Comparison Between 201Tl and 99mTc Sestamibi Uptake During Adenosine-Induced Vasodilation as a Function of Coronary Stenosis Severity
Background Myocardial uptake of either 201Tl or 99mTc-sestamibi (sestamibi) is known to plateau at high coronary flow rates. However, few direct comparisons have been made between these tracers to determine what effect differences in the uptake plateau for the two tracers may have on the detection of coronary stenoses of various severities.
Methods and Results Twenty-two dogs were instrumented with flow transducers on the left anterior descending (LAD) and circumflex (LCx) arteries. In 6 nonstenotic dogs, adenosine was infused directly into the LAD, whereas 16 dogs with either critical (n=7) or mild (n=9) LAD stenoses received an intravenous infusion. At peak flow, 201Tl (0.5 mCi), sestamibi (5 to 8 mCi), and radiolabeled microspheres were injected simultaneously. Five minutes later, dogs were killed, and ex vivo imaging of heart slices and gamma-well counting of multiple myocardial samples was performed. Neither 201Tl nor sestamibi uptake increased in direct proportion to flow. In the 6 nonstenotic dogs, a fivefold increase in LAD flow increased 201Tl and sestamibi uptake by only 202±6% and 138±4%, respectively (P<.0001). In the dogs with critical stenosis, the ratios of stenotic to normal activity by well counting for 201Tl (0.37±0.05) and sestamibi (0.53±0.06) underestimated the microsphere-determined flow disparity (0.17±0.03) (P<.005), but the degree of underestimation was greater for sestamibi (P=.001). Similarly, in the dogs with mild stenosis, the stenotic-to-normal ratio for 201Tl (0.62±0.04) approximated the flow ratio (0.43±0.04) better than sestamibi (0.79±0.03) (P<.0001). Sestamibi defects, however, were visually identifiable on the images of the myocardial slices. By image quantification, sestamibi defect magnitude (LAD-to-LCx count ratio) in the critical stenosis group (0.62±0.05) was significantly less than in the mild stenosis group (0.80±0.02) (P<.01).
Conclusions Thus, with adenosine-induced hyperemic flow, both 201Tl and sestamibi significantly underestimated the magnitude of the flow disparity between stenotic and normal perfusion beds. The degree of underestimation was greater for sestamibi. The clinical implication of these experimental findings for vasodilator perfusion imaging remains to be determined, since factors such as greater redistribution and scatter with 201Tl could offset its advantages.
The radiopharmaceutical 99mTc-2-methoxy-2-methylpropyl isonitrile (sestamibi) is a new myocardial perfusion imaging agent being used for the noninvasive detection and assessment of coronary artery disease.1 When sestamibi is administered under basal conditions, it becomes distributed in the myocardium in proportion to regional myocardial blood flow,2 but when sestamibi is administered at flow rates >2 to 2.5 mL · min−1 · g−1, there is a plateau in extraction.3 This property of decreased myocardial extraction at high coronary flow rates is characteristic of all diffusible flow tracers, including 201Tl.4 5 6 Since the extraction of sestamibi is less than that of 201Tl at resting flow rates,7 it is unknown whether the further decrease in extraction at high flow, as with pharmacological vasodilatation, will lead to an underestimation of the size and magnitude of perfusion defects compared with that obtained with 201Tl. Although a number of clinical studies conducted with relatively large numbers of patients showed comparable sensitivity and specificity between 201Tl and sestamibi for the detection of coronary artery disease with exercise stress,8 9 10 11 12 13 14 15 16 17 some observers have reported a smaller defect size with sestamibi than with 201Tl in the same groups of patients.18 19 Furthermore, sestamibi defect size underestimated the mass of the underperfused region relative to 201Tl in dogs undergoing partial coronary occlusion and adenosine single positron emission computed tomographic (SPECT) imaging.20
The goals of the present study were (1) to compare the myocardial uptake of 201Tl with sestamibi during adenosine-induced hyperemia by use of quantification of flow and tracer activities and (2) to compare defect magnitudes on sestamibi images in dogs with mild versus severe stenoses. The hypothesis tested in this study was that myocardial 201Tl uptake patterns should be less affected by the plateau in tracer extraction at high flows than sestamibi.
Twenty-two fasted adult mongrel dogs (mean weight, 23.3 kg) were anesthetized with sodium pentobarbital (30 mg/kg), intubated, and ventilated on a respirator (Harvard Apparatus) with positive end-expiratory pressure of 4 cm H2O. Arterial blood gases were monitored (model 158, Ciba-Corning), and pH, Po2, Pco2, and HCO3 levels were maintained at physiological levels. Lead II of the ECG was monitored continuously. The right femoral vein was cannulated with an 8F polyethylene catheter for the administration of fluids, medications, and 201Tl and sestamibi. Both femoral arteries were also isolated and cannulated with 8F polyethylene catheters to serve as sites for the collection of arterial blood samples and for microsphere reference blood withdrawal. A 7F catheter was placed in the right femoral artery for continuous monitoring of systemic arterial pressure.
A thoracotomy was performed at the level of the fifth intercostal space, and the heart was suspended in a pericardial cradle. A flare-tipped polyethylene catheter was inserted into the left atrial appendage for continuous left atrial pressure measurements and as a site for the injection of radiolabeled microspheres. The left anterior descending coronary artery (LAD) was then dissected free of the epicardium, and an ultrasonic flow probe (T201, Transonic Systems, Inc) and snare ligature were placed around the vessel. A similar flow probe was placed around the left circumflex artery (LCx). In the 6 dogs receiving intracoronary infusions of adenosine, the 27-gauge needle of a lymphangiography catheter (catalog No. 6656, Becton, Dickinson & Co) was carefully inserted through the LAD wall proximal to the flow probe, and the catheter was anchored in place with several sutures to the myocardial wall. The hemodynamic parameters of heart rate, systemic arterial and left atrial pressures, and LAD and LCx flows were continuously recorded on an eight-channel strip-chart recorder (model 7758D, Hewlett-Packard) throughout each protocol. All experiments were performed with the approval of the University of Virginia Animal Research Committee in compliance with the position of the American Heart Association on use of research animals.
Protocol 1: Intracoronary Adenosine (n=6)
An intracoronary infusion of adenosine was begun into the lymphangiography catheter at a rate of 0.25 mL/min, and the rate was adjusted to maintain mean ultrasonic LAD flow at two to three times baseline without raising LCx flow above 1.5 times its baseline value. Three dogs were then given a simultaneous injection of 8 mCi of 99mTc-sestamibi and 0.5 mCi of 201Tl. Three additional dogs were injected solely with 99mTc-sestamibi to determine whether flow-uptake values were comparable to those seen with the dual-isotope protocol. The tracer injections were followed within 1 to 2 minutes by a left atrial injection of radiolabeled microspheres. All dogs were killed 5 minutes after tracer administrations.
Protocols 2 and 3: Intravenous Adenosine, Either Critical or Mild LAD Stenoses
Protocols 2 and 3 are shown schematically in Fig 1⇓. During the baseline period, before the coronary stenosis was set, microspheres were administered to measure baseline flow. Microspheres were labeled with either 103Ru, 95Nb, or 46Sc. The selection order of microspheres was randomized to minimize bias. To measure the normal reactive hyperemic response, the LAD was then briefly occluded by tightening the snare occluder for 10 seconds and releasing it to produce a reactive hyperemia flow tracing on the strip-chart recorder. In protocol 2, the snare was then adjusted to produce a critical stenosis. A critical stenosis was defined as the point at which baseline flow was unchanged but the reactive hyperemic response was completely abolished. A second injection of microspheres was administered after the stenosis was set. Next, an intravenous infusion of adenosine was begun at a rate of 300 μg · kg−1 · min−1 and continued until LCx flow was maximal. This dose of adenosine was chosen empirically to produce high coronary flow without decreasing systemic arterial pressure below 85 mm Hg. When maximal LCx flow was achieved, 0.5 mCi of 201Tl, 5 mCi of sestamibi, and microspheres were injected simultaneously. Five minutes later, before the thallium could undergo appreciable redistribution, the dogs were killed.
In protocol 3, the snare was adjusted to produce a mild LAD stenosis. A mild stenosis was defined as no change in baseline flow but a 50% reduction in the reactive hyperemic test response. The remainder of protocol 3 was identical to that of protocol 2.
Ex Vivo Slice Imaging (Protocols 2 and 3 Only)
After the dogs were killed, the hearts were removed and sliced into four rings ≈1 cm thick from apex to base. The slices were trimmed of excess fat and adventitia and placed on a thin piece of cardboard covered with cellophane wrap. The slices were imaged directly on the collimator of a conventional planar gamma camera for maximal count time (1638 seconds). The slices were imaged with a 20% window centered on the 99mTc photopeak, and image quantification was performed on a standard nuclear medicine computer system (Sopha Medical Systems). Regions of interest were drawn on the defect area visible in the anteroapical region of the left ventricular wall and on the normal posterior wall of the sestamibi images. Quantification was performed only on the two center slices, since the basal slice was above the stenosis and hence was always normal, whereas the apical slice lacked a quantifiable normal region. The defect magnitude was calculated as the ratio of the average counts in the LAD region of interest divided by the average counts in the normal LCx region of interest. Because of spilldown of 99mTc into the 201Tl window, quantification of 201Tl defect magnitude could not be performed accurately.
Quantification of Myocardial 201Tl, Sestamibi, and Microsphere Flow
To measure 201Tl and sestamibi activities and microsphere-determined flow in the myocardial tissue samples, each of the four myocardial slices was divided into eight transmural sections, which were then further subdivided into epicardial, midwall, and endocardial segments, resulting in a total of 96 myocardial segments for each dog. The myocardial tissue samples were counted in a gamma-well scintillation counter (MINAXI 5550, Packard Instruments) for both 201Tl and 99mTc activities within 24 hours of collection. The samples were recounted for microsphere flows 2 weeks later when the 201Tl and 99mTc had decayed. The window settings on the gamma counter were 201Tl, 50 to 100 keV; 99mTc, 130 to 170 keV; 103Ru, 450 to 550 keV; 95Nb, 640 to 840 keV; and 46Sc, 842 to 1300 keV. The tissue counts were corrected for background, decay, and isotope spillover, and regional myocardial blood flow was calculated with specialized computer software (pcgerda, Packard Instruments). Transmural activity and flow values were calculated as the average of the corresponding epicardial, midwall, and endocardial samples. The microsphere technique used in our laboratory has been described.21
Data Analysis and Statistics
All statistical computations were made with systat software (SYSTAT Inc). The results were expressed as the mean±SEM. Differences between means within a group were assessed by a paired t test, with values of P<.05 considered significant. Comparisons between the two groups were made with one-way ANOVA and Tukey’s post hoc testing.
Protocol 1: Assessment of 201Tl and Sestamibi Uptake After Intracoronary Adenosine Infusion
The mean hemodynamic parameters of heart rate, systemic arterial pressure, and left atrial pressure measured at baseline and at the peak adenosine response when the tracers were administered are shown in Table 1⇓. The intracoronary infusion of adenosine in these normal dogs had no effect on heart rate, mean arterial pressure, or left atrial pressure. Mean LAD blood flow, as measured by the ultrasonic flow probes, was 31±6 mL/min at baseline and increased to 89±17 mL/min (P=.007 versus baseline) at the time of tracer administration. Mean LCx blood flow was 34±2 mL/min at baseline and did not increase significantly during adenosine infusion (44±6 mL/min).
Regional Flow Versus 201Tl and Sestamibi Activities
In this protocol, adenosine was injected directly into the normal LAD, after which 201Tl and sestamibi were administered intravenously. Myocardial segments from all dogs were grouped into flow ranges according to regional flow values at the time of tracer injection (Fig 2⇓). Neither 201Tl nor sestamibi activity increased in direct proportion to blood flow. Sestamibi activity was significantly less than 201Tl activity in all flow ranges, and the difference between the tracers became progressively larger as flow increased. A fivefold to sixfold increase in regional myocardial blood flow increased 201Tl uptake by 202±6% but increased sestamibi uptake by only 138±4% (P<.0001) (Fig 2⇓).
Protocol 2: Assessment of 201Tl and Sestamibi Uptake During Intravenous Adenosine Infusion in the Setting of a Critical LAD Stenosis
The hemodynamic variables measured throughout the protocol are summarized in Table 1⇑. In all dogs in this protocol, there was a biphasic mean arterial pressure response to the intravenous adenosine infusion. At the onset of the infusion, mean arterial pressure fell rapidly at first, but after a nonsignificant reflex rise in heart rate, mean pressure increased and then remained near 85 mm Hg. In this group, adenosine produced a slight increase in left atrial pressure. The mean LAD flow was 17±1 mL/min at baseline and was not significantly altered after the critical stenosis was set (15±1 mL/min). As seen in Fig 3⇓, there was also no change in the stenotic LAD flow after adenosine infusion (13±2 mL/min). Mean LCx flow was 28±6 mL/min at baseline and 29±7 mL/min after the LAD stenosis was set (P=NS). With adenosine, there was a fourfold increase in LCx flow, to 106±18 mL/min (P=.003).
Regional Flow Changes With Setting of the Critical LAD Stenosis and After Adenosine Infusion
In this protocol, adenosine was administered intravenously to dogs with a critical stenosis. Table 2⇓ summarizes the regional myocardial blood flow for this group of dogs during baseline, after the stenosis was set, and during adenosine-induced hyperemia, when 201Tl and sestamibi were administered. As shown, there was no change in resting epicardial, midwall, or endocardial flow after the LAD stenosis was set. In this group, flow in the epicardial region of the LAD zone did not change during adenosine-induced hyperemia (1.2±0.1 versus 1.1±0.1 mL · min−1 · g−1). However, in the midwall and endocardial regions of the LAD zone, flow fell significantly, from 1.1±0.2 and 1.0±0.2 mL · min−1 · g−1 to 0.6±0.1 and 0.3±0.1 mL · min−1 · g−1, respectively (P<.04), with a resultant decrease in the endocardial-to-epicardial flow ratio from 0.8±0.1 to 0.3±0.1 (P=.02), presumably because of coronary steal. Flow increased significantly in all layers of the normal LCx zone after adenosine administration. There was a significant decrease in the endocardial-to-epicardial flow ratio in the LCx zone during adenosine (from 1.1±0.0 to 0.6±0.1 mL · min−1 · g−1) (P<.05) resulting from a greater increase in epicardial flow relative to endocardial flow in the nonstenotic LCx zone.
Myocardial 201Tl and Sestamibi Activities After Tracer Administration During Adenosine Infusion
In Fig 4⇓, 201Tl and sestamibi activities in 32 transmural segments are plotted against the corresponding microsphere flow at the time of injection in a representative dog in this group with a critical LAD stenosis. As shown, the myocardial uptake of both 201Tl and sestamibi plateaued with increasing flow. Note that myocardial sestamibi activity was significantly less than that of 201Tl at flows >150% of normal.
Fig 5⇓ (left) displays the mean ratios of stenotic to normal zones for microsphere flow, 201Tl, and sestamibi determined from gamma-well counting of myocardial segments in the group of dogs that received adenosine in the setting of a critical LAD stenosis. These ratios are reflective of the relative decrease in microsphere and tracer concentrations in the stenotic coronary bed of the LAD. In the stenosis group, the ratios for both 201Tl (0.37±0.05) and sestamibi (0.53±0.06) significantly underestimated the actual ratio of stenotic to normal flow (0.17±0.03) (P<.005). The degree of this flow underestimation was greater for sestamibi versus 201Tl (P=.001).
Protocol 3: Assessment of 201Tl and Sestamibi Uptake During Intravenous Adenosine Infusion in the Setting of a Mild LAD Stenosis
As observed in the dogs with a critical stenosis, a decrease in arterial pressure and increase in heart rate were seen in these dogs with a mild LAD stenosis. The mean LAD and LCx flows as measured by the ultrasonic flowmeter in the mildly stenotic dogs in protocol 3 are summarized in Fig 3⇑. Mean LAD flow was unchanged after the mild stenosis was set (25±2 versus 24±2 mL/min). With adenosine, a significant increase in LAD flow to 47±5 mL/min (P<.001) occurred. Flow in the LCx bed was 32±3 mL/min at baseline (P=NS versus LAD baseline flow) and 32±3 mL/min after the LAD stenosis was set (P=NS). During the adenosine infusion, there was a nearly fourfold increase in LCx flow, to 115±10 mL/min (P<.001).
Regional Flow Changes With Setting of the Mild LAD Stenosis and After Adenosine Infusion
As shown in Table 2⇑, no significant change in epicardial, midwall, or endocardial flow was seen with setting of the mild LAD stenosis. In this group, adenosine increased epicardial flow from 1.5±0.2 to 3.6±0.3 mL · min−1 · g−1 (P<.001) in the stenotic zone. Endocardial flow increased from 1.4±0.1 to 1.7±0.1 mL · min−1 · g−1, although this change did not reach statistical significance (P=.086). In contrast, epicardial and endocardial flows increased to 4.6±0.4 and 3.7±0.3 mL · min−1 · g−1 in the normally perfused LCx zone after adenosine administration. Thus, hyperemic flow distal to the mild LAD stenosis consequent to adenosine-induced vasodilatation occurred, but it was less of a flow increase than seen in the LCx bed. In this group, no endocardial-to-epicardial coronary steal was apparent in the LAD perfusion zone.
201Tl and Sestamibi Activities in the LAD Zone After Tracer Administration During Adenosine Infusion
As shown in Fig 5⇑ (right), for this group of dogs, the stenotic-to-normal count ratios for both 201Tl (0.62±0.04) and sestamibi (0.79±0.03) significantly underestimated the stenotic-to-normal flow ratio (0.43±0.04; P<.001). As observed in the critical stenosis dogs, the underestimation was greater for sestamibi than for 201Tl (P<.001).
Quantification of Defects on Images of Myocardial Slices in Dogs in Protocols 2 and 3
In all 7 dogs with critical stenoses that made up protocol 2, prominent sestamibi defects were readily apparent in the anteroapical region on the gamma-camera images of the myocardial slices. Likewise, in all 9 dogs with mild LAD stenoses in protocol 3, mild anteroapical defects could be visually identified on sestamibi images. An example set of images from 1 dog in each group may be seen in Fig 6⇓. A comparison between the mean sestamibi defect count ratio (stenotic to normal) in dogs with severe versus mild LAD stenoses is shown in Fig 7⇓. The mean sestamibi defect count ratio in the critical stenosis group (0.62±0.05) was significantly less than in the mild stenosis group (0.80±0.02) (P<.01).
When adenosine is infused directly into the normal LAD, there is preferential vasodilatation of the LAD over the LCx without lowering mean arterial and coronary perfusion pressures. Maintenance of coronary perfusion pressure resulted in microsphere-determined LAD coronary flows in some myocardial regions that were 7 to 9 times those in the corresponding LCx bed. Although flows in this range are considerably higher than those reported in humans with intravenous adenosine,22 it was the goal of this protocol to achieve as wide a range of coronary flow as possible to better characterize the myocardial extraction of 201Tl and sestamibi at high flow. We found that both sestamibi and 201Tl significantly underestimated the flow increase after vasodilatation of the LAD perfusion bed, with the underestimation being greater for sestamibi.
In dogs with either critical or mild LAD stenoses (protocols 2 and 3), adenosine was administered intravenously to more closely mimic the clinical situation. The dose administered (300 μg · kg−1 · min−1) was chosen empirically to give the largest increase in coronary flow without lowering mean arterial pressure below 85 mm Hg. In these dogs, there was a decrease in mean arterial pressure and a slight but statistically insignificant reflex rise in heart rate during adenosine infusion. Similar findings have been observed in patients undergoing vasodilatation with adenosine.23 In the dogs with critical LAD stenoses, no change in LAD coronary flow with adenosine stress was seen. In dogs, the total abolition of coronary flow reserve corresponds to an approximately 90% coronary stenosis.24 The fourfold increase in LCx flow resulted in a 4:1 flow disparity between the LCx and LAD coronary supply zones at the time when 201Tl, sestamibi, and microspheres were administered. As expected, the magnitude of flow heterogeneity between stenotic and normal perfusion beds was underestimated by both sestamibi and 201Tl, but more so by sestamibi. In the dogs with mild LAD stenoses, although the change in LCx flow was similar to that in the critical stenosis group, the LAD flow doubled after adenosine, resulting in a 2:1 flow disparity between the LCx and LAD. In this situation of a milder LAD stenosis, both tracers again underestimated the flow disparity between the two perfusion beds.
Comparison Between 201Tl and Sestamibi Activities With Microsphere Flow
Previous studies have examined the myocardial extraction of either 201Tl or sestamibi at high coronary flow rates. Using pharmacological stress, several groups of investigators have demonstrated a progressive decrease in 201Tl5 6 or sestamibi3 extraction with increasing coronary flow. In this study, we sought to directly compare the uptakes of 201Tl and sestamibi in the same experimental animals under several experimental conditions simulating clinical situations. As in previous studies, we found that the early uptake of both 201Tl and sestamibi plateaus with increasing coronary flow, leading to an underestimation of flow at high flow rates. However, in this study we also determined that the extent of the underestimation was greater for sestamibi than for 201Tl in dogs with either critical or mild LAD stenoses. The decreased extraction of sestamibi relative to 201Tl at high coronary flow rates raises the theoretical possibility of reduced defect contrast with sestamibi compared with 201Tl, particularly with mild stenoses in which it is necessary to resolve the difference between two high flow rates. Several recent studies support this possibility. In an experimental study by Melon et al,25 the myocardial retentions of sestamibi and 201Tl were compared at various times after the tracers were injected in dogs with a totally occluded LAD during dipyridamole infusion. These investigators found that myocardial 201Tl tissue retention matched myocardial blood flow more closely than did sestamibi retention, and for both of these tracers, at high flow rates the blood flow match was best early after injection. In another experimental study, Leon et al20 compared polar map displays of 201Tl and sestamibi administered to dogs with either moderately stenotic or totally occluded LADs during adenosine infusion. These investigators showed that sestamibi underestimated the size of the stenosis territory compared with 201Tl, whereas both tracers showed similar defect sizes in dogs with totally occluded coronary arteries. Our study differs from that of Leon et al in several important aspects. In their experiments, the severity of coronary stenoses was substantially greater than in our study. In 3 of the 4 dogs composing their “severe stenosis” group, resting baseline flow was reduced before adenosine infusion. Also, they did not inject 201Tl and sestamibi simultaneously at identical flows during hyperemia. In the study by Leon et al, 201Tl images were obtained 8 minutes after injection, whereas sestamibi images were obtained 60 minutes or longer after tracer injection, which may have resulted in some sestamibi redistribution.26 Furthermore, these investigators did not perform postmortem gamma-well scintillation counting of tracer activities in the myocardium or compare tracer uptake patterns with an independent measure of regional blood flow. Nevertheless, both studies are consistent in showing underestimation of defect size and magnitude of hypoperfusion with sestamibi compared with 201Tl when these tracers are administered during adenosine-induced flow heterogeneity between stenotic and normal myocardial perfusion beds.
Similar results have been reported in clinical studies by Narahara et al.18 Patients undergoing symptom-limited exercise stress testing had smaller sestamibi SPECT defect sizes than with stress 201Tl SPECT imaging (42±39.9 versus 52±46.2 g, P<.05). Using another method for quantifying SPECT defect size, Maublant et al19 also found smaller exercise stress defect sizes in patients with sestamibi (4.6±5.2%) than with 201Tl (6.7±5.2%) (P<.05).
Comparison Between 201Tl and Sestamibi Activity by Gamma-Camera Imaging
In the present study, we obtained gamma-camera images of ex vivo slices of the heart directly on the collimator of the camera. Sestamibi images obtained in this fashion represent ideal conditions, since there was less scatter and attenuation than with in vivo imaging to degrade image quality. Since both 201Tl and sestamibi were administered simultaneously, separation of the pure 201Tl image from the dual isotope mixture is problematic. A fraction of the 99mTc isotope counts spilling down into the 201Tl window cannot simply be subtracted, as undertaken with in vitro well counting, because with imaging, spatial factors must also be taken into account. Although methods for performing this subtraction are currently being investigated,27 further validation is necessary before these techniques can be implemented. Therefore, in this study, quantification was performed only on the sestamibi images. As shown in Fig 7⇑, sestamibi defect magnitude was significantly greater in critically stenotic than in mildly stenotic dogs. In the latter group, only an average 20% reduction in sestamibi counts was observed in the defect region. Nevertheless, defects were still visually apparent on the scintigrams of the myocardial slices in all dogs with a mild stenosis.
Fig 4⇑ shows how the extraction of sestamibi and 201Tl varies with myocardial blood flow. These curves are characteristic of any extractable tracer in that the amount of tracer extracted is proportional to flow only at low flow rates (where tracer extraction is flow limited) and changes to a plateau at high rates, at which the extraction of tracer becomes limited by membrane transport.
The solid lines of Fig 4⇑ are curve fits of the function b(1−exp−ps/b), where b is blood flow per unit volume and ps is capillary permeability times surface area product. This function is from the solute transport model as detailed by Gosselin and Stibitz.28 This is a general model describing an extractable tracer passing through the capillary bed. The model accounts for back-diffusion of tracer as well as direct extraction. At flow rates <1.0 mL · min−1 · g−1, tracer extraction is limited by blood flow and the amount of extracted tracer is proportional to blood flow. As flow increases, less tracer is directly extracted during the capillary transit, and tracer that diffuses back into the capillary channel is less likely to be reabsorbed before it reaches the venules. Thus, at very high flow rates, tracer extraction is determined only by membrane permeability and is no longer dependent on blood flow.
The model described above fits our experimental data. In this model, the extractable tracer behaves like a perfect (ie, microsphere) tracer only at very low flow rates and progressively underestimates flow relative to the perfect tracer at increasingly higher myocardial flow rates. In clinical imaging, we do not have a perfect flow tracer for comparison and normally use the uptake of the injected tracer in a sample of normally perfused myocardium as a reference standard. Compared with the extraction of 201Tl or sestamibi by normally perfused myocardium, the extraction of these tracers is greater in underperfused beds. Thus, the extraction appears to be enhanced at low flows. The tracer overestimates blood flow in ischemic regions compared with normally perfused regions and underestimates blood flow in regions of relatively elevated blood flow.
The blood flow dependence of extraction shown in Fig 4⇑ and discussed above will result in mild underestimation of the amount of flow reduction when normally perfused myocardium is compared with ischemic myocardium. Underestimation of relative flow will be greater when segments with vasodilator-induced hyperemia are compared. The amount of underestimation depends on tracer extraction, being more pronounced with tracers that have lower myocardial first-pass extraction.
Limitations of the Study
One potential limitation to the present study concerns the use of anesthesia. Anesthesia may have some unknown effect on blood flow distribution or tracer transport. This was not considered to be a serious limitation, since the measurements of tracer uptake were relative rather than absolute measurements. A second limitation was that in vivo imaging was not performed. With closed-chest in vivo imaging, attenuation and scatter would have more closely mimicked the clinical setting. However, unless the two tracers were administered at different times with separate imaging protocols undertaken, the dual isotope spill correction problem would still exist. A strength of this study is that 201Tl and sestamibi were administered simultaneously to be certain that flow at the time of injection was identical for both tracers. This permitted valid comparison of myocardial uptake patterns of these agents during adenosine-induced hyperemia by use of highly accurate in vitro gamma-well counting techniques.
The results of the present study suggest that it is theoretically possible for mild stenoses to be better detected by 201Tl rather than sestamibi in conjunction with vasodilator stress. However, it should be pointed out that with clinical imaging, ideal conditions do not exist and other factors such as scatter and attenuation degrade overall image quality, particularly with the lower energy of 201Tl. In addition, in this study there was no appreciable 201Tl redistribution, since the animals were euthanatized several minutes after tracer injection. Redistribution will diminish defect magnitude. These degradative factors are greater for 201Tl than for sestamibi and may offset the difference in extraction between the tracers at high coronary flow. This may explain why clinical studies have found similar sensitivities, specificities, and overall diagnostic accuracy for detection of coronary artery disease between 201Tl and sestamibi.8 9 10 11 12 13 14 15 16 17 Further clinical studies are warranted that compare 201Tl and sestamibi SPECT imaging in the same patients for detection of mild coronary artery stenoses by use of vasodilator stress.
A second implication of the present study is related to the use of higher doses of dipyridamole or adenosine to achieve greater flow disparity between normal and stenotic regions. Measurements of coronary flow in humans by use of a Doppler catheter have shown that the standard dose of dipyridamole currently administered (0.56 mg/kg) may not be achieving maximal coronary vasodilation29 and that adenosine is perhaps more effective than dipyridamole for achieving maximal hyperemia.30 Parodi et al31 recently proposed using a higher dose of dipyridamole with sestamibi to improve image contrast. The present study suggests that, because of the plateau in extraction of both 201Tl and sestamibi at coronary flow rates as low as two to three times normal flow, additional pharmacological vasodilatation will add minimal enhanced contrast to the images for defect detection.
This study was supported in part by grant ROI HL-26205-11 from the National Heart, Lung, and Blood Institute of the National Institutes of Health, Bethesda, Md, and by a research grant from E.I. Dupont de Nemours & Co, Inc, North Billerica, Mass.
Reprint requests to David K. Glover, Division of Cardiology, Department of Medicine, Box 158, University of Virginia Health Sciences Center, Charlottesville, VA 22908.
- Received February 3, 1994.
- Accepted August 19, 1994.
- Copyright © 1995 by American Heart Association
Okada RD, Glover D, Gaffney T, Williams S. Myocardial kinetics of technetium-99m-hexakis-2-methoxy-2-methylpropyl isonitrile. Circulation. 1988;77:491-498.
Glover DK, Okada RD. Myocardial kinetics of Tc-MIBI in canine myocardium after dipyridamole. Circulation. 1990;81:628-636.
Mays AE, Cobb FR. Relationship between regional myocardial blood flow and thallium-201 distribution in the presence of coronary artery stenosis and dipyridamole-induced vasodilation. J Clin Invest. 1984;73:1359-1366.
Weich HF, Strauss HW. The extraction of thallium-201 by the myocardium. Circulation. 1977;56:188-191.
Leppo JA, Meerdink DJ. Comparison of the myocardial uptake of a technetium-labeled isonitrile analogue and thallium. Circ Res. 1989;65:632-639.
Wackers FJTH, Berman DS, Maddahi J, Watson DD, Beller GA, Strauss HW, Boucher CA, Picard M, Holman BL, Fridrich R, Inglese E, Delaloye B, Bischof-Delaloye A, Camin L, McKusick K. Technetium-99m hexakis 2-methoxyisobutyl isonitrile: human biodistribution, dosimetry, safety, and preliminary comparison to thallium-201 for myocardial perfusion imaging. J Nucl Med. 1989;30:301-311.
Kahn JK, McGhie I, Akers MS, Sills MN, Faber TL, Kulkami PV, Willerson JT, Corbett JR. Quantitative rotational tomography with Tl-201 and Tc99m-2-methoxyisobutyl-isonitrile: a direct comparison in normal individuals and patients with coronary artery disease. Circulation. 1989;79:1282-1293.
Iskandrian AS, Heo J, Kong B, Lyons E, Marsch S. Use of technetium-99m isonitrile (RP-30A) in assessing left ventricular perfusion and function at rest and during exercise in coronary artery disease, and comparison with coronary arteriography and exercise thallium-201 SPECT imaging. Am J Cardiol. 1989;64:270-275.
Sinusas AJ, Beller GA, Smith WH, Vinson EL, Brookeman V, Watson DD. Quantitative planar imaging with technetium-99m methoxyisobutyl isonitrile: comparison of uptake patterns with thallium-201. J Nucl Med. 1989;30:1456-1463.
Tartagni F, Dondi M, Limonetti P, Franchi R, Maiello L, Monetti N, Magnani B. Dipyridamole technetium-99m-2-methoxy isobutyl isonitrile tomoscintigraphic imaging for identifying diseased coronary vessels: comparison with thallium-201 stress-rest study. J Nucl Med. 1991;32:369-376.
Marwick T, Willemart B, D’Hondt A, Baudhuin T, Wijns W, Detry J, Melin J. Selection of the optimal nonexercise stress for the evaluation of ischemic regional myocardial dysfunction and malperfusion: comparison of dobutamine and adenosine using echocardiography and Tc-99m-MIBI single photon emission computed tomography. Circulation. 1993;87:345-354.
Narahara KA, Villanueva-Meyer J, Thompson CJ, Brizendine M, Mena I. Comparison of thallium-201 and technetium-99m hexakis 2-methoxyisobutyl isonitrile single photon emission computed tomography for estimating the extent of myocardial ischemia and infarction in coronary artery disease. Am J Cardiol. 1990;66:1438-1444.
Maublant JC, Marcaggi X, Lusson JR, Boire JY, Cauvin JC, Jacob P, Veyre A, Cassagnes J. Comparison between thallium-201 and technetium-99m methoxyisobutyl isonitrile defect size in single-photon emission computed tomography at rest, exercise and redistribution in coronary artery disease. Am J Cardiol. 1992;69:183-187.
Leon AR, Eisner RL, Martin SE, Schmarkey LS, Aaron AM, Boyers AS, Burnham KM, Oh DJ, Patterson RE. Comparison of single-photon emission computed tomographic (SPECT) myocardial perfusion imaging with thallium-201 and technetium-99m sestamibi in dogs. J Am Coll Cardiol. 1992;20:1612-1625.
Heymann MA, Payne BD, Hoffman JIE, Rudolph AM. Blood flow measurements with radionuclide-labeled particles. Prog Cardiovasc Dis. 1987;20:55-79.
Wilson RF, Wyche K, Christensen BSN, Zimmer S, Laxson DD. Effects of adenosine on human coronary arterial circulation. Circulation. 1990;82:1595-1606.
Verani MS, Mahmarian JJ, Hixson JB, Boyce TM, Staudacher RA. Diagnosis of coronary artery disease by controlled coronary vasodilation with adenosine and thallium-201 scintigraphy in patients unable to exercise. Circulation. 1990;82:80-87.
Sinusas AJ, Bergin JD, Edwards NC, Watson DD, Ruiz M, Makuch RW, Smith WH, Beller GA. Redistribution of 99mTc-sestamibi and 201Tl in the presence of a severe coronary artery stenosis. Circulation. 1994;89:2332-2341.
Smith WH, Kurowski CJ, Sobczak M, Beller GA, Watson DD. Enhanced spill correction techniques enable recovery of pure 201Tl image in dual-isotope imaging of Tc-99m and Tl-201. J Nucl Med. 1991;32:1059. Abstract.
Wilson RF, Laughlin DE, Ackell PH, Chilian WM, Holida MD, Hartley CJ, Armstrong ML, Marcus ML, White CW. Transluminal subselective measurement of coronary artery blood flow velocity and vasodilator reserve in man. Circulation. 1985;72:82-92.
Parodi O, Marcassa C, Casucci R, Sumbuceti G, Verna E, Galli M, Inglese E, Marzullo P, Pirelli S, Bisi G, Giubbini R, Scopinaro F. Accuracy and safety of technetium-99m hexakis 2-methoxy-2-isobutyl isonitrile (sestamibi) myocardial scintigraphy with high dose dipyridamole test in patients with effort angina pectoris: a multicenter study. J Am Coll Cardiol. 1991;18:1439-1444.