Hepatoprotective and antiviral functions of silymarin components in hepatitis C virus infection

There exists a widely held view that silymarin (a.k.a. milk thistle) promotes liver health through antioxidant, antiinflammatory, antiproliferative, and immunomodulatory effects.1 In fact, silymarin is one of the top 10 most popular natural products consumed by Western society, and is the most commonly consumed botanical medicine reported in patients with chronic hepatitis C.2, 3 Presently, there is no clear evidence that any of the currently available over-the-counter preparations have efficacy in the treatment of liver disease. While there are compelling in vitro and animal data supporting the hepatoprotective effects of silymarin and inhibition of in vitro HCV infection,4 clinical data are equivocal, with some studies suggesting a protective effect of silymarin against progression of liver disease in subjects with hepatitis C,5 while other studies found no such effect.6, 7 Thus, there is clinical controversy around whether silymarin and silymarin-derived compounds protect the liver. It is the intention of this review article to summarize the current state of knowledge on whether and how silymarin (and the mixture of silymarin components known as silibinin) protects the liver and modulates hepatitis C virus (HCV) infection, and to make recommendations for areas of further research. NASH, nonalcoholic steatohepatitis; PegIFN, pegylated interferon alpha; RBV, ribavirin; SA, silybin A; SB, silybin B; SVR, sustained virologic response; SyNCH, Silymarin for NASH and C Hepatitis. The milk thistle plant originates from the Mediterranean but is now cultivated in Asia and Europe. Silymarin first appears as Silubon in book four of the five volume treatise on medicine known as (Peri Ylis Ialikis; PYI) or De Materia Medica (On Medical Matters). This compendium from the ancient Greek physician Pedanios Dioskurides (latinized as Pedanius Dioscorides, 20-70 CE) was written around 65 CE. While silymarin is sometimes considered part of Traditional Chinese Medicine (TCM), other thistles such as Da Ji (Large Thistle) and Xiao Ji (Thistle Root) are more frequently cited in ancient Chinese medical texts such as Bencao Gangmu (Compendium of Materia Medica). Moreover, Da Ji and Xiao Ji have different chemical compositions than silymarin, which primarily consists of flavonolignans (see below). Silymarin is known as Shui Fei Ji in China, but is not cited in Shen Nong Ben Cao Jing, the earliest Material Medica from China (Jane Saxton, pers. commun.). Thus, silymarin is derived from ancient European medicinal practices. Using the PubMed search term "silymarin" returns over 1,750 publications, the earliest of which date back to a series of German publications from 1968 that focus on the chemical evaluation and hepatoprotective functions of silymarin.8, 9 In 1969, silymarin was shown to protect against toxic mushroom poisoning.10 In 1975, the first reference to silybin dihemisuccinate was made, as a potential antidote for mushroom poisoning.11 Today, this mixture is licensed in Germany for toxic mushroom poisoning, is undergoing a clinical trial in the U.S. for mushroom poisoning (NCT00915681), and has been shown to reduce HCV RNA levels in HCV-infected subjects when administered intravenously.12 In the last 5 years alone, there have been over 700 publications on silymarin indexed on PubMed. The extract and its components display remarkable pleiotropism in biological activities, from growth inhibition of many types of cancer cells,13 to reduction of oxidative stress in multiple cell types including hepatocytes,14 macrophages,15 and neurons,16 to inhibition of many intracellular signal transduction pathways.17, 18 While a plethora of molecular mechanisms have been ascribed to silymarin and its components, no unifying mechanism of action has been forwarded. Silymarin is an extract from the seeds of the milk thistle plant, Silybum marianum L. Gaertn. It is a member of the Asteraceae, a large and widespread family of Angiosperms that include daisies, asters, and sunflowers. The most common name for Silybum marianum is milk thistle or silymarin. However, just like the biological activities ascribed to silymarin, there exist a plethora of names including Bull thistle, cardo blanco, Cardui mariae fructus, Cardui mariae herba, Cardum marianum L., Carduus marianus L., Chardon-Marie, Emetic root, Frauendistel, Fructus Silybi mariae, fruit de chardon Marie, heal thistle, Holy thistle, Kanger, Kocakavkas, kuub, lady's thistle, Marian thistle, mariana mariana, Mariendistel, Marienkrörner, Mary thistle, mild thistle, milk ipecac, pig leaves, royal thistle, S. marianum, St. Mary's thistle, Silybi mariae fructus, snake milk, sow thistle, variegated thistle, Venus thistle, and wild artichoke. An excellent resource is the link found at: http://www.naturalstandard.com/monographs/herbssupplements/milk thistle.asp Silymarin, registered on the Chemical Abstracts Service (CAS) number 84604-20-6, is an extract from the seeds of the milk thistle plant. The major bioactive components consist of seven flavonolignans with the same molecular weight (MW 482) derived from the single flavonoid taxifolin (MW 326). The structure of taxifolin reveals that flavonoids are polyphenolic compounds possessing 15 carbon atoms, with two benzene rings (A and B) joined by a linear three-carbon chain (C) (Fig. 1, left panel), which can be abbreviated C6-C3-C6. Structural basics of silymarin. Left panel shows the basic structure of a flavonoid. The two benzene rings are designated A and B, while the three carbon chain forms ring C. The wavy line that links ring C to ring B indicates the many flavonoids are actually racemic mixtures because the carbon at position 2 on ring C (C2) can be chiral provided there is a single bond between C2 and C3. If there is a double bond between C2 and C3, the flavonoid is not chiral. Right panel shows how taxifolin (a chiral flavonoid) and coniferyl alcohol condense to form silybin A. The blue color-coding shows how conferyl alcohol is incorporated into the larger flavonolignan molecule. As shown in Fig. 1 (right panel), the condensation between taxifolin and coniferyl alcohol has two possible outcomes, thereby generating two different molecules that differ based on the position of the primary alcohol, which can either extend out of the page, or face into the page. Thus, silybin B has the alcohol protruding from the page and the benzene ring faces into the page, while silybin A has the opposite configuration. Since there are two different stereocenters that differ between silybin A and silybin B, they are called diastereoisomers. Silymarin is the entire extract containing seven flavonolignans and taxifolin. Silibinin is a mixture of silybin A and silybin B in a 1:1 ratio. The remaining flavonolignans are: isosilybin A, isosilybin B, silychristin, isosilychristin, and silydianin. The flavonoid found in silymarin is taxifolin. The structures of the eight compounds are shown in Fig. 2. In a typical preparation of silymarin, the approximate percentages of the eight major components are silybin A (16%), silybin B (24%), isosilybin A (6%), isosilybin B (4%), silydianin (16%), silychristin (12%), isosilychristin (2%), and taxifolin (2%).20 The product composition totals are less than 100% because a typical extract is usually labeled as 70% to 80% silymarin, with the remainder consisting of uncharacterized polyphenols and aliphatic fatty acids such as oleic and palmitic acids. Structures and nomenclature of compounds found in silymarin. Please refer to text for details. Silymarin is fat soluble, with an oral bioavailability of 30%-50%. In healthy subjects, oral dosing of silymarin results in low plasma concentrations of flavonolignans between 50-300 ng/mL because the parent compounds are rapidly metabolized to glucuronide and sulfate conjugates which are not thought to have biological activity.21-25 Plasma levels typically achieve maximum concentration after ∼1-2 hours, with an estimated half-life of 4-6 hours.21, 24, 25 Approximately 20%-40% is conjugated and excreted in the bile, 3%-8% excreted in the urine, and the remainder excreted unabsorbed in the feces.26 In cirrhosis patients, the peak serum concentration and time to peak concentration are delayed, suggesting that impaired biliary excretion may be associated with reduced clearance of conjugated silibinin.26 This has also been observed for cirrhosis patients with chronic HCV, who have 3- to 5-fold higher plasma concentrations of flavonolignan conjugates as compared to healthy subjects.6, 27, 28 Silymarin may induce cytochrome P450, isoform 3A429 and p-glycoprotein,30 which may alter the pharmacokinetics of oral contraceptive pills. Metabolism of other drugs may also be affected, including antihistamines, benzodiazepines, protease-inhibitors, and cholesterol-lowering agents. One could argue that in vitro concentrations of silymarin14 are 2-10 times higher than the concentrations achieved in human subjects6 and therefore biological effects of silymarin should never be seen in the clinic. However, prostate cancer subjects given high doses of silipide (a formulation of silibinin with phospholipids) achieved flavonolignan plasma levels of up to 40 μg/mL,31 which are within the biologically active in vitro dose ranges. Moreover, patients with advanced liver disease achieve 3-5 times higher plasma concentrations.28 These data indicate that with oral dosing, plasma (and presumably hepatic) levels of flavonolignans that are equivalent to in vitro doses can be achieved. However, it is clear from the discussion above that orally administered silymarin is not highly bioavailable, which likely contributes to its equivocal effects in clinical trials. Therefore, improvements in oral delivery of silymarin or development of other routes of administration should improve the bioactivity of silymarin in humans. Perhaps the most comprehensive animal safety study on silymarin was the recent publication of the National Toxicology Program report on the toxicology and carcinogenesis of milk thistle extract in rats and mice.32 This study examined the effects of silymarin as part of rat and mouse chow for 3-month and 2-year studies. The key findings of the study were that during the 2-year feed studies, there was no evidence of carcinogenic activity of silymarin in male or female F344/N rats or B6C3F1 mice. Moreover, female rats showed decreased incidences of mammary gland neoplasms, while male mice showed decreased incidences of hepatocellular neoplasms. As an herbal remedy, silymarin is available in various forms and dosages. Therefore, the general public is undoubtedly taking a wide range of doses. Even in clinical studies, silymarin doses have varied widely, ranging from 210 mg through over 1,200 mg daily, with the majority of studies using a daily dose of 420 mg (Table 1). The safety profile of silymarin in published studies has been excellent. In a review of the use of silymarin in the treatment of liver disease, the available safety data using adverse events published in nine clinical trials were summarized.26 Daily doses ranged from 240 mg33 to 600 mg.34 As part of the Silymarin for nonalcoholic steatohepatitis (NASH) and C Hepatitis (SyNCH) phase I studies (clinicaltrials.gov, NCT00389376), subjects receiving as much as 1,680 mg daily for 7 days had similar adverse effects as the placebo group.6 For 296 subjects treated in randomized, double-blinded, placebo-controlled clinical trials, no deaths or serious adverse events were attributed to silymarin. In fact, the incidence of deaths and serious adverse events was lower in the silymarin- than in the placebo-treated population. In these studies, the overall incidence of adverse events was 2.4% of 296 subjects treated in randomized trials. Of 3,612 subjects treated in unblinded studies, the incidence of adverse events was 1%. Considering all published randomized trials, uncontrolled studies, and case reports, only one serious adverse event has been considered related to silymarin (diarrhea, vomiting, and collapse in a 57-year-old woman).35 The most common symptom associated with silymarin use is a laxative effect. Other symptoms have included nausea, epigastric discomfort, arthralgia, pruritis, and urticaria, although in clinical trials the incidence of these side effects is similar between treatment and placebo arms.36 Because silymarin has been reported to decrease bilirubin conjugation and to inhibit the cytochrome P450 enzyme system,37 clinical investigators should be aware of the potential for jaundice or drug interactions. In summary, the various formulations of milk thistle taken as a whole have an excellent safety track record. Nevertheless, particularly at the higher dose ranges, side effects, laboratory parameters, and concomitant medications should be monitored closely. There are compelling data from animal models indicating that silymarin and silymarin-derived compounds protect the liver against injury by a wide array of insults including carbon tetrachloride,38 ischemia-reperfusion,39 the toxic components of death cap mushrooms (Amanita phalloides) phalloidin,10 and alpha-amanitin,40 acetaminophen,41 alcohol,42 and the chemotherapy drug doxorubicin.43 Table 1 includes eight published studies where oral silymarin was administered to patients with chronic hepatitis C. Five of these studies included a placebo control6, 7, 33, 44-46 while three trials included either a multivitamin control group47 or no control group.48, 49 Results are inconsistent among the trials, with three showing alanine aminotransferase (ALT) improvement33, 45, 48 and five demonstrating no effect of silymarin on serum ALT.6, 44, 46, 47, 49 One small trial showed histological improvement in the absence of biochemical response.44 Thus, limited data from published studies in patients with chronic hepatitis C do not uniformly demonstrate hepatoprotection with low to high doses of silymarin. Moreover, the results from the SyNCH trial, which administered the highest oral doses of silymarin to date, were recently published.7 This study used a carefully standardized silymarin preparation, Legalon, available by prescription only, and employed a double-blind, placebo-controlled design. The patients achieved 2-2,000 ng/mL of silymarin flavonolignans and there was no significant change in serum ALT activity or RNA levels in the silymarin treatment arms during the 24-week treatment period.7 As described above, silymarin extract contains silibinin, which is a mixture of the flavonolignans silybin A (SA) and silybin B (SB). Silibinin has antioxidant, immunomodulatory, antiproliferative, antifibrotic, and antiviral activities4, 14, 50 in a wide range of tissues and organs.51 Moreover, both silymarin and silibinin inhibit HCV infection in cell culture by variably blocking viral entry, viral fusion, viral RNA and protein synthesis, HCV N5SB RNA dependent RNA polymerase activity, and virus transmission.4, 14, 52-55 Despite silibinin's promising in vitro activities and demonstrated efficacy in animal models,56 most of the clinical studies in subjects with chronic hepatitis C have administered silymarin, while silibinin has only been tested in four studies.33, 45, 57, 58 An intravenous formulation of silibinin (Silibinin-C-2′, 3-dihydrogen succinate, disodium salt), marketed as Legalon SIL, has been tested in HCV-infected patients. At present, all published data on the use of Legalon SIL are uncontrolled series or case reports in the following three different clinical scenarios. The first report on the clinical use of SIL in chronic hepatitis C demonstrated a dose-dependent decline of HCV RNA over 7 days of daily intravenous infusion in subjects who were prior nonresponders to pegylated interferon alpha (PegIFN) and ribavirin (RBV) therapy. With triple SIL, PegIFN, and RBV therapy, HCV RNA further decreased and became undetectable at week 12 in seven patients who received 15 and 20 mg/kg SIL (50%).12 Treatment with PegIFN/RBV in responders was continued for up to a further 60 weeks. A sustained virologic response was obtained in three patients. This seminal study showed that intravenous silibinin suppresses HCV infection in vivo in patients who failed conventional PegIFN + RBV therapy. In a proof of concept study,59 27 treatment-naïve patients who did not respond to PegIFN/RBV were treated with intravenous silibinin. The majority of patients had the unfavorable IL-28B T-allele (CT = 22; TT = 4). Patients received 20 mg/kg Legalon SIL for either 14 days (n = 12) or 21 days (n = 15), followed by PEG/RBV retreatment. After 7 days of Legalon SIL, 17 (62.9%) patients had undetectable HCV RNA. At the end of intravenous treatment, 23 patients (85.1%) were HCV RNA-negative. After stopping silibinin infusions, HCV RNA returned in five patients, and the viral rebound was associated with baseline viral loads. At the end of the PegIFN/RBV treatment, 17 patients (63%) were HCV RNA-negative. During the 24 weeks of treatment-free follow-up, 12 patients remained HCV RNA-negative (intention-to-treat [ITT] sustained virologic response [SVR] rate: 45%), while five patients experienced virologic relapse (final update).59 Further analysis showed that sustained HCV RNA negativity could only be achieved if HCV RNA became undetectable during silibinin infusions. If HCV RNA persisted after Legalon SIL treatment, no patient went on to achieve SVR. In a recent study, Biermer et al.60 reported on 20 subjects who failed various IFN-based regimens (including four patients receiving triple therapy with a protease inhibitor). The subjects received 1,400 mg/day Legalon SIL on just 2 consecutive days. Complete viral suppression was induced in 13 of 20 subjects within the first week after the short-term silibinin infusion, and all but one of them remained HCV RNA-negative during the subsequent follow-up period during which PegIFN/RBV was administered. Collectively, these studies further demonstrate the clinical efficacy of SIL in patients who fail conventional IFN-based therapies. While intravenous silibinin has been used in the transplant setting in the treatment of Amanita phalloides-induced acute liver failure for over 20 years,61, 62 the mixture has only recently been evaluated in patients with hepatitis C undergoing liver transplantation. In the first case report, Legalon SIL was shown to prevent HCV reinfection of the graft.63 A second case of Legalon-SIL prevention of graft reinfection has also been reported.64 In this case report, cholestatic posttransplant hepatitis was also successfully treated by silibinin infusion. However, intravenous silibinin was unable to prevent graft reinfection in a third patient infected with a genotype 2 virus.65 Finally, Eurich et al.66 reported four patients with graft reinfection with a significant decrease of HCV RNA (mean 2.8 log) after 10 days of intravenous silibinin monotherapy. One patient cleared HCV RNA under silibinin monotherapy while a second patient cleared viremia following additional PEG/RBV therapy. Thus, intravenous silibinin appears to have clinical utility in the prevention of HCV reinfection following liver transplantation. Moreover, a recent study indicates that silibinin is well tolerated and can reduce viral loads in patients waiting for liver transplantation.67 Fortunately, intravenous silibinin appears to be well tolerated. Patients tend to experience a heat sensation and mild sweating during the first infusion. On the first infusion day, approximately half of the patients experience gastrointestinal symptoms including nausea, abdominal pain, and diarrhea. The symptoms appear to be self-limiting and decrease on subsequent treatment days. Following Legalon SIL treatment, significant increases of bilirubin have been reported, with normalization 2 weeks after dosing. The mechanism of the increase in bilirubin is unknown, but data point to the inhibition of anion transporters OATP B1 and OATP B3 by silibinin.67, 68 Finally, SIL has recently been shown to inhibit HCV and, to a lesser extent, HIV-1, in an HCV/HIV coinfected patient.69 Indeed, it has recently been found that SIL inhibits in vitro HIV infection of multiple cell types including peripheral blood mononuclear cells.70 In summary, intravenous silibinin could be an alternative for patients with HCV-associated endstage liver disease who have no further treatment options due to side effects of PegIFN/RBV therapy. Additional systematic studies that explore the length of silibinin treatment, the optimal dose, and the duration of post-silibinin PegIFN/RBV therapy should be performed. Legalon SIL retains many of the same in vitro antiviral properties of the parent silibinin including inhibiting HCV fusion, replication, and production of infectious progeny virus.55 Moreover, both Legalon SIL and silibinin inhibit in vitro NS5B polymerase activity.14, 53-55 Despite clear effects of silibinin on several points of the HCV lifecycle, it is not clear which is the dominant mechanism for virus suppression. Two recent studies shed new light on how silibinin may impinge on HCV. Blaising et al. (submitted) suggest that HCV enters human liver cell cultures primarily by clathrin-mediated endocytosis, and both silibinin and Legalon SIL hinder HCV entry into cells by slowing trafficking through clathrin-coated pits and vesicles. Esser-Nobis et al. (submitted) selected for resistance to Legalon SIL in cell culture and isolated a mutation in the HCV nonstructural 4B (NS4B) protein conferring partial resistance to Legalon SIL treatment. These in vitro results were supported by the identification of distinct mutations affecting highly conserved amino acid residues within NS4B in a liver transplant patient who experienced viral breakthrough while on intravenous silibinin therapy. Transfer of in vivo NS4B mutations into HCV replicons conferred SIL resistance in vitro, and altered the structure of the NS4B-induced membranous web,71 the intracellular site of HCV replication. These new studies add to the evolving story of how silibinin inhibits HCV infection, and also raise important questions. For example, is NS4B a direct viral target of silibinin, or does NS4B resistance to silibinin arise through accessory host cell targets? Given the plethora of ways by which silibinin can modulate cellular functions,13 it is likely that the targets of silibinin action lie within the cell, and that viral resistance is a secondary outcome of this measure. Since HCV-host interactions are required for the HCV lifecycle, silibinin (and other components of silymarin) could impact viral functions including entry, replication, and exit by modulating host cell factors. Moreover, viral resistance to silibinin has been demonstrated and is reflected by mutations in the viral genome, but this does not prove that the viral genome or proteins are direct targets of silibinin. Therefore, the targets of silibinin could be cellular proteins or membranes/lipids. In this regard, the common link between the articles by Blaising et al. and Esser-Nobis et al. is the clear effect of silibinin on cellular processes involving membranes. As listed at ClinicalTrials.org, there are multiple clinical trials on silymarin that are actively recruiting patients. At the University of Maryland, a randomized placebo-controlled trial is under way to evaluate the safety and efficacy of silymarin treatment in patients with acute viral hepatitis (NCT00755950). At the University of North Carolina at Chapel Hill, a study on the combined effects of silymarin and green tea extract in patients with chronic hepatitis C is also enrolling patients (NCT01018615). In Italy, a trial is evaluating Legalon SIL on HCV recurrence in liver transplant recipients (NCT01518933). Legalon SIL is also being evaluated in the U.S. for efficacy against mushroom poisoning (NCT00915681). Even though it appears that orally administered silymarin as a monotherapy has little effect on liver enzymes and viral loads, many individuals still self-prescribe silymarin as a botanical medicine or complementary and alternative medicine practice.2 Clinicians need to be aware that their patients, who may be taking pegylated interferon-alpha (IFNα), ribavirin, and directly acting antiviral (DAA) compounds may also be taking silymarin. Thus, the impact of oral silymarin on current standard of treatment therapies should be considered and even investigated. Moreover, intravenously administered SIL should be further studied as a salvage therapy for previous nonresponders to IFN plus ribavirin therapy, as well as in combination with IFN, ribavirin, and DAA compounds. Further research on SIL therapy in the context of orthotopic liver transplantation is also warranted, as is continued investigation into how silibinin's interactions with cells affect virus infection and replication. It is clear that in vitro and in animal models, silymarin and silymarin-derived pure compounds and mixtures protect cells from injury by numerous agents, in addition to providing cytoprotection against inflammatory sequelae. The recent clinical studies of Fried et al.,7 while sobering, should not detract away from additional research on this interesting class of compounds. Additional randomized clinical trials are required before oral silymarin products can be endorsed as treatments for liver disease. Basic research should continue to define the mechanisms for preventing inflammatory sequelae as well as the cytoprotective mechanisms that are induced by these natural products. In doing so, the cellular targets of silymarin will be identified, which might lead to the design of more selective, potent, and orally deliverable antiinflammatory compounds that could prove clinically useful in liver diseases of both viral and nonviral origins. We thank Nicholas Oberlies (University of North Carolina at Greensboro) and Toni Kline (University of Washington) for assistance with chemical structures and for critical reading of the article; Jane Saxton (Bastyr University) for research on silymarin history; Chia Wang (Virigina Mason Medical Center) and Chihiro Morishima (University of Washington) for initial compilation of silymarin clinical safety and efficacy data.