acids (palmitic-, linoleic-, oleic-, stearic acid) were performed using heptadecanoic acid as internal standard. Results: During the postoperative infusion period glucose and insulin concentrations in KII and Kill were significantly (p < 0.05) lower than in KI. Highest lactate concentrations (up to 9.3 mmol/l) were observed in KI after 6 hours of infusion. In KI all measured fatty acids fell under preoperative concentrations, when glucose was infused. Fatty acids in KI remained at significantly (p < 0.05) lower levels compared to KII and Kill until glucose infusion was stopped. Conclusions: In contrast to hypocaloric glucose infusion after CABG a mixture of glucose-xylitol (1:l) directs whole body metabolism as well as energy metabolism of myo- cardium to reasonable physiologic pathways. Highly elevated blood glucose levels associated with enhanced endogenous lactate production did not occur. The concomitant rise of insulin during glucose infusion reduced endogenous free fatty acid release and oxidation still further (1). On the basis of physiologic balanced fatty acid concentrations hypocaloric glucose-xylitol showed beneficial effects with respect to myocardial energy supply. (1) Gelfand RA, Barrett J. Effect of physiologic hyper- insulinemia on sceletal muscle protein synthesis and break- down in man. J Clin Invest 1987; 80: 1-6. P.3 Cachexia in pancreatic cancer - is islet amyloid polypeptide (IAPP) a patogenetic fac- tor? J. Permerth. T. Adrian” and J. Larsson Department of Surgery, Linkijping, Sweden and ??Department of Biomedical Science, Creighton University, Omaha, USA The cachectic state in pancreatic cancer is characterized by severe weight loss and altered glucose metabolism. The diabetic state that occurs in the majority of patients with pancreatic cancer patients has previously been shown to be an effect of insulin resistance rather than of reduced insulin secretion. IAPP, a recently discovered hormonal factor that is normally co-secreted with insulin from the pancreatic p- cells, has experimentally been shown to reduce insulin sensitivity and food intake. In this study we investigated plasma levels of IAPP in pancreatic cancer in relation to diabetic status. Material and methods: Fasting plasma levels of IAPP were determined in 30 patients with pancreatic cancer and in 25 comparable healthy controls. Diabetic history was registered and glucose tolerance was investigated by an oral glucose tolerance test. Results: Eighteen of the pancreatic cancer patients were diabetic. Seven of the diabetic patients needed insulin treat- ment. All patients were weight-losing. IAPP was markedly elevated in the pancreatic cancer patients compared to the normal subjects (22.3 k 2.6 pM vs 8.0 + 1 .O pM, p < 0.001). The levels in the diabetic patients were grossly elevated compared to non diabetics (28.9 + 3.5 pM vs 12.1 f 0.9, p < O.OOl), however, IAPP also in non diabetic patients was significantly higher compared to the controls (12.2 + 0.8 vs 8.0 + 1.0, p = 0.007). No difference was observed between the insulin and non insulin treated diabetics. Conclusions: We found grossly elevated plasma levels of IAPP in patients with pancreatic cancer, particularly in these with diabetes. It is tempting to speculate that the marked increase may contribute to the pronounced insulin resistance and weight loss that are characteristic for cachexia in pan- creatic cancer. P.4 Regulation of lipid metabolism during con- tinuous total parenteral nutrition A. Soulier’, C. Prudhon, C. Ricour. J.L. Bresson * H6pital Saint-Antoine, H6pital Necker, Paris, France Total parenteral nutrition (TPN) sometime leads to visceral overload syndromes. This has been attributed to an admin- istration of energetic substrates exceeding the patient’s energy expenditure. A maximal oxidative disposal rate is indeed proven for glucose. Yet, such a limitation of lipid oxidation has not been studied during TPN. We thus assessed lipid oxidation during TPN with varying amounts of lipid supply and resting energy expenditure (REE). Methods: Twenty-five children receiving total parenteral nutrition were studied. Two protocols were carried out concurrently. In the first study, fifteen 5.2 & 2.2 years old children with the same REE (56 k 7 kcal kg-’ d-‘) and glucose intake (9 g kg-’ d-‘) were divided into 3 groups according to their lipid intake: 1.5, 4.0 and 6.5 g kg-’ d-’ in groups I to III. In the second study, ten other children were divided into two groups according to their age and REE: 16 + 6 months old - 60 + 4 kcal kg-’ d-’ (group IV) and 18 k 2 years old - 29 + 2 kcal kg-’ d-’ (group V); all received the same energy intake as glucose alone (9 g kg-’ d-‘). without fat. VO1 and VCOl (open-circuit indirect calorimetry) and total urinary nitrogen (micro-Kjeldahl method) were used to calculate fat metabolism; glucose metabolism was determined using D2 glucose infusion. Results (g kg-’ d-‘; means _+ SD) were analyzed using Mann-Whitney tests between groups. Results: Fat storage increased from group I to group Ill, without change in fat oxidation (table I). When REE was greater than energetic supply (group IV), there was a net oxidation of endogenous lipids (table II). When REE was smaller than energy intake (group V), there was a net fat storage from glucose infused in excess (table II). Fat oxidation was significantly correlated with REE (n = 25; r = 0.99; p < 0.01). Table I group I group II group Ill p fat oxidation 1.92kO.5 1.77LO.7 2.23+1.35 NS fat storage -0 42kO.44 2.36+0.66 4.16k1.15 co.01 Table II group IV group v glucose oxidation 8.55k3.28 7.04& 1.43 Rs fat oxidation synthesis < 0.01 2.95+1 0.27 +0.58 Conclusion: During TPN, increasing fat intake over the part of energy expenditure not covered by glucose oxidation only results in increased fat storage, without change in fat oxidation, with potential untoward effects. Fat oxidation seems inversely correlated to the part of REE covered by glucose oxidation. References: J. Clin. Invest.: 1985, 76, 1019; Pediatr. Res.: 1989, 25, 645. P.5 Triglyceride hydrolysis of soy vs fish oil LCT emulsions F. Oliviera, Y.A. Carpentier, I. Hansen andR.J. Deckelbaum Columbia University, New York. USA Fish oil triglycerides (TD) are being considered as a com- ponent of i.v. lipid emulsions but their potential for lipase mediated clearance from plasma is unknown. Free fatty acids (FFA) release was used to assess, in vitro, hydrolysis of soy oil LCT emulsions (LCT) and fish oil LCT emulsions (a-3) with lipoprotein (LPL) and hepatic (HL) lipases. Emulsions were prepared using standard industry methods and Q-3 contained 18% and 28% of total TG fatty acid as EPA and DHA respectively. Both LCT and Q-3 emulsions required apo 44