890 SURGERY PRIMARY AND SECONDARY MALIGNANCIES of the liver represent a major cause of clinical cancer death, particularly when diagnosed at an unresectable stage of disease. 1 Because systemic treatment options have shown disappointing results, multiple regional treatment strategies have been explored with the use of a wide variety of different agents and techniques. 2 Isolated hepatic perfusion (IHP) has been under clinical evaluation for patients with primary or secondary cancers of the liver in recent years. 3 In general, isolation perfusion is a specialized sur- gical technique in which the vascular bed of a can- cer-bearing organ (liver) or region (leg) of the body is isolated and perfused with a recirculating circuit. The procedure has been well defined tech- nically and has resulted in high-response rates for a variety of hepatic neoplasms. 4 Complete vascular isolation of the liver results in the best potential pharmacokinetic advantage of any regional appli- cation of cytotoxic anticancer drugs because regional dose intensity can be achieved without causing fatal systemic side effects. 5 In addition to pharmacologic considerations, the separation of Isolated hepatic perfusion for lapine liver metastases: Impact of hyperthermia on permeability of tumor neovasculature Michael F. X. Gnant, MD, Linda A. Noll, BS, Richard E. Terrill, BS, Peter C. Wu, MD, Adam C. Berger, MD, Hung Q. Nguyen, Titia E. Lans, Bernard M. Flynn, DVM, Steven K. Libutti, MD, David L. Bartlett, MD, and H. Richard Alexander, Jr, MD, Bethesda, Md, and Vienna, Austria Background. Hyperthermic isolated hepatic perfusion (IHP) has been shown to cause significant regres- sion of advanced unresectable liver metastases in patients. Although there are different agents and treat- ment modalities used in IHP, the contribution of perfusion hyperthermia is unknown. Purpose. A large animal model of unresectable liver metastases and a technical standard for IHP in this model were established. This model was used to assess the effects of hyperthermia on vascular permeabili- ty of tumors and normal liver tissue during IHP. Methods. Sixty-five New Zealand White rabbits were used in a series of experiments. Disseminated liver tumors were established by direct injection of 1 × 10 6 VX-2 cells into the portal vein by laparotomy in anesthetized animals. Several surgical perfusion techniques were explored to determine a reliable and reproducible IHP model. Vascular permeability in tumor versus liver was then assessed with Evan’s Blue labeled bovine albumin under normothermic (tissue temperature 36.5°C ± 0.5°C), moderate hyperther- mic (39°C ± 0.5°C), or severe hyperthermic (41°C ± 0.5°C) conditions. Results. Tumor model and perfusion techniques were successfully established with inflow through the portal vein and outflow through an isolated segment of the inferior vena cava. A gravity driven perfu- sion circuit with stable perfusion parameters and complete vascular isolation was used. Vascular perme- ability was higher in tumor than in normal tissues (P = .03) at all time points during IHP. Hyperthermia resulted in a significant (up to 5-fold) increase in permeability of neovasculature; when severe hyperthermia was used, tumor vascular permeability was increased even more than normal liver permeability (P = .01). Conclusions. The VX-2/ New Zealand White rabbit system can be used as a reproducible large-animal model for IHP of unresectable liver metastases. It can be used to characterize the contribution and mech- anism of action of different treatment parameters used in IHP. Hyperthermia preferentially increases vas- cular permeability in tumors compared with liver tissue in a dose-dependent fashion, thus providing a mechanism for its presumed benefit during isolated organ perfusion. (Surgery 1999:126:890-9.) From the Surgery Branch and the Laboratory Animal Medicine Section, Animal Sciences Branch, National Cancer Institute, National Institutes of Health, Bethesda, Md; and the Department of Surgery, University of Vienna, Austria M.G. is a recipient of a research grant by the Max Kade Foundation, New York, NY . Accepted for publication May 28, 1999. Reprint requests: H. Richard Alexander, MD, FACS, Surgical Metabolism Section, Surgery Branch, National Cancer Institute, National Institutes of Health, Building 10 Room 2B07, 9000 Rockville Pike, Bethesda, MD 20892. 11/ 56/ 100668