STATINS AND MYOTOXICITY:  POTENTIAL MECHANISMS AND

John A. Farmer, M.D. and Guillermo Torre-Amione, M.D., Ph.D.

 

Abstract

            The advent of pharmacologic agents which lower cholesterol by the competitive inhibition of the rate limiting enzyme in cholesterol synthesis (hydroxymethylglutyral coenzyme A reductase) have markedly improved the ability of the clinician to optimize the lipid profile in subjects with dyslipidemia.  Statin therapy has been available for almost 15 years and has been documented to reduce cardiovascular mortality in large-scale controlled clinical trials.  The safety profile of statins has been extensively evaluated and the major clinically significant adverse effects encountered with the use of these agents has been induction of elevation of liver enzymes and muscle toxicity.  The extreme form of muscle toxicity (rhabdomyolysis) has been reported both with statin monotherapy and in combination with other pharmacologic agents.  The precise mechanism of statin induced myotoxicity is unknown but may involve problems associated with intrinsic (albeit poorly defined) properties of the specific drug utilized.  Additionally,  genetic, pharmacokinetic and pharmacodynamic interactions may also be involved.  Rhabdomyolysis with statin monotherapy is rare but may be increased significantly with the coadministration of other agents such as gemfibrozil or inhibitors of the cytochrome P450 enzyme system.  Cerivastatin was determined to be associated with an increased risk of myotoxicity despite inherent pharmacologic properties which would appear to be protective.  Cerivastatin has a dual mechanism of excretion via the utilization of the cytochrome P450 enzyme system (2C9 and 3A4) which theoretically imparted a reduced risk for drug interaction.  However, despite the favorable metabolic pathways, rhabdomyolysis cases were reported with increasing frequency following the release of cerivastatin.  The risk for rhabdomyolysis was especially increased with the combination of cerivastatin and gemfibrozil, which carried a specific contraindication by the United States Federal Drug Administration and subsequently resulted in the removal of cerivastatin from the market.  This review will discuss the clinical spectrum and comparative aspects of statin associated myotoxicity.

I.  Introduction

            The age adjusted mortality for cardiovascular disease has been decreasing in the United States for the past two decades [1].  The precise cause for this encouraging improvement in the prognosis for coronary artery disease is

multifactorial and has been variably attributed to both advancements in the management of acute coronary syndromes and risk factor modification.  Dyslipidemia plays a central role in the atherosclerotic process and was difficult to optimize prior to the advent of inhibitors of hydroxymethylglutaryl coenzyme A reductase (statins).  Statin therapy has been available for 15 years and has been proven to be of clinical benefit in efficacy trials which utilized these agents in both the primary and secondary prevention of atherosclerosis.  Additionally, statin therapy has been demonstrated to improve total mortality in studies which were performed with adequate statistical power.  However, both pharmacologic and nonpharmacologic therapies entail a certain degree of risk and therapeutic interventions are most appropriately utilized with a precise knowledge of a clinically validated benefit to risk ratio.  This review will focus on the muscular toxicity associated with statin use and provides a comparative overview of the various agents.

II.            Mechanisms of Rhabdomyolysis

            Muscle toxicity was one of the first major adverse drug reactions which was associated with statin therapy.  The clinical spectrum of muscle abnormalities associated with the use of statins ranges from a mild clinical syndrome consisting of nonspecific myalgias to life threatening rhabdomyolysis.  Rhabdomyolysis is a clinical syndrome with a variety of potential causes including trauma, infection, toxins, genetic abnormalities and drugs.  The clinical consequences of rhabdomyolysis are secondary to diffuse damage of the myocyte sarcolemmal membranes and subsequent massive cell lysis.  The resultant myonecrosis is characterized by the release of a variety of enzymes  into the circulation including creatine kinase (CK) and aldolase.  Additionally,  potassium, myoglobin, creatinine and other predominantly intracellular constituents also enter the plasma compartment.  The clinical sequelae of muscular breakdown includes myoglobinuria and the potential for irreversible renal failure, complex cardiac arrhythmias and local compartment syndromes due to intracellular fluid shifts.  Creatine kinase elevations and myoglobinuria are frequently utilized for the preliminary diagnosis of rhabdomyolysis.

            The potential risk of developing rhabdomyolysis due to any cause may be at least partially genetically mediated.  Enzyme deficiencies may be associated with increased risk for the development of rhabdomyolysis and include genetically mediated abnormalities in phosphorylase, phosphofructokinase, carnitine palmitoyltransferase and myoadenalate deaminase.  The diagnosis of these rare enzyme abnormalities is difficult and generally requires biopsy evidence of muscle histopathology and the subsequent determination of enzyme activity.  However, approximately 25% of patients with recurrent rhabdomyolysis may have an underlying genetic predisposition [2].

III.            Statins and Rhabdomyolysis

            The precise mechanism by which statin therapy is associated with muscle toxicity is unknown, although intrinsic pharmacologic properties of the various statins and potential interactions with coadministered drugs have been implicated.  The primary mechanism of action of statin therapy is to competitively inhibit the activity of hydroxymethylglutaryl coenzyme A (HMG Co A) reductase which is the rate limiting enzyme in cholesterol synthesis.  Reduction in the activity of this key enzyme also results in a secondary intracellular depletion of a variety of metabolic intermediates which are generated in the process of cholesterol synthesis.  The depletion of metabolic intermediates (e.g., mevolonate, ubiquinone, farnesol, geranylgeraniol) have been postulated to potentially play a role in statin associated myotoxicity [3].  Statins with increased lipophilic properties, which allow increased penetration of the cell, may result in a significant reduction of these metabolic intermediates which are required for the post-translational (isoprenylation) modification of a variety of regulatory proteins.  The mouse C2-C12 myoblast has been utilized as a model to determine the effects of statin therapy on cellular  viability and the potential association with reduced cholesterol synthesis [4].  Pravastatin is relatively hydrophilic and has been demonstrated to penetrate striated muscle cells poorly.  The administration of pravastatin at a dose level of 200  M/l had little effect on the synthesis of  cellular cholesterol levels when compared to the relatively lipophilic lovastatin.  Additionally, simvastatin and lovastatin administration decreased the viability of the myoblast by 50%.  However, the results of animal studies utilizing in vitro preparations are difficult to extrapolate to humans.  Human studies are limited in the direct elucidation of the mechanisms involved in  statin associated muscle toxicity.  Determination of circulating levels of a variety of synthetic intermediates in the cholesterol synthetic pathway have been performed.  However, the results do not conclusively determine a direct relationship between statins and rhabdomyolysis.  Ubiquinone (coenzyme Q) is an important cofactor for cellular mitochondrial respiration.  Coenzyme Q-10 has been demonstrated to be a redox link between flavoproteins and the cytochrome system which is required for the synthesis of adenosine triphosphate and is essential in cellular energy production.  Coenzyme Q-10 is lipophilic and distributed in both skeletal muscle and the myocardium and may also play a role in membrane stabilization.  Simvastatin therapy administered to dyslipidemic subjects has been demonstrated to reduce the circulating plasma level of coenzyme Q.  Additionally, the coenzyme Q to cholesterol ratio was also lower when compared to healthy controls [5].  However, pravastatin and atorvastatin differ considerably in tissue penetration but were not demonstrated to significantly alter the circulating levels of coenzyme Q-10 despite significant reductions in lipid levels.  The clinical implications of the statin mediated reduction in plasma levels of coenzyme Q levels are unclear.  The hypothesis that reduced circulating levels of plasma coenzyme Q-10 translate into intramitochondrial abnormalities and cellular dysfunction as a mechanism to increase the risk of  rhabdomyolysis has been investigated in human subjects.  Simvastatin was administered to 19 hypercholesterolemic patients who subsequently underwent direct muscle biopsy following a six-month treatment period [6].  Muscle high energy phosphate determinations and coenzyme Q levels were assayed but were found not to be different from baseline or from healthy control which did not support the hypothesis that the potential for altered isoprenylate synthesis or energy generation in the myocytes represents a significant clinical factor following simvastatin administration in hypercholesterolemic subjects.   Large-scale clinical studies also have demonstrated that clinically significant myopathy following the administration of lipophilic statins is associated with a very low incidence of myopathy.  A recent analysis of the simvastatin megatrials demonstrated an overall incidence of muscular toxicity of 0.025% [7].  Additionally, rhabdomyolysis has been reported with all statins irregardless of lipophilicity and large-scale clinical safety trials which directly compare statins with differing properties relative to tissue penetration have not been performed.

            Considerable interest has been generated concerning the potential role of statin interaction with the cytochrome P450 enzyme system and the risk of muscle toxicity [8].  The cytochrome P450 system is a ubiquitous group of related enzymes that oxidatively modify pharmacologic agents with conversion to a more water soluble form which allows renal excretion.  The cytochrome P450 enzyme system is predominantly localized within the liver and intestinal tract.  The cytochrome enzyme system, which accounts for the metabolism of a significant proportion of clinically  utilized medications, is accounted for by six isoforms (CYP 1A2, 2C9, 2C19, 2D8, 2E1, 3A4).  However, the cytochrome 3A4 isoform accounts for the metabolism of approximately 50% of all commonly used drugs.  Lovastatin, simvastatin and atorvastatin are metabolized by the CYP 3A4 isoform.  Fluvastatin is predominantly metabolized by the CYP 2C9 while cerivastatin has a dual mechanism of excretion which utilizes the CYP 2C9 and 3A4 pathway.  Pravastatin is unique in that the cytochrome P450 system is not utilized and this agent is metabolized by a non-p450 mechanism. 

            The potential for myotoxicity is significantly increased when drugs which are metabolized by the cytochrome P450 3A4 isoform are coadministered with inhibitors of this enzyme system.  Mibefradil, which was developed as a calcium channel blocking agent, significantly inhibited the activity of the 3A4 enzyme system.  The combination of mibefradil and simvastatin was associated with a significant incidence of severe rhabdomyolysis and resulted in the removal from the market of this agent [9].  Additionally, the coadministration of drugs which are both metabolized by the identical cytochrome P450 enzyme system may also result in a potentially adverse drug interaction due to an increase in circulating levels.  However, despite the increase in the amount of data relative to pharmacokinetics,  pharmacodynamics and genetics, the mechanism by which statins cause myopathy is not precisely known which resulted in increased reliance on clinical experience and post-release surveillance studies to determine the relative risk of induced myopathies alone and in combination. 

            The interaction between statins and the fibric acid derivatives has received considerable clinical interest due to the frequent utilization of both classes in subjects with combined hyperlipidemia.  The metabolism of the fibric acid derivatives is complex and the precise pathways are controversial.  The fibrates have been reported to be metabolized by the 3A4 pathway [10].  However,  this is not universally accepted and the exact mechanism involved in the metabolism and subsequent renal excretion of the fibrates is not universally agreed upon.  The coadministration of gemfibrozil and statins which are metabolized by the CYP 3A4 isoform (e.g., simvastatin) have been demonstrated to result in an increase in the total area under the curve for both simvastatin and its active form simvastatin acid [11].  Gemfibrozil has not been demonstrated to inhibit cytochrome 3A4 in vitro and the pharmacologic interaction with statins may be mediated by a mechanism which does not involve the P450 system.  Cerivastatin has a dual mechanism of excretion which involves more than one hepatic cytochrome P450 pathway which had been proposed as a mechanism which presumably would result in a low propensity for drug interaction.  Additionally, fluvastatin which is predominantly metabolized by the 2C9 system may have a different spectrum of drug interactions when  compared to statins which exclusively utilize the 3A4 system.  Pravastatin is unique in the HMG Co A reductase inhibitors in that it does not utilize the cytochrome P450 system for metabolism and may provide a basis for reduced incidence of drug interaction.   Pravastatin is excreted in the bile or by renal mechanisms following the formation of a 3-alpha-hydroxy isomeric metabolite.  However, despite theoretic mechanisms involved in the predisposition for statin induced myotoxicity, the mechanism involving the statin-fibrate interaction is unknown.  The risk of rhabdomyolysis in combination therapy utilizing a fibric acid derivative and a statin ranges from 1-5% and the mechanism is potentially multifactorial [12].  The interaction may be due to pharmacodynamic considerations rather than a predominantly pharmacokinetic mechanism.  Fibric acid derivatives may adversely affect hepatic function resulting in reduced clearance of orally administered statins from the portal circulation with a secondary increase in concentration and risk for rhabdomyolysis.  Additionally, renal insufficiency may be associated with decreased clearance of the fibric acid derivatives which are predominantly eliminated by the kidney and result in an increased risk of rhabdomyolysis.

IV.            Evidence from Clinical Trials

            The lack of a unifying hypothesis which explains the mechanism of a potential for a rare but life threatening complication such as rhabdomyolysis  requires extensive clinical experience to estimate the risk benefit ratio for statin therapy.  The first generation statins (lovastatin, pravastatin, simvastatin) were derived as metabolites from fungal cultures and are similar in chemical structure.  The first generation statins have been extensively evaluated in large-scale clinical endpoint trials which clearly demonstrated a significant reduction in cardiovascular endpoints.  Additionally, the clinical trials provided a large database which permits safety analysis for both common and relatively rare adverse effects (e.g., liver function abnormalities and muscle toxicity).  The five major clinical trials which evaluated the first generation statins enrolled over 29,000 patients and the occurrence of rhabdomyolysis was analyzed as a safety parameter in all studies.  The largest experience is with pravastatin and the cumulative safety results were recently combined in the Prospective Pravastatin Pooling Project (PPPP) which accumulated a total experience in excess of 112,000 patient years of monitored drug exposure by combining the results of the Cholesterol and Recurrent Events (CARE), West of Scotland Coronary Prevention Study (WOSCOPS), and the Long-term Intervention with Pravastatin in Ischemic Disease (LIPID) [13].  The pooled results of these three landmark trials were monitored for myotoxicity and documented that three pravastatin and seven placebo subjects were withdrawn  due to an elevation in CK levels.  Despite the withdrawal from the studies, no cases of rhabdomyolysis were reported in the Pravastatin Pooling Project.  The Scandinavian Simvastatin Survival Study (4S) was the first large-scale statin trial to demonstrate a reduction in total mortality in concert with declines in cardiovascular endpoints [14].  The 4S study compared simvastatin to placebo over a six year period in 4,444 patients in a secondary prevention trial.  Creatine kinase levels were determined every six months on a routine basis and an increase of 10 times above the upper limits of normal occurred in 6 simvastatin subjects although the enzyme level was not associated with a symptom complex compatible with rhabdomyolysis.  Definite muscle toxicity which fit both clinical and enzymatic criteria for rhabdomyolysis occurred in 1 subject who received simvastatin and was reversible following discontinuation of the drug.  Rhabdomyolysis did not occur in the placebo group and 1 patient was withdrawn due to asymptomatic elevations of creatine kinase.  The Air Force/Texas Coronary Atherosclerosis Prevention Study (AFCAPS/TEXCAPS) was a large-scale primary prevention trial which evaluated a cohort of 6,605 subjects with a relatively modest risk factor profile.  The average cholesterol in the AFCAPS/TEXCAPS group was 221 mg/dl (5.71 mmol/L) which was associated with an HDL cholesterol of 36 mg/dl (0.94 mmol/L).  Lovastatin therapy resulted in a decrease in the primary endpoint which was a composite outcome measurement and included first major coronary event (fatal or nonfatal myocardial infarction, unstable angina or sudden cardiac death).  Creatine kinase levels were monitored during the study and an incidence of CK elevations which were in excess of 10 times the upper limit of normal was documented to be identical in the lovastatin and placebo groups (0.6%) [15].  Further safety analysis documented two cases of rhabdomyolysis in patients who received placebo and one case in the group randomized to receive lovastatin.  Interestingly, the single case of rhabdomyolysis which occurred in a patient randomized to the lovastatin group occurred when the subject was off active therapy but was included for statistical purposes since the trial was analyzed on an intent to treat basis [16].  Muscle toxicity did not occur in patients who received lovastatin therapy in combination with drugs which were known inhibitors of the cytochrome P450 3A4 system.  Fluvastatin and atorvastatin have not been studied in similar long-term clinical endpoint trials although safety data is available from the MIRACL and LCAS studies.  The MIRACL trial randomized 3,086 adults to receive atorvastatin or placebo in a trial involving acute ischemic syndromes but was short term in that an evaluation period of 16 weeks was utilized due to ethical reasons.  Myositis was not documented in any patients in this relatively brief trial [17].  The Lipoprotein and Coronary Atherosclerosis Study (LCAS) evaluated fluvastatin in a relatively small (429 subjects) trial over a 2.5 year period [18].  Rhabdomyolysis was not reported in association with fluvastatin therapy.  Thus, large-scale clinical trial with the first generation statins demonstrated a significant reduction in cardiovascular endpoints with a risk of muscle toxicity which was not significantly different from placebo.  Additionally, while not as well studied, the lipophilic synthetic second generation statins (fluvastatin and atorvastatin) have not been associated with significant rates of rhabdomyolysis.

V.        The Special Case of Cerivastatin

            Cerivastatin was developed as a highly potent inhibitor of HMG Co A reductase activity and could be administered in microgram doses.  Cerivastatin is a pure enantiomeric reductase inhibitor which has been evaluated in clinical efficacy in safety studies since 1993 [19].  Cerivastatin utilizes the cytochrome P450 3A4 and 2C8 isozymes and have been advocated as an agent which would be associated with a low propensity for drug interactions [20].  Clinical studies have demonstrated that cerivastatin is totally absorbed following oral administration and subsequently undergoes moderate first pass hepatic metabolism.  Cerivastatin was determined to be exclusively metabolized by the cytochrome P450 enzyme system with the resultant metabolites being cleared by biliary and renal excretion.  However, despite the theoretic benefits imparted by the dual mechanism of excretion, clinical studies demonstrated a potential for drug interactions with cyclosporine, erythromycin and intraconazole.  Pre-marketing clinical trials utilized doses up to 0.4 mg/day and did not indicate increased risk for rhabdomyolysis.  However, surveillance studies which were performed following the release of cerivastatin documented an increased incidence of rhabdomyolysis, especially when cerivastatin was coadministered with gemfibrozil and resulted in the United States Food and Drug Administration issuing a specific contraindication to the combination of cerivastatin and gemfibrozil.  The incidence of rhabdomyolysis continued to increase and a total of 52 deaths were recorded worldwide which resulted in the removal of cerivastatin from the marketplace [21].  The United States accounted for 31 deaths and 12 of the case fatalities involved the combination of gemfibrozil and cerivastatin despite the issuance of clinical warnings concerning the dangers of combination therapy.  Additionally, a significant number of the rhabdomyolysis cases occurred with cerivastatin monotherapy which was employed at a higher dose (0.8 mg/day) which had recently been approved for clinical usage.  Interestingly, a long-term efficacy and safety study which analyzed the effectiveness of the 0.8 mg dose of cerivastatin has been published following the removal of cerivastatin from the marketplace [22].  The investigators analyzed 1,170 subjects over a 52-week period and randomized the individuals to placebo, cerivastatin 0.4 mg/day and cerivastatin 0.8 mg/day.  Following an 8-week trial period, the placebo group was switched to pravastatin 40 mg/day.  Creatine kinase and muscular symptoms were analyzed as a safety endpoint.  Creatine kinase elevations and symptomatology were divided into three categories (CK>5-10x upper limits of normal without symptoms, CK>10x upper limits of normal without symptoms, CK>10x upper limits of normal with symptoms).  Following the 52-week trial period, no patient who received the 8 weeks of placebo followed by 44 weeks of pravastatin therapy had a CK elevation > 10x the upper limits of normal  associated with symptoms of myositis.  Conversely, 8 patients who were dosed for 52 weeks with 0.8 mg/day of cerivastatin had a CK elevation > 10x the upper limits of normal associated with symptoms.  Additionally, a CK elevation >10x the upper limits of normal which was asymptomatic occurred in one of the patients in the pravastatin group and 16 patients who received cerivastatin 0.8 mg/day.  The authors concluded that the long-term use of cerivastatin 0.8 mg/day effectively and safely bring the majority of patients to National Cholesterol Education Program goals, although it would appear that myotoxicity was increased in the cerivastatin group. 

            The fatal rhabdomyolysis cases associated with cerivastatin have prompted the European Medicine Evaluation Agency to undertake the first comprehensive evaluation of statins since their debut into clinical use almost 15 years ago.  Clinical benefits will be correlated with the risk of therapy and an appropriate clinical benefit to risk ratio will be established which should provide important information as to role of the administration of statins in the treatment of patients with either increased risk for the development of symptomatic coronary disease or in the presence of established atherosclerosis.  However, the need for further basic research into the mechanisms involved in statin associated rhabdomyolysis is necessary and the implications of genetic, pharmacodynamic, pharmacokinetic and structure-function relationships of these agents should be encouraged in light of the potential for statin usage to increase from 13 to 36 million recipients in the United States if the recommendations of the Adult Treatment Panel - III of the National Cholesterol Education Program are implemented [23].

VI.            Summary

            Muscle toxicity was one of the first major events associated with the administration of statin therapy.  Rhabdomyolysis with statin monotherapy had a low incidence (<0.1%) although the risk is increased in combination with agents such as gemfibrozil.  The mechanism by which statin therapy alone or in combination is associated with rhabdomyolysis has not been elucidated.  Genetic predisposition may play a role.  However, drug interactions by coadministration of statins with agents that inhibit the cytochrome P450 3A4 enzyme system are associated with a marked increase in overall risk for rhabdomyolysis.  The precise role of genetic, pharmacodynamic, pharmacokinetic and structure function relationships have not been elucidated and will require further basic research directed at unraveling this mechanism in order to insure maximum safety.