Higher docosahexaenoic acid, lower arachidonic acid and reduced lipid tolerance with high doses of a lipid emulsion containing 15% fish oil: a randomized clinical trial.

Clinical Nutrition xxx (2014) 1e8

Contents lists available at ScienceDirect

Clinical Nutrition journal homepage: http://www.elsevier.com/locate/clnu

Randomized control trials

Higher Docosahexaenoic acid, lower Arachidonic acid and reduced lipid tolerance with high doses of a lipid emulsion containing 15% fish oil: A randomized clinical trialq Rita D’Ascenzo a, *, Sara Savini a, Chiara Biagetti a, Maria P. Bellagamba a, Paolo Marchionni a, Adriana Pompilio a, Paola E. Cogo b, Virgilio P. Carnielli a a Division of Neonatology, Department of Clinical Sciences, Polytechnic University of Marche and Azienda Ospedaliero-Universitaria Ospedali Riuniti, Ancona, Italy b DMCCP, Pediatric Hospital “Bambino Gesù”, Rome, Italy

a r t i c l e i n f o

s u m m a r y

Article history: Received 5 September 2013 Accepted 13 January 2014

Background & aims: Lipid emulsions containing fish oil, as source of long chain omega 3 fatty acids, have recently became available for parenteral nutrition in infants, but scanty data exist in extremely low birth weight preterms. The objective of this study was to compare plasma fatty acids and lipid tolerance in preterm infants receiving different doses of a 15% fish oil vs. a soybean oil based lipid emulsion. Methods: Preterm infants (birth weight 500e1249 g) were randomized to receive parenteral nutrition with MOSF (30% Medium-chain triglycerides, 25% Olive oil, 30% Soybean oil, 15% Fish oil) or S (S, 100% Soybean oil) both at two levels of fat intake: 2.5 or 3.5 g kg1 d1, named 2.5Fat and 3.5Fat respectively. Plasma lipid classes and their fatty acid composition were determined on postnatal day 7 and 14 by gas chromatography together with routine biochemistry. Results: We studied 80 infants. MOSF infants had significantly higher plasma phospholipid Docosahexaenoic acid and Eicosapentaenoic and lower Arachidonic acid. Plasma phospholipids, triglycerides and free cholesterol were all significantly higher in the MOSF-3.5Fat group, while cholesterol esters were lower with MOSF than with S. The area under the curve of total bilirubin was significantly lower with MOSF than with S. Conclusions: The use of a lipid emulsion with 15% FO resulted in marked changes of plasma long-chain fatty acids. Whether the benefits of increasing Docosahexaenoic acid outweigh the potential negative effect of reduced Arachidonic acid should be further studied. MOSF patients exhibited reduced lipid tolerance at 3.5 g kg1 d1 fat intake. The trial was conducted between January 2008 and December 2012 so we had not registered it in a public trials registry as it is now required for trials that started after July 2008. Ó 2014 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.

Keywords: Fish oil Parenteral nutrition Preterm infants Docosahexaenoic acid Arachidonic acid Lipid tolerance

1. Introduction Abbreviations: ARA, Arachidonic acid; AUC, area under the curve; BPD, bronchopulmonary dysplasia; CE, cholesterol esters; DHA, Docosahexaenoic acid; ELBW, extremely low birth weight infants; EPA, Eicosapentaenoic acid; FA, fatty acid; FC, free cholesterol; FO, fish oil; IV-FAT, intravenous fat intake; LLA, linoleic acid; LNA, alpha-linolenic acid; LE, lipid emulsion; LC-PUFA, long chain PUFA; NICU, neonatal intensive care unit; PL, phospholipid; PN, parenteral nutrition; MCT, medium-chain triglycerides; PN, parenteral nutrition; PMA, post-menstrual age; RCT, randomized clinical trial; sTG, serum triglyceride; TG, triglyceride. q Data of this manuscript were presented at the 54th Annual Meeting of the European Society for Paediatric Research e ESPR in Porto, Portugal, October 11e14, 2013. * Corresponding author. Neonatologia, Dipartimento di Scienze Cliniche, Azienda Ospedaliero-Universitaria Ospedali Riuniti di Ancona, via Corridoni 11, 62123 Ancona, Italy. Tel.: þ39 0715962014; fax: þ39 0715962044. E-mail address: [email protected] (R. D’Ascenzo).

Preterm infants, especially the extremely low birth weight (ELBW) ones, are usually on parenteral nutrition (PN) as part of their routine care during the first weeks of life. However most of the commercially available lipid emulsions have a very different fatty acid (FA) profile than human milk, which is thought to have the preferred fat and FA composition.1 Indeed, LE contain a much larger proportion of polyunsatured fatty acids, namely Linoleic acid (LLA) and alpha-Linolenic acid (LNA), and are devoid of 20e22 carbon atoms long chain-PUFA (LC-PUFA). The surplus of n-6 FA could result in excess of pro-inflammatory mediators and enhanced lipid peroxidation.2,3 It could also inhibit the endogenous biosynthesis of LCPUFA, including Docosahexaenoic acid (DHA),4,5 which is known to

0261-5614/$ e see front matter Ó 2014 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved. http://dx.doi.org/10.1016/j.clnu.2014.01.009

Please cite this article in press as: D’Ascenzo R, et al., Higher Docosahexaenoic acid, lower Arachidonic acid and reduced lipid tolerance with high doses of a lipid emulsion containing 15% fish oil: A randomized clinical trial, Clinical Nutrition (2014), http://dx.doi.org/10.1016/ j.clnu.2014.01.009

2

R. D’Ascenzo et al. / Clinical Nutrition xxx (2014) 1e8

be limited in the preterm infant.6 As DHA accretion in the brain and retina occurs primarily during the last trimester of pregnancy7 and as DHA is preferentially transported by the placenta from the pregnant woman to the fetus, infants born before term are deprived of the intrauterine supply of DHA. Reduced DHA status in newborn preterm infants has been related to impaired visual function.8 LC-PUFA can be provided to newborn infants on PN by using fish oil (FO) LE as FO contains DHA, Eicosapentaenoic acid (EPA) and Arachidonic acid (ARA). There is however limited experience with the use of FO-LE in the preterm infants. To the best of our knowledge lipid tolerance and fatty acid profile have been tested in only three randomized clinical trials (RCTs) Two commercially available preparations containing 10% or 15% FO were studied, and these have been reported to be well tolerated. Two of these studies described a better DHA status with FO-LE than with conventional products.9,10One study, but not others,10,11 reported a significantly lower ARA concentration with FO-LE. If this finding was to be confirmed, it would rise the question whether or not a depressed ARA could be associated with poor growth.12 If FO-LE could exert “less of negative effect” on liver function than conventional LE in uncomplicated non-surgical infants is also an open issue. Tomsits et al. found a significantly lower plasma gamma-glutamyl transpeptidases and Rayyan et al. reported lower total bilirubin in preterms receiving FO-LE vs. controls.10,11 In a recent retrospective study on preterm infants on routine PN with 5 different LE conducted by our group, there were not found any differences in liver function tests and in conjugated bilirubin at 6 weeks of age between infants who received FO-LE (n ¼ 64) and infants who received conventional LE (n ¼ 84).13 The main aim of this trial was to compare plasma phospholipids FA composition and plasma lipid classes of preterm infants randomized to receive a 15% FO-LE or conventional LE with 100% soybean oil, both at a standard (2.5 g kg1 d1) or at a high fat intake (3.5 g kg1 d1). 2. Materials and methods 2.1. Study protocol and participants In this parallel-group RCT premature neonates were recruited from the Neonatal Intensive Care Unit (NICU) at the Salesi Children’s Hospital, between January 2008 and December 2012. Inclusion criteria were birth weight between 500 and 1249 g and being inborn. We excluded infants with severe malformations, inborn errors of metabolism, severe congenital sepsis or lack of parental consent. The local ethical committee, in accordance with the principles of the Helsinki Declaration as revised in Edinburgh 2000, reviewed and approved the project, and written informed parental consent was obtained before enrollment. 2.2. Randomization and treatment Infants were randomized into 4 study groups, in a 1:1:1:1 ratio, to receive PN with MOSF (SMOFlipidÒ, Fresenius Kabi, 30% soybean oil, 30% medium-chain triglycerides or MCT, 25% olive oil and 15% fish oil) or S (IntralipidÒ, Fresenius Kabi, 100% soybean oil) at 2.5 g kg1 d1 or 3.5 g kg1 d1 (MOSF-2.5Fat: MOSF-3.5Fat: S-2.5Fat: S-3.5Fat). At birth, the caring neonatologist randomized the study infants by a simple randomization method (sealed envelope system). The PN bags containing the study LE were of the same size and of identical appearance. Both the caregivers involved with data collection and the laboratory personnel were blind to group assignment. All infants started PN within the first hour of life. Lipids were infused at 1.0,1.5, 2.0, 2.5 g kg1 d1 from postnatal day 0 to day

4 and kept at 2.5 g kg1 d1 until day 7 in the 2.5Fat groups and at 1.0, 1.5, 2.0, 2.5, and 3.0 g kg1 d1 from day 0 to day 4 and kept at 3.5 g kg1 d1 until day 7 in the 3.5Fat groups. From the first to the 7th day of life glucose was increased respectively from 6 to 12 g kg1 d1 and from 8 to 14 g kg1 d1 and amino acids from 1.0 to 2.5 g kg1 d1. After day 7, PN was gradually reduced, and was stopped on the 18th day of life, while oral feeding was progressively increased. Infants from day 0 to day 7 were on minimal enteral feeding with human milk, when available, or infant milk formula at a maximum intake of 8 ml kg1 d1, from day 1e4 and 16 ml kg1 d1 from days 5e8. According to local NICU guidelines, parenteral lipid intake was temporally lowered by 1 g kg1 d1 if serum triglyceride (sTG) concentrations were between 250 and 350 mg/dL and by 2 g kg1 d1 if they were between 350 and 450 mg/dL. If sTG were above 450 mg/dL lipid intake was stopped for at least 24 h and resumed when sTG were below 250 mg/Dl. Lipid intake was restarted at 50% of the intravenous fat intake before the hyper-triglycerdemia had occurred. 2.3. Primary outcome The primary outcome variable was plasma phospholipid (PL) DHA (mol%) measured on postnatal day 7. 2.4. Secondary outcomes Secondary outcomes were plasma PL FA (Oleic acid, LLA, LNA, ARA, EPA, mol%) and plasma lipid classes concentration (PL, triglycerides (TG), free cholesterol (FC) and cholesterol esters (CE), mg/dL) measured on postnatal day 7 and 14. The lipid classes plasma concentration were assumed to be indicators of lipid tolerance in the way that the lower the plasma lipid concentration, the better the tolerance. Plasma bilirubin was also determined at least once daily as part of routine care. The area under the curve (AUC) of total bilirubin and other kinetic parameters (maximum value, rate up, rate down, area up to the limit, and time up to the limit) were calculated using standard calculations. Moreover, all patients had routine biochemistry (sTG, blood urea, and creatinine) on day 3, 5 and 7 or more frequently when necessary. Acidebase balance, glucose, total bilirubin and electrolytes were obtained at least once a day, or more frequently as clinically indicated. Complete blood counts were also obtained at least once weekly during the first two weeks. Conjugated bilirubin was measured in all study patients during the 6th week of life. Cholestasis was defined as conjugated bilirubin greater than 2.0 mg/dL. We recorded daily weight, weekly head circumference and recumbent length. Weight gain (g kg1 d1) from regained birth weight to 36 weeks postmenstrual age (PMA) was calculated for all study patients. Clinical outcome parameters were prospectively defined and collected as part of the routine care of all admitted infants. 2.5. Analytical methods for measurements of plasma FA, total plasma lipids and LE composition EDTA-blood (0.5 ml) was collected on postnatal days 7 and 14. Plasma lipid classes and their FA composition analysis were carried out as previously described.14 Briefly plasma lipids were extracted from plasma after the addition of appropriate internal standards for each lipid class, according to Folch et al.15 Lipid classes were isolated by thin-layer chromatography, plasma phospholipid and triglyceride fatty acids were transesterified, and the separation and identification of FA methyl esters were performed by capillary gas-chromatography.14 Data for plasma FA from C8 to C24 carbon atom were calculated in mole percent and in absolute plasma concentrations (mg/dL).


R. D’Ascenzo et al. / Clinical Nutrition xxx (2014) 1e8 Table 1 Composition of the two lipid emulsions used in the study.a Composition (in 1000 ml)

MOSF

S

Soybean oil (g) Medium-chain triglycerides (g) Olive oil (g) Fractionated fish oil (g) Vitamin E, a-tocopherol (mg) Egg yolk phospholipids (g) Glycerol (g) Fatty acid composition (% wt/wt)b Caprilyc acid Capric acid Palmitic acid Stearic acid Oleic acid Linoleic acid Alfa-Linolenic acid Arachidonic acid Eicosapentaenoic acid Docosahexaenoic acid

60 60 50 30 200 12 25

200 e e e 38 12 23

20.5 13.0 9.0 2.8 25.2 17.8 2.0 0.4 2.9 2.0

e e 11.8 3.8 22.6 52.6 5.6 0.3 0.0 0.2

a MOSF (SMOFÒ; Fresenius Kabi): 30% SO, 30% MCT, 25% olive oil and 15% FO. FO, fish oil; MCT, medium-chain triglyceride; SO, soybean oil; S (IntralipidÒ; Fresenius Kabi): 100% SO. b Analyzed by gas chromatographyemass spectrometry in our laboratory.

The fat and the FA composition of the study LE were also measured at our laboratory by the same gas-chromatography method as described above (Table 1). 2.6. Data collection and monitoring Clinical laboratory assessments, such as hematological parameters and total bilirubin were obtained at fixed time intervals according the NICU policy using micro-methods (Radiometer

3

Analytical). Total blood bilirubin AUC and standard kinetic parameters were determined through Matlab R2012a, Natick Massachusetts, using a script suitable for this study. AUC was computed as an approximation of the integral of data via the trapezoidal rule.16 Total and conjugated bilirubin and other liver function tests were measured using a Spectrophotometric Chemistry Analyzer (ADVIA 1200 Siemens), by the Routine Chemistry Laboratory. Body weight was measured daily, according to a standard procedure of the NICU, using an electronic scale with a precision: 5 g. Head circumference and length (crown-heel) were measured at birth and weekly thereafter, using a non-stretchable tape and a length board respectively. Individual SD scores were computed using Italian reference data with dedicated proprietary software (NeotoolsÒ, Interactive.com srl, Milano, Italy). Clinical outcome parameters were available for all study patients according to predefined criteria. Bronchopulmonary dysplasia (BPD) was defined by the physiological criteria of Walsh et al.17; neonatal sepsis as positive blood culture or as clinical syndrome with systemic signs and symptoms of infection and abnormalities of laboratory findings18; necrotizing enterocolitis (NEC) was defined as Bell Stage 2 or 3.19 Other complications of prematurity were all classified according to International and/or National definitions.20

2.7. Statistical analysis Sample size was calculated on the primary outcome: phospholipid DHA. Sample size of 16 neonates per group gave 80% power to detect a difference of 1 SD of phospholipid DHA concentration between the study groups with a significance level of 0.05. Data are given as group means and SD for clinical variables and as group means and SD or median and interquartile range (IQR) for

Screened for eligibility (n= 92)

Excluded (n= 12) No informed consent (n=4) Missed opportunity for obtaining consent < 1 h after birth (n=2) Severe congenital malformations (n=6)

Randomized (n= 80)

Allocated MOSF-2.5F (n=21)

Allocated MOSF-3.5F (n=18)

Discontinued intervention: Death (n=1) Missed blood sampling (n=4)


Discontinued intervention: Missed blood sampling (n=5)


Analysed for fatty acid and plasma lipids (n=16)




Allocated S-2.5F (n=22)

Allocated S-3.5F (n=19)

Fig. 1. Flow diagram showing the progress of patients through the trial for each of the 4 study groups: MOSF-2.5Fat, MOSF-3.5Fat, S-2.5Fat and S-3.5Fat.


4


Table 2 Demographics.a

Male/female (n) Gestational age (d) Birth weight (kg) Z-score Birth length (cm) Z-score Birth head circumference (cm) Z-score Apgar score at 1 min Apgar score at 5 min Prenatal steroids (%) Age at 1st blood sampling (d) Age at 2nd blood sampling (d) a b

MOSF-2.5Fat (n ¼ 21)


S-2.5Fat (n ¼ 22)

S-3.5Fat (n ¼ 19)

pb

14/17 197 14 909 189 0.7 1.2 34.5 3.2 0.7 1.1 24.9 1.9 0.4 1.2 7 (5e9) 8 (7e9) 77 7.1 0.8 14.5 0.9

10/8 188 19 888 245 0.2 0.8 34.6 4.2 0.1 0.6 24.1 2.7 0.0 0.9 6 (5e8) 8 (7e8) 90 7.1 0.9 13.8 1.2

15/17 198 14 946 197 0.6 0.9 35.3 2.6 0.5 0.9 25.1 1.7 0.4 0.9 7 (6e9) 8 (7e8) 92 6.8 0.8 13.8 1.1

8/11 196 17 936 225 0.5 1.1 35.3 2.9 0.4 1.1 25.3 1.9 0.1 1.0 7 (5e8) 8 (7e9) 83 6.8 0.9 13.9 1.1

0.59 0.28 0.81 0.56 0.79 0.26 0.44 0.62 0.71 0.98 0.44 0.78 0.81

Data are expressed as Mean SD or Median (IQR). By One way Anova or Kruskall Wallis test as appropriate, p < 0.0.

similar clinical characteristics than those who had complete lipid analyses (not shown).

analytical variables. Data were compared between all four groups and between MOSF and S with independent t-test, two way ANOVA, Mann Whitney test, Kruskall Wallis, as appropriate. Prevalence of the major complications of prematurity was expressed as percentage and analyzed by chi-square (c2) test. Significance was set at 0.05. All statistical analyses were performed using the statistical package SPSS (version 15.0; SPSS Inc, Chicago, IL).

On day 7 (and 14) mol% PL DHA was significantly higher in the two groups assigned to MOSF compared with the S groups (Table 4).

3. Results

3.2. Secondary outcomes

92 preterm infants were screened; 12 were excluded because they did not meet inclusion criteria or parental consent was not obtained. Eighty infants were randomized to the 4 study groups (21 MOSF-2.5Fat: 18 MOSF-3.5Fat: 22 S-2.5Fat:19 S-2.5Fat, Fig. 1 diagram flow). Clinical characteristics of the study infants were not different between groups (Table 2). Total mean macronutrients and energy intakes during the first week of life are reported in Table 3. The two groups assigned to MOSF received significantly higher daily parenteral amounts of ARA, DHA and EPA compared with the S groups (Table 3). Oral intake of human milk or infant milk formula in the first week of life was not different between groups and never exceeded 15% of the total energy intake. During the second and third week of life, oral and parenteral intakes were also not different between groups (data not shown). Plasma lipid classes and their and FA composition could be reliably measured in 62 infants. Lipid analyses could not be obtained in 18 patients because of difficulties in drawing blood samples (n ¼ 9) or because of sample loss during laboratory analysis (n ¼ 5). Four patients (3 in the MOSF groups and 1 in S) died before day 7 of life because of severe NEC, acute pulmonary hypertension, sepsis and severe IVH. Infants in whom blood sampling was not possible or blood samples were lost during processing had

On day 7 and 14 PL Oleic and EPA were higher in MOSF (2.5Fat and 3.5Fat) than S groups, while LLA and ARA were lower as result of a significant lipid type effect at two-way ANOVA analysis (Table 4). Plasma concentration of lipid classes is shown in Fig. 2. PL were significantly higher in MOSF-3.5Fat group compared with MOSF2.5Fat and both S-3.5Fat and S-2.5Fat groups on day 7, while on day 14 they were higher in both MOSF-2.5Fat and MOSF-3.5Fat (data not shown). Plasma TG of MOSF-3.5Fat patients were significantly higher than MOSF-2.5Fat and S-2.5Fat and tended to be higher when compared with the S-3.5Fat group on day 7. CE were significantly lower in the MOSF-2.5Fat infants compared with both S-2.5Fat and S-3.5Fat ones. MOSF-3.5Fat had significantly lower CE compared with S-2.5Fat, but no differences were found between MOSF groups, neither between 3.5Fat groups (Fig. 2). FC was significantly higher in MOSF-3.5Fat than in the other groups. The FC/CE weight ratio was: 0.43 0.10 in MOSF-2.5Fat, 0.47 0.11 in MOSF-3.5Fat, 0.29 0.10 in S-2.5Fat and 0.31 0.09 in S-3.5Fat (p < 0.01, by one way ANOVA). MOSF patients had significantly higher FC/CE ratio than S by Bonferroni analysis (p < 0.01). On day 7 sTG, as determined by the Routine Biochemistry Laboratory, were significantly higher in MOSF-3.5Fat (Table 5) and

3.1. Primary outcome

Table 3 Mean nutrient and daily energy intake during the first week of parenteral nutrition. MOSF-2.5Fat (n ¼ 21) 1

1

IV Glucose intake (g kg d ) IV Proteins intake (g kg1 d1) IV FAT intake (g kg1 d1) IV ARA IV DHA IV EPA Parenteral non protein energy intake (Kcal kg1 d 1) Enteral non protein energy intake (Kcal kg1 d 1)

9.0 2.3 1.9 7.1 37.3 54.5 52.3 8.7

0.8 0.4 0.2 1.0 4.8 8.1 5.6 3.1

MOSF-3.5Fat (n ¼ 18) 10.6 2.7 2.1 8.2 42.7 62.5 58.6 7.6

1.3 0.2 0.5 1.7 8.9 13.1 7.8 3.5

S-2.5Fat (n ¼ 22)

S-3.5Fat (n ¼ 19)

9.3 0.7 2.5 0.3 2.0 0.2 4.9 0.5 1.2 0.12 2.9 0.3 53. 4.4 7.9 2.0

11.1 2.7 2.3 5.7 1.4 3.2 62.9 8.4

1.0 0.1 0.2 0.6 0.13 0.8 5.2 1.1

pLE

pIN

0.14 0.15 0.06 0.00 0.00 0.00 0.07 0.98

0.00 0.00 0.01 0.00 0.11 0.10 0.00 0.52

Data are expressed as Mean SD, mol%. IV:intravenous; ARA:Arachidonic Acid, DHA: Docosahexaenoic acid; EPA, Eicosapentaenoic acid. p values (two-way ANOVA, significant p < 0.05) refer to lipid emulsion comparison (LE); fat intake comparison (IN). The interaction was not significant for all the variables in the table.



5

Table 4 Phospholipid fatty acids profile on day 7 and day 14. MOSF-2.5Fat (n ¼ 16) Day 7 C18:1n-9 C18:2n-6 C18:3n-3 C20:4n-6 C20:5n-3 C22:6n-3 Day 14 C18:1n-9 C18:2n-6 C18:3n-3 C20:4n-6 C20:5n-3 C22:6n-3


S-2.5Fat (n ¼ 17)

S-3.5Fat (n ¼ 15)

pLE

pIN

15.82 15.67 0.13 9.71 1.58 3.06

1.45a 2.15a 0.05a 1.12a 0.50a 0.44a

16.49 15.69 0.13 9.41 1.67 3.18

1.5a,b 1.71a 0.05a 1.06a 0.27a 0.51a

14.42 18.78 0.17 11.18 0.27 2.19

1.76a,g 2.38b 0.08a 1.51b 0.13b 0.42b

13.14 19.99 0.21 10.71 0.21 2.14

0.89d 1.15b 0.07b 0.72b 0.06b 0.20b

0.00 0.00 0.00 0.00 0.00 0.00

0.09 0.09 0.08 0.55 0.79 0.44

13.57 16.47 0.14 8.63 1.31 2.74

2.57a 2.07a 0.08a 1.81a 0.37a 0.66a

14.25 16.54 0.11 8.96 1.46 2.85

1.43a 1.89a 0.05a 1.42a 0.25a 0.65a

12.51 18.63 0.12 9.81 0.51 2.19

1.64a 1.88a 0.08a 2.56a 0.46b 0.58a

12.26 19.99 0.15 10.77 0.39 1.99

1.81a 1.7a 0.04a 1.20a 0.10b 0.48a

0.00 0.03 0.63 0.02 0.00 0.00

0.58 0.52 0.98 0.24 0.86 0.94

Data are expressed as Mean SD, mol%. p values (two-way ANOVA, significant p < 0.05) refer to lipid emulsion comparison (LE); fat intake comparison (IN). The interaction was not significant for all the variables in the table.

hypertriglyceridemia (sTG >250 mg/dL) was more common in MOSF-3.5Fat group then in the other study groups (18% in MOSF3.5Fat vs. 9%, 9% and 3% in MOSF-2.5Fat, S-2.5Fat and S-3.5Fat respectively, p ¼ 0.214 Pearson chi square test between groups). There were no significant differences between groups in the other biochemical and hematological parameters on day 7 (Table 5). The total bilirubin AUC of the MOSF infants was significantly lower than that of the S group (43.3 6.5 vs. 46.5 7.8, p ¼ 0.035 by Mann Whitney test for two independent samples). There was no difference between groups for total and conjugated bilirubin and

for the others liver tests obtained in at 6 weeks of life (Table 6). Anthropometry at birth was similar between groups (Table 2). We found a significantly greater postnatal weight loss (14.3 5.8% vs. 11.1 5.7%, MOSF vs. S, p ¼ 0.015) and longer time from birth to the day of the regained birth weight (13.4 5.6 d vs. 10.5 5.1 d, MOSF vs. S, p ¼ 0.021, by independent t-test) in MOSF groups than in S. Weight gain from regained birth weight to 36 weeks PMA was rather similar between MOSF and S patients (17.1 2.6 g kg1 d1vs. 16.6 2.0 g kg1 d1, MOSF vs. S, p ¼ 0.72, by independent t-test).

Fig. 2. Total plasma lipid classes on day 7 (mg/dL). MOSF-2.5Fat (gray-bar, thin border), MOSF-3.5Fat (grey bar, thick border), S-2.5Fat (white bar, thin border), S-3.5Fat (white bar, thick border). The box plot displays median, IQR, maximum and minimum. The p value indicates comparison between all groups by Kruskall Wallis test. The graphs indicate the significant difference between groups by Mann Whitney test, *p < 0.05.


6


Table 5 Biochemical and hematological parameters on day 7.a MOSF-2.5Fat (n ¼ 20)


RBC, 1012/L

S-2.5Fat (n ¼ 22)

S-3.5Fat (n ¼ 18)

pb

2.97 0.82

3.49 0.61

2.87 0.53

0.16

12.5 3.9

15.2 3.7

12.1 1.93

0.28

30 9

35 9

26 11

0.25

20.7 7.9

13.1 7.5

11.4 3.5

0.17

245 150

215 60

222 121

0.89

7.27 0.03

7.29 0.04

7.31 0.05

0.12

48 9

42 7

43 6

0.19

21 3

20 2

21 2

0.48

5 2

6 2

5 1

0.62

136 4

138 5

137 4

0.88

3.42 0.82 Hemoglobin, g/dL 13.3 3.5 Hematocrit, % 32 9 9

WBCs, 10 /L Platelets, 109/L

17.7 11.6 208 82

pH 7.29 0.04 pCO2, mmHg HCO 3 , mmol/L

44 7 21 2

Base excess, mmol/L 5 2 Sodium, mmol/L 137 4

4.9 0.6

Potassium, mmol/L

4.9 0.4

4.7 0.3

0.62

5.0 0.5 107 56

Clorum, mmol/L

111 5

108 2

0.09

109 4 Calcium, mg/L

5.5 0.2

5.5 0.3

5.4 0.2

0.11

126 24

103 30

103 28

0.16

231b 59

119 a,g 87

166 a,b,g 47

0.01

67 38

59 43

53 12

0.97

0.79 0.25

0.77 0.21

0.64 0.22

0.29

5.9 1.5

6.9 1.7

6.5 1.7

0.41

42.5 6.8

47.1 6.4

45.6 6.4

0.17

5.3 0.4 Glucose, mg/dL Serum Triglycerides, mg/dLc

113 27 132a 50

Urea, mg/dL 61 36 Creatinine, mg/dL 0.86 0.30 Total bilirubin, mg/dL Total bilirubin AUCc

6.3 1.6 45.6 6.4

a b c

Data are expressed as Mean SD or Median (IQR). Values with different superscript letters are significantly different, p < 0.05 (Mann Whitney test). By one-way ANOVA. Kruskall Wallis test.

The incidence of late onset sepsis was significantly different among the four study groups, with a tendency to be higher in 3.5Fat groups than in the others (Table 6). Death and other clinical outcomes were not different (Table 6). 4. Discussion This study demonstrated that the use of MOSF was associated with better DHA status, and lower plasma phospholipid ARA. We also found that MOSF was less tolerated than S at the 3.5 g kg1 d1 fat intake and that the infants receiving MOSF had lower total bilirubin than those with S. Higher plasma DHA in children and infants on PN with FO-LE in comparison with conventional LE has been reported in most9,10,21 but not all pediatric RCT.11 These studies differed in study design, in IV-FAT intake, in postnatal age at PN start and in lipid assays (plasma vs. RBC FA). We studied a well-controlled homogeneous group of very small preterms who all received a full PN with LE from the first hour of life, with a IV-FAT intake up to 3.5 g kg1 d1. With IV-FAT at 2.5 g kg1 d1 and 3.5 g kg1 d1 the infants in MOSF groups received a mean daily DHA intake up to 46 mg kg1 d1 and up to 63 mg kg1 d1, respectively, which is markedly higher than the DHA intake reported in preterm infants fed human milk (w 19 mg kg1 d1)14 and also similar or even higher to the fetal DHA accretion rates reported by Clandinin et al.22

Plasma DHA mol% values found in the MOSF patients were slightly higher than those found in preterm infants fed their mother’s milk.14 The clinical relevance of these findings remains to be studied and the neurodevelopmental follow up of all our study infants is ongoing and will be completed in 2 years. The EPA intake of our infants was also markedly higher than the estimated intake in preterm infants fed human milk. As a result, we observed a marked elevation of plasma EPA and this finding needs further research. Whether or not the elevated EPA, in association with reduced ARA, results in a reduction of inflammation in small preterm infants is unknown. If this were to be the case, it would be extremely important to study the anti-inflammatory effect of FO on the complications of prematurity. A reduced plasma ARA in the infants receiving the MOSF emulsion has not been reported in previous studies with MOSF,10,11 but this finding is in line with a previous study where D’Ascenzo et al reported significantly lower plasma phospholipid ARA in preterms on PN using a LE with a fat blend containing 10% FO.9 Reduced ARA concentrations have been associated with poor growth in preterm infants receiving an infant milk formula containing marine oil12 and nowadays a well balanced DHA/ARA ratio is recommended in infant formulas containing LCPUFA.23 The associations between low ARA and impaired growth have also been documented in fetuses with intrauterine growth retardation and postnatal growth failure.24 A recent meta-analysis reported no statistically significant association between the type of



7

Table 6 Neonatal outcome and liver function test at 6 weeks postnatal age.a

Clinical outcome RDS, n (%) PDA, n (%) BPD, n (%) NEC grade 2, n (%) Cholestasis, n (%) Late onset sepsis, n (%) IVH grade 3, n (%) PVL, n (%) ROP grade 3, n (%) Mortality, n (%) Liver function Total bilirubin (mg/dL) Conjugated bilirubin (mg/dL) ALP (U/L) AST (U/L) ALT (U/L) GGT (U/L)



S-2.5Fat (n ¼ 22)

S-3.5Fat (n ¼ 19)

pb

pc

pd

17 (80) 10 (48) 6 (30) 0 1 (4) 4 (17) 0 0 0 2 (9)

15 (83) 9 (53) 4 (25) 2 (14) 0 6 (36) 4 (23) 0 0 2 (11)

19 (85) 10 (46) 2 (12) 0 1 (4) 1 (4) 4 (18) 0 0 2 (9)

15 (77) 9 (50) 3 (17) 0 3 (15) 7 (38) 2 (10) 0 0 1 (5)

0.91 0.99 0.40 0.15 0.28 0.02 0.23 1.00 1.00 0.64

0.88 0.90 0.10 0.53 0.61 0.30 0.33 1.00 1.00 0.92

1.00 0.81 1.00 1.00 0.60 0.08 0.46 1.00 1.00 0.68

1.37 (0.59e4.9) 0.41 (0.20e0.52) 731 (649e1365) 22 (19e40) 11 (7e13) 61 (35e146)

0.73 (0.41e0.89) 0.23 (0.16e0.37) 1261 (944e2212) 33 (23e37) 8 (6e11) 40 (24e114)

1.19 (0.58e2.18) 0.40 (0.23e0.60) 1068 (736e1481) 28 (22e40) 10 (7e13) 97 (351e122)

1.81 (0.44e4.62) 0.90 (0.19e1.13) 1201 (967e1558) 26 (18e31) 7 (6e10) 43 (42e68)

0.18 0.13 0.07 0.73 0.51 0.43

0.99 0.12 0.46 0.72 0.53 0.42

0.30 0.91 0.04 0.98 0.63 0.12

a ALP: alkaline phosphatasis; ALT: alanine aminotransferase; AST: aspartate aminotransferase; GGT: gamma glutamyl transpeptidase; BPD: bronchopulmonary dysplasia; BW: birth weight; GGT: gamma glutamyl transpeptidase; IVH: intraventricular hemorrhage; NEC: necrotizing enterocolitis; PDA: patent ductus arteriosus; PVL: periventricular leukomalacia; RDS: respiratory distress syndrome; ROP: retinopathy of prematurity .Continuous variable data are expressed as Median (IQR). Data are compared by chi-square test or Mann Whitney test as appropriate, p < 0.05. b Data are compared between all groups. c Data are compared between MOSF and S groups. d Data are compared between 2.5Fat and 3.5Fat groups.

lipid emulsion and growth.25 In the present study, MOSF groups had a greater postnatal weight loss during the first week of life and took longer to regain birth weight, but weight gain from birth to 36 weeks PMA were similar to the S infants. We do not have a clear explanation for these findings. A study with sufficient statistical power on the effect of IV FO on the growth of the preterm infants and on the common complications of prematurity is needed. MOSF was less tolerated at the dose of 3.5 g kg1 d1 than at 2.5 g kg1 d1 and than S at both fat intakes. In this study, we found that plasma TG were significantly higher in the MOSF-3.5Fat compared with the MOFS-2.5Fat group and they tended to be higher when compared with S groups. We are not aware of similar data in a well-controlled population of ELBW on routine PN. Rayyan et al. administered a 15% FO-LE in infants with birth weight ranging from 500 to 2000 g, but detailed information on lipid tolerance of the smallest infant especially in the first days of life was not available.10 The higher PL and FC in MOSF-3.5Fat infants are also in line with a reduced tolerance of this LE at high fat intakes. In our study, unlikely other studies, we sampled the patient during the continuous infusion of PN. It is therefore plausible that high fat intake in 3.5Fat groups was associated with high plasma triglycerides.26 However MOSF has much higher sTG and over all higher plasma lipids than S. We do not have a clear explanation of this finding. This could be due to the markedly different fatty acid composition27 or to different physical and chemical characteristics of the emulsions. Of note the 3.5Fat groups received statistically significantly higher glucose and protein intakes than the 2.5Fat groups .The differences, albeit statistically significant (glucose intake difference 1.7 0.2 g kg1 d1 which correspond to w15% of the daily intake; amino acid intake difference 0.22 0.07 g kg1 d1) are, in our opinion, biologically small and unlikely to exert a significant effect s TG. To the best of our knowledge there is no data, on the effect of varying glucose intake on plasma lipids in preterm neonates. We speculate from one side that a high-carbohydrate intake could enhance the de novo lipogenesis,28,29 and on the other side that the higher plasma insulin induced by higher glucose and amino acid intakes could enhance lipid clearance by lipases.30 However, despite the previous considerations, we found significantly higher plasma

lipids only in the MOSF-3.5Fat group and not in the S-3.5 patients and both 3.5Fat groups received similar glucose and amino acid intakes. As regards the CE, they were significantly lower in MOSF groups while the FC/CE weight ratio was higher. There was no difference between standard or high fat intakes. We speculate that the very high LLA content of S could enhance the esterification of cholesterol by lecithin cholesterol acyltransferase (LCAT), thus producing higher CE in S than in MOSF group.31,32 The finding of significantly lower AUC of total bilirubin in MOSF than in S pointed in the same direction of the report of Rayyan et al.10 Lower bilirubin could result from a favorable effect of FO on liver function or to reduced peroxidation-inflammation of MOSF compared with the soybean based product. Improved liver function in surgical infants with liver failure receiving 100% FO-LE was reported by Puder et al.33 Unfortunately, in this study we could not obtain enough blood to perform liver function tests and measure inflammatory markers. Further work is warranted to elucidate why intravenous fish oil lower plasma bilirubin. The clinical outcomes of the study groups were not different with the exception of higher incidence of sepsis in the infants receiving the high fat intake. We prefer not to discuss this finding as it could be a chance finding in an underpowered study. As mentioned before we strongly believe that a large study in preterm infants with FO-LE is much needed. Such study should answer if FO-LE may have an effect on weight gain and on the complications of prematurity, including sepsis. In conclusion, the use of a lipid emulsion with 15% FO resulted in marked changes of plasma long-chain fatty acids. Whether the benefits of increasing DHA outweigh the potential negative effect of reduced ARA on growth should be further studied. MOSF at 3.5 g kg1 d1 was associated with reduced lipid tolerance. Funding There was no funding source. Conflict of interest All the authors had no conflicts of interest to declare.


8


Acknowledgments The authors are grateful to the infants’ parents and to the NICU staff. There were 8 authors, who contributed to the work. We report below the contribution of each author:

VPC and RD designed research; RD, CB, MPB conducted research; VPC and RD wrote the paper; VPC had responsibility for final content; SS and AP provided essential materials; PM, PEC performed statistical analysis. All the authors read and approved the final manuscript.

14.

15.

16. 17.

18.

19.

Appendix A. Supplementary data

20.

Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.clnu.2014.01.009.

21.

References 1. Carnielli VP, Verlato G, Pederzini F, Luijendijk I, Boerlage A, Pedrotti D, et al. Intestinal absorption of long-chain polyunsaturated fatty acids in preterm infants fed breast milk or formula. Am J Clin Nutr 1998;67:97e103. 2. Heird WC. Biological effects and safety issues related to long-chain polyunsaturated fatty acids in infants. Lipids 1999;34:207e14. 3. Varsila E, Hallman M, Andersson S. Free-radical-induced lipid peroxidation during the early neonatal period. Acta Paediatr 1994;83:692e5. 4. Lehner F, Demmelmair H, Roschinger W, Decsi T, Szász M, Adamovich K, et al. Metabolic effects of intravenous LCT or MCT/LCT lipid emulsions in preterm infants. J Lipid Res 2006;47:404e11. 5. Rubin M, Moser A, Naor N, Merlob P, Pakula R, Sirota L. Effect of three intravenously administered fat emulsions containing different concentrations of fatty acids on the plasma fatty acid composition of premature infants. J Pediatr 1994;125:596e602. 6. Carnielli VP, Simonato M, Verlato G, Luijendijk I, De Curtis M, Sauer Pieter JJ, et al. Synthesis of long-chain polyunsaturated fatty acids in preterm newborns fed formula with long-chain polyunsaturated fatty acids. Am J Clin Nutr 2007;86:1323e30. 7. Innis SM. Essential fatty acid transfer and fetal development. Placenta 2005;26(Suppl. A):S70e5. 8. Carlson SE, Werkman SH, Rhodes PG, Tolley EA. Visual-acuity development in healthy preterm infants: effect of marine-oil supplementation. Am J Clin Nutr 1993;58:35e42. 9. D’Ascenzo R, D’Egidio S, Angelini L, Bellagamba MP, Manna M, Pompilio A, et al. Parenteral nutrition of preterm infants with a lipid emulsion containing 10% fish oil: effect on plasma lipids and long-chain polyunsaturated fatty acids. J Pediatr 2011;159:33e8. e1. 10. Rayyan M, Devlieger H, Jochum F, Allegaert K. Short-term use of parenteral nutrition with a lipid emulsion containing a mixture of soybean oil, olive oil, medium-chain triglycerides, and fish oil: a randomized double-blind study in preterm infants. J Parenter Enteral Nutr 2012;36:81Se94S. 11. Tomsits E, Pataki M, Tolgyesi A, Fekete G, Rischak K, Szollar L. Safety and efficacy of a lipid emulsion containing a mixture of soybean oil, medium-chain triglycerides, olive oil, and fish oil: a randomised, double-blind clinical trial in premature infants requiring parenteral nutrition. J Pediatr Gastroenterol Nutr 2010;51:514e21. 12. Carlson SE, Werkman SH, Peeples JM, Cooke RJ, Tolley EA. Arachidonic acid status correlates with first year growth in preterm infants. Proc Natl Acad Sci U S A 1993;90:1073e7. 13. Savini S, D’Ascenzo R, Biagetti C, Serpentini G, Pompilio A, Bartoli A. The effect of 5 intravenous lipid emulsions on plasma phytosterols in preterm infants

22.

23.

24.

25.

26.

27.

28.

29.

30. 31.

32.

33.

receiving parenteral nutrition: a randomized clinical trial. Am J Clin Nutr 2013;98:312e8. Carnielli VP, Pederzini F, Vittorangeli R, Luijendijk IH, Boomaars WE, Pedrotti D, et al. Plasma and red blood cell fatty acid of very low birth weight infants fed exclusively with expressed preterm human milk. Pediatr Res 1996;39:671e9. Folch J, Lees M, Sloane Stanley GH. A simple method for the isolation and purification of total lipides from animal tissues. J Biol Chem 1957;226:497e 509. Atkinson KA. An introduction to numerical analysis. New York: John Wiley & Sons; 1989. Walsh MC, Yao Q, Gettner P, Hale E, Collins M, Hensman A, et al. Impact of a physiologic definition on bronchopulmonary dysplasia rates. Pediatrics 2004;114:1305e11. Stoll BJ, Hansen N, Fanaroff AA, Wright LL, Carlo WA, Ehrenkranz RA, et al. Lateonset sepsis in very low birth weight neonates: the experience of the NICHD Neonatal Research Network. Pediatrics 2002;110:285e91. Bell MJ, Ternberg JL, Feigin RD, Keating JP, Marshall R, Barton L, et al. Neonatal necrotizing enterocolitis. Therapeutic decisions based upon clinical staging. Ann Surg 1978;187:1e7. Horbar JD, Badger GJ, Carpenter JH, Fanaroff AA, Kilpatrick S, LaCorte M, et al. Trends in mortality and morbidity for very low birth weight infants, 1991e 1999. Pediatrics 2002;110:143e51. Goulet O, Antebi H, Wolf C, Talbotec C, Alcindor LG, Corriol O, et al. A new intravenous fat emulsion containing soybean oil, medium-chain triglycerides, olive oil, and fish oil: a single-center, double-blind randomized study on efficacy and safety in pediatric patients receiving home parenteral nutrition. J Parenter Enteral Nutr 2010;34:485e95. Clandinin MT, Chappell JE, Leong S, Heim T, Swyer PR, Chance GW. Extrauterine fatty acid accretion in infant brain: implications for fatty acid requirements. Early Hum Dev 1980;4:131e8. Agostoni C, Buonocore G, Carnielli VP, De Curtis M, Darmaun D, Decsi T, et al. Enteral nutrient supply for preterm infants: commentary from the European Society of Paediatric Gastroenterology, Hepatology and Nutrition Committee on Nutrition. J Pediatr Gastroenterol Nutr 2010;50:85e91. Koletzko B, Sauerwald U, Keicher U, Saule H, Wawatschek S, Böhles H, et al. Fatty acid profiles, antioxidant status, and growth of preterm infants fed diets without or with long-chain polyunsaturated fatty acids. A randomized clinical trial. Eur J Nutr 2003;42:243e53. Vlaardingerbroek H, Veldhorst MA, Spronk S, van den Akker CH, van Goudoever JB. Parenteral lipid administration to very-low-birth-weight infantseearly introduction of lipids and use of new lipid emulsions: a systematic review and meta-analysis. Am J Clin Nutr;96:255e68. Brans YW, Andrew DS, Carrillo DW, Dutton EB, Menchaca EM, PuleoScheppke BA. Tolerance of fat emulsions in very-low-birthweight neonates. Monitoring of plasma lipid concentrations. Am J Perinatol 1988;5:8e12. Mattson FH, Grundy SM. Comparison of effects of dietary saturated, monounsaturated, and polyunsaturated fatty acids on plasma lipids and lipoproteins in man. J Lipid Res 1985;26:194e202. Pierro A, Carnielli V, Filler RM, Smith J, Heim T. Metabolism of intravenous fat emulsion in the surgical newborn. J Pediatr Surg 1989;24:95e101. Discussion 101e2. Tappy L, Schwarz JM, Schneiter P, Cayeux C, Revelly JP, Fagerquist CK, et al. Effects of isoenergetic glucose-based or lipid-based parenteral nutrition on glucose metabolism, de novo lipogenesis, and respiratory gas exchanges in critically ill patients. Crit Care Med 1998;26:860e7. Watt MJ, Steinberg GR. Regulation and function of triacylglycerol lipases in cellular metabolism. Biochem J 2008;414:313e25. Sheril A, Jeyakumar SM, Jayashree T, Giridharan NV, Vajreswari A. Impact of feeding polyunsaturated fatty acids on cholesterol metabolism of dyslipidemic obese rats of WNIN/GR-Ob strain. Atherosclerosis 2009;204:136e40. Subbaiah PV, Monshizadegan H. Substrate specificity of human plasma lecithin-cholesterol acyltransferase towards molecular species of phosphatidylcholine in native plasma. Biochim Biophys Acta 1988;963:445e55. Puder M, Valim C, Meisel JA, Le HD, de Meijer VE, Robinson EM, et al. Parenteral fish oil improves outcomes in patients with parenteral nutrition-associated liver injury. Ann Surg 2009;250:395e402.


Erythrocyte eicosapentaenoic acid versus docosahexaenoic acid as a marker for fish and fish oil consumption.

Docosahexaenoic Acid and Periodontitis in Adults: A Randomized Controlled Trial.

The Effect of Fish Oil-Based Lipid Emulsion and Soybean Oil-Based Lipid Emulsion on Cholestasis Associated with Long-Term Parenteral Nutrition in Premature Infants.

Reduced ratio of eicosapentaenoic acid and docosahexaenoic acid to arachidonic acid is associated with early onset of acute coronary syndrome.

Dietary manipulation with high marine fish oil intake of fatty acid composition and arachidonic acid metabolism in rat cerebral microvessels.

Pretreatment with Fish Oil-Based Lipid Emulsion Modulates Muscle Leukocyte Chemotaxis in Murine Model of Sublethal Lower Limb Ischemia.

Fish Oil Lipid Emulsion-Associated Sea-Blue Histiocyte Syndrome in a Pediatric Patient.

Docosahexaenoic Acid and Arachidonic Acid Nutrition in Early Development.

Reduced macrophage selenoprotein expression alters oxidized lipid metabolite biosynthesis from arachidonic and linoleic acid.

Breast milk from women living near Lake Malawi is high in docosahexaenoic acid and arachidonic acid.

Developmental Outcomes at 24 Months of Age in Toddlers Supplemented with Arachidonic Acid and Docosahexaenoic Acid: Results of a Double Blind Randomized, Controlled Trial.

Characteristics of lipid and fatty acid of marine gastropod Turbo cornutus: high levels of arachidonic and n-3 docosapentaenoic acid.

Human prefrontal cortex phospholipids containing docosahexaenoic acid increase during normal adult aging, whereas those containing arachidonic acid decrease.

Impact of Providing a Combination Lipid Emulsion Compared With a Standard Soybean Oil Lipid Emulsion in Children Receiving Parenteral Nutrition: A Systematic Review and Meta-Analysis.

Four-year placebo-controlled trial of docosahexaenoic acid in X-linked retinitis pigmentosa (DHAX trial): a randomized clinical trial.

AA to impaired arterial stiffness.

Dietary fish oil interferes with renal arachidonic acid metabolism in rats: correlations with renal physiology.

Low doses of eicosapentaenoic acid and docosahexaenoic acid from fish oil dose-dependently decrease serum triglyceride concentrations in the presence of plant sterols in hypercholesterolemic men and women.

A randomized, controlled clinical trial of intravenous lipid emulsion as an adjunctive treatment for permethrin toxicosis in cats.

Intravenous lipid emulsion improves recovery time and quality from isoflurane anaesthesia: a double-blind clinical trial.

Lipid-lowering and anti-inflammatory effects of omega 3 ethyl esters and krill oil: a randomized, cross-over, clinical trial.

Fish oil, contained in eicosapentaenoic acid and docosahexaenoic acid, attenuates testicular and spermatological damage induced by cisplatin in rats.

Lipid Body Organelles within the Parasite Trypanosoma cruzi: A Role for Intracellular Arachidonic Acid Metabolism.

Induction of aortic lipid deposition in a high-response (ExHC) rat fed a diet containing cholesterol and cholic acid.