SEARCH

SEARCH BY CITATION

Keywords:

  • bone loss;
  • clopidogrel;
  • osteoporosis;
  • platelet aggregation inhibitors;
  • spinal fracture;
  • thienopyridine

Abstract.

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Acknowledgements
  9. References

Jørgensen NR, Grove EL, Schwarz P, Vestergaard P (Research Center for Aging and Osteoporosis, Copenhagen University Hospital Glostrup; Aarhus University Hospital, Skejby; University of Copenhagen, Copenhagen; The Osteoporosis Clinic, Aarhus University Hospital, Denmark). Clopidogrel and the risk of osteoporotic fractures: a nationwide cohort study. J Intern Med 2012; 272: 385–393.

Objectives:  The P2Y12 inhibitor clopidogrel inhibits platelet aggregation and is used in the treatment and prevention of coronary artery disease. It is widely used and, in combination with acetylsalicylic acid, is the standard of care for acute coronary syndrome and percutaneous coronary intervention. The mode of action of clopidogrel involves pathways that are important to the metabolic activity in bone cells, although to our knowledge whether P2Y12 receptors are involved in the regulation of bone metabolism has not yet been investigated. Therefore, the objective of the present study was to investigate the association between clopidogrel use and risk of fractures.

Methods:  We investigated the association between clopidogrel use and fracture incidence in a nationwide cohort study within the Danish population of approximately 5.3 million individuals. All patients who were prescribed clopidogrel during the years 1996–2008 were included in the study (= 77 503), and three nonusers were randomly selected, matched for age and gender (= 232 510), for each clopidogrel-treated subject.

Results:  Treatment with clopidogrel was associated with both increased overall fracture risk and increased risk of osteoporotic fractures, especially in subjects with a treatment duration of more than 1 year. However, individuals with low exposure to clopidogrel (<0.01 defined daily dose) had a lower risk of fracture than never users.

Conclusions:  Use of the P2Y12 inhibitor clopidogrel is associated with risk of fractures. There seems to be a biphasic relation so that lower doses are associated with decreased fracture risk, whereas higher doses (recommended dose range) are associated with increased risk. More studies are warranted to determine the potential in vivo effect of platelet aggregation inhibitors on bone metabolism.


Introduction

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Acknowledgements
  9. References

Antiplatelet drugs are widely used in the treatment and prevention of coronary artery disease, especially in the setting of acute coronary syndrome (ACS) and percutaneous coronary intervention [1]. Acetylsalicylic acid and clopidogrel are the most widely used platelet function inhibitors, and dual antiplatelet treatment with these drugs is the most common treatment for ACS and percutaneous coronary intervention [1, 2]. Furthermore, these drugs are also commonly used for secondary prevention of stroke and peripheral artery occlusive disease. Clopidogrel is a thienopyridine and inhibits platelet function by irreversibly inhibiting the P2Y12 adenosine diphosphate (ADP) receptor. Previously, ticlopidine was used, but this drug has significant side effects, including bone marrow toxicity [3]. Newer agents are now available such as the thienopyridine prasugrel and the nonthienopyridine ticagrelor. These drugs also inhibit the P2Y12 receptor, but in contrast to the other antiplatelet drugs, the binding sites for ticagrelor and ADP are different; it is thus an allosteric antagonist, and blockade is reversible. Both prasugrel and ticagrelor provide a more uniform and potent platelet inhibition and a reduced risk of recurrent cardiovascular events compared with clopidogrel [4–6].

We have previously investigated the relationship between clopidogrel treatment and the occurrence of fractures in a large Danish register-based study, but were unable to detect a convincing association [7]. However, our previous study was based on data from the year 2000 when guidelines did not provide recommendations for long-term treatment with clopidogrel. In recent years, recommendations have changed so that treatment for up to 1 year is recommended for patients with ACS and for at least 1 year in patients with drug-eluting stents [1].

Thienopyridines irreversibly inhibit the pivotal P2Y12 receptor, which belongs to the family of G-protein-coupled P2 purinergic, nucleotide-activated receptors [8]. ADP is the natural ligand for P2Y12, but it also acts as an agonist for P2Y1 receptors. P2 receptors are widely expressed in most tissues in the body, but the expression pattern varies from tissue to tissue and from cell type to cell type. A number of P2 receptors are expressed in osteoblastic cells including the P2Y12 receptor [9], although the importance of this receptor in relation to bone metabolism remains to be proven. Binding of thienopyridines to the P2Y12 receptor blocks normal intracellular signalling, resulting in inhibition of the Rho-associated kinase and inositol triphosphate-/mitogen-activated protein kinase (MAP kinase) pathways [10]. As a result of irreversible receptor binding, the effects last for the entire lifespan of the platelets, that is about 1 week [11]. Furthermore, bone cell precursors are present in the bone marrow and share common precursors with cells of the immune and haematopoietic systems. The potential toxic effects of thienopyridines on the bone marrow [3, 10] could further affect bone cell recruitment, proliferation and differentiation, with implications for normal bone turnover and bone mass.

In the present study, we evaluated the association between the platelet inhibitor clopidogrel and fracture incidence in a population-based nationwide case–control study. The aim of the study was to investigate whether use of clopidogrel is associated with an increased risk of fractures.

Materials and methods

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Acknowledgements
  9. References

Study design

In the present cohort study, all individuals who were prescribed clopidogrel during the years 1996–2008 in Denmark were included as exposed subjects (= 77 503). Three subjects of the same age (same birth year) and gender were randomly selected for each exposed subject from the general population as controls (nonusers) (= 232 510). Nonusers were selected and matched according to age and gender to the users through an intensity sampling technique.

Study end-points

The study end-points were occurrence of any fracture (International Classification of Diseases (ICD) 10 codes: S02.0–S02.9, S07.0–S07.9, S12.0–S12.9, S22.0–S22.9, S32.0–S32.8, S42.0–S42.9, S52.0–S52.9, S62.0–S62.9, S72.0–S72.9, S82.0–S82.9 and S92.0–S92.9), hip fracture (ICD10: S72.0 and S72.1), forearm fracture (ICD10: S52.5 and S52.6) and spine fracture (ICD10: S12.0, S12.1, S12.2, S12.7, S12.8, S12.9, S22.0, S22.1, S32.0, S32.7 and S32.8) between 1 January 1996 and 31 December 2008. In Denmark, almost all patients with fractures are managed within the hospital system (including in the emergency department) [12]. For insurance reasons, even fractures sustained abroad are registered upon return. The capture of fracture information is thus nearly complete [13, 14]. All clinical fractures according to hospital records of patients who were referred to emergency or other departments and were diagnosed with a vertebral fracture were included as spinal fractures in the analyses.

Exposure variables

Clopidogrel use was the primary exposure variable. Patterns of drug use were analysed for the period from 1 January 1996 to the date of fracture or the corresponding dummy date amongst nonusers. Data were collected systematically and information on whether the drugs were used systematically or temporarily was included in the analyses of drug use through the defined daily dose (DDD), using the dates of prescription. According to the definition of the World Health Organization, DDD is the assumed average maintenance dose per day for a drug used for its main indication in adults. Other exposure variables were the use of drugs known to be associated with fracture risk (corticosteroids), and the occurrence of comorbidities known to affect fracture risk (prior fracture [15] and alcoholism [16]). These factors were chosen as it is known that they may affect fracture risk, and they are regarded as important potential confounders in a setting in which many variables besides the main factor may influence the risk of fractures (confounding by indication). To analyse for confounding from use of other cardiovascular drugs, these were also included [17–19].

Because of the imbalances in confounders typical for observational studies, analyses were performed with extensive control for confounding factors including for all unevenly distributed known confounders: Charlson index, income, living alone, prior fracture, alcoholism and use of bronchodilators or smoking cessation medication (proxy for smoking), systemic corticosteroids, statins, angiotensin-converting enzyme (ACE) inhibitors, ACE inhibitors plus diuretics, combined alpha- and beta-blockers, beta-blockers plus any other drugs, beta-blockers alone, calcium channel blockers, spironolactone, thiazide diuretics, loop diuretics, other diuretics, dipyridamole and acetylsalicylic acid. Analyses for interaction with these variables were performed. Information on body weight or body mass index was not available.

Registers

The information on fracture occurrence and occurrence of other diseases, prior fractures and alcoholism came from two registers: the National Hospital Discharge Register [13], and the Psychiatric Central Register [20]. These have been described in detail previously [7, 13, 14, 19].

The information on clopidogrel use came from the Danish Medicines Agency, which keeps a nationwide register of all drugs sold at pharmacies throughout the country from 1996 onwards (the National Pharmacological Database run by the Danish Medicines Agency: http://www.dkma.dk). All purchased drugs are registered with ATC code, dosage sold and the date of sale for the period 1 January 1996 to 31 December 2008. A more detailed description has previously been published [7].

It is possible to link these sources of information through the Central Person Register number, which is unique to all Danish citizens thus enabling registration on an individual basis.

The project was approved and controlled by the National Board of Health, the Danish Data Protection Agency and the directory board of the Psychiatric Central Register.

Statistical analyses

Mean and standard deviation were used as descriptive statistics. Crude and adjusted hazard ratios and 95% confidence intervals were calculated. A propensity-based analysis (propensity score matching) [21] was also performed, but did not change the results. Cox proportional hazard regression models were used to analyse the time to fracture in exposed versus nonexposed subjects. The proportional hazard assumption was checked through inspection of survival plots. In addition, age and gender interactions were tested for in Cox regression models. Analyses were performed using STATA 9.0 (STATA Corp., College Station, TX, USA) and IBM SPSS 19.0 (IBM Corp., Armonk, NY, USA).

Results

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Acknowledgements
  9. References

Baseline characteristics of users and nonusers

Clopidogrel-treated and control (nonusers) subjects were well matched for age and gender; Table 1 shows the baseline characteristics of the two groups. The proportion of subjects with a previous fracture was slightly but significantly higher in the clopidogrel-treated group (Table 1). It is not surprising that the proportion of individuals with previous comorbidity from cardiovascular disease was clearly higher amongst clopidogrel users than amongst nonusers, and the proportion of subjects with prior concomitant treatment with statins, beta-blockers, ACE inhibitors, diuretics and other platelet inhibitors was also higher amongst users (Table 1).

Table 1. Baseline characteristics of clopidogrel users and nonusers. Data are expressed as numbers (%) or mean ± SD
ParameterClopidogrel users (= 77 503)Nonusers (= 232 510) P
  1. ACE, angiotensin-converting enzyme; ARB, angiotensin-receptor blocker; ASA, acetylsalicylic acid.

Age, years65.7 ± 12.765.7 ± 12.7
Men50 118 (64.7)150 354 (64.7)
Women27 385 (35.3)82 156 (35.3) 
Prior fracture16 730 (21.6)47 795 (20.6)<0.01
Prior alcoholism2469 (3.2)6562 (2.8)<0.01
Prior acute myocardial infarction45 365 (58.5)10 709 (4.6)<0.01
Prior angina pectoris53 523 (69.1)21 428 (9.2)<0.01
Prior heart failure11 243 (14.5)8129 (3.5)<0.01
Prior peripheral atherosclerosis7356 (9.5)7193 (3.1)<0.01
Prior ischaemic stroke7003 (9.0)5772 (2.5)<0.01
Prior cerebral atherosclerosis2784 (3.6)2900 (1.2)<0.01
Prior diabetes10 863 (14.0)11 894 (5.1)<0.01
Prior atrial fibrillation8036 (10.4)12 080 (5.2)<0.01
Prior use of systemic corticosteroids20 050 (25.9)47 067 (20.2)<0.01
Prior use of statins30 345 (39.2)26 735 (11.5)<0.01
Prior use of ACE inhibitors/ARBs38 658 (49.9)46 380 (20.0)<0.01
Prior use of ACE inhibitors/ARBs plus diuretics6529 (8.4)13 129 (5.7)<0.01
Prior use of combined alpha-/beta-blockers2897 (3.7)3187 (1.4)<0.01
Prior use of other diuretics8737 (11.3)16 926 (7.3)<0.01
Prior use of beta-blockers alone36 288 (46.8)43 114 (18.5)<0.01
Prior use of beta-blocker combinations562 (0.7)1249 (0.5)<0.01
Prior use of calcium antagonists27 953 (36.1)41 744 (18.0)<0.01
Prior use of thiazide diuretics24 895 (32.1)53 675 (23.1)<0.01
Prior use of loop diuretics18 105 (23.4)31 725 (13.7)<0.01
Prior use of dipyridamole8123 (10.5)7331 (3.2)<0.01
Prior use of low-dose ASA44 015 (56.8)54 610 (23.5)<0.01
Prior use of bronchodilator drugs23 900 (30.)54 703 (23.5)<0.01
Prior use of smoking cessation medication1969 (2.5)2240 (1.0)<0.01
Prior use of spironolactone11 292 (14.6)11 889 (5.1)<0.01
Charlson index2.1 ± 1.90.8 ± 1.5<0.01
Income in index year (Danish crowns)209 223 ± 315 460224 149 ± 227 691<0.01
Not living alone29 030 (37.5)88 901 (38.2)<0.01

Effects of clopidogrel on fracture risk

Table 2 shows the crude relative risk of any fracture or selected osteoporotic fractures (hip, forearm and spine). As can be seen from Table 2, the relative risk of any fracture or an osteoporotic fracture is reduced for individuals exposed to low doses of clopidogrel, whereas the risk of fracture is significantly increased with higher doses (up to prescribed doses). Moreover, this finding is consistent for any fracture and for all three types of osteoporotic fractures (Table 2). As there was a small, yet significant difference in the occurrence of prior fractures between users and nonusers of clopidogrel, Cox proportional hazard regression was performed to adjust for this factor. Even after adjustment, there was a significantly increased risk of any fracture or osteoporotic fractures of up to 50% amongst clopidogrel users (Table 2). By comparison, the hazard ratio for subjects with prior fractures, which are known to considerably increase the risk of a subsequent fracture, was between 2.3 and 2.9 (Table 2). As the function of other P2 receptors is affected by gender, we stratified accordingly. Of interest, we found that the fracture risk was increased by clopidogrel from low doses (from 0.10–0.39 DDD) in women for all fractures and hip fractures, whereas fracture incidence amongst men was only increased for higher doses (Table 3).

Table 2. Risk of fractures by dose of clopidogrel. The frequency of any fractures and of the three major groups of osteoporotic fractures (hip, forearm and spine) is presented. Risk of fractures is presented as crude relative risk, hazard ratio adjusted for prior fractures and hazard ratio adjusted for multiple confounders
 Clopidogrel doseObs. yearsAny fractureHip fractureForearm fractureSpine fracture
n IRR (95% CI) n IRR (95% CI) n IRR (95% CI) n IRR (95% CI)
  1. Multiple confounders: Charlson index, income, living alone, prior fracture, alcoholism and use of spironolactone, bronchodilator drugs or smoking cessation medication (proxy for smoking), systemic corticosteroids, statins, angiotensin-converting enzyme (ACE) inhibitors, ACE inhibitors plus diuretics, combined alpha-plus beta-blockers, beta-blockers plus any other drugs, beta-blockers alone, calcium channel blockers, thiazide diuretics, loop diuretics, other diuretics, dipyridamole and acetylsalicylic acid.

  2. CI, confidence interval; DDD, defined daily dose; IRR, incidence rate ratio; Obs, observational. *Two-tailed < 0.05.

Crude relative riskNever user (control)700 25316 211 3881 3660 887 
<0.10 DDD89 69719480.94 (0.90–0.98)*4160.84 (0.76–0.93)*4370.93 (0.84–1.03)1291.14 (0.94–1.37)
0.10–0.39 DDD68 78716421.03 (0.98–1.08)3861.01 (0.91–1.12)3390.94 (0.84–1.05)840.96 (0.77–1.21)
0.40–0.99 DDD47 37715271.39 (1.32–1.47)*3921.49 (1.35–1.66)*3311.34 (1.19–1.50)*951.58 (1.28–1.95)*
≥1 DDD23 8548181.48 (1.38–1.59)*2361.79 (1.57–2.03)*1721.38 (1.18–1.61)*461.52 (1.13–2.04)*
   HR (95% CI)HR (95% CI)HR (95% CI)HR (95% CI)
Adjusted for prior fracture<0.10 DDD0.97 (0.92–1.01)0.86 (0.78–0.95)*0.95 (0.86–1.05)1.13 (0.94–1.36)
0.10–0.39 DDD1.02 (0.97–1.07)1.00 (0.90–1.11)0.93 (0.84–1.04)0.96 (0.77–1.20)
0.40–0.99 DDD1.33 (1.26–1.40)*1.41 (1.27–1.57)*1.28 (1.14–1.43)*1.52 (1.23–1.88)*
≥1 DDD1.38 (1.28–1.48)*1.63 (1.43–1.86)*1.30 (1.12–1.52)*1.46 (1.09–1.97)*
Prior fracture2.43 (2.36–2.49)*2.92 (2.76–3.08)*2.34 (2.20–2.48)*2.64 (2.35–2.96)*
Adjusted for multiple confounders<0.10 DDD0.84 (0.79–0.88)*0.64 (0.58–0.72)*0.88 (0.79–0.98)*0.87 (0.71–1.06)
0.10–0.39 DDD0.88 (0.83–0.93)*0.75 (0.67–0.84)*0.87 (0.77–0.98)*0.73 (0.57–0.93)*
0.40–0.99 DDD1.07 (1.01–1.14)*0.95 (0.85–1.06)1.12 (0.99–1.27)1.03 (0.82–1.29)
≥1 DDD1.08 (1.01–1.16)*1.09 (0.95–1.25)1.11 (0.95–1.30)0.95 (0.70–1.30)
Table 3. Risk of any fracture and of the three major types of osteoporotic fractures (hip, forearm and spine) in users of clopidogrel compared with nonusers stratified by gender. Cox proportional hazard regression; values are hazard ratio (95% confidence interval) compared with never users
GenderDose of clopidogrelAdjusted for prior fractureAdjusted for multiple confounders
Any fractureHip fractureForearm fractureSpine fractureAny fractureHip fractureForearm fractureSpine fracture
  1. Adjusted for multiple confounders: prior fracture, alcoholism and use of systemic corticosteroids, statins, angiotensin-converting enzyme (ACE) inhibitors, ACE inhibitors plus diuretics, combined alpha-plus beta-blockers, beta-blockers plus any other drugs, beta-blockers alone, calcium channel blockers, thiazide diuretics, loop diuretics, other diuretics, dipyridamole and acetylsalicylic acid.

  2. DDD, defined daily dose.

  3. *Two-tailed < 0.05.

Men<0.10 DDD0.96 (0.90–1.03)0.83 (0.71–0.97)*0.99 (0.83–1.17)1.05 (0.80–1.37)0.93 (0.87–1.00)0.73 (0.62–0.86)*0.99 (0.83–1.19)0.97 (0.73–1.29)
0.10–0.39 DDD1.02 (0.95–1.10)0.91 (0.77–1.07)0.96 (0.79–1.17)0.99 (0.73–1.34)1.00 (0.92–1.07)0.82 (0.69–0.98)*0.96 (0.79–1.18)0.91 (0.66–1.25)
0.40–0.99 DDD1.28 (1.18–1.38)*1.40 (1.18–1.65)*1.17 (0.95–1.46)1.29 (0.92–1.79)1.14 (1.05–1.24)*1.04 (0.87–1.23)1.11 (0.89–1.39)1.07 (0.75–1.51)
≥1 DDD1.31 (1.16–1.47)*1.58 (1.25–2.00)*1.40 (1.03–1.89)*1.57 (0.99–2.48)1.12 (0.99–1.27)1.06 (0.83–1.35)1.29 (0.94–1.76)1.24 (0.74–1.99)
Women<0.10 DDD1.00 (0.94–1.07)0.91 (0.79–1.04)0.97 (0.86–1.10)1.25 (0.97–1.62)0.89 (0.83–0.96)*0.77 (0.67–0.89)*0.89 (0.78–1.01)1.02 (0.78–1.34)
0.10–0.39 DDD1.12 (1.04–1.20)*1.19 (1.04–1.36)*1.03 (0.90–1.18)1.02 (0.73–1.41)0.99 (0.91–1.06)1.00 (0.87–1.15)0.93 (0.81–1.08)0.82 (0.59–1.15)
0.40–0.99 DDD1.43 (1.33–1.53)*1.47 (1.28–1.68)*1.33 (1.17–1.52)*1.80 (1.36–2.37)*1.19 (1.11–1.29)*1.12 (0.97–1.29)1.18 (1.02–1.35)*1.32 (0.99–1.77)
≥1 DDD1.30 (1.19–1.42)*1.49 (1.27–1.75)*1.07 (0.89–1.27)1.31 (0.89–1.94)1.06 (0.97–1.16)1.09 (0.92–1.28)0.94 (0.78–1.13)0.92 (0.62–1.38)

Next, hazard ratios were calculated using Cox proportional hazard regression analysis for different dose strata after adjusting for prior fracture, alcoholism and use of systemic corticosteroids, statins, ACE inhibitors, ACE inhibitors plus diuretics, combined alpha- plus beta-blockers, beta-blockers plus other drugs combined, beta-blockers alone, calcium channel blockers, thiazide diuretics, loop diuretics, other diuretics, dipyridamole and acetylsalicylic acid (Table 2). No interactions with age and gender were observed. The analysis showed a dual effect of clopidogrel with a significant decrease in hazard ratio for most fractures for low doses, and an increased risk for higher doses. Again, we stratified for gender without any effect on the results (Table 3).

As we have shown previously [7], the hazard ratio was also increased in users of acetylsalicylic acid and dipyridamole in the present cohort (Table 4). Finally, we stratified clopidogrel users according to treatment duration. Subjects treated with clopidogrel for a duration of more than 1 year had a significantly increased risk of any fracture and of osteoporotic fractures (hip, forearm and spine) (Table 5).

Table 4. Risk of any fracture and of the three major types of osteoporotic fractures (hip, forearm and spine) in users of clopidogrel compared with nonusers. Cox proportional hazard regression; values show hazard ratio (95% confidence interval) compared with never users
VariableAny fractureHip fractureForearm fractureSpine fracture
  1. Adjusted for prior fracture, alcoholism and use of systemic corticosteroids.

  2. DDD, defined daily dose; ASA, acetylsalicylic acid.

  3. *Two-tailed < 0.05.

Dose of clopidogrel
<0.10 DDD0.82 (0.78–0.86)*0.60 (0.54–0.67)*0.83 (0.75–0.93)*0.90 (0.74–1.10)
0.10–0.39 DDD0.87 (0.83–0.92)*0.71 (0.63–0.79)*0.82 (0.72–0.92)*0.77 (0.61–0.98)*
0.40–0.99 DDD1.17 (1.11–1.24)*1.08 (0.96–1.20)1.14 (1.02–1.29)*1.27 (1.02.1.58)*
≥1 DDD1.30 (1.21–1.39)*1.44 (1.26–1.64)*1.24 (1.06–1.44)*1.34 (0.99–1.81)
Dose of ASA
≤ 75 mg per day1.27 (1.22–1.33)*1.62 (1.49–1.76)*1.21 (1.11–1.32)*1.52 (1.28–1.79)*
76–100 mg per day1.20 (1.16–1.26)*1.67 (1.54–1.81)*1.21 (1.11–1.32)*1.21 (1.01–1.44)*
>100 mg per day1.27 (1.22–1.32)*1.86 (1.72–2.02)*1.14 (1.04–1.25)*1.42 (1.20–1.69)*
Dose of dipyridamole
≤ 300 mg per day1.29 (1.18–1.42)*1.82 (1.56–2.12)*1.21 (0.99–1.48)1.30 (0.89–1.88)
301–400 mg per day1.45 (1.31–1.60)*1.90 (1.61–2.24)*1.18 (0.94–1.48)1.59 (1.09–2.33)*
> 400 mg per day1.20 (1.08–1.33)*1.41 (1.18–1.69)*0.99 (0.78–1.26)1.24 (0.81–1.88)
Table 5. Hazard ratio of any fracture and of the three major types of osteoporotic fractures (hip, forearm and spine) by total cumulated dose for those patients who were observed for ≥1 year
VariableAny fractureHip fractureForearm fractureSpine fracture
  1. Adjusted for prior fracture, alcoholism and use of systemic corticosteroids.

  2. DDD, defined daily dose.

  3. *Two-tailed < 0.05.

Dose of clopidogrel
≤180 DDD1.00 (0.96–1.05)1.00 (0.91–1.09)0.96 (0.87–1.06)1.15 (0.96–1.38)
181–365 DDD0.94 (0.88–1.00)0.83 (0.72–0.96)*0.84 (0.72–0.96)*0.89 (0.67–1.19)
>365 DDD1.23 (1.18–1.28)*1.24 (1.13–1.35)*1.24 (1.13–1.35)*1.22 (1.02–1.47)*

Discussion

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Acknowledgements
  9. References

In the present study, we observed a significant association between risk of all fractures and use of clopidogrel. However, there was a dual effect on fracture risk with low exposure associated with decreased risk and higher exposure associated with increased risk of fracture, compared with controls. Of note, the risks of all fractures and of the three types of osteoporotic fractures were increased for the higher-dose groups.

The expression of the P2Y12 receptor in osteoblasts has already been demonstrated [9]. By contrast, to date, the effects of the P2Y12 receptor on osteoblast and osteoclast functions have not been elucidated. However, in platelets, clopidogrel inhibits pathways that are known to be involved in bone formation in osteoblasts. Therefore, theoretically, clopidogrel could have an inhibitory effect on bone metabolism. First, binding of thienopyridines to the P2Y12 receptor in platelets blocks normal intracellular signalling, thus inhibiting the RHO-associated kinase and inositol triphosphate/MAP kinase pathways [10]. If clopidogrel binds irreversibly to the receptor in osteoblasts, as in platelets, the effect will last the lifespan of the osteoblasts. This could lead to irreversible inhibition of osteoblast P2Y12 receptor signalling, which subsequently would affect osteoblast activity. Furthermore, bone cell precursors are present in the bone marrow as well as, in some cases, in cells of the immune and haematopoietic systems. Thus, the known toxic effects of thienopyridines on the bone marrow [3, 10] might also affect recruitment, proliferation and differentiation of osteoblasts from the marrow compartment. This may have implications for normal bone turnover and bone mass.

To our knowledge, this is the first study to demonstrate that use of clopidogrel is associated with fracture risk, although the effect seems to be dependent on dose. We have no clear explanation for this dose-dependent dual effect of clopidogrel. However, P2Y12 receptors are expressed by both osteoblasts and osteoclasts, and clopidogrel might therefore have effects on both cell types and thereby on both bone formation and resorption. Furthermore, the potency and effects on the two cell types may be different, thus explaining the dose-dependent effects observed in this study.

Our findings could also be explained by the presence of chronic low-grade inflammation associated with atherosclerosis. Chronic inflammatory conditions are often accompanied by bone loss and subsequently osteoporosis, and the inflammation related to cardiovascular disease could thus be a contributing factor to the bone loss in the clopidogrel-treated group.

In contrast to the findings of the present study, we previously failed to demonstrate an association between clopidogrel use and fracture risk [7]. This may be explained by study differences with respect to clopidogrel. As clopidogrel was granted marketing authorization in July 1998 and the analysis in the previous study was based on fracture data collected in the year 2000, patients could only have had a brief duration of exposure to the drug at the time of data collection. In the present study, which was designed as a cohort study with long-term follow-up for most subjects, longer exposure times were possible. Furthermore, significantly more clopidogrel-treated subjects were included in the present study. Another reason for the potentially longer exposure times compared with the previous study is that in recent years guidelines have changed so that a treatment duration of up to 1 year is now recommended for patients with ACS and those with drug-eluting stents, and of more than 1 year for patients with a high risk of thromboembolic events [22].

The most important strength of our study is the large-scale population-based design. The collection of data from a cohort of this size is possible because of the centrally organized healthcare system and the unique identification number for each citizen in Denmark. Furthermore, collection of data on confounders and exposure before the occurrence of fracture, so that recall bias had no influence, also strengthens the data. Selection bias is also minimized as all patients prescribed clopidogrel were included, together with information on all fractures. This was possible as all collected prescriptions are registered with the Danish Medicines Agency and practically all fractures are treated in public emergency departments and hospitals that are covered by the database. However, the possibility of some coding errors by the discharging doctors cannot be excluded.

The study also has some limitations. First, the optimal design would have been a prospective, randomized trial. However, it would not be practically or economically feasible to conduct such a study, especially with a study population this large. Additionally, it would not be ethical to include a placebo-controlled group in this patient population, as clopidogrel is part of the recommended standard regimen for treatment of several types of cardiovascular disease. Secondly, despite the fact that we adjusted for several variables, the results may still be influenced by other potential confounding factors that were not included in the analyses, such as differences in body weight, physical activity, intake of calcium and vitamin D and smoking habits. As in all observational studies, the risk of residual confounding is present because of possible differences in comorbidities. Therefore, we have extensively controlled for all known major health determinants related to fractures. Finally, information bias may have influenced the results, although most probably to a similar extent amongst all treatment groups. We had no information regarding compliance. However, most patients pay for part of their medication so it seems reasonable to expect that they would use a drug that had been collected and paid for at the pharmacy.

In conclusion, treatment with the widely used platelet inhibitor clopidogrel at doses recommended for the treatment and prevention of cardiovascular events is associated with an increased risk of fractures. By contrast, suboptimal doses of clopidogrel are associated with a decreased fracture risk. The physiological importance of the P2Y12 receptor in bone metabolism and the pathophysiological mechanisms whereby clopidogrel affects bone turnover remain largely unknown. Further studies are warranted to elucidate these mechanisms and investigate the potential clinical importance of our findings. Additionally, it is possible that the newer and more potent antiplatelet drugs such as prasugrel and ticagrelor could have even greater effects on fracture risk. Future studies should investigate the effects of ticagrelor and prasugrel on bone metabolism and the risk of fractures.

Conflicts of interest

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Acknowledgements
  9. References

None of the authors has any conflicts of interest to declare.

Acknowledgements

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Acknowledgements
  9. References

This work was supported by the European Commission under the 7th Framework Programme (proposal #202231) as a collaborative project amongst the members of the ATPBone Consortium (Copenhagen University, University College London, University of Maastricht, University of Ferrara, University of Liverpool, University of Sheffield and Université Libre de Bruxelles) and is a substudy of the main study ‘Fighting osteoporosis by blocking nucleotides: purinergic signalling in bone formation and homeostasis’. This work was also supported by a grant from the Jacob and Olga Madsen Foundation.

References

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Acknowledgements
  9. References