Clopidogrel: a pharmacoeconomic review

Summary

Recent randomised, prospective studies have demonstrated the clinical efficacy and safety of clopidogrel in the treatment of patients with acute coronary syndromes, including those treated with primary percutaneous intervention. In these settings clopidogrel decreased the risk of atherothrombotic events by 20–30%.

The use of clopidogrel instead of aspirin, for secondary prevention in patients with stable cardiovascular disease, results in a modest 8.7% relative risk reduction. However, clinical trial failed to show benefit in patients treated with clopidogrel added to aspirin in this population.

Economic analyses conducted in the US and Europe demonstrated the cost effectiveness of clopidogrel in combination with aspirin in patients with acute coronary syndromes, and showed a favourable economic profile for clopidogrel compared with other cardiovascular interventions. On the other hand, the data for the use of clopidogrel in patients with stable cardiovascular disease either alone or in combination with aspirin is not as compelling.

Keywords::

• clopidogrel
• cost effectiveness
• acute coronary syndrome

Introduction

Atherothrombosis is the leading cause of death worldwide1. Platelets play an important role in atherothrombosis2 and are commonly targeted by drugs from several classes as part of prevention and treatment strategies in patients with atherothrombotic diseases.

Thienopyridine antiplatelet agents, including clopidogrel, are prodrugs that, when metabolised, achieve its antiplatelet effects by inhibiting the binding of ADP to its G-protein coupled P2Y12 receptor. When active, the P2Y12 receptor inhibits platelet disaggregation, an effect mediated by intracellular cyclic AMP. Inhibition of this receptor, with the resultant impairment in platelet aggregation, represents the mechanism of action of clopidogrel3,4, a distinct mechanism from that of aspirin, which inhibits the cyclooxygenase pathway. The clinical efficacy and safety of clopidogrel, in addition to standard therapy including aspirin, in patients with atherothrombotic disease has been evaluated in multiple clinical trials5–13.

These trials enrolled patients with different forms of atherothrombotic disease, including patients with acute coronary syndromes (ACS; unstable angina, non-ST elevation myocardial infarction (NSTEMI) and ST elevation myocardial infarction (STEMI)), those undergoing percutaneous coronary intervention (PCI) or those with chronic atherothrombotic disease.

As a result of these trials the American College of Cardiology/American Heart Association and the European Society of Cardiology Guidelines for the management of patients with ACS recommended the use of clopidogrel14–17. Clopidogrel is an acceptable alternative for patients with aspirin allergy. In patients with ACS it is recommended to continue clopidogrel for at least 1 month and preferably 1 year in patients treated conservatively and in those who receive a bare metal stent. In patients with a drug eluting stent it is recommended to continue clopidogrel for at least 1 year.

It is, therefore, important to assess the economic effect of clopidogrel use. This review discusses the clinical and economic benefits of clopidogrel in the different forms of atherothrombotic disease.

Clinical effectiveness of clopidogrel

The design and results of clinical effectiveness trials of clopidogrel are summarised in Table 1 and 2, respectively.

Table 1. Summary of clopidogrel clinical effectiveness trials: design and end points.

Study	Population (n)	Design
CURE^5	UA and NSTEMI (12,562)	ASA vs. ASA and clopidogrel (75 mg) for average of 9 months
PCI-CURE^6	UA and NSTEMI with PCI (2,658)	ASA vs. ASA and clopidogrel pretreatment (300 mg) and long term (75 mg) for average of 8 months^*
CREDO^7	Patients undergoing elective PCI (2,116)	ASA vs. ASA and clopidogrel pretreatment (300 mg) and long term (75 mg) for average of 12 months^*
CLARITY-TIMI 28^8	STEMI treated with fibrinolytics (3,491)	ASA vs. ASA and clopidogrel 300-mg loading dose followed by 75 mg/day
CLARITY-PCI^9	STEMI treated with fibrinolytics then with PCI^† (1,863)	ASA vs. ASA and clopidogrel 300-mg loading dose followed by 75 mg/day until the time of angiography followed by long-term open-label use of clopidogrel in both groups after PCI
COMMIT^10	STEMI (50% fibrinolytic use, no PCI; 46,000)	ASA vs. ASA and clopidogrel 75 mg/day for up to 4 weeks
CAPRIE^11	Patients with recent ischaemic stroke, recent MI, or symptomatic peripheral arterial disease (19,185)	ASA vs. clopidogrel (follow-up for up to 3 years)
CHARISMA^12	High-risk patients without known cardiovascular disease or stable cardiovascular disease (15,603)	ASA vs. ASA and clopidogrel (followed for median of 28 months)

UA, unstable angina; NSTEMI, non-ST elevation myocardial infarction; PCI, percutaneous coronary intervention; MI, myocardial infarction; STEMI, ST elevation myocardial infarction, ASA: aspirin.

^* All patients received open-label thienopyridine for 28 days post-PCI.

^† Patients underwent angiography 48–192 hours after start of study. PCI was performed at the discretion of the investigators.

Table 2. Summary of clopidogrel clinical effectiveness trials: results.

		Results
Study	End point	Clopidegrol	ASA	RRR/OR	p
CURE^5	Cardiovascular death, nonfatal MI or stroke (mean 9 months)	9.3%	11.4%	0.20%	<0.0001
PCI-CURE^6	Cardiovascular death, MI or target vessel revascularisation at 30 days	4.5%	6.4%	0.30	<0.0001
	Cardiovascular death and MI (mean 8 months)	8.8%	12.6%	0.31	<0.002
	Cardiovascular death, nonfatal MI or stroke^* (mean 8 months)^34	9.6%	13.2%	0.27	(95% CI 0.10–0.41)
CREDO^7	Death, MI or stroke (12 months)	8.5%	11.5%	0.27	0.02
CLARITY-TIMI 28^8	Cardiovascular death, MI or occluded infarct-related artery (at time of angiography)^†	15%	21.7	0.36	<0.001
	Cardiovascular death, MI or recurrent ischaemia leading to urgent revascularisation (30 days)	11.6%	14.1%	0.20	0.03
CLARITY-PCI^9	Cardiovascular death, MI or stroke (30 days)	7.5%	12%	0.31	0.001
COMMIT^10	Death, reinfarction or stroke (at time of discharge)	9.2%	10.1%	0.09	0.002
	All-cause death (at time of discharge)	7.5%	8.1%	0.07	0.03
CAPRIE^11	First occurrence of ischaemic stroke, MI, or vascular death (up to 3 years)	5.32%/year	5.83%/ year	0.087	0.043
CHARISMA^12	Cardiovascular death, MI, or stroke (mean 28 months)	6.8%	7.3%	0.07	0.22

MI, myocardial infarction; ASA, aspirin; RRR, relative risk reduction; OR, odds reduction; CI, confidence interval.

^* Stroke was not part of the primary end point.

^† Patients underwent angiography 48–192 hours after start of study. PCI was performed at the discretion of the investigators.

CURE trial

The Clopidogrel in Unstable Angina to Prevent Recurrent Events (CURE) trial evaluated the efficacy and safety of clopidogrel plus aspirin compared with aspirin alone when initiated early and continued long term (average 9 months) in patients presenting with unstable angina or a NSTEMI5. The CURE trial enrolled 12,562 patients at 482 centres in 28 countries.

The primary end point, composite of cardiovascular death, nonfatal myocardial infarction (MI) or stroke, occurred in 9.3% of patients in the clopidogrel arm and 11.4% in the placebo group (relative risk (RR) 0.80; 95% confidence interval (CI) 0.72–0.90; p<0.0001). The absolute risk reduction was 2.1% and the number needed to treat (NNT) to prevent one event was 48 patients.

PCI-CURE study

Percutaneous Coronary Intervention Substudy of Cure (PCI-CURE)6 is a prospective study of 2,658 patients, who underwent PCI at the discretion of the treating physician, from the CURE trial of pretreatment and long-term therapy with clopidogrel vs. no pretreatment, and only up to 4 weeks of open-label therapy with thienopyridine. A 300-mg loading dose of clopidogrel or matching placebo was administered immediately after randomisation. This was followed by 75-mg clopidogrel or matching placebo. After PCI, stented patients received 2–4 weeks of open-label thienopyridine in combination with aspirin. Subsequently, the randomly assigned study medication was resumed for a mean of 8 months.

At 30 days after PCI, the primary composite end point of cardiovascular death, MI or target vessel revascularisation occurred in 4.5% of the clopidogrel arm and 6.4% of the placebo arm (RR 0.70; 95% CI 0.50–0.97; p=0.03). Cardiovascular death or MI occurred in 2.9% of the clopidogrel arm and 4.4% of the placebo arm (RR 0.66; 95% CI 0.44–99; p=0.04). The clopidogrel group had superior long-term outcomes in terms of MI or cardiovascular death compared with placebo (8.8 vs. 12.6%; RR 0.69; 95% CI 0.54–0.87; p=0.002). The absolute RR reduction was 3.8% over the duration of the study and the NNT was 26 patients. Although stroke was not part of the primary end point, it affects life expectancy and cost. If cardiovascular death, MI and stroke were used as the composite end point the clopidogrel arm would still have had significantly lower events (9.6 vs. 13.2%), and the NNT would be 28³⁴.

CREDO trial

In the Clopidogrel for Reduction of Event during Observation (CREDO) trial7 2,116 patients undergoing elective PCI were randomly assigned to 300-mg clopidogrel loading dose plus aspirin 325 mg vs. aspirin alone 3–24 hours before PCI. At the time of PCI all patients were given clopidogrel 75 mg and aspirin 325 mg for 28 days. Subsequently, the study drug was resumed according to the initial randomisation and continued for 12 months. The primary end point of death, MI or stroke at 1 year occurred in 8.5% of the clopidogrel arm and 11.5% of the placebo arm (RR reduction 26.9%; 95% CI 3.9–44.4%; p=0.02). Absolute risk reduction was 3.0 and the NNT was 33.

CLARITY-TIMI 28 trial

In the Clopidogrel as Adjunctive Reperfusion Therapy (CLARITY-TIMI 28) trial8 3,491 patients with STEMI were randomised to a clopidogrel 300-mg loading dose, clopidogrel 75 mg/day or placebo in addition to aspirin. The choice of fibrinolytic agent and anticoagulant were at the discretion of the treating physician. Patients underwent angiography 48–192 hours after the start of study.PCI was performed at the discretion of the investigator.

The primary end point, a composite of cardiovascular death, recurrent MI or occluded infarct-related artery at the time of angiography, occurred in 15% of the clopidogrel group and 21.7% in the placebo group (36% reduction in odds; 95% CI 24–47%; p<0.001). The absolute RR reduction was 6.7% and the NNT was 15. At 30 days, cardiovascular death, MI or recurrent ischaemia leading to urgent revascularisation occurred in 11.6% of the clopidogrel arm vs. 14.1% in the placebo arm (20% reduction in odds, p=0.03). Absolute risk reduction was 2.5% and the NNT was 40 patients.

CLARITY-PCI study

Among the 3,491 patients in the overall CLARITY-TIMI 28 trial9, 1,863 (53.4%) underwent PCI during the index hospitalisation at a median of 3 days after randomisation. An open-label thienopyridine loading dose was given at the time of PCI, followed by a maintenance dose in the majority of patients (77.6 and 90%, respectively). Overall, at 30 days the rate of cardiovascular death, MI or stroke was lower in the clopidogrel arm (7.5 vs. 12.0%; odds ratio 0.59; 95% CI 0.43–0.81; p=0.001).

COMMIT

In the large Chinese Clopidrogel and Metoprolol in Myocardial Infarction Trial (COMMIT), 46,000 patients with STEMI were randomised to up to 4 weeks of treatment with clopidogrel (75 mg/day) or placebo in addition to aspirin 162 mg/day10. Fibrinolytics were used in 50% of patients and those treated with primary PCI were excluded. The primary end point of death, reinfarction or stroke at the time of discharge from hospital was lower in the clopidogrel group compared with placebo (9.2 vs.10.1%; p=0.002). All-cause mortality at discharge was also lower in the clopidogrel group (7.5 vs. 8.1%; p=0.03).

CAPRIE trial

In the Clopidogrel vs. Aspirin in Patients at Risk of Ischaemic Events (CAPRIE) trial11, 19,185 patients with recent ischaemic stroke, recent MI, or symptomatic peripheral arterial disease at 384 centres in 16 countries were randomised to clopidogrel (75 mg/day) or aspirin (325 mg/day) and were followed long term (up to 3 years). The primary end point was the first occurrence of an event in the outcome cluster of ischaemic stroke, MI, or vascular death. There were 939 events in the clopidogrel group during 17,636 patient-years at risk, an average rate per year of 5.32%, and there were 1,021 events in the aspirin group during 17,519 patient-years at risk, an average rate per year of 5.83%. These rates reflect a statistically significant (p=0.043) RR reduction of 8.7% (95% Cl 0.3–16.5) in favour of clopidogrel.

CHARISMA trial

In the Clopidogrel for High Atherothrombotic Risk, Ischaemic Stabilization, Management, and Avoidance (CHARISMA) trial12, 15,603 high-risk patients without known cardiovascular disease or stable cardiovascular disease were randomised to either clopidogrel (75 mg/day) or placebo on top of background therapy with aspirin (75–162 mg/day). The primary end point was a composite of cardiovascular death, MI, or stroke. There was not a statistically significant difference in the primary end point of cardiovascular death, MI, or stroke between the clopidogrel plus aspirin group and the placebo plus aspirin group (6.8 vs. 7.3%; RR 0.93; p=0.22).

Although no statistically significant benefit was found in the overall broad population of stable patients studied, the rate of cardiovascular death, MI, or stroke was significantly lower in the clopidogrel plus aspirin arm than in the placebo plus aspirin arm (7.3 vs. 8.8%; HR 0.83; 95% CI 0.72–0.96; p=0.01) in the subgroup of patients in the CHARISMA trial who were enrolled with documented prior MI, ischaemic stroke, or symptomatic peripheral arterial disease13.

Cost effectiveness of clopidogrel

The cost effectiveness of clopidogrel in the prevention of cardiovascular events in patients with cardiovascular disease including those with ACS has been evaluated in a number of analyses. There are two ways of performing an economic evaluation based on the results of a clinical trial.

The first method is to perform ‘within-trial analysis’ using data collected prospectively in the trial. This has the advantage that all the data are taken from the same source and are consistent. Some of the drawbacks are that the trial may have short follow-up and the study population may differ from the general population.

The second method is to estimate the long-term cost effectiveness using data from clinical trials combined with estimates of life expectancy from various databases or incorporating the clinical risk reduction observed in the trial in a simulation, such as a Markov model. Markov models consist of a number of discrete health states representing the disease with which costs and effects are associated. Disease progression or occurrence is modelled as transitions between states over time. The probability of transition from state to state depends on patient characteristics, but not on the previous events. The advantage of the long-term analysis is that the consequences for life expectancy can be calculated for relevant patient populations at different levels of risk18,19.

CURE trial

Within-trial analysis

During the CURE trial the resources used were prospectively collected for the mean 9-month study duration. This included study drug, hospitalisation, procedures and concomitant medication. Hospitalisation costs were evaluated through a diagnosis-related group (DRG) approach, when available, and local country costs were applied.

Lamy et al20 estimated the incremental cost effectiveness of clopidogrel based on the CURE trial in five countries (Canada, France, Sweden, the US and the UK). The average cost during the initial hospitalisation was reduced for the clopidogrel group in all countries. Total costs during the follow-up period, including subsequent hospitalisation and medication costs, were marginally higher in the clopidogrel group.

Annemans et al21 reported results for five European countries (Belgium, Italy, The Netherlands, Spain and Switzerland). In all countries there were savings during the initial hospitalisation in the clopidogrel arm as fewer complications and fewer procedures offset the additional cost of clopidogrel. Over 9 months, treatment with clopidogrel plus aspirin was associated with an incremental cost per patients compared with aspirin alone. However, as clopidogrel resulted in 2.1% absolute risk reduction, the resulting incremental cost per event avoided ranged from €4,790 in The Netherlands to €19,484 in Spain.

Other analysis in different countries including the study by Lindgren et al22 (Denmark, Finland, Norway and Sweden), Pietrasik et al23 (Poland) and Belousov et al24 (Russia) reached similar conclusions as shown in Table 3.

Table 3. Within-trial analysis of cost effectiveness of clopidogrel based on the CURE trial.

Study	Country	Incremental cost/CV event avoided
Lamy et al^20	Canada	CAD$7,973 (€5,754)^*
	France	€16,186
	Sweden	SEK127,951 (€13,980)^*
	UK	£10,366 (€14,978)^*
	US	US$22,484 (€15,929)^*
Annemans et al^21	Belgium	€18,149
	Italy	€18,348
	The Netherlands	€4,790
	Spain	€19,484
	Switzerland	€15,652
Lindgren et al^22	Denmark	€14,447
	Finland	€24,700
	Norway	€13,391
	Sweden	€15,652
Pietrasik et al^23	Poland	€23,076
Belousov et al^24	Russia	US$27,000 (€–19,139)^*

CV, cardiovascular; CAD, Canadian dollar; SEK, Sweden Kroner.

^* Converted from the published currency to Euro at the rate on the 9^th October 2007.

Long-term cost effectiveness

Since the main outcomes of the study (MI and stroke) have long-term consequences on both survival and cost, this type of analysis is performed to assess the lifetime benefits of treatment with clopidogrel. Weintraub et al25 estimated the long-term cost effectiveness of clopidogrel from a healthcare payer perspective based on within-trial efficacy and resource-use data combined with life expectancy estimates from the Framingham Heart Study and the Saskatchewan Health database. Costs included in the analysis were direct medical costs for hospitalisation and the cost of drugs. The initial and subsequent hospitalisations were assigned a DRG. Costs for each DRG were estimated using average Medicare reimbursement rates for patients >65 years of age and average private payer reimbursement rates, which were obtained from the MEDSTAT database for patients <65 years of age. Costs beyond the trial period were estimated as the average per capita participant Medicare reimbursement of $4,370 in 2001.

Life expectancy estimates for patients with and without an event were estimated from two sources: the Framingham Heart Study and the Saskatchewan Health database. The cost effectiveness of clopidogrel was expressed as an incremental cost-effectiveness ratio (ICER). Overall, patients in the clopidogrel arm were estimated to have gained an average of 0.0699 life years relative to the control arm using the Framingham data and 0.0682 life years using the Saskatchewan data.

When including clopidogrel costs ($766 greater for the clopidogrel arm), average total costs were $442 higher for the clopidogrel arm (95% CI $62–$820). The ICER was <$10,000 (Table 4) whether using Medicare, MEDSTAT or a blend approach for costing, and whether using the Framingham Heart Study or Saskatchewan database for survival data. More than 90% of estimates were <$50,000.

Table 4. Long-term cost effectiveness of clopidogrel based on the CURE trial: US analysis²⁵.

	Cost	Life years	ICER	% <50,000/life years gained
Framingham
Medicare	$639	0.0699	$9,144	92.8%
MEDSTAT	$535	0.0699	$7,654	93.4%
Blend	$541	0.0699	$7,742	93.4%
Saskatchewan
Medicare	$637	0.0682	$9,343	97.0%
MEDSTAT	$534	0.0682	$7,833	97.6%
Blend	$540	0.0682	$7,921	97.5%

ICER, incremental cost-effectiveness ratio.

Lindgren et al26 estimated the ICER of clopidogrel on top of aspirin in patients with ACS without ST-segment elevation using a Markov model. The model developed for this analysis had six health states (at risk, first year with stroke, following years with stroke, first year with new MI, following years with new MI, and death). Patients start in the ‘at risk’ state where they receive clopidogrel during the first year. They are followed as they move from one state to another over particular time intervals called cycles, in this case 12 months. The transition probabilities were calculated using three main data sources: the Swedish Hospital Discharge Register and the Cause of Death Register were used to calculate the risk of suffering an event and the mortality both prior to and following an event. The risk reduction from treatment was taken from the CURE trial.

Two scenarios were analysed: one with a patient population similar to the CURE trial (61.3% male, mean age of 64.2 years) and one similar to the Swedish population extracted from the registers (62.2% males, mean age of 68.4 years). Costs (direct and indirect) caused by different events were taken from published sources, whereas the cost for the initial hospitalisation was estimated on the basis of data from the CURE study. Indirect costs included costs associated with lost productivity after MI or stroke. In the first scenario, the predicted net direct cost was €160 and the net total cost of –€54, which with incremental survival of 0.12 years (9.71 vs. 9.65 years) give an ICER of €1,365 per life gained from the healthcare payer perspective (direct cost) and cost savings from the societal perspective (including indirect cost).

In the second scenario, using patient population with risk factors similar to the Swedish population extracted from the registers, the net cost was €149 (no indirect cost was present as the population was older than 65 years), with incremental survival of 0.15 years (8.04 vs. 7.89 years) giving an ICER of €1,009 for both perspectives.

Similar analyses22,26–29 in other countries using the same model using the CURE trial data have also predicted that clopidogrel on top of standard treatment with aspirin is cost effective (Table 5).

Table 5. Long-term cost effectiveness of clopidogrel plus aspirin vs. aspirin based on the CURE trial: European analyses.

Study	Country	Net direct cost	Net total cost	Cost/life year gained (direct)	Cost per life year gained (total)
Lindgren et al^26	Sweden	€160	–€54	€1,365	Dominant
Lindgren et al^22	Denmark	€64	–€145	€549	Dominant
	Finland	€510	€296	€5,048	€2,935
	Norway	€137	–€82	€1,171	Dominant
	Sweden	€160	–€54	€1,365	Dominant
Schwartz et al^27	Austria				€4,146
Badia et al^28	Spain	€953		€8,132
Bruggenjurgen et al^29	Germany			€3,113
Van Hour et al^30	The Netherlands	–€17		Dominant

Van Hout et al30 reported results of another analysis based on the CURE trial. They demonstrated that clopidogrel on top of standard therapy results in an annual cost saving of €17. There was a gain of 0.122 life years and 0.137 quality-adjusted life years (QALYs)

PCI-CURE study

Within-trial analysis

Similar within-trial analysis as described above were conducted in the subgroup of patients who underwent a PCI. During the trial the resources used were prospectively collected for the study duration including study drug, hospitalisation, procedures and concomitant medication. Hospitalisation costs were evaluated through a DRG approach, when available, and local country costs were applied.

Overall, the clopidogrel arm was associated with an excess cost per patients over 8 months compared with the aspirin alone arm, except in The Netherlands. However, the ICER was within an economically favourable range. It was cost saving in The Netherlands and ranged from €4,732 in Switzerland to $12,000 in the US (Table 6)31–33.

Table 6. Incremental cost per patient and per event avoided with clopidogrel plus aspirin vs. aspirin alone based on PCI-CURE trial31–33.

Country	Net cost	Incremental cost/event avoided
Austria	€385	€11,000
Belgium	€346	€9,851
Italy	€279	€7,931
Spain	€390	€11,065
Switzerland	€166	€4,732
The Netherlands	–€62	Cost saving
France	€323	€9,229
Sweden	SEK3,062 (€335)^*	SEK87, 486 (€9,558)^*
US	US$420 (€298)^*	US$12,000 (€8,503)^*
Canada	CAD$215 (€155)^*	CAD$6,143 (€4,432)^*
UK	£180 (€260)^*	£5,143 (€7,431)^*

CAD, Canadian dollar; SEK, Sweden Kroner.

^* Value between parentheses is cost per event converted to Euro at the rate on the 9th October 2007.

Long-term cost effectiveness

The long-term cost effectiveness of clopidogrel in the US using data from PCI-CURE was evaluated by Mahoney et al34. The analysis was similar to that conducted for the main CURE study. Costs included in the analysis were direct medical costs for hospitalisation and cost of the drugs. The initial and subsequent hospitalisations were assigned a DRG. Costs for each DRG were estimated using average Medicare reimbursement rates for patients >65 years of age and average private payer reimbursement rates, which were obtained from the MEDSTAT database for patients <65 years of age.

Average total costs were higher with clopidogrel. The estimated life expectancy gain with clopidogrel was 0.0885 years. Incremental cost per life year gained with clopidogrel ranged from $2,856 to $4,775. For patients with PCI at initial hospitalisation the difference in cost ranged from $188 lower to $90 higher with clopidogrel. The estimated life expectancy gain with clopidogrel was 0.0962 years. Incremental cost per life year gained with clopidogrel ranged from dominant to $935.

Lindgren et al35 used a Markov model using data from PCI-CURE, similar to the analysis carried out for the CURE study. The clopidogrel arm incurred additional direct costs of €449 and €332 in total cost. Clopidogrel addition to aspirin resulted in 0.04 life years gained. ICER was €10,993 (from health payer perspective) and €8,127 (from societal perspective).

In summary, both short- and long-term analysis conducted in Europe and North America demonstrate that clopidogrel is cost effective in the setting of ACS including patients who undergo PCI.

CREDO trial

Beinart et al36 used clinical outcome data and resources from the CREDO trial to estimate the long-term cost effectiveness of clopidogrel vs. placebo before and after PCI. All hospitalisations were assigned a DRG. Associated costs were estimated in three ways: Medicare costs, MEDSTAT costs and a blend with Medicare for those aged ≥65 years and MEDSTAT for those aged <65 years. Life expectancy in trial survivors was estimated using external data.

The number of life years gained with clopidogrel was 0.1526 using Framingham data and 0.1920 using Saskatchewan data. Average total costs were $664 higher for the clopidogrel arm. The ICERs based on Framingham data ranged from $3,685 to $4,353 per life years gained, with over 97% of bootstrap-derived ICER estimates <$50,000/life years gained. The ICERs based on Saskatchewan data ranged from $2,929 to $3,460/life years gained, with over 98% of estimates <$50,000/life years gained. Using a ceiling ratio of US$18,000 over 90% of simulations will be cost effective.

Cowper et al37 used decision analytic methods to compare the outcomes and cost of prolonging clopidogrel treatment from 1 month to 1 year after PCI with the alternative strategy of discontinuing therapy 1 month after the procedure. Event rates were based on 3,976 PCI patients who were treated at the Duke Medical Center and received no more than 1 month of clopidogrel after the procedure. Event rates were obtained from Duke clinical information systems. The effect of prolonged clopidogrel therapy on event rates was based on CREDO trial data. Unit costs and the effect of MI on life expectancy were based on published sources. Extending clopidogrel therapy from 1 month to 1 year after PCI cost $879 per patient and reduced the risk of MI by 2.6%. Assuming MI decreases life expectancy by 2 years, prolonged therapy would cost $15,696 per year of life saved.

Ringborg et al38 studied the long-term cost effectiveness of clopidogrel in Sweden based on the CREDO trial. A Markov model was developed, which assumed a hypothetical cohort of patients in a post-PCI state to have certain risks of suffering one of the end points of the CREDO trial: stroke, MI, or death. The model predicted a mean survival of 12.098 years in the 12-month arm vs. 12.026 in the 28-day arm, an incremental gain of 0.072 life years. The gain in survival came at a predicted incremental cost of €217, resulting in an ICER of €3,022/life years gained.

Overall, platelet inhibition with clopidogrel loading before PCI followed by therapy for 1 year is highly cost effective.

CAPRIE trial

Schleinitz et al39 constructed a Markov model based on the results from the CAPRIE trial, assuming a societal perspective. In patients with peripheral arterial disease, clopidogrel increased life expectancy by 0.55 QALY at an ICER of $25,100 per QALY, as compared with aspirin. In post-stroke patients, clopidogrel increased life expectancy by 0.17 QALY at a cost of $31,200 per QALY. However in post-MI patients, aspirin was less expensive and more effective.

Haldemann et al40 conducted a cost-effectiveness analysis from the perspective of Swiss third-party payers. The discounted ICER was 22,837 Swiss Franc/life years gained (≈ €14,000/life years gained). The long-term cost effectiveness of clopidogrel for Belgium was evaluated from a healthcare payer's perspective41; the use of clopidogrel instead of aspirin for 2 years was associated with an ICER of €13,390/life years gained. In The Netherlands42 treatment with clopidogrel for 1 year resulted in an ICER of €19,462/life years gained.

Gaspoz et al43 used the coronary heart disease policy model, a computer simulation of the US population, to estimate the ICER of different strategies of antiplatelet therapy for 25 years in patients >35 years of age with coronary artery disease. Clopidogrel alone in all patients or in routine combination with aspirin had an incremental cost of >$130,000 per QALY.

COMMIT and CLARITY-TIMI 28 trial

Zhang et al⁴⁴ performed cost-effectiveness analysis based on COMMIT. The number of initial hospitalisations for death, nonfatal MI and PCI within 28 days was estimated based on COMMIT. The CURE trial was used to estimate the number of subsequent hospitalisations for death, nonfatal MI, stroke, bleeding, angina and revascularisation between 29 days and 1 year. Hospitalisations were assigned a DRG. Costs for each DRG (year 2003) were estimated from Medicare reimbursement rates. Physician costs were estimated as a percentage of hospital costs by DRG. Clopidogrel was assumed to be given as 1 year and priced at $3.80/day. Life expectancy gains as a result of the prevention of the major clinical events of death, MI and stroke were estimated using Framingham data. Cost-effectiveness analysis was performed as incremental costs per life year gained.

Within 28 days, adding clopidogrel to aspirin is likely to be a dominant strategy, lowering event rate without an increase in costs. For the lifetime period, the ICER of clopidogrel is $4,896/life years gained. Sensitivity analyses, which were performed assuming an additional increase or decrease of 10, 20, 30 and 40% based on the estimated reduction rates of death, MI and stroke for clopidogrel, showed that the ICERs range from $4,070/life years gained to $6,034/life years gained.

Berg et al45 used a combined decision tree and Markov model to assess the cost effectiveness of clopidogrel in the short- and long-term treatment of STEMI in Sweden, Germany and France. Data from the long-term NSTEMI CURE trial were combined with 1-month data from CLARITY and COMMIT to model the effect of treatment for up to 1 year. The risks of death, MI, and stroke in an untreated population and long-term survival after all events were derived from the Swedish Hospital Discharge and Cause of Death register. The model was run separately for the two STEMI trials.

In a patient cohort with the same characteristics and event rates as in the CLARITY population, this strategy was dominant in Sweden and France with estimated cost savings of €111 and €367, respectively. In Germany, clopidogrel treatment had an ICER of €92/life years gained. Based on data from COMMIT the corresponding ICERs were €2,772/life years gained, €4,144/life years gained and €2,786/life years gained in Sweden, Germany and France, respectively.

The addition of clopidogrel to aspirin, given for up to 1 year, in the setting of STEMI is a highly cost-effective strategy.

CHARISMA trial

No cost-effectiveness data was available from this trial.

Cost effectiveness of clopidogrel compared with other cardiovascular interventions

Compared with other types of interventions, treatment with clopidogrel is attractive according to the standards of cost effectiveness. In addition, the cost effectiveness of clopidogrel seems to be highly competitive with most medication and intervention used for ACS patients, including beta-blockers, angiotensin-converting enzyme inhibitors, lipid-lowering therapy and glycoprotein IIb/IIIa inhibitors46–54 (Table 7).

Table 7. Reported long-term cost-effectiveness ratios for cardiovascular interventions.

Intervention	Cost-effectiveness ratio per life year gained
Smoking cessation post-MI^46	US$220
Lisinopril post-MI^47	US$2,080
Perindopril in stable CAD^48	US19,816^*
Beta-blocker post-MI^49	US$4,500
Statin (simvastatin and pravastatin)50,^51	US$5,400–US$3,200
Clopidogrel for PCI: CREDO^36	US$2,929–US$4,353
Clopidogrel for PCI in ACS: PCI-CURE^34	US$2,856–US$4,775
Clopidogrel for ACS: CURE^25	US$7,654–US$9,144
Tirofiban for early invasive ACS^52	US$12,739
Eptifibatide for ACS^53	US$16,491
Alteplase versus streptokinase^54	US$32,678

ACS, acute coronary syndrome; MI, myocardial infarction; PCI, percutaneous coronary intervention.

^* Per quality-adjusted life year.

Data from other countries are also consistent with the US data as the cost per life years gained compares favourably with other cardiovascular interventions, whether including direct cost (healthcare payer perspective) or total cost (societal perspective).

Discussion

Pharmacoeconomic analyses, including those carried out for clopidogrel, can be imprecise as it relies on assumptions and extrapolation of data from many sources, including clinical trial results, to the general population. Limitations to this approach include not only the limitation of the economic analysis methods, but also the limitations of the clinical studies.

The clinical effectiveness of clopidogrel has been documented from a number of well-designed clinical trials, although with varying designs and end points5–11. However, data from clinical trials might not be generalisable to all patients with ACS given the strict inclusion and exclusion criteria of most clinical trials. Furthermore, the management of patients in clinical trials can be either more thorough than contemporary standards of care, or may become outdated as the standards change and improve. Nonetheless, the benefit of clopidogrel has been consistently demonstrated in these different trials and within the different subgroups in the same trial.

For within-trial analyses, clinical outcome and resource utilisation for each country has generally been calculated based on the trial-wide clinical outcome, resource utilisation and unit costs in individual countries. This approach assumes a uniform treatment benefit between different countries. However, practice style may differ between different countries. For the long-term analyses, life expectancy was estimated from external databases; data from these databases might not reflect the risks and life expectancies in patients managed using current standards of care.

In spite of these limitations the cost effectiveness of clopidogrel has been widely studied using different and generally well-conducted trials using different methodologies in many countries. Although the exact ICER may differ from one study to another, clopidogrel has been found to be cost effective regardless of the methodology or country. Furthermore, sensitivity analyses consistently prove the results to be robust and well below societal willingness to pay thresholds.

Conclusions

Clinical data from the CURE, PCI-CURE and CREDO studies show that clopidogrel significantly reduces the risk of cardiovascular events on top of standard therapy for patients with ACS (unstable angina and NSTEMI) and in patients undergoing PCI. Economic analyses across a range of countries using various methodologies show that the cost-effectiveness profile of clopidogrel is consistent and favourable in this group of patients. Furthermore, the cost-effectiveness ratio for clopidogrel compares favourably with the other cardiovascular interventions including beta-blockers, angiotensin converting-enzyme inhibitors and glycoprotein IIb/IIIa inhibitors.

The benefit of using clopidogrel instead of aspirin for secondary prevention was tested in the CAPRIE trial. Clopidogrel resulted in a modest 8.7% RR reduction in the combined end point. The cost-effectiveness ratio of clopidogrel as an alternative to aspirin for secondary prevention was less favourable compared with that in patient with ACS.

In patients with STEMI, clopidogrel use was tested in two trials (CLARITY-TIMI 28 and COMMIT). These show that clopidogrel significantly reduces the risk of cardiovascular events on top of standard therapy. The addition of clopidogrel to aspirin, given for up to 1 year, in the setting of STEMI is a highly cost-effective strategy.

In conclusion, clopidogrel has proven in multiple clinical trials to be highly effective in reducing cardiovascular events in a wide variety of patients across the spectrum of cardiovascular disease, including patients with ACS (STEMI, NSTEMI and unstable angina), as well as for secondary prevention. While recognising the limitation of comparing economic evaluations in different populations, the favourable cost-effectiveness profile of clopidogrel is highly competitive with other cardiovascular treatments.

Acknowledgements

Declaration of interest: William Weintraub has grant support from Sanofi-Aventis and Bristol Myers Squibb. Nowwar Mustafa has no relationships to disclose.