Single Photon Emission Computed Tomography (SPECT) Myocardial Perfusion Imaging without Attenuation Correction: An Experience in a Community Hospital

Introduction

Radionuclide myocardial perfusion imaging (rMPI) with single-photon emission computed tomography (SPECT) for the diagnosis and risk stratification of coronary artery disease (CHD) is widely used in United States. However, its full clinical potential has not been realized due to numerous factors producing attenuation artifacts. These artifacts degrade image quality, increase the risk of misinterpretation of the results and increase the false positive rates of the nuclear stress test. ¹

Soft tissue attenuation due to breast, lungs, and diaphragm are the most common causes of rMPI artifact. Attenuation correction is one such method which decreases the artifacts, thus decreasing the false positive rates and increasing the specificity of radionuclide myocardial perfusion imaging. The SPECT that we do in our hospital is without attenuation correction software. The aim of our study was to study the false positive rate caused by attenuation artifact in our stress lab.

Methods

This study was done at Saint Barnabas Hospital which is one of the busiest inner city hospitals located in the Bronx, New York. A total of 119 patients (From January 1^st to December 31^st 2009) underwent SPECT rMPI followed by coronary angiography, they were selected based on this criteria of inclusion (having this two test performed). A retrospective chart analysis was performed on all these patients. Clinical characteristics including age, sex, BMI and reason for nuclear stress tests were noted. False positive rate, true positive rate, false negative rate, true negative rate, specificity and sensitivity of the radionuclide myocardial perfusion imaging were calculated. The study was compared with several meta-analyses of SPECT imaging done in the past.

Results

Out of 119 patients, twenty four (20%) patients had indeterminate results on sub maximal study. The remaining 95 patients had either positive or negative nuclear stress test. All these 95 patients were followed up by cardiac catheterization. Nineteen (20%) had false positive stress test, thirty five (37%) had true positive test, 21 (22%) had false negative test results and 20 (21%) patients had true negative results. Specificity of our stress test was 51% and sensitivity was 62%. Out of total 19 patients who had false positives, 14 (74%) were male and only 5 (26%) were female. Symptoms wise 16 patients had presented with chest pain and 3 had shortness of breath. The mean BMI of these patients was 31, fifteen (79%) out of the total 19 patients were obese (BMI > 30 kg/m²). Radionuclide myocardial perfusion imaging revealed artifact in the right coronary artery territory in 5 (35%) out of 14 males while the rest of the 9 males and all the 5 (100%) females had artifact in the left anterior descending or left circumflex artery. In the true positive group 2 had mild CAD (defined as coronary artery occlusion by 30% or less), 7 had moderate CAD (defined as coronary artery occlusion of 30% to 60%) and 26 patients had severe CAD (defined as coronary artery occlusion of more than 60%).

Of the 24 patients who had indeterminate results, eight had coronary artery disease and 16 patients were normal. Diabetes and Hypertension were the most common risk factors in all the groups.

Table 1: Test results

Table 2: Body Mass Index (BMI) of the patients with attenuation artifact.

Discussion

The heart is surrounded by tissues of varying densities (e.g. bone, lungs, and breast). Radionuclide imaging of the thorax results in nonuniform myocardial photon activity due to these intervening structures. This leads to production of artifacts called “attenuation artifact” on the myocardial images which might be read as an area of ischemia or infarction. This again might subject patients to unnecessary invasive procedures like cardiac catheterization and its inherent risks. Attenuation correction refers to methods that adjust the intensity of the myocardial perfusion image to reflect the estimated magnitude of soft tissue attenuation on different regions of the heart. This increases the normalcy rate and the specificity of the radionuclide myocardial perfusion imaging. Studies in the past have proven the poor specificity of SPECT without attenuation correction compared to invasive testing. ^{2 3 4}

Our study population had a high false positive rate and low specificity of nuclear stress test because of attenuation artifact. In a study done by Ficaro et al, ⁵ 60 patients with known coronary artery disease (CAD) and 59 controls were subjected to the attenuation correction after normal rMPI, the normalcy rate increased from 88 to 98 % while specificity of the test increased significantly from 46 to 82%. In yet another study done by Hendel⁶ et al normalcy rate was increased from 86% to 96% by the addition of attenuation correction. These studies, including ours clearly indicates the superiority of rMPI with attenuation correction for the evaluation of the patients for CAD.

Although there are multiple sources of attenuation artifact, diaphragmatic attenuation, breast attenuation, liver attenuation and attenuation due to obesity ⁷ are the most common causes of soft tissue attenuation. Not surprisingly, 79% of our patients with attenuation artifact were obese. Thompson ⁷ et al evaluated 119 patients who underwent rMPI with SPECT followed by coronary angiography within 60 days. The specificity for clinically significant CAD (stenosis of >70%) was significantly improved with attenuation correction (76% versus 41 % in obese patients). Our findings mirrored this study. Hansen ⁸ et al too concluded that high BMI was associated with lower accuracy for diagnosing CAD.

Diaphragmatic attenuation is common in men while breast attenuation is the most common reason of attenuation artifact in females. Because the left hemi diaphragm causes attenuation in the inferior myocardial wall, perfusion defects are seen in the Right Coronary Artery (RCA) territory, this is estimated to be present in about 25% of the patients ⁹ . In our study it was slightly higher (35%). Soft tissue attenuation from breast causes perfusion artifact in the left anterior descending (LAD) or left circumflex (LCx) artery territory in up to 40% of the females¹. ^{10 11} In our study we found all the females (100%) to be having artifact in LAD and LCx region.

Conclusion

Attenuation correction improves the diagnostic accuracy of radionuclide myocardial perfusion imaging, particularly with regard to specificity. It also decreases the false positive test results thus decreasing unwarranted cardiac catheterization and related costs and patient morbidity. The American Society of Nuclear Cardiology and the Society of Nuclear Medicine encourage the incorporation of attenuation correction into routine clinical practice.