The Pharmaceutical
Journal Vol 267 No 7172 643-649
3 November 2001
|
|
The Pharmaceutical
Journal Vol 267 No 7172 643-649 |
|
Supervision
|
TrialsClarification, please, for CURE studyFrom Mr M. I. Hird MRPharmS The points made by Dr Flather and Dr Bakhai (PJ, 13 October, p509) about the CURE study1 require clarification. Given their estimate of 120,000 admissions due to acute coronary syndrome (ACS) per year, then with an NNT of 48 we would expect to prevent 2,500 outcomes (120,000/48), not up to 10,000 as suggested. They also fail to differentiate all admissions from those that would have been included in the trial — high risk ACS and other patients were excluded, eg, those on oral anticoagulants. National Institute for Clinical Excellence guidance on glycoprotein IIb/IIIa inhibitors (GPIs) suggests that high risk ACS makes up about a third of all cases, their definition being similar to that used as inclusion criteria in CURE. Thus the number of cases suitable for this intervention may fall to 40,000, and could then be expected to prevent about 830 outcomes (40,000/48). According to MIMS (October 2001), clopidogrel costs £344 for nine months (average follow up in CURE). Thus treating 40,000 patients would cost about £14m. If we use the 120,000 figure proposed, the cost would exceed £40m a year. Clearly this is a decision of national importance in a cash-limited service, and should probably be considered by an agency such as NICE before widespread adoption. The possible underestimation of benefit is important and deserves further analysis. Although the figure quoted for transfusion may be correct, it is misleading. The "major bleeding" outcome recorded in CURE resulted in an NNH of 100. So for 833 outcomes prevented in 40,000 cases, we would cause 400 major bleeding episodes along with associated costs and interventions. They are also correct in stating that there was no significant increase in the risk of stroke, surgery or death due to bleeding. Yet, as they reminded us for the single outcome of cardiovascular death, the study was not powered to detect such a difference. It does not rule out that it exists; more work is needed. CURE only considers bleeding as an adverse effect and although this is important, clopidogrel can cause other significant adverse effects, eg, neutropenia. Their example of cost-effectiveness may be a benchmark but it is inappropriate. A better comparison would be GPIs, which on the face of information from NICE, may be preferred. However, longer term outcome data are required to clarify this. This applies equally to clopidogrel. We are unlikely to prevent these events from ever happening, so the questions must be, for how long are they deferred and what is the optimum duration of dual treatment? Do we stop (when?) or continue indefinitely? Hopefully the manufacturers will ensure that long-term follow-up of trial subjects occurs wherever consent remains. The last point is that this combination treatment is currently unlicensed for most patients. Doctors should think carefully before initiating it and ensure their patients match the inclusion and exclusion criteria of the original study. CURE has shown us there are more options to be explored in ACS but we must adopt them carefully to ensure we make the best of limited resources. Widespread and misleading media attention has done little to help clinicians do this. Magnus Hird |
|
|
Dr MARCUS FLATHER and Dr AMEET BAKHAI reply: The use of the number needed to treat, as the sole guide to therapeutic efficacy is incomplete. The NNT is the reciprocal of the absolute difference in event rates between the treatment and control groups. There is a confidence interval around these estimates also. In our letter we stated that we could avoid between 3,000 and 10,000 deaths, strokes and myocardial infarctions each year in the UK if clopidogrel was used in patients similar to those enrolled in the CURE trial.1 The lower estimate of 3,000 is not materially different from 2,500 estimated from the crude NNT calculation. The baseline risk of UK patients with ACS without ST elevation can be estimated from the PRAIS-UK registry.2 Death, stroke or myocardial infarction occurred in 13 per cent at six months and we estimate a mean event rate of 20 per cent at 12 months. If the minimum number of admissions is 120,000 per annum, then we expect 24,000 such events. The upper limit of the confidence interval of the proportional reduction in CURE is 28 per cent, giving an estimate of 6,720 avoided. If the number of admissions is 180,000, which is still plausible, then we estimate 10,080 events avoided. Thus our estimate of 3,000 to 10,000 events avoided is a reasonable guide. CURE enrolled a wide range of patients, not only high-risk, but excluded patients receiving glycoprotein IIb/IIIa receptor blockers. Thus we expect that many more than a third of UK patients would be expected to benefit from a combination of clopidogrel and aspirin. Magnus Hird talks about a "cash limited" service. This is one of the reasons why the UK has a sub-optimal record in managing patients with cardiovascular diseases. The Government has, through the National Service Framework for Coronary Heart Disease, shown a commitment to improving this record. We should consider the total health economic impact of new, effective treatment, not just its initial cost. Thus from a societal viewpoint, it is often more expensive not to use effective treatments than it is to use them. We would welcome the review by NICE of the use of clopidogrel for acute coronary syndromes. There is an excessive bleeding in the CURE trial and this has been stated in the original publication. There was no excessive fatal bleeding and most major bleeds were managed with a blood transfusion. Surgical interventions were rare. The prospective economic analysis of the CURE trial will take into account the extra cost of bleeding. The rate of neutropenia was not greater in the clopidogrel group compared with the control group. Most treatments for chronic illnesses such as coronary heart disease defer complications and death. We are continuing to follow patients beyond the study treatment period, to see if the treatment differences remain. The main point of the publication and its associated media activities was to draw attention to the results within the health care community and society at large. The approval process for effective treatments can take up to a year. We believe it is up to clinicians to make informed decisions about the treatment of their patients, particularly in life-threatening conditions such as acute coronary syndromes. Clopidogrel is already licensed as an alternative to aspirin in patients with a history of stroke, myocardial infarction or peripheral vascular disease. Thus the use of combination with aspirin is a logical step given the CURE results. |
||
|
1. Yusuf S, Zhao F, Mehta SR, Chrolavicius S, Tognoni G, Fox KK. Effects of clopidogrel in addition to aspirin in patients with acute coronary syndromes without ST-segment elevation. N Engl J Med 2001;345:494–502. 2. Collinson J, Flather MD, Fox KA, Findlay I, Rodrigues E, Dooley P et al. Clinical outcomes, risk stratification and practice patterns of unstable angina and myocardial infarction without ST elevation: Prospective Registry of Acute Ischaemic Syndromes in the UK (PRAIS-UK). Eur Heart J 2000;21:1450– 7. |
||
|
Previous Topic (Community pharmacy) |
Home | Journals | News | Notice-board | Search | Jobs Classifieds | Site
Map | Contact us
©The Pharmaceutical Journal
