Intravenous Lipid Emulsions in Parenteral Nutrition^,,

Abstract

Fat is an important macronutrient in the human diet. For patients with intestinal failure who are unable to absorb nutrients via the enteral route, intravenous lipid emulsions play a critical role in providing an energy-dense source of calories and supplying the essential fatty acids that cannot be endogenously synthesized. Over the last 50 y, lipid emulsions have been an important component of parenteral nutrition (PN), and over the last 10–15 y many new lipid emulsions have been manufactured with the goal of improving safety and efficacy profiles and achieving physiologically optimal formulations. The purpose of this review is to provide a background on the components of lipid emulsions, their role in PN, and to discuss the lipid emulsions available for intravenous use. Finally, the role of parenteral fat emulsions in the pathogenesis and management of PN-associated liver disease in PN-dependent pediatric patients is reviewed.

Introduction

For patients with intestinal failure who are unable to absorb sufficient nutrients via the enteral route, parenteral nutrition (PN)⁷ is a lifesaving therapy. In the 1960s, PN formulations were designed to provide intravenous carbohydrates, amino acids, electrolytes, and minerals to meet nutritional needs. These initial formulations allowed infants to survive and even meet certain growth markers over a finite period of PN dependence (1). However, patients sustained on fat-free PN in the long term developed essential FA deficiency (EFAD), characterized by growth impairment, developmental delay, dermatitis, and renal and pulmonary abnormalities. Early on, when intravenous delivery of fat was still in its infancy and the first intravenous lipid emulsions were not yet approved for use in the United States, plasma, blood products, and topical oils were often used to provide essential FAs (EFAs) (2). Lipid emulsions for intravenous administration became available in the United States in the 1970s to supply appropriate fat requirements to patients with intestinal failure who require PN for long-term nutritional support. In the present day, these products continue to evolve with the goal of achieving optimal nutrition via the intravenous route.

Nutritional Fat Requirements

Fats serve many purposes in the body that render them biologically vital. They serve primarily as a dense source of cellular energy. However, they are also components of cell membranes, second messengers in cellular signaling cascades, precursors of modulators of inflammation and platelet function, and serve as substrate for de novo biosynthesis of cholesterol and endogenous steroids.

Adequate fat intake for enterally fed infants up to 2 y old is considered ∼30 g/d (3). Average enteral fat requirements decrease with increasing age but remain substantial at 30–40% of total energy for 2- to 4 y-old children, 25–35% for up to 18-y-old adolescents (3, 4), and 20–30% for adults (5). For parenterally fed infants, recommended intravenous fat doses range from 2.5 to 3 g ⋅ kg⁻¹ ⋅ d⁻¹. In children 1–10 y of age parenteral fat doses are typically 2–2.5 g ⋅ kg⁻¹ ⋅ d⁻¹ and decrease to 1–2 g ⋅ kg⁻¹ ⋅ d⁻¹ in adolescents (6). Maintaining fat intake within an appropriate range to meet physiologic demands is important. Consumption of excess fat or carbohydrate that can be converted to fat results in increased de novo lipogenesis in which fat is produced from acetyl-CoA molecules that are common intermediates of protein and carbohydrate metabolism (Figure 1). FAs synthesized endogenously in the setting of nutrient excess are packaged as TGs, which accumulate in adipocytes and the liver, and can lead to nonalcoholic fatty liver disease. Intake of insufficient amounts of fat can result in EFAD.

FIGURE 1.

Acetyl-CoA is a common intermediate of protein, carbohydrate, and fat metabolism, facilitating the conversion of excess protein and carbohydrate calories to fat, which is the densest form of energy storage. Nonbold black text indicates the important macronutrient metabolites. Red text indicates the cellular processes mediating each conversion.

Caloric requirements can be met with parenteral carbohydrates and protein alone; however, high-carbohydrate PN or PN devoid of fat calories can result in EFAD, with the development of hepatic steatosis in both animal models (7, 8) and patients (9, 10). Experimental evidence has suggested that in high-carbohydrate PN administration, excess acetyl-CoA generated from glucose oxidation provides substrate for upregulation fat synthesis within the liver or hepatic de novo lipogenesis (11, 12). Increased expression of mediators and positive regulators of de novo lipogenesis, including FA synthase and the transcription factor sterol regulatory element-binding protein 1, has been identified in mouse models of PN-induced hepatosteatosis (13). The addition of intravenous fat to PN regimens has been shown to prevent the development of histologic hepatosteatosis (7, 13) and normalize de novo lipogenesis (13) compared with formulations lacking fat in isocaloric PN administration.

Other metabolic abnormalities observed in high-carbohydrate PN administration include hyperglycemia and insulin resistance (14). Replacement of 20% of the carbohydrate calories with fat in a rat PN model resulted in normalized serum glucose, lower serum insulin, and improvements in insulin resistance as measured by the HOMA-IR index compared with groups administered isocaloric fat-free PN (15). These studies highlight the important contribution of intravenous lipid emulsions to meeting nutritional requirements in PN.

EFAs

Most FAs can be synthesized in the liver, adipose tissue, and lactating mammary glands (16) from acetyl-CoA, a common product of glucose and protein breakdown (Figure 1). However, mammals are incapable of synthesizing omega-3 (n–3) and ω-6 (n–6) long-chain PUFAs. The n–3 FAs contain double bonds at the third, sixth, and ninth carbons from the terminal methyl group, and n–6 FAs contain double bonds at the sixth and ninth carbons from the terminal methyl group (Figure 2). These FAs cannot be synthesized because of the lack of desaturase enzymes capable of catalyzing double bond formation at the third and sixth bonds from the terminal carbon in long-chain FAs. Therefore, n–3 and n–6 FAs are EFAs and must be obtained in the diet. Insufficient consumption or supply of EFAs can result in EFAD. Clinical manifestations of EFAD include characteristic dermatitis, growth impairment, developmental delay, fatigue, infertility, derangements of hepatic fat metabolism, susceptibility to infections, and, in some cases, pulmonary insufficiency (17–22).

FIGURE 2.

Structure of the EFAs. (A) The parent n–3 FA ALA. All n–3 PUFAs contain more than one double bond, one of which is always at the n–3 carbon position. (B) The parent n–6 FA LA. All n–6 PUFAs contain more than one double bond, one of which is at the n–6 carbon position. ALA, α-linolenic acid; LA, linoleic acid.

Dietary α-linolenic acid (ALA; 18:3n–3) and linoleic acid (LA; 18:2n–6) serve as the parent n–3 and n–6 EFAs and give rise to other n–3 and n–6 PUFAs through the actions of elongases, which lengthen the hydrocarbon FA tail, and desaturases, which insert additional double bonds. These enzymes act on LA to generate arachidonic acid (ARA) (eicosatetraenoic acid; 20:4n–6) and on ALA to generate eicosapentaenoic acid (EPA; 20:5n–3) and docosahexaenoic acid (DHA; 22:6n–3). These enzymes also act on the nonessential ω-9 (n–9) FA oleic acid (OA) (octadecenoic acid; 18:1n–9) to generate mead acid (MA) (eicosatrienoic acid; 20:3n–9). Of their substrates, these enzymes act preferentially on n–3 and n–6 FAs over n–9 FAs (Figure 3).

FIGURE 3.

Metabolic processing of n–3, n–6, and n–6 PUFAs by shared elongases and desaturases. These enzymes have the highest affinity for the n–6 FAs and the lowest affinity for the n–9 FAs. FAs in red are key intermediates for generating modulators of inflammation, coagulation, and cell signaling.

EFAD is biochemically diagnosed by measuring the serum ratio of trienes (PUFAs with 3 double bonds) to tetraenes (PUFAs with 4 double bonds). MA is a triene, and ARA is a tetraene. In the setting of EFAD, when there is a paucity of n–3 and n–6 FAs, the shared enzymes act on the only substrate available, n–9 FAs, to generate MA. In this setting the triene:tetraene ratio is high. When n–3 and n–6 FAs are abundant, the enzymes act preferentially on the n–6 FAs and the triene:tetraene ratio is low. The threshold for a biochemical diagnosis of EFAD is a triene:tetraene ratio of >0.2 (23), although clinical manifestations are often not observed until a triene:tetraene ratio of >0.4 (24).

Lipid emulsions intravenously administered in PN must provide an appropriate amount of EFAs to prevent EFAD. Efforts have focused on identifying the amount of EFAs required to prevent the development of EFAD. In 1960, experiments in rats fed diets with varying doses of high- and low-PUFA fat sources demonstrated an LA requirement of at least 1% of total daily calories in order to prevent EFAD (25). Years later, Barr et al. (26) reported that provision of at least 3.2% of total calories as a soybean oil–based lipid emulsion to PN-dependent patients could prevent the development of EFAD. Although LA and ALA, as the parent n–6 and n–3 FAs, have historically been considered the EFAs, evidence has suggested that consumption of the downstream metabolites of LA and ALA, including DHA, EPA, and ARA, are sufficient to prevent clinical EFAD (27). In a murine model, biochemical EFAD was reversed with DHA comprising 2% of calories in a DHA:ARA ratio of 20:1 (28).

Composition of Intravenous Lipid Emulsions

Normally, dietary fats are emulsified by bile salts in the intestinal lumen and hydrolyzed by pancreatic lipases for uptake by intestinal enterocytes. The liberated long-chain FAs are packaged into chylomicrons for entry into lymphatic channels and, ultimately, into the venous system. Short- and medium-chain FAs may be absorbed from enterocytes directly into the circulation as free FAs bound to albumin. These processes are carefully regulated to ensure that FAs can be transported in the bloodstream in a stable form for energy needs. Intravenous fats bypass the intestinal lumen and therefore do not undergo hydrolysis by lipases, solubilization by bile, uptake into enterocytes, or packaging into chylomicrons compatible for circulatory transport to target organs. Fats administered intravenously must therefore be prepackaged into particles compatible for traveling through the aqueous physiologic environment.

Phospholipid emulsifiers.

Emulsions are systems that consist of 2 immiscible liquid phases, 1 of which is dispersed throughout the other in the form of fine droplets. Lipid emulsions for intravenous use consist of oil suspended in an aqueous dispersion. The dispersion consists of phospholipid that is typically in the form of lecithin from egg yolks, glycerol, and water. Oil is introduced to the dispersion under continuous mixing conditions, and globules are formed consisting of TGs from the oil surrounded by a phospholipid monolayer. These globules are structurally similar to chylomicrons formed in intestinal enterocytes.

Globule size is an important feature of parenteral lipid emulsions for intravenous injection, because globules that are too large can coalesce into large droplets that can result in lipid deposition in various organs along with oxidative damage (29). Hepatosplenomegaly caused by accumulation of fat in Kupffer cells and in the reticuloendothelial system, in addition to respiratory collapse with fat deposits in the pulmonary circulation, have been reported because of coalescence of intravenously administered lipid emulsions (30, 31). Furthermore, oxidative stress and FA peroxidation have been identified after administration of unstable large-globule lipid emulsions in animal models (32).

Smaller globule sizes provide optimal stability and solubility for transport of lipids in the blood. Two globule size specifications that the United States Pharmacopeia requires of intravenous lipid emulsions are a mean globule size of <500 nm and a percentage of fat globules >5 μm of <0.05%. Globule size can be made smaller by increasing the phospholipid:TG ratio (33). This has been confirmed in analyses of commercially available lipid emulsions demonstrating that those containing higher percentages of oil in a given amount of phospholipid emulsifier have a higher percentage of fat globules of >5 μm values and mean globule sizes than emulsions with lower oil concentration (34). In addition to the phospholipid emulsifier, sodium oleate is added as a stabilizing agent and glycerin is added as an osmotic agent.

FA content in intravenous fat emulsions.

Oils are the principle source of TGs in parenteral lipid emulsions. Each oil has a unique FA composition and profile of additives that influence the physiologic effects of parenteral lipid emulsions. First and foremost, the n–3 and n–6 FA contents of intravenous lipid emulsions in PN must be sufficient to prevent EFAD. Clinical studies have also suggested that the n–6:n–3 FA ratio is also important when considering dietary lipid sources. The n–3 FAs, including ALA and its principle metabolites EPA and DHA, are precursors for anti-inflammatory 3-series prostaglandins, 5-series leukotrienes, and several factors that promote resolution of inflammation including resolvins, protectins, and maresins (35, 36). In vitro, high levels of DHA and EPA in macrophage membranes can attenuate the release of IL-6 and TNF-α in response to a lipopolysaccharide challenge (37, 38). The n–6 FAs, LA and its principle metabolite ARA are precursors of 2-series prostaglandins and thromboxanes and 4–series leukotrienes, which are proinflammatory mediators. Diets high in an n–6:n–3 FA ratio correlate with high serum levels of proinflammatory biomarkers (39) and enhance ADP- and thrombin-mediated platelet aggregation (40). Diets high in an n–6:n–3 FA ratio have also been reported in nonalcoholic fatty liver disease (41), coronary artery disease, and atherosclerosis (42). Evidence in animal models has suggested a 1:1 n–6:n–3 FA ratio may be ideal (43, 44); thus, EFA content is an important consideration in lipid emulsions for PN.

Biochemically active moieties in lipid emulsions.

The most physiologically important additives to consider in intravenous lipid emulsions are phytosterols and α-tocopherol. Phytosterols are plant-based sterols that are structurally similar to cholesterol but cannot be metabolized by the human body. When consumed enterally, ~5% of the phytosterols consumed are intestinally absorbed (45). However, intravenous administration of oils containing phytosterols results in 100% uptake into the systemic circulation. Phytosterols, when intravenously administered in animal models, accumulate in the circulation, liver, and bile and may precipitate cholestasis (46, 47). Mechanistic studies have demonstrated that they inhibit the hepatic farnesoid X receptor, a nuclear receptor that promotes bile flow (48, 49). Administration of farnesoid X receptor agonists can prevent phytosterol-mediated cholestasis and liver injury (50).

α-Tocopherol is an antioxidant that is a member of the vitamin E family. PUFAs, such as the n–3 and n–6 FAs, are particularly susceptible to peroxidation and oxidative damage. Oxidation of PUFAs results in the formation of oxygen free radicals, which can bind to DNA and proteins and result in cell damage and death. α-Tocopherol can scavenge free radicals from peroxidized lipids to prevent propagation of oxidative lipid damage (51).

Oil Sources for Parenteral Lipid Emulsions

There are several oils used in the production of lipid emulsions for intravenous administration. Each differs with respect to FA content, EFAD potential, and phytosterol and α-tocopherol content. These properties affect the suitability of an oil as a lipid source for PN. Table 1 compares the major oils available for intravenous emulsions.

TABLE 1.

Oils in intravenous lipid emulsions¹

	Soybean oil	Safflower oil	Olive oil	Fish oil	Coconut oil
FA composition, %
LA (n–6)	50	77	4	1–3	2
ARA (n–6)	0	0	0	0	0
α-ALA (n–3)	10	0	0	1.3–5.2	0
EPA (n–3)	0	0	0	5.4–13.9	0
DHA (n–3)	0	0	0	5.4–26.8	0
Oleic acid (n–9)	25	15	85	16–20	6
MCT	0	0	0	0	65
SFAs	15	8	11	10–20	27
Phytosterol concentration, mg/100 mg oil	300	450	200	Trace	70
α-Tocopherol concentration, mg/100 mg oil	6.4–7.5	34	10–37	45–70	0.2–2

¹ALA, αlinolenic acid; ARA, arachidonic acid; LA, linoleic acid; MCT, medium-chain TG.

Soybean oil.

Soybean oil was the TG source of the first lipid emulsion generated and successfully administered for intravenous use in PN (52). Soybean oil has an abundance of n–6 PUFAs, with LA comprising ∼50% of its FAs. Other PUFAs are also present, with n–3 ALA comprising ∼10% and OA comprising ∼25% of its FAs. SFAs comprise ∼15% of total FAs. Soybean oil can provide a sufficient supply of EFAs to avoid EFAD, thus making it an appropriate choice as a fat source for patients who obtain nutrition exclusively through PN. However, the n–6:n–3 FA ratio is high at ~5:1, raising concerns regarding inflammatory potential. One study in guinea pigs demonstrated increases in serum inflammatory markers and oxidative stress with administration of pure soybean oil emulsions vs. emulsions containing other sources of EFAs (53). However, in comparison with enteral consumption or intravenous administration of oils high in saturated fat, soybean oil has a favorable inflammatory effect profile in animal models (54, 55).

Soybean oil has a high phytosterol concentration, with ∼300 mg phytosterol/100 g oil (56). β-Sitosterol is the predominant phytosterol in soybean oil, comprising ∼60% of the total phytosterols. The phytosterols campesterol and stigmasterol comprise the majority of the remainder at ∼20% each.

The α-tocopherol concentration of soybean oil is low at ∼6.4–7.5 mg/100 g oil. There are higher amounts of other tocopherols (i.e., γ-tocopherol), but they do not have the same antioxidant potential of α-tocopherol.

Safflower oil.

Safflower oil is another plant-based oil that is rich in n–6 PUFAs and contains phytosterols. Its FA composition consists of 77% n–6 LA, ∼15% n–9 oleic acid, and virtually no n–3 PUFAs. Although it contains sufficient PUFAs to prevent EFAD, given its greater abundance of n–6 PUFAs and lack of n–3 PUFAs compared with soybean oil, it is believed that safflower oil may be more proinflammatory than soybean oil as a fat source in parenteral emulsions (57).

Safflower oil, like soybean oil, is high in phytosterols containing ∼450 mg/100 g oil. However, unlike soybean oil, safflower oil has greater antioxidant potential with substantially more α-tocopherol, at ∼34 mg/100 g oil (57). Over recent years safflower oil supply has been limited and is no longer used as an oil source for intravenous lipid emulsions.

Olive oil.

Olive oil is a third plant-based oil used in intravenous lipid emulsions; however, unlike soybean and safflower oils, the FA composition of olive oil is ∼85% of nonessential n–9 FA oleic acid, only ∼4% of essential n–6 FA LA, and is essentially devoid of essential n–3 PUFAs. Olive oil has a paucity of EFAs, and, currently, to our knowledge, there is no commercially available emulsion for parenteral use in which olive oil is the sole fat source.

The phytosterol concentration of olive oil is lower than soybean and safflower oils, but still significant at ∼200 mg/100 g oil. The majority of the phytosterol content in olive oil is composed of β-sitosterol, and the very small remaining fraction is composed of campesterol. α-Tocopherol concentration varies widely depending on region and cultivation methods, ranging from ∼10 to 37 mg/100 g oil (58).

Fish oil.

Fish oil is more abundant in n–3 FAs than plant-based oils; however, the precise FA content depends on the species of fish and the diet composition of the fish from which the oil is derived. In general, fish oil contains a higher proportion of n–3 essential PUFAs than n–6 PUFAs, with ∼1–3% LA and ∼20–40% of n–3 FAs in the form of ALA, EPA, and DHA. The n–9 FAs comprise 16–20%, and SFAs comprise ∼10–20%. Fish oil, therefore, not only contains abundant essential PUFAs to render its potential to reduce EFAD, but it also contains a very low n–6:n–3 FA ratio, which may provide anti-inflammatory benefits. Intravenous fish oil can mitigate responses to inflammatory stimuli both in vitro and in vivo (59–61).

Because fish oil is animal based, it contains minimal plant-based phytosterols. It is also abundant in α-tocopherol, containing ∼45–70 mg/100 g oil.

Coconut oil.

Coconut oil is principally composed of TGs containing six- to twelve-carbon medium-chain FAs [medium-chain TGs (MCTs)]. Lauric acid (dodecanoic acid; 12:0) comprises ∼50% of its FAs, and the other MCTs, caprylic acid (octanoic acid; 8:0) and capric acid (decanoic acid; 10:0), comprise ∼8% each. Other FAs in coconut oil include stearic acid (octadecanoic acid; 18:0) (2%), OA (6%), and LA (2%). Coconut oil contains a negligible amount of essential PUFAs and therefore has a high EFAD potential if used as the sole source of fat calories.

MCTs are smaller in molecular weight compared with the long-chain TGs (LCTs) that are abundant in other oil sources for intravenous emulsions. Their smaller size renders them more soluble in aqueous environments such as the circulation, and, unlike LCTs, which must be transported in the circulation packaged in chylomicrons or similar globules, MCTs can be transported in the blood as free FAs bound to albumin. By virtue of their improved solubility profile, compared with LCTs, MCTs display emulsifier-like properties that improve the stability of total nutrient admixtures in which the fat is mixed directly into the dextrose and protein admixture (62).

In addition, unlike LCTs, which are the principle lipid mediators involved in cell signaling, inflammation, platelet function, and coagulation, MCTs are biologically inert and can thus provide calories for energy without additional systemic effects. In vivo studies have demonstrated improvements in sepsis-mediated liver damage and oxidative stress with use of an MCT/LCT combination compared with LCT lipids alone (63, 64). Evidence has also indicated that MCTs may be a better source of energy than LCTs, despite LCTs having a higher calorie content at 9 kcal/g compared with MCTs at 8 kcal/g. MCTs more easily access mitochondria, where FA catabolism occurs, and are more efficiently oxidized to ATP than LCTs (65, 66). Fat utilization in PN-dependent patients is improved in those who are administered MCT-containing lipids compared with those administered solely LCT-containing lipids (67, 68).

Coconut oil has a low phytosterol concentration at ∼70 mg/100 g oil. β-Sitosterol comprises ∼70% of the coconut oil phytosterols, and stigmasterol and campesterol comprise the remainder. The α-tocopherol concentration of coconut oil is very low, at ∼0.2–2 mg/100 g oil.

Fat Emulsions Used in PN

Fat emulsions available for intravenous use are each unique in oil composition and, therefore, FA and additive content. Each can be parenterally administered; however, both experimental and clinical evidence has indicated that not all emulsions are equally appropriate in all clinical circumstances. A discussion of the properties and clinical applications of the emulsions available in North America is reviewed in the following sections.

Soybean-based lipid emulsions.

To our knowledge, Intralipid (Fresenius Kabi) was the first lipid emulsion formulated and approved for intravenous use (52) and is a 20-g/100-mL pure soybean oil emulsion. The phytosterol concentration of Intralipid is high at ∼350 mg/L, and the α-tocopherol concentration is low at ∼38 mg/L (69). Liposyn III (Hospira Inc.) is another emulsion containing 100% soybean oil, and Liposyn II (Hospira Inc.) contains 50% soybean oil and 50% safflower oil. Like Intralipid, Liposyn is rich in phytosterols, at ∼600 mg/L in Liposyn III (70) and ∼380 mg/L in Liposyn II (69), and low in α-tocopherol, at ∼16 mg/L in Liposyn III and ∼40 mg/L in Liposyn II (71).

In the United States, pure soybean oil lipid emulsions are the predominant fat source in PN, and Intralipid is the most commonly used emulsion in PN. They contain sufficient amounts of EFAs to avoid EFAD. However, given a high ratio of n–6:n–3 FAs and an abundance of phytosterols, the inflammatory and hepatotoxic potential of these emulsions is high. After the development of Intralipid, it was demonstrated that parenteral soybean oil emulsions resulted in lipid accumulation in hepatocytes and in the reticuloendothelial system of patients (72, 73). Studies in animal models showed an exacerbation of sepsis-induced hepatotoxicity with the use of soybean oil emulsions (74). Thus, although soybean-based emulsions may provide the necessary fat calories and EFAs to be an appropriate parenteral fat source, they may not be suitable for use in certain clinical circumstances.

One particular clinical circumstance in which soybean oil–based emulsions may not be the ideal parenteral fat source is in the setting of PN-associated liver disease (PNALD) in PN-dependent infants and children with intestinal failure. PNALD is characterized by the development of cholestasis and hepatic inflammation that can progress to cirrhosis, end-stage liver disease, and a need for liver transplantation if unmanaged. Risk factors for the development of PNALD include prematurity, low birth weight, lack of enteral feeding, long-term PN, and gastrointestinal surgical procedures (75, 76). Although it can be diagnosed by liver biopsy, PNALD is typically identified based on the presence of a serum direct bilirubin level sustained above 2 mg/dL. Javid et al. (7) demonstrated in a mouse model of PN-induced steatosis that changing the route of administration of the lipid from intravenous to oral prevented the development of steatosis, suggesting that the intravenous route of lipid delivery may play a role in the development of PNALD. Studies further showed that substituting the parenteral soybean oil emulsion with an alternative parenteral emulsion containing fish oil could also prevent the development of PN-induced steatosis in mice (77) and reverse PNALD in children (78, 79). More recent studies have focused on understanding the role of soybean oil in PNALD. There has been a particular focus on the role of phytosterols in PNALD onset given the known inhibitory effect of phytosterols on bile flow (48, 49). PN-dependent infants who develop PNALD while administered soybean oil emulsions have a greater accumulation of serum phytosterols than PN-dependent infants who do not develop PNALD (80, 81). Despite a particular research emphasis on phytosterols, an understanding of the mechanism of parenteral soybean emulsion’s role in PNALD remains incomplete and may also include its high n–6:n–3 FA ratio and its relative lack of antioxidants.

Fish oil lipid emulsions.

To our knowledge, Omegaven (Fresenius Kabi) is the only commercially available pure fish oil emulsion. The oil concentration is 10 g/100 mL with an oil composition of 100% fish oil. The fish oil used to make Omegaven is produced from fish off the Western coast of South America and contains ∼30% of the n–3 FAs EPA and DHA. It also contains negligible phytosterols and is abundant in α-tocopherol, containing ∼150–300 mg/L (69).

In the United States, fish oil emulsions are not FDA approved for routine use in PN; however, their use is permitted under a compassionate-use allowance through the FDA for treatment of PNALD in PN-dependent patients. It has been consistently demonstrated over the last 10 y that replacement of soybean oil emulsions with fish oil emulsion monotherapy can reverse PNALD in PN-dependent pediatric patients unable to achieve enteral autonomy (78, 79, 82–86).

Efforts have focused on understanding the basis of the hepatoprotective effects of fish oil. In both animal models and PN-dependent infants, use of parenteral fish oil emulsions is associated with improved serum inflammatory profiles and increased anti-inflammatory mediators compared with parenteral soybean oil emulsions (8). In vitro, fish oil suppresses the lipopolysaccharide-induced inflammatory response in cultured monocytes as well as TGF-β1–induced epithelial-to-mesenchymal transition that characterizes fibrosis in cultured liver epithelial cells (87). A recent study in PN-administered preterm pigs showed that adding 251 mg/L α-tocopherol to a parenterally administered soybean oil lipid emulsion attenuates soybean oil–mediated biochemical evidence of liver injury and reduction in bile acid clearance, thus suggesting a hepatoprotective role for antioxidants in fish oil emulsions (88). Similar to our understanding of parenteral soybean oil hepatotoxicity, our understanding of parenteral fish oil hepatoprotection in the setting of PNALD remains incomplete but is likely multifactorial.

There has been some debate regarding the suitability of fish oil monotherapy for treatment of PNALD. Although fish oil emulsions contain abundant EFAs in the form of n–3 FAs, the n–6 FA content is low. The manufacturer’s label for Omegaven recommends coadministration with emulsions containing n–6 FAs to satisfy EFA requirements and does not recommend fish oil monotherapy. One case study reporting significant bleeding in an infant administered fish oil monotherapy suggested that a lack of ARA metabolites involved in clotting, such as the thromboxane A₂ that is involved in platelet aggregation, could precipitate devastating hemorrhage (89, 90). However, although serum levels of LA and ARA do decline during fish oil monotherapy, they do not reach nonsignificant levels and ARA:MA ratios remain stable at ∼0.02 (91). In addition, children administered fish oil monotherapy in larger cohorts have not demonstrated clinical evidence of EFAD or significant coagulopathies (79, 92, 93). Thus, despite concerns that fish oil contains insufficient n–6 FAs to be the sole fat source in PN, outcome reports of fish oil monotherapy have not had clinical or biochemical evidence of EFAD.

SMOFlipid.

SMOFlipid (Fresenius Kabi) is a mixed lipid emulsion containing 20 g/100 mL emulsion composed 30% soybean, 30% MCT, 25% olive oil, and 15% fish oil. This emulsion contains less phytosterols than soybean oil emulsions alone, at ∼50 mg/L (69), and more α-tocopherol than soybean oil, at ∼200 mg/L (69). Given that only 15% of the oil is fish oil and 30% is soybean oil, the n–6:n–3 FA ratio is high with ∼30% LA and ∼7% ALA, EPA, and DHA combined. SMOFlipid has also been shown to be safe and well tolerated in short-term trials in PN-dependent preterm infants (94).

SMOFlipid may have promise as a preventative therapy for PNALD. In a PN-fed neonatal pig model, SMOFlipid was able to prevent PN-associated decreases in bile flow, biochemical evidence of hepatic dysfunction, and increases in inflammatory markers observed with parenteral soybean oil (95). In randomized controlled clinical trials of SMOFlipid, compared with parenteral soybean oil, SMOFlipid improved liver function as measured by γ-glutamyl transferase (96) and total bilirubin (97). However, these studies took place over 7–14 d, whereas PNALD typically develops over a longer period. In a murine model of P-induced steatosis, SMOFlipid was not able to prevent steatosis as well as parenteral fish oil (69). In one small retrospective cohort study of children who developed PNALD, transition from a soybean oil lipid source to SMOFlipid resulted in resolution of cholestasis in 5 of 8 children, whereas cholestasis was reversed in only 2 of 9 children with PNALD who continued receiving soybean oil (98). One larger retrospective study reported outcomes of using SMOFLipid in 71 PN-dependent children with intestinal failure. Twenty of these patients were switched from a soybean oil emulsion to SMOFLipid after developing biochemical PNALD, and 51 patients started using SMOFLipid at the time of PN initiation because they were at high risk of developing PNALD. In the 51 patients for whom SMOFLipid was the initial parenteral fat source, 4 developed cholestasis, 45 maintained normal bilirubin levels, and 2 died. Among the 20 patients switched to SMOFLipid for treatment of PNALD, cholestasis was reversed in 12, 4 remained cholestatic, and 4 died (99). Lee et al. (100) reported development of biochemical PNALD (direct bilirubin of >2 mg/dL) in 2 PN-dependent infants administered SMOFlipid who subsequently reversed PNALD on transition to fish oil monotherapy. Thus, although SMOFlipid has shown promise in animal models and in early clinical studies in preventing PNALD, long-term data in larger cohorts are required to draw definitive conclusions.

Clinoleic/Clinolipid.

Clinolipid (Baxter Healthcare Corporation) is an emulsion that was recently approved in the United States for parenteral use, although it has been used elsewhere in the world under the brand name Clinoleic (Baxter/Clintec parenteral SA). Its oil concentration is 20 g/100 mL and the oil composition is 80% olive oil and 20% soybean oil. Given its pure plant-oil composition, it is high in phytosterols, at ∼330 mg/L. It is also low in α-tocopherol, at ∼30 mg/L. The FA content of this olive oil/soybean oil emulsion is unique in that there is an abundance of MUFAs, particularly n–9 oleic acid. The olive oil portion of Clinoleic/Clinolipid contains minimal EFAs, and the soybean oil component contains EFAs but comprises only 20% of the emulsion. In an animal model, Clinoleic as a sole fat source resulted in EFAD (69). In a study of 30 adults dependent on PN for a majority of their calories, Clinoleic as the parenteral fat source did not result in EFAD (101).

Clinoleic/Clinolipid contains low levels of the biologically active n–3 and n–6 PUFAs relative to other lipid emulsions. It has therefore been widely tested and has promise for its potential use in inflammatory states including trauma, sepsis, and burn (102). In PN-dependent infants and children with intestinal failure, use of Clinoleic/Clinolipid as the parenteral fat source has some expected benefits with regard to systemic inflammation and oxidative stress (103). However, less is understood regarding the therapeutic or preventive potential of this olive oil/soybean oil combination in PNALD. In a mouse model of PN-induced steatosis, Clinoleic was not able to prevent PN-associated hepatic injury (69). In PN-dependent infants, a 1:1 mixture of Clinoleic and fish oil was able to reverse PNALD in 14 of 16 patients compared with a historical cohort of soybean oil–treated patients in which only 2 of 6 reversed PNALD (104). Another case series of 5 PN-dependent infants who were administered a 1:1 intravenous fat mixture of Clinoleic and fish oil reported reversal of PNALD in 4 of the 5 infants (105). To date, to our knowledge, no studies have investigated Clinoleic/Clinolipid monotherapy in the management of PNALD. One recent retrospective cohort study of 517 PN-dependent infants administered either a pure soybean oil emulsion or Clinoleic as their fat source had no significant difference in the development of cholestasis between the 2 groups (106). Although this study suggests that Clinoleic as a sole fat source may not confer benefits in preventing the development of cholestasis, it must be noted that the infants who were administered Clinoleic collectively had more risk factors for the development of PNALD, including more days on PN, lower gestational age, and greater need for mechanical ventilation.

Conclusions

Intravenous lipid emulsions have constituted an important element in PN over the past 50 y. The emulsions currently available for use in PN all satisfy the requirement of providing sufficient amounts of EFAs to avoid EFAD. Pure soybean oil emulsions administered over the long term may have adverse effects on inflammatory markers that may be caused by the high content of ARA precursors, the paucity of antioxidant tocopherols, or the impact of phytosterols on bile flow. These features may contribute to the development of PNALD and are limiting factors in dosing intravenous fat. Many of the more newly developed fat emulsions available for parenteral use contain lower levels of n–6 FAs and phytosterols that may offer advantages in treating or preventing PNALD and other systemic inflammatory insults associated with long-term PN use. The optimal balance of fat and other macronutrients in PN to support growth, development, and the physiologic responses to illness and injury for which PN is often administered remains unknown. As access to these alternative lipid emulsions improves, it may foster further investigations aimed at understanding the optimal balance of parenteral nutrients and the impact of parenteral fats in PN-dependent patients.