UPMC Physician Resources
Sorry CME is not currently available for this activity.
Regional Perfusion in the Treatment of Patients with Metastatic Melanoma
Dr. James F. Pingpank discusses the utility of regional perfusion strategies in the treatment of patients with metastatic melanoma.
Upon completion of this activity, participants should be able to:
- Describe the role of regional therapy for the management of hepatic metastases for ocular melanoma
- List risk factors present in primary tumors leading to the development of hepatic tumors
- Describe the role of immunotherapy for metastatic ocular melanoma
- Alexander HR, et al. J Clin Oncol. 2010 Jan 1;28(1):114-8.
- Avital I, et al. Ann Surg Oncol. 2010, 17(1):163-70.
- Zeh HJ, Bartlett DL. Ann Surg Oncol 2009; 16:385-94.
Dr. Pingpank has no relationships with proprietary entities producing health care goods or services.
The University of Pittsburgh School of Medicine is accredited by the Accreditation Council for Continuing Medical Education (ACCME) to provide continuing medical education for physicians.
The University of Pittsburgh School of Medicine designates this enduring material for a maximum of 0.75 AMA PRA Category 1 CreditsTM. Each physician should only claim credit commensurate with the extent of their participation in the activity. Other health care professionals are awarded (0.075) continuing education units (CEU) which are equivalent to 0.75 contact hour.
For your credit transcript, please access our website 4 weeks post-completion at http://ccehs.upmc.edu and follow the link to the Credit Transcript page. If you do not provide the last 5 digits of your SSN on the next page you will not be able to access a CME credit transcript. Providing your SSN is voluntary.
Release Date: 3/5/2013 | Last Modified On: 3/5/2013 | Expires: 3/5/2014
Good morning. Today I’d like to talk about the utility of regional perfusion strategies in the treatment of patients with metastatic melanoma of multiple etiologies. I have no relevant disclosures to discuss. For patients with established metastatic melanoma traditional chemotherapies have been woefully inadequate until recent advances. Specifically looking at DTIC, combination chemotherapy and biochemotherapy we see that the overall response rate is mild to moderate but very few patients have complete response and of those patients a sustained complete response is very difficult to achieve.
Down here when look at the role, the impact of high dose IL2 we see that the overall response rate is quite modest but in those patients who obtain a response frequently this associated with a complete response and a complete response of very durable nature, often resulting in a cure. Since these patients tend to present at least in part with disease limited to an extremity or a body cavity we, we and other investigators have looked at the utility of regional therapy in the advancement or the treatment of these patients with advanced disease and have noticed an overall response rate which approaches 100% in multiple series and with a complete response rate of over 50%. What’s interesting to note is when you look at these patients with disease which is unresectable but confined to a single area of the body the sustained complete response rate is obtained in 1 out of 5 patients.
When we look at the long term results of isolated limb perfusion for intransient melanoma we get a good feel for what is possible for the treatment of patients with limited metastatic disease. In a series out of the National Cancer Institute from the ‘90s and early 2000s, 91 patients were treated who had stage 3b or 3c stage melanoma based upon the 2002 AJCC staging system. Approach used to treat this disease was via the femoral or popliteal artery in a minority of patients who had upper extremity disease. Patients were treated on a variety of prospective surgical trials using either Melphalan in half the patients or Melphalan plus tumor necrosis factor in the other half of the patients. The number of lesions seen in these patients ranged from a median of 6 to 10 lesion with a time of intransit disease from the primary of just over a year and a half. The overall response rate was 69% achieved and 62 patients had a complete response and a partial response was seen in 26 patients.
If we look over here we look at the survival curves and if we go to the bottom first we see the over – the progression free survival of patients treated on this therapy and a progression free survival at 4 years of approximately in 25% of patients illustrated in the potential ability to obtain durable responses. Additionally in these patients who presented with unresectable disease we see a sustained overall survival achieved in these patients with over 30% of patients alive greater than 10 years after their initial therapy.
Other investigators have utilized a more minimally invasive approach to the delivery of Melphalan treated within the catheter. This is a slide courtesy of Jon Zager at the Moffitt Cancer Institute looking at a multi-institutional experience of isolated limb infusion looking at response and toxicity. The major difference of infusion versus perfusion is that the limb is treated with an infusion dwell dose of chemotherapy as opposed to a recirculating oxygenated circuit. 166 isolated limb infusions were performed at 8 institutions with dosing corrected in the final 36% for ideal body weight in order to mitigate toxicity. When the overall response rates are examined an overall response rate of 61% is achieved with a complete response rate of 33% and a partial response rate of 20%. In this un-randomized assessment versus isolated limb perfusion these numbers are slightly lower but can represent a difference in patient selection, so no direct experience is available comparing isolated limb perfusion with isolated limb infusion.
So what is possible? This is a patient who was treated at the NCI many years ago who presented with intransit metastases all below the inguinal ligament from metastatic melanoma. It is a 32 year old woman who had progressed through the Dartmouth regimen of biochemotherapy as well as isolated, as well as high dose IL2. Her disease was limited to her proximal thigh and thus she was felt to be a candidate for isolated limb perfusion.
Here you can see her postoperative scan one year later with complete resolution of the disease in her thigh and here we see her primary site which had been resected with a skin graft. Ultimately she recurred 13 months after the limb perfusion and this disease progression was outside of her perfused extremity and she died 20 months later of metastatic disease outside of the extremity with a sustained complete response in her lower extremity.
Moving onto a different type of melanoma we can look at patients who – sorry, different site of disease, we can look at patients who have hepatic metastases from cutaneous melanoma. These are patients who are traditionally considered to be unresectable but in the face of isolated hepatic disease this represents a potentially curable situation. So again the group at the NCI examined 35 patients who underwent resection of their hepatic disease. The median survival from the primary tumor was 17 months, and the location of tumors, of the primary tumors were cutaneous in the majority of patients. Only one patient had metastatic ocular melanoma, a disease we’ll talk about later on in the discussion.
Patients had multiple lesions and one of the things that should be addressed the rationale for these resections, the majority of these patients were undergoing resection in order to obtain tissue on a research protocol for adopted immunotherapy strategies and thus the reason for these more aggressive surgical resection. 31% had all of their disease resected at the time of surgical resection and the actuarial 3 year survival in these patients was 53% and when you take this subset of patients who had an R0 resection an incredibly highly selective group of patients, an R0 resection resulted in a 3 year survival of 80% indicating that in properly selected patients controlling metastatic disease to even a distant organ such as the liver can be associated with prolonged survival.
Now to go to patients with metastatic ocular melanoma a disease which is different than cutaneous melanoma in both its predilection for metastases and also its response to systemic therapy. The incidence of ocular melanoma is 3500 or 4000 patients a year, and 50 to 60% of those patients will recur with metastatic disease. In the majority of those patients 80% the liver is the sole or dominant site of disease and this represents approximately 2000 patients a year. Metastatic disease is associated with a very poor prognosis with a median survival of 2 to 7 months and a 1 year survival of approximately 10% when we look at historical controls.
As yet systemic chemotherapy or immunotherapy do not appear to alter the natural history of this disease but this does not necessarily represent a lack of effect but a lack of these patients being included on major national trials. Because of the protensity of these patients to die of liver toxicity and liver progression multiple hepatic therapies have been justified in order to control progression to the liver including hepatic artery infusion, chemoembolization and isolated hepatic perfusion. The rational for regional therapy to the liver is that it allows dose escalation of an agent to the cancer bearing region or organ within the body while minimizing systemic exposure and toxicity. This is achieved through complete separation of the regional and systemic circulations. This reduction or elimination of systemic toxicity and dose escalation of therapeutic agents is really limited by what the exposed and perfused organ can tolerate. And this brings us into a much higher dose response area of the curve, allowing for improved efficacy and tumor response.
The liver is similar to the limb as a unique vascular anatomy that allows us to control the entire inflow and outflow of the organ so that all drug and all blood can be absolutely 100% controlled. Additionally, studies by Ridge and Sigerson at Memorial Sloan Kettering have established that the liver metastases derive the majority of the blood flow from the hepatic artery where the liver survives off of a portal venous flow predominantly. This allows us to even concentrate drug delivery within the liver by utilizing the hepatic artery as opposed to the portal vein.
This model also allows us to deliver potentially, deliver clinically relevant levels of hyperthermia and biologic agents which will either be too toxic or too difficult to delivery to the body as a whole. And it allows us to treat the entire tumor burdened area, not just disease – focal areas of disease as would be seen with ablation. As you can see, picking off specific tumors in this MRI above or in this patient with ocular melanoma would really not be possible and treating the entire organ both macroscopic and microscopic disease is vitally important to preserve the patient’s function.
So for metastatic ocular melanoma what are our, our options? The gold standard for any metastatic disease is surgical resection and this has been much more extensively studied in Europe than in the United States. At a center of excellence in Europe 3800 patients were treated with ocular melanoma and 798 of them developed metastatic disease to the liver. 255 were able to undergo hepatic resection with a very prolonged disease free interval from the primary tumor indicating a degree of selection bias. 76 of those 255 patients were able to be resected to an R0 zero representing a small but significant minority of those patients. The median overall survival is 14 months and those obtaining an R0 resection had a median overall survival of 27 months. When examinations of factors associated with improved survival were looked at, the ability to obtain an R0 resection, a prolonged disease free interval and absence of miliary or significant disease in the liver were associated with improved survival.
Chemoembolization is another therapy which has been investigated in multiple centers. In a series reported in the American Journal of Clinical Oncology in 2009 265 procedures were performed on 125 patients with no complete responses visualized. Partial responses were seen in 11% of patients and 16% of patients had a minor response. The median overall and progression free survival seen in these patients was 6.7 and 3.8 months respectively and this was not significantly greater than what we would expect to see in historical controls. Factors which were associated with poor survival included increasing liver involvement and high LDH both representing significant tumor disease.
A separate series, a smaller series has been more recently published which looks at, which saw progressive disease in 65% of patients at the initial evaluation. Medial progression free survival was – overall and progression free survival was 271 days and 185 days representing a minimal treatment impact associated with therapy. What was noted however was that – that patients who had a nodular appearance on angiography and more diffuse disease appeared to respond less well to therapy than patients who had discreet disease.
Regional administration of Fotemustine has also been examined as a way to potentially treat these patients. As noted above the – or as noted previously patients have – tend to have multiple areas of disease within the liver and treating the entire liver isn’t very desirable. 30 patients had a total of 36 ports placed for Q weekly administration of Fotemustine and with a prolonged disease. These patients had a prolonged disease free interval from their primary tumor. Number of patients – the number of infusions per patient was 8 and response data was somewhat better with approximately 1/3 of the patients having a partial response. 11 patients though, an additional third, had no evidence of any therapeutic benefit. The median overall survival was 14 months and for patients no responding the median overall survival was quite short at 5 months. The toxicity was very manageable however and so this at least was performed in a very safe manner but the efficacy of the drug does not appear to be what we had hoped it to be.
Finally the group at Jefferson has had extensive experience looking at immuno-embolization to treat these patients. Immuno-embolization is performed with GM-CSF every 4 weeks and in 34 patients treated a median overall survival of 14 ½ months was obtained. Factors associated with survival included gender and GM-CSF dose and in the 31 evaluable patients 2 CRs and one partial response were noted. The duration of response was not reported with this series.
Because of the desire to treat the liver with extensive disease through a comprehensive fashion and the overall positive impact of Melphalan on metastatic cutaneous melanoma we undertook a significant program at the National Cancer Institute looking at metastatic melanoma to the liver treated with intraarterial Melphalan. We’ll go over the circuit in just a minute, bu the first report of this that was published in 2003 showed 29 patients with an overall response rate of 62% with a median overall survival of 12 months. This is an un-randomized series and in patients who were nonresectable patients who underwent surgical laparotomy.
This is a schematic of the dissection which was undertaken in a total of 290 patients at the National Cancer Institute with a variety of metastatic disease confined to the liver, and here we can see it’s a dual circuit model in which surgically the entire vena cava is mobilized away from the retro-hepatic tissue in order to completely isolate the liver and the vena cava from the remainder of the body. The suprahepatic vena cava is completely controlled and a comprehensive portal dissection assuring complete inflow and outflow of the liver is obtained. This allows us to run two circuits, an intraarterial circuit in which the gastroduodenal artery is used to access the liver, and a liver perfusion with high dose chemotherapy is obtained on a cardiopulmonary bypass circuit. The blood flow in the vena cava is clamped above the renal veins and above the liver so that a second circuit is necessary which shunts blood from below the liver including the kidneys up through a veno-veno bypass circuit and into the systemic circulation above the diaphragm. The original group of patients treated at the NCI also had portal venous blood shunted into the circuit but subsequently this has been modified at the University of Pittsburgh to take out this circuit and clamp the portal vein.
The efficacy of this is demonstrated by utilizing pharmacokinetic analysis. So patients are perfused for one hour at temperatures of approximately 40 degree. Tumor necrosis factor is added at 1 mg in patients where this is obtainable and flow rates of approximately 600 cc to a liter are obtained. Physiologic arterial blood pressure is maintained and very robust veno-veno bypasses are obtained.
When we look at the pharmacokinetics of both tumor necrosis factor and Melphalan administered through this circuit we see that in the case of tumor necrosis factor no detectable systemic levels are able to be obtained and only minimal post-washout Melphalan is able to be demonstrated in the systemic circulation indicating complete isolation of the live from the remainder of the body with delivery of drug.
As noted before, this dissection circuit has, or this perfusion circuit has now been modified here at the University of Pittsburgh to make this a less toxic procedure and is associated with shorter hospital stays and more prolonged – sorry, shorter hospital stays and shorter operative times. So again the catheter is placed in the gastric duodenal artery for access to the hepatic circuit but a percutaneously placed catheter in the retro-hepatic vena cava is used to empty the hepatic – empty the liver through the vena cave, no longer is a venacaval incision necessary. The vena-vena bypass is additionally maintained through percutaneously placed catheters and no clamp is maintained on the portal vein.
The downside of this circuit is that as you can imagine it is very difficult to repeat this, this is an extraordinarily well tolerated procedure and in the patients with ocular melanoma close to 100 patients have been treated without a mortality. The length of stay in the hospital now averages 5 to 6 days, but this dissection is very difficult to repeat both for the portal dissection as well the retrohepatic venacaval dissection. So we have been examining other ways to potentially treat these patients either prior to or after they’ve had an open liver circuit, a liver perfusion.
So in the form of disclosures the following data that I will be presenting is from multiple trials which have been run at the NCI and at multiple other institutions including the University of Pittsburgh and all participating institutions received research support from the study’s sponsor, Delcath Systems in New York. In addition intraarterial Melphalan and the Delcath Double Balloon Catheter are under FDA I&D status at this time.
So the strategy here was to continue to isolate the liver but to do so without a surgical approach. And to allow this we utilized a double balloon catheter which was again commercial made and placed into the retro-hepatic vena-cava through a percutaneous approach. Balloons were inflated to completely isolate the hepatic venous outflow into the vena cava as demonstrated here by this vena cavagram with both balloons will with contrast and contrast retrograde injected into the hepatic veins. Once this performed and the venous outflow is able to be isolated drug can then be infused through the hepatic artery – or into the hepatic artery through a percutaneously placed arteriography catheter and the entire liver perfused with drug. The drug – the blood is then removed from the body and brought out through two inline activated filters and the chemotherapy is filtered and then the blood is returned to the patient. This is performed over a 60 minute process with the first 30 minutes being drug infusion and the second 30 minutes being a venous washout.
So what we were able to demonstrate with this model is for patients with metastatic melanoma treated on the Phase I trial is a significant evidence of response. Mind you that Phase I trials are designed to look at dose response and not look at – or look at dosing but not at response, even in that trial a very small group of patients with melanoma we were able to see an overall response rate of 50% in treated melanoma patients and two of those patients experienced durable complete responses. When we look at this very mature group of patients we see that the site of hepatic progression still remains either in the liver of liver plus systemic in the majority of patients indicating that controlling disease in the liver is really a clinically relevant strategy in these patients.
So why does this therapy work? If we look at what we can do with Melphalan while administered systemically we see that the maximum tolerated dose of Melphalan administered through a systemic approach on this bottom graph is .1 mg/ml. After you get above this level you start to see significant bone marrow toxicity which precludes the drug’s effective use. This is below the therapeutic window for metastatic melanoma. When we look at both the isolated open liver perfusion and the percutaneous approach to the liver perfusion what we see is that we are able to get 2-fold increases in drug deliver, again here this is now above 15 mg/ml so 2 orders of magnitude greater than what we were able to obtain with the systemic administration. What we do see is not completely effective filtering however, and that we do see some drug which leaks out into the systemic circulation when we do this percutaneously. However this comes with also a greater dose of drug which safely can be delivered so although we do see a little more systemically than we’ve seen with the open surgical procedure we are able to deliver twice the amount of drug into the liver.
So what do we see in the, in the trial design related to toxicity? If we look back at our open liver perfusion data we see no systemic toxicity, all the toxicity is related to the surgical procedure. This because of the incomplete filtering of drug we do see some systemic toxicity. We see no – we see neutropenia, thrombocytopenia and anemia which were our dose limiting toxicities in these patients, these were all treatable toxicities that were observed and patients were able to undergo multiple treatments. What we did not see though is any evidence of significant hepatic toxicity. All the patients enrolled in this Phase I trial had too much disease to be treated via an open approach and yet utilizing this percutaneous approach we were able to affectively treat them with really minimal self reversing liver toxicity.
This data was then used to design a Phase III random assignment trial of percutaneous hepatic perfusion versus best available care under a FDA special protocol assessment, 92 patients were enrolled at 10 institutions and this was designed as a crossover trial so that at hepatic progression patients in the control arm would be able to be crossed over to the treatment arm. The Melphalan dose administered was 3 mg/kg based upon ideal body weight and patients were stratified for ocular versus cutaneous disease. As mentioned the primary end point was hepatic progression free survival and the secondary end points were response rates, disease free survival with best available alternative care and response rates for all patients treated with percutaneous hepatic perfusion. This trial was initiated at the National Cancer Institute as an intramural study within the surgery branch and had – underwent a planned transition to a multicenter trial. This was done under external DSMB monitoring.
Treatment strategy for these patients, patients were placed on study and randomized. They would then have imaging at very frequent intervals irrespective of their treatment. This was felt to be very important because these patients tend to progress outside – within and outside of the liver very quickly and patients who were not responding to either therapy the option for additional therapies was thought to be potentially necessary in a relatively straightforward and rapid fashion. So patients had baseline 6 week, 12 week, 20 week, 28 week and 36 week interval exams and staging and then subsequently went on every 4, every 8 weeks after that.
Patients on the treatment arm were treated with Melphalan and had an angiogram at each treatment and the assessment of the gastric duodenal artery in order to assure that there was no perfusion of drug outside the liver. Anything that did not go into the liver would escape the filtering system and would be associated with significant toxicity. Patients were scheduled to be treated every 4 to 5 weeks with time left for recovery of bone marrow between each of the treatments.
Inclusion criteria for this patient population was very loose, patients needed biopsy proven ocular cutaneous metastatic melanoma predominantly in the liver parenchyma but limited extrahepatic disease was allowed. This was felt because the majority of patients with metastatic ocular melanoma of the liver will die of liver disease and small volume lung or bone disease was thus allowed to be present in patients enrolled in the study. Patients needed adequate hepatic function and this was a very liberal definition, serum bilirubin less than 3, prothrombin time within 2 seconds of the upper limit of normal, a liver function test less than 10 times the upper limit of normal. No portal hypertension or reversal of flow in the portal vein was permitted on the study.
93 patients were randomized to the study at 10 institutions in a 1 to 1 fashion with 44 patients in the percutaneous perfusion arm or the study arm, and 49 patients in the control arm. Hepatic progression was noted and then 55% of the patients in the control arm crossed over in a planned crossover trial to the treatment arm. When we looked at the demographics of these patients with regard to age, gender, race, ecog status and the location of the primary tumor there was no significant difference between the two groups. When we looked at prior chemotherapy a large number of patients had had prior radiation therapy, this was to their primary tumor but there was an equal amount of chemotherapy, immunotherapy, image directed local therapy between both groups of patients.
Toxicity of this treatment was not insignificant however, 40 patients underwent 116 treatments at the time of the primary analysis for efficacy. When we look at the treatment related Grade 3, Grade 4 and Grade 5 toxicities we see that hematologic were significant. The majority of these patients had treatable toxicity but we did see 61% evidence of neutropenia, Grade 3 or 4 neutropenia and a 75% incidence of Grade 3 or 4 thrombocytopenia. Of note, two patients died of complications of neutropenic sepsis and we also – I wanted to show Grade 5 toxicity or mortality of a patient who died of liver failure. This is the first liver related death we had seen with this trial and represented a very massive disease present in the patient who also progressed through the therapy and into liver failure within the first month after treatment. Other GI toxicity was very self limited or minor with minor acute cholecystitis and gastric ulceration, and one patient had arterial pseudoaneurysm with development of AV fistula on catheter placement.
So we examined the mortality rate of this looking at it in two different ways we see the overall mortality rate is 7.5% when we look at the number of patients treated or 2.6% procedure related mortality rate, a number that represents for the majority of institutions their first use of this device. when we look at the treatment related deaths 2 of them were form neutropenic sepsis, 1 from hepatic failure. The patient with hepatic failure on autopsy had a complete replacement of his liver with tumor, and this is something we have now learned to screen for when MRI and CT underestimate the amount of disease by doing dye laparoscopies on any patient with a question of extensive disease.
As noted the primary end point of the study was hepatic progression free survival on an intention to treat basis. When we look at that we see a clear prolongation of the target end point versus the control with the medial and the best alternative care a control arm of 49 days versus 245 days in the experimental arm. This resulted in a log ranked P value of less than .0001 and an associated hazard ratio of .301.
If we were correct in assuming that treating the liver would be beneficial to patients then the overall progression free survival rate would also be positively impacted on by treating what we felt was the dominant source of disease, and in fact that turned out to be true. The overall progression free survival rate including all disease within the patient was significantly improved of – from a control arm of 46 days to 186 days in the treatment arm when seen in an intention to treat analysis. Again the P value was less than .0001 and the hazard ratio for analysis was .04. Due to the fact of a high crossover rate there was no evidence seen on overall survival. In fact, patients who crossed – who were in the best alternative care arm had an exactly similar survival of just under a year versus patients who were treated on the control – on the treatment arm and this is due to the high crossover rate.
When we looked at an analysis of factors associated with improved survival we saw that the only thing that played out as being significant was the ability to obtain Melphalan as positively impacting upon survival. Even when we looked at this in a multivariant fashion, again Melphalan, the ability to obtain Melphalan for a patient was the only thing associated with an increased survival. On an exploratory analysis we looked at all the control patients trying to figure out what the impact was of the crossover. There are a variety of reasons patients did and did not crossover and this is nonrandomized data, so it needs to be taken with all the usual caveats looking at a secondary analyses or post hoc analyses. But what we saw was that patients who were either too ill or unwilling to crossover to therapy did significantly worse than those patients who were able to crossover. And when we look at that group of patients we see that the patients who underwent crossover to percutaneous hepatic perfusion had an overall survival – a overall survival of 400 days versus 124 days in those patients who were unable to crossover.
Secondary end point was also looking on an intention to treat basis looking at response. No control arm – no patients in the control or the treatment arm were able to obtain a complete response but the overall objective response rates were 34% in patients on the treatment arm versus 2% or a single patient on the control arm. This resulted in a very significant P value comparing the two of less than .0001.
And this is what we can expect to see with regard to responses. Here is a patient treated at the University of Pittsburgh who had significant disease in both lobes of the liver and had had extensive prior treatment on multiple chemotherapy and systemic based options. Her baseline LDH was 999 and 9 months post treatment we see a very gratifying response. There is still a small amount of nonviable tumor seen here but the majority of the tumor has completely resolved in her liver. Ultimately this patient progressed at 12 months within her liver and underwent an open liver perfusion as at that time she was felt healthier and able to tolerate a more vigorous approach and at that time obtained an additional 12 months of progression free survival.
Here is a patient treated with less high volume disease but you can see the diffuse nature of the disease with multiple metastases presenting throughout both lobes of the liver. This is the typical natural progression of this disease and the reason why embolization strategies frequently are not effective in this disease.
Here we see a very early and gratifying response just 6 week after the initial therapy in which we see the very strong efficacy of Melphalan versus this histology with a real clearing of disease even at just the – after just the initial treatment seen.
So in conclusion increased drug delivery achieved through novel regional therapeutic approaches may increase the efficacy of a given agent versus systemic administration but overcoming a low therapeutic index. Melphalan is a perfect drug to be used in this setting for a variety of - for a variety of approaches including isolated limb and isolated liver perfusions. High dose Melphalan delivered via intraarterial administration with subsequent hemofiltration as seen on the Phase III trial is effective against hepatic metastases from ocular cutaneous melanoma and provides improved disease control when compared to standardly available treatment regimens and really should be considered a frontline therapy for these patients. The expansion of this novel therapy to multiple clinical centers has proven safe and effective and shows the translatability of this from a high volume research center into other university settings. For patients with hepatic disease which might represent the life limiting extent of disease high dose regional chemotherapy provides meaningful control of hepatic disease without compromising overall disease control and overall total body toxicity. Thank you very much for your time.