SEATTLEDelivery of chemotherapy to hepatic artery branches that only or mainly feed a tumorso-called superselective chemoembolization and chemoinfusionyields favorable outcomes in patients with hepatocellular carcinoma (HCC), according to a pair of studies reported at the Society of Interventional Radiology annual meeting.
Daniel Yung-Ho Sze, MD, PhD, of Stanford University, explained the rationale for superselective chemoembolization: "We can give very high local doses of chemotherapy, prolong the dwell time of the chemotherapy, induce ischemia in the tumor, and limit the amount of organic as well as systemic toxicity of the chemotherapeutic agents."
Dr. Sze cited two landmark randomized trials showing that chemoembolization reduces the risk of death by about half relative to supportive care in patients with HCC (Lancet 359:1734-1739, 2002, and Hepatology 35:1164-1171, 2002).The technical aspects of superselective chemoembolization are still debated, Dr. Sze noted. "Our protocol at Stanford is to maximize the superselectivity with microcatheters and minimize the amount of solid particulate embolization we use," he said. The Stanford protocol is derived from those widely used in Asia.Angiographically, this procedure produces a characteristic "tumor on a stick" appearance (see Figure).
In this retrospective study (abstract 153), the investigators assessed outcomes in 165 patients with HCC who underwent 345 superselective chemoembolization procedures, either to palliate symptoms or to maintain candidacy for transplantation. Dr. Sze recognized Dr. Bo Yoon Ha, of Seoul National Medical Center, and Dr. Mindie Nguyen, of Stanford, for compiling and analyzing the data.
Up to 20 mL of ethiodized oil (Ethio-dol), 50 mg cisplatin, and 50 mg doxorubicin were used. Particulate material, a slurry of absorbable gelatin sponge (Gelfoam), was added to treatment in only 12% of patients.
After censoring of patients who went on to liver transplant, 69%, 46%, 39%, and 30% of patients were alive at 1, 2, 3, and 4 years, respectively. "With multivariate analysis, all the things we thought were importanttotal bilirubin and albumin, prothrombin/INR, and plateletsreally didn't have much of an effect," he said.The only significant predictor of survival was TNM stage. The 3-year overall survival value (39%) was roughly similar to values obtained in the two landmark trials (26% to 29%), Dr. Sze commented.
After 99% of the superselective procedures, patients were discharged within 24 hours. The 30-day rate of treatment-related mortality was 0.3% of procedures and 0.6% of patients. Severe complications, excluding nausea, vomiting, or pain requiring prolongation of hospitalization, occurred in about 1% of procedures and 3% of patientsvalues that were lower by an order of magnitude than those of the previous two landmark trials (17% of procedures, 23% of patients).