点击显示 收起
1 From Indiana University, Indianapolis (BBP and JAL); Yale University, New Haven, CT (RAE); Emory University, Atlanta (BJS); the Research Triangle Institute, Research Triangle Park, NC (MAK); the National Institute of Child Health and Human Development, Bethesda, MD (LLW); the Women and Infant’s Hospital, Providence, RI (WO); the University of New Mexico, Albuquerque (L-AP); the University of Miami (CRB); the University of Alabama, Birmingham (WAC); the University of Cincinnati (EFD); Case Western Reserve University, Cleveland (AAF); the University of Tennessee, Memphis (SBK); the University of Texas Southwestern Medical Center, Dallas (ARL); Wayne State University, Detroit (SS); Stanford University, Stanford, CA (DKS); and the University of Texas at Houston (JET).
2 Supported by cooperative agreements with the National Institute of Child Health and Human Development (U10 HD27856, U10 HD27871, U10 HD27851, U01 HD36790, U10 HD27904, U10 HD27881, U10 HD21397, U10 HD34216, U10 HD27853, U10 HD21364, U10 HD21415, U10 HD40689, U10 HD21385, U10 HD27880, U10 HD21373, and HD 19089), by the General Clinical Research Centers (MO1 RR 00750, MO1 RR 06022, MO1 RR 00997, MO1 RR 08084, and MO1 RR 00070), and by the Riley Memorial Association. 3 Address reprint requests to BB Poindexter, Riley Hospital for Children, 699 West Drive RR 208, Indianapolis, IN 46202–5210. E-mail: bpoindex{at}iupui.edu.
ABSTRACT
Background: Glutamine is one of the most abundant amino acids in both plasma and human milk and may be conditionally essential in premature infants. However, glutamine is not provided by standard intravenous amino acid solutions.
Objective: We assessed the effect of parenteral glutamine supplementation on plasma amino acid concentrations in extremely low-birth-weight infants receiving parenteral nutrition (PN).
Design: A total of 141 infants with birth weights of 401–1000 g were randomly assigned to receive a standard intravenous amino acid solution that did not contain glutamine or an isonitrogenous amino acid solution with 20% of the total amino acids as glutamine. Blood samples were obtained just before initiation of study PN and again after the infants had received study PN (mean intake: 2.3 ± 1.0 g amino acids · kg-1 · d-1) for 10 d.
Results: Infants randomly assigned to receive glutamine had mean plasma glutamine concentrations that increased significantly and were 30% higher than those in the control group in response to PN (425 ± 182 and 332 ± 148 µmol/L for the glutamine and control groups, respectively). There was no significant difference between the 2 groups in the relative change in plasma glutamate concentration between the baseline and PN samples. In both groups, there were significant decreases in plasma phenylalanine and tyrosine between the baseline and PN samples; the decrease in tyrosine was greater in the group that received glutamine.
Conclusions: In extremely low-birth-weight infants, parenteral glutamine supplementation can increase plasma glutamine concentrations without apparent biochemical risk. Currently available amino acid solutions are likely to be suboptimal in their supply of phenylalanine, tyrosine, or both for these infants.
Key Words: Glutamine • phenylalanine • tyrosine • extremely low-birth-weight • premature infants • low-birth-weight infants • parenteral nutrition • neonatology • neonatal care
INTRODUCTION
Glutamine is one of the most abundant amino acids in both plasma and human milk (1–3), yet it is not a component of standard intravenous amino acid solutions because of its instability in solution. The importance of glutamine, particularly during periods of stress, injury, and illness, is increasingly being recognized. Studies in critically ill adults provide compelling evidence that, in addition to reducing episodes of sepsis, parenteral glutamine supplementation significantly reduces mortality (4).
Extremely premature neonates are at high risk for developing late-onset sepsis (5). The National Institute of Child Health and Human Development Neonatal Research Network is currently conducting a multicenter, randomized, double-masked controlled clinical trial to assess the efficacy and safety of parenteral glutamine supplementation with early parenteral nutrition (PN) in reducing the incidence of mortality or late-onset sepsis in extremely low-birth-weight (ELBW) infants.
The primary purpose of the current study, which is a subset of the main trial, was to determine the effect of parenteral glutamine supplementation on plasma amino acid concentrations as a means of assessing the safety of parenteral glutamine supplementation in ELBW infants. We hypothesized that substituting 20% of the total amino acid intake as glutamine would result in a significant increase in plasma glutamine concentrations and would not result in significant aberrations in the remaining amino acid profile or significantly increase plasma ammonia concentrations.
Although PN is widely used in the neonatal intensive care unit, there is a paucity of data on plasma amino acid concentrations in ELBW infants, the population that most frequently receives prolonged PN. Consequently, a second aim of the current study was to collect data on amino acid concentrations in ELBW infants receiving PN. To address these issues, plasma amino acid and ammonia concentrations were measured both before and during provision of PN.
SUBJECTS AND METHODS
Study subjects and design
A priori, we determined that plasma amino acid and ammonia concentrations would be obtained from the first 10 infants randomly assigned at each of the 14 participating National Institute of Child Health and Human Development Neonatal Research Network centers. Inclusion criteria were a birth weight between 401 and 1000 g and enrollment at or before 72 h of age. We excluded infants with major congenital anomalies or congenital nonbacterial infection, those thought to have terminal illness (as indicated by a pH below 6.80 or by the presence of hypoxia with bradycardia for > 2 h), and those for whom a decision had been made that full support would not be provided. The Institutional Review Board at each center approved the study, and written informed consent was obtained from the parents of each infant.
The infants were stratified according to center and birth weight (401–750 or 751–1000 g) and were assigned to the control or glutamine group by a hospital pharmacist using a randomization list provided by the data coordinating center (Research Triangle Institute).
Infants in the control group received TrophAmine (B Braun, Irvine, CA) as their intravenous amino acid solution. Infants in the glutamine group received an isonitrogenous study amino acid solution with 20% glutamine; this solution consisted of TrophAmine and nonpyrogenic L-glutamine powder (Ajinomoto, Raleigh, NC). A Food and Drug Administration–approved drug manufacturer compounded the study amino acid solution under controlled, clean-room conditions (Central Admixture Pharmacy Services Inc, Irvine, CA). Before beginning the study, both Central Admixture Pharmacy Services and the laboratory at Indiana University School of Medicine documented the sterility and stability of the glutamine-enriched amino acid solution to 17 wk. The hospital pharmacist labeled all bags of PN with the total amount, in g · kg-1 · d-1, of study amino acids (as ordered by the attending physician). The glutamine-enriched solution was visually indistinguishable from standard PN. A standard dose of cysteine hydrochloride (40 mg/g amino acid; 120 mg · kg-1 · d-1 maximum) was added to the final compounded bag of PN in both groups (6, 7).
Although the study protocol gave specific guidelines for the use of PN, including early initiation and rapid advancement of amino acid administration to 3.0–3.5 g · kg-1 · d-1, the neonatologist caring for the infant determined the final prescription for total PN and made all decisions related to the introduction and advancement of enteral feedings.
Blood sample collection
Blood samples were obtained from each infant at 2 specific time points: just before initiation of study PN (baseline sample) and again after the infant had received study PN for 10 d (PN sample). A minimum of 0.5 mL whole blood was collected in a heparin-containing microtainer and was immediately centrifuged at room temperature for 5 min. The plasma was separated and then frozen at -70 °C for later analysis.
Plasma amino acid analysis
All samples were shipped frozen on dry ice to a central amino acid laboratory at Indiana University. Amino acid concentrations were determined by using standard ion-exchange chromatography methodology with post-column Ninhydrin detection and an automated amino acid analyzer (model 6300; Beckman Instruments, Fullerton, CA).
Plasma ammonia analysis
For plasma ammonia analysis, arterial samples were preferentially obtained and were transported on ice to the local hospital laboratory at each of the participating centers. The laboratory technique for ammonia analysis was not standardized between the centers. The study coordinators recorded the values obtained on the data collection forms.
Statistical analyses
The statistical analyses were performed with SAS, release 8.2 (SAS Institute Inc, Cary, NC). To analyze the differences in baseline characteristics, we used the Wilcoxon rank-sum test for the continuous variables (eg, birth weight) and the Pearson chi-square test for the categorical variables (eg, sex).
The 2 groups were compared according to the intention to treat; subjects were included in the analysis in their original groups regardless of the amount of PN received before or at the time of the PN sample. A logarithmic transformation of the data was applied to minimize the effect of skewness of the amino acid data before statistical analysis. The Wilcoxon signed-ranks test was then used for the within-group comparisons of relative change from baseline sample to PN sample. To assess the primary hypothesis regarding the effect of treatment on the relative change from baseline sample to PN sample, a general linear model multiple regression analysis was performed separately for each of the amino acids. The variables included in the regression were center and birth weight stratum, and the outcome was the change score calculated on the log scale. All data are presented in the original units.
An independent Data Safety and Monitoring Committee analyzed the data after the samples were obtained from the first 10 infants at each center and determined that no further monitoring of amino acid or ammonia concentrations was necessary to monitor safety.
RESULTS
Plasma amino acid samples were obtained from 141 infants (69 in the control group and 72 in the glutamine group). The 2 groups did not differ significantly with regard to any of the baseline characteristics (Table 1). Parenteral amino acid intake did not differ significantly between the groups at either of the 2 times when plasma samples were obtained (Table 2). Parenteral amino acid intake ranged from 0 to 3.9 g · kg-1 · d-1 (median: 2.52 g · kg-1 · d-1) in the control group and from 0 to 4.1 g · kg-1 · d-1 (median: 2.68 g · kg-1 · d-1) in the glutamine group at the time of the PN sample. Total energy intake from PN was not significantly different between the 2 groups (268 ± 105 and 251 ± 113 kJ · kg-1 · d-1 or 64 ± 25 and 60 ± 27 kcal · kg-1 · d-1 in the glutamine and control groups, respectively; P = 0.39) at the time of the PN sample. Although energy intake from enteral feedings was slightly higher in the control group than in the glutamine group at the time of the PN sample (67 ± 105 and 109 ± 126 kJ · kg-1 · d-1 or 16 ± 25 and 26 ± 31 kcal · kg-1 · d-1 in the glutamine and control groups, respectively; P = 0.04), enteral feedings provided 30% of total energy intake in both groups. Plasma amino acid concentrations before initiation of study PN (baseline sample) and during PN (PN sample; average age: 11.6 ± 2.8 d) are shown in Table 3.
View this table:
TABLE 1 . Characteristics of the study infants at baseline1
View this table:
TABLE 2 . Age and nutritional intake of the study infants1
View this table:
TABLE 3 . Plasma amino acid concentrations at baseline and during parenteral nutrition (PN)1
Essential amino acids
In both groups, the plasma concentration of total essential amino acids increased significantly between the baseline sample and PN sample. The PN-sample median was 20% higher than the baseline-sample median in the control group and was 18% higher in the glutamine group. There was no effect of treatment group on the relative change in concentration of total essential amino acids. There were significant increases in the concentrations of the essential amino acids leucine, isoleucine, valine, threonine, and methionine in both groups between the baseline sample and PN sample. Phenylalanine was the only essential amino acid that decreased significantly between the baseline and PN samples, with the median value decreasing by 11% in both groups. There was no effect of treatment group on the relative changes in concentrations of any of the 8 individual essential amino acids.
Nonessential amino acids
The plasma concentration of total nonessential amino acids increased in response to PN by 9% in the control group and by 19% in the glutamine group. This increase was significant in the glutamine group only, in which the increase was accounted for by the significant increase in glutamine concentration. There was no effect of treatment on the relative change in the concentration of total nonessential amino acids. There were significant increases in the concentrations of histidine, cysteine, proline, serine, arginine, glycine, aspartate, glutamate, and citrulline in both groups. There were significant increases in the concentrations of alanine and glutamine in the glutamine group only. Concentrations of tyrosine and asparagine decreased significantly in both groups. There was an effect of treatment on the relative changes in the concentrations of glutamine and tyrosine between the baseline sample and PN sample.
Plasma glutamine and glutamate concentrations
In the infants randomly assigned to receive glutamine, plasma glutamine concentrations increased significantly, by 31% on average, between the baseline sample and PN sample. Plasma glutamine concentrations were unchanged in the control group between the baseline sample and PN sample.
In both groups, most of the infants had an increase in plasma glutamate concentration in response to PN. There was no difference between the groups in the relative change in plasma glutamate concentrations.
Plasma tyrosine concentrations
Plasma tyrosine concentrations decreased significantly in both groups (by 39% in the control group and by 59% in the glutamine group) between the baseline and PN samples. There was also a statistically significant effect of treatment; the glutamine group had a significantly greater decrease in plasma tyrosine than did the control group.
Plasma ammonia concentrations
We measured plasma ammonia concentrations in 110 infants at each of the 2 time points; these results are shown in Figure 1. In the control group, the mean plasma ammonia concentration decreased by 6% (from 68 to 64 µmol/L) between the baseline sample and PN sample. In the glutamine group, there was no significant change in the mean plasma ammonia concentration between the baseline sample and PN sample. Comparison of the 2 groups regarding the relative change in ammonia concentration between the baseline and PN samples (with adjustments for center and birth weight group) showed a significant but clinically inconsequential difference between the groups. There was no difference between groups regarding the number of infants with an ammonia concentration > 100 µmol/L at the time of the PN sample (glutamine group: n = 12/59; control group: n = 6/51; P = 0.23).
FIGURE 1. . Mean (± SD) plasma ammonia concentrations at baseline (open bars) and during parenteral nutrition (PN; solid bars) in the control group (n = 51) and glutamine group (n = 59). The between-group comparison regarding the effect of treatment on the relative change between the baseline sample and PN sample was significant at P = 0.023. The within-group comparison regarding the change from the baseline sample to the PN sample was significant in the control group at P < 0.05.
Comparative data
Amino acid concentrations determined in 4 previously published studies are shown in Table 4. In 3 of the studies, amino acid concentrations were measured in plasma; the subjects were normally growing, breast-fed term infants (8), premature low-birth-weight infants receiving unsupplemented human milk (9), and low-birth-weight infants receiving TrophAmine (7). In the fourth study, amino acid concentrations were measured in cord blood from neonates at 29 wk gestation (10). The mean amino acid concentrations from the current study are shown in Table 4 and represent the values obtained from both treatment groups for most of the amino acids. Data from the individual treatment groups are shown for glutamine and tyrosine, the only 2 amino acids for which there was a significant treatment effect, and for threonine, because the treatment effect was nearly significant (P = 0.055).
View this table:
TABLE 4 . Amino acid concentrations of infants in the current study and of 4 comparative groups1
DISCUSSION
In the present study, we evaluated the effect of parenteral glutamine supplementation on plasma amino acid and ammonia concentrations in ELBW infants. The amount of glutamine supplementation in the current trial, 20% of the total amino acid intake, is similar to that administered in a study by Lacey et al (11). The latter is the only other published study in which parenteral glutamine was given to premature infants and their glutamine concentrations increased in response to the glutamine-supplemented PN.
In the current study, at the time of the PN sample, mean plasma glutamine concentrations were 30% higher in the infants who received supplemental glutamine and were similar to those measured in the trial by Lacey et al (439 µmol/L) (11). Although an increase in plasma glutamine concentration suggests increased availability of this amino acid, the clinical efficacy of this increase remains to be determined by the primary and secondary clinical outcomes of the main randomized clinical trial.
Currently available intravenous amino acid solutions such as TrophAmine and Aminosyn-PF do provide glutamate. The metabolic interrelationship between glutamine and glutamate necessitates careful monitoring of glutamate as a measure of potential toxicity of glutamine supplementation (12). In our study, there was no significant difference between the 2 groups in the increase in glutamate, providing assurance that the amount of glutamine administered did not cause a significant increase in glutamate, which is potentially neurotoxic. Nonetheless, neurodevelopmental follow-up assessment will also play an important role in further evaluating the safety of glutamine supplementation in this population.
Although plasma ammonia concentrations decreased by 6% in the control group, there was no significant change in plasma ammonia in response to parenteral glutamine supplementation in the glutamine group. In addition, there was no significant difference between the 2 groups in the number of infants with an ammonia concentration > 100 µmol/L during the study PN. We found no evidence of a clinically significant change in ammonia concentration related to glutamine supplementation or to this level of PN support in the ELBW infant during the first 2 wk of age. However, note that ammonia was not measured in all of the infants. In addition, those samples analyzed for ammonia were not analyzed in a central laboratory, nor was the technique for analysis of ammonia standardized.
Currently available parenteral amino acid solutions may not be optimal for ELBW premature infants. One of the most commonly used mixtures for premature infants in the United States, TrophAmine, was formulated to result in plasma amino acid concentrations similar to those of full-term, growing, 1-mo-old breast-fed infants (7, 8). Whether this should be the standard in extremely premature infants is unknown.
Several amino acids are considered to be conditionally essential or indispensable in premature infants. That is, the infant’s ability to synthesize these amino acids de novo is less than adequate to meet functional demands (13). Glutamine is one such amino acid that is potentially indispensable in premature infants, yet is not supplied by standard intravenous amino acid solutions.
Phenylalanine, leucine, isoleucine, valine, threonine, lysine, tryptophan, and methionine are considered essential amino acids in humans. As expected, the concentration of total essential amino acids increased in both groups in response to PN. Despite the fact that 20% of the total essential amino acid supply was replaced by glutamine in the treatment group, there was no significant difference in the magnitude of the increase in total essential amino acids between the 2 study groups.
The decrease in phenylalanine concentration in both groups was unexpected. Plasma concentrations of phenylalanine in both study groups were comparable to those measured by Heird et al (7) in their subgroup of infants who weighed < 1250 g at study entry and received 2 g TrophAmine/kg daily (68.9 ± 16.1 µmol/L). As previously reported by Heird et al (7), this concentration of phenylalanine is higher than that measured in healthy, term infants receiving breast milk. However, previous studies in premature infants generally have not measured the change in amino acid concentrations in response to parenteral amino acid solutions. It is possible that plasma phenylalanine concentrations decrease with increasing postnatal age irrespective of the phenylalanine supply. Alternatively, it is also possible that the combined supply of phenylalanine and tyrosine in currently available amino acid solutions is inadequate, resulting in declines such as that measured in this study. Both animal and human studies of premature infants have suggested that the combined phenylalanine and tyrosine supply may be a limiting factor in protein accretion during the provision of PN (14–17).
Phenylalanine and tyrosine are related in that tyrosine is synthesized endogenously from phenylalanine via phenylalanine hydroxylase. Several investigators have shown that premature infants have the capacity for phenylalanine hydroxylation (14, 18, 19). Consequently, tyrosine is not thought to be an essential amino acid in the classic sense of the definition. Nonetheless, although tyrosine is widely considered to be a conditionally essential amino acid in premature infants, it is not present in appreciable amounts in currently available amino acid solutions because of its low solubility. In some currently available amino acid solutions, tyrosine is supplied as the peptide N-acetyl tyrosine in amounts much lower than the tyrosine content/g protein supplied by enteral formulas and human milk. However, the bioavailability of N-acetyl tyrosine has been questioned (6, 17, 20).
To address the issue of limited tyrosine supply, one potential solution would be to provide an amino acid solution containing sufficient phenylalanine not only for protein anabolism, but also to meet the need for tyrosine (via phenylalanine hydroxylation). However, this strategy has not been effective in animal models (21). In the current study, median tyrosine concentrations while the infants were receiving PN were 39% lower in the control group and 59% lower in the glutamine group compared with the baseline median (P = 0.014 for treatment effect). In both groups, the decline in tyrosine may reflect the limited supply and bioavailability of the tyrosine source provided by the amino acid solution. We speculate that, to the extent that glutamine supplementation increases overall protein anabolism, it may exacerbate specific amino acid deficiencies such as the combined supply of phenylalanine and tyrosine; this may explain the greater decline in tyrosine concentration in the glutamine group.
Relatively little research has been done on plasma amino acid concentrations in ELBW infants. Studies have reported plasma amino acid concentrations in low-birth-weight infants receiving unsupplemented human milk (9) or moderate amounts of TrophAmine (7). To our knowledge, the current study is the largest published collection of data on plasma amino acid concentrations in ELBW infants. Although caution must be used in interpreting these data as normative for ELBW infants, it is somewhat reassuring that the amount of PN in the current study resulted in plasma amino acid concentrations that were not vastly different from those of the comparison groups (Table 4). Although these values reflect current clinical practice, the parenteral amino acid solution provided, and the amount of PN administered, may not have been optimal (22, 23).
Although there is no apparent biochemical risk posed by parenteral glutamine supplementation in ELBW infants, the potential clinical effects of glutamine supplementation remain to be elucidated. In addition, careful neurodevelopmental follow-up of these infants is currently in progress; these results will be critical and are necessary before a recommendation regarding supplementation of PN solution with glutamine can be made.
ACKNOWLEDGMENTS
We are indebted to William C Heird for his help in reviewing the amino acid and ammonia data for the Data Safety and Monitoring Committee, to Edward A Liechty and Larry Auble for their performance of the amino acid analysis, to Bill Buss for his help with the pharmacy procedures, to Lisa Wrage for her assistance with the statistical analyses, and to the medical and nursing staff, parents, and infants in the centers for participating in the study.
For this multicenter trial, each author contributed to protocol development, implementation, and drafting of the manuscript as specified by the publication policy of the National Institute of Child Health and Human Development Neonatal Research Network. None of the authors have any financial or personal interest in any company or organization sponsoring the research.
REFERENCES