TNF-a HAS NO DIRECT IN VIVO METABOLIC EFFECT ON HUMAN MUSCLE Ivo DE BLAAUW 1 *, Alexander M.M. EGGERMONT 2 , Nicolaas E.P. DEUTZ 1 , Marc DE VRIES 2 , Wim A. BUURMAN 1 and Maarten F. Von MEYENFELDT 1 1 Department of Surgery, Maastricht University, Maastricht, The Netherlands 2 Department of Surgery, Dr. Daniel den Hoed Cancer Institute, Rotterdam, The Netherlands T umor necrosis factor alpha (T N F-a) is thought to have a key role in metabolic changes of muscle tissue during inflam- matory diseases. It isunknown whether T N F-a affects muscle metabolism directly or whether these changes are mediated by secondary mediators. W e studied 6 patients undergoing isolated limb perfusion with TN F-a for irresectable soft- tissue sarcoma or in-transit melanomas. Glucose, lactate, ammonia and amino-acid consumption or production were measured in the perfusate during 3 perfusion periods: before, after TNF-a and after the combined administration of T N Fa and melphalan. Arterial glucose, lactate, ammonia and amino- acid concentrations were monitored to detect metabolic effects of T N F-a after it entered the systemic circulation. Glucose uptake and lactate release by the limb remained unchanged after the injection of T N F-a alone, as well as after the combination of TNF-a and melphalan. Furthermore, glutamine, alanine, phenylalanine, tyrosine and total amino- acid release into the perfusate did not increase during T N F-a and melphalan treatment, indicating that muscle metabolism was not changed. After the isolated limb perfusion, TNF-a entered the systemic circulation and induced metabolic changes resulting in a doubling of arterial lactate concentra- tions, decreased arterial glucose concentrations and de- creased arterial amino-acid concentrations. Our study shows that regional administration of T N F-a alone or in combina- tion with melphalan does not directly affect muscle glucose and protein metabolism. The data suggest that systemic metabolic changes induced by T N F-a are mediated through secondary, centrally produced, factors. Int. J. Cancer 71:148– 154, 1997. r 1997 Wiley-Liss, Inc. Cytokines such as tumor necrosis factor alpha (TNF-a) are implicated as humoral mediators of metabolic disturbances ob- served in response to severe infections and in cancer cachexia (Tracey and Cerami, 1993; Starnes et al., 1988). These metabolic disturbances (ultimately) result in loss of body mass, which adversely affects the clinical outcome of such patients (Fearon, 1992). The disturbances are characterized by increased glucose turnover (Frayn, 1986) and increased net protein breakdown of peripheral-muscle tissue, with a subsequent increased efflux of amino acids to visceral organs (Fearon, 1992). The diverse metabolic effects of TNF-a are considered to be the result of directly mediated cellular effects and of indirect effects induced by secondary mediators such as IL-1, IL-6, IFN-g, platelet-activating factor and stress hormones (Tracey and Cerami, 1990). In vitro studies suggest that TNF-a has no direct effects on muscle protein metabolism (Moldawer et al., 1987; Goodman, 1991). However, the findings of in vitro studies on glucose metabolism are conflicting. In rat-muscle preparations, glucose conversion into lactate and alanine remained unchanged by the addition of TNF-a (Rofe et al., 1987), whereas in a later study TNF-a induced high glycolytic activity with increased lactate produc- tion in cultured muscle cells (Zentella et al., 1993). TNF-a also induced electrophysiological changes in the cell membranes of these cells, suggestive of increased ion fluxes across the membranes. The present study examined in vivo the direct and indirect metabolic effects of TNF-a on human muscle. Glucose and protein metabolism was studied in patients undergoing isolated limb perfusion with TNF-a. This procedure has been used successfully to treat irresectable limb sarcomas and in-transit metastases of melanomas (Lie ´nard et al., 1992; Swaak et al., 1993). Evidence was obtained that the in vivo effects of TNF-a on human muscle are indirect, and are mediated by secondary factors. SUBJECTS AND METHODS Patients The direct effects of TNF-a on human muscle glucose and protein metabolism was studied in 6 patients undergoing isolated limb perfusion. The study was performed according to the guide- lines of the declaration of Helsinki. Patient characteristics are summarized in Table I. We studied 2 male and 4 female patients with a mean age of 51 years (range 15–82 years). All subjects underwent isolated limb perfusion with TNF-a and melphalan as treatment for irresectable soft-tissue sarcomas or multiple in-transit melanoma metastases of the extremities (MD Anderson classifica- tion: stage IIIA or IIIAB) (Klaase et al., 1994). The indirect effects of TNF-a on systemic glucose and protein metabolism were studied in a second group of patients, described in Table II. The patients, 2 males and 4 females, with a mean age of 47 (range 20–76 years), were studied before, during and after perfu- sion. Treatment was administered for similar conditions as those of the first group. Drugs Recombinant human TNF-a (0.2 mg/ampoule, 4.9–5.8 3 10 7 U/mg) was a gift from Boehringer Ingelheim (Ingelheim/Rhein, Germany). The cytostatic drug melphalan (Alkeran) was obtained as a sterile powder (100 mg) that was dissolved aseptically using solvent and diluent provided by Burroughs Wellcome (London, UK). Treatment schedule Isolated limb perfusions were carried out under general anesthe- sia and generally took 4 to 5 hr. Isolation of the circulation of a limb was achieved by clamping the major artery and vein, by ligating collateral vessels and by applying a tourniquet to compress the remaining minor vessels in cutis, subcutis and muscles. Perfusion was carried out at the axillary, brachial, iliac, femoral or femoro- popliteal level. Isolated limb perfusions (ILP) consisted of a 90-min perfusion with 3 (arm) to 4 (leg) mg TNF-a and 10 mg/l (leg) or 13 mg/l (arm) volume of melphalan at mild hyperthermia (39–40°C). The composition of the perfusate consisted of 400 to 500 ml blood (50% red blood cells, 50% plasma), 200 to 400 ml 5% dextran 40 in glucose 5% (Isodex, Pharmacia, Uppsala, Sweden) of Haemacel, 10 to 30 ml 8.4% sodium bicarbonate and 0.5 ml 2500–5000 IU heparin. TNF-a was injected as a bolus into the arterial line provided limb tissue temperature was .38°C. Melpha- lan was administered 30 min later at limb temperatures between 39 to 40°C. At the end of the perfusion, the limb was washed with at least 1 l of Haemacel and 1 l of dextran 6% 70 (Macrodex, Pharmacia). Contract grant sponsor: Netherlands Organization for Scientific Re- search, contract grant number 900-562-125. *Correspondence to Department of Surgery, Maastricht University, PO Box 616, NL-6200 MD Maastricht, The Netherlands. Fax: 31-433670964. Received 10 September 1996; revised 5 December 1996 Int. J. Cancer: 71, 148–154 (1997) r 1997 Wiley-Liss, Inc. Publication of the International Union Against Cancer Publication de l’Union Internationale Contre le Cancer