1. The effects of essential fatty acids (γ-linolenic acid, eicosapentaenoic acid, docosahexaenoic acid), at a dose of 4.8 g/day, given in combination as dietary supplements, on cytokine production were investigated in patients with colorectal cancer. 2. Total serum cytokines - interleukin (interleukin-1β, 2, 4 and 6), tumour necrosis factor-α and interferon-γ - were analysed using the enzyme-linked immunosorbent assay technique at different time intervals during the course of essential fatty acid supplementation. 3. Fatty acid uptake and patient compliance were confirmed by a significant increase in serum levels of γ-linolenic acid, eicosapentaenoic acid and docosahexaenoic acid in all three fractions: triacylglycerol, cholesterol and phospholipid. 4. There was no significant alteration in total serum cytokine concentration/levels in the first 2 months of essential fatty acid ingestion, but the levels of serum cytokines steadily declined thereafter, reaching minimum levels after 6 months of essential fatty acid supplementation. 5. Essential fatty acids, at the dose and duration (6 months) used in this study, reduced total serum interleukin-1β levels by 61% ( P = 0.044), interleukin-2 by 63% ( P = 0.05), interleukin-4 by 69% ( P = 0.025), interleukin-6 by 83% ( P = 0.030), tumour necrosis factor-α by 73% ( P = 0.040) and interferon-γ by 67% ( P = 0.050). 6. Three months after cessation of essential fatty acid intake, however, these cytokine levels returned to presupplementation values. 7. This present study has shown that long-term n -3 and n -6 EFA ingestion results in a significant reduction in circulating key cytokines. The precise mechanism of this reduction is unclear.

1. The T-cell-derived cytokine interleukin-2 may be used to reverse the immune suppression associated with major surgery. However, both major surgical procedures and recombinant interleukin-2 therapy are known to induce renal dysfunction. 2. Eighteen patients were randomized to receive either recombinant interleukin-2 (18 × 10 6 i.u./day) or placebo, given subcutaneously for 3 days before undergoing curative colorectal cancer surgery. Indices of renal function were determined pre-operatively and for 21 days after surgery. 3. Pre-operative recombinant interleukin-2 was found to significantly increase, compared with placebo controls, N -acetyl-β- D -glucosaminidase [peak levels 28 (SEM 2) versus 11 (SEM 3) i.u./mmol of Cr] and γ-glutamyltransferase [peak levels 5.3 (SEM 0.6) versus 2.4 (SEM 0.2) i.u./mmol/l] and decrease urinary fractional excretion of sodium [peak difference 0.32 (SEM 0.06) versus 0.76 (SEM 0.08)] (all P < 0.05). Significantly increased urinary excretions of creatinine, N -acetyl-β- D -glucosaminidase and γ-glutamyltransferase were also identified after surgery. All variables returned to pretreatment limits by the seventh day post-operatively, except N -acetyl-β- D -glucosaminidase, which was still significantly elevated 21 days after surgery. No differences in the serum concentrations of sodium, creatinine or urea were observed before or after surgery in either group. 4. Recombinant interleukin-2, when given in the preoperative period, was associated with significant renal dysfunction. However, routine monitoring of serum indices (i.e. sodium, urea, creatinine and albumin) failed to detect such renal damage. These results suggest that, with the use of recombinant interleukin-2 to enhance natural cytotoxicity in the peri-operative period, such therapy may potentiate the renal impairment occurring after surgery.

1 Rates of protein synthesis have been measured from the incorporation of 57 mg of l-[1- 13 C]leucine/kg for 90 min into muscle tissue and colorectal tumours removed at surgery from cancer patients. 2. For the 20 h preceding surgery and during the measurement of protein synthesis, the patients received intravenous saline, conventional intravenous nutrition (0.2 g of N and 103 non-protein kJ/kg body weight) or intravenous nutrition enriched with the branched-chain amino acids leucine, isoleucine and valine (0.2 g of N with 30% from branched-chain amino acids and 103 non-protein kJ/kg body weight). 3. Conventional intravenous nutrition resulted in a significant stimulation of the rate of protein synthesis in both muscle tissue (2.64 ± 0.75%/day versus 1.78 ± 0.51%/day in saline control, means ± SD) and tumour tissue (43.9 ± 10.3%/day versus 22.6 ± 5.6%/day in saline control). 4. Pre-operative nutrition enriched with branched-chain amino acids was less effective than conventional intravenous nutrition in stimulating protein synthesis in both muscle and tumour. The rates of protein synthesis were 2.12 ± 0.41%/day in muscle and 33.7 ± 5.3%/day in the tumours. 5. The expression of proliferating cell nuclear antigen in sections of the tumours showed changes with intravenous feeding of the two different amino acid mixtures that were similar to the changes in protein synthesis, and these two variables were significantly correlated. This is evidence that feeding with conventional mixtures and mixtures enriched with branched-chain amino acids stimulates tumour growth. 6. In this study the mixture enriched with branched-chain amino acids provided no clear advantage for cancer patients, since a smaller response to branched-chain amino acids was observed in both tumours and host muscle tissue.

1. A method is described for measuring the rates of protein synthesis in vivo in human colorectal and breast tumours by the intravenous injection of l -[1- 13 C]leucine as a ‘flooding dose'. 2. The incorporation of isotope into colorectal tumour protein was measured in six patients, whose tumours were biopsied after the injection. Fractional rates of protein synthesis were calculated from the enrichment of leucine in protein and the average free leucine enrichment in plasma. The range of rates obtained was 17.2–33.9%/day, with a mean rate (± sem ) of 22.5 ± 2.6%/day. 3. Tumour protein synthesis rates were also measured in 15 patients with breast cancer. The range of rates obtained was 5.3–15.9%/day, with a mean rate (± sem ) of 10.3 ± 0.8%/day. These rates are significantly lower than those obtained with colorectal tumours ( P < 0.001). 4. In 9 of the breast cancer patients, protein synthesis was measured in multiple random biopsies taken from the same tumour. The mean (± sem ) difference between the highest and lowest rates in biopsies from the same tumour was only 1.1 ± 0.3%/day. Only 13% of the variation in protein synthesis between separate tumours could be explained by sampling error because of variation within the tumour itself, the remainder being genuine variation between individual tumours.

1. Rates of protein synthesis were measured, in vivo , in lung, liver, heart and skeletal muscle of young male rats. Groups of rats were exposed for 1 h duration to one of the following anaesthetic regimens: 1.4% halothane, 2.2% halothane, 1.4% halothane in 66% nitrous oxide, intravenous pentobarbitone (20 mg/kg) and intravenous midazolam (18 mg/kg) combined with fentanyl (2 μg/kg). Fractional rates of protein synthesis were determined by injecting [ 3 H]phenylalanine (150 μmol/100 g body weight) 2. Liver protein synthesis was depressed significantly by all regimens, except midazolam/fentanyl, by up to 37.7% of control values. Lung protein synthesis was significantly reduced by all the anaesthetic agents by up to 30% of control rates 3. The effects of the anaesthetic agents on skeletal muscle and heart were small and not statistically significant 4. There was no evidence of ventilatory depression as manifested by changes in arterial blood gas partial pressures of CO 2 and O 2 , except in the group treated with 2.2% halothane.