- Meet Consumers/Patients
|Newsletters: HPN-associated Liver Disease in Infants|
HPN-associated Liver Disease in Infants
Kathleen M. Gura, PharmD, Sang Lee, MD, Mark Puder, MD, PhD, Children’s Hospital, Boston
Patients suffering from short bowel syndrome require life-saving parenteral nutrition (PN). One of the major complications of this essential therapy is PN-associated liver disease (PNALD), especially in infants and children. This liver injury affects the majority of infants on long term PN and may progress to liver failure, liver transplantation or ultimately death. Many theories have been brought forward as to the cause(s) of this complication. Here we will discuss the role of lipids and the success we’ve had in treating PNALD in infants.
Patients diagnosed with PNALD typically have elevations in their serum aminotransferases, bilirubin, and alkaline phosphatase. Histologic alterations include steatosis (fatty liver), steatohepatitis (inflammatory response), cholestasis (plugged bile ducts), and in some cases, progression to fibrosis and cirrhosis. Risk factors for PNALD include young age, premature birth, low birth weight, long-term use of PN, absence of enteral nutrition, gastrointestinal mucosal disease, bacterial sepsis, and multiple operative procedures.
A number of causes of PNALD have been proposed including nutrient deficiencies (taurine, choline, vitamin E, zinc and essential fatty acids) and excesses (energy, carbohydrates, amino acids, and fats). Intravenous fat emulsions may predispose patients to PNALD due to their phytosterol content. Phytosterols, such as those contained in soybean oils, are thought to damage the liver.
Care of PN-dependent children is focused on increasing enteral nutrition while the residual bowel increases in size and function so that PN may be discontinued. However, liver injury frequently occurs before bowel adaptation and growth is complete, and irreversible liver damage can pose a serious threat to patient survival. It is often a race between bowel adaptation and growth, and the development of end stage liver disease. Thus, it is essential to prevent, or at least slow, the progression of liver injury for patient survival. In severe cases of refractory hepatic failure, liver transplantation with or without accompanying small bowel transplantation remains the only treatment option.
Treatment for liver disease in children is extremely limited. Often times patients are given ursodeoxycholic acid, hepatotoxic trace elements (manganese, copper) are removed from their PN and/or their PN is supplemented with choline or molybdenum. Some practitioners advocate discontinuing the lipid infusion totally when patients develop signs of PNALD. This practice can be detrimental in infants who are already at risk of developing essential fatty acid deficiency due to their limited fat stores, and need fatty acids for growth and brain development. The most effective treatment is to discontinue PN and begin full enteral feeds. Mortality rates as high as 100 percent have been reported in infants diagnosed with PNALD who are unable to come off of PN within 1 year of diagnosis.
Fatty Acids and their Role in PNALD
Essential fatty acids (EFA) are termed such because they cannot be synthesized by the human body and thus must be derived from the diet. There are two EFA groups: omega-6 and omega-3. They are both characterized as poly-unsaturated fatty acids (PUFA).
Patients need a balance of the products made by both omega 6 and omega 3 fatty acids. Omega 6 (typically given in PN solutions as linoleic acid) breaks down into arachidonic acid, which is essential for bilipid membranes but also produces pro-inflammatory, immuno inhibiting cytokines (see Figure 1). Omega 3 (typically given as a-linolenic acid in PN solutions) breaks down into eicosapentaenoic acid, docosapentaenoic acid and docosahexaenoic acid — which are anti-inflammatory, less immuno inhibiting products. Patients with too much omega 6 product are at risk of developing PNALD, patients with too much omega 3 products are at risk of bleeding disorders.
The metabolic pathways for breaking down Omega 3 fat (a-linolenic acid) to its essential products is impaired in infants, and this leads to a relative excess of omega 6 product. To correct this imbalance, we proposed to give infants the omega 3 products already broken down as they are found in fish oil.
Parenteral lipid emulsions in the United States are soy-based products which are rich in omega-6 fatty acids. Elsewhere in the world, synthesized lipids are available, that provide not only omega 6 but also medium chain triglycerides (similar to MCT oil used in enteral products, such as Pregestimil) and olive oils. Due to regulatory and manufacturing limitations, these products are not available to patients in the U.S. unless their physician petitions the Food and Drug Administration for permission to import them on a compassionate use basis.
There is only one pure fish oil lipid emulsion available for PN use, Omegaven®, and currently it is only approved in Europe, South America and Asia. The manufacturer, Fresenius Kabi, says it is not to be administered as the only lipid product, and when Omegaven is administered with Intralipid®, it is given at a dose of 0.2 g/kg/day. Thus far it hasn’t been released for use in children and patients with liver disease.
We began our studies on liver injury based on a mouse model where animals were allowed to drink PN without lipids. These animals developed fatty livers and elevated liver enzymes. There are at least 2 risk factors for liver injury: a high glucose diet and essential fatty acid deficiency. This, however, would not explain why children develop elevated enzymes in the absence of EFA. We postulated that the route of Intralipid administration may contribute to the problem. We then administered the standard Intralipid emulsion orally, intravenously, and subcutaneously. We found that the group receiving the Intralipid intravenously had severe fatty infiltrates, while the mice receiving the lipid enterally had normal livers.
Based on these findings, we concluded that the PN fat may be a major contributor to PN liver injury. We then repeated the experiments described above, comparing Intralipid to Omegaven head to head. The differences in liver histology were dramatic. Both groups received the lipid solutions intravenously. The Intralipid group had fatty livers while the Omegaven group was normal.
It is quite a leap to go from mouse studies to humans. In most cases, it takes up to 10 years before a laboratory discovery is put into clinical practice. In our case, this leap occurred within months of our initial discovery when one of our 5-month-old patients developed severe PN liver injury that included bridging fibrosis. He was listed for a liver transplant. We were asked by our colleague to try our intravenous fish oil protocol on this patient since it worked so well on the mice. Originally, the goal was to help the child survive until organs for transplant became available. After discussing the nonstandard approach with the parents, treatment was started at 0.2 g/kg/day and increased over a 2 week period to a goal dose of 1 g/kg/day. Within 2 months of starting Omegaven the child’s liver function improved and his jaundice resolved. He was removed from the transplant list shortly thereafter. He is still TPN dependent and on Omegaven two years later, but his bilirubin remains low and is he free from jaundice. (See photo above.)
As Omegaven or any similar product is not available in the United States, compassionate approval from our institutional review board (IRB) and from the FDA is required for each patient. We have treated 22 patients with PN liver injury with very good results. Our criteria for treatment include a direct bilirubin of greater than 2 and no other cause for the liver injury. The patient can have any other medical or surgical problem. The exclusion criteria include active bleeding or allergies to fish or eggs.
Theoretical risks of Omegaven alone include bleeding and EFA deficiency. Based on our previous experience of using Omegaven to treat EFA deficiency in a patient with soy allergy, we knew that if Omegaven is dosed at 1 g/kg/day (5 times the manufacturer’s recommended dose) there is sufficient arachidonic acid and linoleic acid that it can be safely used as monotherapy. Other lipid emulsions do not contain arachidonic acid, thus they require considerably more linoleic acid so that the body can produce it. There have been no complications due to hemostatic changes; to date, there have not been any active cases of bleeding in our patients. Furthermore, most of our patients undergo extensive surgical procedures after starting Omegaven without developing this complication. The EFA profiles are followed weekly and no patients have developed EFA deficiency. In fact, we have used Omegaven to treat EFA deficiency in patients who are unable to receive conventional lipid products such as those with a soy allergy or hypertriglyceridemia.
From our experience, we now use Omegaven at a goal dose of 1 g/kg/day. This is five times higher than the manufacturer’s (Fresenius Kabi) suggested dose of 0.2 g/kg/day and we do not mix it with any other lipid emulsion. Laboratory monitoring includes EFA profiles, hepatic enzymes, lipid panels, C-reactive protein, electrolyte, mineral, albumin/prealbumin prior to the initiation of therapy, and weekly until the bilirubin level normalizes. We also check complete blood counts, coagulation parameters, lipid panels and vitamin status. We then test every 2 weeks, and eventually, monthly. Patients may go home on this treatment and be followed as outpatients.
There have been no complications directly related to this new treatment. Our central line infection rate is lower once on the Omegaven, but may be due to improvement in liver function and close attention to indirect markers of sepsis such as C-reactive protein and platelet counts.
Since this continues to be an investigational therapy, costs of therapy must be born by the sponsoring institution and cannot be passed along to the patient or their insurer. At Children’s Hospital, Boston, the surgeons have funded the costs for their patients. At other centers, unrestricted grants have been obtained through donations.
Despite these dramatic findings, the research does not end here. Our next step is to determine if using intravenous fish oil at the start of PN nutrition may prevent the development of this often fatal complication. The March of Dimes recently awarded us a grant to investigate this, comparing the use of Omegaven to the currently available commercial products. This 3-year randomized controlled clinical trial will be starting within the next several months and will be conducted at Children’s Hospital, Boston.