Introduction

According to the World Health Organization (WHO), cardiovascular disease (CVD) remains the leading cause of death in adult men and women worldwide [1]. Most of these deaths are related to ischemic heart disease (IHD). IHD seems to be a more appropriate term, since obstructive and non-obstructive coronary artery disease (CAD) are only part of spectrum and many ischemic syndromes in women may result from significant endothelial dysfunction, coronary spam, and microcirculatory failure, without significant obstructive CAD [2, 3].

At the world level, IHD accounts for 38 % of deaths in women and cerebral vascular disease for 37 %, while for men 46 and 34 %, respectively. This is not different in Latin America, where most of these deaths in men and women are related to IHD or CAD [4]. The mortality rate due to IHD varies within Latin America, being highest in Honduras (152.4 per 100.000 of the population), Guyana (151.4), Cuba (111.3), Jamaica (110.7), and Venezuela (107). This is in contrast to what is found in France (29.2), Japan (31.2), and Canada (66.2). Brazil and the USA have similar IHD mortality rates at 81.2 and 80.5, respectively. When we look at gender-specific mortality, Brazil, Colombia, Venezuela, Bolivia, Paraguay, and the Guyanas are the countries with the highest mortality rates due to IHD in women compared to other countries in the Americas, North America included [1].

Women, on average are affected by CVD later in life and live approximately 6–7 years more than men, in most countries, with some rare exceptions around the world (Tuvalu, Tonga, Kuwait, and Qatar) were men live longer. Nevertheless, for women living in Brazil, the largest and most populated Latin American country, life expectancy is on average 79.8 years, which is much lower than in Japan, for example, where women are expected to live an average of 87.3 years [1]. This is also in contrast to the life expectancy of men in Brazil at 72.6 years and of men in Japan at 85 years. Considering that CAD is the leading cause of mortality in Brazil, but the third cause of mortality in Japan, could prevention, appropriate diagnosis and management help Brazilian women to live longer, perhaps even 7.4 years longer like the Japanese women? Perhaps so, if the main reason for this difference is not genetic but related to modifiable factors that will be discussed in details in this review.

Inadequate nutrition habits and sedentary lifestyle, so common in the modern occidental life, are contributing to the increase of important risk factors for CAD, such as type 2 diabetes mellitus (DM), observed in Latin America [5]. The WHO and the International Diabetes Federation (IDF) predict that the number of diabetics in the world will double by the year 2030. Also, according to the IDF, nearly 80 % of diabetics live in low to mid-income countries, contributing to an overwhelming, increasing burden of CAD. Having this in mind, it is not surprising the WHO estimates that 80 % of all the 17.3 million deaths due to CVD in the world every year occur in low to mid-income countries. Women in Latin America, and especially women with DM, are at a high risk of developing and dying due to CAD. In this review, we will focus our attention to the hearts of Latin America women, the overwhelming burden of DM and CAD, and the potential role of nuclear cardiology to stratify risk and guide management in our region.

Special Attention to Diabetic Women in Latin America

One of the challenges to detect CAD in women, particularly women with DM, is the atypical presentation or the absence of symptoms, frequently delaying diagnosis and appropriate treatment [6]. Once CAD is established, women tend to be older, have more diffuse disease, and a worse prognosis compared to men [7], including a higher mortality rate after acute myocardial infarction and after coronary bypass surgery.

The IDF estimates a prevalence of 8.1 % of Latin American population currently affected by DM [8]. This prevalence is expected to increase by at least 60 % in the next 15 years. Similarly to other regions in the world, Latin America population is expected to age significantly in the next few decades. Urbanization is a growing phenomenon in this region. All of these, coupled with progressive changes in nutrition habits and adoption of a sedentary modern lifestyle, are contributing to increase the prevalence of DM in Latin America. A similar phenomenon is observed in Latino immigrants living in the USA [9]. In fact, a study conducted by the American Heart Association in 2012 demonstrated that Latinos living in the USA constitute the group that lacks the most awareness about heart disease as the number one killer in women [10].

There is no question that DM is an important risk factor for CAD, especially in women, with major implications for the use of imaging, including the appropriate use of nuclear cardiology, to diagnose, estimate risk and guide treatment [2, 5–9]. At our institution, Quanta Diagnóstico e Terapia, a referral center for nuclear cardiology procedures in Latin America located in Brazil, we observe that of all women sent for SPECT-MPI, 22.5 % already have DM (Table 1), which is potentially underestimated considering that under diagnosis of DM is a worldwide phenomenon. Importantly, 41 % of women with DM at our institution are found to have an abnormal MPI, which is much higher than the 23 %, for women without DM. Data from this registry will be discussed in this review to illustrate the role of SPECT-MPI to evaluate CAD in women from our region.

Table 1 Gender differences in 41,671 patients undergoing SPECT-MPI in a referral center in Brazil
Full size table

Nuclear Cardiology Registry—a Focus on Latin American Women, Gender Differences, and Comparison with Other World Regions

To the best of our knowledge, this nuclear cardiology registry is the largest in Latin America and one of the largest in the world. Gender differences in Latin American patients undergoing CAD evaluation will be shown. Currently, data collected prospectively for nearly 10 years from 41,671 patients undergoing SPECT-MPI is available for review (Table 1). Of these, 55.3 % are men. Women at 44.7 % are surprisingly well represented in this registry, considering that most large trials on CAD from the USA and Europe show an underrepresentation of women [2, 7, 8]. The prevalence of DM was similar for both genders at around 22 %. Women had less known CAD prior to SPECT compared to men (20.9 vs 35.5 %). The rate of SPECT-MPI abnormalities was 27 % for women and 35 % for men. These findings are surprisingly high when compared to some laboratories in the USA, reporting abnormalities rates of less than 10 % [11]. Women in our registry were a mean of 2 years older. Of all the perfusion abnormalities found they were mainly ischemia. In women: ischemia in 85 %, mixed ischemia and scar in 8 %, and scar in 7 %. In men: ischemia in 64 %, mixed ischemia and scar in 19 %, and scar in 17 %. Men had more scar compared to women, which was consistent with a higher rate of known CAD prior to SPECT (Table 1). Women had a lower prevalence of more severe perfusion defects, defined as SSS >13 (4.9 vs 11 %).

When we focus on the 5027 women with abnormal SPECT-MPI, some interesting observations can be made, especially in relation to their symptoms that led to SPECT (Fig. 1). Most of these patients (87 %) had either no chest pain (n = 2488, 50 %) or had atypical pain (n = 1870, 37 %). This is striking, considering that all of them were found to have abnormal MPI, and the overwhelming majority ischemic results that would intuitively makes us think of angina. But angina was infrequent in ischemic patients. In fact, rarely these women presented with typical, exercise induced, chest pain (n = 669, 13 %). The small number of patients with typical angina seen in nuclear cardiology may reflect a common practice in our region of sending these patients directly to invasive coronary angiography but it also points to the frequent atypical presentation of IHD.

Fig. 1
figure 1

Symptoms at presentation in women and men found to have abnormal perfusion on SPECT-MPI. All of these 13,067 patients (5027 women and 8040 men) had abnormal studies; the overwhelming majority showing ischemia. Both genders presented either with no chest pain or had atypical pain at the time of the study and only a minority of them had typical pain. For patients with chest pain, that was the indication for SPECT. For patients with no chest pain, SPECT was appropriately indicated to evaluate a positive or inconclusive exercise test, the presence of angina equivalent, non-interpretable rest ECG, DM, contrasted CT angiography or elevated calcium score showing advanced CAD, and preoperative evaluation. The lower rate of typical angina in patients found to have abnormal SPECT-MPI may be the result of a pattern in this region to send typical angina patients directly to invasive coronary angiography and not preferentially to the nuclear laboratory

Full size image

Atypical presentation of IHD in women, such as angina equivalents, non-specific discomforts and fatigue have previously been described [2, 6, 7] and may have been the reason many of the SPECT in our registry were performed in patients with no pain (Fig. 2). Even during acute coronary syndrome women behaves in atypical ways. In a large registry including over 1 million patients hospitalized with myocardial infarction in the USA, women often (42 %) presented without chest pain, which was more frequent compared to men (31 %). Women also had higher mortality rates than men (14.6 vs 10.3 %) [12]. Most women having sudden death (63–64 %) have no prior symptoms [12, 13].

Fig. 2
figure 2

Gender differences in clinical presentation and abnormality rate on SPECT-MPI. In 41,671 Latin American patients sent to nuclear cardiology, gender differences can be observed, regarding symptoms and rate of perfusion abnormalities on SPECT-MPI related to each symptom category. Although no pain or atypical pain was seen at a much higher number of women, abnormality rate was higher for men. More striking was the difference in abnormality rate in women and men presenting with typical pain, suggesting that typical angina would be more predictive of abnormal SPECT in men. *p < 0.001 women compared to men

Full size image

In our registry, men having an abnormal SPECT-MPI also frequently presented with no pain or atypical symptoms (Figs. 1 and 2). The resulting abnormal tests demonstrate that the indications for SPECT were indeed appropriate despite the absence of typical pain or no pain at all in many cases. SPECT indications included a positive or inconclusive exercise test, non-interpretable rest ECG, the presence of angina equivalent, DM, contrasted CT angiography showing CAD, and preoperative evaluation. Typical pain seemed to have a different meaning between genders (Fig. 2). Note that only 39 % of women with typical pain had abnormalities on SPECT, which is much lower than the rate observed for men (61 %) with the same symptoms. Therefore, one can conclude that most of the patients with abnormal SPECT, independent of gender, will not have typical pain (Fig. 1) but when presenting with typical pain (Fig. 2), it is more likely for men than for women to have an abnormal SPECT result. This observation from Latin America goes along with what has been shown in other parts of the world, where women frequently present with CAD with no chest pain or atypical symptoms. Nevertheless, from our experience in the nuclear laboratory, this does not seem to be exclusive of women.

We have previously shown, using part of this registry, that the presence of any myocardial perfusion abnormality, leading to any summed stress score (SSS) above 4, is a predictor of death, in the largest published prognostic study from Latin America women using nuclear cardiology [14]. That study involved 2225 women, mean age 64.5 ± 5.6 years, followed for nearly 4 years for all-cause mortality. Women with abnormal SPECT-MPI had 3.02 times higher incidence of all-cause death during the follow-up period, when compared with women with normal results. Interestingly, in that study, in a similar way to what we are now observing for the entire registry, there is no relationship between symptoms and the degree of perfusion abnormalities in women, although perfusion deficits were highly predictive of death (Table 2).

Table 2 Clinical characteristics, LVEF, perfusion scores, absolute mortality, and survival rates stratified by the SPECT-MPI results in 2250 women followed for nearly 4 years
Full size table

One additional gender difference in our registry regards pharmacological stimulation for stress. A total of 6670 (36.3 %) out of 18,628 women compared to 5580 (24.2 %) out of 23,043 men needed pharmacological stress (Table 1). This clearly shows that women in our region require more pharmacological stress compared to men. Nevertheless, men, for the same stress modality, had consistently a higher abnormality rate on SPECT-MPI compared to women (Fig. 3). Both genders had a higher rate of abnormalities on SPECT when pharmacological stress was performed, but this rate was more than doubled in women requiring pharmacological stress compared to exercise. For both genders, this probably has to do with sicker and older patients not being able to exercise on the treadmill and having comorbidities. Regarding women and exercise capacity, the WOMEN randomized trial (What is the Optimal Method for Ischemia Evaluation in Women) has shown that female patients at low to intermediate risk of having IHD, who are able to exercise more than five metabolic equivalents, may benefit from an exercise test alone as a first test compared to nuclear imaging. A similar event rate was found for both groups after a 2-year follow-up [15]. In the WOMEN trial 1 in every five women, initially randomized to exercise test, required a follow-up nuclear imaging. The conclusion of that study, supported by current guidelines [6], is that nuclear imaging should be reserved for women suspected of IHD, with a prior positive exercise test, or sent directly to nuclear if limited exercise capacity, intermediate to high risk of IHD, or abnormal/non-interpretable resting ECG (class I, level of evidence B).

Fig. 3
figure 3

Gender differences in stress modality and abnormality rate on SPECT-MPI. Observe that among the 41,617 Latin American patients in our registry, women tend to require more pharmacological stress compared to men but men had a higher abnormality rate on SPECT-MPI. Both genders had a higher rate of abnormalities on SPECT when pharmacological stress was performed. This relates to a sicker and older sample of the population undergoing pharmacological stress. Women requiring pharmacological stress more than double their chance of SPECT abnormalities compared to women who were able to exercise. *p < 0.001 women compared to men

Full size image

Extent and Severity of Perfusion Abnormalities—Risk Estimates in Women

Risk estimations from nuclear cardiology are based on a principle, supported by the literature, that the absence of myocardial perfusion defects confers a good prognosis while increasing magnitudes of perfusion abnormalities correlates to progressively worse prognosis [6, 14]. A normal myocardial perfusion and left ventricular function on GATED SPECT-MPI is associated with a low cardiac event rate of less than 1 % in 1 year. On the other hand, the larger and more severe the perfusion abnormality, the worse the outcome, especially for summed stress score (SSS), above 13.

When we look at our registry and focus on patients with severe perfusion abnormalities, defined as a SSS above 13, we observe that 2529 men, out of 23,043 (11 %) and 918 women, out of 18,628 (4.9 %) are in this higher risk category. Interestingly, for women, only two clinical variables, known CAD and a high Duke treadmill score were predictive of a higher than 10 % chance of identifying SSS >13 (Fig. 4). Nevertheless, for men, there were seven variables that were predictive: dyslipidemia, DM, intermediate risk Duke, high-risk Duke, need for pharmacological stress, known CAD, and having typical pain. These data gives us a perspective of the gender differences among patients referred to nuclear cardiology in our region, as well as which clinical variables are more likely associated with severe findings on SPECT-MPI. Although women referred to us are less likely than men to have severe perfusion abnormalities, these patients are well known to have a bad outcome as demonstrated previously, by us included. Our data demonstrated that mortality rate increases progressively with increasing degrees of SPECT abnormalities (Table 2) and that women with abnormal SPECT-MPI had a 13.1 % mortality rate after 4 years, much higher compared to 4 % of women with normal SPECT-MPI, but the worse prognosis was for patients with severe perfusion deficits [14].

Fig. 4
figure 4

Clinical variables, gender differences, and percentage of patients with severe perfusion defects. Overall, 4.9 % of women and 11 % of men in our registry of 41,617 patients were found to have a severe perfusion abnormality with a summed stress score (SSS) above 13. Observe gender differences and the probability of finding a severe perfusion defect, considering the same clinical variable. Only known CAD and a high-risk Duke treadmill score were associated with a higher than 10 % chance of finding SSS >13 in women

Full size image

Severe CAD in Latin America Women: Management Guided by Risk Estimates

Among patients seen in our institutions, there were cases like these: a 54-year-old, obese, and hypertensive women, with no prior history of CAD, sent to SPECT-MPI due to atypical symptoms and fatigue on exertion. As part of the study, she underwent a Bruce protocol, developing significant shortness of breath at low workload. Exercise had to be interrupted due to symptoms and important downsloping ST segment depression on the ECG. At 2 min and 42 s into recovery, she develops an episode of ischemia induced torsades des pointes ventricular tachycardia (Fig. 5) and arrests. Her arrest was successfully managed. There is no question that these are high-risk findings. Coronary angiography revealed severe obstructive diffuse three vessels CAD. Surgical revascularization was successfully performed. This case exemplifies some important aspects of CAD in woman which is pertinent also for our region: atypical symptoms at presentation, low workload capacity, and advanced severe disease at the time of diagnosis. This same pattern of clinical presentation with atypical symptoms can be seen in the other case examples shown in this review (Figs. 6 and 7) and illustrates what our nuclear cardiology registry is showing: the vast majority of women found to have abnormal perfusion present with atypical symptoms and/or no chest pain (Figs. 1 and 2).

Fig. 5
figure 5

Ischemia-induced ventricular tachycardia (torsades des pointes) during the recovery phase after exercise in a 54-year-old patient presenting with atypical symptoms for CAD evaluation using SPECT-MPI. Observe the significant downsloping ST segment depression just before the tachycardia. Invasive coronary angiography was performed leading to successful three vessels surgical myocardial revascularization. This is a clear example of how severe myocardial ischemia can induce malignant and potentially fatal ventricular arrhythmia

Full size image
Fig. 6
figure 6

SPECT-MPI of a 75-year-old woman presenting with shortness of breath and no chest pain. Observe the large moderate to severe defect, mostly reversible defect (ischemia), involving the infero-septal, septal, and apical regions of the heart. There is also evidence of transient ischemic dilation. These are high-risk markers for death. This patient had sudden death at home within days following this study

Full size image
Fig. 7
figure 7

SPECT-MPI images of a 57-year-old female presenting with atypical symptoms. Observe the severe reversible defect (ischemia) involving a large area of the left ventricle, in the anterior wall, septum, and apex. The finding is consistent with severe obstructive disease of the proximal left anterior descending branch (LAD). The calculated summed stress score was 22, and there is transient ischemic dilation. High-risk markers are present. The patient was successfully managed with one stent to the proximal LAD and optimal medical therapy

Full size image

A 75-year-old woman with hypertension, dyslipidemia, with no chest pain but shortness of breath on exertion, presents for a dobutamine stress SPECT-MPI. Dobutamine was used because of chronic obstructive pulmonary disease. Tracer was injected at a heart rate of 139 beats per minute during dobutamine. She had no angina and no ECG changes during the test. SPECT-MPI (Fig. 6) revealed a large moderate to severe perfusion defect, mostly reversible (ischemia) in the infero-septal, septal, and apical regions of the heart. SSS was calculated at 24. There is also evidence of transient ischemic dilation induced by stress. All of these are high risk markers. Unfortunately, this patient had sudden death at home within days following this study. This case demonstrates the relationship of high risk markers in nuclear cardiology and the risk of cardiac death.

A-57-years-old female was presented for SPECT-MPI with atypical chest pain (Fig. 7). She has hypertension, dyslipidemia, and a family history, with her mother at age of 58 years old having CAD requiring revascularization. Three years prior to the current study, she underwent a normal SPECT-MPI. She now exercised 8 min on the Bruce protocol, had ECG changes with 3.5 mm horizontal ST depression, and developed some anterior cervical pain suggestive of myocardial ischemia. Her Duke treadmill score was calculated (+8 − 5 × 3, 5 − 4 × 1) and resulted in −13.5. This finding was considered of high risk. Only 234 women in our entire registry had a high-risk Duke score; nevertheless, 144 of them (61 %) where found to have abnormal perfusion. Observe that the SPECT-MPI of this patient shows a severe reversible defect (ischemia) involving a large area of the left ventricle, in the anterior wall, septum, and apex of the heart. For purposes of semiquantitation the heart was divided into 3 coronary territories and 17 segments. The left anterior descending (LAD) branch territory corresponds to 7 of them, which are all compromised in this patient, giving us the extent of the defect, which in this case is large. Considering the degree of severity in most of these segments, varying from moderate to severe, receiving a grade varying from 2 till 4 in a scale of 0 to 4 and that 7 segments were involved, the SSS calculated for this patient was 22. High-risk finding demonstrated by SPECT-MPI with superior capacity to discriminate patients more likely to have events compared to the exercise variables alone. In addition, nuclear imaging provided localization, extent, and severity of myocardial ischemia, which is usually not possible by exercise testing alone. The SPECT-MPI finding in this patient is consistent with obstructive CAD involving the proximal LAD, which was confirmed on coronary angiography and the patient was treated with implant of one stent coupled with optimal medical therapy.

Observe that in our registry 4.9 % of women have a SSS >13 (Table 1) similar to the SPECT-MPI images shown on Figs. 6 and 7. Looking at the correlation between exercise test variables and perfusion, a high-risk Duke treadmill score was infrequent in our patients sent for SPECT. Only 234 women and 499 men had a high-risk Duke, but most of them showed abnormal perfusion 61 and 69 %, respectively. We observe that only 15 % of the women with a high-risk Duke treadmill score were found to have a perfusion defect with SSS >13 (Fig. 4), similarly to the case shown on Fig. 7. Although these severe myocardial perfusion findings are less frequent in women compared to men, it is important to know that they carry a bad prognosis. We have previously shown that severe reversible defects (SSS >13), correlate with high mortality rate in women (Table 2).

The three case examples shown emphasize the message that women with severe CAD may present with atypical symptoms, as most of the patients with abnormal SPECT in our institution (Fig. 1). It also demonstrates how information and the integration of variables from the exercise testing and the perfusion images can be used, impacting treatment decisions. One of these patients had cardiac arrest due to severe exercise-induced ischemia, which could have been fatal if no immediate medical attention was available, which led to surgical revascularization. The other patient, with severe ischemia and high-risk markers on SPECT had sudden death at home within days of that study, not even allowing time for a more comprehensive treatment that the case deserved. Lastly, the third patient had an impressive extensive and large ischemic defect on SPECT, at a relatively young age. This information was useful to guide treatment with revascularization combined with optimal medical therapy, and to initiate aggressive secondary prevention, as per current guidelines, attempting to reduce ischemia-related mortality.

Conclusions

IHD is the leading cause of mortality for Latin American women. Unfortunately, with aging of the population, urbanization, and adoption of modern lifestyle, the prevalence of DM and, consequently, IHD mortality is expected to increase in Latin America. Prevention will be essential to fight this overwhelming burden but risk stratification to guide appropriate and cost-effective management will become increasingly important. Nuclear cardiology offers a very important tool for risk stratification and to identify women at high risk of cardiac death. In this review, we showed, from a nuclear cardiology perspective and using data from the largest registry from our region, which are the variables most useful to predict SPECT-MPI abnormalities, to identify IHD and to guide patient management in this part of the world.