Moving toward the next paradigm for cardiovascular prevention
Invited editorial forThanassoulis G, Williams K, Altobelli K, Pencina M, Cannon C and Sniderman A. Individualized statin benefit for determining statin eligibility in the primary prevention of cardiovascular disease. Circulation. 2016;in press.
Jennifer G Robinson MD MPH
University of Iowa
Kausik Ray, MBChB, FRCP, MD, MPhil
Imperial College London
Corresponding author
Jennifer G. Robinson, MD, MPH
Professor
Departments of Epidemiology & Medicine
Director, Prevention Intervention Center
Department of Epidemiology
College of Public Health
University of Iowa
145 N Riverside Drive S455 CPHB
Iowa City, IA 52242-2007
Office 319.384.1563
Assistant 319.384.1540
Fax 319.384.4155
Words:2408 including references
References: 17
Figures: 2
Conflict of interest disclosures
Jennifer G. Robinson, MD, MPH in the past year:
Research grants to Institution: Amarin, Amgen, Astra-Zeneca, Eli Lilly, Esai, Glaxo-Smith Kline, Pfizer, Regeneron, Sanofi, and Takeda.
Consultant: Akcea/Isis, Amgen, Eli Lilly, Esperion, Merck, Pfizer, Regeneron, Sanofi
Kausik Ray, MBChB, FRCP, MD, MPhil in the past year:
Research grants to Institution: Pfizer, Amgen, MSD, Sanofi, and Regeneron
Consultant: Aegerion, Amgen, Astra Zeneca, BoehringerIngelheim, Cerenis, Eli Lilly, Kowa, Merck, Pfizer, Regeneron, Resverlogix, Sanofi, Takeda
Guidelines change based on new evidence and methods. The Adult Treatment Panel (ATP) guidelines evolved over their three iterations and focused on a treat-to-goal approach. Following a rigorous systematic review of randomized trials of cholesterol-lowering drug therapy with cardiovascular outcomes, the 2013 ACC/AHA cholesterol guideline moved away from the approach recommended in earlier ATP guidelines, to a new paradigm (Figure 1) for cholesterol treatment based on the use of statin therapy in 4 “statin benefit groups”, in which there was a clear net atherosclerotic cardiovascular disease (ASCVD) reduction benefit.1 For primary prevention, strong evidence from 3 exclusively primary prevention trials supportedthe initiation of statin therapy in those with 7.5%10-year ASCVD risk, and moderate evidence supported statin therapy for those with 5 to <7.5% 10-year ASCVD risk. However, the guidelines stated that decision to start statin therapy in these patients and in lower risk patients should be deliberated within the context of shared decision-making in a Clinician-Patient Discussion, due to the potentially narrower margin of benefit and need to elicit patient preferences. Consideration could also be given to other characteristics that may increase ASCVD risk, includingfamily history of premature ASCVD, elevated CAC score, high sensitivity C-reactive protein 2 mg/L, ankle brachial index <0.9, or increased lifetime ASCVD risk, or an LDL-C 160 mg/dl.
Since the completion of the 2013 ACC/AHA cholesterol guidelines,several important analyses have been publishedthat should inform future primary prevention guideline updates. Based on a Cholesterol Treatment Trialists meta-analysis that found each 39 mg/dl reduction in LDL-C was associated with a 21% relative risk reduction (RRR) in major cardiovascular events,2Soran et al evaluated the impact of the relative risk reduction (RRR) expected from the magnitude of LDL-C reduction, which is a function of both the starting LDL-C level and the intensity of statin therapy.3 Based on their analysis, fairly similar numbers-needed-to-treat to prevent one cardiovascular event (NNT) of 48 and 49, respectively, would occur when adding atorvastatin 20 mg in a patient with a 5% 10-year ASCVD risk and LDL-C 193 mg/dl (5mmol/L) as in a patient with 7.5% 10-year ASCVD risk and LDL-C 115 mg/dl (3 mmol/L). Another analysis by the Cholesterol Treatment Trialists further showed that the RRR is greater for a given magnitude of LDL-C lowering in low risk patients (34% RRR when <10% 5-year major cardiovascular disease risk, which approximately translates to <10% 10-year “hard” ASCVD risk4) than in higher risk patients (17-19% RRR in those with 10% 5-year major cardiovascular disease risk).5 This means that the potential for an ASCVD risk reduction benefit from statin therapy may be higher than was previously appreciated in lower risk patients.
Thanassoulis, et al have now synthesized the information from these 2 papers, along with the net benefit a concept first introduced in the 2013 ACC/AHA cholesterol guideline, to take the next step in refiningthe estimation of net benefit for an individual patient.6 The “individualized net benefit approach” of Thanassoulis, et al used an absolute risk reduction (ARR) benefit threshold of 2.3% (for a10-year NNT 44), the minimum expected ASCVD risk reduction benefit from the 2013 ACC/AHA cholesterol guideline. Absolute risk reduction was based on (1) the absolute 10-year ASCVD risk estimated by the Pooled Cohort Equations for white and African American women and men7, and (2) the relative risk reduction from a 40% reduction LDL-C by a mid-potency statin, adjusted for absolute 10-year ASCVD risk level < or 10%.
Using data representative of the US population from the National Health and Nutrition Survey, this new individualized net approach would identify 34% (24.6 million) US adults as eligible for statin therapy, compared to the 21% (15.0 million) identified by the 2013 ACC/AHA guideline. The mean 10-year predicted ASCVD risk was 10.9% (average 10-year NNT=25) with the new approach, and 13.9% (average 10-year NNT=21) for the 2013 ACC/AHA guideline approach, both of which would be considered acceptable by both clinicians and patients. An NNT 50 is considered reasonable by many clinicians and an NNT 30 is considered reasonable by many patients.8 Importantly, while a 7.5% 10-year ASCVD treatment threshold would prevent over 728,000 ASCVD events over 10 years, expansion of statin treatment to those with ARR 2.3% would prevent an additional 265,000 events over 10 years.
Several sensitivity analyses were performed. If a high intensity statin were used (50% LDL-C reduction), an additional 12.3 million US adults would be eligible for statin therapy at the ARR 2.3% threshold.A more liberal ARR threshold of 1.5% (10-year NNT67; average 10-year NNT=30), would treat 11 million more US adults than the 2.3% ARR threshold. A more conservative strategy with an ARR threshold 3.0% (NNT 30) would treat 17.3 million US adults, still more than the 2013 ACC/AHA cholesterol guideline, but would also exclude about 0.8 million individuals who would be treated under that guideline. The authors also examined whether simply treating all individuals with 5% or 3% 10-year ASCVD risk would be sufficient to capture those likely to experience a net ARR benefit at the 2.3% ARR threshold. A 5% cut-point would add 7.7 million US adults eligible for statin therapy, yet still miss 2 million who could benefit at that level. A 3% cut-point would add another 1.8 million who could benefit, but would also add 10.4 million with an ARR <2.3%.
In comparison to the example from the Soran, et al analysis cited above, the individualized net benefit approach of Thanassoulis, et al further refines the levels of absolute risk and LDL-C that might identify patients likely to experience a net risk reduction benefit at the ARR 2.3% level. For example, in the Thanassoulis, et al analysis, an individual at 7.5% 10-year ASCVD risk with an LDL-C of 85 mg/dl (2.2 mmol/L) would experience the same net benefit (ARR=2.3%; NNT=44) from statin therapy as an individual with 5% 10year ASCVD risk and an LDL-C of 125 mg/dl. These findings further support the move away from LDL-C treatment goals, since the treat-to-goal approach does not take the starting level of LDL-C into consideration when considering the potential for benefit. Notably,both of these patients would be considered “at goal” before statin initiation, and have average on-statin LDL-C levels of 51 mg/dl and 75 mg/dl, respectively, and so would not be treated under treat-to-goal primary prevention guidelines.9
It should also be noted that statin eligibility based on clinical trial inclusion criteria and a hybrid approach using clinical trial inclusion trial performed worse than both the 2013 ACC/AHA cholesterol guideline and the ARR-based individualized net benefit approach. Inclusion criteria over-identified low risk individuals for statin eligibility, and a hybrid approach applying inclusion criteria after 10-year ASCVD risk estimation failed to identify a large proportion of those who would likely experience a net benefit from statin therapy. Application of clinical trial eligibility-based approaches in other populations have had similar findings.10
Moving toward a new paradigm of net benefit guided preventive therapy
The analysis by Thanassoulis, et al does supportthe current paradigm of statin initiation for those with 7.5% 10-year ASCVD risk, which was considered too aggressive by some critics of the 2013 ACC/AHA cholesterol guideline. However, the analysis further provides important new evidencesupporting development of a new prevention paradigm based on the potential for net benefit using ARR thresholds, or NNT. This analysis supports extending statin therapy to at least an additional 13% of the US population based on the potential for net benefit at a threshold of ARR 2.3%. Lower ARR thresholds could be desirable, and cost-benefit analyses should be helpful in determining reasonable ARR cut-points.
Unfortunately, “net benefit” calculators are not yet available. Until such estimators are available for clinical use, the analysis by Thanassoulis, et alalso suggests that a net benefit consistent with an ARR 2.3% will accrue to individuals with 5% 10-year ASCVD risk and LDL-C 130 mg/dl. Notably, a 5% 10-year ASCD risk threshold for LDL-C levels of 70-189 mg/dl is already considered cost-effective at US$57,000 per quality adjusted life year.11
However, more evidence is needed to implement a third paradigm based on individualized estimation of net benefit (Figure 1). A net benefit calculator requires not just an estimation of benefit, as provided by the ARR, but also needs an estimate of harm. The 2013 ACC/AHA cholesterol guideline definition of “net benefit” included a consideration of the potential for adverse events and drug-drug interactions from cholesterol-lowering drug therapy. Thanassoulis, et al. did not take adverse events into consideration in their calculation of individualized net benefit. This is likely appropriate for healthy individuals similar to those enrolled in the randomized trials of statin therapy, where the risk of adverse events appears to be confined to rare cases of serious myopathy or hemorrhagic stroke.5 Statin-associated diabetes appears to be confined to those already at risk of diabetes, and the additional 2-4 months of a diabetes diagnosis does not appear to increase ASCVD risk.12 However, adverse event risk is higher in those who likely would have been excluded from the statin trials based on age or comorbidities.13 The next important step for improved prediction of net benefit will therefore be the development of risk estimation tools for adverse events (Figure 2).
In addition, the improved estimation of the absolute risk of a patient from which to calculate the individualized ARR is also needed (Figure 2), Refinements in ASCVD risk estimation will likely occur as analyses are performed using the new 10-year ASCVD risk cut-points to evaluate the incremental information added by noninvasive assessments of atherosclerotic burden and new biomarkers, including genetic risk scores.14, 15 Emerging evidence also suggest genetic risk scores may inform estimates of relative risk reduction from statin therapy.15
The concept of net benefit can also be extended to preventive therapies that may or may not lower LDL-C (Figure 2). The incremental additional net benefit from adding ezetimibeor aspirin in statin-treated patients are current clinical issues that could be addressed through an estimation of net benefit using NNT and NNH.16, 17 Ongoing cardiovascular outcomes trials are evaluating a number of drugs, including proproteinconvertasesubtilisin/kexin 9inhibitors, cholesteryl ester transfer protein inhibitors, omega-3 fatty acids, and anti-inflammatory agents. If these drugs further reduce ASCVD events in statin treated patients, risk prediction equations in statin-treated patients with and without clinical ASCVD will be needed to refine estimates of net benefit to guide their use.
In conclusion, evidence is accumulating for yet another paradigm shift in the approach to cardiovascular risk reduction therapies. Individualized estimates of ARR are the first step toward this new paradigm. Evidence gaps in individualized net benefit estimation will likely be addressed in the next few years asdata become available from cardiovascular outcomes trials of new preventive therapies, meta-analyses facilitating the individualized estimation of adverse events, and from studies improving the individualized estimation of absolute ASCVD risk. Putting this evidence together in a net benefit calculator will inform shared decision-making and should allow better utilization of preventive therapies.
Figure legends
Figure 1. Treatment paradigms for cholesterol treatment guidelines evolve after incorporating accumulated new evidence and methods
Figure 2. Estimating the potential net benefit from additional preventive therapy in primary or secondary prevention. Potential for net benefit is a function of the absolute reduction in ASCVD risk (ARR) reflected in the number-needed-to treat to prevent one ASCVD event (NNT) and the number-needed-to-harm from one adverse event (NNH). The ARR results from the estimated absolute risk of the patient, and the relative risk reduction (RRR) expected from the added therapy.
References
1.Stone NJ, Robinson JG, Lichtenstein AH, Bairey Merz CN, Blum CB, Eckel RH, Goldberg AC, Gordon D, Levy D, Lloyd-Jones DM, McBride P, Schwartz JS, Shero ST, Smith Jr SC, Watson K and Wilson PWF. 2013 ACC/AHA Guideline on the Treatment of Blood Cholesterol to Reduce Atherosclerotic Cardiovascular Risk in Adults: A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol. 2014;63:2889-2934.
2.Cholesterol Treatment Trialists' (CTT) Collaborators. Efficacy and safety of cholesterol-lowering treatment: prospective meta-analysis of data from 90,056 participants in 14 randomised trials of statins. Lancet 2005;366:1267-1278.
3.Soran H, Schofield JD and Durrington PN. Cholesterol, not just cardiovascular risk, is important in deciding who should receive statin treatment. European Heart Journal. 2015;36:2975-2983.
4.Robinson JG. Starting Primary Prevention Earlier With Statins. The American Journal of Cardiology. 2014;114:1437-1442.
5.Cholesterol Treatment Trialists Collaborators. The effects of lowering LDL cholesterol with statin therapy in people at low risk of vascular disease: meta-analysis of individual data from 27 randomised trials. Lancet. 2012;380:581-590.
6.Thanassoulis G, Williams K, Altobelli K, Pencina M, Cannon C and Sniderman A. Individualized statin benefit for determining statin eligibility in the primary prevention of cardiovascular disease. Circulation. 2016;in press.
7.Muntner P, Colantonio LD, Cushman M, Goff DJ, Howard G, Howard V, Kissela B, Levitan E, Lloyd-Jones D and Safford M. Validation of the atherosclerotic cardiovascular disease pooled cohort risk equations. JAMA. 2014;311:1406-1415.
8.Steel N. Thresholds for taking antihypertensive drugs in different professional and lay groups: questionnaire survey. 2000;320:1446-1447.
9.Grundy SM. An International Atherosclerosis Society Position Paper: Global recommendations for the management of dyslipidemia. J Clin Lipidol. 2013;7:561-565.
10.Mortensen MB, Afzal S, Nordestgaard BG and Falk E. Primary Prevention With Statins: ACC/AHA Risk-Based Approach Versus Trial-Based Approaches to Guide Statin Therapy. Journal of the American College of Cardiology. 2015;66:2699-2709.
11.Pandya A, Sy S, Cho S, Weinstein MC and Gaziano TA. COst-effectiveness of 10-year risk thresholds for initiation of statin therapy for primary prevention of cardiovascular disease. JAMA. 2015;314:142-150.
12.Robinson J. Statins and diabetes risk: How real is it and what are the mechanisms? Curr Opin Lipidol. 2015 26:228-235.
13.Brooks JM, Cook E, Chapman CG, Schroeder MC, A. Chrischilles E, Schneider KM, Kulchaitanaroaj P and Robinson J. Statin Use After Acute Myocardial Infarction by Patient Complexity: Are the Rates Right? Medical Care. 2015;53:324-331.
14.Yeboah J, Polonsky TS, Young R, McClelland RL, Delaney JC, Dawood F, Blaha MJ, Miedema MD, Sibley CT, Carr JJ, Burke GL, Goff DC, Psaty BM, Greenland P and Herrington DM. Utility of Non-Traditional Risk Markers in Individuals Ineligible for Statin Therapy According to the 2013 ACC/AHA Cholesterol Guidelines. Circulation. 2015.
15.Mega JL, Stitziel NO, Smith JG, Chasman DI, Caulfield MJ, Devlin JJ, Nordio F, Hyde CL, Cannon CP, Sacks FM, Poulter NR, Sever PS, Ridker PM, Braunwald E, Melander O, Kathiresan S and Sabatine MS. Genetic risk, coronary heart disease events, and the clinical benefit of statin therapy: an analysis of primary and secondary prevention trials. The Lancet. 2015;385:2264-2271.
16.Robinson J and Stone N. The 2013 ACC/AHA guideline on the treatment of blood cholesterol to reduce atherosclerotic cardivoascular disease risk: A new paradigm supported by more evidence. Eur Heart J. 2015 36:2110-2118.
17.Miedema MD, Lopez FL, Blaha MJ and et al. Eligibility for statin therapy according to new cholesterol guidelines and prevalent use of medication to lower lipid levels in an older us cohort: The atherosclerosis risk in communities study cohort. JAMA Internal Medicine. 2015;175:138-140.