Speaker: Dr. Jeff (Jean-Francois) Geschwind, Associate Professor of Radiology, Oncology, and Surgery, Johns Hopkins University School of Medicine, and Director of Interventional Radiology, Johns Hopkins Hospital, Baltimore, Maryland
Background on the speaker. A biosketch of Dr. Geschwind is available at http://www.hopkinsvascular.com/doctors/geschwind.cfm .
An excellent website on chemoembolization. This site is provided by Beth Israel Deaconess Medical Center, associated with Harvard University Medical School (reference chosen by a CACS member, not by Dr. Geschwind): http://www.bidmc.harvard.edu/mainframe.asp?node_id=453
At Johns Hopkins, you can read about chemoembolization at http://www.hopkinsvascular.com/procedures/chemo_embol.cfm .
Liver tumors in general. Dr. Geschwind explained that most of the data on chemoembolization are based on experience with liver cancers, especially hepatomas; he will include some carcinoid tumor data as well.
Hepatocellular carcinoma (HCC). This is the most lethal cancer worldwide, with an estimated 1.25 million deaths per year (especially in Africa and Asia). The rate in the U.S. has doubled; individuals with hepatitis C viral infection have a 1 to 4% risk per year of getting HCC (for a rate of 0.25 million per 10 years).
Colorectal carcinoma. This is the second leading cancer in the U.S. with 180,000 per year---of whom 70,000 have liver metastases. Survival of these colon cancer patients with liver metastases tended to be 6 to 12 months before the use of chemotherapy, which has doubled life expectancy.
Note: Carcinoid cancer is very different with a much better survival rate of 70% at 5 years if there are no liver metastases (and the primary tumor in carcinoid is almost never in the liver). With liver metastases, survival varies with the type of tumor and presence of carcinoid syndrome. Chromogranin A is the marker of choice for following how well treatment for carcinoid is working and can be used for measuring tumor aggressiveness.
Treatments. For any tumor the treatment of choice, when possible, is surgical excision. Transplantation of an organ is another option---and one that is being looked at again for carcinoid. Excision cannot be done unless enough liver tissue (perhaps 50%) can be left to provide some functioning liver. There is also a high recurrence rate for tumors in the remnant liver in HCC and colorectal cancer; for carcinoid, the return is at a much slower rate (although there are not enough data to be specific about the rate). Still, while surgeons like to think of the surgery as curative, it is not.
Hepatic artery chemoembolization. This is a fairly new procedure. Dr. Geschwind first worked with animal models under a National Cancer Institute grant. He decided that it was not wise to completely occlude an artery because in that situation, the tumors recruit new blood vessels. This approach of his differs from that of some others doing chemoembolization.
Interventional radiology. Other treatments besides chemoembolization that rely on interventional radiology for guidance include radioembolization and radiofrequency ablation (RFA). As in chemoembolization, radioembolization uses a catheter to deliver materials via arteries. At University of Maryland, Dr. Andrew Kennedy has been using beads called theraspheres coated with radioactive yttrium-90. These are not tumor-specific. Studies elsewhere involve Y-90-labeled octreotide (the generic name for the drug more familiarly known as Sandostatin) that is expected to be attracted to carcinoid tumors. (It is Dr. Geschwind’s impression that similar studies with indium-111 did not work as well as Y-90 [note added by CACS in 9/03: now lutetium-177 octreatate seems to be better than both in studies in Rotterdam].
While alcohol can be used to ablate some soft tumors, it is not useful for the hard tumors of colon cancer, and it is not used for carcinoid tumors for the same reason. RFA (basically using radiowaves to cook tumor tissue) can be used for tumors that are not too large. Dr. Geschwind uses a cutoff of 3 centimeters. (Some others have said 5 or 7 cm; basically if the tumor is too large, not all the tissue can be ablated. Unfortunately, it is rare to find patients with just a few small but detectable tumors. It is a challenge during treatments to reach tumor tissue without affecting normal tissue.
Other materials under study include acetic acid (for hepatomas), microwaves and lasers. In Dr. Geschwind’s opinion, cryotherapy (freezing tissue) does not work well.
The interventional radiologist helping a carcinoid patient with treatment choices needs to do a pre-procedure evaluation that covers patient history, histologic diagnosis, physical exam, and a risk/benefits discussion. One histologic stain that should be done is the ki-67 test measuring mitotic index (cell division) to determine the tumor’s aggressiveness.
Chemoembolization (CE) in detail. The basis for being able to do chemoembolization in the liver is that tumors in the liver tend to get nearly 100% of their blood supply from the hepatic artery while normal liver tissue receives blood as well from the hepatic portal vein. The dual treatment involves infusing concentrated doses of chemotherapy, then embolizing (partially blocking) the hepatic artery to decrease blood flow.
Why a partial block vs. a full one? Dr. Geschwind’s concern is that a full block, creating an “ischemic” tumor, will stimulate new blood vessel growth to resupply the tumor. He therefore modified the original procedure in which a full block was used.
The chemotherapy going directly to the liver can be used at a concentration 100x that of standard chemotherapy that is circulating throughout the body and should stay in contact with the tumors in the liver for a longer time than if blood were flowing normally through the liver. This approach also should mean less toxicity and less nausea than standard chemotherapy.
From results of chemoembolization studies over 10 to 15 years, Dr. Geschwind has selected a combination of 3 agents to use: cisplatin, doxorubicin (also known as adriamycin), and mitomycin. The carrier for these drugs is poppyseed oil, which Japanese researchers have shown to accumulate in tumors.
Embolic agent. Types of embolic agents include polyhydroxyl alcohol particles, a gelatin sponge (brand name, Gelfoam), and embospheres. Based on animal research done by Dr. Geschwind, the embospheres are now the preferred agent. (Previously, he used Gelfoam.) The polyhyroxyl alcohol particles tend to clump and occlude too much area of the blood vessel. The embospheres do not clump but can damage normal liver tissue.
Ineligibility. Individuals are not eligible for the chemoembolization procedure if they have a combination of several factors: high bilirubin, a high AST (the liver enzyme aspartate transaminase) level, hepatic encephalopathy, and tumors taking up more than 50% of the area of the liver. The presence of any one of the following may preclude the chemoembolization procedure: biliary obstruction, jaundice, or hepatic encephalopathy. Other factors that may make one ineligible include extrahepatic disease (disease outside the liver) and portal vein thrombus.
CE technique. At Hopkins when Dr. Geschwind became aware that full occlusion of the blood vessel could limit the number of chemoembolizations to only one and could stimulate tumors to recruit new blood vessels, they changed the process to the partial occlusion. Johns Hopkins now first mixes the chemotherapy and a product called Ethiodol containing the poppyseed oil, then uses a slurry of Ethiodol, the chemotherapeutic agents, and a brand of particles called Ivalon. This mixture was first studied in rabbits.
Dr. Geschwind prefers to use MRI scanning to identify where the tumors are---and to follow up after treatment to identify dead vs. living tumors.
After the CE, the patient stays overnight. The next day a non-contrast CT scan of the liver shows how well the Ethiodol has been distributed in the liver. The patient is given antiemetics (anti-nausea drugs) and anti-pain medication to take with him/her on discharge.
Side effects. Dr. Geschwind has found more side effects after CE of the left lobe of the liver vs. the right lobe, perhaps because the left lobe is nearer the stomach. The side effects tend to start 24-48 hours after the procedure. There can be gastric pain, hiccups, shoulder pain that is referred from the liver, nausea, and vomiting. Inflammation can occur 3 to 5 days after the procedure. Some of the pain is due to irritation of the nerves in the liver area. Low-grade fever is normal. [Note: see the website reference at the beginning of this summary for more details.] Side effects can last from 1 to 6 weeks. Some of the pain is due to the necrosis (dying) of tumors, a desirable but uncomfortable side effect. Neuroendocrine tumors do tend to shrink significantly.
Special note: The writer’s notes end here (because of departing for a plane flight), but the discussion continued for another hour at the April 2002 CACS meeting.
Additional note: Dr. Geschwind has a large database on patients undergoing CEs for liver tumors. He hopes to be able to determine prognostic factors that can predict which patients will have the most success with this procedure.