The polypill in the primary prevention of cardiovascular disease

doi: 10.1111/j.1472-8206.2009.00795.x REVIEW ARTICLE The polypill in the primary prevention of cardiovascular disease David S. Wald*, Nicholas J. Wald Wolfson Institute of Preventive Medicine, Barts and the London School of Medicine, Queen Mary University, London EC1M 6BQ, UK Keywords heart attack, polypill, primary prevention, stroke Received 29 June 2009; revised 29 September 2009; accepted 23 October 2009 *Correspondence and reprints: d.s.wald@qmul.ac.uk ABSTRACT Coronary heart disease (CHD) and stroke are the most common causes of death worldwide. The polypill, a multi-component tablet or capsule designed to reduce several cardiovascular causal risk factors simultaneously, has the potential to reduce the incidence of heart attacks and stroke by about 80%, the precise reduction depending on the formulation of the polypill and adherence to preventive treatment. The full public health impact can only be realized if the polypill is used to prevent first cardiovascular disease events, because most heart attacks and strokes are first events. A simple and effective method of selecting who should be offered the polypill is needed. Even though serum cholesterol and blood pressure are important causes of CHD and stroke and lowering them has a large preventive effect (a 1.8 mmol/l reduction in LDL cholesterol reduces the incidence of CHD events by about 60% and a 10 mmhg reduction in diastolic blood pressure also reduces the incidence of stroke by about 60%), they are poor screening tests. They add little to age in discriminating between individuals who will and will not have a CHD event or stroke. Including them in a risk assessment algorithm needlessly complicates the screening process and tends to medicalize the preventive strategy. Age alone is effective and a simpler means of selecting people for preventive treatment using the polypill. In 2003 Wald and Law, with colleagues, published a set of three papers [1 3], based on a patent application filed in 2000 [4]. They proposed a combination fixed dose medical formulation designed to modify several causal cardiovascular risk factors simultaneously and called this the polypill. They proposed a new preventive approach (the polypill approach) in which preventive treatment would be used by everyone above a specified age (e.g. 55) regardless of their starting risk factor levels and showed that a six component polypill could, with full adherence to preventive treatment, reduce a person s risk of coronary heart disease (CHD) and stroke by about 80%. The principle cardiovascular risk factors lowered by a polypill are LDL cholesterol (e.g. with a statin) and blood pressure (e.g. with ACE-inhibitors, angiotensin receptor blockers, thiazides, beta blockers, or calcium channel blockers). Reducing platelet aggregation (e.g. with aspirin) and serum homocysteine (e.g. with folic acid) are options. All the components have been shown to reduce the risk of CHD and stroke in randomized trials, except folic acid, which is inferred from epidemiological studies. Since 2003 several groups have developed polypills, demonstrating the feasibility of incorporating multiple drug components in a single tablet or capsule [5,6]. Efficacy in preventing heart attacks and strokes is likely to be similar for the different proposed polypills (about an 80% preventive effect on CHD events and 70% on stroke for a polypill with aspirin, a statin and two or more blood pressure components). The variations in formulation and dose will, however, affect tolerability and the incidence of serious adverse effects (such as gastric and intracranial bleeding with aspirin), which will influence the choice of polypill [7,8]. A polypill without aspirin may be preferred for primary prevention because the absolute risk of haemorrhage may be too great in relation to the absolute 29

30 D.S. Wald & N.J. Wald risk reduction [9], but it confers a net benefit in a polypill for people who have had an ischaemic event. The inclusion of folic acid in the polypill is an open question given the uncertainty over the benefits of homocysteine reduction. Did not die of CHD Died of CHD DR = 13% THE POLYPILL APPROACH FPR = 5% The proposed polypill approach recognizes the fact that, within the range of usual risk factor levels, there are no risk factor thresholds above which treatment is needed and below which treatment is not needed. Cohort studies show that within the usual physiological range of blood pressure, cholesterol and serum homocysteine, there are no thresholds below which a further decrease in the risk factor is not associated with a further decrease in risk [10 13]. The polypill approach also recognizes that there is benefit in reducing causal risk factor levels in people at high risk of cardiovascular disease, whatever the reason for their high risk and regardless of the level of the risk factor. In other words, treatment should not be limited to people with hypertension or hypercholesterolaemia. The proposed approach differs from the standard method of selecting individuals for preventive treatment by not using serum cholesterol or blood pressure measurements to determine who receives preventive treatment. This is perhaps surprising, as reducing these risk factors is so important in preventing heart disease and stroke. The reason they are not very effective in determining who should receive preventive treatment, i.e. as screening tests [14,15], is that nearly all adults have high risk factors, so the risk factors, whilst responsible for most cases of the diseases they cause, do little to distinguish between those who will and those who will not become affected by these diseases. SINGLE CARDIOVASCULAR RISK FACTORS AS SCREENING TESTS Figure 1 shows the relative distribution of diastolic blood pressure in people who did (right curve) and did not (left curve) die of CHD in a large cohort study [16]. There is substantial overlap between the two distributions. If a blood pressure cut-off were chosen that classified 5% of individuals who did not later die of CHD over a given period, as being high risk (a commonly used falsepositive rate in screening), then only 13% of those who did die of CHD in the same period would be identified. The screening performance is poor; a detection rate of 13% for a 5% false positive rate. 40 50 60 70 80 90 100 110 120 130 140 150 Diastolic blood pressure (mmhg) Figure 1 The distribution of diastolic blood pressure levels in people who did and did not die of coronary heart disease (CHD) in the BUPA cohort study [16]. The detection rate (DR, sensitivity) is shown for a 5% false positive rate (FPR). In clinical practice, a blood pressure cut-off is usually specified (e.g. a diastolic of 90 mmhg), above which individuals are regarded as high risk (hypertensive) and offered treatment, and below which they are not. However, the substantial overlap between the two distributions means that it is not possible to select a blood pressure level that identifies most people who develop CHD without including a large proportion of people who do not. If a blood pressure level of 90 mmhg were selected as the cut-off, as in Figure 2, 52% of all people who develop CHD would be classified as high risk (and offered treatment), but so too would 38% of those who do not; the discrimination is little better than selecting people for treatment at random. Moreover, 48% of people who will develop CHD (in the dark shaded area), would be classified as low risk and so be excluded FNR = 48% Did not die of CHD Died of CHD DR = 52% FPR = 38% 40 50 60 70 80 90 100 110 120 130 140 150 Diastolic blood pressure (mmhg) Figure 2 The distribution of diastolic blood pressure levels in people who did and did not die of coronary heart disease (CHD) in the BUPA cohort study [16]. The dark shaded area shows the proportion of people with CHD who would be misclassified as low risk using a diastolic blood pressure cut-off of 90 mmhg. (DR = detection rate; FPR = false-positive rate; FNR = false negative rate).

The polypill in the primary prevention of cardiovascular disease 31 from useful treatment. Screening performance is similar for other cardiovascular risk factors, like serum cholesterol, homocysteine, body mass index or HbA1c [17 20]. MULTIPLE CARDIOVASCULAR RISK FACTORS AS SCREENING TESTS It may appear intuitive that combining several risk factors (e.g. based on a Framingham, SCORE or QRISK2 algorithm) [21 23] to derive an individual s risk of a cardiovascular disease event may overcome the poor screening performance of individual risk factors. This may have prompted the UK Government, in April 2009, to introduce population screening for cardiovascular disease based on multiple risk factor measurement in people aged 40 74 without prior cardiovascular disease or diabetes [24]. The programme, initially called Vascular Checks and now termed NHS Health Checks, involves the calculation of a person s risk of developing a CHD event or stroke, based on age, weight, smoking, blood pressure and blood cholesterol using either the Framingham or QRISK2 algorithm (QRISK2 adds information on family history, ethnicity and social deprivation). Whichever risk assessment is used, the risk estimate itself is the screening test. If a person s risk of cardiovascular disease is judged to be high (>20% over 10 years), he or she is screen-positive and life-style advice (mainly to reduce weight and increase exercise) is given and interventions such as cholesterol lowering drugs may be offered. At moderate risk (10 20% over 10 years), lifestyle advice alone is given. If the risk estimation is low (<10% over 10 years), individuals are reassured and encouraged to maintain healthy life-style behaviour. The problem with this approach is that weakly predictive risk factors combining risk factors in this way has a small effect on improving screening performance compared to using one. Table I (based on men who did and did not die in the UK BUPA cohort study) [15] shows that adding a second or third risk factor adds little to the ability to separate individuals who will develop CHD from those who will not. For example, either systolic blood pressure (SBP) or apob detects 17% of individuals who die of CHD at a 5% false-positive rate. If both risk factors are used together and a person is designated screen-positive if either or both are positive, the detection rate is slightly less than twice as great (31%, calculated as 1 [1 0.17] 2 ), but the falsepositive rate is also about twice as great (9.75%). The key question is whether this detection rate of 31% is substantially higher than that from using either blood pressure or apob alone, at a 10% false-positive rate. It is Table I Illustration of the effect of combining risk factors on screening performance using estimates of the detection rate for various combinations of cardiovascular risk factors at 5% and 10% false-positive rates (adapted from Wald et al. [15]). Screening variable Detection rate (%) 5% false-positive rate Systolic blood pressure (SBP) alone 17 ApoB alone 17 SBP and apob 22 SBP, apob, apoai, apo(a) 24 SBP, apob, apoai, apo(a), smoking 27 SBP, apob, apoai, apo(a), smoking, family history 28 10% false-positive rate SBP alone 26 apob alone 28 SBP and apob 34 SBP, apob, apoai, apo(a) 36 SBP, apob, apoai, apo(a), smoking 40 SBP, apob, apoai, apo(a), smoking, family history 41 apob (protein component of LDL cholesterol). apoai (protein component of HDL cholesterol). apo(a) (protein component of lipoprotein a). not; the detection rate is 26% for blood pressure alone and 28% for apob alone (Table I). The gain in detection at the same false-positive rate from using both together over the more discriminatory of the two measurements (apob) alone is only about 4 percentage points. Adding the second risk factor adds relatively little to the first and adding a third or fourth adds even less. ApoB is the main surface protein on LDL particles and has been shown to be more strongly associated with CHD than its lipid fraction, LDL cholesterol [25], so measurement of LDL (or total) cholesterol instead of ApoB would be less, rather than more informative about an individual s future risk. Assessment of subclinical atherosclerosis, for example using carotid ultrasound to measure intima-media thickness, has been proposed as an alternative screening method, but the discrimination is only marginally better (19% detection rate for a 5% falsepositive rate) [26]. Figure 3 shows that about 20 risk factors, comparable with serum cholesterol or blood pressure as predictors of cardiovascular disease events, would be needed to detect 80% of cases at a 5% falsepositive rate [27]. Even if there were this many risk factors, their use would probably be complex and costly. Although multiple risk factor screening is widely adopted it is a questionable policy. Another issue is that the 20% risk cut-off of having a heart attack or stroke over 10 years that is used to trigger preventive treatment is too high. Whilst an

32 D.S. Wald & N.J. Wald 100 80 Detection rate (%) 60 40 20 0 0 5 10 15 Number of risk factors 20 Figure 3 Detection rate (sensitivity) for a 5% false positive rate according to the number of independent risk factors (each with a detection rate of 15% for a 5% false positive rate) used in combination (adapted from Wald et al. [27]). individual, at this risk cut-off, has a 1 in 50 chance of a heart attack or stroke each year and would be offered preventive treatment, someone below the risk cut-off, say with a 1 in 100 risk (10% over 10 years), would not be offered the treatment. If someone knew he or she had a 1 in 100 chance of a heart attack or stroke over the next 12 months, that person is unlikely to be satisfied with reassurance or dietary advice when simple and effective treatment could reduce this by about 75%. Limiting treatment to a minority of the population may be defendable if the preventive treatments were toxic or expensive, but they are not. Statins and blood pressure lowering drugs are remarkably safe and generic, so cost little. People willingly pay much more for other precautionary measures, like home insurance, when the risk of serious damage is far lower and less devastating than having a heart attack or stroke. AGE AS A SCREENING ENQUIRY As about 95% of all heart attack and stroke deaths in England and Wales occur in people aged 55 and over [28], age 55 has been proposed as the entry point for preventive treatment [1]. Figure 4 (based on England and Wales registry data [16]) shows that at age 55, the risk of a first coronary heart disease event or stroke over Figure 4 Risk of a first coronary heart disease event or stroke according to age based on published one-year risks, derived from registry data in England and Wales [16] (allowing for the increase in risk over the 10 year period). the next 10 years is about 4% (or about 1 in 250/year) and doubles every 7 years. Offering preventive treatment to all over a specified age is simple and equitable. It recognizes that an effective preventive strategy needs to be conceptually inclusive, given that heart attacks and strokes are common, killing one in three people [28]. The proposed approach is similar to that adopted in most vaccination programmes, where there is no attempt at selecting high and low risk groups who should or should not be offered vaccination. An age-based approach has the benefit of not medicalizing the population; individuals are not selected as having hypertension or hypercholesterolaemia and thus do not become patients to receive preventive treatment. Multiple risk factor screening does medicalize people and serves only to delay the start of treatment in some by a few years because everyone becomes high risk as they become a few years older. RECENT DEVELOPMENTS AND FUTURE CHALLENGES Several trials assessing the effect of polypills on risk factor reduction are underway and one has been reported [5]. The Indian Polycap Study (TIPS) trial on about 2000 people showed that a Polypill (Polycap) can be made and

The polypill in the primary prevention of cardiovascular disease 33 can be used safely by healthy people. The mean age was 54 years, mean blood pressure was 135/86 mmhg and mean LDL cholesterol was 3.0 mmol/l. The serum cholesterol and blood pressure reductions in TIPS were thought to be smaller than the estimates from metaanalyses of trials of the individual components [5]. However, there was no inconsistency with respect to LDL reduction; (a reduction of 0.7 mmol/l rather than 1.8 mmol/l). The lower pre-treatment cholesterol in the TIPS trial (3.0 mmol/l compared with 4.8 mmol/l in the meta-analysis of trials) taken together with the lower dose of simvastatin used in TIPS (20 mg, not 40 mg) and non-adherence to prescribed therapy explains this difference. The blood pressure reductions observed in TIPS (7.4/5.6 mmhg) were smaller than expected allowing for pre-treatment blood pressure and non-adherence [29,30]. This is unexplained and the TIPS authors acknowledge this. Trials assessing the effect of Polypills on reducing first heart attacks and strokes are planned or underway [30]. The main question in these trials is not on efficacy and safety but on adherence and tolerability. Intolerance to one component (for example, an ACE-inhibitor, which causes cough in about 1 in 10 people) [31] may result in the Polypill being stopped. Encouraging continuation with preventive treatment, even in the absence of sideeffects, is a challenge, and strategies to incentivize adherence need to be developed and tested. Polypills are new medicinal products but their components are not new. The regulatory standards that apply to the approval of new chemical entities should not automatically apply to polypills [32]. Bioequivalence studies (studies that show that taking a polypill is equivalent or nearly equivalent in terms of absorption and blood levels as taking the individual components) should be sufficient for a Regulatory Authority to approve the drug for use with indications of use that apply to use of the components in medical practice. If the components are effective in cardiovascular disease prevention, the indications for use of a polypill containing such components should extend to this, even if there is already no such licensed use (usually because no company has applied for one). So, for example, it should be possible to license a polypill with a statin and three blood pressure-lowering drugs for the primary prevention of cardiovascular disease, not just the treatment of hypercholesterolaemia or hypertension. It is in the public interest for regulators and scientists to be flexible and to work together to simplify the development, delivery and approval of polypills, so preventive treatment is accessible and affordable to all who stand to benefit. COST-EFFECTIVENESS OF THE POLYPILL APPROACH A polypill, based on generic drugs that lower cholesterol and blood pressure, might cost about 400 per person per year. If all people aged 55 and older took such treatment for 25 years (average life expectancy from age 55), the total cost per person would be about 10 000. Approximately, one in three people who take treatment would benefit and he or she would gain about 11 years of life without a heart attack or stroke [1], that is about 2500 for each year of life gained without a heart attack or stroke [(10 000 3)/11]. There would be almost no costs attached to screening, allowing general practitioners and other health care professionals to focus attention on other health matters instead of engaging in relatively ineffective selection and administrative activities. This is a simple calculation to indicate what the approximate cost could be. Adherence to preventive therapy is uncertain and hence the overall benefit may be less, but even if the polypill prevented only half the estimated proportion of events, it would still be cost-effective. With direct access to the public, for example, through pharmacist prescribing, the cost could, in part, be moved from national health care budgets to the individual. Employers, who wish to retain healthy employees over age 55 at work would also have an incentive to subsidize the preventive treatment costs to their staff. In a few years, polypills are likely to be available in many countries. They are already available in India. While formulation, tolerability and cost are important, the extent to which polypills will achieve their full preventive potential will depend largely on specifying a policy that ensures that all who may benefit are able to obtain them. There is general agreement that people who have already had a heart attack or stroke will benefit. However, the current neglect in preventing first events is a mistake; every person who has had a second or subsequent heart attack or stroke has had a first event and the priority should be to prevent the first event. If a first event is prevented, there is no second event to prevent. KEY POINTS The polypill approach (offering it to people above a certain age, say 55) without medical examination or risk factor measurement is simple and could save millions of lives a year worldwide. Preventive treatment should reduce several causal risk factors, even if they are not raised.

34 D.S. Wald & N.J. Wald Causal risk factors such as serum cholesterol and blood pressure are important aetiologically, but are poor screening tests, both individually and in combination with other risk factors. Current cardiovascular disease prevention strategies based multiple risk factor measurement should be reviewed. Primary prevention of cardiovascular disease is the priority. FUNDING SOURCES None. DECLARATION OF INTEREST Nicholas J Wald holds patents (EU1272220 and GB2361186) for a combination pill for the prevention of cardiovascular disease and, together with David Wald, has interests in its development. ACKNOWLEDGEMENT We are grateful to Mark Simmonds for his comments on the manuscript and for preparing Figure 4. REFERENCES 1 Wald N.J., Law M.R. A strategy to reduce cardiovascular disease by more than 80%. BMJ (2003) 326 1419 1423. 2 Law M.R., Wald N.J., Morris J.K., Jordan R.E. Value of low dose combination treatment with blood pressure lowering drugs: analysis of 354 randomised trials. BMJ (2003) 326 1427 1431. 3 Law M.R., Wald N.J., Rudnicka A.R. Quantifying effect of statins on low density lipoprotein cholesterol, ischaemic heart disease, and stroke: systematic review and meta-analysis. BMJ (2003) 326 1423 1427. 4 European patent specification: formulation for the prevention of cardiovascular disease. EP 1 272 220 B1. Proprietors Wald N.J., Law M.R.; priority 10.04.2000 5 The Indian Polycap Study (TIPS). Effects of a polypill (Polycap) on risk factors in middle aged individuals without cardiovascular disease (TIPS): a phase II, double-blind, randomised trial. Lancet (2009) 373 1341 1351. 6 Phase 2 study of heart polypill safety and efficacy in primary prevention of cardiovascuolar disease. http://clinicaltrials.gov/ ct2/show/nct00603590 DOI 15/6/2009. 7 Derry S., Loke Y.K. Risk of gastrointestinal haemorrhage with long term use of aspirin: meta-analysis. BMJ (2000) 321 1183 1187. 8 McQuaid K.R., Laine L. Systematic review and meta-analysis of adverse events of low-dose aspirin and clopidogrel in randomized controlled trials. Am. J. Med. (2006) 119 624 638. 9 Antithromobotic Trialists (ATT) Collaboration. Aspirin in the primary and secondary prevention of vascular disease: collaborative meta-analysis of individual participant data from randomised trials. Lancet (2009) 373 1821 1822. 10 MacMahon S., Peto R., Cutler J. et al. Blood pressure, stroke and coronary heart disease. Part 1, prolonged differences in blood pressure: prospective observational studies corrected for the regression dilution bias. Lancet (1990) 335 765 774. 11 Neaton J.D., Wentworth D. Serum cholesterol, blood pressure, cigarette smoking, and death from coronary heart disease. Arch. Intern. Med. (1992) 152 56 64. 12 Prospective Studies Collaboration. Age-specific relevance of usual blood pressure to vascular mortality: a meta-analysis of individual data for one million adults in 61 prospective studies. Lancet (2002) 360 1903 1913. 13 Wald D.S., Law M.R., Morris J. Homocysteine and cardiovascular disease: evidence on causality from a meta-analysis. BMJ (2002) 325 1202 1206. 14 Wald N.J., Hackshaw A.K., Frost C.D. When can a risk factor be used as a worthwhile screening test? BMJ (1999) 319 1562 1565. 15 Wald N.J., Law M., Watt H. et al. Apolipoproteins and ischaemic heart disease: implications for screening. Lancet (1994) 343 75 79. 16 Law M., Wald N.J., Morris J. Lowering blood pressure to prevent myocardial infarction and strokes: a new preventive strategy. Health Technol. Assess. (2003) 7 No 31. 17 Law M.R., Wald N.J., Wu T., Hackshaw A., Bailey A. Systematic underestimation of association between serum cholesterol concentration and ischaemic heart disease in observational studies: data from the BUPA study. BMJ (1994) 308 363 366. 18 Law M., Morris J. The dose response relationship between serum homocysteine and cardiovascular disease:implications for screening and treatment. Eur J Cardiovasc Prev Rehabil (2004) 11 250 253. 19 Prospective Studies Collaboration. Body-mass index and causespecific mortality in 900 000 adults; collaborative analyses of 57 propsective studies. Lancet (2009) 373 1083 1096. 20 Khaw K.T., Wareham N., Luben R. et al. Glycated haemoglobin, diabetes and mortality in men in Norfolk cohort of European Prospective Investigation of Cancer and Nutrition (EPIV-Norfolk). BMJ (2001) 322 15. 21 D Agostino R.B., Vasan R.S., Pencina M.J. et al. General cardiovascular risk profile for use in primary care. the Framingham heart study. Circulation (2008) 117 743 753. 22 De Backer G., Ambrosioni E., Borch-Johnsen K. et al. Executive summary. European guidelines on cardiovascular disease prevention in clinical practice. Third Joint Task Force of European and other Societies on Cardiovascular Disease Prevention in Clinical Practice (constituted by representatives of eight societies and by invited experts. Eur. Heart J. (2003) 24 1601 1610. 23 Hippisley-Cox J., Coupland C., Vinogradova Y. et al. Predicting cardiovascular risk in England and Wales: prospective deri-

The polypill in the primary prevention of cardiovascular disease 35 vation and validation of QRISK2. A new cardiovascular disease risk score for the UK. BMJ (2008) 336(7659) 1475 1482. 24 Department of Health. Putting prevention first. NHS Health Check: Vascular Risk Assessment and Management. http:// www.dh.gov.uk/en/publicationsandstatistics/publications/ PublicationsPolicyAndGuidance/DH_097489. 25 Parish S., Peto R., Palmer A. et al. The joint effects of apolipoprotein B, apolipoprotein A 1, LDL cholesterol, and HDL cholesterol on risk: 3510 cases of acuts myocardial infarction and 9805 controls. Eur. Heart J. (2009) 30) 2137 2146. 26 Wald D.S., Bestwick J.P. Carotid ultrasound screening for coronary heart disease: results based on a meta-analysis of 18 studies and 44,861 subjects. J. Med. Screen. (2009) 16 147 154. 27 Wald N.J., Morris J.K., Rish S. The efficacy of combining several risk factors as a screening test. J. Med. Screen. (2005) 12 197 201. 28 Office for National Statistics. Mortality Statistics Cause. 2003 29 Wald N., Law M. The Indian Polycap Study (TIPS). Lancet (2009) 374 781. 30 Yusuf S., Pais P., Xavier D. The Indian Polycap Study (TIPS): Authors reply. Lancet (2009) 374 782. 31 The Heart Outcomes Prevention Evaluation Study Investigators. Effects of an angiotensin-converting-enzyme inhibitor, ramipril, on cardiovascular events in high-risk patients. N. Engl. J. Med. (2000) 342 145 153. 32 Sleight P., Pouleur H., Zannad F. Benefits, challenges, and registerability of the polypill. Eur. Heart J. (2006) 27 1651 1656.