Article Text
PDF
Abstract
Background There is limited evidence regarding the optimal time to commence parenteral nutrition (PN) in term and late preterm infants.
Design Single-centre, non-blinded, exploratory randomised controlled trial.
Setting A level-3 neonatal unit in a stand-alone paediatric hospital.
Patients Infants born ≥34 weeks of gestation and ≤28 days, who needed PN. Eligible infants were randomised on day 1 or day 2 of admission.
Interventions Early (day 1 or day 2 of admission, N=30) or late (day 6 of admission, N=30) PN.
Main outcome measures Plasma phenylalanine and F2-isoprostane levels on day 4 and day 8 of admission. Secondary outcomes were amino-acid and fatty-acid profiles on day 4 and day 8, and clinical outcomes.
Results The postnatal age at randomisation was similar between the groups (2.3 (SD 0.8) vs 2.3 (0.7) days, p=0.90). On day 4, phenylalanine levels in early-PN infants were higher than in late-PN (mean (SD) 62.9 (26.7) vs 45.5 (15.3) µmol/L; baseline-adjusted percentage difference 25.8% (95% CI 11.6% to 39.9%), p<0.001). There was no significant difference in phenylalanine levels between the two groups on day 8. There was no significant difference between the groups for F2-isoprostane levels on day 4 (early-PN mean (SD) 389 (176) vs late-PN 419 (291) pg/mL; baseline-adjusted percentage difference: −4.4% (95% CI −21.5% to 12.8%) p=0.62) and day 8 (mean (SD) 305 (125) vs 354 (113) pg/mL; adjusted mean percentage difference −16.1 (95% CI −34.1 to 1.9) p=0.09).
Postnatal growth restriction for weight was less severe in the early-PN group (change in weight z-score from baseline to discharge: −0.6 (0.6) vs −1.0 (0.6); p=0.02). The incidence of hyperglycaemia was greater in the early-PN group (20/30 (66.7%) vs 11/30 (36.7%), p=0.02).
Conclusions The timing of the commencement of PN did not seem to affect the degree of oxidative stress in critically ill term and late preterm infants. The effect of transiently high plasma phenylalanine with early PN on clinical outcomes requires further investigation.
Trial registration number ACTRN12620000324910.
- Neonatology
Data availability statement
Data are available on reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
The subgroup analysis of a paediatric randomised controlled trial found that late commencement of parenteral nutrition (PN) in term infants is associated with earlier discharge from the intensive care unit but increased risk of hypoglycaemia.
There are no studies comparing the plasma amino acid and fatty acid profiles and markers of oxidative stress in infants receiving different timing of the commencement of PN.
WHAT THIS STUDY ADDS
Early commencement of PN results in similar plasma F2-isoprostane levels and transiently higher phenylalanine and total amino acid levels compared with late PN in term and late preterm infants.
Provides preliminary information on plasma and red blood cell fatty acid levels of critically ill term and late preterm infants receiving early versus late PN.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
This study provides valuable data informing the design of future trials.
Background
Term and late preterm infants (≥34 weeks of gestation) with critical illness may not tolerate enteral nutrition (EN), which necessitates the use of parenteral nutrition (PN). However, currently, there is limited evidence regarding the optimal time to commence PN in term and late preterm infants.1 A subgroup analysis of a paediatric multicentre randomised controlled trial (RCT) concluded that early commencement of PN was associated with longer intensive care unit (ICU) stay but lower risk of hypoglycaemia compared with late PN in term neonates.2 Given the limited evidence in this area, guidelines from professional bodies have varied. The NICE guidelines recommend commencing PN within 72 hours after birth if no progress is made with EN in term and late preterm infants.3 In contrast, the ESPGHAN/ESPEN/ESPR/CPEN guidelines recommend withholding parenteral amino acids for 1 week in critically ill term infants; they do not make a recommendation on the timing of parenteral lipids.4 Recent surveys have also reported variations in practice among clinicians.5–7
Early commencement of PN has the potential to minimise catabolism, prevent essential fatty acid and amino acid deficiency, improve weight gain and clinical outcomes. However, there are concerns about the harmful effects of PN. Parenteral amino acids may lead to hyperammonaemia, azotaemia, metabolic acidosis, free radical injury and slower growth of head circumference and metabolic syndromes later in life.8–12 Parenteral lipids may exacerbating free radical activity,13 14 increasing the risk of sepsis,15 16 worsening respiratory function by decreasing arterial oxygenation,17–19 increasing the risk of kernicterus in infants with hyperbilirubinaemia due to free fatty acids displacing bilirubin from albumin binding sites20 and may worsen pulmonary hypertension.21
Many reports have described abnormal blood amino acid profiles, especially phenylalanine, in critically ill neonates,22–26 suggesting metabolic derangements of certain amino acids during critical illness. If such infants are administered parenteral amino acids early during their critical illness, plasma levels may reach very high levels, leading to potential harm. Observational studies have demonstrated that high plasma amino acid concentrations are associated with adverse clinical outcomes.23 27 28 Experimental studies have shown that hyperphenylalaninaemia can induce oxidative stress, inflammation, mitochondrial dysfunction and neuronal apoptosis.29–33 Hence, it is important to weigh the risks and benefits of commencing PN very early in the neonatal period.
Oxidative stress is a major characteristic of various diseases in humans. In the context of critically ill term and late preterm infants, sources of oxidative stress are birth trauma, reperfusion injury from hypoxia, oxygen therapy, acidosis, phototherapy, mechanical ventilation, infection and inflammation.34–36 Oxidative stress is present in conditions that predispose to critical illness in term and late preterm infants. The examples are persistent pulmonary hypertension of the newborn (PPHN),37 hypoxic ischaemic encephalopathy (HIE),38 necrotising enterocolitis (NEC)39 and sepsis.40 41 Oxidative stress can lead to the shortening of telomeres in vivo.42
In the presence of oxygen, nutrients such as lipids, amino acids and vitamins are potent electric donors that promote the generation of peroxides.11 Polyunsaturated fatty acids present in lipid emulsions are susceptible to oxidation.43 44 Furthermore, various studies have shown that oxidation of amino acid solutions, vitamins and trace elements can also result in increased oxidative stress.13 45–48 Hence, early intake of PN (amino acid solution and lipids) may worsen the oxidative stress in critically ill infants.
Given that oxidative stress plays a significant role in the pathogenesis of diseases in critically ill term and late preterm infants and that PN is known to worsen free-radical injuries, it is important to compare the oxidative stress levels in infants receiving different PN strategies such as timing of the commencement.
F2-isoprostanes are prostaglandin-like substances that are formed in vivo by free radical-induced peroxidation of arachidonic acids.49 There is considerable literature demonstrating F2-isoprostanes are highly reliable measures of in vivo oxidative stress.50 51 F2-isoprostanes are (a) chemically stable, (b) specific products of peroxidation, (c) formed in vivo, (d) present in detectable amounts in all normal tissues and biological fluids and (e) unaffected by lipid contents in the diet.52
Recent studies found that increased F2-isoprostane levels, an indicator of oxidative stress,49 are associated with adverse clinical outcomes.53–55
To build evidence in this area, we conducted an exploratory RCT to compare biochemical parameters between early versus late PN in term and late preterm infants. We hypothesised that late commencement of PN will achieve lower (and more physiological) levels of plasma phenylalanine and plasma F2-isoprostanes on day 4 and day 8 compared with early commencement of PN in term and late preterm infants.
Methods
Design and oversight
A single-centre, non-blinded exploratory RCT in the neonatal ICU (NICU) of Perth Children’s Hospital, Australia. The trial was preregistered with Australia and New Zealand Clinical Trials Registry (ACTRN12620000324910), and the protocol has been published.56 Data and safety monitoring committee (DSMC) charter57 was established prior to commencing the trial. The DSMC conducted two interim analyses and advised continuation to full recruitment. Patients or the public were not involved in the design, or conduct, or reporting, or dissemination plans of our research.
Participants
Infants born ≥34 weeks of gestation and ≤28 days, who had a high likelihood of being unable to tolerate enteral feeds for at least 3–5 days, were eligible to participate. Written parental/guardian consent was obtained prior to inclusion. We excluded infants (a) who had received PN at a referring hospital, (b) with suspected inborn error of metabolism or (c) who had established more than 50% of intake as EN for >12 hours since birth.
Interventions
Infants randomised to early-PN were commenced PN on day 1 or day 2 of admission to the NICU. Late-PN infants were commenced on day 6 of admission and received 10% glucose and electrolytes till then. Group assignment was allocated by the clinical trial pharmacist using a computer-generated randomisation sequence in a 1:1 ratio in block sizes of 2 and 4 and stratified according to the main diagnostic category (medical or surgical). Allocation concealment was achieved using opaque sealed envelopes. Blinding of families, healthcare providers and investigators were not possible. The laboratory personnel and the trial biostatistician were blinded to group allocation.
Prescriptions of PN were determined by the treating clinician daily, depending on the infant’s condition and using hospital guidelines. EN was commenced as early as feasible in both groups (online supplemental materials 1-3). PN was stopped when EN provided 80%–100% of target (150 mL/kg/day).
Supplemental material
Data collection
Non-fasting venous bloods (1.4 mL) were collected at baseline, day 4 and day 8 of admission, or at discharge if before day 8, and transferred on ice for centrifugation at 4°C within 2 hours. Plasma and red blood cell (RBC) samples were extracted and transferred to cryovials and stored at −80°C. All samples were finally transported to laboratories for analyses.56
Daily intakes of PN and EN were calculated based on actual intake (online supplemental material 4). Clinical data were collected from the medical records of study infants.
Study outcomes
The primary outcomes were plasma phenylalanine and plasma F2-isoprostane levels on day 4 and day 8 of admission. Secondary outcomes included amino acid profiles, fatty acid profiles and clinical outcomes.56
Statistical analysis
With 30 infants in each arm, the study was powered to detect an effect size of 0.75, being the difference in means relative to the pooled SD on the log scale, translating to approximately 80% power (alpha=0.05) to detect a differential of 23% in the phenylalanine concentrations and 39% in F2-isoprostane levels using observed estimates of variability. Baseline characteristics are reported as counts (percentages) for categorical outcomes, mean (SD) or median (IQR) for continuous outcomes.
A linear regression framework was used to analyse differences between the two groups in log-transformed phenylalanine and F2-isoprostane levels at days 4 and 8, with and without controlling for baseline F2-isoprostanes/phenylalanine, sex, gestational age, medical/surgical group and severity of illness (Score for Neonatal Acute Physiology with Perinatal Extension-II).58 Regression model coefficients, representing the mean of log-transformed early-PN levels minus the mean of log-transformed late-PN levels, were multiplied by 100 to estimate symmetric percentage differences (sympercents). Infants randomised to late-PN who did not require PN by day 6 were analysed as being in the late PN group. Values were missing for 6.7% of primary outcome data, primarily for technical reasons and 1.1% of main secondary outcomes; all demographic and clinical data at baseline were collected. Multivariate imputation by chained equations was used to accommodate missing outcome data under an assumption of missing at random. Further detail is included in online supplemental material 5. All analyses were based on the principle of intention to treat. Key between-group differences are presented with 95% CIs, and as analyses were primarily exploratory in nature adjustments for multiplicity were not performed. P values <0.05 were considered significant for primary outcomes.
Results
Study recruitment commenced in June 2020 and was completed in January 2022 after achieving a full sample size. 30 infants were randomised to early-PN and 32 to late-PN (figure 1); consent was withdrawn for two in the late-PN group. Baseline characteristics are given in table 1.
CONSORT flow diagram showing participant flow through the trial. CONSORT, Consolidated Standards of Reporting Trials; EN, enteral nutrition; PN, parenteral nutrition.
Patient characteristics
The mean postnatal age at randomisation was similar between the two groups (median (Q1, Q3): 2.0 (2.0, 3.0) days vs 2.0 (2.0, 3.0) days; mean (SD): 2.3 (0.8) days vs 2.3 (0.7) days, p=0.90).
All infants in the early-PN group received parenteral amino acids and lipids, with 100% (n=30) and 87% (n=26), respectively, commencing by day 2 of admission. In the late-PN group, 63% (n=19) and 60% (n=18) of infants received parenteral amino acids and lipids, respectively. The remaining late-PN infants did not require PN because they were tolerating EN by day 6. Two infants in the late-PN group received PN on day 4 as the neonatologist deemed PN was necessary.
EN was commenced as early as day 2 of admission for both groups, and 75% of infants received some amounts of EN by day 7.
Table 2 summarises daily nutrient intakes on day 1, day 4 and day 8 of admission. Marked differences in parenteral amino acids, lipids, parenteral and total energy between the two groups were observed on day 4.
Daily nutrient intakes on day 1, day 4 and day 8 of admission
Detailed daily nutrient intakes until day 21 are shown in online supplemental tables S1.1–S4.2 and online supplemental figure S1.
Primary outcomes
Plasma phenylalanine
There was no difference between the two groups in baseline plasma phenylalanine levels. On day 4, phenylalanine levels in the early group were significantly higher than the late-PN group (table 3, mean (SD) 62.9 (26.5) vs 45.5 (15.3) µmol/L; percentage difference (95% CI) adjusted for baseline characteristics: 25.8% (11.6% to 39.9%), p<0.001), with an average 11% increase from baseline in the early group contrasting with an average 19% decrease in the late group. The initial increases in the early-PN group were followed by an average 20% decrease after day 4, and by day 8, the difference between the two groups had dissipated (p=0.94) (table 3).
Plasma phenylalanine and F2-isoprostane levels
Transient elevation above 112 µmol/L, the maximum level in a cohort of healthy breast-fed infants,59 was observed in two infants in the early group (days 1/4/8: 110/155/83 and 35/139/45 µmol/L). Eight infants in the late PN group had levels <35 µmol/L, the minimum level observed in the cohort.59 On day 8, three infants in the late-PN group and five in the early-PN group had levels <35 µmol/L.
Plasma F2-isoprostanes
Plasma F2-isoprostane levels were similar between the two groups at baseline and generally declined after admission in all infants (table 3). There was no statistically significant difference between the two groups on day 4 (early-PN mean (SD) 389 (176) vs late-PN 419 (291) pg/mL; adjusted percentage difference: −4.4% (95% CI −21.5% to 12.8%) p=0.62) and day 8 (early PN mean (SD) 305 (125) vs late-PN 354 (113) pg/mL; adjusted mean percentage difference: −16.1 (95% CI −34.1 to 1.9) p=0.09).
Secondary biochemical endpoints
The total amino acid levels were similar between the two groups at baseline and day 8. However, on day 4, average levels in the early-PN group were 53% (37%–70%) higher than the late-PN group. Transient elevation in concentration of phenylalanine, methionine and serine was observed in early-PN infants, whereas for other amino acids increased at day 4 were sustained and then matched at day 8 in late-PN infants (online supplemental table S5, online supplemental figure S2).
The plasma and RBC fatty acid profiles were similar between the two groups at baseline. However, relative compositions of n-6 and n-3 plasma and RBC fatty acids differed between the two groups on days 4 and 8. In particular, n-3 fatty acid levels were typically increased in early-PN infants at day 4, with differences between the two groups remaining at day 8 for eicosapentanoic acid (online supplemental tables S6 and S7, online supplemental figures S3 and S4).
Secondary clinical endpoints
There was no mortality. The overall incidence of hyperglycaemia (plasma glucose >150 mg/dL or 8.3 mmol/L) was 52% (early-PN 66.7% vs late-PN 36.7%; p=0.02), and incidence of hypoglycaemia (plasma glucose <2.6 mmol/L) was 10% (early-PN 13.3% vs late-PN 6.7%; p=0.39). The degree of postnatal growth restriction (PNGR: decline in weight z score from baseline to discharge) was less severe in the early-PN group compared with late PN (mean (SD): −0.6 (0.6) vs −1.0 (0.6)) p=0.02). PNGR for head circumference (early PN −0.41 (1.14) vs late PN −0.36 (1.00), p=0.87) and other clinical outcomes were similar between the two groups (online supplemental table S8).
Discussion
This RCT found that early PN has no significant effects on plasma F2-isoprostane (an important marker of oxidative stress) levels but can lead to transiently higher phenylalanine levels compared with delayed commencement in term and late preterm infants. In addition, on day 4, total amino acid levels in the early-PN group were higher than the late-PN group. PNGR for weight was less severe in the early-PN group head growth was similar. The incidence of hyperglycaemia was higher in the early-PN group.
To our knowledge, the only other RCT addressing the issue of early versus late PN in term infants was the neonatal subgroup analysis2 of the PEPaNIC trial.60 They found that early PN was associated with delayed discharge from the ICU but a decreased risk of hypoglycaemia.
We found that the plasma phenylalanine and total amino-acid levels on day 4 were higher in the early-PN than late-PN group. The most likely explanation is that the late-PN group was not on PN on day 4. The other reason could be the inability of the liver to handle parenterally administered amino acids given that liver dysfunction is common in critical illness.27 61–63
Higher plasma phenylalanine and total amino acids in the early-PN group on day 4 may be of concern because some studies have shown that high plasma phenylalanine and amino acid levels are associated with worse clinical outcomes.27 28 In an RCT of preterm infants, higher plasma phenylalanine levels were associated with poor growth and adverse neurodevelopment at 18–24 months.28 In an observational study, children (including neonates) undergoing cardiac surgery, non-surviving children had higher plasma amino acid levels.27 We will evaluate neurodevelopmental outcomes of our cohort when the last participant reaches 2 years.
Similar plasma F2-isoprostane levels between the early and late PN groups suggest that the timing of commencing PN may not affect the severity of oxidative stress in critically ill neonates. This was an unexpected finding that was not in line with our hypothesis. Polyunsaturated fatty acids, one of the main constituents of lipid emulsions, are susceptible to oxidation.36 43 44 64 Amino acid solutions, vitamins and trace elements also contribute to oxidative stress, especially when exposed to ambient light.13 45 46 48 65 66 Aggressive energy intake during the early phase of illness may induce oxidative stress.67 Infants in our early-PN group had all the above risk factors but indication of higher oxidative stress compared with the late-PN group is lacking. Our findings may be explained by the following reasons. The early lipids may have facilitated antioxidant effects by providing n-3 fatty acids from SMOF lipids.68 69 This is supported by the finding that the percentage of n-3 fatty acids in total plasma and RBC/plasma eicosapentaenoic acid in our study infants were higher in the early PN group on day 4 and day 8. Studies have shown that n-3 fatty acids are associated with a reduction in oxidative stress (F2- isoprostanes).70 Furthermore, it is possible that the study was underpowered to detect differences in oxidative stress.
Our study found that, overall, PNGR for weight was more severe in the late-PN group. This may be because they received lesser amounts of PN in the first week of admission. There was no effect on z-scores for head circumference between the two groups. Our finding differs from the PEPaNIC trial that showed late PN did not lead to weight z-score deterioration at discharge (136 term infants).71 The lack of between-group differential growth seen in the PEPaNIC trial may be explained by the fact that children in that study received substantial amounts of EN. The relation between PNGR and adverse neurodevelopmental outcomes is an area of debate, with observational studies showing an association,72–74 whereas interventional studies (RCTs) have not shown benefits of aggressive nutrition.10 75 Importantly, most studies are in extremely preterm infants rather than late preterm and term infants. The 4-year follow-up of 121 infants <1 year in the PEPaNIC trial found that growth restriction was not associated with adverse neuropsychological outcomes.76 Large RCTs are needed to evaluate the effect of timing of the commencement of PN on long-term neurodevelopment in term and late preterm infants.
In contrast to the findings of the PEPaNIC trial, the incidence of hypoglycaemia was low in our study infants, and the late PN did not worsen the risk of hypoglycaemia. This could be attributed to how glucose homoeostasis in study infants was achieved. In two of the study sites of the PEPaNIC trial, insulin infusion was used to tightly control blood glucose levels, whereas we do not apply that practice. The ‘tight glycaemic control’ has been reported to be associated with significant increase in hypoglycaemia in critically ill infants and children admitted to PICU.77–79
Our study found a higher incidence of hyperglycaemia in the early-PN group, the likely reason being the PN contains 12% glucose and early-PN group would have received it from the very first day. In contrast, the late-PN group received 10% glucose as the maintenance fluid for the first 5 days. Given that hyperglycaemia is associated with mortality and morbidities in neonates,80 particular attention to glycaemic control should be exercised in future studies.
The strength of our study is that it is the first to compare biochemical outcomes of early versus late PN in exclusive late preterm and term infants. We have compared key biochemical outcomes such as plasma phenylalanine and F2-isoprostanes, amino acid and fatty acid profiles.
The limitations are (a) we did not have an internal reference group (healthy breastfed infants) to compare the plasma phenylalanine, other amino acids, fatty acids and F2-isoprostane levels of the study infants; (b) the effects of parenteral amino acids and lipids could not be assessed separately; (c) due to the small sample size, the study was not powered to compare clinical outcomes and (d) the study population was heterogeneous. Future studies should overcome the above limitations and be of large size to evaluate long-term outcomes and consider separately evaluating the effects of lipids and amino acids.
Conclusion
The timing of the commencement of PN did not seem to affect the degree of oxidative stress in critically ill term and late preterm infants. The effect of transiently high plasma phenylalanine with early parenteral amino acids on clinical outcomes requires further investigation.
Data availability statement
Data are available on reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and was approved by Child and Adolescent Health Service Human Research Ethics Committee RGS 0000003537University of Western Australia Human Research Ethics Committee RA/4/20/5825. Participants gave informed consent to participate in the study before taking part.
Acknowledgments
We sincerely thank the following people: Infants and parents for providing consent and participating in the RCT; Nursing, medical and clerical staff of NICU and PCC at Perth Children’s Hospital for collecting blood samples and for providing support; Ms Margaret Shave (Pharmacy Department) for randomisation; Dr Chris Gorman and Ms Sarah Herold (Biospecimens Service, Telethon Kids Institute, Western Australia) for processing and storing blood samples; Dr David Cantor (Australian Proteome Analysis Facility, Macquarie University, New South Wales) for performing plasma amino acid analysis; Dr Henrietta Koch (The University of Western Australia) for performing plasma F2-isoprostane analysis; and Dr Abhijeet Rakshasbhuvankar and Ms Nabeelah Mukadam, members of the data safety and monitoring committee.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors KM, the guarantor, contributed to the conception and design, acquisition of data, analysis, interpreted data and wrote the first and final draft of the manuscript. EM contributed to the conception and design, analysis, interpreted data and final draft of the manuscript. KC contributed to the biochemical analyses (F2-isoprostanes) and final draft of the manuscript. TAM contributed to the biochemical analysis (plasma and RBC fatty acids), interpreted data and final draft of the manuscript. SP contributed to the conception and design and final draft of the manuscript. KS contributed to the conception and design and final draft of the manuscript. SR contributed to the conception and design, interpreted data and final draft of the manuscript. All authors read and approved the final manuscript.
Funding Supported by the Centre for Neonatal Research and Education, University of Western Australia.
Disclaimer The content of this manuscript is solely the responsibility of the authors.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.