Effects of Low Flow and Hypoxia on a New Ischemia-Avid Myocardial Imaging Agent
Background 99mTc-HL91 is a potential imaging agent that has demonstrated increased uptake in hypoxic tumor cells. The purpose of this study was to determine if 99mTc-HL91 demonstrates increased uptake and retention in ischemic and hypoxic myocardium.
Methods and Results 99mTc-HL91 (11.1 MBq) was infused over 10 minutes, followed by a 60-minute clearance phase. Activity was monitored by using an NaI detector. Three groups were studied using Krebs-Henseleit buffer (KH): controls (12 mL/min, n=6), low-flow ischemic (1 mL/min, n=7), and hypoxic (12 mL/min, n=8). Two groups were perfused with KH, red blood cells, and albumin: controls (6 mL/min, n=6) and low-flow ischemic (0.5 mL/min, n=6). For the KH hearts, the 99mTc-HL91 peak uptake progressively increased from control (6.3±0.5 μCi, mean±SEM) to hypoxic (9.1±1.0 μCi) to low flow (44.0±2.6 μCi; P<.01). The peak uptake low-flow/control ratio was 7:1. Final retention increased progressively from control (0.8±0.1 μCi) to hypoxic (2.9±0.5 μCi) to low flow (10.9±1.3 μCi; P<.01). The final low-flow/control activity ratio was 13.6:1. Similar results were observed in the red blood cell–perfused control and low-flow groups.
Conclusions This study introduces a new myocardial “hot spot” imaging agent, 99mTc-HL91. This agent demonstrates increased myocardial uptake and retention in hypoxic and low-flow ischemic models. Further in vivo imaging studies are warranted to determine the clinical potential of this agent.
A “hot spot” myocardial ischemia–imaging agent detectable with gamma camera technology could be very valuable in cardiac patient care. Nitroimidazole-containing compounds have been described as markers of tissue hypoxia. Initial reports of hypoxia detection in tumors employed either fluorine-18–labeled misonidazole and positron imaging1 2 3 4 5 or the iodine-123–labeled misonidazole analogue iodoazomycin arabinoside and gamma camera imaging.6 7 8 Initial cardiac studies in myocytes, perfused rat hearts, and dogs describe the detection of hypoxia using 18F-labeled misonidazole and positron imaging.9 10 11 12 13 More recent studies have reported research on new 123I- and 99mTc-labeled compounds that contain 2-nitroimidazole functionalities,14 15 16 17 18 19 20 thus allowing the possibility of gamma camera imaging. However, attempts to image the heart using nitroimidazoles and gamma camera imaging have been frustrated by high hepatic uptake.
Recently, a 99mTc-labeled compound that has demonstrated both increased uptake by hypoxic tumor tissue and reduced hepatic uptake has been synthesized.21 The purpose of the current study was to determine if this agent, 99mTc-HL91 (HL91=4,9-diaza-3,3,10,10-tetramethyldodecan-2,11-dione dioxime), would demonstrate increased uptake and retention in ischemic and hypoxic myocardium.
Preparation of Isolated Perfused Rat Heart and 99mTc-HL91
The agent was supplied as a freeze-dried solid containing the ligand HL91, stannous chloride dihydrate, stabilizing agent, and buffer salt in a 10-mL glass vial sealed with a rubber closure under an inert nitrogen atmosphere. Each vial was reconstituted with 5 mL of a sterile sodium pertechnetate solution containing ≈1 GBq (25 mCi) of 99mTc activity. The vial was then shaken gently to ensure complete dissolution of the lyophilized powder and allowed to stand at room temperature (15°C to 25°C) for at least 15 minutes. The radiochemical purity of the reconstituted preparation was determined22 and was always >95% before use. Fig 2⇓ shows the chemical structure of 99mTc-HL91.
Method for Measuring Tissue Oxygen Pressure
Tissue oxygen concentration was quantified by measuring phosphorescence using noninvasive optical techniques. This measurement technique has been well documented23 relative to other available methods for measuring oxygen in mammalian tissue and blood. Specifically, the addition of a phosphorescent probe to the perfusate entering the heart allowed sensitive measurements of oxygen pressure within the physiologically important range. The oxygen probe used in these studies was a Pd complex of meso-tetra (4-carboxyphenyl) porphine (35 mg/L) (Porphyrin Products).
Fig 3⇓ shows the experimental protocols used in this study. The hearts were perfused in a retrograde manner in the Langendorff mode with either nonradioactive KH buffer at 12 mL/min or RBCs at 6 mL/min for a 20-minute stabilization period. Low-flow or hypoxic conditions were then instituted for 15 minutes in groups 2, 3, and 5. 99mTc-HL91 (300 μCi, 11.1 MBq) was infused constantly over a 10-minute period by switching to the tank containing the tracer. This constant infusion technique was used instead of a bolus in order to accurately identify peak activity. 99mTc activity was monitored by using a NaI probe. Perfusion was then switched to the third tank containing nonradioactive perfusate, and 99mTc-HL91 clearance was monitored for 60 minutes. Three groups of KH buffer–perfused hearts were studied. Group 1 (KH control, n=6) hearts were perfused at 12 mL/min and group 2 (KH low-flow–ischemic, n=7) hearts were perfused at 1 mL/min for 15 minutes after baseline and during 10 minutes of uptake and 60 minutes of clearance. For group 3 (KH normal-flow–hypoxic, n=8) hearts, the perfusate was switched at the end of baseline from 95% O2/5% CO2–bubbled KH to 95% N2/5% CO2–bubbled KH for 15 minutes after baseline and during 10 minutes of uptake and 60 minutes of clearance. Flow was 12 mL/min throughout the experiment.
KH buffer has a lower oncotic pressure than blood and thus, as shown in pilot studies in our laboratory, produces edema formation. To determine whether these differences would alter 99mTc-HL91 myocardial kinetics, two groups of KH buffer plus RBC (hematocrit, 35% to 40%) and albumin (3 g percent serum albumin) hearts were studied. Pilot studies showed no significant edema formation with this perfusate. Group 4 (RBC control, n=6) hearts were perfused at 6 mL/min and group 5 (RBC low-flow–ischemic, n=6) hearts were perfused at 0.5 mL/min for 15 minutes after baseline and during 10 minutes of uptake and 60 minutes of clearance. In group 4 and 5 hearts, tissue oxygen concentration was quantified by measuring phosphorescence using the noninvasive optical technique described above. Because this technique could not be reliably used in the KH-perfused hearts, perfusate oxygen content was measured by using Clark-type electrodes. At the end of the experiments, the remaining activity in all groups was counted by using a dose calibrator (Squibb, CRC-17) and gamma well counter (LKB 1282 Compugamma).
Different flow rates were used to perfuse hearts with KH buffer and RBC plus albumin buffers as described above. KH buffer hearts were perfused at 12 mL/min, which provides adequate oxygenation with this buffer. Perfusion rates for the RBC plus albumin hearts also provided adequate oxygenation due to the increased oxygen-carrying capacity of the RBCs. The same relative flow reduction was chosen for the low-flow groups.
99mTc activity in the hearts was measured at 1-minute intervals by using a collimated NaI detector positioned 3 cm from the left side of the heart. The detector system was calibrated before the experiments by being exposed to a series of Tc sources of known activity. The detector was interfaced with a window discriminator, and the data were transferred to a multichannel analyzer (Canberra Nuclear) that displayed time-activity curves. Myocardial 99mTc-HL91 activity was monitored during uptake and clearance.
Data Analysis and CK Assay
MUCs were plotted by using mean multichannel analyzer counts corrected for background and decay. Background activity was recorded after the heart was removed from the apparatus at the end of the experiment. This background was always <0.5% of the final activity. MCCs were plotted by using multichannel analyzer activities normalized to peak activity. The final 60-minute postclearance activities shown in Figs 5⇓ and 7⇓ were generated by using activities determined by the well counter. Peak uptake activities for Figs 5⇓ and 7⇓ were generated by using the final counted activities and the probe-determined clearances. For example, if the heart had 10 μCi of activity measured in the well counter at the end of the experiment and the probe-determined clearance was 50%, then the peak activity was calculated as 20 μCi. Percent injected dose was this value divided by the total administered dose ×100%.
The CK assay technique has been reported.22
An ANOVA procedure (Crunch Statistical Software) was used to analyze group differences. When the assumption of homogeneity of variance among groups was violated, the equivalent nonparametric (Kruskal-Wallis) analysis was used. Tests of mean differences were conducted by analysis with t tests by using the Bonferroni correction for multiple tests. Data are reported as mean±SEM. A value was considered significant when P<.05.
The myocardial retention curves from each of the individual experiments were fit by nonlinear estimation using Tablecurve 2D software (Jandel Scientific). The best-fit statistic used was the coefficient of determination.
The KH and RBC control groups demonstrated no significant changes in HR, systolic and diastolic LV pressures, CPP, or perfusate oxygen content during the experiments. In addition, the RBC control group demonstrated no significant change in tissue oxygenation.
For the KH perfusion studies, baseline HR, systolic and diastolic LV pressures, CPP, and oxygen content of the perfusate did not differ among the three groups. During the experiments, HR fell (216.7 and 103.3 bpm), systolic pressure fell (32.0 and 44.9 mm Hg), perfusate oxygen content fell (91% and 283.5% saturation), and diastolic pressure rose (7.6 and 9.9 mm Hg) significantly compared with baseline for the low-flow–ischemic and normal-flow–hypoxic groups, respectively. CPP fell 36.6 mm Hg in the low-flow–ischemic hearts, but no significant change was observed in the normal-flow–hypoxic hearts.
For the RBC perfusion studies, baseline HR, systolic and diastolic LV pressures, CPP, and tissue oxygen content did not differ between the control and low-flow–ischemic groups. However, diastolic pressure rose significantly (32.4 mm Hg) and CPP (24.6 mm Hg), HR (135 bpm), and tissue oxygen (20.2 torr; 1 torr=1 mm Hg) all fell significantly in the low-flow–ischemic group during the experiment.
No significant increases in CK developed in any of the five groups. Compared with baseline, mean normalized CK values were 98.9±56.6% for KH control, 87.1±20.1% for KH low-flow–ischemic, 55.3±11.5% for KH normal-flow–hypoxic, 85.9±49.1% for RBC control, and 130.2±45.5% for RBC low-flow–ischemic hearts.
99mTc-HL91 Myocardial Uptake
Fig 4⇓ demonstrates the 99mTc-HL91 MUCs for the three KH groups. Normal-flow–hypoxic hearts demonstrated a significant increase in uptake compared with the KH control (P<.05) starting at 7 minutes. The KH low-flow–ischemic group demonstrated marked and significantly increased 99mTc-HL91 uptake compared with the control group (P<.01) starting at 1 minute. Fig 5⇓ demonstrates the peak myocardial 99mTc-HL91 activities for the three KH-perfused groups. Peak activities were 6.3±0.5 μCi (2.1% of the injected dose) for control, 9.1±1.0 μCi (3.0% of the injected dose) for normal-flow–hypoxic, and 44.0±2.6 μCi (14.6% of the injected dose) for the low-flow–ischemic hearts (P<.01 between groups). The peak normal-flow–hypoxic/control activity ratio was 1.4:1. The peak low-flow–ischemic/control activity ratio was 7.0:1.
Fig 6⇓ demonstrates the 99mTc-HL91 MUCs for the two RBC-perfused groups. The low-flow–ischemic group demonstrated significantly increased 99mTc-HL91 uptake compared with the control group (P<.01) starting at 1 minute. Fig 7⇓ demonstrates the peak myocardial 99mTc-HL91 activities for the two RBC-perfused groups. Peak activities were 7.6±0.7 μCi (2.5% of the injected dose) for control and 33.7±8.2 μCi (11.2% of the injected dose) for the low-flow–ischemic hearts (P<.01). The peak low-flow–ischemic/control activity ratio was 4.4:1.
99mTc-HL91 Myocardial Clearance and Retention
Fig 8⇓ demonstrates the MCCs for the KH perfusion studies. MCCs were biphasic for the control, low-flow–ischemic, and normal-flow–hypoxic groups. Clearance for all three groups demonstrated a rapid early phase and a slower late phase. The clearance curves for both the low-flow–ischemic and the normal-flow–hypoxic groups became significantly slower than that of the control group after 2 minutes of clearance and remained so until the end of 60 minutes (P<.05 for both). There was significantly more retention in the low-flow–ischemic group than the normal-flow–hypoxic group during the first 15 minutes of clearance (P<.05). During the entire 60-minute clearance phase, fractional retention was 12.5±1.9% for control, 30.4±3.9% for low-flow–ischemic, and 33.8±3.2% for normal-flow–hypoxic hearts (P<.01 among groups). Control group retention differed significantly from those of the low-flow–ischemic and normal-flow–hypoxic groups (P<.01), which were not significantly different from each other (P=NS). Fig 5⇑ demonstrates the final myocardial 99mTc-HL91 activities after 60 minutes of clearance for the three KH-perfused groups. Final activities were 0.8±0.1 μCi (0.3% of the injected dose) for control, 2.9±0.5 μCi (1.0% of the injected dose) for normal-flow–hypoxic, and 10.9±1.3 μCi (4.0% of the injected dose) for low-flow–ischemic hearts (P<.01 between groups). The final normal-flow–hypoxic/control activity ratio was 3.6:1. The final low-flow–ischemic/control activity ratio was 13.6:1.
Fig 9⇓ demonstrates the MCCs for the RBC perfusion studies. MCCs were biphasic for the control and low-flow–ischemic groups. Clearance curves for both groups demonstrated a rapid early phase and a slower second phase. The low-flow–ischemic group clearance curve became significantly slower than that of the control group after 2 minutes of clearance and remained so until the end of 60 minutes (P<.05). During the entire 60-minute clearance phase, the fractional retention of peak value was 6.1±1.4% for control and 18.1±2.2% for low-flow–ischemic hearts (P<.01). Fig 7⇑ demonstrates the final myocardial 99mTc-HL91 activities after 60 minutes of clearance for the two RBC-perfused groups. Final activities were 0.4±0.1 μCi (0.2% of the injected dose) for control and 5.4±1.8 μCi (1.8% of the injected dose) for low-flow–ischemic hearts (P<.01). The final low-flow–ischemic/control activity ratio was 13.5:1.
Both the KH- and RBC-perfused hearts demonstrated biexponential clearances. The KH-perfused low-flow–ischemic (4.2±0.8 versus 1.5±0.1 minutes; P<.05) and RBC-perfused low-flow–ischemic (5.3±0.4 versus 2.3±0.1 minutes; P<.05) hearts demonstrated an initial t1/2 longer than the respective control hearts (P<.05). The late t1/2 values were significantly longer for the KH-perfused low-flow–ischemic (387.0±86.9 minutes) and normal-flow–hypoxic (385.7±32.1 minutes) and the RBC-perfused low-flow–ischemic (201.9±30.6 minutes) hearts compared with the control hearts (202.2±18.9 and 92.9±20.5 minutes, respectively; P<.05).
The current study describes 99mTc-HL91, a newly synthesized 99mTc-labeled compound that demonstrates increased uptake by normal-flow–hypoxic and low-flow–ischemic hearts compared with normal control hearts. Many studies with putative hypoxia-imaging agents have used nitroimidazoles, ie, compounds with a 2-nitroimidazole group. Initial cardiac studies that used 18F-labeled misonidazole required positron imaging9 10 11 12 13 and did show increased uptake by ischemic myocardium. Subsequent studies described nitroimidazoles labeled with either 123I or 99mTc, thus allowing the potential for gamma camera imaging.14 15 16 17 18 19 20 These agents were also found to demonstrate increased uptake in hypoxic myocardium. However, all of these initial agents had high hepatic uptake, making cardiac imaging difficult.
Amersham International has synthesized a number of new nitroimidazoles containing Tc ligands. The 99mTc complex of one of these compounds, HL91M, has demonstrated improved hypoxic/normal count ratios compared with previously described compounds. Furthermore, when the 2-nitroimidazole moiety was removed from the core ligand of HL91M, the 99mTc complex of the resulting compound (HL91) exhibited significantly improved hypoxic/normoxic uptake ratios in preliminary studies of several in vitro and in vivo models of hypoxia.21
Hemodynamic and CK Data
HR, systolic and diastolic LV pressures, CPP, and oxygen content remained constant throughout the experiment for the control groups, which indicates that the control preparations remained stable during the study period. Ventricular HR fell in the low-flow–ischemic and normal-flow–hypoxic hearts despite attempts to pace, probably due to hypoxia-induced partial electrical-mechanical dissociation. Systolic LV pressure also fell for the low-flow–ischemic and normal-flow–hypoxic hearts, probably due to ischemic LV dysfunction. LV diastolic pressure rose in the low-flow–ischemic and normal-flow–hypoxic groups, probably due to decreased compliance as a result of inadequate energy supply. However, the diastolic pressure rise was greater in the low-flow–ischemic RBC-perfused hearts than in the low-flow–ischemic KH-perfused hearts. This would suggest a greater insult in the former group. As expected, CPP fell for the low-flow–ischemic hearts and perfusate oxygen saturation fell for the KH-perfused normal-flow–hypoxic hearts. There was a small fall in perfusate oxygen saturation for the low-flow–ischemic hearts, since Clark-type electrodes are also somewhat sensitive to flows due to lack of adequate movement of the perfusate past the electrode surfaces.23 Nevertheless, the oxygen saturations all remained >200%, which would be considered fully saturated. Tissue oxygen tension fell as expected in the RBC-perfused low-flow–ischemic hearts.
There was no significant increase in CK release in any of the five groups, which further supports the hemodynamic evidence of heart stability in the controls during the study. This would also indicate that the low-flow–ischemic and normal-flow–hypoxic models did not cause significant irreversible myocardial injury. However, small amounts of injury cannot be excluded on the basis of the CK analysis, particularly in the RBC low-flow–ischemic hearts, in which there was a trend toward higher values than those observed at baseline.
99mTc-HL91 Myocardial Uptake
99mTc-HL91 myocardial uptake was only 2.3% of the injected dose in the control hearts. However, uptake increased severalfold in the low-flow–ischemic hearts. This confirms the results of studies in isolated fibroblasts that show a 15-fold increased uptake in hypoxic cells after 4.5 hours of incubation.21 The current study demonstrated increased 99mTc-HL91 uptake in normal-flow–hypoxic compared with control myocardium. However, the magnitude of this increased uptake was not as great as for low-flow hearts. Martin and associates13 and Kusuoka and associates16 have made similar observations using 18F-labeled misonidazole and 99mTc-nitroimidazole (BMS181321), respectively. These observations suggest that this agent is diffusion-limited in high-flow states in its uptake and that uptake depends on flow as well as tissue hypoxia at normal to high flows.
The uptake curves shown in Figs 4⇑ and 6⇑ were obtained during steady-state 10-minute 99mTc-HL91 infusions. Thus, these uptake curves represent a combination of wash-in and wash-out. Since the subsequent clearance phase demonstrated increased retention for the low-flow–ischemic hearts, some of the increased accumulation of the tracer during the steady-state infusion could be accounted for by the increased retention. Analysis of the first few minutes of the accumulation curve is probably a better assessment of pure uptake. Figs 4⇑ and 6⇑ demonstrate significantly greater tracer activity in the low-flow–ischemic than in the control hearts 1 minute after tracer administration. This would suggest that the low-flow–ischemic hearts do demonstrate a true increase in tracer accumulation.
In a related class of agents, the mechanism of myocardial nitroimidazole uptake has been thought to be the following. Nitroimidazoles are lipophilic, with an octanol:water partition coefficient of .42. Thus, these agents diffuse easily from the blood stream to tissues. In cells, the nitroimidazoles are metabolized by nitroreductase enzymes.24 In the presence of oxygen, the nitroimidazoles are reformed, but in the absence of oxygen, the metabolites are trapped after binding to intracellular proteins.25 This protein binding appears to be covalent.1 26 27 The nitroimidazoles are not trapped by metabolically inactive myocardium, since the initial metabolism is dependent on nitroreductase enzymes.1 Rumsey and associates15 have shown that mitochondria are involved in the trapping process. The 99mTc-HL91 complex described in the current study has previously demonstrated improved hypoxic/normoxic uptake ratios compared with other nitroimidazoles. In fact, 99mTc-HL91 is an agent produced by removing the 2-nitroimidazole moiety from the parent nitroimidazole ligand (HL91M). Thus, it appears that the 2-nitroimidazole moiety in these complexes is not essential for localization in hypoxic tissue, and our previous concept regarding the mechanism of nitroimidazole retention is too simplistic. There are probably additional reducing steps that occur beyond the cleavage of the nitroimidazole moiety that result in alterations in the charge, shape, or size of the parent compound that in turn result in prolonged retention in hypoxic tissue.
99mTc-HL91 Myocardial Clearance
This study demonstrated biexponential myocardial clearance from normal, low-flow–ischemic, and normal-flow–hypoxic tissues. This clearance consisted of an initial short fast phase followed by a longer slow phase. Furthermore, slower clearances from low-flow–ischemic and normal-flow–hypoxic myocardium were shown. MCC modeling demonstrated early t1/2 values that were not significantly different for control and normal-flow–hypoxic hearts. However, the mean early t1/2 for the low-flow–ischemic hearts was significantly prolonged, which means that flow is an important component of the early clearance t1/2. In contrast, the late t1/2 values for the low-flow–ischemic and normal-flow–hypoxic hearts were significantly prolonged compared with control, which indicates that the late t1/2 is influenced predominantly by the hypoxic state rather than flow per se. The final normal-flow–hypoxic/control retention ratio was 3.6:1; the final low-flow–ischemic/control retention ratio was 13.6:1.
A 13.6-fold increase in 99mTc-HL91 ischemic uptake was found in the current study, which used a model of reduced blood flow of 1/12th normal. It is interesting to speculate concerning the potential relationship between these two numbers. Regarding the wash-in side of the 99mTc-HL91 curve, studies using another hypoxia-avid agent, BMS181321, have also demonstrated increased tracer uptake with decreased flow.18 28 Studies using flow agents such as teboroxime have demonstrated the opposite relationship, ie, increased tracer uptake with increased flow. Regarding the wash-out side of the 99mTc-HL91 low-flow–ischemic curve, it is possible that reduced flow per se rather than tissue hypoxia could contribute to the increased tracer retention. However, since the current study also demonstrated increased 99mTc-HL91 retention in normal-flow–hypoxic hearts, hypoxia per se also contributed to the increased retention.
Comparison With Other Hypoxic Agents
Initial hypoxia-imaging agents required positron camera imaging. Martin and associates13 studied the related compound 18F-labeled misonidazole in isolated rat myocytes. Both anoxia and hypoxia caused increases in misonidazole accumulation. However, in their model of complete anoxia, 60 minutes were required before an 8.4-fold increase in uptake could be achieved. In contrast, a 7.0-fold increase was achieved in the current study after only 10 minutes of uptake. Shelton and associates,11 using perfused rat heart models of ischemia and hypoxia, have demonstrated two- to threefold increases in uptake of 18F-labeled misonidazole after 20-minute uptake and 20-minute clearance phases. In a canine model of circumflex occlusion or stenosis, Martin and associates9 12 have shown that 18F-labeled misonidazole accumulates in inverse proportion to myocardial blood flow, indicating enhanced binding in hypoxic tissues. Maximum tissue concentrations in ischemic myocardium were only eightfold greater than in normal myocardium after 4 hours. This again is in contrast to the current study using 99mTc-HL91, in which more than sixfold increased activity was achieved after only 10 minutes. Shelton and associates,10 using a 3-hour canine occlusion model, reported 23% retention of misonidazole in the occlusion zone versus 2% retention in the normal zone after 3 hours.
More recently, several laboratories have reported results using the 99mTc-labeled nitroimidazole BMS181321. Rumsey and associates15 17 report a good correlation between tracer retention and perfusate oxygen in a perfused rat heart model. Ng and associates20 also found that BMS181321 retention increased with decreasing perfusate oxygen in a perfused rat heart model. Tracer retention varied from 0.61 in normoxic conditions to 5.94 in the most severe hypoxic conditions. Kusuoka and associates,16 who studied BMS181321 wash-out kinetics in perfused rat hearts, have demonstrated increased retention in ischemic myocardium using a protocol consisting of tracer administration followed by occlusion and reperfusion. Our laboratory has demonstrated28 serial increases in BMS181321 uptake and retention with serial reductions in flow in a perfused rat heart model. The peak ischemic/normal heart uptake ratio was 12:1; the final 1-hour ischemic/normal zone retention ratio was 30:1. Stone and associates18 and Shi and associates,19 who have studied BMS181321 in swine and dog models of ischemia, have shown increased tracer uptake and retention detectable with in vivo gamma camera imaging. However, unfavorable hepatic/heart ratios were felt to limit the clinical use of this agent.19 Preliminary reports describing BMS194796, a modification of BMS181321, suggest reduced liver uptake compared with BMS181321.29 30
When comparing the current study with the studies noted above, one should recall that those studies used different models and conditions. Whereas we used a perfused rat heart model, some of the above studies used cultured cells (with no flow), and others used a canine model. The studies that did use perfused rat hearts used different flows and conditions. Thus, a direct comparison of the 99mTc-HL91 uptake ratios achieved in the current study with the results of previous studies is impossible. However, some general comparisons can be made between the current study and results using BMS181321 with a similar protocol.28 Although the infusion times were different, the flow rates and clearance times were similar. At a flow of 1 mL·min−1·g−1, the BMS181321 1-hour fractional retention was 0.59 (versus 0.32 control), whereas the 99mTc-HL91 1-hour fractional retention was 0.30 (versus 0.12 control).
The current study demonstrated increased 99mTc-HL91 myocardial uptake and retention in low-flow–ischemic compared with control myocardium using KH perfusate. Because the KH perfusate requires higher than physiological flow rates to ensure adequate tissue oxygenation, we also examined the effects of low-flow ischemia on 99mTc-HL91 kinetics in an RBC plus albumin–perfused model. The RBC groups also demonstrated increased 99mTc-HL91 myocardial uptake and retention in the low-flow–ischemic hearts. Thus, the properties of 99mTc-HL91 in myocardial tissue observed in this study appear to be relatively independent of perfusate composition. However, reduced uptake and faster clearance compared with KH study results do indicate binding to RBCs, albumin, or both.
The current study examined 99mTc-HL91 uptake and retention in normal, low-flow–ischemic, and normal-flow–hypoxic myocardium. These protocols were designed to produce myocardial ischemia with minimal injury. The CK results indicate that significant amounts of irreversible injury were avoided. Consequently, the uptake and retention kinetics of 99mTc-HL91 in nonviable myocardium are unknown. Only after it is determined that 99mTc-HL91 can distinguish viable from nonviable ischemic myocardium will its full potential for clinical use become clear.
We used a perfused isolated heart model, which has no recirculation, and infusion rather than bolus injection. To more closely simulate the clinical situation, 99mTc-HL91 tracer kinetics need to be studied in vivo in large animal models that can be imaged.
Different flow rates were used to perfuse hearts with KH and RBC buffers to compensate for the greater oxygen-carrying capacity of the RBCs. Although generally successful, it should be noted that some of the hemodynamic parameters and CK releases tended to be different for the two perfusates during low flow.
Currently available myocardial imaging agents such as thallium, sestamibi, teboroxime, and tetrofosmin demonstrate decreasing tracer uptake as flow falls. This produces defects on clinical images. The newer ischemia-avid nitroimidazoles demonstrate increasing tracer uptake as flow falls, thus producing a “hot spot” on images. Previous investigators have reported increased nitroimidazole uptake in ischemic models using positron imaging. Attempts to label nitroimidazoles with radioisotopes suitable for imaging with more widely available gamma cameras have been frustrated by increased hepatic uptake. In contrast, 99mTc-HL91 has low hepatic uptake, and uptake and retention are increased in low-flow–ischemic and normal-flow–hypoxic myocardium. The final low-flow–ischemic/normal-flow heart activity ratio was 13.6:1. Although the final normal-flow–hypoxic/normal-flow heart ratio was lower, it was still significant at 3.6:1. Gamma camera imaging of ischemic hot spots should be attainable with ratios of this magnitude. The ability to positively image myocardial ischemia could be helpful in several clinical situations: (1) the detection of chronically ischemic (hypoxic) and nonfunctional but viable myocardium (hibernating myocardium) prior to cardiac transplantation versus coronary artery bypass; (2) the detection of salvageable ischemic myocardium prior to reperfusion therapy in the setting of acute myocardial infarction; and (3) the detection of transient myocardial ischemia produced by exercise in the diagnosis of coronary artery disease. The latter may be technically difficult due to relatively short periods of ischemia.
The current study demonstrated increased 99mTc-HL91 initial accumulation in normal-flow–hypoxic and low-flow–ischemic models. The results also demonstrated differential clearance kinetics for normal-flow–hypoxic and low-flow–ischemic compared with normal myocardium that were apparent after only 2 minutes of clearance. These findings suggest that clinical imaging protocols producing quantitative clearance kinetics may result in rapid diagnoses of ischemia. However, until further in vivo imaging studies are performed, the clinical possibilities mentioned above remain speculative.
This is the first full study on the myocardial imaging potential of 99mTc-HL91, an agent that shows increased myocardial accumulation and retention in normal-flow–hypoxic and low-flow–ischemic viable myocardial models. After 60 minutes of clearance, a low-flow–ischemic/control heart activity ratio of 13.6:1 can be achieved. However, the uptake and retention kinetics in nonviable tissue are still unknown. Further studies are warranted based on these findings.
Selected Abbreviations and Acronyms
|CPP||=||coronary perfusion pressure|
|MCC||=||myocardial clearance curve|
|MUC||=||myocardial uptake curve|
|RBC||=||red blood cell|
We wish to acknowledge the excellent secretarial assistance of Andrea Lightfoot. This study is dedicated to William K. Warren, Sr, and family for their support of medical research.
- Received May 23, 1996.
- Revision received November 26, 1996.
- Accepted November 27, 1996.
- Copyright © 1997 by American Heart Association
International Conference on Dose, Time, and Fractionation. Prediction of response in radiation therapy. In: Paliwal BR, et al, eds. Symposium Proceedings of the American Association of Physicists in Medicine. New York, NY: American Institute of Physics; 1989;2:1-757.
Groshar D, McEwan AJB, Parliament MB, Urtasun RC, Golberg LE, Hoskinson M, Mercer JR, Mannan RH, Wiebe LI, Chapman JD. Imaging tumor hypoxia and tumor perfusion. J Nucl Med. 1993;34:885-888.
Mannan RH, Somayaji VV, Lee J, Mercer JR, Chapman JD, Wiebe LI. Radioiodinated 1-(5-iodo-5-deoxy-β-D-arabinofuranosyl)-2-nitroimidazole (iodoazomycin arabinoside: IAZA): a novel marker of tissue hypoxia. J Nucl Med. 1991;32:1764-1770.
Martin GV, Caldwell JH, Graham MM, Grierson JR, Kroll K, Cowan MJ, Lewellen TK, Rasey JS, Casciari JJ, Krohn KA. Noninvasive detection of hypoxic myocardium using fluorine-18-fluoromisonidazole and positron emission tomography. J Nucl Med. 1992;33:2202-2208.
Shelton ME, Dence CS, Hwang D-R, Herrero P, Welch MJ, Bergmann SR. In vivo delineation of myocardial hypoxia during coronary occlusion using fluorine-18 fluoromisonidazole and positron emission tomography: a potential approach for identification of jeopardized myocardium. J Am Coll Cardiol. 1990;16:477-485.
Shelton ME, Dence C, Hwang D-R, Welch MJ. Myocardial kinetics of fluorine-18 misonidazole: a marker of hypoxic myocardium. J Nucl Med. 1989;30:351-358.
Martin GV, Caldwell JH, Rasey JS, Grunbaum Z, Cerqueira M, Krohn KA. Enhanced binding of the hypoxic cell marker [3H] fluoromisonidazole in ischemic myocardium. J Nucl Med. 1989;30:194-201.
Martin GV, Cerqueira M, Caldwell JH, Rasey JS, Embree L, Krohn KA. Fluoromisonidazole: a metabolic marker of myocyte hypoxia. Circ Res. 1990;67:240-244.
Martin GV, Biskupiak JE, Caldwell JH, Rasey JS, Krohn KA. Characterization of iodovinylmisonidazole as a marker for myocardial hypoxia. J Nucl Med. 1993;34:918-924.
Kusuoka H, Hashimoto K, Fukuchi K, Nishimura T. Kinetics of a putative hypoxic tissue marker, technetium-99 m-nitroimidazole (BMS181321), in normoxic, hypoxic, ischemic and stunned myocardium. J Nucl Med. 1994;35:1371-1376.
Rumsey WL, Patel B, Linder K. Effects of graded hypoxia in retention of technetium-99 m-nitroheterocycle in perfused rat hearts. J Nucl Med. 1995;36:632-636.
Stone CK, Mulnix T, Nickles RJ, Renstrom B, Nellis SH, Liedtke J, Nunn AD, Kuczynski BL, Rumsey WL. Myocardial kinetics of a putative hypoxic tissue marker, 99 m-Tc–labeled nitroimidazole (BMS181321), after regional ischemia and reperfusion. Circulation. 1995;92:1246-1253.
Shi CQ, Sinusas AJ, Dione DP, Singer MJ, Young LH, Heller EN, Rinker BD, Wackers FJT, Zaret BL. Technetium-99 m nitroimidazole (BMS181321): a positive imaging agent for detecting myocardial ischemia. J Nucl Med. 1995;36:1078-1086.
Ng CK, Sinusas AJ, Zaret B, Souter R. Kinetic analysis of technetium-99 m–labeled nitroimidazole (BMS181321) as a tracer of myocardial hypoxia. Circulation. 1995;92:1261-1268.
Archer CM, Edwards B, Kelly JD, King AC, Burke JF, Riley ALM. Technetium labelled agents for imaging tissue hypoxia in vivo. In: Nicolini M, Bandoli G, Mazzi U, eds. Technetium and Rhenium in Chemistry and Nuclear Medicine. Padova, Italy: SGE Ditoriali Publishers; 1995:535-539.
Vanderkooi JM, Erecinska M, Silver IA. Oxygen in mammalian tissue: methods of measurement and affinities of various reactions. Am J Physiol. 1991;260:C1131-C1150.
Rumsey WL, Patel B, Kuczynski B, Hood C, Linder E, Nunn A, Strauss HW. Comparison of two novel technetium agents for imaging ischemic myocardium. Circulation. 1995;92:181. Abstract.
Stone CK, Mulnix T, Nickles RJ, Carlson KJ, Renstrom B, Nellis SH, Liedke AJ, Nunn AD. Comparison of the myocardial and hepatic uptake of two technetium-labeled nitroimidazoles after regional ischemia and reperfusion using a dual isotope technique. J Nucl Med. 1995;36:138. Abstract.