Comment on:Safety and Efficacy of a Fish-Oil-Based Fat Emulsion in the Treatment of Parenteral Nutrition-Associated Liver Disease
Pediatric Gastroenterology, Hepatology, and Nutrition, Medical College of Wisconsin, Milwaukee, Wisconsin Background: Parenteral nutrition-associated liver disease can be a progressive and fatal entity in children with short-bowel syndrome. Soybean-fat emulsions provided as part of standard parenteral nutrition may contribute to its pathophysiology. Methods: The authors compared safety and efficacy outcomes of a fish-oil-based fat emulsion in 18 infants with short-bowel syndrome who developed cholestasis (serum direct bilirubin level of >2 mg/dL) while receiving soybean emulsions with those from a historical cohort of 21 infants with short-bowel syndrome who also developed cholestasis while receiving soybean emulsions. The primary end point was time to reversal of cholestasis (3 consecutive measurements of serum direct bilirubin level of >2 mg/dL). Results: Among survivors, the median time to reversal of cholestasis was 9.4 and 44.1 weeks in the fish-oil and historical cohorts, respectively. Subjects who received fish-oil-based emulsion experienced reversal of cholestasis 4.8 times faster than those who received soybean emulsions and 6.8 times faster in analysis adjusted for baseline bilirubin concentration, gestational age, and the diagnosis of necrotizing enterocolitis. A total of 2 deaths and 0 liver transplantations were recorded in the fish-oil cohort and 7 deaths and 2 transplantations in the historical cohort. The provision of fish-oil-based fat emulsion was not associated with essential fatty acid deficiency, hypertriglyceridemia, coagulopathy, infections, or growth delay. Conclusions: Parenteral fish-oil-based fat emulsions are safe and may be effective in the treatment of parenteral nutrition-associated liver disease.
Parenteral nutrition (PN) can be a lifesaving intervention, but its prolonged use is associated with hepatic dysfunction that can eventually lead to cirrhosis.1 These complications tend to occur more often in premature infants who have undergone numerous surgical interventions.2 Currently, there is some evidence that the phytosterols or the  -6 fatty
acids that are predominant in standard U.S. parenteral lipid formulations may
play a role in the pathogenesis of parenteral nutrition-associated liver
dysfunction
(PNALD)3,4;
however, the precise etiology of PNALD remains poorly understood and there is
no effective treatment. This article addresses the important issue of PNALD in
infants by evaluating the use of a fish-oil-based fat emulsion vs standard
soybean-based fat emulsion.
The efficacy and safety of a fish-oil-based fat emulsion was evaluated in
this clinical trial with an external historical control. The authors selected
18 infants under a compassionate-use treatment protocol at Children's Hospital
Boston (CHB) who developed cholestasis while being treated with standard
soybean-based fat emulsion, to be treated with a fish-oil-based fat emulsion.
The fish-oil-based emulsion contained predominantly
The primary outcome of the study was the time to reversal of cholestasis,
which was defined as the time to the first of 3 consecutive serum direct
bilirubin measurements of The result of this comparison demonstrated that individuals receiving the fish-oil-based PN had dramatic improvements in serum direct bilirubin levels compared with the historical controls. Specifically, the fish-oil group had 16 (89%) reverse cholestasis during the study period, with 11 (61%) having reversal of cholestasis while on PN. In addition, only 2 (11%) individuals in the fish-oil cohort died during the study period of causes unrelated to treatment. In comparison, the historical controls had a total of 7 (33%) individuals reverse cholestasis during the study period, with no individuals reversing cholestasis while on PN. In addition, 7 (33%) controls died during the study period. Based on these findings, Kaplan Meier estimates demonstrated the mean time to reversal of cholestasis was 9.4 weeks for the fish-oil group and 44.1 weeks for the controls, resulting in a 4.8 times faster reversal of cholestasis for the fish-oil group compared with the historical controls. Regression model analysis demonstrated that baseline serum bilirubin level, gestational age, and history of necrotizing enterocolitis impacted these findings. When these factors were taken into consideration, repeat hazard ratio testing demonstrated that the fish-oil group had an even faster rate of reversal of cholestasis compared with the historical controls. Lastly, safety analysis demonstrated no clinical evidence of fatty acid deficiency (although 1 patient did briefly develop laboratory evidence of EFA deficiency), alterations in coagulation profile, or increase in infections in the fish-oil-treated cohort. From a nutrition perspective, the children in the fish-oil group were given 1 g/kg/day of a lipid emulsion, with the rest of the nonprotein calories provided as carbohydrate. Despite this approach, the majority of children had no evidence of EFA deficiency during the study. Inasmuch as the fish-oil-based lipid emulsion contains only minimal EFA, one must assume that most of these children were receiving sufficient enteral EFA to prevent deficiency; EFA deficiency may still prove to be a concern with long-term use of this emulsion. Because these children were receiving smaller than usual amounts of lipid, they were receiving much larger amounts of dextrose. This should theoretically have been associated with an increased rate of steatosis and worsening liver function, which was not documented in the case patients. Despite these impressive results, several factors must be taken into consideration when interpreting these results. First, although the baseline analysis of both cohorts was not statistically different, this study is not a randomized, controled trial. As a result, unforeseen biases may account for some of the observed differences. For example, the fact that the historical control group contained individuals being treated nearly 2 decades ago may have influenced the number of individuals in the historical control group who reversed cholestasis and/or survived. This was highlighted by Kubota et al in 2000, who demonstrated significant differences in mortality over the past 25 years in neonates requiring PN.5 More important, we are unable to determine the exact cause for the improved outcome of the infants treated with the fish-oil-based lipid emulsion. The relationship between cholestasis and lipid emulsions has been well described in both adults and children.6,7 Cavicchi et al demonstrated in a prospective adult study that limiting the lipid infusion to <1 g/kg/day decreased the likelihood of individuals developing cholestasis when treated with home PN.6 Other studies have demonstrated that cycling or limiting the exposure of PN or lipid emulsion decreases the degree and severity of PNALD.8 Therefore, although this study demonstrates dramatic differences in reversal of PNALD, it is unclear if the infants improved because of the fish-oil-based lipid emulsion, the limited quantity of the lipid emulsion they received compared with controls, or because the soybean-based lipid emulsion was removed. Accordingly, one recommendation would be to repeat the study prospectively and randomly assign individuals to one of 3 different groups. The specific groups would include (1) a standard soybean-based lipid emulsion to act as the control group, (2) a fish-oil-based lipid emulsion treatment group, where each child receives 1 g/kg/day of lipid, and (3) a standard PN solution with either no fat emulsion or soybean-based lipid emulsion limited to 1 g/kg/day. With this study design, the true impact of this fish-oil-based lipid emulsion would become much clearer. Overall, despite the potential biases resulting from study design, this study is the first human clinical study to demonstrate a potential alternative treatment for PNALD. The results of this study describe an unparalleled success in the treatment of a potentially life-threatening disease and eventually may pave the way for making fish-oil-based fat emulsion the standard of care for infants with PNALD.
Contributing Editor—Praveen Goday, MBBS, CNSP Pediatric Gastroenterology and Nutrition, Division of Pediatrics, Medical College of Wisconsin, Milwaukee, Wisconsin
Nutrition in Clinical Practice, Vol. 23, No. 4,
445-446 (2008)
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Top
Comments
2 mg/dL and that the patient was predicted to require PN for >30 days;
children with other forms of liver disease were excluded. These
"case" patients subsequently received 1 g/kg/day of the lipid,
with the rest of the nonprotein calories provided as dextrose. Each case
infant was then followed prospectively for a total of 36 weeks and compared
with 21 historical "control" infants retrospectively identified at
CHB. The inclusion criteria for the control infants required that each infant
be PN dependent for >90 days with 2 consecutive serum direct bilirubin
levels of >2 mg/dL. These infants had received between 1 and 4 g/kg/day of
the soybean-based fat emulsion. 
2 mg/dL. Laboratory values were prospectively
measured in the cases approximately weekly, whereas the historical controls
had all available tests retrospectively recorded. In addition, safety outcomes
including fatty acid levels, coagulation profiles, weight-for-age z
scores, and blood stream infections were prospectively recorded in the case
patients. Statistical analysis was then performed to assess differences
between the 2 groups, and Kaplan Meier survival curves were used to estimate
the time to reversal of cholestasis, whereas hazard ratios were used to assess
the rate of cholestasis reversal between the 2 groups. Lastly, regression
models were developed to identify any additional confounding variables that
may have inadvertently resulted in the differences between the 2 groups.