Fatty emulsions are an essential part of parenteral nutrition, both as a part of energy supply, and as a source of essential fatty acids . Furthermore, they are involved in the structure and function of cell membranes and receptors, modifying gene expression, and modulating the inflammatory and immune response [2,3]. In addition, fatty acids are precursors of prostaglandins and other eicosanoids and have therefore important metabolic functions . It has been demonstrated that intravenous lipid emulsions show varying immunomodulatory effects dependent on the omega-6/omega-3 (ω-6/ω-3) fatty acid ratio . Experimental and clinical studies showed that the most favourable ω-6/ω-3 fatty acid ratio is proposed to range between 2: 1 and 3: 1 [5–9]. On the basis of this knowledge, a new lipid emulsion was developed soybean oil, medium-chain triglycerides, olive and fish oil (SMOFlipid20%) based upon a physical mixture of soybean oil, medium-chain triglycerides (MCTs), olive and fish oil  as well as 200 mg α-tocopherol (vitamin E). However, parenteral nutrition, including lipids might be associated with liver disease [11,12]. The mechanisms leading to parenteral nutrition-related liver dysfunction remain largely unknown but are likely to be multifactorial [12,13]. Standard enzyme markers, such as gamma-glutamyl transferase (γ-GT), alanin-aminotransferase (ALT), aspartate aminotransferase (AST), are most commonly used in clinical routine to assess hepatic alterations. However, they are relatively insensitive for detecting early hepatic dysfunction . Beside these routine parameters, new markers of hepatic dysfunction have recently become available. Alpha-glutathion S-transferase (α-GST) is abundant in the cytosol of hepatocytes as well as in proximal renal tubular cells and small intestinal mucosa . Unlike ALT and AST, α-GST is found in high concentrations in centrolobular cells, and, therefore, this parameter is more sensitive to injury in this metabolic zone of the liver . Changes in α-GST have never been investigated secondary to the administration of the above-mentioned new lipid emulsion. Consequently, the present study was performed to assess the effects of SMOFlipid in parenteral nutrition compared with that of a lipid emulsion based on olive and soybean oil (ClinOleic20%) on hepatic integrity in postoperative ICU patients using serial monitoring of α-GST serum concentrations.
After approval of the ethics committee and obtaining written informed consent, 44 patients aged over 18 years, who were expected to receive parenteral nutrition over 5 postoperative days following major abdominal surgery or large cranio-maxillo-facial resections for cancer, were enrolled in this study. Postoperatively, all patients were transferred to the ICU, and controlled mechanical ventilation was continued at least during the following 2 h. Sedation was maintained using continuous infusion of midazolam and, if necessary, boli of piritramide, whereas propofol including a lipid emulsion as carrier was avoided. The following exclusion criteria were defined for the present study: renal (creatinine > 1.4 mg dl−1; dialysis) or hepatic (AST > 40 U l−1 and/or alanine aminotransferase > 40 U l−1) insufficiency, acute pulmonary oedema, decompensated cardiac insufficiency, New York Heart Association (NYHA) status IV or V, hyperlipidaemia, insulin-dependent diabetes mellitus, overweight (body mass index > 30 kg m−2), cachexia (body mass index less than 18 kg m−2), pregnancy, psychiatric disorders, known alcohol or drug abuse, known hypersensitivity to egg, fish, olive, coconut or soy proteins and patients taking chronic corticoids.
The preoperative management (including preoperative feeding) was not standardized and decided upon by the surgeons.
Randomization was performed with closed envelopes containing the study assignment, which were opened before admission to ICU (concealed allocation). Patients were thus allocated to one of two nutritional regimens: group A (n = 22) received SMOFlipid20% (Fresenius Kabi Deutschland GmbH, Bad Homburg, Germany; SMOF group) and group B (n = 22) a lipid emulsion based on olive and soybean oil (ClinOleic20%, Baxter Deutschland GmbH, Unterschleißheim, Germany; control group). Parenteral nutrition including lipids was given continuously via a central venous catheter for 5 postoperative days, corresponding to the observation time. Nonprotein calories were given as 60% glucose and 40% lipid emulsion. The nonprotein energy intake per day was adjusted to 25 kcal kg−1 body weight. Crystalloid and colloid solutions, trace elements, electrolytes and vitamins were added according to the patients' needs. Packed red blood cells (PRBCs) were administrated when haemoglobin was less than 8.0 g dl−1. Blood glucose was maintained at a level between 80 and 140 mg dc−1, administering a continuous infusion of insulin, if appropriate.
According to the European Society for Clinical Nutrition and Metabolism (ESPEN) guidelines , all patients received a supplementary enteral nutrition (250 ml/24 h feeding solution) via a nasogastric or a jejunal feeding tube as of the first postoperative day.
Measurements, blood sampling, and analytical procedures
Arterial blood samples were taken before the start of parenteral nutrition (d0) and 48 h (d2) and 120 h (d5) after the start of lipid emulsion. At the same data points, mean arterial blood pressure (MAP), heart rate (HR), and central venous pressure (CVP) were documented. Serum was separated by centrifugation for 7 min at 4000 U min−1, and 1 ml aliquots were frozen at −30°C until analysis. The following parameters were measured: α-GST concentrations were analysed with the Biotrin Hepkit-Alpha Human GST-Alpha (Biotrin International GmbH, Sinsheim-Reihen, Germany; normal value < 7.5 μg l−1), a quantitative enzyme-linked immunosorbent assay (ELISA) based on the sequential addition of sample, antibody–enzyme conjugate and substrate to microtitre wells coated with anti-α-GST immunoglobulin G. The resulting colour is proportional to the amount of α-GST present in the sample. The activities of ALT (normal range < 31 U l−1) and AST (normal range < 34 U l−1) were measured with routine hospital laboratory tests, and serum triglycerides (normal value: <200 mg dl−1) were measured using routine enzymatic colorimetric test triglycerides GPO-PAP (Roche Diagnostics GmbH, Mannheim, Germany).
All statistical analyses, including the power analysis, were performed by an external professional statistician (M.A.R.C.O. mind, Düsseldorf, Germany). Our primary defined outcome was the increase of α-GST serum concentrations. The number of study patients required to detect a 125% increase in α-GST concentrations during parenteral nutrition with an α-error of 0.05 (two-sided), and a β-error was set at 0.2 (80% sensitivity) was calculated to include a minimum of 21 patients in each group using the recently published standard deviation of α-GST concentration of 19.5 μg l−1 postoperative patients receiving parenteral nutrition . We decided to include at least 22 patients per group to take potential dropouts into account.
Data are shown as mean and standard deviation (SD). Normal distribution was determined with the Kolmogorov–Smirnov test. Demographic data, duration of surgery and anaesthesia were analysed with the Student's t-test. Comparisons between the groups were performed by use of the unpaired t-test. A Bonferroni correction for multiple tests was used whenever necessary. P values of less than 0.05 were considered significant.
The enrolment of the patients, randomization, and dropouts are given in Fig. 1, according to the recommendations of the Consolidated Standards of Reporting Trials (CONSORT) statement .
The two groups did not differ significantly with respect to demographic characteristics, anaesthetic, and surgical data (Table 1). Furthermore, we did not detect any significant differences concerning MAP, HR, CVP, haemoglobin (Hb) levels, and platelet count throughout the study period (Table 2).
There was no significant difference at baseline (d0), but at d2 and d5, significantly lower AST (d2: group A: 27 ± 13 vs. group B: 47 ± 36 U l−1, P < 0.02; d5: group A: 31 ± 14 vs. group B: 56 ± 45 U l−1, P < 0.02), ALT (d2: group A: 20 ± 12 vs. group B: 42 ± 39 U l−1, P < 0.03; d5: A: 26 ± 15 vs. B: 49 ± 44 U l−1, P < 0.03) (Fig. 2), and α-GST levels (d2: group A: 5 ± 6 vs. group B: 17 ± 21 U l−1, P < 0.03; d5: A: 6 ± 7 vs. B: 24 ± 27 μg l−1, P < 0.01) were found in the SMOF group compared with the control group (Fig. 2).
There were no significant differences in triglyceride levels at baseline (d0: group A: 119 ± 35 vs. group B: 120 ± 45 mg dl−1; P = 0.87), whereas at d2 (group A: 151 ± 52 vs. group B: 202 ± 108 mg dl−1; P < 0.03) and at d5 (group A: 163 ± 72 vs. group B: 233 ± 94 mg dl−1; P < 0.01), the triglyceride levels in the SMOF group were significantly lower than in the control group (Fig. 3). Glucose levels and insulin dosage did not differ significantly between the groups throughout the study period (Fig. 3).
The main results of the present study were that a lipid emulsion based on soybean oil, MCTs, olive and fish oil (SMOFlipid20%) as well as α-tocopherol had no measurable effect on hepatocellular integrity during parenteral nutrition, whereas in patients receiving a lipid emulsion based on olive and soybean oil, liver enzymes were elevated indicating a lower liver tolerability in postoperative patients. Furthermore, triglyceride levels were significantly lower during the study period of 5 days in patients receiving SMOF in comparison to the lipid emulsions based on olive and soybean oil.
Serum ALT and AST activity levels are the most frequently relied-upon laboratory indicators of hepatotoxic effects . Unfortunately, these parameters show infrequent false negative signals of liver histopathologic injury as well as limited false positive signals, though they are still considered as the gold standard clinical chemistry markers of liver injury . Although the overall clinical utility of these markers are exceptional, they do not always correlate well with histomorphologic data . Consequently, ideal attributes of new markers of hepatic response include organ specificity for liver disorders, strongly correlating with well defined hepatic histomorphologic changes, or added information to ALT or AST values or both. Determination of the cytosolic liver enzyme α-GST is a sensitive and highly specific test for hepatocellullar damage and correlates better with drug-induced liver disorders . Unlike ALT and AST, α-GST is found in high concentration in centrolobular cells, and, therefore, it is more sensitive to injury in this metabolic zone of the liver . Furthermore, its low molecular mass of 45–50 kDa and its short plasma half-time of about 90 min allow a fast return to normal values, when the active phase of hepatic alteration is over .
Dietary lipids are important factors in affecting xenobiotic metabolism activities . Hepatic microsomal cytochrome P450 plays a key role in the metabolism of various endogenous and exogenous compounds, such as hormones and drugs. α-GST is an important enzyme in conjugation of these compounds . Consequently, hepatic complications related to the administration of total parenteral nutrition (TPN), in particular lipids, have been widely reported [11–13,22,23]. There are a lot of potential causes of parenteral nutrition-associated liver disease (PNALD), but the cause remains largely unclear and likely to be multifactorial [12,22]. The progress of PNALD is often self-limiting but can lead to liver failure in a minority of patients , and deaths have been reported as well . Described risk factors are a shortage of essential fatty acids or carnitine, septic events, length of intestinal resection, hepatic fat deposit, production of endotoxins and lithocholic acid secondary to intestinal bacterial overgrowth, and the absence of enteral nutritional intake [12,13,22–24].
In order to avoid such a disadvantageous influence on our patients and because there is no doubt that enteral feeding over parenteral nutrition improves the outcome in critically ill patients by maintaining the immunological and barrier integrity of the gut , leading to reductions in infectious complications , all patients of the present study received a supplementary enteral nutrition. Daily volumes up to 250 ml (5–10 ml h−1) are considered adequate to preserve the integrity of the gut. However, in Germany, a recent evaluation on ICUs revealed the induction of a combined parenteral and enteral nutrition regimen predominantly . Reasons might be that patients often suffer from gastric paresis; therefore, an increase in enteral feeding volume is not reasonable or patients are considered for reoperation. The patients are consequently considered not to be adequately enterally nourished during subsequent days, and, therefore, physicians may not delay caloric intake and administer a sufficient nutrition parenterally from the beginning.
There are some concerns regarding the role of intravenously administered fat emulsions in the development of hepatic abnormalities, including the fat source and the dose. Long-chain triglyceride (LCT) emulsions have been implicated causing hepatomegalia, pigment accumulation in Kuffer's cells and reticuloendothelial overload . Alwayn et al.[28,29] have shown that polyunsaturated fatty acids (PUFAs), in particular omega-3 (ω-3) PUFA, can prevent as well as decrease hepatic steatosis in a mouse model. It is well known that fish oil is a suitable source of ω-3 PUFA. However, specific attention has to be focused on the risk of their higher susceptibility to oxidation due to the unsaturated level of these fatty acids. Therefore, it makes sense adding α-tocopherol, one of the most lipophilic antioxidants to fish-oil-containing emulsions . Heller et al. reported that ω-3 PUFA in fish oil improves liver and pancreas function in postoperative cancer patients. In a small study, including 10 ICU patients in each group, Antébi et al.  could demonstrate that SMOF has beneficial effects on liver function when compared with a conventional soybean oil-based emulsion. Sheth and Bankey  reported on an immunostimulant effect of fish oil, including an amplification of the ‘positive’ hepatic acute-phase response, as measured by C-reactive protein (CRP) and fibrinogen. However, in the present study, we showed for the first time that SMOF compared with a lipid emulsion based on olive and soybean oil had a better liver tolerability in postoperative ICU patients.
Another explanation of our findings is the improvement of splanchnic blood flow secondary to the administration of fish-oil-containing emulsions: Pscheidl et al.  reported on normalizing splanchnic perfusion and improving killing of translocated bacteria in a low-dose endotoxin rat model after administration of a fish-oil-supplemented parenteral diet. On the other hand, it is well known that a reduction of hepatic blood flow based on haemodynamic instabilities with subsequent decreased oxygen availability might cause an increase in α-GST concentrations. In a previous study, we reported on a significant increase of α-GST at the end of surgery in patients undergoing controlled hypotensive anaesthesia . Therefore, one limitation of the present study is the lack of measurements of splanchnic blood flow or hepatic vein catheterization. However, the haemodynamic parameters of our patients were without significant differences between both study groups.
Furthermore, it is presently unclear whether the beneficial effects of SMOF are caused by one or two of its components (such as fish oil, coconut oil, α-tocopherol) or by the exact mixture of these components in the commercial solution.
In the present trial, SMOF-treated patients showed a significantly weaker increase in plasma triglyceride concentrations compared with patients receiving a lipid emulsion based on olive and soybean oil. This can probably be interpreted as a sign of a quicker metabolic utilization: SMOFlipid contains coconut oil as a source of MCT. In contrast to LCTs, MCTs cannot be stored in fatty tissues, because they rapidly undergo complete β-oxidation . The effect of their faster utilization can be seen in the smaller increase of triglyceride levels (also shown in the present study), the increase of free glycerol, and the higher concentrations of acetoacetate and β-hydoxybutyrate . The induced ketonaemia is tolerable and does not cause a ketoacidois .
Hyperglycaemia is a common metabolic complication during parenteral nutrition and may contribute to hepatic dysfunction . In our study, the glucose levels of all study groups were similar, and, consequently, it is very unlikely that the increase of liver enzymes was significantly influenced by this pathophysiological approach.
In summary, secondary to parenteral nutrition with SMOFlipid, hepatic integrity was well preserved whereas in postoperative surgery, in patients receiving a lipid emulsion based on olive and soybean oil, the liver enzymes including α-GST were elevated indicating a lower liver tolerability.
However, it has not been proven to date that a therapy associated with lower α-GST levels, meaning less damage on a cellular level which is usually clinically irrelevant, will also cause a lower incidence of clinically relevant damage to the liver. To prove a subclinical damage, histopathological measurements could have been undertaken but were not performed due to both ethical and financial concerns. Furthermore, the study period was too short to expect a serious liver damage in patients without known preoperative hepatic disorder.
The present study was not undertaken to detect differences in outcome parameters. For this purpose, our investigation was clearly underpowered. Therefore, either a very large multicentre study in patients such as ours or a medium-sized study in patients with increased risk of liver damage (i.e. liver surgery or preoperative hepatic problems) is needed to investigate whether parenteral nutrition with SMOFlipid or similar solutions might improve the outcome in critically ill patients.
The study was supported in part by Fresenius Kabi, Germany.
This article was presented in part at the 28th International Symposium on Intensive Care and Emergency Medicine at Brussels (18.03.2008–21.03.2008), at German Anesthesiologists Congress (DAC) at Nuremberg (26.04.2008–29.04.2008), and at the 30th ESPEN Congress at Florence (13.09.2008–16.09.2008).
The corresponding author (S.N.P.) and one of the coauthors (R.B.B.) have received travel gifts and lecture fees from Fresenius Kabi, Germany (the manufacturer of SMOFlipid), respectively. The other authors declare that they have no conflicts of interest.
1 Adolph M. Lipid emulsions in parenteral nutrition
. Ann Nutr Metab 1999; 43:1–13.
2 Mayer K, Meyer S, Reinholz-Muhly M, et al
. Short-time infusion of fish oil
-based lipid emulsions, approved for parenteral nutrition
, reduces monocyte proinflammatory cytokine generation and adhesive interaction with endothelium in humans. J Immunol 2003; 171:4837–4843.
3 Singer P, Shapiro H, Theilla M, et al
. Anti-inflammatory properties on omega-3 fatty acids
in critical illness: novel mechanisms and an integrative perspective. Intensive Care Med 2008; 34:1580–1592.
4 Antébi H, Mansoor O, Ferrier C, et al
function and plasma antioxidant status in intensive care unit patients requiring total parenteral nutrition
: comparison of 2 fat emulsions. J Parenter Enter Nutr 2004; 28:142–148.
5 Grimm H, Tibell A, Norrlind B, et al
. Immunoregulation by parenteral lipids: impact of the n-3 to n-6 fatty acid ratio. J Parenter Enteral Nutr 1994; 18:417–421.
6 Morlion BJ, Torwesten E, Wrenger K, et al
. What is the optimum ω-3 to ω-6 fatty acid ratio of parenteral lipid emulsions in postoperative trauma? Clin Nutr 1997; 16(Suppl 2):49.
7 Grimm H. A balanced lipid emulsion
. A new concept in parenteral nutrition
. Clin Nutr Suppl 2005; 1:25–30.
8 Kinsella JE, Broughton KS, Whelan JW. Dietary unsaturated fatty acids
: interactions and possible needs in relation to eicosanoid synthesis. J Nutr Biochem 1990; 1:123–141.
9 Grimm H, Kraus A. Immunonutrition: supplementary amino acids and fatty acids
ameliorate immune deficiency in critically ill patients. Arch Surg 2001; 386:369–376.
10 Mertes N, Grimm H, Fürst P, Stehle P. Safety and efficacy of a new parenteral lipid emulsion
(SMOFlipid) in surgical patients: a randomized, double-blind, multicenter study. Ann Nutr Metab 2006; 50:253–259.
11 Buchman AL, Ament ME, Sohel M, et al
. Choline deficiency causes reversible hepatic abnormalities in patients receiving parenteral nutrition
: proof of a human cholin requirement: a placebo-controlled trial. J Parenter Enteral Nutr 2001; 25:260–268.
12 Kumpf VJ. Parenteral nutrition
disease in adult and pediatric patients. Nutr Clin Prac 2006; 21:279–290.
13 Porayko MK. Liver
dysfunction and parenteral nutritional therapies. Clin Liver
Dis 1998; 2:133–147.
14 Röhm KD, Suttner SW, Boldt J, et al
. Insignificant effect of desflurane-fentanyl-thiopental on hepatocellular integrity: a comparison with total intravenous anaesthesia using propofol-remifentanil. Eur J Anaesthesiol 2005; 22:209–214.
15 Ozer J, Ratner M, Shaw M, et al
. The current state of serum biomarkers of hepatotoxicity. Toxicology 2008; 245:194–205.
16 Weimann A, Braga M, Harsanyi L, et al
. ESPEN guidelines on enteral nutrition: surgery including organ transplantation. Clin Nutr 2006; 25:224–244.
17 Piper SN, Röhm KD, Boldt J, et al
. Hepatocellular integrity in patients requiring parenteral nutrition
: comparison of structured MCT/LCT versus a standard MCT/LCT- and a LCT- emulsion. Eur J Anaesthesiol 2008; 25:557–565.
18 Moher D, Schulz KF, Altman DG. The CONSORT statement: revised recommendations for improving the quality of reports of parallel-group randomized trials. Ann Intern Med 2001; 134:657–662.
19 Anmacher DE. A toxicologist's guide to biomarkers of hepatic response. Hum Exp Toxicol 2002; 21:253–262.
20 Beckett GJ, Hayes JD. Glutathione S
-transferases: biomedical applications. Adv Clin Chem 1993; 30:281–380.
21 Yoo JSH, Hong JY, Ning SM, Yang CS. Role of dietary corn oil in the regulation of cytochromes P450 and glutathione S
-transferase in rat liver
. J Nutr 1990; 120:1718–1726.
22 Grau T, Bonet A, Rubio M, et al
dysfunction associated with artificial nutrition in critically ill patients. Crit Care 2007; 11:R10.
23 Baker AL, Rosenberg IH. Hepatic complications of total parenteral nutrition
. Am J Med 1987; 82:489–497.
24 Zamir O, Nussbaum MS, Bhadra S, et al
. Effect of enteral feeding on hepatic steatosis induced by total parenteral nutrition
. J Parenter Enteral Nutr 1994; 18:20–25.
25 Lewis SJ, Egger M, Sylvester PA, Thomas S. Early enteral feeding versus ‘nil by mouth’ after gastrointestinal surgery: systematic review and meta-analysis of controlled trials. BMJ 2001; 323:773–776.
26 Heyland DK. Nutritional support in the critically ill patients. A critical review of the evidence. Crit Care Clin 1998; 14:423–440.
27 Röhm KD, Schöllhorn T, Boldt J, et al
. Nutrition support and treatment of motility disorders in critically ill patients: results of a survey on German intensive care units. Eur J Anaesthesiol 2008; 25:58–66.
28 Alwayn IP, Javid PJ, Gura KM, et al
. Do polyunsaturated fatty acids
ameliorate hepatic steatosis in obese mice by SREPB-1 suppression or by correcting essential fatty acid deficiency. Hepatology 2004; 39:1176–1177.
29 Alwayn IP, Andersson C, Zauscher B, et al
. Omega-3 fatty acids
improve hepatic steatosis in a murine model: potential implications for the marginal steatotic liver
donor. Transplantation 2005; 79:606–608.
30 Heller AR, Rössel T, Gottschlich B, et al
. Omega-3 fatty acids
and pancreas function in postoperative cancer patients. Int J Cancer 2004; 111:611–616.
31 Sheth K, Bankey P. The liver
as an immune organ. Curr Opin Crit Care 2001; 7:99–104.
32 Pscheidl E, Schywalsky M, Tschaikowsky K, Böke-Pröls T. Fish oil
-supplemented parenteral diets normalize splanchnic blood flow and improve killing of translocated bacteria in a low-dose endotoxin rat model. Crit Care Med 2000; 28:1489–1496.
33 Piper SN, Haisch G, Kumle B, et al
. Effects of esmolol- and sodium nitroprusside-induced controlled hypotension on hepatocellular integrity in patients undergoing endonasal sinus surgery. Anaesthesiol Intensivmed Notfallmed Schmerzther 2003; 38:781–786.
34 Sailer D, Müller M. Medium chain triglycerides in parenteral nutrition
. J Parenter Enteral Nutr 1981; 5:115–119.
35 Kim H, Son E, Kim J, et al
. Association of hyperglycemia and markers of hepatic dysfunction with dextrose infusion rates in Korean patients receiving total parenteral nutrition
. Am J Health Syst Pharm 2003; 60:1760–1766.