Prognostic Implications of Combined Prone and Supine Acquisitions in Patients with Equivocal or Abnormal Supine Myocardial Perfusion SPECT

Study Population

We identified 4,466 consecutive patients who underwent dual-isotope MPS with either exercise or adenosine stress between January 1, 1994, and December 31, 1995. During this time, our standard protocol was to perform a prone MPS acquisition only in patients in whom an inferior wall perfusion defect observed in the supine study was considered to be possibly artifactual and would have resulted in an overall interpretation as equivocal or abnormal.

Patients with known valvular heart disease or cardiomyopathy (n = 64), as well as those who had myocardial revascularization within 60 d of the index MPS (8) (n = 395), were excluded. Of the initial population, 173 patients (4.5%) were lost to follow-up. The study therefore consisted of 3,834 patients, of whom 3,179 (83.0%) underwent supine imaging only and 655 (17.0%) underwent both supine and prone imaging.

Rest ²⁰¹Tl Imaging

All patients underwent rest ²⁰¹Tl/stress ^99mTc-sestamibi MPS as previously described (9). For rest imaging, weight-adjusted ²⁰¹Tl (111–167 MBq [3.0–4.5 mCi]) was injected intravenously and acquisition was performed 10 min after injection.

Exercise Myocardial Perfusion Protocol

Patients performed a symptom-limited treadmill exercise test using a standard Bruce protocol. Leads aVF, V₁, and V₅ were continuously monitored and a 12-lead electrocardiogram (ECG) was recorded at each minute of exercise. At peak exercise, a weight-adjusted dose of ^99mTc-sestamibi (925–1,480 MBq [25–40 mCi]) was injected, and exercise was continued for 1 min after injection. MPS acquisition was initiated 30 min after isotope injection. Whenever possible, β-blockers and calcium channel blockers were discontinued 48 h before testing, and nitrates were discontinued at least 6 h before testing.

Adenosine Myocardial Perfusion Protocol

Patients were instructed not to consume caffeine-containing products for 24 h before the test. Adenosine was infused intravenously at a rate of 140 μg/kg/min for 6 min. At the end of the third minute of infusion, weight-adjusted ^99mTc-sestamibi was injected. Image acquisition was initiated approximately 60 min after isotope injection.

During both types of stress, blood pressure was measured and recorded at rest, at the end of each stress stage, and at peak stress. Maximal degree of ST segment change at 80 ms after the J point of the ECG was measured and assessed as horizontal, upsloping, or downsloping. ECG responses were considered positive when ≥1 mm of horizontal or downsloping or ≥1.5 mm of upsloping ST segment depression developed at 80 ms after the J point in patients with normal baseline ECG, negative when these changes were absent, and nondiagnostic when baseline changes in the ECG precluded interpretation during exercise.

MPS Acquisition Protocol

MPS acquisitions (Fig. 2) were performed with a large-field-of-view gamma camera and a high-resolution collimator, obtaining projections over a semicircular 180° arc. For ²⁰¹Tl imaging, 2 energy windows were used: a 30% window centered on 70 keV and a 20% window centered on the 167-keV peak. For ^99mTc-sestamibi, a 15% window centered on the 140-keV peak was used. Acquisition protocols for poststress supine imaging used 32 projections, with 25 s per projection for a dual-head camera, and 64 projections, with 20 s per projection for a single-head camera. For poststress prone acquisition, 15 s per projection were used, with 32 or 64 projections for dual- or single-head, respectively. For rest ²⁰¹Tl, 32 projections and 35 s per projection were used for the dual-head and 64 projections with 25 s per projection for single-head.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Injection and acquisition protocol for prone and supine MPS study. *15 min for exercise, 1 h for no walk adenosine (adapted and reprinted with permission of (24)).

Image Interpretation

Semiquantitative visual interpretation was performed with short-axis and vertical long-axis myocardial tomograms and a 20-segment model, as previously described (10). These segments were assigned on 6 evenly spaced regions in the apical, midventricular, and basal slices of the short-axis views and 2 apical segments on the midventricular long-axis slice. Each segment was scored by consensus of 2 experienced observers using a 5-point scoring system (0 = normal, 1 = equivocal, 2 = moderate reduction of isotope uptake, 3 = severe reduction of uptake, and 4 = absence of detectable tracer uptake in a segment). When a segment demonstrated a lower score on either supine or prone imaging than in the other position, the lower score was assigned to that segment. On the basis of the number and severity of segments with scores >2, the observers judged the study results as normal, probably normal, equivocal, probably abnormal, or definitely abnormal (10).

To further define the study results as normal or abnormal, the summed stress score (SSS) was considered, from addition of the scores of the 20 segments of the stress ^99mTc-sestamibi images (10). SSSs of <4 were considered normal, between 4 to 8 mildly abnormal, and >8 moderately to severely abnormal. For purposes of this study, the SSS had to be <4 and the final interpretation had to be normal or probably normal for the study to be considered normal. Similarly, patients with equivocal, probably abnormal, and definitely abnormal final MPS interpretations were grouped as abnormal.

Patient Follow-Up

Patient follow-up was performed by scripted telephone interview by individuals blinded to the patients’ test results. Hard events were defined as either cardiac death, as noted and confirmed by review of death certificate and either hospital chart or physician’s records, or nonfatal myocardial infarction, as evidenced by the appropriate combination of symptoms, ECG, and enzyme changes. Total events were defined as hard events or late revascularization (>60 d after the index MPS study) (8). Patients were monitored for 24.2 ± 6.0 mo (all for at least 1 y).

Likelihood of CAD

The prescan likelihood of CAD was calculated with CADENZA (Advanced Heuristics Inc.), using Bayesian analysis of prescan patient data (11). For the patients who underwent adenosine stress testing, the history and resting ECG information were considered, whereas for those who underwent exercise stress testing, the prescan likelihood of CAD included history and exercise test information.

Statistical Analysis

Continuous variables were expressed as mean value ± SD and compared using a Student t test with correction for multiple comparisons when appropriate. Comparisons of categoric variables were computed by a χ² statistic. A P value of 0.05 was used to define statistical significance in univariate analysis. A Cox proportional hazards model was used to identify variables independently associated with cardiac events. Multivariable models were developed using hard or total events as the endpoints of interest. In this study, the goal of modeling was to determine all potential confounders of the relationship between MPS results and adverse outcomes. This approach included setting a threshold of P < 0.20 for entry into the model. Hence, the likelihood of finding a significant relationship between prone imaging and increased risk of adverse events would be maximized.

Patient Population

The 3,179 patients who underwent supine-only imaging and 655 patients who underwent supine and prone imaging are characterized in Table 1. Patients who had prone and supine acquisitions were more frequently male than patients who had supine-only MPS. By all other comparators, the groups were not significantly different.

Table 2 demonstrates patient characteristics by MPS acquisition types and normal or abnormal MPS results. Within each acquisition type, differences were as follows: for patients undergoing prone and supine imaging, those with abnormal studies were more frequently male, had a higher prescan likelihood of CAD, and had more frequent abnormal rest ECG than patients with normal scans. For patients studied in the supine position only, those with abnormal studies were significantly older and more frequently male, with a history of myocardial revascularization, were on digitalis, and had a higher prescan likelihood of CAD. Comparing patients with normal scans by the 2 acquisition types, the only significant difference was the higher frequency of males in the prone and supine group. Comparing patients with abnormal scans, not only were the patients undergoing prone and supine more often male than those undergoing supine MPS alone, but also they less frequently had prior revascularization, were on digitalis or had rest ECG abnormality, as well as had a lower SSS.

Outcome Events

During follow-up, there were 132 hard events (52 myocardial infarctions and 83 cardiac deaths) and 375 total events. Overall, the annualized hard event rate was 1.8% and the total event rate was 5.2%. Event rates were higher for patients undergoing adenosine stress: 3.2% of hard events per year and 7.4% of total events per year, versus 1.1% and 4.2% of hard and total events per year, for exercise stress (P < 0.001). Patients imaged only in the supine position had, with adenosine or exercise stress, hard event rates of 3.6% and 1.0% and total event rates of 7.8% and 4.0%, respectively (P < 0.001 for comparison between adenosine and exercise stress). Patients who had prone and supine imaging had, with adenosine or exercise stress, hard event rates of 1.6% and 1.1% and total event rates of 5.5% and 5.3%, respectively (P > 0.05).

Table 3 demonstrates annualized event rates according to scan results and type of stress for patients imaged in prone and supine or supine-only positions. Of note, 368 of 655 patients with equivocal or abnormal supine MPS who underwent additional prone imaging had normal final interpretations. In both acquisition types, hard and total event rates increased with worsening scan results, although this did not reach statistical significance in patients who underwent adenosine stress and prone and supine acquisition. Importantly, the annualized hard event rate was as low in patients with normal prone and supine scans as in those with normal supine studies. Furthermore, for each category of SSS, hard and total event rates, with either exercise or adenosine stress, were not significantly different between patients in the prone and supine and supine-only groups.

Univariate Predictors of Cardiac Events

Descriptive characteristics and nuclear variables of patients with events versus those without events during follow-up are presented in Tables 4 and 5 for hard events and total events, respectively. Patients with hard events demonstrated greater abnormality in all variables assessed with the exception of hypertension (borderline higher in the event group), smoking, family history of early CAD, and angina. Of importance, there was also no difference in the frequency of prone and supine or supine-only acquisitions in the patients with and without hard events. Regarding total events (Table 5), those with events demonstrated a greater abnormality in most variables. The variables that were not significantly different between these groups were hypercholesterolemia, smoking, family history of early CAD, angina, shortness of breath, and abnormal rest ECG. As with hard events, there was also no difference in the frequency of prone and supine or supine-only acquisitions in the patients with and without total events.

Multivariable Predictors of Events

In the overall study population, Cox proportional hazards analysis revealed that age (χ² = 9.2, P = 0.002), a history of myocardial infarction (χ² = 4.5, P = 0.03), digitalis use (χ² = 12.1, P < 0.0001), adenosine stress (χ² = 10.0, P = 0.002), and the SSS (χ² = 38.2, P < 0.0001) constituted the model most predictive of hard events. For total events, age (χ² = 13.4, P = 0.001), digitalis use (χ² = 15.3, P = 0.0001), adenosine stress (χ² = 9.2, P = 0.003), and the SSS (χ² = 84.2, P < 0.0001) constituted the most predictive model. Figures 3 and 4 illustrate that, after risk adjustment using the multivariable model, hard and total event rates were not significantly different between patients who underwent supine-only or prone and supine acquisitions.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Risk-adjusted hard event rates in patients undergoing supine-only or prone and supine acquisition, according to MPS interpretation. P < 0.0001 across scan categories; P > 0.05 for comparisons between supine-only or prone and supine. Mod-sev = moderately to severely abnormal.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Risk-adjusted total event rates in patients undergoing supine-only or prone and supine acquisition, according to MPS interpretation. P < 0.0001 across scan categories; P > 0.05 for comparisons between supine-only or prone and supine. Mod-sev = moderately to severely abnormal.