Highlights
-
•
CAC score of 0 has 98% NPV for ruling out obstructive CAD in stable chest pain.
-
•
CAC = 0 linked to 81% lower MACE risk versus CAC > 0 in symptomatic patients.
-
•
CAC = 0 remains effective in intermediate/high PTP groups with 98% NPV.
-
•
Annualized MACE rate: 0.4% for CAC = 0 versus 1.9% for CAC > 0 (p < 0.01).
-
•
CAC testing is cost-effective, simpler, and uses less radiation than CCTA.
Obesity, defined as Body Mass Index (BMI) exceeding 30 kg/m 2, is the epidemic of our century. Obesity is estimated to affect more than 40% of the adult population in the United States. Due to the common associations with other prevalent cardiovascular risk factors (diabetes, metabolic syndrome, hypertension, obstructive sleep apnea, hyperlipidemia), obesity is encountered very frequently in patients with established cardiovascular disease or in patients who undergo evaluation for obstructive coronary artery disease. High BMI can be a challenge not only for many of the traditional noninvasive methods used to diagnose coronary artery disease, including treadmill stress test, nuclear imaging, but also coronary computed tomography angiography due to the unique characteristics in those patients, including but not limited to inability to exercise, attenuation artifact, and need for higher radiation dose. With this review, we hope to shed light on the published literature and the gaps for future research.
Graphical Abstract
In the last few decades, the obesity epidemic has increased and is now a significant burden on global health. Patients with obesity are classified by body mass index (BMI): Class 1 mild obesity (30.0–34.9 kg/m²), Class 2 moderate obesity (35.0–39.9 kg/m²), and Class 3 severe/extreme obesity (≥ 40.0 kg/m²), while patients with BMI of 25.0–29.9 kg/m² are classified as overweight. Obesity affects about 40% of the US population and has a causative association with cardiovascular disease. Patients with obesity have an increased risk for atherosclerosis, coronary artery calcification, and subsequent development of obstructive coronary artery disease (CAD). Prompt diagnosis and management are essential to limit the acceleration of CAD and the disease severity. Noninvasive cardiac imaging constitutes a useful tool to investigate the presence of CAD and its significance, and acts as a gatekeeper to invasive coronary angiography, without affecting safety outcomes. Noninvasive cardiac imaging can also help guide treatment decisions for preventative medications. , Because of excess adiposity, obesity can potentially undermine image interpretation and diagnostic accuracy, not only increasing the false positive but also the false negative rates. At the same time, obesity is known to be associated with higher radiation uptake during nuclear stress imaging or computed tomography imaging, which is an important consideration, especially for younger patients or for patients who will undergo serial imaging in the future. In this narrative review, we will examine if and to what extent obesity influences the diagnostic performance of the whole spectrum of noninvasive cardiac imaging for coronary evaluation, ranging from exercise stress testing to nuclear imaging, cardiac MRI, and we will especially focus on coronary CT.
Obesity and Cardiovascular Disease (CVD)
High BMI is associated with several cardiometabolic and inflammatory risk factors such as hypertension, diabetes, and dyslipidemia. Obesity can also lead to left ventricular hypertrophy and increase the risk for the development of cardiomyopathies and clinical heart failure. On the other hand, there is also data showing that among patients with established cardiovascular disease, patients who are overweight or with obesity tend to have a lower risk for mortality and negative cardiovascular outcomes when compared with normal weight patients. This well-documented phenomenon is referred to in literature as the “obesity paradox”. The risks faced by individuals with frailty, limitations of BMI in defining obesity (BMI does not account for the percentage of lean mass), and the potential for lead time bias in diagnosing cardiovascular disease in patients with obesity have been proposed as possible explanations for the obesity paradox. ,, In a substudy of the prospective STABILITY trial consisting of 15,828 patients with stable CAD, a U-shaped curve was observed for the association of BMI and the risk for cardiovascular death, total death, and hospitalization for heart failure. The highest risk was seen in the low and high ends of BMI (BMI of <20 kg/m 2 and BMI of ≥35 kg/m 2), and the lowest risk was seen in patients with a BMI of around 27 kg/m 2 .
Obesity can further exacerbate the atherosclerotic process in CAD through mechanisms of insulin resistance and inflammation. Cholesterol depositions into the vessel walls, increased oxidation of low-density lipoprotein, as well as endothelial dysfunction in the setting of inflammation and oxidative stress, can lead to increased disease burden in patients with obesity. Indeed, for every 1 kg/m 2 increase in BMI, a 5%−7% increase in the incidence of CAD across all BMI categories has been reported. In addition, the presence of metabolic syndrome and abdominal obesity has been shown to increase the coronary artery calcium (CAC) score by about 45%, which is an indicator for coronary atherosclerosis and a strong predictor of CAD severity.
The result of the above is that a great number of patients who undergo noninvasive cardiac imaging for the diagnosis of obstructive CAD suffer from obesity. In several patient cohorts, this proportion ranges from 42% to 61% with the rates being on the higher end of the spectrum among high pretest probability cohorts. Thus, optimizing the imaging of those patients is of paramount importance not only because they constitute a large portion of the total noninvasive cardiac imaging cohort, but also that their pretest probability for obstructive CAD is higher than the average.
Noninvasive Cardiac Imaging for Assessing Coronary Artery Disease in Patients With Obesity
Exercise treadmill test (ETT)
Exercise treadmill testing with a treadmill requires patients to be able to walk fast enough on a treadmill and be able to reach the required heart rate for the ETT to be diagnostic. Patients with obesity may not be able to reach the maximal heart rate and the necessary threshold for inducible ischemia, not only because of general deconditioning but also because of common comorbidities such as osteoarthritis or other orthopedic problems, which are known to be more common in this population. Besides, obesity is known to be associated with several other comorbidities such as diabetes, which increases the pretest probability of CAD and thus makes ETT far from ideal for this- often high-risk patient population.
Exercise stress echocardiography (SE)
SE (either with exercise or with dobutamine) is a widely available tool for diagnosis and risk stratification of patients suspected of CAD. SE holds the advantage of not exposing the patient to ionizing radiation, while providing functional information about the left ventricular response to stress induced by exercise or with dobutamine-induced tachycardia. Image quality can be challenging in patients with obesity because of the often difficult-to-obtain apical and parasternal echocardiographic windows. These acoustic windows may be suboptimal for diagnosis, undermining the test accuracy. Nevertheless, the use of ultrasound contrast agents allows for better endocardial border definition and has been shown to significantly improve the interpretability, diagnostic accuracy, and reader confidence of stress echocardiography. In the Stress Ultrasonography in Morbid Obesity (SUMO) study, 209 patients (BMI > 35 kg/m²) suspected of CAD underwent SE, with a contrast agent used in 96% of the cases. Excellent test feasibility was reported with a high success rate in obtaining diagnostic-quality images. SE demonstrated a high positive predictive value (PPV) of 88% and provided appropriate risk stratification, highlighting the usefulness of this noninvasive test in the obese population. Dobutamine SE is preferred over exercise SE in patients with poor exercise capacity, a condition commonly seen in the obese population. In the SUMO study, 60% of the total patients underwent dobutamine SE. These patients were significantly older and with more comorbidities compared with patients who underwent exercise SE, likely reflecting their low capacity for exercise.
Single-photon emission computed tomography (SPECT)
SPECT imaging assists in the diagnosis of CAD by providing an assessment of the myocardial perfusion. In patients with obesity, soft tissue attenuation occurs when gamma rays emitted by the radioactive tracer are absorbed by surrounding tissue before reaching the detectors, leading to artifacts that interfere with result interpretation. Soft tissue attenuation artifacts can resemble perfusion defects and are erroneously attributed to CAD, thus increasing false positive results and decreasing SPECT specificity. , Previous studies have reported decreased accuracy for diagnosing CAD in patients with obesity. , Hansen et al. demonstrated significantly decreased area under the receiver operating characteristic curve (AUC) in patients with BMI > 30 kg/m 2 compared to patients without obesity (AUC: 0.86 ± 0.03 vs 0.92 ± 0.02, p < 0.05).
Attenuation correction using iterative reconstruction algorithms, now standard in contemporary SPECT systems, have been shown to significantly improve specificity in patients with BMI > 30 kg/m 2. , Adding the anatomical evaluation with the CT component (SPECT-CT) further improves the attenuation correction. Of note, in morbidly obese patients (BMI > 40 kg/m 2), SPECT has shown lower accuracy and specificity for detecting obstructive CAD compared to PET/CT, rendering the former more appropriate for myocardial perfusion imaging in these patients. , Lastly, safety concerns arise regarding radiation exposure, since patients with obesity frequently require increased doses to obtain acceptable image quality.
Positron emission tomography (PET)
PET myocardial perfusion imaging (MPI) offers enhanced spatial resolution, attenuation correction and iterative reconstruction techniques, making it diagnostically accurate, even in patients with obesity. Cardiac rubidium 82 (Rb-82) PET MPI was shown to have excellent prognostic value in 6,037 patients with obesity in a study by Chow et al. Bateman et al. demonstrated better image quality, interpretive certainty, and diagnostic accuracy of PET vs SPECT. Likewise, Harnett et al. reported increased diagnostic accuracy of PET with more definitive scan interpretation and less artifact compared to SPECT. Of note, SPECT imaging was without attenuation correction in these studies.
Stress cardiac magnetic resonance imaging (MRI)
MRI can provide optimal images and spatial resolution for evaluation of cardiac structure and function without carrying the risks of ionizing radiation or contrast medium. Vasodilator stress perfusion cardiac MRI is feasible in patients with obesity, with previously reported diagnostic imaging quality close to 94%. Shah et al. investigated 285 patients with obesity and identified a strong association between inducible ischemia and major adverse cardiac events (MACE). In contrast, patients without ischemia or scarring demonstrated a very low incidence of MACE during long-term follow-up. Results from the multicenter, observational Stress CMR Perfusion Imaging in the United States (SPINS) study, which included 2349 patients (half of them with obesity), demonstrated no negative impact on the diagnostic quality and the effectiveness of risk stratification of patients by elevated BMI.
Cardiac computed tomography angiography (CCTA)
Obesity is well known for creating potential obstacles in coronary CT image acquisition. First, there are scanner physical limitations. Patients must fit inside the gantry aperture and the usable portion of it (accounting for the entry of the table into the gantry). A typical gantry diameter is 70 cm, with the usable anteroposterior (vertical) diameter being 52 cm. Larger, bariatric CT scanners have been developed with gantry diameters of up to 85cm. In addition, the periphery of obese patients may not lie inside the reconstruction or display Field of View (FOV), excluding relevant anatomy. However, this probably does not affect the imaging of the heart, given its position in the middle part of the chest. Finally, although an unlikely event, the CT scanner table may break if the patient’s weight exceeds the weight limit.
Most importantly, though, from a cardiac standpoint, obesity can lead to artifacts in the CT image. As the body thickness increases, photon attenuation (defined as the decrease in photon intensity as they pass through a material) increases exponentially. Insufficient number of X-ray photons reach the detector, a phenomenon known as photon starvation artifact, with consequent reduced contrast-to-noise ratio (CNR) and signal-noise-ratio (SNR), and inferior image quality. , In addition, due to greater blood distribution volume, obese patients will have reduced peak vascular contrast enhancement.
For cardiovascular imaging, increasing the tube potential or the tube current enhances the CT image quality in patients with obesity, but simultaneously raises the radiation exposure. The Society of Cardiovascular Computed Tomography (SCCT) suggests a tube potential of 120 kV for patients over 90 kg and with a BMI above 30 kg/m 2, and a higher tube potential for severely obese patients compared to 100 kV required for patients with a BMI < 30 kg/m 2. Indeed, several studies have demonstrated a significant increase in radiation effective dose in obese patients, most commonly exceeding 15 mSv in patients with BMI > 40 kg/m 2. ,,
Methods like iterative reconstruction (IR) have been studied in both lean and obese patients as potential ways of decreasing radiation burden while maintaining sufficient image quality. , Results from the randomized multicenter REALISE trial demonstrated that implementation of IR techniques in lower tube voltages CCTAs reduced the effective radiation dose and the iodine load while providing comparable image quality. Importantly, there was no difference in the SNR and CNR in the subgroup analysis of patients with BMI ≥ 25 kg/m 2, and acceptable image noise was reported in the 5 included patients with BMI > 30 kg/m 2.
Impact on image quality
The influence of increased BMI on CCTA image quality has been investigated in many observational studies. An overview of relevant studies can be shown in Table 1 .
Table 1
Studies reporting on CCTA image quality in relation to BMI
| Study | Year | Objective | Sample size | Mean BMI (kg/m²) | BMI subgroups (kg/m²) | Key findings on image quality |
|---|---|---|---|---|---|---|
| Lowenstern et al. | 2023 | Investigate the association between BMI, CAD, and clinical outcomes | 5014 | NR | <25, 25-29.9, ≥30 |
|
| Mangold et al. | 2016 | Investigate diagnostic accuracy of 3rd-generation DSCT in obese and non-obese patients | 76 | 32.1 | <30, ≥30 |
|
| Latif et al. | 2016 | Evaluate image quality dependence on heart rate and BMI using new single-beat volumetric CT scanner | 439 | 29.95 | <30, ≥30 |
|
| Zhang et al. | 2015 | Investigate image quality, radiation dose, and diagnostic performance of high-pitch CCTA at 70 kVp | 43 | 23.3 | >25, ≤25 kg/m² |
|
| Zimmerman et al. | 2014 | Evaluate the diagnostic quality of dual-source coronary CT exams without heart rate control | 160 | 31.4 | <25, 25-30, ≥30 |
|
| Hausleiter et al. | 2012 | Compare image quality and radiation dose of prospectively ECG-triggered axial scanning versus helical scanning | 400 | 25.9 | <30, ≥30 |
|
| Ho et al. | 2010 | Evaluate the effect of increasing BMI on image quality and diagnostic accuracy of 100-kV CTA | 914 | 27.6 | <25, 25-30, 30-35 |
|
| Brodoefel et al. | 2008 | Assess impact of BMI on image quality and diagnostic accuracy using dual-source CT | 125 | 28.4 | <25, 25-30, ≥30 |
|
| Burgstahler et al. | 2006 | Assess the impact of obesity on image quality and diagnostic accuracy of 16-slice MDCT | 125 | 28.4 | <25, 25-30, >30 |
|
| Yoshimura et al. | 2006 | Evaluate the correlation between image noise and body weight/BMI in CCTA | 36 | 28.1 | NR |
|
BMI, body mass index; CCTA, coronary computed tomography angiography; PET, positron emission tomography; SPECT, single photon emission computed tomography.
Good correlation ( r = 0.74, p = 0.015) between image noise and body weight was reported in a study of 36 patients with mean BMI = 28.1 kg/m 2, while another study with 439 patients demonstrated significantly decreased contrast-to-noise ratio (CNR) in patients with BMI > 30 kg/m 2 compared to BMI < 30 kg/m 2 (6.4 ± 2.4 vs 7 ± 2.8, p = 0.002). BMI > 30 kg/m 2 was reported to be a significant influence on dual-source computed tomography (DSCT) image quality, with significant degradation in the group of patients with BMI > 30 kg/m 2 along with heart rate variability and calcium score. Another similar study examined 117 patients undergoing Multislice spiral computed tomography (MSCT) and found that patients with BMI ≥ 25 kg/m 2 had significantly worse image quality than patients with BMI < 25 kg/m 2.
It is interesting that while in numerous studies normal weight patients have a significantly greater number of segments with very good image quality, the segments with sufficient quality for diagnosis (diagnostic image quality) are comparable between the BMI subgroups. , Accordingly, although higher BMI tends to reduce image quality, diagnostic quality images can still be obtained in the vast majority of patients, regardless of their size, ,,, even when using low-voltage protocols (100-kV).
Impact on diagnostic performance
Although obesity produces image noise and worsens image quality, it does not appear to undermine CCTA diagnostic performance. When investigating predictors of inaccurate CCTA results, studies showed that BMI was not a predictive variable for CCTA-ICA disagreement. , Additionally, most of the studies that investigated measures of diagnostic performance between obese and nonobese patients revealed no significant difference between the 2 groups ( Table 2 ).
Table 2
Studies reporting on diagnostic performance of CCTA in relation to BMI
| Study | Year | Objective | Sample Size | Mean BMI (kg/m²) | BMI subgroups (kg/m²) | Key findings on diagnostic performance |
|---|---|---|---|---|---|---|
| Mangold et al. | 2016 | Evaluate diagnostic accuracy in non-obese and obese populations using 3rd-generation dual-source CT | 76 | 32.1 ± 6.7 | <30, ≥30 |
|
| Ho et al. | 2010 | Assess the impact of BMI on image quality and positive predictive value of 100-kV Coronary CT Angiography | 914 | 27.6 ± 4.5 | <25, 25-30, ≥30 |
|
| Brodoefel et al. | 2008 | Assess the impact of BMI on image quality and diagnostic accuracy using dual-source CT | 125 | 28.4 ± 4.1 | <25, 25-30, ≥30 |
|
| Alkadhi et al. | 2008 | Investigate the diagnostic accuracy of dual-source CT coronary angiography in relation to BMI, calcium load, and heart rate | 150 | 26.5 ± 4.2 | ≤26, >26 |
|
| Budoff et al. | 2008 | Evaluate the diagnostic accuracy of 64-slice CT | 230 | 31.4 ± 6.2 | <30, ≥30 |
|
| Raff et al. | 2005 | Assess diagnostic accuracy of 64-slice CT | 70 | NR | <25, 25-30, ≥30 |
|
| Burgstahler et al. | 2005 | Evaluate the impact of BMI on image quality and diagnostic accuracy using 16-slice CT | 117 | 28.4 ± 3.7 | <25, 25-30, ≥30 |
|
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
