Biochemical Parameters in Extremely Preterm Infants Receiving Mixed Lipid Emulsions

Abstract

OBJECTIVE

A mixture of soybean, medium-chain triglycerides, olive, and fish oils (SMOF) contains higher α-tocopherol and n-3 polyunsaturated fatty acids and lower phytosterol content compared with conventional soybean oil lipid emulsions (SOLE). We sought to characterize plasma total fatty acid profiles (FAPs) and assess the tolerability of long-term SMOF therapy in extremely preterm infants.

METHODS

We retrospectively evaluated infants born <28 weeks gestational age who received at least 30 consecutive days of SMOF between July 2016 and June 2019. We evaluated monthly FAPs and biochemical tolerance to SMOF using direct bilirubin (DB) and triglyceride (TG) levels. Growth parameters were evaluated longitudinally until discharge.

RESULTS

Sixteen patients with median gestational age 24 weeks (IQR, 23–25 weeks) received SMOF for median 76 days (IQR, 52–130 days). Fourteen patients had necrotizing enterocolitis (NEC) requiring surgical intervention and 15 patients received SOLE for median 19 days (IQR, 14–26 days) prior to switching to SMOF. Median docosahexaenoic acid (DHA) and eicosapentaenoic acid (EPA) levels were elevated, whereas the remaining fatty acid levels fell within reported reference ranges. There were no incidents of essential fatty acid deficiency (triene to tetraene ratio >0.2) or hypertriglyceridemia (TG >200 mg/dL) with a general downtrend in DB after the first month on SMOF. All growth Z-scores declined throughout hospital stay.

CONCLUSIONS

Infants who received SMOF had a more pronounced elevation in DHA than EPA, of which the clinical significance remains unknown. Growth Z-scores declined with SMOF but were confounded by a high prevalence of surgically treated NEC.

Introduction

Parenteral nutrition (PN) sustains the surgical neonate's life, while carrying significant risks of morbidity with long-term use. Lipids are the most calorie-dense component but may cause the most harm. Since the introduction of PN more than 50 years ago, soybean oil lipid emulsion (SOLE) was the most widely used product in the United States for use in neonates until 2016, when mixed lipid emulsions, such as soybean, medium-chain triglycerides (MCTs), olive, and fish oils (SMOF; SMOFlipid, Fresenius Kabi, Toronto, ON, Canada), were introduced.^1,2 Lipid emulsions are an integral part of PN as a source of high-density non-protein calories and essential fatty acids, such as linoleic acid (LA) and α-linolenic acid (ALA).^3,4 Various manufacturers produce different lipid formulations, each derived from a variety or mixture of parent oils and comprising varying fatty acid contents. However, an intravenous lipid emulsion designed specifically for the unique needs of preterm infants has yet to be developed.

Current evidence suggests long-term therapy with SOLE may be harmful, contributing to intestinal failure–associated liver disease (IFALD) via induction of inflammatory pathways due to a high content of n-6 polyunsaturated fatty acids (PUFAs) and provision of phytosterols, a hepatotoxic plant sterol.^5–7 Prolonged phytosterol administration leads to a disruption in red blood cell membrane fluidity, causing cell lysis and leaked bilirubin.⁸ Additionally, phytosterols antagonize cholesterol 7-α-hydroxylase, an enzyme that converts cholesterol to bile acids, and hepatic farnesoid X receptor, an important modulator of bile acid homeostasis, thus contributing to the development of cholestasis.^5,6,9

SMOF contains 30% soybean oil, 30% MCT oil, 25% olive oil, and 15% fish oil, and it offers several theoretical advantages versus SOLE. Compared with SOLE, SMOF contains α-tocopherol, the most biologically active form of vitamin E, and a higher content of n-3 PUFAs, such as docosahexaenoic acid (DHA) and eicosapentaenoic acid (EPA). Higher DHA content may benefit preterm neonates who have a reduced capacity to synthesize DHA from its precursor, ALA.^7,10 The antioxidant, α-tocopherol, may play a role in the prevention of cholestasis, and a higher n-3 PUFA content may impart more anti-inflammatory properties.⁷ Additionally, SMOF contains less phytosterols and n-6 PUFAs, such as LA, when delivered at similar caloric rates.¹¹ Decreasing the amount of LA, an essential n-6 PUFA, may therefore reduce inflammatory mediators but also theoretically increase the risk for developing essential fatty acid deficiency (EFAD). Because of limited fat stores, preterm neonates are already at a higher risk of developing EFAD as early as 2 to 3 days after birth.^7,12 Because more mead acid is produced as a compensatory response to decreased LA and ALA concentrations, a ratio of triene (mead acid) to tetraene (arachidonic acid [ARA]), also called the Holman Index, is used as a measurement for biochemical EFAD.^13,14

SMOF offers several theoretical benefits and its use in preterm neonates has been investigated.^{10,11,15–20} However, the appropriate dose in this patient population has yet to be established. Over-dosing SMOF may lead to hypertriglyceridemia and IFALD, whereas under-dosing SMOF may lead to EFAD.^9,21 Other studies have reported the serum fatty acid profiles (FAPs) of preterm neonates receiving SMOF on a short-term basis for up to 2 weeks.^10,11,15,16 However, the effects of long-term SMOF administration on the FAP and the likelihood of developing EFAD in preterm infants remain unknown. Our study aims to describe essential FAPs of extremely preterm infants receiving long-term SMOF in addition to therapy tolerance and growth patterns throughout the neonatal intensive care unit (NICU) stay.

Materials and Methods

Study Design and Patient Population. This was a retrospective evaluation of extremely preterm infants who received at least 30 consecutive days of SMOF in the NICU between July 2016 and June 2019. Inclusion criteria were birth gestational age <28 weeks and postmenstrual age <45 weeks at the time of SMOF initiation. Patients with human immunodeficiency virus, hepatitis B or C, cystic fibrosis, liver injury attributed to cytomegalovirus infection, primary liver disease, and inborn metabolic disorders were excluded. The institution's investigational review board approved this study under the internal designation of “IRB18-00705” and waived the need for written informed consent.

Institutional Guideline. Because of a lack of robust evidence supporting the use of SMOF in preterm infants, SMOF use in our center's NICU was not implemented until Fall 2017. During the time of the study, the primary lipid was SOLE for at least 2 weeks, and the criteria for switching to SMOF included anticipated long-term PN beyond 3 weeks or if the patient developed a direct bilirubin (DB) level >2 mg/dL. SMOF was initiated at 2 g/kg/day and advanced by 0.5 to 1 g/kg/day daily increments as tolerated to a goal of 3 g/kg/day and max of 3.5 g/kg/day if needed for additional growth. There was no guideline for specific PN weans or discontinuation as enteral feeds were advanced, so daily titrations were left to the discretion of the practitioners based on patient-specific caloric goals and growth trends. However, practitioners generally began weaning the lipid dose by daily increments of 1 g/kg once the feeds were increased to at least 60 mL/kg/day and discontinued lipids once feeds reached 100 mL/kg/day. Fortification of feeds occurred at 80 to 100 mL/kg/day of enteral feeds, so discontinuation of PN therapy usually occurred within 3 days of discontinuing lipid therapy.

Objectives and Study Variables. The primary objective was to assess the occurrence of clinically significant EFAD, defined as a triene to tetraene ratio >0.2, and to characterize other findings of the FAP. Secondary objectives were to assess SMOF tolerance and growth Z-scores. Tolerance to SMOF was defined as DB <2 mg/dL and triglyceride (TG) <200 mg/dL. Patient demographics included biologic sex, ethnicity, gestational age, birth weight, birth length, birth head circumference, cesarean delivery, Apgar scores at 1 and 5 minutes, administration of surfactant, maternal hypertension, chorioamnionitis, preterm premature rupture of membranes, prenatal care, and surgical history. Study variables included FAP, DB, TG, and growth Z-scores at monthly intervals.

The FAPs were measured monthly following SMOF initiation until SMOF discontinuation. Although each FAP reports values for more than 20 parameters, we focused on ALA, DHA, EPA, LA, ARA, mead acid, and the triene to tetraene ratio because of their clinical relevance. The FAPs obtained at our institution were sent to Mayo Clinic laboratories to be analyzed by gas chromatography mass spectrometry stable isotope dilution analysis. The DB measurements included the level obtained 1 month prior to SMOF initiation, the baseline level at the time of SMOF initiation, and monthly levels thereafter until 2 months after SMOF discontinuation. If SMOF was initiated prior to 1 month of age, the first DB level was recorded. The TG levels were measured at baseline and monthly intervals during SMOF treatment until SMOF discontinuation. The SMOF dose associated with the monthly lab values was the dose used for most of the month prior to obtaining the labs. SMOF duration and associated lab values were based on consecutive days of SMOF therapy without an interruption in therapy for more than 1 day. Total PN duration was based on total days of PN therapy during the hospital stay, including resumption of PN therapy after previously weaning off. Z-scores for weight, length, and head circumference on the Fenton curve were trended monthly from SMOF initiation to discharge or transfer out of NICU. If patients transitioned from the Fenton to the World Health Organization curve at 50 weeks postmenstrual age prior to discharge or transfer out of the NICU, the last Fenton Z-score was used as the last measurement.

Statistics. Patient data were summarized using descriptive statistics. Categoric data were reported as percentages and continuous variables as medians with an interquartile range (IQR).

Results

Seventeen patients met inclusion criteria. One patient was excluded because of an inherited fructosemia diagnosis. Sixteen patients were evaluated with a median gestational age at birth of 24 weeks (IQR, 23–25 weeks), median birth weight of 680 g (IQR, 583–753 g), median birth length of 32 cm (IQR, 29–33 cm), and median birth head circumference of 22 cm (IQR, 21–23 cm; Table 1). Most of the patients were male (88%) and were born via cesarean delivery (88%). Fourteen patients (88%) had necrotizing enterocolitis (NEC) requiring surgical intervention. The median PN duration was 114 days (IQR, 89–175 days), and median length of hospital stay was 171 days (IQR, 152–250 days). Fourteen patients (88%) were on SOLE for a median of 19 days (IQR, 14–26 days) prior to switching to SMOF. The median SMOF duration was 76 days (IQR, 52–130 days), and only 1 patient remained on SMOF for more than 4 consecutive months. No patients were discharged home on PN therapy.

Table 1.

Patient Demographics

Parameter	Value
Male, n (%),	14 (88)
Female, n (%)	2 (12)
Race, n (%)
White	6 (37.5)
Black or African American	6 (37.5)
Other	4 (25)
Gestational age, median (IQR), wk	24 (23–25)
Birth weight, median (IQR)
Grams	680 (583–753)
Z-score	0.3 (−0.4 to 0.7)
Birth length, median (IQR)
Cm	32 (29–33)
Z-score	0.5 (−0.9 to 0.7)
Birth head circumference, median (IQR)
Centimeters	22 (21–23)
Z-score	0.05 (−0.3 to 0.4)
Cesarean delivery, n (%)	14 (88)
Apgar score at 1 min, median (IQR)	3 (1–5)
Apgar score at 5 min, median (IQR)	5 (3–6)
Received surfactant, n (%)	16 (100)
Maternal hypertension, n (%)	1 (6)
Chorioamnionitis, n (%)	2 (13)
Preterm premature rupture of membranes, n (%)	7 (44)
Received early prenatal care (starting in the first trimester), n (%)	15 (94)
Previous significant SOLE exposure prior to SMOF use, n (%)	14 (88)

SMOF, soybean, medium chain triglycerides, olive, and fish oils; SOLE, soybean oil lipid emulsion

There were no incidents of triene to tetraene ratios >0.2, and median mead acid concentrations fell within the reference range (Table 2) even with SMOF doses less than 3 g/kg (Figure 1). Median DHA and EPA levels of the first 4 FAPs were on average 1.5 and 4 times the upper limit of reference ranges, respectively, whereas median ALA, LA, and ARA levels fell within reference ranges for patients between 1 month and 1 year of age (Table 2). Median levels of ALA, DHA, EPA, and ARA remained stable over time, compared with LA levels, which decreased during SMOF therapy (Figure 2).

Table 2.

Serum Fatty Acid Concentrations

Fatty Acid	Reference Range, nmol/mL	Month 1 (n = 16)	Month 2 (n = 8)	Month 3 (n = 3)	Month 4 (n = 5)
ALA, median (IQR)	10–190	65 (48–72)	48 (40–72)	43 (33–54)	42 (32–42)
DHA, median (IQR)	10–220	321 (258–384)	306 (255–371)	369 (308–492)	354 (214–386)
EPA, median (IQR)	2–60	282 (172–304)	221 (191–270)	261 (203–283)	259 (131–322)
LA, median (IQR)	1000–3300	1446 (1190–1745)	1461 (1081–1660)	1341 (1175–1682)	1313 (1201–1324)
ARA, median (IQR)	110–1110	346 (266–404)	247 (209–378)	298 (281–322)	265 (259–342)
Mead acid, median (IQR)	3–24	19 (15–28)	18 (15–25)	19 (15–19)	12 (12–21)
Triene to tetraene ratio, median (IQR) [range]	0.013–0.05	0.053 (0.047–0.081) [0.035–0.126]	0.054 (0.045–0.063) [0.029–0.077]	0.046 (0.041–0.06) [0.04–0.112]	0.048 (0.045–0.072) [0.042–0.074]

ALA, α-linolenic acid; ARA, arachidonic acid; DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid; LA, linoleic acid

Figure 1.

Median lipid dose (g/kg/day) and associated median mead acid concentrations and median triene:tetraene ratios.

Figure 2.

Median serum fatty acid concentrations.

There were no incidents of hypertriglyceridemia throughout SMOF therapy (Figure 3). However, DB was elevated during the first month after SMOF initiation (median, 2.6 mg/dL [IQR, 1.1–4.2 mg/dL]) with a subsequent decrease thereafter (Figure 4). Z-scores for weight, length, and head circumference all declined throughout the NICU stay (Figure 5).

Figure 3.

Median (IQR) triglyceride concentrations during SMOF therapy.

Figure 4.

Median (IQR) direct bilirubin concentrations during SMOF therapy.

Figure 5.

Anthropometric Z-scores for weight, length, and head circumference at birth, and throughout SMOFlipid therapy until discharge.

Discussion

We found no essential fatty acid deficiencies in patients receiving SMOF, even at doses less than 3 g/kg. Additionally, patients in the cohort had increased levels of DHA and EPA. Despite patients tolerating SMOF at 3 g/kg without hypertriglyceridemia and decreasing direct bilirubin levels, overall growth trajectories exhibited a downward trend throughout the inpatient stay.

Our study characterized plasma total FAPs of extremely preterm infants receiving long-term SMOF therapy. There are no reported reference ranges of essential fatty acid levels in this specific patient population. We used fatty acid plasma ranges reported by Lagerstedt et al²² as a point of reference, which are the values adopted by Mayo Clinic laboratories. These data were compiled from healthy individuals without any metabolic or nutritional abnormalities, and infants in this study were most likely term. The reported clinical reference considers a triene to tetraene ratio elevated at values >0.05, which may not necessarily correlate with clinically significant elevations leading to biochemical or symptomatic EFAD. As previously described by Holman et al,¹³ we chose a triene to tetraene ratio of >0.2 to indicate clinically significant biochemical EFAD. Biochemical EFAD did not occur, which suggests that patients received adequate exogenous supplementation of LA and ALA. Median mead acid concentrations did not increase above reference ranges, even with SMOF doses less than 3 g/kg (Figure 1). However, it is important to note that there was a low number of patients receiving SMOF doses less than 3 g/kg, with 6 patients at the third month and 5 patients at the fourth month of SMOF. An increased mead acid concentration with an increased triene to tetraene ratio may indicate EFAD, because increased metabolism of oleic acid to mead acid occurs when there is insufficient LA and ALA.¹⁴ Although there was a general decline in median LA levels during the span of 4 months, none of the values fell below the reference range. Our findings did not demonstrate the development of EFAD in our study population because there were no clinically significant triene to tetraene ratio elevations and no concurrent mead acid elevations throughout each month of SMOF therapy. However, these results were confounded by a portion of patients who also received enteral nutrition that provided an additional source of PUFAs.

Infants in this study showed elevated DHA and EPA levels. Without supplementation of either, studies have shown rapidly declining DHA levels after birth and no change in EPA levels.^16,23 This finding is consistent with previously published studies showing higher DHA and EPA levels in patients receiving SMOF compared with SOLE.^10,11,24 Although elevated DHA and EPA levels may be attributed to SMOF therapy, these findings were confounded by a portion of the patients who were receiving concomitant enteral feeds that could have provided supplemental DHA and EPA.

The DHA and ARA status of preterm infants who miss most of the third trimester is of particular interest because DHA accretion occurs during that period and is essential for brain and visual development.^6,7 Preterm infants are likely to have a reduced capacity to convert sufficient DHA and ARA from their precursors ALA and LA, respectively, to meet estimated requirements and thus require exogenous supplementation.¹⁰ It is estimated that infants receive 5% to 10% weight/volume (w/v) of DHA and 16% to 21% (w/v) of ARA in utero during the last trimester, whereas SMOF only provides 1% to 3.5% (w/v) and 0.5% (w/v), respectively.^10,25 Despite lower DHA and ARA delivery with SMOF compared with in utero delivery, DHA serum levels exceeded reference ranges, whereas ARA serum levels fell within reference ranges. This highlights an incomplete understanding of DHA and ARA regulation in extremely preterm neonates.

This study did not directly assess the clinical impact of elevated DHA and EPA status. Many theorized benefits of DHA have been postulated which include improved brain accretion rates with subsequent positive impacts on neurodevelopment, growth, retinopathy of prematurity, and bronchopulmonary dysplasia due to its anti-inflammatory effects. However, very little in vivo data support these claims and should be interpreted with caution.^17,18,26,27 Further studies are needed to understand the regulation of DHA status and to evaluate its long-term effects. The elevated EPA levels in patients of this study may play a role in the initial decline in ARA levels over time because of an inhibition of ARA production by EPA. However, a decline in ARA has been demonstrated with both SMOF and SOLE use as well as in healthy infants within the first month of life, so the extent of the interaction remains unknown.^13,16,28 It is unclear whether the decline in ARA in our study influenced the negative growth trends, because ARA supplementation has been associated with growth and development. However, it is reassuring that the ARA levels still fell within the reported reference range despite the initial decline. Although other studies showed improved growth with mixed lipid emulsions or no difference in growth when compared to SOLE, our study showed a declining trend.^10,16,19,29 The negative Z-scores for weight, length, and head circumference in our study were likely related to the high prevalence of surgical NEC in the study population, which is a risk factor for poor growth.^30,31

Biochemical tolerance was measured by the incidence of hypertriglyceridemia and cholestasis. There were no detected periods of hypertriglyceridemia, likely because of the faster metabolism of MCT oil and increased lipoprotein lipase activity as the infant matures over time. SMOF also contains α-tocopherol to prevent lipid peroxidation of its PUFA contents and has demonstrated a potential benefit on IFALD in animal models.^6,18 Although the median DB was elevated the first month after initiation of SMOF, there was a subsequent decrease over time. Although most patients received SOLE prior to switching to SMOF, the first month's DB was likely more reflective of underlying medical conditions rather than a specific lipid product. Most patients had NEC requiring surgical intervention. Surgical NEC is commonly accompanied by sepsis, bacterial translocation, bacterial overgrowth, and intestinal failure, which are all factors that play a role in the development of cholestasis.^32,33 Additionally, cholestatic preterm infants have longer phytosterol half-lives, which may explain the delayed improvement in DB as the phytosterol concentration slowly decreased over time, in addition to stabilization of their medical conditions. Although SMOF does not seem to have a role in the prevention or treatment of cholestasis in preterm infants, it may aid the recovery process through decreased phytosterol and n-6 PUFA administration.^9,18,20

This observational study provided insight on the biochemical tolerance of SMOF in extremely preterm infants during the span of 4 months. However, there were a few limitations to the study. The small sample size limits the conclusions that can be drawn from this study. The study population was primarily composed of surgical NEC patients, so our findings may not apply to the more general extremely preterm population that we initially targeted. Additionally, some patients in our study received partial enteral nutrition at the time the FAPs were drawn, which may impact the levels and confound the results. This was a retrospective analysis without a comparator group. Although some assumptions can be made about fatty acid concentrations in SMOF compared with SOLE, there were no FAPs at baseline prior to SMOF or with SOLE therapy for comparison, because historically there was not a need for this. Additionally, currently available fatty acid reference ranges were not meant to be used in extremely preterm infants, and thus comparisons made in this study should be interpreted with caution. The absence of fields within our electronic medical record for documenting prospective and discrete short-term adverse effects associated with SMOF therapy limited the descriptive abilities of this retrospective study. Lastly, we did not evaluate long-term follow-up outcomes. Future studies are needed to assess the long-term impact of increased DHA and vitamin E status on neurodevelopment, growth, retinopathy of prematurity, bronchopulmonary dysplasia, and intraventricular hemorrhage.

Conclusion

Use of long-term SMOF in extremely preterm infants appears to be well tolerated based on the biochemical profiles of our cohort. SMOF delivers DHA to infants who would have otherwise missed in utero accretion and received suboptimal supplementation with prolonged feeding advancement. However, it is unknown whether postnatal elevations in DHA and EPA levels are associated with improved neonatal outcomes. Future studies are needed to assess the long-term impact of SMOF use in both surgical and non-surgical populations, as well as the resulting increase in DHA and EPA levels. Prospective randomized controlled trials are needed to compare FAPs, growth patterns, morbidities, and neurodevelopmental outcomes between patients receiving SMOF and SOLE to determine the optimal intravenous fat emulsion for neonates.

ABBREVIATIONS

ALA

α-linolenic acid

ARA

arachidonic acid

DB

direct bilirubin

DHA

docosahexaenoic acid

EFAD

essential fatty acid deficiency

EPA

eicosapentaenoic acid

FAP

fatty acid profile

IFALD

intestinal failure–associated liver disease

LA

linoleic acid

MCT

medium chain triglycerides

NEC

necrotizing enterocolitis

NICU

neonatal intensive care unit

PN

parenteral nutrition

PUFA

polyunsaturated fatty acids

SMOF

soybean oil, medium-chain triglycerides oil, olive oil, and fish oil

SOLE

soybean oil lipid emulsions

TG

triglyceride

w/v

weight/volume percent