Admission of ignorance is the first step toward knowledge. —Benjamin Disraeli
Cardiovascular disease is responsible for most death and disability associated with diabetes. Those with diabetes suffer two to four times more myocardial infarctions, strokes, and sudden deaths than those without this condition (1–3). The medical community has recognized the increased vulnerability of these patients by assigning diabetes “coronary heart disease equivalent” status in its prevention guidelines (4). Since coronary syndromes tend to be clinically atypical and difficult to diagnose in diabetes, noninvasive approaches in asymptomatic or mildly symptomatic subjects have been proposed for screening and diagnosis (5,6).
Calcium screening as an approach
Pathologic studies have demonstrated a strong correlation between the presence of coronary calcium and the amount of atheromatous plaque (7). Since sensitive and noninvasive radiographic techniques are capable of detecting and quantitating calcium, calcium screening is an obvious candidate for screening for subclinical atherosclerosis. Morrison (8) has defined the purpose of screening for chronic diseases. “Screening for disease control can be defined as the examination of asymptomatic people in order to classify them as likely or not likely to have the disease that is the object of screening […]. The goal of screening is to reduce morbidity and mortality from disease among the people screened by early treatment of the cases discovered.”
If one accepts this definition, one can then surmise that three distinct conditions must be satisfied before a test can be considered adequate for screening.
The test must identify a subset of the population with the underlying pathology
The subset identified by the test must be highly likely to suffer adverse clinical consequences without interventions
Test-guided interventions must reduce the likelihood of these adverse consequences
Coronary calcium screening fulfills the first of these criteria in the general population. That is, coronary calcium presence and amount has been shown to be associated with the presence and amount of coronary atherosclerosis (7). It is also likely, though unproven, that calcium screening will fulfill this same first criterion in those with diabetes. In this issue of Diabetes Care, Starkman et al. (9) present evidence that coronary calcium and therefore coronary atherosclerosis is more extensive in type 1 diabetes than would be expected.
The second criterion for an effective screening test, prediction of incident events, is also fulfilled in the general population (10–12). Based on the clear association with events in multiple studies, the American College of Cardiology and the American Heart Association (13) have given a limited endorsement to the use of coronary calcium screening in nondiabetic asymptomatic subjects at intermediate disease risk.
However, there is no evidence for a clear and strong association between coronary calcium and future events in diabetic individuals. In fact, data from our lab (14) suggest that diabetic patients demonstrate a reduced correlation between calcium amount and risk of future events. In a group of 269 persons with diabetes undergoing computed tomography scanning for calcium and with 64 cardiovascular events during a follow-up period of 6.3 years, there was no significant relationship between increasing tertile of calcium score and event incidence. This finding may largely explain the difference between the relatively low relative risk reported in our earlier report (12), which regarded a cohort including these 269 diabetic subjects and the higher relative risks of coronary calcium reported by others who included fewer diabetic patients in their study samples (10).
The third criterion for effective screening, that is, proven efficacy in reducing morbidity and mortality, is not strictly fulfilled for the general population or for the diabetic subgroup of the population. There has been no clinical trial of the efficacy of calcium screening as there has been for mammography for breast cancer. The need for such a clinical trial is just beginning to be discussed (15).
Why might those with diabetes be different?
Why might the relationship between coronary calcium and future incident events be reduced or nullified in those with diabetes? Current knowledge regarding atherosclerosis allows pathologists to distinguish vulnerable atherosclerotic plaques, which are likely to rupture and cause coronary syndromes (acute myocardial infarction, sudden coronary death, and unstable angina), from those that are stable and unlikely to result in acute events. A model used to explain the relationship between coronary calcium and events relates calcified plaque, thought to be stable, with total plaque burden consisting of both stable calcified and vulnerable noncalcified lesions (16). This model assumes a relatively constant proportion of stable calcified to vulnerable noncalcified plaques across individuals in the population. If this ratio is indeed constant, those with much calcium will have an abundance of both types of plaque (stable and vulnerable) and will have an increased risk of events compared with those with less calcium. This model explains the increased risk of future events in those with increased calcium scores and is consistent with the premise that coronary calcium reflects “arterial age” (17).
This model, which assumes constant ratios of stable to vulnerable plaque, may not be valid in persons with diabetes. Plaque development is probably more rapid in diabetes (18), and this increase in activity may not be constant but rather sporadic, possibly accelerating during periods of poor glucose control (19,20). If this is the case, the proportional relationship between stable calcified and unstable noncalcified plaque amount may be uncoupled by diabetes. Those with diabetes, with little calcium, might thus be very vulnerable to events, and other diabetic persons with much calcium might be less vulnerable. This is of course hypothetical conjecture, and further research is needed to clarify how calcium relates to vulnerable plaque and to events in population groups with and without diabetes.
Our understanding of diabetic atherosclerosis is embryonic. Though we are beginning to work out the relationships between calcium and events in the general population, current epidemiologic follow-up studies have just begun in diabetes. The Epidemiology of Diabetes Interventions and Complications (EDIC) study in type 1 diabetes and the Risk Factors, Atherosclerosis and Cardiovascular Events in Diabetes (RACED) study in type 2 diabetes involve collecting computed tomography scan data on large cohorts of diabetic individuals and subsequent incident disease follow-up. These studies should be completed before we can know if calcium can accurately risk-stratify persons with diabetes. Then, the question of therapeutic efficacy must be addressed. Clinical trials may be needed to answer this question.
Those who promote the unproven application of calcium screening for diabetic patients do them a disservice. Unwarranted proselytizing of technology to those unfortunate enough to suffer from a condition will not assist in finding answers as to how best to prolong and enrich their lives. It is far better to be patient with our patients and to offer only those tests and therapies that we know are likely to help them.
References
Address correspondence to Robert Detrano, 1124 W. Carson St., Harbor UCLA Medical Center Research and Education Institute, Torrance, CA, 90502. E-mail: [email protected].
