Tendinous disorders attributed to statins: A study on ninety-six spontaneous reports in the period 1990–2005 and review of the literature
To date, statins have more often been considered a safe medication. However, with the wider use of statins, severe side effects have also been reported to occur in statin-treated patients, especially myositis and rhabdomyolysis. Currently, however, statin-associated tendon impairment has only been described anecdotally. The aim of this retrospective study was to evaluate tendon manifestations occurring in statin-treated patients.
All reports in which a statin was listed spontaneously as a causative suspect medication of tendon complications in the network of the 31 French Pharmacovigilance Centers from 1990–2005 were included in this study. Data collection included patient characteristics and tendon adverse effects (time to onset of adverse effects, pattern, site of injury, and outcome). The percentage of the reports was further calculated for each statin.
Data were collected from 96 patients with a median age of 56 years; patients exhibited tendinitis (n = 63) and tendon rupture (n = 33). Tendinopathy more often occurred within the first year after statin initiation (59%). Tendon manifestations were related to atorvastatin (n = 35), simvastatin (n = 30), pravastatin (n = 21), fluvastatin (n = 5), and rosuvastatin (n = 5). Statin was reinitiated in 7 patients, resulting in recurrence of tendinopathy in all cases.
Our series suggests that statin-attributed tendinous complications are rare, considering the huge number of statin prescriptions. We suggest that prescribers should be aware of tendinous complications related to statins, particularly in risky situations, including physical exertion and association with medications known to increase the toxicity of statins.
Statins (3-hydroxy-3-methylglutaryl-coenzyme A reductase inhibitors) are the most effective therapeutic modalities for reducing serum cholesterol and low-density lipoprotein (LDL) cholesterol levels. Large clinical trials have also demonstrated that statins decrease cardiovascular events in patients with coronary heart disease with LDL cholesterol elevations and in previously healthy patients with high/normal baseline LDL cholesterol values (1–26). To date, statins have more often been considered a safe medication with minor side effects such as headache, mild gastrointestinal manifestations (dyspepsia, constipation), and moderate increase of transaminases (0.2–3.1% of cases) (1–27). However, with the wider use of statins, more severe side effects have also been reported to occur in statin-treated patients, especially musculoskeletal complications such as myositis with muscle weakness and increased creatine phosphokinase (>10 times the upper limit of normal) and rarely rhabdomyolysis as well as inflammatory myopathies (1–32); in previous clinical trials, the incidence of myositis (0.09–0.63%) and rhabdomyolysis (1 case per 100,000 patients yearly) has indeed been estimated to be low (8, 11, 16, 17, 32).
Currently, however, statin-associated tendon impairment has not been reported in pre- and postmarketing studies; statin-associated tendon disorders have, in fact, only been described anecdotally (33–36). These data prompted us to conduct this retrospective study to evaluate tendon complications occurring in statin-treated patients.
PATIENTS AND METHODS
We retrospectively identified all spontaneous cases of statin-attributed tendinous manifestations from the French Pharmacovigilance database. In essence, cases of statin- attributed tendinous complications have been reported by physicians in the network of all 31 French Pharmacovigilance Centers between 1990 and 2005; all of these reports, in which a statin was listed as either a causative suspect or a concomitant medication from January 1, 1990 through December 31, 2005, were included in this analysis. A causal relationship between tendon involvement and statin use was considered probable if symptoms resolved or significantly decreased after discontinuation of the statin.
Inclusion criteria in this study were statin-attributed tendinous manifestations (tendinitis, tendon rupture) occurring during the course of statin therapy. Tendon injury manifested most commonly with pain that was severe and of sudden onset. Other frequent signs/symptoms included tenderness on palpation, swelling and erythema, and difficulty with movement of the involved tendinous area; warmth and stiffness were also reported (37). Diagnoses of tendinous manifestations were made primarily by physical examination, although radiography and occasionally ultrasonography and magnetic resonance imaging were used.
For these patients with statin-attributed tendinous manifestations, data were reviewed and collected using a standardized questionnaire, which included patient characteristics (age; sex; previous medical history, especially tendinopathy and comorbid illnesses that might predispose onset of tendinous complications; other concomitant therapy such as drugs known to increase statin concentrations [e.g., cyclosporin, gemfibrozil, CYP3A4 inhibitors]); characteristics of tendon adverse effect (time to onset of adverse effect after the institution of statin therapy, patterns of tendon side effect as well as tendon involved); outcome categories of statin-associated tendon involvement, including death, significant disability, and life-threatening reactions, as determined by the Pharmacovigilance system; and implicated statin (the percentage of the reports of statin-associated tendon impairment was calculated for each statin; dose of statin therapy was also identified when tendon injury occurred).
Overall, reports of 115 patients were collected and analyzed during the study period as possible cases of statin-associated tendinopathy. Of these patients, 96 were included in the study after analysis. Indeed, a causal link between tendon manifestations and statin use was not considered probable in the excluded 19 patients. In these latter 19 patients, another drug had been added to statin therapy and the tendinopathy occurred in the following days; obviously, the suspected therapy was the novel drug and the potential role of statin therapy was missed. These medications were as follows: fluoroquinolone (n = 15), steroids (n = 2), isotretinoin (n = 1), and metformin (n = 1).
Interestingly, during the period 1990–2005, we observed that the number of statin-attributed tendinous complications was as follows: 8 during 1990–1995, 32 during 1996–2000, and 56 during 2001–2005. From 1990 to 2005, a total of 4,597 statin-associated global side effects were reported in the network of all 31 French Pharmacovigilance Centers; the prevalence of statin-attributed tendinous manifestations was, in fact, as high as 2.09% (n = 96 in 4,597) (Table 1).
Table 1. Prevalence of statin-attributed tendinous manifestations among the total number of spontaneous reports of adverse effects of statins that have been reported in the network of 31 French Pharmacovigilance Centers during 1990–2005
Patients' general characteristics.
The patient population consisted of 67 men and 29 women; the ratio of men to women was 2.3:1. Patients' median age was 56 years (range 20–89 years). Regarding patients' age range, statin-associated tendinopathy occurred in 7 patients (7.4%) age 20–40 years, in 22 patients (23.4%) age 41–50 years, in 32 patients (34%) age 51–60 years, in 27 patients (28.7%) age 61–70 years, in 5 patients (5.3%) age 71–80 years, and in 1 patient (1.1%) age ≥81 years.
Twenty-seven patients (28.1%) had associated conditions, which have been reported to be a favoring factor of onset of tendinous complications; in essence, these 27 patients had a history of diabetes (n = 7) and hyperuricemia (n = 5), as well as sport practice (n = 15). Moreover, we also observed that 11 patients had a history of tendinopathy involving Achilles tendon in most cases (90%).
None of our 96 patients received concomitant drugs known to increase statin toxicity, e.g., gemfibrozil or other fibric acid derivatives and CYP3A4 inhibitors. The characteristics of patients who developed statin-associated tendinous complications are shown in Table 2.
Table 2. General characteristics of patients with statin-associated tendon complications in our series and in the Medline literature: 1966–2007*
|Sex, no.|| || |
|Age, median years||56||58|
|Associated comorbidities|| || |
|History of tendon disorder||11.5||–|
Time to onset of statin-associated tendinous manifestations.
The median time to onset of tendinous symptoms after the institution of statin therapy was 243 days (range 0–5,659 days). In 1 patient, tendinitis was, in fact, described to occur as early as within the first 24 hours after the initiation of statin therapy. However, tendinopathy more often occurred within the first year after statin initiation (in 59% of cases); time to onset of tendinopathy in patients who developed such a complication is shown in Table 3.
Table 3. Characteristics of tendon complications after the institution of statin therapy in our series and in the Medline literature: 1966–2007*
|Time to onset, median months||8||10.2|
Characteristics of statin-associated tendon manifestations.
In our population, the patients exhibited the following statin-associated tendon manifestations: tendinitis (n = 63), tendinitis with subsequent tendon rupture (n = 12), and de novo tendon rupture (n = 21). The following symptoms were more commonly found in our 96 patients: pain (n = 96), swelling (n = 59), warmth and erythema (n = 28), and stiffness and difficulty with movement of the involved tendinous area (n = 62).
The Achilles tendon was the most frequent site of tendon injury (n = 50; 52.1% of cases). Other locations of statin-associated tendon manifestations were also observed in our patients, i.e., quadriceps femoris tendon (n = 14), tendon of musculus gluteus medius (n = 3) and tibialis anterior (n = 1), rotator cuff (n = 1), tendon of subscapularis terrea (n = 1), musculus supraspinatus (n = 2) and infraspinatus (n = 3), epicondyles (n = 9), tendon of musculus biceps brachii (n = 7), and flexor/extensor tendon sheath (n = 5). Moreover, statin-induced tendon manifestations were concomitantly bilateral in 26 patients (41.3% of cases).
Severity and outcome of statin-associated tendon manifestations.
Interventions included discontinuation of the suspected statin in all patients, as well as concomitant nonsurgical procedures (e.g., analgesics, immobilization). In our population, the median delay of regression of tendon manifestations was 23 days after disruption of statin therapy.
However, statin-attributed tendinous complications were considered to be serious in 36 patients (37.5% of cases). In fact, 17 patients were admitted to the hospital. Moreover, 19 patients had significant functional sequelae (difficulty walking, decreased flexion, bruising, and pain) related to statin-attributed tendon rupture. None of our patients died.
Finally, after the onset of statin-associated tendon complications, statin therapy was reinitiated in 7 patients, which resulted in recurrence of statin-associated tendon manifestations (100% of cases).
Statins responsible for tendinous complications.
As shown in Table 4, atorvastatin was the most commonly involved (n = 35 [37%]). The remaining patients exhibited statin-associated tendinous complications related to other agents such as simvastatin (n = 30 [31%]), pravastatin (n = 21 [22%]), fluvastatin (n = 5 [5%]), and rosuvastatin (n = 5 [5%]).
Table 4. Statins responsible for tendon manifestations in our series and in the Medline literature: 1966–2007*
We further found that doses of statins used were within the recommended ranges in all 96 patients. Indeed, the median dosages of statin at onset of tendinous manifestations were as follows: atorvastatin: 26.3 mg/day (range 10–80), simvastatin: 18.3 mg/day (range 10–40), pravastatin: 21.8 mg/day (range 5–40), fluvastatin: 40 mg/day (range 20–80), and rosuvastatin: 10 mg/day.
Tendinous manifestations may be an as yet unreported side effect of statin therapy. In essence, no cases were reported during the large therapeutic trials, which included more than 30,000 patients (1–26). However, the fact that statin-related tendon complications have not been described in these large clinical trials may be explained in part by control of factors predisposing to tendinopathy and regular followup of trial patients (1–26). Our literature search using the Medline database (1966–2007), reviewing all English- and French-language articles, identified 43 cases of statin-attributed tendon complications. These 43 statin-attributed tendon injury cases have been reported in 2 articles and 2 abstracts (33–36); the first case report was published in 2000. Interestingly, we have observed that these previous reports all come from France, as was the case initially with fluoroquinolone-related tendinopathies; the reason for this remains unclear.
The present retrospective series involved 96 spontaneous cases of statin-attributed tendon complications that were reported in the 31 French Pharmacovigilance Centers between 1990 and 2005. In our series, tendon disorders could be reasonably attributed to statin therapy because 1) there was a temporal relationship between the onset of tendinous signs and the initiation of statin therapy; 2) disappearance/improvement of tendon manifestations occurred after disruption of statins; and 3) tendinopathy recurred in the 7 patients (100% of cases) in whom statin therapy was reinstituted, which strongly supports a relationship with the use of these drugs. Interestingly, we observed that the frequency of reported statin-attributed tendinous complications did not differ between the 31 Pharmacovigilance Centers; this finding suggests that a selection bias of cases, according to Pharmacovigilance Centers, could be, to a certain extent, excluded. Moreover, data from our patients and those from the literature were comparable (33–36), which also suggests that they may reflect the current situation; these similarities also indicate that a bias related to the background of patients may be, in part, excluded. In our study, we have further observed an increased number of statin-attributed tendinous complications reported in the network of 31 Pharmacovigilance Centers from 1990 to 2005; these data are obvious, as the number of statin prescriptions has increased in France during the same period. Interestingly, in this instance, we have found that tendinous manifestations accounted for 2.09% of overall statin-related adverse effects. Nevertheless, our retrospective study cannot define the prevalence/incidence of statin-associated tendon disorders in statin-treated patients, although it suggests that the prevalence may be low considering the huge number of statin prescriptions filled worldwide; in turn, in 2003, 3.2 million subjects received statins in France, and only 13 statin-attributed tendon disorders were reported in the network of 31 French Pharmacovigilance Centers. Only epidemiologic studies will indeed allow the prevalence/incidence of such side effects to be determined. Our literature review failed to find articles describing comparisons of patients who have received statins with those who have not been treated in an attempt to quantify incidence of statin-associated complications; such previous studies have shown that fluoroquinolone-treated patients have an increased relative risk (3.7–10.1) of developing tendinous complications (38, 39).
All statins may be responsible for tendinous complications (33–36). In our Medline analysis of 43 cases of statin-attributed tendinous complications, we in fact found that the percentages of cases associated with each statin were as follows (Table 4): simvastatin (86.2%), atorvastatin (10.4%), and fluvastatin (3.4%) (33–36). In our series, we observed the following percentages of cases associated with each statin: atorvastatin (37%), simvastatin (31%), pravastatin (22%), fluvastatin (5%), and rosuvastatin (5%). We suggest that the higher number of reports regarding atorvastatin-, simvastatin-, and pravastatin-attributed tendinous manifestations may be related to the greater use of these 3 statins from 1990 to 2005 in France. In turn, our data underscore that tendinous complications appear to be associated with each of the available statins.
Another finding of the present series is that tendinous manifestations occurred in patients receiving statins at recommended dose ranges, as the median daily dosages of statin were as follows: atorvastatin: 26.3 mg/day, simvastatin: 18.3 mg/day, pravastatin: 21.8 mg/day, fluvastatin: 40 mg/day, and rosuvastatin: 10 mg/day. Our findings therefore suggest that statin-attributed tendon complications may not be dose dependent. Furthermore, none of our 96 patients received combined therapy of statin and fibric acid derivative, statin and ezetimibe, or statin and a drug known to increase statin toxicity (particularly CYP3A4 inhibitors, fluoroquinolone, or steroids).
The time to onset of statin-attributed tendinous complications has been found to be variable. The Medline analysis of the 43 statin-associated tendon manifestations shows that the median time interval between the first intake of statin and the onset of tendinous manifestations was 10.2 months (Table 3) (33–36). Our series confirms results of previous studies, with tendinous symptoms occurring from within hours to as long as months after the initiation of statin therapy. In this instance, 59% of events occurred within the first year after starting statin therapy; interestingly, 80% of these latter patients exhibited tendon manifestations within the first 4 months after initiation of statin therapy. This time interval between the first intake of statins and the onset of tendinous signs was similar for all 5 statins. Our study suggests that regular tendinous clinical examination may be required in statin-treated patients, particularly during the first year following statin therapy initiation.
Various statin-attributed tendinous manifestations have been described (33–36). Our Medline review reveals that statin-attributed tendinitis is the most common tendon complication; in essence, we found that statin-associated tendon manifestations included tendinitis (n = 37) and tendon rupture (n = 6) (Table 3) (33–36). In this instance, among the 96 patients who developed statin-related tendinous manifestations, we observed tendinitis and tendon rupture in 65.6% and 34.4% of cases, respectively. Nevertheless, among the patients who exhibited statin-attributed tendon complications, we failed to find differences between patients with tendinitis and those with tendon rupture. Moreover, our study demonstrates that Achilles tendinopathy was the most common tendinous location; this finding is not unexpected because Achilles tendon impairment is also encountered more frequently in tendinitis associated with other drugs (especially fluoroquinolone); previous authors have reported that high-impact use of the Achilles tendon leads to the site being the most common for drug-associated tendon injury (37).
From a practical point of view, knowledge of predictive factors of statin-attributed tendon impairment appears essential in order to improve patient treatment. In our series, a small trend indicated that more men experienced statin-attributed tendon injury (69.8% of men versus 30.2% of women; sex ratio 2.3:1); these sex differences were small, and were also reported in the 43 cases in the literature (sex ratio: 1.7:1). These findings have also been described in patients with tendon disorders related to fluoroquinolone (39–41). Furthermore, previous investigators have described comorbid conditions associated with the onset of other drug-associated tendon complications such as diabetes mellitus, renal failure and/or hemodialysis, hyperparathyroidism, history of musculoskeletal disorders, or gout (37, 42, 43). Our Medline analysis of 43 patients with statin-related tendinous manifestations revealed associated comorbid disorders (i.e., diabetes [7% of cases], hyperuricemia [2.3% of cases]), as well as physical exertion (7% of cases) (33–36). In our population of 96 patients with statin-attributed tendon complications, we found similar prevalences of the following comorbid diseases: diabetes (7.3%) and hyperuricemia (5.2%); sporting activity was found in 15.6% of our patients. In fluoroquinolone-treated patients, the majority of investigators recommend discontinuation of fluoroquinolone before events that may exacerbate tendon injury, including prodigious amounts of physical exertion (such as marathon running) (40, 41, 44). We also suggest that statin therapy may be disrupted in patients before such prodigious physical exertion and marathon running; however, no definite conclusion can be drawn from our retrospective uncontrolled series, and our study warrants further investigation.
Nothing is known regarding how statins may produce tendon injury. Theories of statin-associated tendon toxicity, including the following theories, have been proposed based on statin-associated skeletal muscle toxicity: 1) cholesterol is an important component of human cell membranes; indeed, blocking cholesterol synthesis may reduce the cholesterol content of tendon cell membranes, making them unstable and changing the function of tendon cell membranes; 2) reduced levels of regulatory proteins involved in the maintenance of tendon cells may be responsible for tendon injury; and 3) apoptosis produced by statins reduces vascular smooth muscle cell proliferation; indeed, apoptosis of tendon cells with hydroxymethylglutaryl-coenzyme A reductase inhibitors could also lead to tendon damage in statin-treated patients (27–32).
In fluoroquinolone-treated patients, other physiopathologic mechanisms have been proposed to explain tendon toxicity, especially oxidative stress resulting in a decrease of both collagen synthesis and proliferation of fibroblasts (45). Nevertheless, our study warrants further investigation to detect whether statin may induce such a tendon toxicity.
In conclusion, statins are the most effective therapeutic agents for reducing LDL cholesterol and have been documented to reduce the incidence of cardiovascular events in patients. Our retrospective series and our literature review suggest that tendon manifestations may be included within the spectrum of adverse effects of statins; to date, French recommendations include tendon disorders within adverse effects of statins. Nevertheless, future case–control pharmacoepidemiologic studies are required to 1) confirm the relationship between statin therapy and onset of tendinous complications, and 2) determine actual prevalence/incidence of statin-associated tendinopathy. We suggest that prescribers should be aware of tendinous manifestations attributable to statins, particularly in situations at high relevant risk: high level of physical exertion, metabolic disorders, and association with fluoroquinolone treatment or other drugs known to increase the toxicity of statins. We also suggest that patients who are at risk of developing statin-associated tendon manifestations and who require statins be routinely questioned about symptoms consistent with tendon involvement. Finally, recognition of statin-related tendon complications also appears to be important in preventing serious sequelae in statin-treated patients; postmarketing surveillance appears to be a major tool for early detection of safety problems with a new drug.
Dr. Marie had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study design. Marie, Delafenêtre, Massy, Thuillez, Noblet.
Acquisition of data. Marie, Delafenêtre, Massy, Noblet.
Analysis and interpretation of data. Marie, Delafenêtre, Massy, Thuillez, Noblet.
Manuscript preparation. Marie, Delafenêtre, Massy, Thuillez, Noblet.