrospective, single institution, multi-institution retrospective. There's value to that, but when you're talking about clinical practice versus guideline development, that's where you need higher level of evidence for the guidelines.

I think the issue that drives a lot of this is that approvals in interventional radiology things are devices. Even beads are a device from what I understand. The level required from the regulators to get these things approved is not the same as for a drug. You have a negative study with Brivanib, a drug that was being developed by BMS in combination with TACE, in front line, second line. The drug failed, it's not available, it's gone. The TACE study was stopped early because the front line study was negative, so BMS stopped development, so in the TACE study, stopped early, it's not approved.

Without, I think, that requirement, it's going to be hard to change the practice. I know people are academic and they want to do that and generate that data, but it has to be driven either by the government and the NIH to fund these type of studies or industries should have to do it in order to market something.

[Dr. Julius Chapiro]
I think HCC care is a team sport, and I think that it needs to be data-driven in. In all honesty, what's going on in our specialty, it gives me a lot of pause and, actually, headache, and I think we need to be data-driven, and we must not rely on just sort of like level 3B opinion and expert opinion evidence or worse, evidence that is based on conflict of interest or the billing codes that are available to us, right?

I think high-level data is extremely important. As of now, last year really gives us for the first time in a long time for IR, high-level data and support of TACE with immunotherapy, something really new, something that we can and need to investigate more, particularly the sequencing for-- of the combination. I'm enthused because I think that using both instruments that are available for us to induce complete response and think of cure, I think, is really important.

I think we as interventional radiologists really ought to think about the patients in terms of survival and in terms of actually tumor shrinkage, as we said, although I'm a big fan of mRECIST for studies. We need to think about the endpoint of median overall survival rather than a lot of other things. Although I know a lot of people in our community will say this is an imperfect endpoint in progression-free survival and time-to-progression tumor response are good surrogates, but being young in the field overall and having another probably, hopefully, 25 to 30 years of career ahead of me, I'm just thinking how to set myself up for long-term success. I believe that that is through, as Rich says, high-level data and relying on it consistently.

[Dr. Christopher Beck]
In your practice, Julius, so if-- I'm curious like how it does work. Are your oncologists, medical oncologists, interested in combination? They bring you in specifically because they're like, okay, and so I was curious. When do you end up treating these patients? When they're on immunotherapy, when they're on a pause?

[Dr. Julius Chapiro]
There are several patients where we're confident that we can achieve a complete response with a TACE, or we see a complete response, and there's sometimes just no reason to really expose the patients to an additional therapy at this point. Also, I always think about this as a little bit wasting my ammo before they progress or will need it down the road, right?

Secondly, I think that when we think that a TACE may provide a partial response or good complete response but the AFP is not ideal or there is some other small extrepatic disease or-- so there I think we combine a lot. I'm very grateful for having courageous medical oncologists here at Yale who are willing to go for the combination therapies. Sometimes, when there is an indication to start with immunotherapy, and this is started, and then there is a lesion that did not respond paradoxically as compared to other lesions in the liver, and it needs touching up, right, then we bring TACE back in.

It's no longer just one versus the other. It's a lot of really combinations in individual scenarios, but we're more and more trying to, by consensus, agree upon the combinations when they're needed.

[Dr. Christopher Beck]
Richard, this was a scenario that I actually wanted to specifically talk about. In your practice, how often is it that you'll make a referral to IR when you have a patient that's on immunotherapy, 10 lesions, whatever, nine of the 10 are responding great but then you have one lesion that's either growing or not responding at all? How do you think about that scenario?

[Dr. Richard Finn]
It's not such an uncommon scenario, and whether it's all in the liver or even, they develop an oligometastatic lesion in the adrenal or the lung. This is more common in patients who have been on for a long time. They've had some response, some tumor control, and then they have, perhaps, escape of one lesion, or they develop a poor vein thrombosis, in which case I've engaged the radiation oncologist.

I have referred back to IR because there's a data gap there now in second-line systemic treatment. We have a lot of drugs available, all the drugs that we had before IO. Lenvantinib, sorafenib, regorafenib, cabozantinib, ramucirumab, a lot of drugs available, but how well they work after frontline IO, we don't know. The biggest improvement in the modern age of taking care of people with this disease is from immunotherapy.

If patients are tolerating it, and also compared to all those drugs I just mentioned outside of ramucirumab, they're not so well tolerated. They have side effects that we don't see with IO. Hand foot skin syndrome, GI toxicity, anorexia. If a patient's doing really well on IO-based cocktail, and their tumor is very well controlled but they have one lesion that, for lack of a better word, is escaping, I would offer them some procedure with IR at least one time.

Give them a chance, because I think that growth of that lesion is either escape, or it's just a precursor to more progression that's going to happen in due course. I don't think there's a role for picking off five or six lesions, but if there's one lesion that escapes, and it also depends on that disease-free interval. You have that lesion treated, but it's different for a lesion that pops up in two months versus one that pops up a year later. I'm not going to rush and change their systemic treatment if-- I would probably send them back to IR.