.png){kind=small}
.png){kind=small}
BackTable / VI / Article
HEPZATO for Uveal Melanoma Metastasis: Patient Selection, Dosing & Treatment Cycles
Audrey Qian • Updated Aug 25, 2025 • 35 hits
Uveal melanoma metastasis is the spread of uveal melanoma – a rare eye cancer originating from the uveal tract of the eye – to other parts of the body, most often to the liver. About 40-50% of patients with uveal melanoma experience metastasis within 15 years of diagnosis. While treatments like radiation may control the original tumor localized in the eye, systemic treatment including chemotherapy and chemoembolization are needed to treat liver metastases.
One minimally invasive and more targeted treatment for metastatic uveal melanoma is HEPZATO, a therapy delivered via percutaneous hepatic perfusion (PHP) directly to the liver, minimizing systemic chemo-exposure. Interventional radiologists Dr. Altan Ahmed and Dr. Siddharth Padia cover how practitioners may optimize HEPZATO therapy, including determining the most suitable patients and considering the most efficient drug dosages and treatment cycle number.
This article features transcripts for the BackTable Podcast. We’ve provided the highlight reel here, and you can listen to the full podcast below.
The BackTable VI Brief
• HEPZATO (or CHEMOSAT) delivers high-dose melphalan directly to the liver via percutaneous hepatic perfusion (PHP), isolating hepatic circulation with a double-balloon catheter and extracorporeal filtration to mitigate systemic toxic exposure.
• Ideal candidates for HEPZATO have liver-dominant disease with <50% hepatic involvement, ECOG of 0-1 status, and preserved hepatic function (Child-Pugh A). Patients with altered biliary anatomy or significant cardiac comorbidities require careful consideration due to procedural demands and lack of clinical trial data.
• The current dosing model (3 mg/kg for ideal body weight) does not account for individual tumor burden and hepatic volume, prompting for a more physiologically grounded, stratified dosing approach based on hepatic and disease metrics.
• Clinicians are moving away from fixed-cycle regimens, instead tailoring the number and timing of PHP cycles based on patient response, tolerability, and disease progression. This patient-paced strategy maintains control and preserves bone marrow reserve over a longer duration of time.
• PHP’s non-necrotic, liver-sparing profile enables safe combination with other local or systemic therapies, including Y90, Tebentafusp, and immune checkpoint inhibitors, without adding onto overall toxicity. This approach may contribute to the advancement of personalized treatment strategies for patients with uveal melanoma liver metastasis.
Table of Contents
(1) The HEPZATO Kit
(2) Patient Selection for HEPZATO Therapy in Uveal Melanoma Metastasis
(3) HEPZATO Dosing in Percutaneous Hepatic Perfusion
(4) HEPZATO Treatment Cycles in Uveal Melanoma Metastasis
The HEPZATO Kit
The HEPZATO system is used to deliver high-dose melphalan to the liver via a percutaneous hepatic perfusion (PHP) procedure to treat metastatic uveal melanoma. Unlike surgically implanted infusion pumps, PHP adopts a filtration mechanism, using a double-balloon catheter system to deliver melphalan directly into the hepatic artery before immediately withdrawing the drug-laden blood through the hepatic veins. The blood is filtered extracorporeally and then returned systemically, mitigating whole-body toxicities, especially to the bone marrow.
Originally developed from the concept of isolated hepatic perfusion in the 1960s, PHP has progressed into a minimally invasive, repeatable, and safer method of regional chemotherapy delivery. Its 2023 FDA approval for uveal melanoma highlights a critical advancement in liver-directed therapy for this disease.
[Dr. Venkatesh Krishnasamy]
All right. Let's get started. We're going to, like I said, talk about HEPZATO or CHEMOSAT as it's called in Europe. Tell me, Sid, what is HEPZATO? What is CHEMOSAT?
[Dr. Siddharth Padia]
HEPZATO, CHEMOSAT, generic name is PHP or Percutaneous Hepatic Perfusion. I would consider it relatively new in the US just because it got FDA approval last year and is a device related treatment for patients with liver metastases from uveal melanoma. It is a way to deliver high dose of chemotherapy into the liver.
In this case, the chemotherapy drug is called melphalan, which is a DNA alkylating agent. The device is designed for us to deliver this via the hepatic artery, just like we do many other hepatic artery procedures. The difference is that we are withdrawing the drug at the same time via the IVC, filtering the drug out through a filtration circuit, and then returning the filtered blood back in.
This whole device, including the drug, is called-- the trade name is HEPZATO kit in the US, which was FDA approved in 2023 for uveal melanoma metastases, and then became commercially available in 2024.
[Dr. Venkatesh Krishnasamy]
Got it. Probably to the average practitioner in the US, it sounds a little bit like a surgical hepatic pump, but this is obviously very different. How is it different?
[Dr. Siddharth Padia]
I think there's a couple of differences. Number one, when we talk about infusion pumps in surgery, which actually our group does a good number of them in the colorectal cancer setting, the difference is that the drug that we're using in this case, which is melphalan, is very cytotoxic, potentially, and especially in terms of bone marrow suppression.
The systemic dose needs to be minimized. The difference between a surgically implanted pump and this is-- the main difference is that we are withdrawing the drug after the first pass through the liver, and then filtering the drug out through this filtration circuit. That's number one. Number two, this is given all at once during a procedure.
Then what allows us to do this in IR appropriately is that we're able to repeat this down the line because this is all percutaneous. By percutaneous, meaning femoral artery, femoral vein access, things that IRs are very used to and very comfortable with for the last several decades. This allows us to repeat treatments for patients, which in the past, when this same procedure was done in the surgical setting, you really can't do this operatively multiple times.
[Dr. Venkatesh Krishnasamy]
Got it. No implanted device. Then multiple procedures, like you said, so big difference, and lends itself to our skill set, which is great. Altan, Moffitt has the longest running US experience for PHP or HEPZATO. We already talked a little bit about when it was FDA approved, but how long has the device been around? How long have you been using it at Moffitt to treat patients? Great question.
[Dr. Altan Ahmed]
Great question. Yes, we have about over a decade now, of experience with Percutaneous Hepatic Perfusion. The device itself, in one form or another, has been around since the early '90s, actually. First described in Japan and MD Anderson around the same time, early '90s, in animal models, and then '93, '94 is where the first human trials were conducted.
In some form, the double-balloon catheter for hepatic venous isolation has been around and evolved over time, not only the catheter itself, but also the filtration mechanisms with the carbon-based system. Really an evolution of a technology that's been around for quite some time. Then even before that, if you look at isolated hepatic perfusion that Sid had mentioned there, that has been described as far back as 1960.
An evolution of that, a surgical procedure that could only be performed once, great response rates, high morbidity. This is a natural evolution in something that can be repeated and continue to be efficacious for patients.
Listen to the Full Podcast
Stay Up To Date
Follow:
Subscribe:
Sign Up:
Patient Selection for HEPZATO Therapy in Uveal Melanoma Metastasis
Patient selection for PHP with HEPZATO relies on well-established criteria and liver function metrics, including ECOG performance status of 0-1, Child-Pugh A liver function, and liver-dominant disease with ideally <50% hepatic involvement. Unlike embolic therapies, PHP does not cause hepatic necrosis, potentially reducing the risk of infection in altered biliary anatomy. However, clinical trial data for this subgroup is still needed, as patients with prior biliary surgery were excluded from studies.
As general anesthesia and hemodynamic monitoring are required when performing PHP, other considerations for patient selection include cardiac risk assessment, and general pre-surgical clearances. Although transient troponin leaks are often found incidentally during procedures like this, it’s important to note that they are clinically insignificant in the absence of other findings.
In terms of concurrent treatment modalities, PHP appears to be compatible with systemic therapy, as there are no known adverse interactions with the commonly co-prescribed immune checkpoint inhibitors. However, ongoing trials like CHOPAN are still investigating the combined use of PHP and immunotherapy.
[Dr. Venkatesh Krishnasamy]
Over that time, obviously, like all interventions, we evolve, our patient selection gets better. Tell me, how are you choosing your patients? ECOG, Child-Pugh, ALBI scores, give me some details.
[Dr. Altan Ahmed]
Overall, not very different from most of our other local regional therapies. Liver dominant disease number one from a tumor perspective, ideally less than 50% of the liver being involved. ECOG zero to one, we don't want major cirrhosis. Child-Pugh A, B and C would usually be excluded.
Baseline hepatic synthetic dysfunction would usually be an exclusion criteria for patients. Patients who cannot undergo general anesthesia, this is going to be one key difference from our other local regional therapies. This is a pretty intensive procedure for a patient, general anesthesia, perfusion bypass, all that. They have to be relatively healthy otherwise.
[Dr. Venkatesh Krishnasamy]
And cardiac strain. I think patients that have underlying cardiac morbidities, the blood pressure liabilities is there as well. What about things like-- you mentioned tumor burden. What about biliary incompetence, post-whipple, biliary stents, things like that? Those are areas where I will sway away from TACE and ablation, even with long-term antibiotics.
Then Sid and I recently had this conversation about radiation segmentectomy in this type of patient as well. How about that patient population here?
[Dr. Siddharth Padia]
Fortunately, it doesn't really happen that often. Most of these patients, keep in mind, we're talking about a very targeted group of people who have liver metastases from melanoma. These patients in general are not going to have prior biliary surgery. If they did, it's most likely going to be unrelated to their cancer and perhaps related to something else. We haven't really encountered that.
The trials, which we'll talk about later, did exclude those patients just because you want to optimize your safety during a trial. We don't really know what's going to happen the first time we have that scenario. I suspect it's going to be fine because we're not causing tissue necrosis. The challenge in doing radiation segmentectomy or thermal ablation or chemoembolization in the setting of prior biliary surgery is that you're leaving necrotic tissue behind when there's reflux of contents in the bile system. You're going to get that necrotic tissue infected.
This doesn't because any necrosis. I suspect, we would probably put them on antibiotics, but I suspect it's not going to be a big issue. We haven't had any experience with it.
[Dr. Venkatesh Krishnasamy]
For both of you, I guess, if you do see a patient that's on systemic therapy, what are you saying? Ipinevo? Tibentifasp? Anything else?
[Dr. Siddharth Padia]
Those are probably the two most common. There's others. We have not discontinued their systemic therapy during the course of this treatment. To our knowledge, we haven't had any interactions or adverse interactions between the immune checkpoint inhibitors and this. There are ongoing trials. One is in Europe called the CHOPAN trial, looking at the combination of immune checkpoint inhibitors and cutaneous hepatic perfusion, which results will hopefully be out later in 2025.
[Dr. Venkatesh Krishnasamy]
Yes, I think we're all excited to see that. I think the phase 1B data came out last year at ASCO and they're hoping to publish a phase 2B data this year at ASCO, so that could be very impactful for this patient population. Altan, I want to go back to one thing real quick, talked a little bit about cardiac co-morbidities.
Your group recently published troponin leaks after PHP, if I'm not mistaken, showing no long-term consequences and them all being subclinical. Again, obviously, cardiac strain during the intervention, which we'll get to here shortly, can you tell me a little bit more about that?
[Dr. Altan Ahmed]
Sure. You'll often find troponin leaks in these patients if you're checking the labs afterwards. We found that, overall, if there's no other sign other than the biochemical finding that it's inconsequential for these patients. ICOs have not shown anything. Typically, you will see a little bit of troponin leak just from the stress induced by the procedure itself.
There's a lot of hemodynamic shifts, a lot of changes happening during the procedure. Not totally unexpected, but as long as you're selecting patients appropriately, you wouldn't expect to have any adverse effects long-term.
HEPZATO Dosing in Percutaneous Hepatic Perfusion
Melphalan dosing via the HEPZATO system is currently standardized at 3 mg/kg (based on ideal body weight) with maximum dose being 220 mg – per parameters derived from early-phase NIH trials in 2005. While this approach achieved FDA approval in the FOCUS trial, Drs. Padia and Ahmed question its clinical logic and adaptability.
The use of ideal body weight – calculated from height and sex rather than liver size or tumor burden – may be a metric that does not fully consider the extent of the disease or liver physiology. For instance, uniform dosing across patients with 2% versus 40% liver tumor burden highlights the lack of patient case stratification. Given these challenges, there is a need for a shift toward more nuanced dosing algorithms that incorporate tumor burden, hepatic volume, and patient-specific recovery profiles.
[Dr. Venkatesh Krishnasamy]
Sid, going back to what you mentioned about thermal ablation and Y90, I think that's a great segue. Altan, the drug dosing is part of this as well. In the FOCUS trial, which is a landmark trial that got HEPZATO FDA approval, it was just a flat dose above a certain body weight.
Tell me a little bit more about the drug dosing. Do you ever dose reduce in certain situations in these patients? I know that's a little bit of a loaded question because we don't know the answer to that probably at this point. Then is there going to be a role in the future to as opposed to just dividing the dose between right and left liver dosing to tumor burden as we do with Y90?
[Dr. Altan Ahmed]
Great and very important question. Yes, there is a dose that came out of the phase one trial at the NIH back in 2005. That dose was three milligrams per kilogram for ideal body weight. Now they tested anywhere from 2 to 3.5 and 3.5 is where they had the most adverse events.
Three was selected as the dose to a maximum of 220 milligrams. Ideal body weight is the key thing here. Now, when we do dose reduce, there are instances where that is done typically in patients who have adverse events, mainly neutropenia. Delayed recovery in between treatments or they have neutropenia leading to an infection or something like that or fibromyalgia neutropenia.
If they have delayed recovery over six to eight weeks for their white counts to bounce back or recover, then we would consider that patient for a 2.5 milligram per kilogram dose versus the standard three. Sid, I'd like to definitely hear what your practices have been so far.
[Dr. Siddharth Padia]
Thanks for the recommendation on that. I appreciate it because right now in my opinion, the dosing is terrible, to be honest, but that also means it leaves a lot of room for improvement and optimization. As Altan mentioned, it's 3 milligrams per kilogram per and kilogram is measured in ideal body weight.
When you look at the specific instructions for ideal body weight, it makes no logical sense on how they came up with this equation for this specific drug. It's really based on the patient's height and sex, whether they're male or female. In other words, in my inexperienced opinion and not having done several hundred of these, the dosing seems very antiquated.
It reminds me of back 15 years ago when we were using radioembolization and doing it based on BSA formulas and having arbitrary cutoffs for lung shunt fraction. That's what it brings me back to. The good thing is that the FOCUS trial used this dosing data and got their published results. In theory, if we can try to optimize the dosing over the next two to three years, maybe we can even get better results.
Can we optimize the dose based on the patient's tumor burden, based on liver volumes instead of body weight, and so on and really figure out how to drop the dose in those cases of neutropenia, as Altan mentioned. Is 2.5 sufficient or do we keep the same dose? Is it really dose dependent and so on?
All these variables, I think we can optimize in the next two to three years, especially as more centers come online and we try to collaborate and share data, that hopefully, the results will get even better than they are right now.
[Dr. Altan Ahmed]
Yes, definitely a big opportunity there Covey and Sid, as you mentioned, to start looking at tumor volumes and liver volumes and see if there is a better way to do this.
[Dr. Siddharth Padia]
To me, it makes no sense if you have 40% tumor burden versus 2% tumor burden, we're giving the same dose.
HEPZATO Treatment Cycles in Uveal Melanoma Metastasis
The optimal treatment protocol for HEPZATO remains elusive, but several European studies show promise extending the FOCUS report’s 6-cycle framework into a longer 8-9 treatment cycle program. U.S. physicians are beginning to adopt a more adaptive, patient-centered strategy where treatments are spaced based on tumor response and overall tolerance rather than complete responses.
The general clinical strategy in this treatment focuses on holding further treatment after a meaningful response from the patient (usually after two to four cycles).before resuming upon signs of progression to avoid overtreatment and marrow toxicity. Given its key advantage as non-hepatotoxic, PHP can be integrated with other therapies, such as radioembolization or systemic immunotherapies without adding additional hepatotoxicity.
[Dr. Venkatesh Krishnasamy]
What about cycles of treatment? The FOCUS trial was six cycles of treatment or up to six cycles of treatment. Not everybody got all the way to six. There's some patients that in Europe, Europe has had CHEMOSAT, which is what they call it for over 10 years now.
They've got up to, Altan, I don't know off the top of my head, eight to nine maybe cycles, something like that. Where do we envision that going in the US?
[Dr. Altan Ahmed]
This is one place where I feel like there is going to be improvement, especially with CHOPAN and some of the combinatorial therapy trials coming out soon. Is it necessary to go all the way to six treatments or can we get to a steady state where we have response and patients are stable?
They either have, hopefully a complete response, but that's not going to be the most common finding. They have a significant response. You do your combination therapies of PHP plus whatever else may be there. Then do you watch and wait for a little while or do you just keep going until you have progression?
I don't know ultimately what the answer to that is going to be, but Covey, I think, some of our discussions with our European colleagues has been that they've started to err on the side of fewer treatments and watching and waiting. Sid, what has your practice been like so far?
[Dr. Siddharth Padia]
It's funny because I think the three of us with some others had this email conversation several months ago about, when do you pull the trigger for the next treatment? I think we're all over the map and I don't know if any of us are right because I don't know if there's a right answer. We've been playing a more conservative approach to what you're saying.
I'm not trying to get a complete response on everyone and just going, you're going to get six treatments up front. I see this more of a marathon and not a sprint. This is not curative in intent, just like most therapies for stage four cancer. The goal is to get longevity while preserving their quality of life.
To me, the best thing that makes sense is to stretch these out a little bit, to try to get some objective response, but then maybe perhaps hold. We don't even necessarily go to four treatments if we're getting a response after two. Is that the right answer? Only time is going to tell. It does remind me of our evolution in experience in treating patients with neuroendocrine tumors, which I know that, for example, the Moffitt Group has a lot of experience with.
Where we would go-- when we were starting out over a decade ago, going very aggressive, trying to hit every little thing and finding out that you have no options two to three years down the road, or you because liver fibrosis and so on. Now what we do is a much more patient-paced, methodical treatment algorithm for these patients where we're only picking off enlarging tumors or big tumors and so on.
We're not seeing hepatic fibrosis or bad vasculature anymore. We're able to stretch these patients out for over five years, and sometimes now even approaching 10 years. I think we can try to do the same thing with HEPZATO as well.
[Dr. Venkatesh Krishnasamy]
That's a great point. This is a non-embolic, non-liver toxic therapy. We're walking the line a little bit because even in the FOCUS trial, there was a significant portion of patients that required at least two treatments to show response. At the same time, as you said, we're not going after a complete response.
Just loading the drug, loading intervention, and then having bone marrow toxicity down the line can be a problem as well. Spacing those treatments out, as Altan alluded to, we don't know the right answer and hopefully we'll figure it out over the next three to five years.
[Dr. Siddharth Padia]
We still don't know, is this better? Is radioembolization better for uveal melanoma? There was one analysis published out of Europe looking at a propensity score match retrospective of about 35 patients in each group, which suggested that percutaneous hepatic perfusion was superior.
Again, it's really one study. We're still trying to figure out our algorithm. I think one of the big advantages of this therapy of HEPZATO is that it is non-hepatotoxic. The rate of liver decompensation is extraordinarily low. In every single patient I've treated, we've measured their LFTs afterwards, not even their AST or ALT will budge, even the next day. It really does seem to be pretty benign on the normal hepatocytes.
[Dr. Altan Ahmed]
I think the chronic liver toxicity is just not there. Ultimately, what this is going to mean is that we're not really burning bridges for the alternatives, whether it's Tebentafusp or whatever may come down the line later. This is huge for the patients. If we can preserve the liver, get control of the disease, we're just extending their lifespan here.
[Dr. Siddharth Padia]
We're even starting to integrate this with radioembolization. Someone might get radioembolization, they may have a good response, they may progress in the future. Then can we switch to this or vice versa. It's nice to have this other tool that we can use for these patients that does not have overlapping adverse events.
It's not the same adverse event profile. In other words, you can figure out how to mix and match these therapies. Of course, from a data perspective, it's going to be very hard to tease this out. It allows us to just offer more therapies that are individualized and personalized for patients.
Podcast Contributors
Cite This Podcast
BackTable, LLC (Producer). (2025, March 4). Ep. 522 – Advancements in Treatment of Metastatic Ocular Melanoma [Audio podcast]. Retrieved from https://www.backtable.com
Disclaimer: The Materials available on BackTable.com are for informational and educational purposes only and are not a substitute for the professional judgment of a healthcare professional in diagnosing and treating patients. The opinions expressed by participants of the BackTable Podcast belong solely to the participants, and do not necessarily reflect the views of BackTable.
Podcasts
