Fish oil has become the hot topic in nutrition and health care in general over the past few years. With any therapy that receives inordinate attention from the lay press, though, scientists and health care providers must approach this issue with a healthy dose of professional skepticism. However, contrary to previous “fads”, the benefits of fish oil have a strong foundation in physiological theory and peer-reviewed scientific and clinical studies, with the most notable being efficacy in certain aspects of treating heart disease.1 Fish oil is primarily composed of omega-3 fatty acids, which, along with omega-6 fatty acids, make up the essential fatty acids. These cannot be synthesized by the human body and thus must be derived from exogenous sources. Fish, marine animals, and nuts are particularly rich in omega-3 fatty acids whereas omega-6 fatty acids are concentrated in animal products, vegetable oils, and trans-fatty acids that typify the modern Western diet. Research relevant to liver disease has recently been gaining momentum with very encouraging preliminary studies which are establishing benchmarks in the field. The manuscripts highlighted in this issue of HEPATOLOGY are 2 such studies.

The fat-1 transgenic mouse was first described by Kang et al. in 2004.2 Its ability to convert omega-6 to omega-3 fatty acids provided an elegant model to examine membrane composition effects independent of dietary restrictions. These animals have already been used in inflammatory and neoplastic models,3, 4 the latter by the same group that demonstrated remarkable hepatoprotection in this issue. Schmöcker et al displayed amelioration of both biochemical (ALT) and histological hepatotoxicity in the well-established model of lipopolysaccharide/galactosamine-induced hepatitis and that this effect may be attributable to inflammatory cytokine down-regulation via omega-3 fatty acid content. Indeed, if true, it is likely that this is a global anti-inflammatory effect, further supported by this group's previous work in a murine colitis model.5

However, in a limitation recognized by the authors, the selection parameters of fat-1 mice did not include important omega-3 fatty acids such as alpha-linolenic acid and docosahexaenoic acid (DHA). In addition, measuring biochemical hepatotoxicity with only ALT may be excluding an effect seen in other liver function tests (e.g., AST, bilirubin, alkaline phosphatase) or overgeneralizing a single transaminitis event to whole-organ toxicity. Characterization of cholestasis or the lack thereof in this model with bilirubin measurements might better describe the hepatic response to septic insult. Regarding cytokine level measurement in the liver, mRNA levels do not necessarily correlate to biological activity. Direct measurement, as was done in the plasma measurements, or activity assays are better markers. Also, measurement of apoptosis is better noted by direct markers of DNA damage characteristic of this process, such as terminal deoxynucleotidyl transferase biotin-dUTP nick-end labeling (TUNEL). A comparison to baseline levels of apoptosis in the liver due to normal cell turnover would be appropriate.

It would be interesting to see what effect shifting the omega-6/omega-3 ratio even further toward 1:1 would have in this model. Is there a “dose-dependent” response or is there a “threshold” effect that, once achieved, stays at a plateau? It should also be noted that although the fat-1 animals showed less hepatotoxicity, it still existed to some degree. Analysis of the mechanism of this overall effect would be useful in characterizing not only an omega-3 benefit but also toward understanding inflammatory and septic liver disease.

The injury that results from ischemia and subsequent reperfusion of tissue has long been known to be the source of significantly morbid and often mortal insults in cases of acute embolic/thrombotic events, traumatic wounds, and organ transplantation. As blood flow is restored, the inflammatory cascade potentiates excessive cytokine production and free radical formation, with varying degrees of tissue damage, initiation of apoptosis, and local vascular derangement. Susceptibility is largely dependent on the type of cell and its physiologic milieu.

El-Badry et al. studied both of these factors in a model of liver ischemia and reperfusion, specifically examining normal versus steatotic hepatocytes (and microsteatosis versus macrosteatosis) and omega-3 fatty acid supplemented alteration of the hepatic biochemical environment. When the investigators performed analysis with dynamic intravital fluoromicroscopy, the omega-3 fatty acid–supplemented hepatosteatotic animals, as compared to those fed normal chow, demonstrated decreased microvascular dysfunction, biochemical hepatitis (AST), and inflammatory activity (Kupffer cell) after ischemia and reperfusion. The authors' primary conclusion was that this intervention could dramatically increase the viability of donor organs, especially in countries subsisting primarily on a Western diet and where hepatosteatosis is so prevalent,6, 7 by blunting the inflammatory response to ischemic insult during donor hepatectomy and subsequent reperfusion in the recipient. Other significant findings included overall decrease in intrahepatic lipid content and the apparent vasculature-stabilizing effect of omega-3 fatty acid supplementation.

A possible confounding factor is the failure of standardization of caloric intake. Ad libitum feeding does not account for potential differences in chow smell and taste which might affect intake and hence omega-3 fatty acid levels. This could be solved with a pair-feeding technique to ensure equal caloric intake between compared groups. Similar to the previous paper, El-Badry et al. also did not fully analyze all liver function tests, instead choosing to study only AST. The same drawbacks of possibly excluding an effect in other enzymes and overgeneralizing a whole-organ effect exist here. In addition, the finding of decreased polyunsaturated fatty acids of both omega-6 and omega-3 origin in fatty livers as compared to lean livers remains unexplained and counterintuitive. It is also important to note here that the inherent danger of using a ratio as a measure of efficacy is that it does not take into account total amounts of the numerator and denominator.

The authors make important inferences into the evolution of microsteatosis to macrosteatosis and the role of omega-3 fatty acids in possibly reversing this process. Having already demonstrated microvascular dysfunction in macrosteatotic livers even before insult, the authors show this has significant implications in nonalcoholic fatty liver disease (NAFLD). Several hypotheses have been proposed to explain the pathogenesis of steatosis and steatohepatitis, although to date none have been conclusive. One theory is the “two hit” hypothesis, where the first hit involves the development of hepatic steatosis, rendering the liver more susceptible to a second, as yet undefined, hit that results in more severe liver damage.8 The “second hit” in this case appears to be ischemia and reperfusion, whereas in other situations it might be sepsis, the inflammatory response, and so forth. More study into the mechanism of this process and how omega-3 supplementation might reverse it may have a large clinical impact.

These 2 important papers describe a singular phenomenon where hepatitis, generically liver inflammation, is preventable with omega-3 fatty acids. This has been borne out in the recent literature, both in the basic science and clinical arenas, with evidence of protective omega-3 fatty acid derivatives9 and treatment of neonatal liver injury with parenteral formulations,10 respectively. It is a growing body of work that is garnering more interest and support, due in no small part to the fact that the far-reaching benefits of omega-3 fatty acids are becoming obvious11 and the supporting physiological theory is sound. The unprecedented meteoric explosion of technology in all aspects of daily life in the last century has resulted in modernization and industrialization of food processing, which in turn has resulted in an omega-6/omega-3 fatty acid ratio of 16:1 in modern Western diets, a sharp departure from primitive diets that were closer to 1:1.12 Taking into account downstream products of these membrane components and their effect on the inflammatory response (in simple terms, omega-6 being proinflammatory and omega-3 being anti-inflammatory), it is certain that this derangement has had an effect on human pathology and, more specific and pertinent to this discussion, hepatic pathology.

NAFLD is an indolent disease that may affect up to 30% of individuals subsisting on the modern Western diet.6, 7 As noted by El-Badry et al., the steatotic liver is more susceptible to microvascular dysfunction and inflammation, which supports the disease progression of steatosis to steatohepatitis to fibrosis/cirrhosis and then ultimately to liver failure. Recent evidence suggests that NAFLD patients ingest an excessive amount of omega-6 fatty acids and that this promotes inflammation and conversion of simple NAFLD to nonalcoholic steatohepatitis,13 and more importantly that omega-3 fatty acid supplementation improves biochemical, ultrasonographic, and hemodynamic features of liver steatosis.14 Alcoholic liver disease shows a similar progression of severity, ranging from steatosis to end-stage liver disease. Although in the past it was thought that supplementation with polyunsaturated fatty acids might exacerbate inflammation secondary to increased lipid peroxidation,15 more recent literature suggests that the converse may be true, similar to the data on NAFLD.16 Prospective randomized trials are still necessary in both NAFLD and alcoholic liver disease to make definitive conclusions.

Regarding pediatric hepatic disease, our group developed a novel use of parenteral omega-3 fatty acids in the treatment of parenteral nutrition (PN)–associated liver disease. Some studies have reported that infants with PN-associated liver disease have a mortality rate approaching 100% within a year of diagnosis if they are unable to be weaned off PN or fail to receive liver/small-bowel transplantation.17 We hypothesized that the standard omega-6–based lipid emulsion was the culprit, and our laboratory data supported this, as well as the possibility of an omega-3 emulsion as treatment.18 The previously established timeline for normalization of direct bilirubin levels was 3 to 4 months, and only after full enteral nutrition was achieved and PN discontinued.19 We were able to truncate this course to 2 months, and sometimes even shorter, despite less than 100% enteral nutrition and continuation on PN; concomitant with this change, there was an observed associated decrease in C-reactive protein, a known inflammatory marker.10 The latter directly links our clinical findings with the manuscripts describing down-regulation of inflammatory cytokines, discussed previously.

The inflammatory process is the unifying principle that is pervasively involved in the pathophysiology of cholestasis and liver disease. Although the body of evidence is not yet sufficient to make concrete conclusions, there are suggestions that omega-3 fatty acid supplementation may be helpful in blunting this process.

Sepsis induces cholestasis via endotoxin-mediated release of inflammatory cytokines, which negatively affect cellular biliary transport. Although the infectious focus can not always be prevented, perhaps the sequelae can, namely by treatment with omega-3 fatty acids. Converting cellular membranes to an omega-6/omega-3 ratio closer to 1:1 would decrease the substrate availability necessary for production of inflammatory cytokines and replace them with anti-inflammatory mediators. This is not limited to sepsis-induced cholestasis. Viral hepatitis is often characterized by cholestasis and fibrosis/cirrhosis.20 Omega-3 fatty acids might benefit cholestatic viral hepatitis directly and may be beneficial during acute inflammatory episodes as well. If indeed there is a common pathway to cholestasis via the same inflammatory cytokines, there is reason to believe that infection-related cholestasis could respond well to omega-3 supplementation, similar to the effects we have seen in PN-associated liver disease. However, this all remains to be determined through more basic science and clinical research.

The possibilities of this treatment modality in the realm of liver transplantation could be significant. As noted by El-Badry et al., organ shortage is a major problem in the field, and expanding the donor pool could mean life-saving transplants for more patients. In addition, treating recipients proactively in order to decrease reperfusion injury could have the same effect. Data from our laboratory supports this tenet, having shown that omega-3 fatty acids improve hepatic steatosis in mice and may be used to increase the pool of potential live liver donors that are currently excluded because of the presence of macrovesicular steatosis.21 More recent data in an animal model of small-bowel transplantation suggest the same inhibitory effects on inflammatory cytokines in addition to a possible prevention of graft arteriosclerosis.22

There are a myriad of other hepatic disorders of autoimmune, congenital, and neoplastic etiology that may benefit from supplementation, such as autoimmune hepatitis, primary sclerosing cholangitis, primary biliary cirrhosis, familial cholestasis, biliary atresia, hepatoblastoma, and hepatocellular carcinoma. Aside from the potential benefits to cholestasis and inflammation, omega-3 fatty acids may allow patients to better tolerate noxious therapies, such as hepatic resection, liver transplantation, and hepatotoxic chemotherapy.

The biologically active essential fatty acids of most interest are arachidonic acid (AA) in the omega-6 family and eicosapentaenoic acid (EPA) and DHA in the omega-3 family. These fatty acids play an important role in cell membrane composition which, in turn, influences fluidity and cell surface biochemical signaling, and may serve as natural ligands for certain nuclear receptors that affect gene expression. In addition, AA and EPA are important eicosanoid and prostanoid precursors. AA products include 4-series leukotrienes and 2-series prostaglandins (prostaglandin E2, prostacyclin I2, thromboxane A2), and these synthetic processes are mediated by 5-lipooxygenase and cyclooxygenase enzymes. EPA products include 5-series leukotrienes and 3-series prostaglandins (prostaglandin E3, thromboxane A3), and their synthesis is mediated by the same enzymes. EPA provides the substrate for a different array of lipid mediators, which significantly are less biologically active and, thus, less inflammatory than those derived from AA. More recently, it has been suggested that EPA, along with cyclooxygenase inhibition by aspirin, may result in the by-production of resolvin E1, a possible endogenous anti-inflammatory agent.23 In addition, DHA has similar protective metabolites (protectin D1 and 17S-hydroxy-DHA), which have been studied specifically in liver injury.9 In simplified terms, AA products are thought to be pro-inflammatory mediators and EPA products are “anti-inflammatory,” or rather less pro-inflammatory, which makes their interaction critical, especially considering that AA and EPA are competitive substrates (Fig. 1).11

thumbnail image

Figure 1. Omega-6 and omega-3 fatty acid metabolites.

Download figure to PowerPoint

Another facet of the mechanism of omega-3 fatty acids has already been highlighted by Schmöcker et al., namely cytokine modulation. Tumor necrosis factor alpha (TNF-α) has been linked to hepatotoxicity, sepsis, and the inflammatory response. In animal models, TNF-α infusion results in rapid liver failure and death.24 Hepatocyte exposure to TNF-α initiates intracellular signals that may lead to caspase activation and apoptosis, release of reactive oxygen species, and increased mitochondrial permeability.25–27 Omega-3 fatty acids have been shown here and by others to down-regulate the TNF-α response to lipopolysaccharide insult, as would be seen in sepsis, and this may have a direct role in hepatoprotection.28 The mechanism of action is thought to occur via a NF-κB pathway and is also likely to involve the membrane modulatory effects of supplementation.29

The use of parenteral versus enteral administration of omega-3 fatty acids is an important consideration because of the unknown effect that traversing the enteroportal circulation might have on physiological availability. Theoretically, the “first pass” effect may be beneficial for hepatic treatment, but this is speculative and future studies examining potential impact need to be conducted. The parenteral route deposits a lipid emulsion directly into the systemic circulation without packaging or processing (i.e., chylomicrons). This has the potential to more rapidly turn over cell membranes and decrease omega-6/omega-3 ratios, but again, more study is necessary before firm conclusions can be made.

Currently, there are no formal studies examining a population with low omega-3 fatty acid levels, and the need for such studies is questionable, considering that all people on a modern Western diet fulfill this parameter. Epidemiological disease prevalence and incidence studies on the U.S. population are in effect studying an omega-3–deficient cohort. There have been studies regarding disease prevalence comparing Mediterranean diets versus modern Western diets, but with no specific mention of liver diseases; most of the benefit is found in decreased incidence of cardiovascular pathology.30

The most striking evidence thus far supporting use of this treatment is the apparent dearth of side effects. The theoretical complication of bleeding, due to an antithrombotic effect, has not been seen in animal models or in the clinical setting. Similarly, when dosed appropriately, essential fatty acid deficiency does not occur. In light of this and the likelihood that omega-3 fatty acid supplementation is important in not only liver disease but whole-body health, with widespread literature suggesting beneficial effects in multiple disciplines, the potential applications are numerous. Omega-3 fatty acid supplementation offers an attractive alternative to other therapies, because it possesses both nutritional and therapeutic benefits. Future prospective studies using this promising potential therapy are necessary to confirm preliminary findings in animal studies and early clinical reports.


  1. Top of page
  2. References