Intestinal fat digestion requires adequate lipase activity and bile acid concentrations, both of which are low in preterm infants. Bile acids are needed for emulsification of fat before and during lipolysis, and they act as activator for the bile acid–depending lipase in breast milk. Stable solubilization of fat during digestion by bile acids is one of the main prerequisites for sufficient fat absorption (1–3).
Ursodeoxycholic acid (UDCA), a natural bile acid, has been proposed as a treatment option for parenteral nutrition–associated cholestasis (PNAC) in adults, children (4,5), and recently also in preterm infants (6,7). To our knowledge, it has never been used to improve fat absorption in preterm infants. The present study tested the efficacy and safety of UDCA in preterm infants, in terms of its potential impact on fat absorption, advancement of enteral feeding, and the development of PNAC, growth, nutritional status, and metabolic status.
PATIENTS AND METHODS
This prospective, randomized, double-blind, placebo-controlled study consisted of the following periods:
* Period 0: period of randomization.
* Period 1: began on day 3 of life. Infants received total PN and UDCA 5 mg/kg/d or placebo suspension.
* Period 2: began with the initiation of enteral feeding. UDCA 10 mg/kg/d or placebo suspension was given until the last day of PN.
* Period 3: started on the day when full enteral feeding (FEF) was achieved. UDCA 20 mg/kg /d or placebo suspension was given till the last day of the study.
Infants were enrolled and randomly assigned to 1 of the groups—UDCA or placebo treatment (only additives of the suspension)—according to a randomization list. A random permuted block method was used; the block size was 4. For blinding, UDCA and placebo were provided in identical bottles labeled by serial patient number. The intervention began on the third day of life and lasted 4 to 6 weeks.
Preterm infants with birth weight ≥900 g, gestational age ≤ 34 weeks, and requiring total PN during the first days of life were eligible for the study. They were free of major congenital abnormalities, chromosomal aberrations, congenital infections, and severe neonatal diseases. The competent ethical committee (Lombardia Region) approved the study protocol. Informed written consent was obtained from the parents.
All of the infants received PN, starting soon after birth, by use of central venous catheters. Enteral feeding was begun through a gastric feeding tube with human milk (mother's milk or donor milk from the bank). Milk was fortified with human milk fortifier (Eoprotin, Milupa) when the feeding volume reached 90 mL · kg · day. PN was discontinued when enteral feeding volume reached 120 mL · kg · day.
Ursofalk suspension (Dr. Falk Pharma GmbH, Freiburg, Germany) contains the naturally occurring UDCA at a concentration of 50 mg/mL, which has been synthesized from ox bile. UDCA or placebo was administered through a gastric tube in 4 divided doses per day immediately before feedings. Bottles containing the study medication were stored at room temperature not exceeding 25°C.
Changes in fat excretion and time to achieve FEF were considered primary outcome measures. Fat excretion was defined as the percentage of fat in stool collected for 6 hours. Fat in the stool was measured gravimetrically at the end of period 2 and period 3 after lipid extraction by the method of Folch et al (8). FEF day was defined as the time when enteral feeding reached 120 mL · kg · day.
Growth, nutritional status, metabolic status, and markers of cholestasis were considered secondary outcome measures. Body weight, length, and head circumference were measured by experienced nurses. Serum total protein, albumin, calcium, and phosphorus levels were evaluated. Activities of serum alkaline phosphatase, γ-glutamyl transferase (γ-GT), and direct bilirubin concentration were measured.
Complete physical examination was done daily. Adverse events and concomitant diseases were recorded. Blood oxygen saturation, heart rate, and respiratory rate were monitored continuously. Blood gas analysis, serum biochemical indices of renal and hepatic function, serum electrolytes, total blood count, and routine urine analysis were evaluated at study entry and repeated at the third/fourth week (end of period 2) and at the sixth week (end of period 3).
Sample Size and Statistical Analysis
To be clinically relevant, with the UDCA treatment, the fecal fat content should be reduced by 20% and FEF achievement should be shortened by 20% to 25%. Assuming that the fecal fat content would be 20% ± 4% in the control group and the duration of PN would be 7 ± 1.5 days, 15 infants per group were needed to detect such a difference at α = 0.05 with 80% power.
Analysis of variance procedures and t tests were used to compare continuous variables. An exploratory survival analysis was performed for achievement of FEF. A Cox proportional hazard model adjusted for the potential confounder of gestational age was used. Repeated dependent variables were evaluated by repeated measures analysis of variance. Statistical significance was set at the 5% level of probability. Statistical analyses were performed with the SPSS 10.0 program for Windows.
Of 32 enrolled infants, 2 placebo-treated infants left the study (1 Candida infection and 1 transfer to another hospital). The remaining 30 infants completed the protocol. The data from 1 infant in the UDCA group were excluded from analysis, inasmuch as FEF had already been achieved at study entry. For the final analysis, 29 infants (15 in the UDCA group, 14 in the placebo group) were evaluated.
The mean gestational age at birth in the UDCA group was approximately 1 week greater than in the placebo group (31.9 vs 30.9 weeks), resulting in a heavier birth weight (1646.4 vs 1414.5 grams). These differences were not statistically significant. The other demographic characteristics were similar.
No significant differences were found for primary outcome measures. Fat absorption was evaluated indirectly as fat excretion (percentage of fat in stool) at the end of the study periods 2 and 3. The proportion of excreted fat was similar in the 2 groups (46.2% and 42.2% in the UDCA group; 34.5% and 43.8% in the placebo group, at the end of study periods 2 and 3, respectively). Although fecal fat excretion decreased slightly with UDCA treatment and increased in the placebo group, this trend was not statistically significant.
The UDCA-treated infants reached FEF approximately 2 days earlier than did the placebo-treated infants (18.6 ± 5.8 vs 20.4 ± 8.6 days, respectively). When this parameter was adjusted for gestational age, the risk ratio was 1.09, indicating no significant benefit for any treatment. For the secondary outcome measures, growth, nutritional status, and metabolic status were similar in both groups.
Regarding markers of cholestasis, γ-GT activity declined continuously and significantly throughout the intervention period in the UDCA group, and in the placebo group it increased during the PN period and then declined slightly without significance (Fig. 1).
In the UDCA group, aspartate aminotransferase and alanine aminotransferase activities declined significantly (P < 0.05), whereas in the placebo group aspartate aminotransferase activity remained unchanged and alanine aminotransferase activity increased (P > 0.05). Serum direct bilirubin level and serum alkaline phosphatase activity were similar in the 2 groups and over time (values are available from the authors).
Only 1 adverse event was reported in the placebo group: deterioration of clinical condition as a result of systemic Candida infection. Biochemical parameters were normal and similar in the 2 groups.
Major components of intestinal fat digestion (pancreatic lipase and intraduodenal bile acid concentrations) are low in preterm infants (1–3). This pilot study mainly investigated whether UDCA treatment could improve fat absorption in preterm infants, leading to a rapid achievement of FEF. Although fat absorption somewhat improved with UDCA treatment, the difference was not significant. Similarly, with regard to the time of FEF, there was a small but nonsignificant difference in favor of the UDCA group. These findings support that efficient fat absorption in preterm infants depends on alternative mechanisms for dietary fat digestion (3). Indeed, intragastric lipolysis through gastric lipase is of special importance in preterm newborns, accounting for the 25% to 60% of total lipid digestion (3,9). Additionally, products of gastric lipolysis compensate for low bile acid concentrations by emulsifying the lipid mixture (3).
Gastric lipolysis is significantly higher in preterm infants fed with human milk than in formula-fed infants (25% vs 14%). Triglycerides within human milk fat globules are more accessible to gastric lipase than are those in formula (3,9). All of the infants were exclusively fed human milk throughout the study; this also may partly explain the similar fat excretion and FEF achievement in the groups.
In the present study, UDCA proved to be safe and well tolerated in preterm infants. All of the infants in the UDCA group completed the trial without any adverse events. Laboratory and clinical assessments during the study revealed no abnormalities.
PN is an integral part of the care of premature infants, and cholestatic liver disease is a frequent complication of prolonged PN. It has been shown that UDCA treatment alters the course of PNAC in children and adults (4,5). Its mechanism of action includes exchange of hydrophobic for hydrophilic bile acids, leading to an improvement in bile flow (10). Experience with preterm infants is limited; 2 studies (6,7) suggested that UDCA can improve the course of PNAC in preterm infants. To our knowledge, no study has evaluated a potential preventive effect of UDCA on the occurrence of PNAC in preterm infants.
The most interesting finding of this pilot study was the observation of a constant and significant reduction of γ-GT activity over time in the UDCA group. γ-GT is a microsomal enzyme, widely distributed in human tissues involved in secretory and absorptive processes, particularly in the bile canaliculi. γ-GT activity has become a widely used parameter in detecting PNAC and assessing the efficacy of UDCA treatment because it is the earliest sensitive marker of cholestasis (11).
Inasmuch as γ-GT is the early marker of PNAC, this finding may have important implications in practice. The reduction in γ-GT activity with UDCA therapy becomes more important when we consider the relatively short time of PN in our study (approximately 3 weeks). This finding warrants further investigation evaluating the potential prophylactic use of UDCA to prevent cholestasis in infants who receive prolonged PN.
1. Boehm G, Braun W, Moro G, et al
. Bile acid concentrations in serum and duodenal aspirates of healthy preterm infants: effects of gestational and postnatal age. Biol Neonate 1997; 71:207–214.
2. Boehm G, Bierbach U, DelSanto A, et al
. Activities of trypsin and lipase in duodenal aspirates of healthy preterm infants: effects of gestational and postnatal age. Biol Neonate 1995; 67:248–253.
3. Hamosh M. Enteral lipid digestion and absorption. In: Thureen PJ, Hay WW, (eds). Neonatal Nutrition and Metabolism
. 2nd ed. Cambridge, UK: Cambridge University Press; 2006. pp. 350–68.
4. Spagnuolo MI, Iorio R, Vegnente A, et al
. Ursodeoxycholic acid for treatment of inhibition of cholestasis in children on long-term parenteral nutrition: a pilot study. Gastroenterology 1996; 111:716–719.
5. De Marco G, Sordino D, Bruzzese E, et al
. Early treatment with ursodeoxycholic acid for cholestasis in children on parenteral nutrition because of primary intestinal failure. Aliment Pharmacol Ther 2006; 24:387–394.
6. Levine A, Maayan A, Shamir R, et al
. Parenteral nutrition-associated cholestasis in preterm neonates: evaluation of ursodeoxycholic acid treatment. J Pediatr Endocrinol Metab 1999; 12:549–553.
7. Chen CY, Tsao PN, Chen HL, et al
. Ursodeoxycholic acid (UDCA) therapy in very-low-birth-weight infants with parenteral nutrition-associated cholestasis. J Pediatr 2004; 145:317–321.
8. Folch J, Lees M, Sloane-Stanley GH. A simple method for the isolation and purification of total lipids from animal tissue. J Biol Chem 1957; 22:497–509.
9. Armand M, Hamosh M, Mehta NR, et al
. Effect of human milk or formula on gastric function of and fat digestion in the premature infant. Pediatr Res 1996; 40:429–437.
10. Duerksen DR, Van Aerde JE, Gramlich L, et al
. Intravenous ursodeoxycholic acid reduces cholestasis in parenterally fed newborn piglets. Gastroenterology 1996; 11:1111–1117.
11. Cabrera-Abreu JC, Green A. Gamma-glutamyltransferase: value of its measurement in paediatrics. Ann Clin Biochem 2002; 39:22–25.