Correspondence En C. Fung Department of Clinical Biochemistry and Metabolic Medicine, University Hospital Lewisham, London SE13 6LH, UK. Tel.: +44 020 8333 3000, ext.: 6598; Fax: +44 020 8333 3314; E-mail: email@example.com
Introduction: Statin myopathy is a generally encountered side effect of statin usage. Both muscle symptoms and a raised serum creatine kinase (CK) are used in case definition, but these are common manifestations of other conditions, which may not be statin related. Statin rechallenge assuming no contraindication in selected cases is an option before considering a different class of lipid-lowering agent. Aims: We aim to characterize retrospectively the patients referred to our Lipid Clinic with a diagnosis of statin myopathy. The tolerability of different statins was assessed to determine a strategy for rechallenging statins in such patients in the future. Results: Patients with statin myopathy constitute 10.2% of our Lipid Clinic workload. They are predominantly female (62.0%), Caucasian (63.9%), with a mean age of 58.3 years and mean body mass index (BMI) of 29.3 kg/m2. The serum CK and erythrocyte sedimentation rate (ESR) were statistically higher compared to patients with statin intolerances with no muscular component or CK elevations. Secondary causes of statin myopathy were implicated in 2.7% of cases. Following statin myopathy to simvastatin we found no statistical difference between the tolerability rates between atorvastatin, rosuvastatin, pravastatin, and fluvastatin. Fibrates, cholestyramine, and ezetimibe were statistically better tolerated in these patients. Conclusions: Statin rechallenge is a real treatment option in patients with statin myopathy. Detailed history and examination is required to exclude muscle diseases unrelated to statin usage. In patients developing statin myopathy on simvastatin, we did not find any statistical difference between subsequent tolerability rates to rosuvastatin, pravastatin, and fluvastatin.
Statin myopathy generally refers to muscle disease occurring as a result of statin use. It may develop in the form of myalgia which may be accompanied by elevated serum creatine kinase (CK) such as in the case of myositis [1–4]. Left unaddressed, serum CK above 10 times the upper limit of normal (ULN) may be associated with rhabdomyolysis. The prevalence of statin myopathy is around 1.5–3.0% in trial conditions but community-based studies suggest a higher prevalence of up to 33%[1,3]. The commonly cited reason for this discrepancy was a selective inclusion of younger and healthier individuals in statin trials. However, there is generally an underreporting of mild statin myopathy in statin trials, namely myalgias and/or serum CK elevations which are well below the 10 times ULN . These milder forms of statin myopathy are common in clinical practice and may explain to a large degree the discrepancy aforementioned.
Chronic widespread pain arising from musculoskeletal pain is prevalent in the community, occurring in up to 18% of a typical general practitioner (GP) workload [6,7]. It is not surprising myalgia per se is the commonest reason for discontinuing statin . A careful history and examination can help to address common causes such as osteoarthritis, intermittent claudication, nerve impingement, and chronic fatigue syndrome [8,9]. In the absence of a noninvasive diagnostic test for statin myopathy the diagnosis rests firmly on clinical index of suspicion supported by selective laboratory tests. Statin myopathy is thus a diagnosis of exclusion. Where there is doubt about the validity of the initial diagnosis of statin myopathy, a statin rechallenge offers an opportunity for the high cardiovascular risk patient to stay on a statin class drug before non-statin alternatives are considered [4, 10]
Isolated elevations in serum CK do not necessarily preclude statin use. Serum CK is physiologically raised following a recent exercise and it is helpful to note that levels can remain raised for 1 week or more [11,12]. The cut-off for the ULN for serum CK also varies among ethnoracial group such that for African–Carribeans the ULN can be two to three folds that quoted for the Caucasian population [12,13]. This difference was found to be independent to body mass and exercise habits . Clinical trials were generally unhelpful in using a serum CK cut-off of 10 × ULN in case definition of statin myopathy leaving little guidance for the practical management of patients falling short of this cut-off value. Following the above points, it is reasonable to apply a practical cut-off of serum CK five times the ULN [10,14,15] prior to stopping or discontinuing statins and the safety of this notion has been substantiated by a recent prospective case-control trial .
The contemporaneous occurrence of myalgia and/or mild elevations in serum CK (of less than 5 × ULN), may lead to unnecessary withdrawal of statins [17,18]. The aim of this retrospective study was to examine the extent of the role of statin rechallenge in patients with a history of statin myopathy.
The study was conducted in the setting of a Greater London hospital-based Lipid Clinic service and was approved by the local Clinical Audit Department. The clinic was singly managed by a lipidologist for the duration of the audit and the patients are referred from the surrounding GP practices within the area. The lexicon “intolerance” is used to search the electronic patient record—a term used specifically to denote patients experiencing any side effects to their lipid-lowering medications. The list is further hand-searched by the author to select only those fitting the category of statin myopathy. Supplementary clinical information was obtained from the patient physical case file and dietetic records. The study period of this audit was from January 2008 to June 2009.
Demographic data were obtained including cardiovascular risk factors such as smoking and alcohol history, diabetes mellitus, hypertension, and family history of hyperlipidemia. Serial values of serum renal and liver profile, CK, full lipid profile were obtained. In the case of erythrocyte sedimentation rate (ESR), thyroid stimulating hormone (TSH), and CK, the highest recorded values were used for baseline characterization. We contrasted the baseline characteristics of patients with statin myopathy with those intolerant to statins for reasons other than myopathy or raised serum CK, such as nausea and rashes on statins. In so doing, we aim to highlight any characteristics of statin myopathy that set them apart from other causes of intolerance to statins.
For the tolerability profiling, we analyzed all lipid-lowering drugs that were ever taken by the individual patient prior to and during the audit period. We assume the order the drugs were taken bore little effect on subsequent tolerability to others and there was sufficient washout period between each drug. We proceeded to compare the tolerability of different combinations of statins that were challenged to the group during the audit period.
All data were analyzed by SPSS version 15. Nominal data were analyzed by nonparametric student t-test and Mann–Whitney U test and categorical data were analyzed by chi-square analysis and Fisher Exact test where indicated; P < 0.05 was taken as statistically significant.
During the 18 months audit period, there were a total of 1056 attendances in the University Hospital Lewisham Lipid Clinic out of which 208 attendances (19.7%) involved patients unable to tolerate statin for various reasons (n = 163). These statin intolerances include gastrointestinal symptoms such as nausea, deranged liver function tests, erectile dysfunction, rashes, and hair loss. There were a total of 108 patients (10.2%) that finally fitted into the category of statin myopathy (Figure 1). More than half (54.6%, n = 59) were assessed at least twice in the Lipid Clinic.
The clinical characteristics of patients with statin myopathy and statin intolerance without muscle or CK involvement are depicted alongside each other in Table 1. The statin myopathy group consists predominantly of female (62.0%), Caucasian patients (63.9%), with a mean age of 58.3 years and mean BMI of 29.3 kg/m2. A not insignificant proportion (17.6%) had a history of coronary artery disease and the prevalence of common cardiovascular risk factors in this group were as follows: smoking history (43.5%), significant alcohol history above recommended daily units (10.2%), family history of premature ischemic heart disease (7.4%), hypertension (44.4%), and Type 2 diabetes mellitus (19.4%). These features were almost all equally matched by patients with statin intolerance in whom no muscle or CK involvement were found. Statistically the median CK was significantly higher among the statin myopathy group (P= 0.028), as did the ESR (P= 0.035).
Table 1. Patient characteristics in the statin myopathy group in comparison to the statin intolerance group in whom there was no muscle or CK involvement
Statin myopathy n = 108 (66.3%)
Statin intolerance (no myopathy) n = 55 (33.7%)
Age (mean, SD)
Smoking (n, %)
Alcohol (n, %)
History of CHD (n, %)
Family history of premature CHD (n, %)
Diabetes mellitus (n, %)
Hypertension (n, %)
Total cholesterol (mmol/L) (median, range)
Triglyceride (mmol/L) (median, range)
HDL (mmol/L) (median, range)
Creatine kinase (IU/L) (median, range)
ESR (mm/h) (median, range)
TSH (mIU/L) (median, range)
Creatinine (μmol/L) (mean, range)
Simvastatin was implicated in almost all of our patients with statin myopathy (98.1%), commensurating its use as a first-line treatment for hyperlipidemia as recommended by current UK national guidelines. In the initial Lipid Clinic consultation, 17.4% (n = 17) were referred with symptoms of myalgia but mild enough to allow the continued use of Simvastatin (7.1%, n = 7), Atovastatin (4.1%, n = 4), Rosuvastatin (3.1%, n = 3), and Pravastatin (3.1%, n = 3). None of the patients had been prescribed Simvastatin 80 mg daily. In addition none were prescribed combination lipid-lowering therapy either prior to or during the clinic visits. Table 2 showed the frequency of other statin preparations used during rechallenge and their respective statin myopathy recurrence rates. Between 15.0–41.9% were found to tolerate a second statin after simvastatin (Table 3). Interestingly we found no statistical difference between rates of myopathy between atorvastatin, rosuvastatin, pravastatin, and fluvastatin (Figure 2). Similarly, there were also no statistical differences between the rates of myopathy between fibrates, colestyramine, and ezetimibe, although all of these non-statin preparations have a significantly lower myopathy rates than statins (P < 0.05). In the scenarios where a patient was found to have statin myopathy to both simvastatin and atorvastatin, a further 33% were able to tolerate either rosuvastatin or fluvastatin, although none in this group were able to tolerate pravastatin.
Table 2. Amongst our patients with know Simvastatin-induced myopathy, this table shows the proportions developing myopathy on introduction of other lipid-lowering preparations
Number prescribed on (n, %)
Number subsequently developing myopathy (n, %)
Table 3. The proportion of patients developing statin myopathy following different combinations of statin rechallenge;
Notes: (+) occurrence of myopathy; (–) tolerance with no recurrence of myopathy.
aFollowing statin myopathy to simvastatin, the recurrence rates of myopathy following the four other statins were statistically no different.
bFollowing statin myopathy to both simvastatin and atorvastatin, the recurrence rates of myopathy following the remaining three statins were statistically also no different.
(–) 12/34 (35.3%)
(–) 13/31 (41.9%)
(–) 3/20 (15.0%)
(–) 6/19 (31.6%)
(+) 22/34 (64.7%)a
(+) 18/31 (58.1%)a
(+) 17/20 (85.0%)a
(+) 13/19 (68.4%)a
Simvastatin (+) and Atorvastatin (+)
(–) 2/6 (33.3%)
(–) 0/9 (0.0%)
(–) 3/9 (33.3%)
(+) 4/6 (66.7%)b
(+) 9/9 (100%)b
(+) 6/9 (66.7%)b
We were not able to assess the serum CK changes following individual statin rechallenge because of the lack of serial serum CK values. However our records showed that 55.6% (n = 60) had serum CK values within the reference range, 38.9% (n = 42) had the highest serum CK at 1–5 times ULN, 3.7% (n = 4) at 5–10 times ULN, and 1.9% (n = 2) over 10 times ULN. None of these patients had documented evidence of rhabdomyolysis. One patient had a serum CK of 3243 IU/L on simvastatin, which normalized on discontinuation. This patient subsequently chose to remain on dietary and lifestyle management refusing all lipid-lowering agents. The other patient had a serum CK of 2092 IU/L secondary to an undiagnosed hypothyroidism. This last patient together with two patients who were subsequently found to have polymyalgia rheumatica and Type II muscle atrophy, thus making 2.7% (n = 3) of our cohort have a secondary cause to their myopathy other than statin.
This retrospective study has shown that among patients with a history of statin myopathy to simvastatin, the tolerability of subsequent statin rechallenge was similar irrespective of the statin chosen. Simvastatin represents the commonest cause of statin myopathy in our cohort, as it is the first-line therapy for hypercholesterolemia here in the United Kingdom [19,20]. Between 60–80% of these patients subsequently developed statin myopathy to one or more of the other statins.
The similar rates of statin myopathy across statins implied a common pathogenetic mechanism, which supercedes the importance of individual pharmacokinetic characteristics such as lipophilicity and CYP450 metabolism. Statins are competitive, reversible inhibitor of the HMG-CoA reductase. As a result of mevalonate substrate depletion, cholesterol synthesis is reduced but there is also concomitant reductions in the isoprenylation of small guanosine triphosphate (GTP)-binding proteins and ubiquinone (CoQ). Both of these metabolites are involved in mitochondrial function and may account for the development of statin myopathy. Ubiquinone is a key component of the mitochondrial electron transport chain responsible for oxidative ATP generation. Unselective supplementation with oral ubiquinone in symptomatic statin myopathy patients has shown conflicting results in randomized controlled trials with a lack of convincing evidence for its routine use [21–23]. In contrast, evidence of mitochondrial dysfunction was found in a muscle biopsy study showing characteristic abnormal lipid accumulation in cytochrome oxidase stain-negative muscle fibers in symptomatic statin myopathy patients . Furthermore, in a series of mouse in vitro experiments, proapoptotic processes appeared to be induced by statins, which involve atrogin-1 gene overexpression, small GTP-binding proteins such as rap1, and the PI3K/AKT signaling pathways . Apoptotic remnants have been postulated to induce an autoimmune response , and this may ultimately give rise to the experience of muscle aches. The observations that not all patients on statins developed myopathy suggest other susceptibility factors may be at play such as genetic predisposition, preexisting myopathic conditions, and drug interactions. Statins have been known to exacerbate myasthenia gravis , and unmask underlying muscle conditions such as polymyositis  and mitochondrial myopathy .
None of our patients were on combination hypolipidemic therapy as we consider these patients were at increased myopathy risk in view of their previous history of simvastatin-induced myopathy. Statin-fibrates combination therapy is a well-recognized cause of increased myopathy risk [30,31]. However both fibrate [32,33] and ezetimibe [34,35] monotherapy have been independently known to cause myopathy. It is generally accepted that the risk of myopathy to fibrates and ezetimibe is relatively less common compared to statins. In our cohort of patients with a history of statin myopathy, the incidence of fibrate myopathy and ezetimibe myopathy were statistically similar, between 17–31% of cases.
Fibrates are PPARα agonists, which also act at the mitochondrial level by stimulating fatty acid β-oxidation. This is the underlying basis for its beneficial effects on plasma lipids and insulin resistance [36,37]. Although it is unclear the mechanism that fibrates myopathy involves, it is likely mediated by oxidative damage following the increased demand on the mitochondria in fatty acid metabolism. Evidence from a human biopsy study has shown that the oxidatively active, mitochondrial-dense Type 1 muscle fibers are preferentially recruited during exercise with increased levels of PPARα and its cofactor demonstrated . The three genes upregulated during PPARα activation were identified to play a role in muscle lipid catabolism, namely CPT1, MCD, and PDHK4 . PDHK4 is known to inhibit pyruvate dehydrogenase complex (PDC) in turn inhibiting oxidative decarboxylation of pyruvate to acetyl-CoA [38,39]. The relative reduction in the availability of acetyl-CoA for cholesterol synthesis helps to explain its cholesterol-lowering effect . Interestingly in a rat in vitro study, PPARα has been shown to reduce the expression of the NPC1L1 in the intestine . A recent human in vitro study however, demonstrated conflicting result with PPARα positively regulating NPC1L1 transcription . This discrepancy is likely attributable to species differences in PPARα action between rats and humans. Further studies will be required to ascertain the involvement of the NPC1L1 receptor in fibrate action.
Ezetimibe blocks NPC1L1 pathways in intestine and liver thus causing cholesterol depletion, impairing hepatic VLDL secretion, and induction of LDLR gene. Although the mechanism for ezetimibe myopathy is unclear, the similar muscle biopsy findings of these patients to statin myopathy, may also suggest an underlying mitochondrial dysfunction [34,35].
Mitochondrial dysfunction with mitochondrial-mediated apoptosis appears to be the common underlying pathophysiology of myopathy in patients treated with statins, fibrates, and ezetimibe. The higher prevalence of statin myopathy may be related to the unintended reductions in the other products of the mevalonate pathways such as the isoprenylated proteins and N-linked glycosylation. It will be of interest to further characterize these patients with statin myopathy in terms of genetic predisposition such as the SLCO1B1 polymorphism , and detailed investigations for the possibility of underlying muscle pathology. Unfortunately this is beyond the remit and scope of our study.
In the Heart Protection Study (HPS) , up to a third of the placebo group admitted to muscular symptoms during their 5-year follow-up. Muscular symptoms are common in the population and often a contemporaneous use of a statin is taken as proof of causality before appropriate assessment is done. It is possible that a fair proportion of statin and nonstatin myopathy we observed constituted this background level of muscle symptoms in the broader population. Although this poses an intriguing explanation for nonstatin myopathy including the bile acid sequestrant colestyramine, there is insufficient evidence that this is indeed the case, and further investigations are always indicated to exclude an underlying primary muscle disease.
In the absence of a noninvasive diagnostic test for statin myopathy, the assessment is heavily reliant on history, clinical examination, a review of the drug history, and supplementary tests such as serum CK. Although a persistently elevated level of serum CK raises the possibility of statin myopathy it is not a diagnostic test by itself (Table 4). On the other hand, statin myopathy can occur even with a normal serum CK . Muscle biopsy studies have shown intact sarcolemma, which is in keeping with the finding of normal serum CK in such cases [24, 45]. Serum CK is influenced by exercise and ethnoracial differences in resting serum CK levels. To address this, many quote serum CK levels of up to 5 × ULN as a practical limit for safe continuation of a statin [10,14,15]. In our study, 5.6% (n = 6) have serum CK above 5 × ULN. Two of these patients were subsequently found to have untreated hypothyroidism and polymyalgia rheumatica. One patient were assessed and later confirmed to have Type 2 muscle atrophy by the neurologist. None of our patients were given interfering drugs such as the azoles, erythromycin, verapamil, cyclosporin, and red-yeast during their incidence of statin myopathy. Red yeast rice contains monacolins, which are HMG-CoA inhibitors with a modest efficacy in LDL-cholesterol lowering. Up to 10% may develop myalgias with this “natural” product . It is likely to share the same mechanism of myopathy as the statins.
Table 4. Statin myopathy subtypes categorized by the presence or absence of muscle symptoms and/or raised serum CK
Serum CK raised
Serum CK normal
Myalgia: require assessment to ascertain etiology is statin related.
No muscle symptoms
Histological changes described, but of uncertain clinical significance.
Although not assessed during the study period, macro CK is a recognized cause of a falsely elevated serum CK [47,48] and may lead to a misleading diagnosis of statin myopathy with myositis in an otherwise asymptomatic patient. Macro CKs are immunoglobulin-bound CKs which occurred in 1–2% in the population . They are not known to cause any pathology, and to our knowledge there were no studies that investigate the prevalence of macro CKs in statin myopathy patients. A suggested approach to a patient suspected of statin myopathy is shown in Figure 3.
This study has shown that in patients with Simvastatin-induced myopathy, the data suggest nonsuperiority in atorvastatin, rosuvastatin, fluvastatin, or pravastatin in terms of subsequent tolerability on rechallenge. In terms of lipid-lowering efficacy and evidence for cardiovascular outcome benefits [15,50], atorvastatin and rosuvastatin were both reasonable choices. They have an added advantage of long half-lives enabling an alternate day or twice-weekly dosing strategy further minimizing the occurrence of statin myopathy [10,14]. In terms of current cost-effectiveness between the two, atorvastatin stood out as our statin of choice for statin rechallenge. Fibrates or ezetimibe may offer better tolerability but on weighing the strength of evidence on outcome benefit in overall cardiovascular risk reduction, in many circumstances we choose statins over these agents unless there is a history of statin-induced rashes, angioedema, personal or family history of rhabdomyolysis. This is reflected by a greater proportion of our patients who were rechallenged on another statins during the period of the study.
Our audit study is limited by its retrospective nature and the small numbers in each treatment group. We set out examining our experience in managing statin myopathy and seek to identify any areas of best practice. We are not aware of any head-to-head study that compares the relative tolerability of different statin preparations in the setting of statin myopathy. Our study's strengths lie in its inclusiveness of all consecutive patients with statin myopathy and its “real-world” setting in a UK healthcare system. We have demonstrated that a not so insignificant proportion of patients with statin myopathy do tolerate other statins and we feel that this message would stimulate further work on this area of therapeutic dilemma.
One weakness of the retrospective approach in assessing treatment is the lack of systematic approach to treatment allocation a priori. However we would like to highlight some uniformity in our approach to statin rechallenge. The service is singly run by a lipidologist with a uniform approach to statin myopathy. The initial clinical assessment was targeted toward excluding secondary causes of myopathy including drug interactions. Statins were reintroduced at the lowest dosage possible with dose escalation considered in subsequent clinic visits for failure to achieve lipid targets. Only single lipid-lowering agent was introduced in all of these at-risk patients. We cannot exclude physician bias in the treatment allocation but part of this may be influenced by the patient's cardiovascular risk, the degree of hypercholesterolemia, patient comorbidities, and patient choice. A study randomizing the type of statin a patient shall receive will be needed to address this in a more robust manner.
We assumed throughout that the tolerability of a chosen statin was not influenced by the order and the number of different statins previously challenged. As a precautionary step, we have ensured at least 1 month of “wash-out” period elapsed between each change in treatment. Intuitively the potency of HMG-CoA inhibition of a statin should correlate with the incidence of myotoxicity , considering downstream mitochondrial pathways involved. Our study appeared to refute this with statistically similar rates of myopathy observed in all the statins used. A prospective study with larger numbers of patients in each statin group will help to confirm our findings. It will be of interest too to correlate the incidence and severity of statin myopathy with the magnitude of lipid reduction in each case, as suggested in a recent review . Further research is also required to examine the extent at which vitamin D deficiency may play a role in the myalgia/myositis of statin myopathy .
In summary, in patients with a history of statin myopathy to simvastatin, the tolerability of subsequent statin rechallenge was irrespective to the statin chosen. Fibrates and ezetimibe are better tolerated with halved the incidence expected from statins. A careful assessment of the patient's drug history and clinical examination are crucial to exclude common causes of nonstatin causes of myalgias. Statin rechallenge enables some patients to stay on a statin which will help them to achieve lipid targets and ameliorate the risk of cardiovascular disease.