Introduction
Coronary artery disease (CAD) is the leading cause of death in the US in both men and women, and in all ethnic groups that have been evaluated.1 The American College of Cardiology estimated the 1998 annual costs (direct and indirect) of CAD in the US to be US$368.4 billion, compared with cumulative cancer costs of US$189 billion.2 The scope of this problem is expanding as the population ages. The national focus on cardiovascular research, diagnosis and therapeutics has led to a recent trend toward decreased CAD-related mortality1 (see Figure 1). However, this trend may not benefit all groups equally, as African-Americans have experienced a slower decline in mortality than non- Hispanic whites, and women have a smaller decrease than men3 (see Figure 2). This has led to increased disparity between certain subgroups, a challenge recognized in the Healthy People 2010 Objectives.4 Early detection of CAD may be able to address this disparity by providing accurate, cost-effective diagnosis, risk stratification, and guides for treatment strategies to potentially reduce cardiac risk. The identification of ethnic minority patients at higher risk of cardiac events could provide more appropriate treatment- and therapy-impacting outcomes. In addition to primary and secondary risk factor management, myocardial perfusion imaging represents an effective strategy to achieve such a goal.
Ethnic Differences in Cardiovascular Mortality
Epidemiologic data show wide variability between ethnic groups with respect to cardiovascular disease. White males, with historically the highest coronary disease death rates, have experienced the sharpest decline in mortality. The decline has been less dramatic in black males, to the point that in 2001 African-American males had a higher overall CAD mortality rate than Caucasian males, 262.0 versus 228.4 (per 100,000), representing a reversal from the historical relationship. An even greater relative disparity is seen between African-American and Caucasian women1 (see Table 1). Excess cardiovascular mortality has not been demonstrated in all ethnic minorities. From 1999 CDC data, the overall death rate in Hispanics (138 per 100,000), Native Americans (123.9 per 100,000), and Asian/Pacific Islanders (115.5 per 100,000) is much lower than the overall rate in the US (177.8 per 100,000). Research into the importance of the biologic, social, and behavioral issues underlying these differences is on-going.
Traditional Cardiovascular Risk Factors in Ethnic Minorities
Familiar cardiovascular disease risk factors are present in varying percentages of different populations.3 The Third National Health and Nutrition Examination Survey (NHANES III) data show increased rates of obesity, hypertension, smoking, and diabetes among blacks and Hispanics.5 Furthermore, these risk factors may have differing effects in ethnic groups, with hypertension exerting a particularly deleterious effect among Blacks while diabetes disproportionately affects Hispanics.6 The impact of these risk factors is complex - increased cardiovascular mortality has been demonstrated in some ethnic minorities in the presence of less obstructive coronary disease.7 The disparity in cardiovascular mortality is not explained by differences in traditional risk factors.8 Strategies focusing on broad issues, such as socioeconomic status (SES), or individual risk stratification, such as that offered with myocardial perfusion imaging may help clarify the issue in a clinically relevant manner.
SES and Mortality
Indicators of SES have consistently been found to have an inverse relationship to cardiovascular death and all causes of mortality.9 Several studies have supported the concept that SES may have a greater impact on mortality than traditional risk factors.10 The interaction between ethnicity and SES is difficult to untangle. In the US, Blacks and Hispanics have lower mean incomes and a greater percentage of their populations are at or near the poverty level.11 Mortality differences may be explained by differences in risk factors, access to healthcare, environmental factors, psychological stressors, behavioral factors, and other yet to be understood factors.
Challenges of Ethnicity-based Research
Studies involving ethnic groups present many challenges beyond the association of cause and effect between biologic risk factors and outcomes. These challenges include classification of groups, clarification of financial status, and access to healthcare, as well as social and behavioral issues. Of note, there are no universally agreed on or commonly instituted methods of classifying race in the US. For example, many studies have used self-reported race or ethnicity to analyze groups, knowing that many people classify themselves as more than one ethnicity, i.e. African-American and Hispanic. This suggests that most classification schemes are oversimplifications. Addressing the complexity of the situation, the US Census now has more than 100 combinations of race and ethnicity from which a respondent can select.11 Artificial social subgroups may have as many differences within them as between them.
These differences may involve genetic, financial, social, environmental, psychological, behavioral, and personal differences.12,13 These factors, which studies may not account for, might influence medical outcomes at the level of patient choice, patient-provider interaction, access to healthcare, or through overt or subconscious racism. In short, race/ethnicity has a complicated interaction with many other factors regarding cardiovascular outcomes and presents many challenges to research in this area. Despite the inherent challenges, it is important to undertake research designed to identify and eliminate the disparities in cardiovascular outcomes through the early diagnosis and appropriate treatment of CAD. Stress myocardial perfusion imaging provides one strategy to accomplish just such a goal.
Identification of CAD
The tremendous number of patients affected by CAD has driven the development of effective, non-invasive methods to identify and risk-stratify patients with CAD. The premise behind this identification is the appropriate application of treatment strategies to individual patients to prevent future events, i.e. death or myocardial infarction (MI). Historically, exercise treadmill testing (ETT) with electrocardiogram (ECG) monitoring has a sensitivity of 50% to 70% and a specificity of 60% to 80%; relatively low accuracy for the detection of such a substantial disease.14 Stress MPI represents an advancement beyond ETT with improved sensitivity (81% to 84%) and specificity (87% to 91%) and has become an important means of evaluating patients with known or suspected CAD.14
Benefits of MPI
MPI has been demonstrated to be a powerful indicator of future cardiac events with a high diagnostic accuracy for CAD. Specifically, MPI is valuable in assessing risk for MI and cardiac death,15,16 evaluating myocardial viability,17 as a to guide future therapy,18 and for pre-operative risk assessment.19,20 The incremental value of single photon emission computed tomography (SPECT) MPI beyond clinical information and stress ECG data has been demonstrated in patients at intermediate-to-high risk.15,16 In addition to the general population, SPECT MPI has been validated in diabetic patients,21,22 women,23 following MI24,25 or revascularization,26 in the setting of acute chest pain,27 and in those unable to exercise,28-30 or with left bundle branch block.31 Unfortunately, few studies with MPI have addressed the ethnicity of the participants.
Differences in risk factors and obstructive coronary disease prevalence between ethnic groups suggest that there may be a discrepancy in the expression of cardiovascular disease in these groups. Increasing disparity in CVD mortality between ethnic groups underlies the importance of improved detection and treatment of CVD in ethnic minorities with the goal of benefiting all members of society equally. It is thus reasonable to consider whether myocardial perfusion imaging has the ability to provide the same powerful role in risk stratification of ethnic subgroups as it does in the population as a whole, and whether low-, intermediate-, and high-risk patients can be identified.
Myocardial Perfusion Imaging in Ethnic Minorities - Prior Research
Limited data have been published regarding risk stratification with myocardial perfusion imaging in racial or ethnic minorities.32-36 Alkeylani et al. analyzed a multi-center registry for cardiovascular death and myocardial infarction among 864 Caucasian and 222 African-American patients referred for nuclear stress testing.32 Mean follow-up was 2.3 years. Among patients with normal perfusion imaging they found a two-fold higher cardiac death and MI rate in African-Americans (2.3% versus 1.0%). However, the difference was not statistically different. Patients with abnormal perfusion imaging had similar cardiac event rates (Black 11.0%, White 11.8%) (see Figure 3). Both ethnic groups demonstrated a strong relationship of multi-vessel perfusion defects and cardiac events. Despite the relatively small number of patients in this study, it appears that perfusion imaging provides similar risk stratification among blacks and whites. It is not clear whether there is a difference in cardiac event rates between Blacks and Whites after normal perfusion imaging - baseline demographics differed in regard to age, gender, diabetes, and hypertension. Co-morbidities may cluster in or have exaggerated effects in individual ethnic groups. Additionally, no evaluation of socioeconomic status was made. It would be erroneous to assume that this study demonstrates equivalent cardiac event rates in Blacks and Whites, but rather, it does demonstrate similar risk stratification based on perfusion data within each group.
In another study, Akinboboye et al. studied an urban population of African-Americans with normal stress thallium SPECT MPI.33 Overall, they found a 2% cardiac event rate at one year. Among patients undergoing treadmill exercise as the mode of stress they observed a very low (1%) annualized rate of cardiac death or MI. However, among the 35% of patients undergoing pharmacologic MPI, the cardiac event rate was 5% at one year. There was no comparison population, and follow-up was only 81%.
However, the authors concluded that normal exercise MPI in this population was associated with low cardiac event rates, similar to that found in other populations. Patients undergoing pharmacologic rather than exercise stress, had higher cardiac event rates, similar to a trend noted in a recent meta-analysis.27,36
In an analysis of our single center database, encompassing nearly 7,000 patients, no statistical difference was found in the gross or age-adjusted cardiac event rates between White, Black and Hispanic patients referred for stress MPI with either normal (0.7% to 0.8% per year) or abnormal perfusion (2.3% to 3.0% per year)34 (see Figure 4). Abnormal myocardial perfusion (any ischemia or infarct) was associated with a similar magnitude of significantly elevated cardiovascular risk (myocardial infarction or cardiac death) in all ethnic groups as compared with those with normal perfusion. This suggests that SPECT MPI imaging results afford similar risk stratification in all three ethnic groups. Importantly, those with normal perfusion all had very low annual cardiac event rates.
Diabetes was an important risk factor in both White and Hispanic patients with previously normal studies. Among patients with normal perfusion in the two largest subgroups in the study, White (n=4,565) and Hispanic (n=746), there was a similar increase in cardiac event rates among diabetics in each group compared with those without diabetes35 (see Figure 5).There was no statistical difference in cardiac event rates between Whites and Hispanics.
Evidence suggesting higher cardiovascular risk across the spectrum of MPI is emerging. In a multi-center study presented in abstract form, Shaw et al. evaluated the impact of ethnicity in 7,849 patients undergoing SPECT MPI (African Americans=1,993, Hispanic=464, Caucasian=5258).36 Similar to the study by Noble et al.,34 they noted increased cardiac event rates with abnormal perfusion in all ethnic groups. However, in contrast, they also found significantly higher event rates for African-American and Hispanic patients at every level of perfusion abnormality compared with whites. After adjustment for multiple clinical factors, annual ischemic heart disease death-rates ranged from 0.2% to 3% for Whites and 0.8% to 6.5% for African-Americans for mild-to-severely-abnormal perfusion, and combined cardiovascular deaths were also higher in these groups. This represents at least twice the ischemic death rate in African-Americans.
The African-American and Hispanic groups in this study may high risk subsets of these ethnic groups, however, as they had higher prevalence of CAD (prior MI or abnormal perfusion) than Caucasians, a reversal from prior literature.1,32 Deathrates may be different in African Americans and Caucasians due to many factors such as, co-morbidities, socioeconomic status, and risk factor modification or referral bias. Referral for diagnostic testing and revascularization demonstrate divergent ethnic patterns31, 38-40 and may have affected the decision to refer patients for testing in this study imparting selection bias. As noted previously, there are many challenges in assessing the impact of ethnicity on outcomes.
Conclusion
Ethnic groups differ in the incidence and prevalence of cardiovascular disease and its associated risk factors. Many factors confound the relationship between race/ethnicity and cardiovascular outcomes. However, the data currently available support the use of stress MPI for risk stratification in ethnic subgroups. Disparities in absolute cardiac event rates between ethnic groups may exist despite similar imaging results, perhaps reflecting co-morbidities, differences in risk factor modification, or selection bias in the populations studied. The relative value for risk stratification of normal versus abnormal perfusion imaging appears to be consistent and of similar magnitude in all three ethnic groups that have been reported to date. The burden of cardiovascular diseases in certain ethnic groups is increasing, or decreasing at a slower rate than others. The appropriate use of myocardial perfusion imaging could help to risk stratify these groups as it has been shown to do in the general population. Based on available evidence, current guidelines regarding myocardial perfusion imaging should be applied uniformly, regardless of race or ethnicity. Further research is needed to assess disparities in referral patterns for non-invasive testing and to evaluate whether perfusion guided treatment has an effect on minimizing disparities in cardiovascular mortality between ethnic groups.