This article is more than 5 years old. Information may no longer be current.
The wait is over: The new cholesterol treatment guidelines are here
After a long wait, the guideline previously known as ATP IV is here. We would like to be among the first to congratulate the expert panel on completing the new 2013 American College of Cardiology/American Heart Association Cholesterol Treatment Guidelines. There is no doubt that it was a large undertaking to translate a decade’s worth of rich and novel science that transformed the way we approach our patients at increased risk for atherosclerotic CVD (ASCVD).
Indeed, the panel was highly selective in what it felt was worthy of inclusion in the guideline. The panel restricted its consideration to evidence from randomized controlled trials (RCTs) of drug treatments. With these guidelines hot off the press, this article examines some of its key features.
Primary prevention is here to stay
The 2013 guidelines are an emphatic declaration that primary prevention with selective use of statin therapy is here to stay. After years of careful review of statin trials, the expert writing group determined that the role for statins in primary prevention should be expanded beyond CHD to include stroke prevention, whether fatal or nonfatal.
Seth S. Martin
Rather than focusing solely on total mortality , the panel concluded that prevention of major nonfatal ASCVD events can reduce the large burden of disability from nonfatal stroke and nonfatal CHD events. Getting ahead of the curve may be critical to tackling the projected tripling of ASCVD-related health care costs in the next 20 years. Moreover, the new guidelines also debunk the mortality myth, finding that updated, scientifically rigorous meta-analyses of patients without prior ASCVD show an approximate 10% reduction in all-cause mortality.
The Cholesterol Treatment Trialists conducted one such meta-analysis. In this study, the effects of lowering LDL cholesterol with statin therapy were examined in individuals with a less than 10% 5-year risk for major vascular events. The proportional reduction in major ASCVD events was at least as big in the two lowest 5-year risk categories as in the higher risk categories (rate ratio per 1-mmol/L or 39-mg/dL reduction in LDL: 0.62; 99% CI, 0.47-0.81 in <5% risk category; rate ratio: 0.69; 99% CI, 0.6-0.79 in ≥5% to <10% risk category). Moreover, the expansion of statin use is further supported from a cost perspective due to multiple statins becoming generic since ATP III, with several available for $4 to $5 per month.
Risk assessment, treatment threshold and statin therapy
In essence, the new guidelines broaden risk assessment, lower the treatment threshold, and now explicitly identify statins as the first-line treatment for high blood cholesterol and increased CV risk. The updated ASCVD risk calculator adds stroke to MI as an endpoint to look at ASCVD risk, rather than just CHD risk, in the next decade. In addition, separate NHLBI risk-prediction equations were developed for non-Hispanic and black men and women.
Roger S. Blumenthal
Rather than the prior threshold of MI/CHD death risk of 20%, the expert panel strongly recommends statin therapy for patients surpassing a 7.5% threshold of ASCVD risk in the next decade. The panel made this judgment after examining event rates in the placebo groups in randomized trials.
They also weighed the number needed to treat with statins against the number needed to harm, defining the latter as statin-related diabetes. Those who are familiar with current best evidence, however, will quickly realize that the emphasis on statin-related diabetes is likely premature and artificial. Although we do not know the long-term effect of statin-related increases in blood sugar, short-term evidence, however, shows a lack of prognostic significance.
Many individuals without dyslipidemia will qualify for statin Rx
Based on NHANES data, it would seem that many more adults will qualify for statins because the 7.5% 10-year ASCVD risk threshold is exceeded by more than half of black men in their 50s and more than one-third of white men in that age range. Virtually all men will surpass this threshold by age 70 years. In contrast, for women in their 60s, about 70% of black women and 30% of white women will have a 10-year ASCVD risk >7.5%. The increased number of black individuals qualifying for statin therapy is due to their higher prevalence of hypertension.
Given the 7.5% ASCVD risk cutoff is featured so prominently in the guidelines, one would hope that quantification of risk by the new algorithm would be as accurate as possible. However, in our opinion, the precision of the risk calculator at the patient level appears disappointing based on trying to validate the risk score in the MESA and REGARD trials with C-statistics of only about 0.7 and 0.6, respectively.
Despite a wealth of research on additional tools like high-sensitivity C-reactive protein and coronary calcium measurements in the last decade to increase the accuracy of risk assessment in selected patients, this is limited to the small, new intermediate-risk group of 5% to 7.5% ASCVD risk or when risk estimates are still “uncertain.” We agree that such tools should be used selectively in situations of uncertainty. We believe that coronary calcium scanning also can be applied to patients with a predicted risk score of <5% who have a family history of premature CVD or two or more components of the metabolic syndrome as described in the 2010 AHA Cardiovascular Risk Assessment guidelines by Greenland and colleagues.
In our view, tie breakers such as coronary calcium measurement, carotid intima-media thickness determinations or high-sensitivity CRP measurements also should be considered to better personalize therapy in men older than 65 years or in recent-onset diabetics or in those with normal lipid values but whose risk score exceeds the 7.5% threshold due to their age and mild hypertension, where a one-size-fits-all approach is taken by the new guidelines.
New algorithm is simplified
Nevertheless, the new algorithm for initiating statin therapy is substantially simplified compared with prior ATP guidelines, which we view as a welcome change, and should help promote better implementation in clinical practice. Besides the new recommendation to use statin therapy in primary prevention adults with >7.5% ASCVD risk, the panel recommends statin therapy in the following three scenarios:
- High-intensity statin in adults aged 40 to 75 years who have ASCVD.
- High-intensity statin in adults with LDL >190 mg/dL.
- Moderate- to high-intensity statin in adults with diabetes mellitus aged 40 to 75 years.
With these broad sweeping recommendations, the guideline treats groups/populations, not individual patients. The new simplicity is not complemented by careful guidance on the further personalization of care. As Albert Einstein said, “Everything should be made as simple as possible, but not simpler.”
For example, do 100% of diabetics aged 40 to 75 years require moderate- to high-intensity statin therapy? In MESA, 30% of those with diabetes had no coronary calcium and a very low event rate during the next 5 years. Does not the duration of diabetes and diabetes control affect the overall risk assessment and therapeutic decision making? Would selective use of atherosclerosis imaging help promote more judicious use of statins and better personalize patient care? We would respond in the affirmative.
Adequacy of therapy and follow-up management
In contrast to prior ATP guidelines and other lipid guidelines around the world, the 2013 guidelines provide less-explicit guidance on assessing the adequacy of therapy and follow-up management. In ATP III, explicit LDL and non-HDL goals were used to assess the adequacy of therapy and guide ongoing therapies. The justification was the current best evidence from a synthesis of RCTs, epidemiology, genetics, pharmacology and basic laboratory data.
Two of the three core questions addressed by the panel were devoted to answering whether trials have randomly assigned patients to different lipid targets in isolation — not as a strategy of the trial, but as an isolated point of randomization. As expected, the panel found “no evidence.” It is unfortunate that the panel allocated such extensive resources to these questions, since the answer to them is common knowledge and has already been published in the literature. Yet, this does not mean that there is no evidence.
It means that the question must be reframed and other relevant evidence must be examined. The landmark COURAGE trial, not considered by this new cholesterol guideline, targeted explicit lipid goals, including an LDL goal of 60 mg/dL to 85 mg/dL. This was part of a comprehensive strategy to risk reduction and forms the foundation for the way that we approach chronic CHD patients in clinical practice today. Most recently, the AIM-HIGH trial titrated lipid-lowering medications to maintain LDL <70 mg/dL in both arms of the trial. Both groups had relatively low event rates, in fact so low that the trial’s endpoint was expanded mid-trial to accumulate more events.
In addition, the Treating to New Targets trial explicitly stated lipid goals in the protocol: “The occurrence of major cardiovascular outcomes was compared in two groups of patients: one group received 10 mg of atorvastatin daily with the goal of an average LDL cholesterol level of 100 mg/dL, and the other group received 80 mg of atorvastatin daily with the goal of an average LDL cholesterol level of 75 mg/dL.”
After finding no evidence, rather than consider the information from such RCTs and attempt to synthesize it with epidemiology, genetics, pharmacology and basic laboratory research, the panel closed its evaluation of the literature. For this reason, the panel’s conclusions differ from ATP III and dyslipidemia guidelines abroad using the current best evidence, including those from Europe and Canada.
Such guidelines assign the highest level of evidence (Class IA) to LDL goals in the high-risk treatment range (<100 mg/dL or <70 mg/dL) and also recognize non-HDL and apolipoprotein B as treatment goals. In fact, the 2013 guidelines directly contradict a recent AHA/ACC guideline on secondary prevention recommending LDL goals. The shift in scope of the new cholesterol guideline is reflected in its title, Treatment of Blood Cholesterol in Adults.
Compared with ATP III, “detection” and “evaluation” have been dropped. This leaves us wondering, who is now responsible for detection and evaluation? For now we must turn to prior AHA and ACC documents.
When it comes to the question of lipid goals, we tend to favor a more balanced look at the totality of current best evidence. Not just primary reports from trials reported before December 2009. Lipid goals were previously a major focus of the guidelines. Now they are minimized, which will cause confusion, and unfortunately are not based on any proof for the superiority of the new approach. In the future, we predict we will end up somewhere in between.
Age and the 2013 guidelines
Age factors prominently into the new guidelines, which suggest limiting statin use in patients with a chronologic age older than 75 years. In this age group, competing risks and potential for adverse effects certainly take greater precedence. The new document also argues that data are relatively sparse in this age group. In a CTT meta-analysis of 26 randomized clinical trials and 24,323 CV events, 1,872 occurred in those older than 75 years. Although the data are less than in adults in their 60s, it is certainly far more than in many other areas of medicine and, importantly, there was no evidence of heterogeneity of treatment effect by age.
Interpreting the evidence differently, the European guidelines state as a Class IB recommendation: “Treatment with statins is recommended for elderly patients with established CVD in the same way as for younger patients.” Specifically addressing primary prevention, the European guidelines state, as a Class IIb recommendation: “Statin therapy may be considered in elderly subjects free of CVD, particularly in the presence of at least one other CV risk factor besides age.” With increases in life expectancy and growth of the elderly population, further clarity and consensus on this topic will be even more critical.
We struggle with the idea of withholding statin therapy or any dyslipidemia intervention solely on the basis of chronologic age. Patients age differently and, for clinical purposes, we find it more useful to consider physiologic age than chronologic age. A robust elderly patient with high ASCVD risk, but without other major comorbidities, warrants strong consideration of a statin in our book. In the past decade, a phenomenon of undertreatment of high-risk older individuals has become increasingly apparent and, in this context, we worry about the effect of the new recommendation.
In patients younger than 75 years, as providers calculate the new ASCVD risk score, they will find that it is still dominated by chronologic age. We submit: Does sticking to risk scores that are driven by one’s birth certificate (age) leave us playing catch up on a disease that we otherwise may have a head start of 4 decades or so?
Once atherosclerotic plaque has reached an advanced stage, there may be a certain degree of residual risk that is not modifiable. There is a signal in the literature that dyslipidemia treatments may be even more efficacious if we start them earlier in life, in lower-risk patients with less advanced plaque. To this end, we would point readers to a persuasive recent editorial by Drs. Steinberg and Grundy that makes a compelling case for treating hypercholesterolemia at an earlier age.
What is ‘evidence based’?
Like any guideline, the new document states that these clinical recommendations are designed to inform clinical judgment and not to replace it, where the RCT evidence is not robust. The 2013 guidelines contrast with ATP III in implying a more evidence-based approach. However, if the panel were to apply the same definition of “evidence” used for examining lipid goals to the new ASCVD risk calculator, then the answer is the same: no evidence. This speaks to the artificial definition of evidence in this guideline.
The panel certainly has a different definition of “evidence-based” than the traditional definition. David Sackett and the pioneers of evidence-based medicine wrote a helpful article that said, “Evidence-based medicine is the conscientious, explicit and judicious use of current best evidence in making decisions about the care of individual patients.”
In fact, it was further specified that “Evidence-based medicine is not restricted to randomised trials and meta-analyses. It involves tracking down the best external evidence with which to answer our clinical questions.” The limited scope of this guideline and repeated mention of “resource limitations” means that the panel could not determine the “current best evidence” and therefore this guideline cannot be evidence-based.
Also note, “individual patients,” not ASCVD risk groups. This 2013 guideline follows a population health approach rather than a patient-centered therapy approach. RCTs of drug therapies provide an answer about the population average response to a drug. They are indexed to the drug, not to patients. In selecting patients for such drugs, the guideline focuses on the new risk equation. What is the precision of this at the individual patient-level? Would not atherosclerosis imaging improve the precision of risk estimation and therefore better personalize treatment decisions? Would not lipid and/or lipoprotein targets make for more personalized treatment?
In line with the pioneers of evidence-based medicine, the Institute of Medicine Standards for Developing Trustworthy Clinical Practice Guidelines recommend using “relevant available evidence,” not RCTs only. Indeed, well-designed observational studies can be as reliable and informative as RCTs. We submit, isn’t the job of guideline committees to critically appraise and synthesize the totality of evidence? The less guidelines do this, the more they place the burden on busy clinicians to synthesize networks of data and concepts on their own, many of whom will be less equipped than guideline writers to understand the strengths and limitations of studies.
Collaboration
A notable aspect of the updated guidelines is the collaboration between different writing groups in developing the guidelines. The 2013 guidelines were developed in collaboration with obesity, hypertension, lifestyle and risk-assessment writing groups. The guideline authors should certainly be commended for such efforts. We imagine that considering the different perspectives of members from each group was extremely challenging, but also a very fruitful process. The authors have given a lot of their time and service and we again wish to close by congratulating them on completing the herculean task.
In summary, well-intentioned CV specialists often disagree, but we wholeheartedly support at least 90% of the new guidelines, which declare that primary prevention is here to stay. Although many more adults will now qualify for statin therapy, heart-healthy dietary and exercise habits remain the foundation of primary prevention efforts. Following these new guidelines will allow clinicians to markedly reduce CV events in their patients.
We look forward to helping to promulgate important messages from these guidelines to clinicians throughout the United States, and we anticipate publishing key clinical research papers in 2014 to better inform the next set of CV risk-assessment guidelines. – by Seth S. Martin, MD, and Roger S. Blumenthal, MD
For more information:
Baigent C. Lancet. 2010;376:1670-1681.
Boden WE. N Engl J Med. 2007;356:1503-1516.
Boden WE. N Engl J Med. 2011;365:2255-2267.
Concato J. N Engl J Med. 2000;342:1887-1892.
Genest J. Can J Cardiol. 2009;25:567-579.
Greenland P. J Am Coll Cardiol. 2010;56:e50-103.
Ko DT. JAMA. 2004;291:1864-1870.
LaRosa JC. N Engl J Med. 2005;352:1425-1435.
Martin SS. Am J Cardiol. 2012;110:307-313.
Mihaylova B. Lancet. 2012;380:581-590.
Redberg RF. JAMA. 2012;307:1491-1492.
Redberg RF. Arch Intern Med. 2011;171:1594.
Reiner Z. Eur Heart J. 2011;32:1769-1818.
Sackett DL. BMJ. 1996;312;71-72.
Smith SC Jr. Circulation. 2011;124:2458-2473.
Steinberg D. J Am Coll Cardiol. 2012;60:2640-2642.