Emerging Clinical Benefits of New-Generation Fat Emulsions in Preterm Neonates

Abstract

Soybean oil–based intravenous fat emulsions (IVFEs) have been the predominant parenteral nutrition IVFE used in the United States for neonates over the past 45 years. Even though this emulsion has proven useful in supplying infants with energy for growth and essential fatty acids, there have been concerns over its composition in the development of several morbidities, ranging from sepsis to liver disease, bronchopulmonary dysplasia, and impaired neurodevelopment and growth. The exact mechanisms that drive these morbidities in preterm infants are multifactorial, but potential contributors include high ω-6 (n-6) fatty acid composition, low docosahexaenoic acid and antioxidant supplementation, and the presence of potentially harmful nonnutritive components (eg, phytosterols). To address these issues, new-generation IVFEs with various types and amounts of fat have been developed containing greater amounts of the medium-chain fatty acids, long-chain polyunsaturated fatty acid, docosahexaenoic acid, lower concentrations of ω-6 polyunsaturated fatty acids, supplemental vitamin E, and low or negligible amounts of phytosterols. This review examines the clinical outcomes associated with different morbidities of parenteral nutrition in neonates who have received either soybean oil–based or new-generation IVFEs and addresses whether the proposed benefits of new-generation IVFEs have improved outcomes in the neonatal population.

Infants who are unable to feed with diagnoses such as extreme prematurity and short bowel syndrome following congenital or acquired gastrointestinal conditions are now able to survive mainly due to the availability of parenteral nutrition (PN). PN consists of amino acids, glucose, fats, electrolytes, minerals, and vitamins balanced to meet nutritional need. The fat component of PN serves as a source of nonglucose energy, to avoid hyperglycemia from excess glucose load, and to prevent essential fatty acid deficiency. The use of intravenous fat emulsions (IVFEs) based on soy oil for PN was implemented, as they are high in the essential ω-6 (n-6) polyunsaturated fatty acid (PUFA) linoleic acid and contain moderate amounts of the essential ω-3 (n-3) PUFA α-linolenic acid.

In the United States, Intralipid (Fresenius Kabi, Friedberg, Germany), a soy oil–based IVFE (SOFE), has been the predominant IVFE used for preterm infants for the past 45 years. While this emulsion has been very effective in reducing overall mortality, concern has recently developed whether this emulsion is contributing to common morbidities associated with preterm infants receiving PN. In preterm infants, there is strong evidence that preformed n-3 PUFA docosahexaenoic acid (DHA) is necessary for brain and retinal development.^1,2 In addition, there are concerns about the proinflammatory effects of n-6 PUFAs.³ The presence of nonnutritive phytosterol compounds in SOFEs has also come under scrutiny for their potential role in suppressing gene expression pathways in bile acid homeostasis.⁴ Inflammation and bile acid pathway dysregulation are associated with the morbidities of PN-associated liver disease (PNALD), sepsis.

This review provides a brief overview of the clinical experience with the currently available IVFEs. The following sections summarize the current clinical literature on the most common morbidities associated with PN administration, the potential mechanisms of action, and the impact of new-generation IVFEs on incidence of these morbidities.

IVFEs

In 1929, Burr and Burr were the first to describe the essentiality of fat in the diet when they described features of essential fatty acid deficiency in their laboratory animals.⁵ In 1968, Douglas Wilmore and Stanley Dudrick gave intravenous nutrition to a 2300-g infant with intestinal atresia for a period of 44 days and demonstrated improvement in growth.⁶ Soon thereafter, they published a case series of 18 infants with gastrointestinal conditions, including surgical diagnoses, who were supported with PN for up to 400 days.⁷ The parenteral solution administered in these landmark studies did not contain any IVFE, as none were available, so plasma was administered to supply essential fatty acids. Following this, Benda and Babson published their experience with 14 preterm infants weighing <1251 g who were safely given PN for a maximum period of up to 3 weeks.⁸

In the United States, Intralipid (SOFE) was first approved in 1972 for general use. To date, Intralipid remains the most commonly used fat source for PN and is derived from 100% soybean oil rich in n-6 PUFA. Apart from Intralipid, ClinOleic (olive oil–based IVFE [OOFE]; Baxter, Deerfield, IL) is another IVFE approved by the Food and Drug Administration for use in the United States; it is composed of 80% olive oil and 20% soybean oil (SO). Since the 1990s, Omegaven (Fresenius Kabi; fish oil–based IVFE [FOFE])—a IVFE derived entirely (100%) from fish oil rich in n-3 PUFA—has been increasingly used across the United States, strictly under Food and Drug Administration–approved compassionate use protocols for the treatment of PNALD. Other fat preparations that have been used in the United States and elsewhere include Lipofundin (50% MCTs and 50% soybean oil; B Braun, Philippines; coconut oil–based IVFE [COFE]), Lipidem (50% MCTs, 40% soybean oil, and 10% fish oil; B Braun; 3 oil–based IVFE [TOFE]), and newer-generation multicomponent IVFEs, such as SMOFlipid (30% soybean oil, 30% coconut oil–derived MCT, 25% olive oil, 15% fish oil; Fresenius Kabi; mixed oil–based IVFE [MOFE]). As of 2016 in the United States, SMOFlipid has been approved for use in adult patients but is not yet approved for pediatric use.

IVFEs and PNALD

In 1971, just 3 years since the first use of PN in infants, Peden and colleagues reported the death of an infant receiving PN who had developed liver failure and showed signs of cholestasis.⁹ Five years from the first description of PN in infants, Touloukian and Downing described 3 of 18 infants who succumbed to various complications following PN, including cholestasis.¹⁰ Over the next 2 decades as the association of liver disease with use of PN grew stronger, suspected etiologies included hyperglycemic load, deficiencies of various amino acids and trace elements, the fat component of PN, and free radical injury following exposure to light. Allardyce observed that the majority of patients who received higher doses of SOFE developed cholestasis, whereas those who received lower doses were spared from it, thereby suggesting a dose-dependent relationship between SOFE and cholestasis.¹¹ Results originating from laboratory animal studies further strengthened the link between the fat component dose of SOFE and PNALD.¹² Over the next 2 decades, several components of SOFE have been implicated in the pathogenesis of PNALD, including its predominant n-6 PUFA profile, high phytosterol content (particularly the stigmasterol), and low α-tocopherol levels.

PNALD is a complication of the liver in infants receiving PN. PNALD includes the biochemical and histologic alterations representing cholestasis, elevation of transaminases, and alteration of synthetic functions of the liver. The common serum biochemical markers of PNALD are an elevated conjugated bilirubin (CB) >2 mg/dL along with elevations in transaminases. The use of PN in modern neonatal intensive care units is common, as demonstrated by Christensen et al, who showed that nearly 70% of the 10,000 infants in their neonatal intensive care units received PN at some point and 21% of them for >14 days.¹³ In the same report, PNALD was reported to be more common with immaturity and lower birth weights, longer duration of PN, and surgical gastrointestinal conditions.¹³ The pathogenesis of PNALD is complex and multifactorial. Apart from exposure to PN, several other risk factors contribute to the pathogenesis of PNALD, such as absence of enteral feeds, prematurity, sepsis, states of inflammation (eg, necrotizing enterocolitis), gastroschisis, gastrointestinal surgery, and hypoperfusion injury, as seen in the presence of patent ductus arteriosus. Studies have suggested that an exposure to PN for a period as short as 2 weeks can induce PNALD in such high-risk population of infants. Prior to the widespread use of currently practiced interventions to treat PNALD, such as fat-lowering strategy and use of FOFEs, infants whose CB was >10 mg/dL had a nearly 40% chance of death or a liver transplant.¹⁴

Following the discoveries of the associations of PN with liver disease, several strategies were adopted to limit and reverse the damage inflicted by PN. These strategies included cycling of PN and use of choleretic agents, such as ursodeoxycholic acid and cholecystokinin, all of which yielded very limited success.^15–17 One strategy that has been more effective in limiting liver disease is the fat-lowering strategy, where SOFE is given at a reduced dose of 1–1.5 g/kg/d instead of the regular dose of 3–3.5 g/kg/d. Even though the association of lower fat dosages with better liver health was first described by Allardyce,¹¹ rediscovery of the same association by Cavicchi et al a decade later renewed interest in this strategy.^18,19 It was noted that that those adults who received a smaller dose of fats had better outcomes and fewer liver-related complications. This practice was later adopted in the treatment of PNALD in the pediatric population with better but still suboptimal results.^20,21

Gura et al described their experience with 2 soybean-allergic infants with short bowel resections who subsequently developed essential fatty acid deficiency and PNALD. Treatment with Omegaven in these infants resulted in not only reversal of cholestasis but also of essential fatty acid deficiency.²² Since then, several institutions across the United States have successfully treated >1000 infants of PNALD with FOFE using compassionate use protocols.^23,24 In spite of the reported improved outcomes with the use of FOFE, a significant proportion of clinicians are skeptical about its purported benefits. Following the initial reports, FOFE has always been used in a dose of 1–1.5 g/kg/d, the same low-fat dosage at which several groups have shown similar liver-protective actions even with the use of SOFE. This and the absence of rigorous randomized controlled trials (RCTs) comparing the effectiveness of SOFE with FOFE in the resolution of PNALD have been the major criticisms against the supposed beneficial effects of FOFE.

In an attempt to address these deficiencies, a recently concluded RCT in infants compared the effect of FOFE and SOFE in the prevention and treatment of PNALD in which both IVFEs were given at a dose of 1.5 g/kg/d. Though the study failed to show any difference in the reversal of PNALD and was terminated prematurely, the rate of rise for CB and alanine amino transferase was lower in those infants who received FOFE. It was also noted that the infants in the FOFE group appeared to have better tolerance to enteral feeds.²⁵ Nehra et al conducted a double-blind RCT comparing FOFE and SOFE in neonates for prevention of PNALD, both given at a dosage of 1 g/kg/d. By the time that the low incidence of PNALD in both groups forced the investigators to end the study earlier than they had planned, the incidence of PNALD in the 2 groups was not significantly different.²⁶ The limitation of these 2 aforementioned studies is that they were conducted with relatively low-fat infusion rates resulting in low overall rates of PNALD. Several single-center RCTs have been conducted comparing the effects of SMOFlipid (MOFE) with SOFE, but none of these studies showed a difference in any of the biochemical indices suggestive of PNALD between the groups.^26–29 In one such study, Goulet et al compared the effects of MOFE with SOFE in pediatric patients receiving home nutrition and failed to show any benefits in terms of prevention of treatment of PNALD in these children.³⁰ In another study, 5 IVFEs—SOFE, MOFE, COFE, TOFE, and OOFE—were studied in infants with birth weight <1250 g. The rates of cholestasis were not only similar but also low across all treatment arms.³¹

All these single-center RCTs have been plagued by lower rates of cholestasis, thus being inadequately powered to address PNALD as a primary outcome in assessing the treatment with IVFE. Though the obvious solution to this issue would be to perform large multicenter RCTs, studies of this scope are lacking in current literature. To overcome this limitation and to synthesize meaningful information from the available small-scale studies, several meta-analyses have been conducted to study the effect of IVFE in the prevention and treatment of PNALD.

A recent meta-analysis was published comparing the use of FOFE with non-FOFE, assessing its efficacy in the prevention and treatment of PNALD in neonates. FOFEs included Omegaven, SMOFlipid, and Lipidem; non-FOFEs included Intralipid, Lipofundin, and ClinOleic preparations. The criteria for inclusion in this meta-analysis were less stringent, as it included case-control and prospective and retrospective cohort studies. In spite of these lenient criteria, only 7 studies fulfilled the requirements: 3 of which involved 93 subjects and addressed reversal of PNALD, while the other 4 involved 1012 subjects and addressed prevention of PNALD. The analyses suggested that the use of FOFE was more likely to reverse PNALD (odds ratio: 6.14, 95% CI: 2.27–16.6, P < .01) but failed to show any benefit of FOFE in prevention of PNALD. Still, the use of Omegaven was associated with near-significant trend toward decreased development of PNALD (odds ratio: 0.13, 95% CI: 0.02–1.03, P = .05).³²

In the recently concluded Cochrane meta-analysis, with more stringent inclusion criteria, randomized or quasi-RCTs in preterm infants compared newer alternative IVFEs with SOFE. Various alternative IVFEs were studied in these 15 trials—including MCTs/long-chain triglycerides, MCTs and olive/fish/soy oil, MCTs and fish/soy oil, olive/soy oil, and borage/soy oil—in comparison with SOFE. There were no studies that compared 100% FOFE (Omegaven) with SOFE. Apart from 1 study in which alternative IVFE showed a decrease in stage 1 and stage 2 retinopathy of prematurity, there were no other benefits attributed to their use. It was concluded that based on the level of evidence presented in this review, the use of alternative IVFE, including FOFE, over the current standard of SOFE was not supported.³³

The European Society for Pediatric Gastroenterology, Hepatology, and Nutrition performed a meta-analysis aiming to study the pathogenesis of PNALD.³⁴ This meta-analysis included 23 studies involving neonates and older children with either short-duration (<4 weeks) or long-term use of IVFE. This meta-analysis failed to show any beneficial effects with the short-term use of multicomponent fish oil containing IVFEs over the use of Intralipid. A meta-analysis of the long-term use of multicomponent IVFE in the pathogenesis of PNALD could not be performed because of the paucity of studies. However, in one of the studies included in this meta-analysis, the long-term use of multicomponent FOFE was associated with some benefit.²⁶ Noncholestatic infants with intestinal failure who received multicomponent fish oil–based fat emulsion over prolonged periods exhibited a reduction in CB levels.

IVFEs and Sepsis

The actions of n-3, n-6, and n-9 PUFAs include wide-ranging effects on cell membrane structure and function, eicosanoid signaling, nuclear receptor activation, and fat metabolism. These PUFAs have been shown to modify leucocyte activity by altering (1) neutrophil, monocyte, and macrophage migration; (2) leucocyte adhesion to the endothelium; and (3) T-cell proliferative capacity to antigenic stimuli. n-6 PUFA (arachidonic acid and linoleic acid) metabolites—such as cyclooxygenase-derived prostaglandin-E2 and lipoxygenase-derived leukotriene B4—are predominantly proinflammatory, whereas n-3 PUFA–derived metabolites (eg, PG-E3, leukotriene B5) are predominantly anti-inflammatory in their actions. n3-PUFAs also generate protectins, D-series resolvins and maresins, and E-series resolvins that further extend their anti-inflammatory profile. n-3 PUFAs also act at the transcriptional level, thereby inhibiting activation of NF-kB and decreasing the production of various inflammatory cytokines such as tumor necrosis factor alpha and interleukins 1β, 6, and 8.^35,36 Because of these broad proinflammatory and anti-inflammatory effects on the inflammation pathway conferred by their predominant n-6 or n-3 PUFAs, the IVFEs have been thought to alter the risks of sepsis in infants who receive them.

Sepsis independently alters the profile of fat metabolism in critically ill infants. Patients with sepsis have elevated levels of plasma free fatty acids even in the absence of IVFEs, due to enhanced lipolysis from adipocytes, increased de novo lipogenesis in the liver, and decreased muscle fatty acid oxidation. This imbalance in fat metabolism is exacerbated by exogenous therapies used in critically ill infants, such as vasopressors (epinephrine and norepinephrine) and heparin, which activates lipoprotein lipase. Sepsis has also been shown to impair the ability of a premature neonate to handle intravenous fat load, as evidenced in infants with septicemia, in whom a higher mean serum triglyceride and free fatty acid levels are observed.³⁷ This state of dyslipidemia is exacerbated in the presence of morbidities such as growth restriction and prior liver dysfunction.³⁸ In states of sepsis, neutrophils exhibit decreased responsiveness to antigenic stimulation, secretion of leukotriene B4, and generation of platelet-activating factor and phosphatidylinositol.³⁹ As mentioned, various PUFAs also serve as substrates for products of inflammation with pro-inflammatory and anti-inflammatory profiles. Due to this extensive and complex interaction between sepsis and fats, it has been hypothesized that the use of IVFEs alters the risk of sepsis in neonates and infants.

In a retrospective case-control study performed in nearly 900 infants in neonatal intensive care units, the administration of IVFEs was shown to have a strong association with coagulase-negative staphylococcal bacteremia even when corrected for other risk factors. Those infants who received IVFEs were noted to have 5.8-times increased odds (95% CI: 4.1–8.3) of hospital-acquired coagulase-negative staphylococcal bacteremia.⁴⁰ In a study conducted in 40 premature infants with bloodstream infections, the dose of MOFE was compared with time taken for bacterial clearance. The experimental dose of MOFE, involving a restricted dose of 1 g/kg/d, was associated with significantly rapid clearance of bacteremia and reduction in antibiotic use duration in comparison with a standard dose with increases up to 3.5 g/kg/d. This study proposed that the protective effect of this MOFE is perhaps restricted to lower doses of fat load.⁴¹

Beken et al studied the effects of MOFE, in comparison with SOFE, in 80 very low birth weight (VLBW) infants with retinopathy of prematurity as their primary end point and showed that the rate of sepsis, a secondary outcome in this study, was no different between the groups.²⁷ Similarly, there were no significant changes in the rate of late-onset sepsis when D’Ascenzo et al compared various doses of MOFE and SOFE with an intention to study fat tolerance as the primary outcome.⁴² Savini et al extended the comparison by studying the effect of 5 IVFEs—SOFE, MOFE, COFE, TOFE, and OOFE—in 150 infants with birth weight <1250 g. While the primary outcome was the plasma phytosterol levels, rate of sepsis (included as a secondary outcome) was not different across these regimens of fat preparations.³¹ Similarly, several studies have been performed regarding VLBW infants that compared the effects of MOFE with SOFE, albeit with different primary outcomes, and they have failed to show superiority of one fat emulsion over the other in the prevention of sepsis.^29,43

All of these studies were limited by smaller numbers of subjects and were inadequately powered to study sepsis as a primary outcome. A recent meta-analysis assessed the effects of early initiation of fat therapy on the general well-being of VLBW infants and compared the actions of SOFEs with non–soybean oil based fat emulsions. This included 2 studies in which SOFE was provided within the first 48 hours and compared with a delayed initiation^44–46 and 3 studies that compared SOFE with MCTs/long-chain triglycerides,⁴⁶ MOFE,⁴³ and OOFE.⁴⁷ This analysis showed trends toward decreased incidence of sepsis in infants who received IVFEs that were not purely soybean oil based (relative risk: 0.75, 95% CI: 0.56–1.00), thereby suggesting a protective effect of non–soybean oil based fat emulsions against sepsis in VLBW infants.⁴⁸

IVFEs and BPD

BPD is a complex sequence of lung development, injury, and repair that is initiated by an in utero insult, postnatal management, and aberrant lung tissue repair mechanisms.^49,50 The diagnosis of the disease is based on breathing abnormalities that arise from injury to the lung parenchyma with decreased microvascular development, alveolar septation, and airway injury.⁴⁹ Historically, BPD has been clinically assessed by various criteria including an oxygen requirement at 36 weeks corrected postmenstrual age in VLBW infants⁵¹ and a more rigorous definition of preterm infants <32 weeks requiring supplemental oxygen at 36 weeks postmenstrual age or preterm infants >32 weeks requiring supplemental oxygen >28 days but <56 days postnatal age.⁵² Chronic lung disease (CLD)/BPD is of considerable concern due to the long-term effects on growth failure^53–55 and impaired neurodevelopment.^56–58 The mechanisms by which IVFE can contribute to BPD are not fully understood; however, some potential targets include impaired pulmonary oxygen diffusion,^59,60 fat agglutination leading to embolism,⁶¹ oxidant stress,^62,63 and inflammation.⁶⁴

Preterm infants are more susceptible that term infants to oxidant stress,^65,66 and high oxidant stress in preterm infants has been correlated with CLD/BPD.⁶⁷ The higher oxidant stress in preterm infants is likely due to lower levels of glutathione as compared with term infants, which leads to impaired redox potential.⁶⁸ Further connection to this is the observation that levels of glutathione are typically lower in male preterm infants than females, and they have a tendency for a greater incidence of BPD.^68,69 SOFE is prone to oxidation, which can generate fat hydroperoxides, especially in the presence of a high concentration of oxygen from mechanical ventilation, which could facilitate increased oxidant stress in the preterm infant.⁷⁰

Hammerman and Aramuro were the first to show, in a small cohort (n = 43) of VLBW neonates, that by withholding IVFE for 5 days, they could decrease the incidence rate of BPD.⁷¹ Cooke gave further strength to the concept of IVFE-induced BPD by examining >650 preterm infants born ≤30 weeks, out of which 195 infants received supplemental oxygen for >4 days and lived to 28 days.⁷² From the infants receiving supplemental oxygen, 87 developed BPD. There was a strong association with the infants developing BPD and receiving fats during their first 21 days. These early studies presented a strong case for not only the administration of IVFE leading to BPD but that specifically early administration of fat could be the driver of the disease. However, more recent studies have not supported that idea. Alwaidh et al looked at 64 VLBW preterm infants who received fat at 5 or 14 days after birth and saw no difference in BPD incidence.⁷³ Ibrahim et al administered earlier PN at 2 and 48 hours following birth and did not increase the incidence rate of BPD in VLBW infants either.⁷⁴ A 2012 meta-analysis by Vlaardingerbroek et al examined 4 clinical studies stratified by IVFE administration initiated before 2 days and after 2 days and also failed to see any overall effect on the incidence of BPD/CLD (relative risk: 0.88, 95% CI: 0.68–1.14).⁴⁸ The results of these later studies suggest that the very presence of parenteral fat is likely a risk factor for BPD rather than any specific time point when it is administered. In support of this, infants who receive >50% of their total caloric intake from enteral feeds do have a lower incidence of BPD/CLD.⁷⁵

Fats such as OOFE contain much higher concentrations of monounsaturated fatty acids, making them more stable and less likely to undergo peroxidation as compared with soy-only emulsions. IVFEs that contain fish oil (FOFE and MOFE) contain n-3 PUFAs, DHA, and eicosapentaenoic acid, which are at risk for fat peroxidation, but are supplemented with 500 μmol/L of synthetic all-rac-α-tocopherol to prevent oxidation. Several small studies examined the rate of oxidation of OOFE in preterm infants. Pitkanen et al measured the effect of fat peroxidation in VLBW infants with respiratory distress. This small study (n = 13) used a short-term 3-hour administration of VasoLipid, an MCT-soy emulsion, and compared fat oxidation by pentane exhalation with ClinOleic. There were no differences on the extent of fat peroxidation by the administration of either IVFE.⁷⁶ A similar result of no effect on fat peroxidation was seen in a longer-term study by Koskal et al on VLBW infants administered OOFE or SOFE following 7 days.⁷⁷ This study also looked at the outcome of BPD in the infants. There was a significantly lower incidence of BPD, with only 31% of infants receiving OOFE versus 69% of infants receiving SOFE. The beneficial effects of OOFE compared with soy-based emulsions in ELBW infants have not held up in subsequent studies.^31,78 However, the trend for these studies suggests that in extremely low birth weight (ELBW) and VLBW preterm infants, administration of monounsaturated-rich olive oil containing fats does not appear to improve the outcomes on BPD or have much effect on oxidant stress.

Fish oil containing IVFEs could represent a far better option for the prevention of BPD in preterm infants as compared with monounsaturated-rich fat emulsions. As mentioned, FOFE are supplemented with vitamin E to reduce fat peroxidation while in storage and to limit oxidative stress once infused into infants. DHA also has well-established anti-inflammatory properties that may have a role in reduced lung inflammation.⁷⁹ Of particular benefit to very preterm infants, DHA could aid in lung maturation through activating peroxisome proliferator-activated receptor gamma.^80,81 DHA has an important role in the prevention of CLD in infants, as low DHA levels strongly correlate with a high CLD incidence rate.⁸² This finding suggests that DHA supplementation could reduce the incidence of CLD. Two studies found that the administration of MOFE for 8 and 14 days to VLBW preterm infants led to reduced markers of oxidative stress when compared with control IVFE lacking fish oil and vitamin E supplementation.^43,83 In a prospective observational study by Skouroliakou et al, the administration of MOFE in place of SOFE led to a significantly lower incidence of BPD in VLBW infants (6% vs 23%, respectively) but did not affect BPD outcomes in low birth weight infants.⁸⁴ Results in low birth weight are mostly due to the low overall incidence of BPD in this population of neonates. However, this research group has a recent 2016 randomized controlled double-blind study of 71 VLBW preterm infants that does not show a significant difference in BPD incidence between MOFE and SOFE (24% vs 46%, respectively).⁸⁵ This outcome may be due a lack of statistical power as their prospective study had a sample size of 129 neonates and their current study only had 51 neonates. However, studies by Savini et al, Vlaardingerbroek et al, Deshpande et al, and D’Ascenzo et al also failed to see a difference in BPD outcomes in VLBW infants receiving MOFE compared with SOFE.^28,29,31,83 D’Ascenzo et al administered MOFE and SOFE at 2.5 and 3.5 g/kg/d and saw no difference at either concentration nor a dose-related effect in BPD.²⁸ Pawlik et al administered OOFE or a 50:50 mix of OOFE:FOFE emulsion on a mixed VLBW/ELBW population of preterm infants.⁸⁶ The OOFE group had a 27% incidence rate of BPD, compared with a 23% incidence rate in the OOFE/FOFE mix group, which was not significantly different. These incidence rates also fall within the ranges seen for the development of BPD in studies using SOFE. The most recent meta-analysis from the Cochrane database by Kapoor et al fails to show a treatment effect of new-generation IVFE compared with SOFE for BPD.³³ The current data are mixed on the beneficial effect from using a different type of IVFE in parenteral formulas for the prevention of BPD. Fewer studies show positive effects overall. Larger multicenter trials with the treatment of MOFE are needed to draw definitive conclusions.

Pulmonary inflammation is a major contributor to the development of BPD.⁶⁴ Inflammation from IVFE is believed to be derived from the high n-6 PUFA content present in SOFE. As discussed in detail in this review, the n-3 PUFAs in new-generation IVFEs (FOFE and MOFE) are thought to be anti-inflammatory. The most studied aspect of this has been with allergic response through maternal supplementation with DHA, with the best example being the DINO study (Docosahexaenoic Acid for the Improvement of Neurodevelopmental Outcome in Preterm Infant Trial). Lactating mothers were supplemented with DHA to increase milk DHA concentration to 1% of total fatty acids. Manley et al analyzed data from this study and found that preterm males ≤1250 g had reduced incidence of hay fever.⁸⁷ However, this effect was not seen in preterm females or low birth weight males. There were also no differences in other forms of allergic response, including asthma, eczema, or food allergies. There have also been more direct findings that DHA supplementation during pregnancy can lead to downregulation of interleukins 4 and 13 in neonates.^88,89 There is limited research directly showing an effect of n-3 PUFAs on an infant’s inflammatory response as it pertains to lung function. An RCT by Deshpande et al saw increases in inflammatory markers (interleukins 6 and 8) in VLBW preterm infants receiving SOFE day of life 30, whereas MOFE-administered infants had no change or a slight nonsignificant reduction in both cytokines.⁸³ The incidence of BPD failed to reach a significant difference between groups; however, the SOFE group had a 46% BPD incidence and the MOFE group, a 24% BPD incidence. A better-powered study could have seen benefit from MOFE treatment.

IVFEs and Neurodevelopment

Preterm infants, especially ELBW preterm infants, are at a higher risk of having poor neurodevelopmental outcomes.^90–97 DHA incorporation into the brain is necessary for cell signaling through altering fluidity of lipid rafts,⁹⁸ modulating sodium and calcium channels in n eurons,⁹⁹ and neurogenesis.¹⁰⁰ In addition to brain function, DHA serves as a neuroprotective molecule through generation of bioactive compounds such as neuroprotection D1.¹⁰¹ The fetal brain increases to a total fat content of 1.45, 4.74, and 8.79 g at 25, 35, and 40 weeks, respectively.¹⁰² This increase in brain fat is accompanied by an increase in total DHA content to 102, 356, and 682 mg at 25, 35, and 40 weeks, respectively. Interestingly, this DHA increase maintains the total DHA content in the brain to 23% of all bodily DHA stores. The main source of DHA during fetal growth is through maternal placental transfer, so preterm birth represents an abrupt loss of preformed DHA for the infant. As the accretion rates show, preterm infants born at <35 weeks are at risk for a severe shortage of DHA to reach levels seen at 40 weeks. In addition to brain accretion of DHA, adipose tissue undergoes a selective increase in DHA during the final trimester.¹⁰² The loss of this DHA reservoir can be seen by the first postnatal week where serum DHA drops from 7 to 4.5 mol%.⁸² Preterm infants do show some capacity to synthesize DHA from precursor fatty acids, but it is limited and decreases over the first 7 months of life.¹⁰³ So, the need for appropriate supplementation of preformed DHA is potentially very important for long-term outcomes in this infant population.

Most studies looking at the effects of DHA use cognitive assessment tools to determine efficacy. The standard assessment tool for neurodevelopment in children aged 1–42 months is the Bayley Scale of Infant Development II (BSID-II).¹⁰⁴ A main component of the BSID-II is the Mental Development Index (MDI), which is a measure of nonverbal cognitive and language development. The MDI is graded on scores of neurodevelopmental delay: >85%, normal; 70%–84%, mild; 55%–69%, moderate; and <55%, severe. A newer version, the Bayley-III, was published in 2006 and has cognitive and language scores but does not use the MDI.¹⁰⁵ The Bayley-III scores tend to be 10 points higher than the BSID-II scores, and some recent studies showed that an MDI <70 correlates with Bayley-III cognitive and language scores <85.¹⁰⁶ In addition, there is some concern that the BSID is not a good assessment of long-term neurodevelopment.¹⁰⁷ These 2 points are important to remember for outcomes presented after the Bayley update, especially in the context of meta-analyses conducted before 2014 (the date of the most recent cutoff data for Bayley-III).

The majority of DHA studies conducted looked at the use of enteral administration of the fat in preterm infants. Recent studies show that administration of supplemental DHA directly to the infant can help in preventing the drop of DHA levels observed after birth. Baack et al fed preterm infants 50 mg/d of preformed DHA or an MCT control, starting within the first week of birth.¹⁰⁸ Infants receiving the DHA supplement had a significant increase in DHA serum concentrations (2.88–3.55 mol%) as compared with the serum concentrations of the MCT group (2.91–2.87 mol%) by time of discharge or at 38 weeks postmenstrual age. The DHA-supplemented group failed to reach the levels of serum DHA in term infant controls, which were at 4.31 mol%. These values are considerably lower than the DHA levels found in term-born infants in the study by Martin et al discussed earlier.⁸² Similar benefit was seen by Collins et al, who gave doses of 40, 80, and 120 mg/kg/d of DHA starting at days 5–6 of life.¹⁰⁹ The increasing dose of DHA led to significant mol% enrichment of DHA in red blood cells to a level near 6% at 28 days in the 80- and 120-mg/kg/d groups. However, this study did not have gestational age–matched term infants for a control group, so near to normal levels can be assumed only on the basis of preexisting data. The interesting results from this study show that even with the high intake of 120 mg/kg/d of DHA, the infants still cannot reach the levels seen in term infants, and there does not appear to be a dose effect from 80–120 mg/kg/d of DHA. A particular concern with this study and many others is that the administration of fat is too late after birth. If DHA levels are decreased by the end of the first week of birth, then supplementation given at the end of the week would have to not only meet the needs but overcome the deficit that has already occurred.

A number of studies have looked at the effect of DHA enteral supplementation on neurodevelopment outcomes.^110–123 A 2011 Cochrane database meta-analysis failed to show a positive effect from DHA at 18 and 24 months on cognitive function as compared with standard enteral formulas.¹²⁴ This meta-analysis had limited power, with only 2 studies in the 18-month analysis and 1 study in the 24-month analysis. A larger meta-analysis with mixed analysis for term and preterm infants also did not show a difference in MDI or psychomotor development index scores with DHA supplementation.¹²⁵ When the data were stratified to only preterm infants, there was a strong but nonsignificant trend (P = .06) for a positive effect for DHA supplementation. The DINO study has been one of the largest (n = 657) multicenter RCTs to examine neurodevelopment with enteral DHA supplementation in VLBW infants.¹¹⁹ Preterm infants overall receiving the high DHA supplement (1% total fatty acids) compared with standard DHA (0.3% total fatty acids) had a significant reduction in MDI scores <70 (P = .03). Stratified by body weight, infants <1250 g saw a significant benefit in mild metal delay (MDI <85, P = .02) but no benefit in significant mental delay incidence (MDI <70, P = .17). When stratified by sex, girls benefited from DHA supplementation across all MDI scores (MDI <85, P = .01; MDI <70, P = .02), whereas boys did not benefit from DHA treatment in either MDI group (MDI <85, P = .94; MDI <70, P = .47). A follow-up of this study was conducted in 2015.¹¹⁵ No differences were seen in IQ at 7 years of age between children who had received the DHA supplement or no DHA as preterm infants. Stratification by birth weight and sex also failed to show significant differences.

SOFE administration does not effectively maintain DHA levels. ELBW infants who receive SOFE for >28 days can have a 50% reduction of red blood cell DHA concentrations from birth.¹²⁶ This reduction can be partially rescued with the administration of enteral formulation (breast milk or formula) prior to reaching 28 days. Administration of MOFE does not prevent a progressive drop in DHA levels, independent of the time at which the PN feeds have been started.²⁸ However, it does appear that MOFE can prevent as large an initial drop in DHA levels if initiated early, as compared with SOFE feedings. This drop in DHA has been seen in IVFEs supplemented with fish oil as well.^83,127 It may be necessary to consider more aggressive PN feeding strategies within the first day of birth to prevent the initial drop in DHA levels. Once the drop occurs, current PN DHA feeding strategies are inadequate.

No published studies have looked at the effect of new-generation IVFEs containing DHA on neurodevelopmental outcomes. As mentioned, the DHA concentration in preterm infants is sustained at higher levels in fish oil–supplemented PN compared with soy-based emulsions. Yet, these levels fall below DHA concentrations in term infants. It is not clear to what extent a marginal increase in DHA levels would have on overall neurodevelopment. On the basis of studies involving enteral DHA supplementation, it is possible that PN results will show minimal benefit unless DHA levels can be brought significantly higher.

Failure to achieve appropriate growth affects the neurodevelopment of preterm infants. A large cohort study by Ehrenkranz et al of 695 ELBW infants examined the effect of growth on BSID-II MDI scores at 18–22 months of age.⁹² The infants were stratified into quartiles for mean weight gain, and the lowest quartile had the lowest mean MDI scores and the largest number of scores that fell <70. The study showed an even stronger association with smaller head circumference growth and MDI scores <70. Head growth and impaired developmental outcomes have been observed in VLBW infants as well.¹²⁸ Consistent with this result, a study by Stephens et al regarding the MDI score of ELBW infants at 18 months found that infants with higher energy or protein intake per kilogram of body weight in the first week of life had higher MDI scores.⁹⁴ There have not been many positive results to suggest that administration of mixed-fat emulsion is of any benefit to improving weight gain when compared with SOFE alone.^{28,31,77,129–131} One study by Vlaardingerbroek et al did find that the administration of MOFE led to greater weight gain in infants at the time of discharge.²⁹ However, in this study, most infants were on full enteral feeds by day 14, and their dates until discharge were 87 and 95 days for MOFE and SOFE groups, respectively. These results suggest that the infants receiving the MOFE were healthier overall as compared with the SOFE infants; regardless, the weight gain did not occur during MOFE administration.

Conclusion

There are several morbidities currently associated with current-generation IVFE administration that could theoretically benefit from the modified compositions of new generation fat emulsions (Figure 1). However, the current evidence does not strongly support the use of new-generation IVFE with regard to decreasing these common morbidities associated with PN. Despite the current evidence, there is far too much variation among treatment protocols to dismiss the potential benefit of n-3 PUFA–enriched IVFEs. Variation in the start of fat treatments, the overall concentration of fat administered, and the rate at which fat is increased to full dose of 3 g/kg/d all affect study outcomes and greatly contribute to the large variation in study results. What can be determined from the majority of research is that meeting the n-3 PUFA needs for the preterm infant is still a large hurdle that needs to be overcome to optimize outcomes in this population. The implementation of standardized protocols across study groups and larger multicenter RCTs are necessary to reach concrete conclusions on the benefit of IVFEs containing fish oil. The use of preclinical models could also benefit our understanding of IVFE in this population.

Figure 1.

Risk factors from prematurity and current-generation fat emulsions and the potential benefits of new-generation fat emulsions in organ systems. DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid; LA, linoleic acid; LT, leukotriene; PG, prostaglandins; PPAR, peroxisome proliferator-activated receptor. Images © guniita (gastrointestinal tract), Eraxion (brain), tigatelu (liver), alila (lungs) / 123RF.com.

Understanding the impact of different parenteral fat emulsions on the growth, metabolic function, and developmental outcomes in neonates requires not only well-controlled clinical studies but preclinical studies in appropriate animal models. Various models have been used to study IVFE, including the mouse, guinea pig, and domestic pig, but only the domestic pig has been able to reproduce the condition of prematurity in the neonatal stage of postnatal life. Further studies in IVFE administration in these models, with a focus on the administration of DHA, could help facilitate a better understanding of the obligate needs of preterms and how best to meet these needs.