Author + information
- aIcahn School of Medicine at Mount Sinai, New York, New York
- bEmory School of Medicine, Atlanta, Georgia
- cUniversity of Minnesota & VA Medical Center, Minneapolis, Minnesota
- ↵∗Reprint requests and correspondence:
Dr. Jagat Narula, Icahn School of Medicine at Mount Sinai, One Gustave L. Levy Place, New York, New York 10029.
One of the important and least controversial concepts introduced by the 2013 American College of Cardiology and American Heart Association cholesterol guidelines, with a class IIa recommendation, is shared decision making (SDM) (1): informing patients about all available options and seeking their active input in decisions to start or not start therapy. Although this may yet be honored more in the breach than the observance, it is critically important to fully inform patients about the risk and benefits of potentially lifelong statin therapy. Aside from the obvious discussion of the benefits and potential risks of statins, are there other important issues that should also be discussed? One philosophical but as yet unanswered issue here is what data are appropriate for discussion in SDM, especially when major society guidelines are somewhat limited or somewhat controversial; does SDM mean telling patients only what the guidelines recommend (on the basis of randomized controlled trials [RCTs] or in other instances on just expert opinion) regarding risk stratification (whether frontline physicians agree with it or not)? Should patients also be told about what well-informed treating physicians know about all other data out there, outside of RCTs, that may have a bearing on the patients’ decision making? This is particularly relevant for using coronary artery calcium (CAC) to determine eligibility and strategy for statin therapy; CAC seems to predict future risk better than any other risk stratification method out there, and it reclassifies risk (both up and down) in a significant number of patients who were assessed with traditional risk factors (2). A CAC score of zero predicts low risk (that is unlikely to be modified by statin therapy) and also offers reasonably well-quantitated long-term warranty against future events. Transition from zero CAC to any CAC is momentous, significantly changing future risk profile. All of these should ordinarily guarantee CAC a role in SDM, but current guidelines restrict its role to a small group of subjects.
A Wider Role for CAC in SDM?
Whether the risk assessment recommendations derived from large populations are applicable to the risk of an individual patient engaged in SDM deserves further exploration. Should a patient be informed of a test that may provide more accurate individual risk stratification? In particular, CAC predicts risk better than any other risk factor–based assessment paradigm, by measuring the individual atherosclerotic plaque burden and removing the need for extrapolation from groups to an individual patient (3).
In the 2013 clinical practice guidelines, CAC has been relegated to a minor role, removed from the vast majority of decision making, and applied only as 1 of 5 novel risk markers to assist in patients not belonging to the 4 major groups or in those in whom questions still persist (1). Including CAC in SDM would allow patients to decide whether, in view of the available data, they wish to have their treatment individualized according to their CAC findings or to be treated as “average” patients who fall under the rubric of the fourth statin benefit group: “individuals without clinical ASCVD [atherosclerotic cardiovascular disease] or diabetes who are 40 to 75 years of age with [low-density lipoprotein cholesterol] 70-189 mg/dl and an estimated 10-year ASCVD risk of 7.5% or higher” (1). The same question may also apply to those with 10-year ASCVD risk of 5% to <7.5%, for whom statin therapy is “reasonable” (1).
Data Supporting the Use of CAC in SDM
Several studies strongly support considering CAC in SDM (Table 1). In 10-year follow-up of almost 5,000 asymptomatic patients from MESA (Multi-Ethnic Study of Atherosclerosis), 5% sustained hard ASCVD events (3). Of those qualifying for moderate- to high-intensity statin therapy, 40% had CAC scores of zero, with 5 events per 1,000 patient-years; 30% had CAC scores >100, which portended higher risk of 15 events per 1,000 patient-years, with a hazard ratio of 3 compared with patients with zero CAC and a number needed to treat of 28. In contrast, in the moderate-intensity statin group, almost 60% of subjects had CAC scores of zero, with 1.5 events per 1,000 patient years and a number needed to treat of 223, compared with 9 events per 1,000 patient-years and a number needed to treat of 45 for patients with CAC scores >100 and a hazard ratio of 6 compared with the group with zero CAC. Thus, 45% of patients eligible for statin therapy had zero CAC, with a 10-year event rate of only 4 events per 1,000 patient-years, and were in a low-risk group for which there is little evidence for statin benefit. And only 4% of those not considered eligible for statin therapy had CAC scores >100, placing them at least at intermediate risk and therefore potentially appropriate candidates for statins.
In yet another 10-year follow-up study of 5,000 patients and more than 300 events from MESA (4), a 7% rate of upward reclassification was observed in subjects who were considered to carry <7.5% pooled cohort equation (PCE) 10-year risk by CAC score >300 (or >75th percentile); the risk was reclassified to 13% with a relative risk of 4.0 compared with the 3% event rate for those who were not reclassified upward by CAC score. The number needed to scan to reclassify upward was <15. Similarly, in 10-year follow-up of about 2,500 statin-naive patients in the Framingham Heart Study (5), statin eligibility by the 2013 PCE was compared with the Adult Treatment Panel III guidelines; the hazard ratio for statin-eligible patients was 7 by PCE (40%), compared with 3 for the Adult Treatment Panel III statin-eligible (15%) patients. However, one-third of the PCE statin-eligible group had zero CAC, with very low likelihood of 10-year event rate. Comparable overtreatment of low-risk patients was noted in a computed tomography angiographic study of more than 3,000 patients (6) in which statin eligibility by PCE was noted in 35% of patients with neither calcified nor noncalcified plaque compared with 70% in those with any plaque.
There are convincing data supporting the low risk conferred by a CAC score of zero (3). Most recently (7), in 15-year follow-up of 9,715 asymptomatic patients, a mere 3% all-cause mortality rate was reported in patients with zero CAC, with progressive increases in mortality up to 30% in those with CAC >1,000. From the same cohort (8), a 15-year warranty (defined as all-cause mortality event rate <1% per year) was suggested for patients with zero CAC in the low and intermediate Framingham risk score groups and a 5-year warranty for the Framingham high-risk group. In summary, more than one-third of 2013 PCE asymptomatic patients would be considered for lifelong statin therapy but by virtue of their CAC scores of zero are at low risk for cardiovascular events (Table 1); only 5% of those not eligible for statins would be upgraded to eligibility by CAC.
The other side of this equation is whether current statin use guidelines are sufficient for SDM. One of the well-discussed challenges with the current guidelines is their suboptimal prognostic validation and overestimation of risk, which could potentially lead to overtreatment (Table 1). This overestimation of risk has been demonstrated in more than 1 population. As an example, application of the 2013 PCE to 10-year follow-up of the MESA cohort was associated with an 85% overestimation in men and a 65% overestimation in women (9); in men and women with 2013 PCE scores of 7.5% to 10%, the actual event rates were 3% and 5%, respectively. Incorporating CAC either on its own or as a part of newer calculators seems to help mitigate some of this overestimation. It is therefore reasonable for patients to know as part of SDM that: 1) there are limitations to what we currently use for risk stratification; and 2) although we do not as yet have RCT data for using CAC in improving risk stratification, there is a large body of evidence supporting its superior efficacy.
In summary, the data presented here are likely to support the discussion of CAC during SDM regarding statin use for patients in the fourth statin benefit group, and possibly in those with 10-year ASCVD risk of 5.0% to 7.5%. These groups would currently be prescribed statins on the basis of the 2013 guidelines. Although we cannot as yet advocate universal incorporation of CAC screening for risk stratification, the data seem to suggest that there might be tangible benefits to incorporating CAC into at least some SDM scenarios. Offering CAC screening to patients during the SDM encounter would be accompanied by the expectation that there is at least a 33% chance that they will be converted to a lower risk group for which there are no data supporting statin therapy. Of course, randomized trial data would be needed to support the improved effectiveness of using CAC compared with other strategies; such trials have been proposed. In the absence of RCT data, but in the presence of much collateral evidence of its efficacy in risk stratification, offering a discussion about CAC to individualize risk should be considered an important opportunity to engage patients in SDM and seek their preferences. Engaging patients in SDM with regard to the value of CAC in relation to currently recommended strategies would help patients see and weigh the data on which to guide their clinical decisions regarding life-style modifications at the least, and possibly even treatment options. It would also add to health awareness, which happens to be the primary goal of cardiovascular health promotion.
All authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- American College of Cardiology Foundation
- Stone N.J.,
- Robinson J.,
- Lichtenstein A.H.,
- et al.
- Hecht H.S.
- Nasir K.,
- Bittencourt M.S.,
- Blaha M.J.,
- et al.
- Yeboah J.,
- Polonsky T.A.,
- Young R.,
- et al.
- Johnson K.M.,
- Dowe D.A.
- Valenti V.,
- Ó Hartaigh B.,
- Heo R.,
- et al.