Potential for statins in the chemoprevention and management of hepatocellular carcinoma

Authors

  • Amedeo Lonardo,

    Corresponding author
    • Department of Internal Medicine, Endocrinology, Metabolism and Geriatrics, University of Modena and Reggio Emilia and Nocsae Baggiovara, Modena, Italy
    Search for more papers by this author
  • Paola Loria

    1. Department of Internal Medicine, Endocrinology, Metabolism and Geriatrics, University of Modena and Reggio Emilia and Nocsae Baggiovara, Modena, Italy
    Search for more papers by this author

Correspondence

Dr Amedeo Lonardo, Nuovo Ospedale Civile Sant'Agostino Estense (NOCSAE), Via Giardini, Baggiovara Modena 41125, Italy. Email: a.lonardo@libero.it

Abstract

Hepatocellular carcinoma (HCC) is a common, treatment-resistant malignancy with a complex molecular pathogenesis. Statins are a widely used class of cholesterol-lowering drugs with potential anticancer activity.

We reviewed the evidence for a role of statins in primary and secondary chemoprevention of HCC and slowing the course of otherwise incurable primary or recurrent disease. A literature search (key words: Statins, hepatocellular carcinoma) conducted to this end, retrieved 119 references. Here we summarize the history, mechanism of action and cardiovascular use of statins and highlight that statins can affect several pathways implicated in the development of HCC. In vitro and animal studies provide strong evidence for a favorable effect of statins on HCC. However, evidence in humans is conflicting. We discuss in full detail the methodological strengths and pitfalls of published data including three cohort studies suggesting that the use of statins may protect from the development of HCC and of a single trial reporting increased survival in those with advanced HCC randomized to receive statins. A remarkably hepato-safe class of drugs acting on both hepatocyte and endothelial cells, statins also have potentially beneficial effects in lowering portal hypertension. In conclusion, there is strong experimental evidence that statins are beneficial in chemopreventing and slowing the growth of HCC. However, randomized controlled trials are necessary in order to investigate the role of statins in the chemoprevention of HCC and in slowing the course of otherwise incurable disease in humans.

Background and methods

Hepatocellular carcinoma (HCC) is one of the most lethal cancers, and affects many of the world's populations. Various etiologies have been linked to HCC development, the most prominent of which include hemochromatosis, chronic viral hepatitis due to either B or C infection, excess alcohol consumption and aflatoxin-B1-contaminated food.[1] Virtually all cirrhosis-inducing conditions can cause HCC, pointing to important interactions with the host micro-environment.[2] Moreover, the number of multifocal disease stemming from non-cirrhotic disease[3] may be expected to increase as a result of the “explosion” of nonalcoholic fatty liver disease (NAFLD) worldwide and of failure to offer surveillance to patients with clinically occult chronic liver disease developed in the setting of the metabolic syndrome.

The presently available therapeutic weaponry, which includes radical and palliative options,[4] is not applicable to all patients. Therapeutic failures may result from diagnostic delays, particularly in those with underlying non-cirrhotic liver disease, or recurrent HCC in those with poor liver function. In clinical practice, little is usually done for the secondary chemoprevention of disease in those who, already treated for HCC, remain susceptible to disease recurrence elsewhere in their cirrhotic livers.

Based on the observation that the growth of HCC growth critically depends on cholesterol,[5] we discuss the evidence potentially favoring the use of statins in clinical trials aimed at primary and secondary chemoprevention of HCC in those individuals at high risk of developing this condition and slowing the course of otherwise incurable primary or recurrent disease.

A Medline search of the literature conducted on 12 June 2012, (key words: Statins; hepatocellular carcinoma) provided 119 references. Such references, which were all evaluated for potential inclusion, cross-references, and all those references deemed to be relevant by the authors represent the basis of the present review.

What are statins?

Pure cholesterol, the molecular formula of which was established in 1888, was first extracted from gallstones and named cholesterine, namely “solid bile” in ancient Greek.[6] Medical science has progressed from an era when hypercholesterolemia was deemed to be a mere consequence of ageing—and thus atherosclerosis an unpreventable condition—to the present paradigm that atherosclerosis can be prevented through targeting hypercholesterolemia to reduce mortality.[6, 7] This major shift in clinicians' attitude largely results from statins having been made available.

Cholesterol synthesis takes place in four stages:[6]

  • acondensation of three acetate units to form a 6-carbon intermediate, mevalonate;
  • bconversion of mevalonate to activated isoprene units;
  • cpolymerization of six five-carbon isoprene units to form the 30-carbon linear squalene;
  • dcyclization of squalene to form the steroid nucleus, with a further series of changes to produce cholesterol.

The third reaction in the first stages is the committed and rate-limiting step: reduction of Hydroxy-Methyl-Glutaryl-Coenzyme A (HMGCoA) to mevalonate is the major point of regulation on the pathway to cholesterol (Fig. 1).

Figure 1.

Biosynthesis of cholesterol. The diagram shows the site of action of statins along the cholesterol biosynthetic pathway. Statins act on the key regulation step of cholesterol biosynthesis: the reduction of Hydroxy-Methyl-Glutaryl-Coenzyme A (HMGCoA) to mevalonate.[8] Permission obtained to reproduce from Tobert JA Nat Rev Drug Discov 2003. Confirmation number 110191133 Order Date 08/09/2012.

The discovery of statins is due to a substantial extent to the pioneer work by the Japanese researcher Akira Endo influenced by his native rural environment, by the biography of the discoverer of penicillin Alexander Fleming, and by the high rate of heart attacks observed while working in the USA.[6]

In 1985, Brown and Goldstein were awarded the Nobel Prize in Physiology and Medicine for their discoveries on the regulation of cholesterol metabolism[9] and lovastatin received FDA approval to be commercialized in 1987.[6]

Statins (the chemical structure of which is depicted in Fig. 2) have now been tested in many large-scale clinical trials, involving 90 000 subjects who were followed for 5 years.[6] These studies have consistently shown that treatment with statins lowers plasma low-density lipoprotein (LDL) levels by 25–35% and reduces the frequency of heart attacks to the same extent. Statins are deemed to be the largest selling class of drugs currently taken by patients throughout the world.[6]

Figure 2.

Chemical structure of statins. The remarkable similarities existing among the ancestor molecules, compactin and lovastatin and those statins which are commercially available for clinical use.[6] The figure is reused from the Proceedings of the Japan Academy, Ser. B (6), with permission.

How statins may affect the molecular pathogenesis of HCC

General considerations

During disease development, cancer cells acquire multiple key biological capabilities conferring them a competitive survival advantage and culminating in invasion and metastasis.[10] Whether the pathogenesis of HCC is strongly etiology-dependent remains unproven.[11] In any case, fully integrated direct (genome instability leading to mutations occurring in single or multiple oncogenes or tumor suppressor genes) and indirect oncogenic mechanisms (liver inflammation, regeneration and cirrhosis)[10, 12] are likely to concur to a variable extent in the development of disease in the individual patient.

Specific mechanisms

Karyotype abnormalities

Karyotype abnormalities, the morphological hallmark of genetic instability, have been consistently described in human HCC, structural chromosomal abnormalities being found predominantly in the pericentromeric region and in advanced tumors.[13] Key cellular functions are inhibited by statins selectively in various karyotypically abnormal cell types (including colorectal and ovarian cancer cells and human embryonic stem cells, which possess neoplastic-like properties) and this is mediated via a suppression of the stemness pathway.[14, 15]

Lipogenic pathways

Low serum levels of either LDL-[16] or total-cholesterol[5, 17] are major risk factors for HCC suggesting that HCC itself hi-jacks cholesterol away from the bloodstream because its growth is critically cholesterol-dependent.[5] HCC displays perturbed cholesterol metabolism both within mitochondria and in cell membranes.[18] In human HCC, a relatively higher cell membrane cholesterol content contributes to increasing membrane rigidity. This, in turn, alters membrane signal transduction pathways leading to favored cell proliferation.[19] Increased cholesterol levels in mitochondria from either rat or human HCC cells contribute to chemotherapy resistance and cholesterol depletion by inhibition of hydroxymethylglutaryl-CoA reductase enhances sensitivity to chemotherapy.[20]

Cell growth/survival pathways

The proto-oncogene myc (c-myc) codes for a nuclear protein, which controls nucleic acid metabolism and mediates the cellular response to growth factors. The human c-myc gene plays a pivotal role in liver oncogenesis.[21] Truncation of the first exon, which regulates the expression of c-myc, is crucial for tumorigenicity. Given that HMG-CoA reductase is a critical regulator of MYC phosphorylation, activation, and tumorigenic properties, the inhibition of this enzyme by statins may be a useful target for the treatment of MYC-associated HCC. Consistently atorvastatin blocks both MYC phosphorylation and activation and suppresses tumor initiation and growth both in a transgenic model of MYC-induced HCC as well as in cell lines derived from human HCC.[22] The specificity of these findings was proven by showing that the antitumor effects of atorvastatin were blocked by co-administering mevalonate, the product of HMG-CoA reductase.[22]

Cell proliferation, differentiation and angiogenesis

IL-6-STAT3 pathway

As a gender-dependent risk factor for HCC explaining why females are less prone to liver cancer than males,[12, 23] IL-6 is a HCC bio-marker and an ideal molecular target to be aimed at.[24] IL-6 activates the transcription factor STAT3 (signal transducer and activator of transcription 3), an acute-phase response factor, which is next phosphorylated by the receptor associated kinases, and then forms homo- or hetero-dimers that translocate to the cell nucleus where it acts as a transcription activator. STAT-3 directly affects cell proliferation, differentiation[25] and angiogenesis.[26] Thus, the dysregulation of the STAT-3 pathway, which follows HCV infection may take part in HCC development at an early stage of hepatocyte dysplasia.[27] Moreover, STAT3 is a major pathway which mediates signals from IL-6 to the nucleus. At this level, where different genes associated with proliferation and apoptosis are regulated, IL-6 induces cell survival upon drug treatment in HCC cells; a feature that is blunted by inhibition of IL-6/STAT3 pathway.[24] Therefore, it is of major interest that statins reduce IL-6-induced C-reactive protein (CRP) production directly in hepatocytes via inhibition of protein geranylgeranylation.[28] While the potential of STAT-3 as a therapeutic target in different neoplasms has recently been highlighted,[29, 30] evidence that statins might affect STAT3 pathway mainly comes from vascular rather than oncology studies[31, 32] and therefore further research is required.

Apoptosis

Apoptosis is a key mechanism leading to disposal of unwanted, senescent, or damaged cells and therefore plays a major role in cell health and disease.[33] The development and growth of HCC are heralded by overexpression of anti-apoptotic genes permitting cell survival and neoangiogenesis.[33] Thus, strategies aimed at inducing apoptosis might be exploited to manage HCC.[33, 34] In one study simvastatin induced overexpression of the pro-apoptotic gene Bax together with an inhibition of BCL-2, the gene that has the well-known function of protecting cells from apoptosis. Interestingly, the simvastatin-mediated induction of apoptosis occurs selectively in cancer cells but not in normal cells.[35]

Cell migration and metastasis

Rho-dependent pathway is a mechanism promoting cancer cell migration and metastasis.[36, 37] Rho small GTPases, cycle between a guanosine triphosphate (GTP)-bound active and a guanosine diphosphate (GDP)-bound inactive conformation and it is the intracellular GTP/GDP-bound forms ratio that works as molecular switch that controls a wide variety of signal transduction pathways.[38, 39] Once activated, the Rho protein promotes cell motility via assembly of the actin-myosin contractile filaments.[38] Increased expression of RhoC is linked to increased invasion in various cancer types, including HCC, in which it is a marker of ominous prognosis,[40, 41] a risk factor for metastasis and a candidate molecular target for therapy.[42] In reviewing the role of statins in gastrointestinal cancer, Bhuket and Higgins have highlighted that the interaction of prenylated proteins with cell membranes (Fig. 3) is essential for the activity of signaling of the G proteins Ras and Rho, which are involved in cancerigenesis[43] Interestingly, simvastatin treatment inhibits tumor cell growth and adhesion to endothelium in HepG2 and Huh7 cells in a dose-dependent manner, mediated by decreased expression of integrins and ROCK-I.[44]

Figure 3.

Chemical basis for cancer prevention of statins. By inhibiting Hydroxy-Methyl-Glutaryl-Coenzyme A (HMGCoA) Reductase activity, statins block the synthesis of mevalonate, the precursor molecule of cholesterol, ubiquinone and prenyl groups (i.e. farnesyl and geranylgeranyl). Prenylation of proteins is required for their attachment to cell membrane which, in turn, is essential for the signaling of G proteins Ras and Rho.[43] Permission obtained to reproduce from Bhuket Nature Clin Pract Gastro Hepatol 2006. Confirmation number 11019142 Order Date 08/09/2012.

Taken collectively, data summarized in this chapter are the molecular basis accounting for the findings observed both in animal studies and in humans discussed next.

Statins inhibit primary and metastatic HCC: Evidence from in vitro and in vivo experimental studies

Several experimental studies, summarized in Table 1[22, 45-54] support that statins have an inhibitory effect on primary and metastatic HCC. Analysis of these studies pinpoints that the anticancer effect of statins may be more intensive when given pre- and post-implantation of cancer.[45] Such an effect is dose-dependent and more evident in metastases,[46] occurs via increased apoptosis, arrest of the cell cycle,[47] endoplasmic reticulum (ER) stress response, leading to autophagy,[50] and is partly due to the pleiotropic action of statins[55] being mediated by inhibition of synthesis of ubiquinone.[49] Of potential clinical interest, anticancer activity of statins may be potentiated by either enzastaurin,[48] an inhibitor of PKC (deemed to be the receptor protein of tumor-promoting phorbol esters) or celecoxib, a cyclooxygenase-2 (COX-2)-specific inhibitor that blocks the synthesis of prostaglandins from arachidonic acid.[51] Of major importance, further to hepatocytic cells, endothelial cells are an elective target of the action of statins as well.[53] Such an additional cell target accounts for the combined activity against both cancer and portal hypertension shown by statins, which will be discussed in detail in paragraph on gastrointestinal hemorrhage.

Table 1. Anticancer effect of statins. Evidence from experimental studies
Author (ref)Experimental modelFinding
  1. AMPK, adenosine mono-phosphate kinase; ER, endoplasmic reticulum; HCC, hepatocellular carcinoma; PBR, peripheral benzodiazepine receptor; PKC, protein kinase C; TNF, tumor necrosis factor.
Paragh[45]HCC cells were implanted under the left renal capsule of FLF1 hybrid rats. Fluvastatin was administered per os in 0.5, 2 and 20 mg/kg per day doses.Fluvastatin administered pre- and post-tumor implantation showed a more intensive anticancer effect than treatment given post-tumor implantation alone.
Paragh[46]Rats pre-treated with 2 and 20 mg/kg per day fluvastatin, after 6 weeks were inoculated HCC cells under the left renal capsule.Lovastatin had a dose-dependent inhibitory effect on primary and metastatic tumors. The inhibitory effect on growth and pyruvate kinase activity were higher and occurred earlier in metastases than in primary tumors.
Sutter[47]Statin induced apoptosis was characterized by a breakdown of the mitochondrial membrane potential, caspase activation and nuclear degradation. Furthermore, synergistic growth inhibition was obtained by the combination of statins with the PBR ligand FGIN-1–27. PBR ligands induced a downregulation of HMG-CoAR expression which may account for the enhanced sensitivity of HCC cells to statins.Various statins inhibited the proliferation of HCC cells by inducing apoptosis and G1/S cell cycle arrest.
Kim[48]
  • In vitro studies: Huh-7 cells (a well differentiated human HCC cell line) were used.
  • In vivo studies: MH134 cells (a mouse HCC cell line) and 4-week old male C3H/He mice were used
Co-treatment with lovastatin and enzastaurin, a PKC inhibitor, synergistically suppressed HCC cell growth in vitro and in vivo.
Björkhem-Bergman.[49]These authors hypothesized that statins inhibit carcinogenesis via inhibition of ubiquinone synthesis in a rat model for liver cancer.Lovastatin inhibits carcinogenesis despite unaffected cholesterol levels, at least in part, via inhibition of ubiquinone synthesis.
Yang[50]

This study was conducted in various cell lines including

  • human HCC cell lines Hep3B, HepG2, and Huh7;
  • HCT116 wt and p53−/− cells.
  • Murine embryonic fibroblast (MEF) cells.
  • MEF autophagy-related gene 5−/− (Atg5−/−) cells.
Activation of AMPK by atorvastatin enhances p21 expression and ER stress response, leading to autophagy, which promotes the survival of cancer cells.
Gao[51]The in vivo growth inhibitory effects of celecoxib, a cyclo-oxygenase-2 inhibitor, and fluvastatin was investigated in athymic nude mice implanted with HCC cell line BEL-7402 cellsThe combination of celecoxib and fluvastatin enhanced inhibition of tumor growth, induction of apoptosis, inhibition of tumor cell proliferation, and inhibition of tumor angiogenesis compared with either treatment alone. The combination of celecoxib and fluvastatin also increased levels of the cyclin-dependent kinase inhibitor p21 (Waf1/Cip1), decreased levels of p-Akt, myeloid cell leukaemia-1 (Mcl-1) and survivin protein, but had no effect on Akt protein levels in tumors.
Cao[22]Transgenic model of MYC-induced HCC and human HCC-derived cell lines.Iinhibition of HMG-CoA reductase suppresses MYC phosphorylation through Rac GTPase.
Shimizu[52]This study evaluated the effects of pitavastatin on the development of diethylnitrosamine (DEN)-induced liver preneoplastic lesions in C57BL/KsJ-db/db (db/db) obese mice.Pitavastatin is effective in inhibiting the early phase of obesity-related liver tumorigenesis and, therefore, may be useful in the chemoprevention of liver cancer in obese individuals.
Tijeras-Raballand[53]Male db/db mice were administered tap water containing 40 ppm DEN for 2 weeks and were subsequently fed a diet containing 1 ppm or 10 ppm pitavastatin for 14 weeks.

Rosuvastatin reduced the vessel anomalies and tumor growth and prolonged survival in HCC concurrent with

  • activation of hepatic AMPK,
  • decreased steatosis, free fatty acids, and aminotransferases levels, and the expression of TNF-α and interleukin-6 mRNAs in the liver;
  • increased serum adiponectin levels suggesting attenuation of the chronic inflammation induced by steatosis.
Polo[54]

This study conducted on human hepatocarcinoma cell line (Hep G2) determined

  • the action of geraniol, in combination with simvastatin, and
  • the effect of simvastatin, geraniol and the combination of both on the biosynthesis of lipids from [14C]-acetate.
The combination of simvastatin and geraniol synergistically inhibited cholesterol biosynthesis and proliferation of Hep G2 cell line.

Beneficial effects of statins in either preventing or curing HCC. conflicting evidence from human studies

Owing to the declining prevalence of competing etiologies, HCC is increasingly discovered against a background of metabolic disorders.[5, 56] NAFLD is a major player and provides an essential milieu in the development of HCC in those with metabolic syndrome.[3, 57] Owing to their potential therapeutic indication in NAFLD,[58] statins could hinder the development of HCC in this specific setting in as much as these two conditions are pathogenically linked to one another. Moreover, in theory, statins could exert a beneficial role in the chemoprevention/cure of HCC in those cases of HCC occurring in a viral chronic liver disease as well.

Table 2 details the available evidence for the benefits of statins based on studies conducted in humans.[59-63] There are no large multicentre randomized controlled trials investigating the potential benefits of using statins in patients either susceptible to develop HCC or with advanced and otherwise incurable disease.

Table 2. Beneficial effects of statins in either preventing or curing hepatocellular carcinoma (HCC). Evidence from human studies
Author (Ref)MethodFinding
  1. cDDDs, cumulative defined daily doses; CI, confidence interval; DDD, defined daily dose; HBV, hepatitis B virus; OR, odds ratio; RCT, randomized clinical trial; RR, rate ratio; TAE, transcatheter arterial embolization.
Kawata[59]
  • After withdrawal of eight out of 91 cases with advanced disease, 83 adults with unresectable HCC were enrolled in a RCT.
  • Standard treatment with/without the potent HMG-CoA reductase inhibitor pravastatin, was compared with death as the primary endpoint.
  • All patients underwent TAE followed by oral 5-FU 200 mg/day for 2 months. Patients were then randomly assigned to control (n = 42) and pravastatin 40 mg daily (n = 41) groups for 16.5 ± 9.8 months.
  • Compliance with treatment was assessed by detecting pravastatin in urine.
  • Pravastatin was discontinued in all patients because of advancing disease and not because of any adverse effects of the medicine.
  • Patients in the pravastatin group survived significantly longer than those in the control group (Median survival 18 months vs 9 months in controls p = 0.006).
  • Analysis using the Cox proportional-hazards model showed that treatment with pravastatin was the significant factor contributing to prolonged survival (P = 0.02 for univariate analysis and P = 0.005 for multivariate analysis).
  • This study suggests that pravastatin offers a survival advantage in the treatment of advanced HCC.
Friis[60]
  • This population-based cohort study was conducted in Denmark taking advantage of the Central Population Register, offering valid linkage between cancer and prescription histories for each individual.
  • Within the population of North Jutland county, which has nearly 500 000 inhabitants, 334 754 county residents who were 30–80 years of age during the period 1 January 1989 to 31 December 2002 were identified.
  • The age- and gender-standardized incidence rate of cancer overall was 596 per 100 000 person-years among statin users and 645 per 100 000 person-years among non-users, yielding an age- and gender-standardized incidence RR of 0.92 (95% CI, 0.83–1.02).
  • No statistically significant decreased or increased risks were observed for any of the selected site-specific cancers including liver cancer.
El-Serag[61]
  • This matched case-control nested study was conducted within a cohort of patients with diabetes.
  • Cases comprised incident HCC at least 6 months after entry in the cohort.
  • Controls were identified by incidence density sampling from age and gender matched patients.
  • 1303 cases and 5212 controls were examined. The mean age was 72 years and 99% were men. Only 34.3 % of cases had at least one filled prescription for statins vs. 53.1% of controls (adjusted OR for any statin prescription 0.74; 95% CI 0.64–0.87). In contrast, there were no significant associations between HCC and nonstatin lipid-lowering agents indicating that statins alone were associated with reduced HCC incidence. In order to reduce the potential confounding effect of existing liver disease, analyses were repeated in a subgroup of liver disease-free patients. Such analysis confirmed highly significant ORs for any statin prescription (0.63; 95% CI, 0.50–0.78).
  • This study provides strong evidence that the use of statins is associated with a significant reduction in the risk of HCC in diabetics.
Chiu[62]
  • This is a retrospective population-based case-control study conducted in Taiwan using the local National Health Insurance Research Database.
  • Cases consisted of all ≥ 50 year old patients who had a first-time diagnosis of liver cancer between 2005 and 2008.
  • Age and gender paired controls were matched to cases by index date.
  • Adjusted ORs and 95% CIs (95% CI) were calculated through multiple logistic regression analysis.
  • 1166 liver cancer cases and 1166 controls were examined.
  • Using those individuals not taking statins as the reference group, those who had been prescribed statins below 215.4 defined daily dose (DDD) had ORs 0.62 (95% CI = 0.42–0.91) and those with cumulative statin use of 215.4 DDDs or more had ORs 0.63 (95% CI = 0.37–1.06).
  • This study extends the finding that the use of statins may be associated with a reduced risk of liver cancer to a geographic area where viral hepatitis rather than metabolic disease represents the milieu in which this cancer usually develops.
Tsan[63]
  • This is cohort study was conducted in the general Taiwan population using the local National Health Insurance Research Database as in the previous study by Chiu.[62]
  • The study cohort consisted of 33 413 patients with HBV infection. During the follow-up period, 1021 cases of HCC were observed.
  • Compared to no statin use, defined as < 28 cDDDs, the adjusted HRs for statins users were reduced as follows:
  • 0.66 (95% CI = 0.44–0.99); 0.41 (95% CI = 0.27–0.61), and 0.34 (95% CI = 0.18–0.67) for 28–91; 91–365; and > 365 cDDDs, respectively.
  • The risk of HCC in those with HBV infection was reduced by the use of statins in a dose-response manner in this survey from Taiwan.

A single clinical trial[59] supports the hypothesis that the use of statins might contribute to survival in those with unresectable HCC. This study reports an impressive 9-month longer survival in the pravastatin group. Nevertheless, the level of evidence provided by this study is limited by the low number of the 91 individuals recruited with unresectable HCC, all submitted to trans-arterial embolization (TAE) and 5-FU as a common pre-treatment before randomization which, moreover, was not blinded.[59]

Three population studies[61-63] suggest that statin use might be associated with decreased incidence of HCC.[61-63] Interestingly, such a preventive activity might not be limited to those statins-treated patients with diabetes as suggested by a previous study,[61] but could affect individuals living in an area where liver cancers occur in a viral rather than metabolic milieu.[62] Moreover, the finding of a dose-dependent activity of statins[63] gives further strength to the biological rationale for a putative action of statins in preventing HCC. However, results from these cohorts studies[61-63] need to be interpreted with caution. Despite one of their strengths being based on computerized[61-63] and population based database,[62, 63] the papers by El-Serag,[61] Chiu[62] and Tsan,[63] further to being retrospective, failed to including smoking status and coffee consumption in the propensity score to statins prescription. Smoking has been identified as an independent risk factor for HCC.[64] Given that in a British study statins were given less often to current cigarette smokers than to non-smokers,[65] the seemingly protective effect of statins against HCC might be spurious owing to failure to evaluate perceived hepatological “contraindications” to use of statins and smoking status as potentially confounding factors. Coffee consumption is associated with raised serum cholesterol levels[66] on the one hand and protection from developing HCC[67] on the other hand. Therefore, it is plausible that these two populations of coffee drinkers and statins users overlap at least partially, potentially masking the truly beneficial effect of coffee consumption as a deceptive protection of statins.

Recall bias cannot be ruled out in the paper by Tsan; moreover, this paper was based on a random sampling of those carrying HBV infection, which raises the issue of inadvertent selection bias.[63] In addition, the inclusion of cases of HCC occurring as shortly as 6 months after the entry in the cohort[61] suggests the opportunity to conduct a longer follow-up to reveal changes in slow biological processes such as the development of HCC. The study by Chiu[62] was mismatched as far as risk factors for HCC were concerned, the prevalence of the chief risk factors for HCC (HBV and HCV infection,: alcoholic liver disease and diabetes) being significantly more common among cases than in controls (P < 0.001 for all comparisons).

The seemingly protective effect of statins could well result from a selection bias alone. For instance, given that the use of statins is associated with increased liver enzymes (in a quite small subset of individuals),[68] physicians might nevertheless be more prone to prescribe these drugs to individuals with less elevated liver enzymes, such as, for instance, alcoholic cirrhosis and carriers of HBV with a lower viral load, both populations being per se typically less prone to develop HCC.[69, 70] Not surprisingly, the inclusion of the etiology of HCC into the statistical model attenuated the observed inverse association between the use of statins and HCC.[62] Moreover, pharmacy records of statins prescriptions used in some studies provides evidence for dispensing rather than true statins usage and strict adherence to medical prescription of these drugs.[62] In addition, cohort studies tend to discriminate poorly among the various subtypes of statins,[62] despite the recognized major chemical and biological differences that may occur among the various statins[55, 68] Finally, studies have provided inconsistent results concerning the dose-response relationship protection of HCC exerted by statins.[62, 63]

Not surprisingly, given the number of methodological limitations of the studies discussed above, the results from a recent robust meta-analysis conflict with previous cohort studies reported above and fully agree with a previous population-based Danish study with prospectively registered and virtually complete data on drug prescription and cancer diagnosis.[60] The Cholesterol Treatment Trialists' Collaboration study, collecting data from over 10 000 cases of cancer and over 3500 deaths from cancer among 175 000 randomized patients,[71] aimed at ruling out that statin treatment might be associated with increased risk of cancer, has failed to show any decrease in incidence and mortality for liver cancer.[71] The strength of this study results from it being based on individual patient data, so providing a reliable gauge of the potential association between statins and the development of various cancer types, including HCC, to a significantly larger extent than that offered by previous studies.[71]

In conclusion, on the basis of current evidence, the use of statins in the chemoprevention and treatment of HCC in humans cannot be recommended for clinical practice. However, given that preliminary studies are conflicting, further studies are needed.

Relationship between HCC and gastrointestinal hemorrhage: Potential effects of statins

Portal hypertension commonly occurs in patients with HCC in whom it may precede the development of clinically detectable disease. Gastrointestinal bleeding occurs as the initial manifestation of HCC in 4% of cases and accounts for 15% of mortality in untreated HCC.[72] A recent prospective study has reported portal hypertension to be an independent predictor of HCC development, hepatic venous pressure gradient (HVPG) > 10 mmHg at baseline being associated with a sixfold increase of HCC risk during a 4-year follow-up.[73] Although the mechanistic reasons for such a finding are unclear, both structural and functional changes are likely to be involved. From a pathological point of view capillarization of sinusoids and formation of fibrous septa could all be linked to neoangiogenesis, which might precede the development of HCC.[73] From a physiopathological perspective, intrahepatic shunts, unbalanced ratio of vasoactive substances such as endothelin-1 and nitrates/nitrites and hepatic endothelial dysfunction could well play a role in portal hypertension associated with HCC.[73-75]

It is reported that with currently available treatments as few as 35–40% of the patients achieve target reductions in portal pressure (more than 20% from baseline values or to less than 12 mmHg), which supports the plea for improved treatment schedules.[76] In particular, statins are not included in the presently available strategy for the chemoprevention of either primary or recurrent gastrointestinal bleeding and in clinical practice cirrhotic patients are offered either beta blockers or submitted to rubber band ligation.[77] Physiopathological evidence, however, suggests that statins might in future be used to lower portal hypertension associated with cirrhosis.[78]

Zafra et al. were first in reporting that the administration of simvastatin resulted in decreased hepatic resistance in cirrhotic patients via increased hepatosplanchnic output of nitric oxide products.[79] These authors observed in 13 cirrhotic patients that acute simvastatin administration increased the hepatic blood flow and decreased hepatic sinusoidal resistance without altering HPVG. Such changes were associated with increased nitric oxide product levels in hepatic venous (but not in peripheral) blood. Accordingly, systemic hemodynamics were not modified. In the same study, postprandial portal pressure was evaluated in 17 patients randomized to receive placebo or 40 mg simvastatin 12 h and 1 h before the study. Pretreatment with simvastatin significantly attenuated the postprandial increase in HPVG. Hepatic blood flow increased similarly in the two groups. Hepatic nitric oxide products increased in the simvastatin group but not in the placebo group.[78] The same group of researchers provided further evidence for a beneficial activity of statins in cirrhotic patients in a double-blind clinical trial. Abraldes et al. randomized 59 cirrhotic patients with portal hypertension, defined by HVPG ≥ 12 mm Hg, to either simvastatin (20 mg/day for 1 month [increased to 40 mg/day at day 15]) or placebo. The authors were able to demonstrate that treatment with simvastatin significantly decreased HVPG irrespective of whether patients were receiving beta-adrenergic blockers or not and that treatment was not only free of adverse effects but also associated with surrogate evidence of improved liver perfusion and function.[78] The mechanisms underlying the pharmacological effects of statins on cirrhotic portal hypertension are being increasingly elucidated. Simvastatin selectively increases nitric oxide availability in the cirrhotic liver circulation through increased eNOS expression, Akt-dependent eNOS phosphorylation and cyclic guanosine monophosphate (cGMP) liver content.[80] These effects are mediated in part by increased hepatic levels of the transcription factor Kruppel-like factor 2 (KLF2), the endothelium inducing the expression of a variety of vasoprotector genes/proteins and its vasoprotective target genes, eNOS and thrombomodulin.[81]

Usually studies on portal hypertension are conducted on cirrhotic patients and the presence of HCC is a criterion for exclusion. Therefore, it is unlikely that studies might be conducted specifically in HCC patients and the unproven assumption is that these patients have a response rate similar to that observed in those with cirrhosis. Importantly, future evaluation of statins is needed to use clinical (e.g. effective prevention of bleeding) as opposed to physiopathological end points before these drugs may be allowed to enter the clinical arena.

Which statins should be tested? Concerns for safety of statins in patients with chronic liver disease and cirrhosis

Statins are remarkably hepato-safe agents.[55, 68] Lewis et al. conducted a double-blind randomized controlled trial comparing high dose pravastatin (80 mg daily) to placebo in hypercholesterolemic adults with chronic liver disease.[82] These authors found that while being effective in lowering Total and LDL-cholesterol and triglycerides, pravstatin was not associated with primary pre-specified alanine aminotransferase (ALT) elevations.[82] No differences were registered as a function of the etiology of liver disease, or of the pre-treatment ALT values. In a more recent survey, adverse effects were similar across the statin types for each outcome except liver dysfunction where fluvastatin was associated with the highest risks.[83] This is consistent with the general rule that both the cholesterol-lowering activity and the incidence of aminotransferase elevations are tightly associated with the lipophilicity of ortho-substituents and meta-substituents on the aryl/biphenyl moiety.[55]

Conclusions

By acting on both liver stem cells and endothelial cells, statins might specifically affect some of the main molecular pathways which are implicated in the pathogenesis and biological features of HCC, such as inhibition of cell proliferation, induction of apoptosis and inhibition of angiogenesis. Such effects, which may be relatively selective in cancer cells, result from either inhibited synthesis of cholesterol or pleiotropic activity and may be observed also in advanced primary/metastatic disease. Experimental studies and preclinical observations suggest that statins might prevent/inhibit the development of HCC and portal hypertension. Evidence in humans, however, is much more conflicting, limited and mostly observational. Therefore, there is a strong need for randomized controlled trials for the chemoprevention of HCC in categories of individuals with chronic liver disease at a high risk for HCC. In particular, cirrhotic individuals already submitted to treatment for HCC with liver resection or ablation represent the chief candidates to be recruited in such future studies.

Financial and competing interests disclosure

The Institution of the Authors of this review are recipients of funding from the European Community's Seventh Framework Programme (FP7/2007-2013) under agreement no. HEALTH-F2-2009-241762 for the project FLIP and from Regione Emilia Romagna (PRIER “Molecular Signature of HCC”). The authors have no other relevant affiliations or financial interests with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Ancillary