Methods

We studied stable patients referred for SPECT imaging because of suspected CAD prospectively enrolled in an SPECT/computed tomography (CT) registry. It is approved by the Isala Medical Ethics committee and includes a waiver of consent. Consecutive patients, without a known history of CAD, referred for noninvasive CAD detection with SPECT/CT in the Isala Hospital from January 2009 to June 2013 were included in the registry (n = 5,026). All patients underwent 99mTc-tetrofosmin stress-first SPECT imaging combined with simultaneous CAC scoring. In case of normal stress SPECT MPI, no additional rest examination was performed. Patients who demonstrated normal stress SPECT MPI or normal SPECT MPI findings after both stress and rest imaging were included in the present analysis. Patients in whom no CAC score could be acquired (mainly because of high or irregular heart rate, n = 123) and patients with abnormal SPECT findings (n = 1198, 24%) were excluded. No other exclusion criteria were applied. Information regarding the presence of risk factors was prospectively collected by written questionnaires. These data were verified and complemented with demographic and clinical information collected from medical records. Height and weight were measured, and body mass index was calculated.

Stress testing was routinely performed with pharmacological stress using adenosine (140 μg/min/kg for 6 minutes), unless there was a contraindication for pharmacological stress. Because of logistical reasons, this is common practice in our high-volume center. Patients were instructed to refrain from caffeine-containing beverages for at least 24 hours before the test. In case of a contraindication for adenosine, patients underwent dobutamine (starting dose of 10 μg/kg/min, increased at 3-minute intervals to a maximum of 50 μg/kg/min), regadenoson (fixed dose of 400-μg bolus injection over 15 seconds), or bicycle testing. A weight-adjusted dose of 99mTc-tetrofosmin (standard 370 MBq, 500 MBq for patients >100 kg) was administered after 3 minutes (adenosine), after 35 seconds (regadenoson), or when the target heart rate of >85% of predicted maximal was reached (dobutamine, bicycle test). Patients scheduled for rest imaging received a dose of 99mTc-Tetrofosmin (standard 740 MBq, but 1,000 MBq for patients >100 kg).

Both stress and rest SPECT images were acquired 45 to 60 minutes after tracer injection. Time delay between the stress and rest studies was >3 hours. All patients were imaged in the supine position with arms placed above the head. From January 2009 to April 2010, patients (n = 977) were scanned on a conventional dual-detector gamma camera (Ventri-LightSpeed VCT XT; GE Healthcare, Milwaukee, Wisconsin), using a low-energy, high-resolution collimator, a 20% symmetrical window at 140 keV, a 64 × 64 matrix, and an elliptical orbit with step-and-shoot acquisition at 6° intervals over a 180° arc (45° anterior oblique to 45° left posterior oblique) with 15 steps (15 views). Acquisition time was 12 minutes for the stress images and 15 minutes for the rest images as previously described. From May 2010 to June 2013, patients (n = 4,057) were scanned with a cadmium zinc telluride (CZT)–based SPECT/CT camera (Discovery NM/CT 570c, GE Healthcare) with 19 stationary CZT detectors simultaneously imaging 19 cardiac views. Each detector comprised 32 × 32 pixelated (2.46 × 2.46 mm) CZT elements. Acquisition time was 5 minutes for the stress images and 4 minutes for the rest images. This was derived from the recommendations of the manufacturer, published experience, our own qualitative assessment in heart phantom studies, and our initial experience in patients. All SPECT studies were followed by an unenhanced low-dose CT scan during a breath-hold to provide the attenuation map for attenuation correction as previously described.

SPECT MPI was unblindedly and semiquantitatively interpreted using a 17-segment model. Segments were scored by consensus of 2 experienced nuclear cardiology observers using a 5-point scoring system (0 = normal, 1 = equivocal, 2 = moderate, 3 = severe reduction of radioisotope uptake, 4 = absence of detectable tracer uptake). The combination of attenuation-corrected and non–attenuation-corrected images were reviewed. A stress study was interpreted as normal if the summed stress scores was ≤3. Additional rest SPECT was acquired if the stress images did not fulfill these criteria. The perfusion images were reviewed again after both stress and rest SPECT. An ischemic defect was defined as a summed difference score ≥2. Reversible defects not fulfilling these criteria were assessed as equivocal for ischemia. Perfusion defects which demonstrated no reversibility were defined as fixed defects. If, after reviewing both stress and rest SPECT, no reversible of fixed defects were observed, SPECT was considered normal.

A nonenhanced CT study was performed for the CAC score using the 64-section CT scanner of the integrated SPECT/CT scanner (LightSpeed VCT XT). All patients with heart rates >70 beats/min received oral β-blocker therapy, with 50 or 100 mg of metoprolol tartrate (AstraZeneca, Zoetermeer, the Netherlands) before the CAC scan. Images were obtained with electrocardiographic gating at 75% of the R-R interval and with the following scanning parameters: 40 or 48 sections and 2.5-mm section thickness; gantry rotation time, 330 msec; tube voltage, 120 kV; and a tube current ranging from 125 to 250 mA, depending on patient size. Postprocessing was conducted at a dedicated workstation using Smartscore software (GE Healthcare). The CAC score was calculated using the standard Agatston criteria and was categorized as CAC score 0, 1 to 99, 100 to 399, 400 to 999, or ≥1,000.

Follow-up data were based on clinical visits or standardized telephone interviews. Major adverse cardiac events (MACE) were defined as revascularization (angioplasty or coronary artery bypass surgery), nonfatal myocardial infarction, or all-cause mortality. Nonfatal myocardial infarction was defined based on the criteria of typical chest pain, elevated cardiac enzyme levels, and typical changes on the electrocardiography as defined by Thygesen et al.

Continuous variables are expressed as mean ± SD or median (interquartile range), and categorical variables are expressed as frequency (percentage). Differences between groups were assessed by the unpaired Student t test, Mann–Whitney U test, and chi-square test, where appropriate. Patient clinical characteristics were compared across different CAC score categories using the chi-square test for trend. We calculated annualized event rates on the basis of events per patient per year and compared differences in annualized events between men and women in the different CAC score categories by Poisson regression for rate data. We used Cox proportional hazard regression to analyze the association between CAC scores, patient characteristics, and clinical outcomes. The Cox model was adjusted for age, gender, and traditional risk factors. Interaction terms were evaluated between CAC score, gender, and the outcome measure with logistic regression. Two-sided p values <0.05 were considered statistically significant in all tests. All statistical analysis was performed with a commercially available software package (SPSS version 20.0 for Windows); the Poisson regression for rate data was performed with MedCalc (version 16.2 for Windows).