Published online before print March 27, 2008, doi: 10.1161/STROKEAHA.107.485938
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Stroke. 2008;39:1629.)
© 2008 American Heart Association, Inc.
Topical Review |
Applications of Nitroimidazole In Vivo Hypoxia Imaging in Ischemic Stroke
From the Stroke Research Group, Department of Clinical Neurosciences, University of Cambridge, UK.
Correspondence to Professor Jean-Claude Baron, University of Cambridge, Department of Clinical Neurosciences, Addenbrooke’s Hospital, Box 83, Cambridge CB2 2QQ, UK. E-mail jcb54{at}cam.ac.uk
Wolf-Dieter Heiss MD Gregory Sorensen MD Section Editors:
 |
Abstract |
|---|
Top
Abstract
IntroductionBackground and Purpose— Nitroimidazole imaging is a promising contender for noninvasive in vivo mapping of brain hypoxia after stroke. However, there is a dearth of knowledge about the behavior of these compounds in the various pathophysiologic situations encountered in ischemic stroke. In this article we report the findings from a systematic review of the literature on the use of the nitroimidazoles to map hypoxia after stroke.
Summary of Review— We describe the characteristics of nitroimidazoles as imaging tracers, their pharmacology, and results of both animal and clinical studies during and after focal cerebral ischemia. Findings in brain tumors are also presented to the extent that they enlighten results in stroke. Early results from application of kinetic modeling for quantitative measurement of tracer binding are briefly discussed.
Conclusions— Based on this literature review, nitroimidazole hypoxia imaging agents are of considerable interest in stroke because they appear, both in animal models and in humans, to specifically detect the severely hypoxic viable tissue, but not the reperfused nor the necrotic tissue. To fully realize this potential in stroke, however, formal validation by concurrent measurement of tissue oxygen tension, together with development of novel ligands with faster distribution kinetics, faster clearance from normal tissue, and well-defined trapping mechanisms, are important goals for future investigations.
Key Words: imaging • ischemia • PET • stroke • hypoxia • nitroimidazole
|
Introduction |
|---|
The present management of acute stroke is guided by the identification and ultimately rescue of the ischemic penumbra.1 CBF thresholds defining the penumbra and core have been determined using multi-tracer 15O PET and were thereafter extended to other imaging modalities.2 Nonetheless, CBF alone does not indicate the actual penumbral topography because the infarction threshold depends on time since stroke onset.3 Furthermore, concurrent estimation of the oxygen extraction fraction (OEF) is necessary to define the penumbra in conditions of partial reperfusion and secondary hypoperfusion.4
The penumbra can be defined as hypoxic tissue whose function is reversibly affected by the reduction of tissue partial pressure of oxygen (tPO2), itself caused by the scarcity of blood supply relative to oxygen consumption. Experimental and clinical studies measuring tPO2 directly in acute stroke indicate that it falls from normal values around 30 mm Hg to below 10 mm Hg and may even approach zero (see supplemental Appendix for further detail, available online at http://stroke.ahajournals.org). Mapping of the tPO2 noninvasively would provide a direct way to depict salvageable tissue, either alone or in conjunction with CBF. Although the 15O-PET can provide an estimate of end-capillary PO25, comparison with direct tPO2 measurement has shown discrepancies,6 implying that end-capillary PO2 reflects a different, though partly overlapping, entity. A group of oxygen-sensitive compounds known as 2-nitroimidazoles, originally developed as antibacterial agents, have now been adapted as imaging agents that can be radio-labeled and potentially map the viable hypoxic tissue in the live brain of animals and humans7–9 (see supplemental Appendix for a historical note).
An ideal hypoxia imaging agent for use in stroke should: (1) freely cross the blood brain barrier (BBB) and rapidly localize to the target viable hypoxic tissue with high specificity; (2) have a known relationship to true tPO2 allowing its indirect estimation; (3) have suitable tissue kinetics to allow sufficient tissue contrast within a timeframe permitting its use in the acute setting8 and compatible with known time-course of infarct growth; and (4) have favorable dosing and safety profiles for use in humans.
Although, as will be shown below, such an ideal hypoxia imaging agent remains to be found, some currently available nitroimidazoles satisfy several of these prerequisites.
|
Pharmacological Properties of Nitroimidazoles |
|---|
Biochemical Pathways
Several nitroimidazole derivatives have been developed for imaging use (see supplemental Figure I, available online at http://stroke.ahajournals.org). Nitroimidazoles undergo selective bioreduction in hypoxic cells to form reactive products that irreversibly bind to cell components (Figure 1), a property exploited for their antibacterial action. The process is initiated by an enzyme-mediated single electron reduction to form a free radical. Under hypoxic conditions, this reduction is catalyzed by pyruvate (specific enzyme for this step not completely identified).8,10,11 Inside cells, the free radical anion is rapidly reversed to its original compound by intracellular oxygen, which has a higher electron affinity than the nitro group. The rate of oxidation is dependent on the intracellular concentration of oxygen. If tPO2 is low, subsequent steps ensure the irreversible retention of nitroimidazoles in the cells—a prerequisite to differentiating normoxic from hypoxic tissue. The cellular components that covalently bind the free radical-type nitroimidazoles have not been clearly identified, but might be intracellular macromolecules such as proteins or DNA.8 (See the supplemental Appendix for a note on the pharmacokinetics of nitroimidazoles).
|
|
Oxygen Dependency of Nitroimidazole Binding
The sensitivity of nitroimidazole imaging for detecting hypoxic tissue is determined by: (1) the amount of tracer delivered to the target site via CBF and the BBB; (2) the fraction that gets past the initial reversible enzyme-dependant reaction; (3) the washout from normal cells; (4) the rate of trapping in hypoxic tissue, which in turn determines the duration of the experiment sufficient for reaching adequate contrast between normoxic and hypoxic tissue; and (5) the presence of the intracellular macromolecular components necessary for binding.8
Much of the research done on the oxygen dependency of nitroimidazole binding concerned the radiosensitizer misonidazole.8 Studies using multi-cellular spheroids have shown that very little accumulation of misonidazole occurs in the tPO2 range 20 to 60 mm Hg but a steep rise is seen as this approaches 10 mm Hg.12 In healthy brain tissue in the mouse, the Km of misonidazole in vitro is 3.04 mm Hg,13 whereas in animal and human tumor cell lines, it varies between 0.76 and 4.56 mm Hg8,14,15—tPO2 levels relevant to stroke. However, the retention of nitroimidazoles varies widely with the type of tissue and pathology under study.16 Thus it may not be possible to generalize data to other nitroimidazoles, particularly in acute stroke, and investigation of the individual ligands under various conditions in stroke models is therefore necessary to establish their fitness for this application.
In the gerbil, misonidazole binding positively correlates with the severity of stroke symptoms17 6 to 10 hours after carotid ligation. In rodent and canine tumors,18,19 fluoromisonidazole (FMISO) appears to bind to tissue at similar tPO2 levels as misonidazole. In the ischemic porcine liver, Piert et al20 demonstrated a precise nonlinear inverse relationship between 18F-FMISO standardized uptake value (SUV) and tPO2 below 15 mm Hg (Figure 2) and were able to directly estimate the tPO2 from FMISO SUV. However, in murine and human soft tissue tumors, others did not replicate this relationship,21,22 possibly because of the tumoral nature of the tissue, tissue heterogeneities and necrosis in their selected tumor types, or the use of the polarographic microelectrode. Combining 18F-FMISO PET, deoxyglucose autoradiography, histological analysis and tPO2 measurements in murine tumors, high binding of FMISO was found (tumor-reference ratio of 11) in tumors with a median tPO2 of 7 mm Hg, in agreement with histological and autoradiographic patterns documenting cellular viability.16 Increased FMISO binding was also seen in tumors with a median tPO2 of 1 mm Hg, but the tumor-to-reference ratio was only 2, possibly reflecting cell death.
|
These results altogether indicate a close relationship between nitroimidazole binding and tPO2 in normal liver tissue and most tumors, but also point to a role for factors other than hypoxia in tracer uptake and binding, such as the type of tissue (eg, normal versus tumor), cell viability, and local metabolic and enzymatic patterns. These considerations are critical when considering extending the use of nitroimidazoles to stroke.
|
Applications of Nitroimidazoles in Hypoxia Imaging |
|---|
Experimental Stroke Studies
Considerable heterogeneity exists among experimental nitroimidazole stroke studies published to date with regard to stroke model, species, and hypoxia tracer used; time of assessment of tracer uptake after administration and arterial occlusion; method to assess and express tracer uptake; and use of ancillary techniques such as CBF. Furthermore, no experimental stroke study so far has validated nitroimidazole tissue uptake against concurrently obtained tPO2. This heterogeneity precludes any normalization, and hence combined analysis, across studies, and consequently a descriptive approach will be adopted here (see supplemental Table I
for details).
|
|
With only 2 exceptions,23,24 all of these studies have used ex vivo autoradiography. In gerbils,17 3H-misonidazole was shown to have good brain penetration and binding to hypoxic-ischemic tissue in correlation to stroke severity. Subsequently,25,26 a correlation to histological infarction was demonstrated, though direct comparisons could not be made. Trapping was consistently found in regions with reduced rCBF, yet a correlation to CBF was not ascertained. It was also shown that 3H-misonidazole trapping was not secondary to BBB leakage.26
One hour after permanent middle cerebral artery occlusion (pMCAO) in rats, the relative uptake of 99mTc-BMS-181321 in the most affected region was 6.4±2.2, corresponding to a mean 92% reduction of rCBF.24 However, comparison of uptake ratios to rCBF on a pixel-by-pixel basis along the edge of this region revealed a gradual increase of uptake from around 50 mL/100g/min (corresponding to the penumbra threshold in that model) and a rapid rise at lower rCBF levels near the infarction threshold (Figure 2). Importantly, when the tracer was administered 24 hours after pMCAO, no retention was seen in the area of infarction. These results imply that BMS-181321 only localized in areas of severe hypoxia, but not in areas of moderate hypoxia nor in irreversibly damaged tissue. In experiments on cats, BMS-181321 retention was detected by SPECT after 3 hours of MCAO24 and was further confirmed on ex vivo autoradiograms. Interestingly, early SPECT imaging (immediately postinjection), but not late imaging, appeared to closely approximate CBF.
125I-IAZA was subsequently used in rat pMCAO experiments.27,28 The first study involved visual inspection of double-label ex vivo autoradiograms of 125I-IAZA and the semiquantitative perfusion tracer 99mTc-HMPAO. In animals injected 2 hours after the start of occlusion and euthanized 7 hours later, uptake of IAZA was seen in regions of moderately decreased perfusion but was absent in regions of either slight or severe hypoperfusion, probably corresponding to oligemia and infarcted core, respectively, consistent with the above BMS-181321 studies. These findings were confirmed in their follow-up study using the same design plus diffusion-weighted imaging (DWI) at 2 hours after pMCAO.28 They reported four distinct regions: (1) "normal"; (2) "oligemic misery perfusion" showing mildly decreased perfusion (to 66% of normal), normal IAZA relative uptake (1.1) and normal ADC values; (3) "ischemic misery perfusion", showing CBF down to 34% of normal, increased IAZA uptake (2.1), and low ADC values (< 50% of normal); and (4) "core ischemia", with severe hypoperfusion (<7% of normal), normal or slightly reduced IAZA uptake (0.8), and further decreased ADC. The presumed "penumbra" indicated by increased IAZA uptake lay entirely within the area of restricted diffusion and was later found to be part of the infarct on histology. However, there was a region with very low CBF and ADC but no IAZA trapping, likely representing the core (ie, region 4; illustrated in their Figure 2), so the area of tracer trapping did not exactly match the DWI lesion. This lack of IAZA trapping despite probable ongoing hypoxia in Region 4 probably reflects damage to the key enzymes and proteins that underlie trapping in viable tissue and make the important point that hypoxia tracers, in conjunction with DWI, may allow the differentiation between the core (DWI lesion, no tracer trapping) and the penumbra (tracer trapping regardless of DWI). Although the finding that the area of IAZA trapping was embedded within the DWI lesion would agree with the notion that the DWI lesion includes areas of penumbra,29 it is surprising that IAZA uptake was not also seen in the surrounds, because in principle the DWI lesion does not contain the whole penumbra. However, IAZA was injected 2 hours after pMCAO, by which time the entire penumbra perhaps manifested high DWI signal. This later time of administration than in the BMS-181321 study24 probably accounts for the different behavior of the tracers at very low blood flows, ie, in the BMS study the core still demonstrated maintained cellular machinery for hypoxia tracer binding, whereas viability was probably lost in the IAZA study.28 Finally, both the BMS and the IAZA studies seem to concur that hypoxia tracers do not show trapping in mildly hypoperfused tissues (ie, the oligemia), although close inspection of the data in Figures 3 and 4 of the IAZA article28 suggests a mild but definite gradual tracer retention for perfusion values above the penumbra threshold. Nevertheless, both studies agree that nitroimidazoles will show substantial retention in the moderately to severely hypoperfused tissue, provided it is still viable.
|
|
Recently, Saita et al30 reported on 18F-FMISO ex vivo autoradiography in rats subjected to 2-hour MCAO. Rats were injected at varying time points from occlusion (0.5 to 22 hour) and euthanized 2 hour later, and tracer retention was compared to triphenyltetrazolium chloride (TTC) postmortem staining. Consistent with being selectively retained in ischemic tissue, FMISO uptake was seen in large areas within the MCA territory at prereperfusion time points. These became progressively smaller and more localized to the edges of the infarct over time. Only a thin rim of uptake surrounding the TTC-determined infarct was seen at the late time points. All animals showed reduced 18F-FMISO uptake at the core of the infarct. Nonetheless, the finding that 18F-FMISO was retained in periinfarct tissue up to 20 hours despite reversal of MCAO was intriguing. As possible explanations, the authors suggested periinfarct edema, small vessel occlusion, or "no-reflow"31 may have resulted in persistent tissue hypoxia. Furthermore, the thread-up MCAO method used in that study can induce vascular injury or rupture during withdrawal of the thread, and thus may have precluded effective reperfusion,32 which would also account for the TTC-determined infarct growing until the last time point. That retention occurred in the core also cannot be completely excluded because superimposing the autoradiograms and the TTC sections was not technically feasible. This study raised the disconcerting possibility that FMISO could be trapped in normoxic periinfarct or in hypoxic nonsalvageable tissue, in contrast to the above BMS and IAZA studies. However, further experiments dispelled these concerns.33 Using an improved thread-up technique and tracer injection either during the occlusion or 1 hour after 45 to 90 minutes MCAO, 3H-FMISO binding occurred only in ongoing tissue hypoxia and not if it was administered after reperfusion. Interestingly, the long half-life of 3H enabled direct comparison between early tissue hypoxia and histology done 24 hours later.
Overall, the above (mostly ex vivo) experimental studies support the notion that nitroimidazoles can specifically detect the viable severely hypoxic tissue in acute stroke, but not the mildly hypoperfused tissue nor the severely hypoperfused and necrotic core, although the latter point remains incompletely solved. An important limitation is that direct measurement of tPO2 was not performed in any of these studies. Furthermore, CBF was not systematically quantified in conjunction with the hypoxia tracer, so the CBF threshold for significant tracer uptake has not been fully clarified.
We have successfully used animal PET to study 18F-FMISO in the rat distal MCAO model.23 In pMCAO conditions, 18F-FMISO was injected around 30 minutes after occlusion and dynamic scanning performed for 3 hours; a follow-up PET study was performed 48 hours later (all the 18F from the first study would have decayed by then), at a time when pan-necrosis was expected. In the acute stage, there was continuous uptake of the tracer in the affected MCA cortical territory, reaching more than twice that of the unaffected side at 3 hours postinjection (Figure 3). No significant retention was found on the follow-up scan, which corresponded to frank tissue infarction at histology, nor when the tracer was administrated immediately after 45 or 120 minutes MCAO. These results were confirmed by modeling of the full plasma and tissue FMISO kinetics. The findings from this study using in vivo imaging are therefore largely consistent with the above ex vivo studies, further supporting the notion that 18F-FMISO is trapped only in hypoxic viable tissue, but not in necrotic or reperfused tissue, potentially making it a specific marker for the salvageable penumbra. Follow-up PET at shorter intervals after pMCAO are however needed to address whether early irreversibly damaged but still hypoxic tissue still binds F-MISO.
Clinical Studies
Nitroimidazoles have been used in the detection of hypoxia in humans in stroke as well as in a number of other fields, particularly in tumors (see supplemental Appendix for details).
Supplemental Table II tabulates all the nitroimidazole studies performed in acute stroke so far. In the first study, reported in abstract form,34 3 of 6 patients presenting with "acute" ischemic stroke (no further detail provided) showed regions of intense 18F-FMISO retention surrounding the core of the infarction on PET. This presumed "penumbra" had disappeared when the patients were rescanned 1 month later.
|
|
Also in a conference abstract, 99mTc-BMS-181321 SPECT was used in 3 patients presenting with subacute stroke (11 to 38 hours).35 Only 1 patient, imaged within 12 hours, demonstrated a region of (moderately) increased uptake (lesion-to-contralateral ratio 1.33) located in close proximity to the infarct seen on CT at 36 hours, which could have represented penumbral tissue. The other 2 patients, scanned at 22.5 and 38.8 hours, had negative studies. Work with this tracer has not, to our knowledge, been pursued.
99mTc-ethylene dicysteine-metronidazole (99mTc-EC MN) SPECT has been tested in 8 patients scanned in the late subacute stage (6 to 14 days after onset).36 Surprisingly, increased periinfarct uptake (lesion-to-normal ratios 3.5±1.5) was reported in all the subjects. Because scanning was done so late, this tissue is unlikely to represent the penumbra, though may have represented persistent oligemia if ongoing vascular occlusion was present at the time of imaging (information not reported). Validation against established PET techniques is needed to confirm these results.
The Melbourne group have extensively investigated the potential of 18F-FMISO for depicting the penumbra,37–43 using a similar methodology whereby PET scanning was performed at least 2 hours after tracer injection. They first reported definite FMISO retention in the first 48 hours after stroke onset, which subsequently disappeared on follow-up imaging.37 This tissue shared the topographical and temporal characteristics of the penumbra. FMISO retention was based on selecting voxels with uptake 
3SD above contralateral hemisphere mean, a conservative threshold that may have excluded lesser but potentially important tracer trapping. Similar approaches were used in all subsequent studies.
In a second set of patients, the results further suggested that tissue defined by high FMISO uptake represented the ischemic penumbra because: (1) both the volume of hypoxic tissue and the frequency of a positive study declined with time after stroke onset; (2) these areas comprised part of the final infarct volume; and (3) the initial severity of stroke correlated with the "initially affected" tissue volume, defined as the sum of the final infarct and the hypoxic tissue eventually surviving infarction.38 However, they also reported that if a high proportion of the initially affected volume proceeded to infarction, neurological deterioration was likely. This is somewhat unexpected, because if the hypoxic tissue only represented the penumbra, then its progression to infarction would not in principle contribute to clinical worsening. A possible explanation is that the area defined by FMISO PET may have not been fully symptomatic at the time of imaging because it represented oligemic rather than penumbral tissue44—which could also account for the observation of hypoxic tissue up to 43 hours after stroke onset.
In later studies, statistical parametric and nonparametric mapping was used to identify voxels with increased FMISO uptake relative to normal controls, allowing the 3D extent of hypoxic tissue (so-called penumbragram) to be mapped.39,40,41 This tissue was found predominantly superior, mesial and posterior to the center of the final infarct. Furthermore, this tissue was preferentially seen in the center of the final infarct in patients studied <6 hour from onset, while it was mostly peripheral or external to it at later times. These findings were consistent with the known topography of the penumbra and the notion of peripheral infarct expansion.4 Although of clear interest, this method assumes that the geographic center of the infarct corresponds to the true core of infarction, which seems only true for MCA territory infarcts with a convex surface. Validation of the usefulness of the penumbragram is necessary before it is accepted as a clinical or research tool.
A further study on an enlarged series of 27 patients with positive FMISO scans up to 48 hour after onset42 showed that 
46% of the initially hypoxic tissue (7% of the total ischemic volume) survived infarction spontaneously. The chance of tissue survival was independent of time of onset, consistent with the concept of the penumbra. Furthermore, hypoxic tissue survival was associated with clinical improvement, even for the group studied beyond 12 hours, suggesting that it contributed to the deficit. This satisfies an essential operational criterion defining the penumbra45 and further confirms the presence of potentially salvageable tissue >12 hours in a subset of anterior circulation stroke patients.
Overall, therefore, the above seminal FMISO studies are consistent with the animal studies described earlier in that trapping after ischemic stroke occurs in severely hypoxic but viable tissue, whereas its disappearance within 48 hours of onset suggests that necrotic or reperfused tissues do not retain FMISO. Figure 4 illustrates some of these points in 2 patients, 1 showing complete MCA occlusion with marked 18F-FMISO trapping overlapping the MR DWI lesion and extending beyond it to a lesser degree, whereas in the other there is no trapping in the region of the DWI lesion after partial recanalization of the MCA, with only mild trapping posterior to it.
Differences in vulnerability to ischemia between white and gray matter have also been addressed using 18F-FMISO.46 White matter appeared to have a higher proportion of hypoxic tissue than gray at the time of imaging, though the volumes of hypoxic tissue eventually surviving infarction were similar in both compartments. This may indirectly suggest that white matter was initially more resistant to ischemia and proceeded more slowly to infarction than gray matter. Several studies have recently echoed these observations using MR imaging,47–49 yet the results should be considered with the view that: (1) binding of 18F-FMISO was assumed to be independent of tissue type; (2) it was assumed that the observed differences in survival were not the manifestation of differential blood flow to each of those compartments; and (3) the classification of voxels as white or gray matter was based on probabilistic maps generated from healthy controls, rather than on the subjects’ own scans.
|
Discussion |
|---|
Imaging the viable hypoxic tissue in acute stroke is a highly desirable goal. Nitroimidazoles are the prime contenders among imaging tracers as they fulfill the main pharmacological and clinical prerequisites, and have been successfully used in imaging hypoxia in animal stroke models and in humans. Animal studies, including one using PET, generally support the idea that nitroimidazoles have rather high sensitivity and specificity for viable hypoxic tissue. Imaging studies in humans using 18F-FMISO PET also indicate that tracer trapping occurs mostly in viable hypoxic tissue that satisfies the operational criteria for the penumbra and matches its spatial pattern and temporal evolution.
Despite these clear potentials, nitroimidazoles have several significant limitations as in vivo clinical imaging tracers. The main limitation is their slow tissue kinetics, requiring for instance at least 2-hour equilibrium for 18F-FMISO before scanning is commenced, which precludes their use for clinical decision-making. The long scanning time is also incongruous with the dynamics of cerebral ischemia such that the resulting images represent the entire time epoch since injection, during which the hemodynamic and metabolic state of the tissue may have evolved (eg, demise of the penumbra, reperfusion, etc). Furthermore, as FMISO is not readily metabolized in the human body and remains available for tissue uptake for many hours after injection, areas with even mild hypoxia may over time show increased uptake. Development of novel ligands with faster distribution kinetics and clearance from normal tissue than 18F-FMISO would therefore be paramount. The sugar-coupled tracer 18F-FAZA has not been tested yet in stroke, but data in animals50–52 indicate it has fast clearance from normal tissues and might be suitable in providing good contrast in a reasonable time-frame. In this regard, copper-labeled compounds such as 60Cu-, 62Cu, and 64Cu-ATSM have been investigated in tumor hypoxia and also appear promising.53–55 Nonnitroimidazole compounds with more than one bioreductive site may also prove useful, as again these would require less time to bind to the target tissue.56 In addition to improved tracers, shortening of scanning time could be achievable if quantitative modeling allowed rate constants to be estimated reliably with short data collection (see below).
Another limitation derives from the almost exclusive use of PET in acute stroke studies so far. PET is an expensive, logistically complex, and scarcely available tool, making the study of acute stroke patients challenging. SPECT hypoxia ligands would therefore be desirable. Iodinated derivatives labeled with 123I, such as IAZA,27,28,57–59 and perhaps also 99mTc-BMS-181321,24,35 appear promising. On the other hand, the increasingly wider availability of PET in general hospitals for oncology diagnosis may have unexpected implications for hypoxia imaging in stroke.
Formal validation of nitroimidazole trapping against direct tPO2 measurements in the ischemic brain is still lacking. The only formal validation against concurrently measured tPO2 to date refers to the liver only (Figure 2). Knowing the quantitative relationship between nitroimidazole trapping and tPO2 should in turn allow the mapping of tPO2 in stroke.
Another limitation of nitroimidazole imaging agents refers to their somewhat ill-defined trapping mechanisms, with major gaps remaining in our knowledge of the successive steps leading to intracellular retention in situations of hypoxia, particularly the nature of the macromolecules that covalently bind reduced nitroimidazoles. Enhanced knowledge of these mechanisms would lead to improved interpretation of findings and better compartmental kinetic modeling.
An intriguing issue regarding the use of nitroimidazole imaging in stroke refers to the relationship between trapping and the DWI lesion. In the rat IAZA study detailed above,28 tracer trapping was roughly similar in topography and size to the DWI lesion. If nitroimidazole trapping equated the DWI lesion, then MR-DWI would be sufficient to map the hypoxic area in acute stroke without recourse to complex and time-consuming radiotracer-based investigations. However, in that study there was a region (denoted Region 4) that demonstrated very low ADC but no IAZA trapping, likely representing the necrotic core. The area of tracer trapping therefore did not match the DWI lesion, and hypoxia tracers in conjunction with DWI may in fact allow the differentiation between the true core and penumbral tissue. Furthermore, the area of IAZA trapping was embedded within the DWI lesion, which conflicts with the established notion that the latter is not supposed to contain the whole of the penumbra, and this surprising finding was probably attributable to the late timing of the assessment after MCAO. More generally, the DWI lesion is a marker of cellular dysfunction and develops consequent to, but is not a direct marker of, hypoxia. Accordingly, after MCAO: (1) the DWI lesion takes several minutes to develop, whereas hypoxia is immediate; (2) it expands over time in the face of stable ischemia, reflecting progressive metabolic dysfunction; and (3) it generally persists after reperfusion.29 Figure 4 illustrates the clear dissociation between the DWI lesion and the presence of FMISO trapping.
A key issue that remains in the interpretation of nitroimidazole imaging in acute stroke is whether the increased uptake only reflects the viable hypoxic tissue, ie, the penumbra. As detailed in previous sections, the available experimental evidence supports this idea, yet a study28 raised doubts that trapping might also occur in the ischemic core. This is theoretically plausible if the tissue, although inevitably destined to proceed to infarction, still includes cells that are nonnecrotic and consume oxygen at levels higher than oxygen delivery.60 Thus possibly the very high tracer retention seen in Figure 4 (top row) overlying the DWI lesion may have represented core rather than penumbra. It would still need to be established whether tracer retention in such tissue is attributable to slower wash-in and wash-out of the tracer over the scanning period rather than true active uptake. Experimental and clinical studies combining FMISO and multi-tracer 15O PET are needed to elucidate these points—although implementing this sort of studies is challenging.
Likewise, whether the oligemic tissue may demonstrate nitroimidazole trapping is unclear. All the methods used thus far for depicting significant tracer trapping result in a binary classification into positive and negative voxels—often applying conservative thresholds—which differs from more conventional multiparameter classification.4 These dichotomous images lack information on the severity—as opposed to presence—of tracer trapping, and hence of hypoxia. It is possible that less marked, yet pathophysiologically meaningful, levels of tracer trapping exist surrounding areas of stringently significant trapping, beyond simple partial voluming effects. Consistent with this idea, both the BMS52,67 and the IAZA rat studies28 in fact suggest the presence of mild tracer retention at CBF values above the penumbra threshold (see Figure 2). Quantifying the irreversible binding rate constant k3 using compartmental modeling might resolve this issue in the future. Again, directly comparing nitroimidazole binding to 15O-PET would help clarify this point.
As alluded to above, deriving quantitative rate constants for nitroimidazole binding, instead of currently used uptake ratios, may also allow for shortening of the required imaging time. Although there are some reports from oncology studies (see supplemental Appendix), there is still no complete model that takes into consideration all the chemical characteristics of nitroimidazoles, including any influence of their labeled metabolites. The optimum number of compartments and rate constants necessary to describe the observed tissue kinetics in normal and hypoxic conditions remains uncertain. Kinetic modeling may also allow generation of the initial entry of the tracer into tissue, which is proportional to perfusion if BBB passage is normal. Thus, maps of both relative perfusion and tissue hypoxia may be obtained from a single tracer administration, which would be of considerable interest.
|
Conclusion |
|---|
It is hoped that eventually better ligands, together with a better understanding and full quantification of hypoxia imaging and its translation to widely available technologies, will allow its wider application in clinical research and potentially also in clinical routine. Beyond identification of the penumbra, nitroimidazole imaging may also prove useful in monitoring treatments directed at tackling hypoxia, such as normobaric oxygen therapy.
|
Acknowledgments |
|---|
The authors are grateful to the staff at the Wolfson Brain Imaging Centre (University of Cambridge).
Sources of Funding
M.T. was supported by Japan Society for the Promotion of Science (JSPS) Post Doctoral Fellowships for Research Abroad at University of Cambridge (17-307). R.R.M. is supported by a grant from the Cambridge Overseas Trust, UK. The work on hypoxia imaging was supported by Medical Research Council (MRC) Grant G0001219 to J.C.B.
Disclosures
None.
Received February 21, 2007; revision received August 30, 2007; accepted September 21, 2007.
|
References |
|---|
1. Heiss WD. Ischemic penumbra: Evidence from functional imaging in man. J Cereb Blood Flow Metab. 2000; 20: 1276–1293.[CrossRef][Medline] [Order article via Infotrieve]
2. Baron JC. Perfusion thresholds in human cerebral ischemia: Historical perspective and therapeutic implications. Cerebrovasc Dis. 2001; 11 Suppl 1: 2–8.[Medline] [Order article via Infotrieve]
3. Jones TH, Morawetz RB, Crowell RM, Marcoux FW, FitzGibbon SJ, DeGirolami U, Ojemann RG. Thresholds of focal cerebral ischemia in awake monkeys. J Neurosurg. 1981; 54: 773–782.[Medline] [Order article via Infotrieve]
4. Baron JC. Mapping the ischaemic penumbra with PET: Implications for acute stroke treatment. Cerebrovasc Dis. 1999; 9: 193–201.[CrossRef][Medline] [Order article via Infotrieve]
5. Alpert NM, Buxton RB, Correia JA, Katz PM, Ackerman RH. Measurement of end-capillary po2 with positron emission tomography. J Cereb Blood Flow Metab. 1988; 8: 403–410.[Medline] [Order article via Infotrieve]
6. Gupta AK, Hutchinson PJ, Fryer T, Al-Rawi PG, Parry DA, Minhas PS, Kett-White R, Kirkpatrick PJ, Mathews JC, Downey S, Aigbirhio F, Clark J, Pickard JD, Menon DK. Measurement of brain tissue oxygenation performed using positron emission tomography scanning to validate a novel monitoring method. J Neurosurg. 2002; 96: 263–268.[Medline] [Order article via Infotrieve]
7. Ballinger JR. Imaging hypoxia in tumors. Semin Nucl Med. 2001; 31: 321–329.[CrossRef][Medline] [Order article via Infotrieve]
8. Nunn A, Linder K, Strauss HW. Nitroimidazoles and imaging hypoxia. Eur J Nucl Med. 1995; 22: 265–280.[CrossRef][Medline] [Order article via Infotrieve]
9. Hossmann KA. Non-invasive imaging methods for the characterization of the pathophysiology of brain ischemia. Acta Neurochir Suppl. 2003; 86: 21–27.[Medline] [Order article via Infotrieve]
10. Joseph P, Jaiswal AK, Stobbe CC, Chapman JD. The role of specific reductases in the intracellular activation and binding of 2-nitroimidazoles. Int J Radiat Oncol Biol Phys. 1994; 29: 351–355.[Medline] [Order article via Infotrieve]
11. Kedderis GL, Miwa GT. The metabolic activation of nitroheterocyclic therapeutic agents. Drug Metab Rev. 1988; 19: 33–62.[Medline] [Order article via Infotrieve]
12. Gross MW, Karbach U, Groebe K, Franko AJ, Mueller-Klieser W. Calibration of misonidazole labeling by simultaneous measurement of oxygen tension and labeling density in multicellular spheroids. Int J Cancer. 1995; 61: 567–573.[Medline] [Order article via Infotrieve]
13. Van Os-Corby DJ, Koch CJ, Chapman JD. Is misonidazole binding to mouse tissues a measure of cellular po2? Biochem Pharmacol. 1987; 36: 3487–3494.[CrossRef][Medline] [Order article via Infotrieve]
14. Franko AJ, Koch CJ, Garrecht BM, Sharplin J, Hughes D. Oxygen dependence of binding of misonidazole to rodent and human tumors in vitro. Cancer Res. 1987; 47: 5367–5376.
15. Franko AJ. Misonidazole and other hypoxia markers: Metabolism and applications. Int J Radiat Oncol Biol Phys. 1986; 12: 1195–1202.[Medline] [Order article via Infotrieve]
16. Sorensen M, Horsman MR, Cumming P, Munk OL, Keiding S. Effect of intratumoral heterogeneity in oxygenation status on fmiso pet, autoradiography, and electrode po2 measurements in murine tumors. Int J Radiat Oncol Biol Phys. 2005; 62: 854–861.[CrossRef][Medline] [Order article via Infotrieve]
17. Rasey JS, Hoffman JM, Spence AM, Krohn KA. Hypoxia mediated binding of misonidazole in non-malignant tissue. Int J Radiat Oncol Biol Phys. 1986; 12: 1255–1258.[Medline] [Order article via Infotrieve]
18. Rasey JS, Grunbaum Z, Magee S, Nelson NJ, Olive PL, Durand RE, Krohn KA. Characterization of radiolabeled fluoromisonidazole as a probe for hypoxic cells. Radiat Res. 1987; 111: 292–304.[CrossRef][Medline] [Order article via Infotrieve]
19. Rasey JS, Koh WJ, Grierson JR, Grunbaum Z, Krohn KA. Radiolabelled fluoromisonidazole as an imaging agent for tumor hypoxia. Int J Radiat Oncol Biol Phys. 1989; 17: 985–991.[Medline] [Order article via Infotrieve]
20. Piert M, Machulla HJ, Becker G, Aldinger P, Winter E, Bares R. Dependency of the [18f]fluoromisonidazole uptake on oxygen delivery and tissue oxygenation in the porcine liver. Nucl Med Biol. 2000; 27: 693–700.[CrossRef][Medline] [Order article via Infotrieve]
21. Bentzen L, Keiding S, Nordsmark M, Falborg L, Hansen SB, Keller J, Nielsen OS, Overgaard J. Tumour oxygenation assessed by 18f-fluoromisonidazole pet and polarographic needle electrodes in human soft tissue tumours. Radiother Oncol. 2003; 67: 339–344.[CrossRef][Medline] [Order article via Infotrieve]
22. Bentzen L, Keiding S, Horsman MR, Gronroos T, Hansen SB, Overgaard J. Assessment of hypoxia in experimental mice tumours by [18f]fluoromisonidazole pet and po2 electrode measurements. Influence of tumour volume and carbogen breathing. Acta Oncol. 2002; 41: 304–312.[CrossRef][Medline] [Order article via Infotrieve]
23. Takasawa M, Beech JS, Fryer TD, Hong YT, Hughes JL, Igase K, Jones PS, Smith R, Aigbirhio FI, Menon DK, Clark JC, Baron JC. Imaging of brain hypoxia in permanent and temporary middle cerebral artery occlusion in the rat using 18f-fluoromisonidazole and positron emission tomography: A pilot study. J Cereb Blood Flow Metab. 2007; 27: 679–689.[Medline] [Order article via Infotrieve]
24. Di Rocco RJ, Kuczynski BL, Pirro JP, Bauer A, Linder KE, Ramalingam K, Cyr JE, Chan YW, Raju N, Narra RK, et al. Imaging ischemic tissue at risk of infarction during stroke. J Cereb Blood Flow Metab. 1993; 13: 755–762.[Medline] [Order article via Infotrieve]
25. Hoffman JM, Rasey JS, Spence AM, Shaw DW, Krohn KA. Binding of the hypoxia tracer [3h]misonidazole in cerebral ischemia. Stroke. 1987; 18: 168–176.
26. Mathias CJ, Welch MJ, Kilbourn MR, Jerabek PA, Patrick TB, Raichle ME, Krohn KA, Rasey JS, Shaw DW. Radiolabeled hypoxic cell sensitizers: Tracers for assessment of ischemia. Life Sci. 1987; 41: 199–206.[CrossRef][Medline] [Order article via Infotrieve]
27. Lythgoe MF, Williams SR, Wiebe LI, McEwan AJ, Gordon I. Autoradiographic imaging of cerebral ischaemia using a combination of blood flow and hypoxic markers in an animal model. Eur J Nucl Med. 1997; 24: 16–20.[CrossRef][Medline] [Order article via Infotrieve]
28. Lythgoe MF, Williams SR, Busza AL, Wiebe L, McEwan AJ, Gadian DG, Gordon I. The relationship between magnetic resonance diffusion imaging and autoradiographic markers of cerebral blood flow and hypoxia in an animal stroke model. Magn Reson Med. 1999; 41: 706–714.[CrossRef][Medline] [Order article via Infotrieve]
29. Guadagno JV, Warburton EA, Jones PS, Fryer TD, Day DJ, Gillard JH, Carpenter TA, Aigbirhio FI, Price CJ, Baron JC. The diffusion-weighted lesion in acute stroke: Heterogeneous patterns of flow/metabolism uncoupling as assessed by quantitative positron emission tomography. Cerebrovasc Dis. 2005; 19: 239–246.[CrossRef][Medline] [Order article via Infotrieve]
30. Saita K, Chen M, Spratt NJ, Porritt MJ, Liberatore GT, Read SJ, Levi CR, Donnan GA, Ackermann U, Tochon-Danguy HJ, Sachinidis JI, Howells DW. Imaging the ischemic penumbra with 18f-fluoromisonidazole in a rat model of ischemic stroke. Stroke. 2004; 35: 975–980.
31. del Zoppo GJ, Mabuchi T. Cerebral microvessel responses to focal ischemia. J Cereb Blood Flow Metab. 2003; 23: 879–894.[CrossRef][Medline] [Order article via Infotrieve]
32. Spratt NJ, Fernandez J, Chen M, Rewell S, Cox S, van Raay L, Hogan L, Howells DW. Modification of the method of thread manufacture improves stroke induction rate and reduces mortality after thread-occlusion of the middle cerebral artery in young or aged rats. J Neurosci Methods. 2006; 155: 285–290.[CrossRef][Medline] [Order article via Infotrieve]
33. Spratt NJ, Ackerman U, Tochon-Danguy HJ, Donnan GA, Howells DW. Characterization of fluoromisonidazole binding in stroke. Stroke. 2006; 37: 1862–1867.
34. Yeh SH, Liu RS, Hu HH, et al. Ischemic penumbra in acute stroke: Demonstration by pet with fluorine-18 fluoromisonidazole. J Nucl Med. 1994; 35: 205P.
35. Barron B, Grotta JC, Lamki L, Villar C, Ephron V, Patel D. Preliminary experience with technetium-99m bms-181231, a nitroimidazole, in the detection of cerebral ischemia associated with acute stroke. J Nucl Med. 1996; 37: 272P–273P
36. Song HC, Bom HS, Cho KH, Kim BC, Seo JJ, Kim CG, Yang DJ, Kim EE. Prognostication of recovery in patients with acute ischemic stroke through the use of brain spect with technetium-99m–labeled metronidazole. Stroke. 2003; 34: 982–986.
37. Read SJ, Hirano T, Abbott DF, Sachinidis JI, Tochon-Danguy HJ, Chan JG, Egan GF, Scott AM, Bladin CF, McKay WJ, Donnan GA. Identifying hypoxic tissue after acute ischemic stroke using pet and 18f-fluoromisonidazole. Neurology. 1998; 51: 1617–1621.
38. Read SJ, Hirano T, Abbott DF, Markus R, Sachinidis JI, Tochon-Danguy HJ, Chan JG, Egan GF, Scott AM, Bladin CF, McKay WJ, Donnan GA. The fate of hypoxic tissue on 18f-fluoromisonidazole positron emission tomography after ischemic stroke. Ann Neurol. 2000; 48: 228–235.[CrossRef][Medline] [Order article via Infotrieve]
39. Markus R, Donnan GA, Kazui S, Read S, Hirano T, Scott AM, O’Keefe GJ, Tochon-Danguy HJ, Sachinidis JI, Reutens DC. Statistical parametric mapping of hypoxic tissue identified by [(18)f]fluoromisonidazole and positron emission tomography following acute ischemic stroke. Neuroimage. 2002; 16: 425–433.[CrossRef][Medline] [Order article via Infotrieve]
40. Markus R, Reutens DC, Kazui S, Read S, Wright P, Chambers BR, Sachinidis JI, Tochon-Danguy HJ, Donnan GA. Topography and temporal evolution of hypoxic viable tissue identified by 18f-fluoromisonidazole positron emission tomography in humans after ischemic stroke. Stroke. 2003; 34: 2646–2652.
41. Markus R, Donnan G, Kazui S, Read S, Reutens D. Penumbral topography in human stroke: Methodology and validation of the ‘penumbragram’. Neuroimage. 2004; 21: 1252–1259.[CrossRef][Medline] [Order article via Infotrieve]
42. Markus R, Reutens DC, Kazui S, Read S, Wright P, Pearce DC, Tochon-Danguy HJ, Sachinidis JI, Donnan GA. Hypoxic tissue in ischaemic stroke: Persistence and clinical consequences of spontaneous survival. Brain. 2004; 127: 1427–1436.
43. Hirano T, Read SJ, Abbott DF, Sachinidis JI, Tochon-Danguy HJ, Egan GF, Bladin CF, Scott AM, McKay WJ, Donnan GA. No evidence of hypoxic tissue on 18f-fluoromisonidazole pet after intracerebral hemorrhage. Neurology. 1999; 53: 2179–2182.
44. Baron JC. Mapping the ischaemic penumbra with PET: A new approach. Brain. 2001; 124: 2–4.
45. Furlan M, Marchal G, Viader F, Derlon JM, Baron JC. Spontaneous neurological recovery after stroke and the fate of the ischemic penumbra. Ann Neurol. 1996; 40: 216–226.[CrossRef][Medline] [Order article via Infotrieve]
46. Falcao AL, Reutens DC, Markus R, Koga M, Read SJ, Tochon-Danguy H, Sachinidis J, Howells DW, Donnan GA. The resistance to ischemia of white and gray matter after stroke. Ann Neurol. 2004; 56: 695–701.[CrossRef][Medline] [Order article via Infotrieve]
47. Koga M, Reutens DC, Wright P, Phan T, Markus R, Pedreira B, Fitt G, Lim I, Donnan GA. The existence and evolution of diffusion-perfusion mismatched tissue in white and gray matter after acute stroke. Stroke. 2005; 36: 2132–2137.
48. Bristow MS, Simon JE, Brown RA, Eliasziw M, Hill MD, Coutts SB, Frayne R, Demchuk AM, Mitchell JR. MR perfusion and diffusion in acute ischemic stroke: Human gray and white matter have different thresholds for infarction. J Cereb Blood Flow Metab. 2005; 25: 1280–1287.[CrossRef][Medline] [Order article via Infotrieve]
49. Simon JE, Bristow MS, Lu H, Lauzon ML, Brown RA, Manjón JV, Eliasziw M, Frayne R, Buchan AM, Demchuk AM, Mitchell JR; Calgary Stroke Program. A novel method to derive separate gray and white matter cerebral blood flow measures from mr imaging of acute ischemic stroke patients. J Cereb Blood Flow Metab. 2005; 25: 1236–1243.[CrossRef][Medline] [Order article via Infotrieve]
50. Bruehlmeier M, Roelcke U, Schubiger PA, Ametamey SM. Assessment of hypoxia and perfusion in human brain tumors using pet with 18f-fluoromisonidazole and 15o-h2o. J Nucl Med. 2004; 45: 1851–1859.
51. Piert M, Machulla HJ, Picchio M, Reischl G, Ziegler S, Kumar P, Wester HJ, Beck R, McEwan AJ, Wiebe LI, Schwaiger M. Hypoxia-specific tumor imaging with 18f-fluoroazomycin arabinoside. J Nucl Med. 2005; 46: 106–113.
52. Sorger D, Patt M, Kumar P, Wiebe LI, Barthel H, Seese A, Dannenberg C, Tannapfel A, Kluge R, Sabri O. [18f]fluoroazomycinarabinofuranoside (18faza) and [18f]fluoromisonidazole (18fmiso): A comparative study of their selective uptake in hypoxic cells and pet imaging in experimental rat tumors. Nucl Med Biol. 2003; 30: 317–326.[CrossRef][Medline] [Order article via Infotrieve]
53. Obata A, Yoshimi E, Waki A, Lewis JS, Oyama N, Welch MJ, Saji H, Yonekura Y, Fujibayashi Y. Retention mechanism of hypoxia selective nuclear imaging/radiotherapeutic agent cu-diacetyl-bis(n4-methylthiosemicarbazone) (cu-atsm) in tumor cells. Ann Nucl Med. 2001; 15: 499–504.[Medline] [Order article via Infotrieve]
54. Takahashi N, Fujibayashi Y, Yonekura Y, Welch MJ, Waki A, Tsuchida T, Sadato N, Sugimoto K, Itoh H. Evaluation of 62cu labeled diacetyl-bis(n4-methylthiosemicarbazone) as a hypoxic tissue tracer in patients with lung cancer. Ann Nucl Med. 2000; 14: 323–328.[Medline] [Order article via Infotrieve]
55. Yuan H, Schroeder T, Bowsher JE, Hedlund LW, Wong T, Dewhirst MW. Intertumoral differences in hypoxia selectivity of the pet imaging agent 64cu(ii)-diacetyl-bis(n4-methylthiosemicarbazone). J Nucl Med. 2006; 47: 989–998.
56. Falzon CL, Ackermann U, Spratt N, et al. F-18 labelled n,n-bis-haloethylamino-phenylsulfoxides - a new class of compounds for the imaging of hypoxic tissue. J Label Comp Radiopharm. 2006; 49: 1089–1103.[CrossRef]
57. Parliament MB, Chapman JD, Urtasun RC, McEwan AJ, Golberg L, Mercer JR, Mannan RH, Wiebe LI. Non-invasive assessment of human tumour hypoxia with 123i-iodoazomycin arabinoside: Preliminary report of a clinical study. Br J Cancer. 1992; 65: 90–95.[Medline] [Order article via Infotrieve]
58. Stypinski D, McQuarrie SA, Wiebe LI, Tam YK, Mercer JR, McEwan AJ. Dosimetry estimations for 123i-iaza in healthy volunteers. J Nucl Med. 2001; 42: 1418–1423.
59. Vinjamuri S, O’Driscoll K, Maltby P, McEwan AJ, Wiebe LI, Critchley M. Identification of hypoxic regions in traumatic brain injury. Clin Nucl Med. 1999; 24: 891–892.[CrossRef][Medline] [Order article via Infotrieve]
60. Baron JC. Stroke research in the modern era: Images versus dogmas. Cerebrovasc Dis. 2005; 20: 154–163.[CrossRef][Medline] [Order article via Infotrieve]
Supplemental Appendix
|
Direct Measurement of Brain Tissue Partial Pressure of Oxygen |
|---|
Local tPO2 can be measured directly with oxygen-sensitive probes, using either: (1) the classic polarographic (Eppendorf) microelectrode, which samples tPO2 only in the immediate vicinity of the probe tip, induces local tissue trauma and cannot differentiate necrotic from viable tissue1–3; or (2) electron paramagnetic resonance (EPR) oximetry, which has recently been used in experimental stroke research4,5 and involves stereotaxic intraparenchymal implantation of oxygen-sensitive paramagnetic crystals and allows repeated, continuous and highly accurate measurements of tPO2. However, measurement of tPO2 using either method has rarely been feasible in stroke patients because of their invasiveness. Less invasive methods, such as optical techniques, still have in their present versions limited resolution, restricted coverage, and do not reliably separate cerebral from extracerebral tissue to be utilized clinically.6
A few studies have assessed tPO2 in experimental stroke4,5,7–9 with quite consistent results. For instance, after middle cerebral artery occlusion (MCAO) in the baboon,7 tPO2 fell rapidly from 24 mm Hg to 3.6±5.9 mm Hg in the ischemic core, with a gradient of tissue hypoxia following the gradation of CBF. In the parietal areas, however, tPO2 was less affected than predicted by CBF, suggesting that above a certain flow threshold, tPO2 is possibly sustained by mechanisms such as increasing oxygen dissociation from hemoglobin and/or reduction of oxygen utilization.9 On reperfusion, tPO2 recovered in the core to 80% of control levels, whereas in the parietal area there was overshoot of tPO2 above the normal values, ie, tissue hyperoxia. In cats with MCAO, Farrar8 reported a constant tPO2 threshold for infarction at around 7 to 8 mm Hg over 6 hours, whereas as expected, the CBF threshold for infarction was dependant on time.
In patients undergoing temporary cerebral artery occlusion, Hoffman et al1011 reported that intraoperative tPO2 immediately fell to below 10 mm Hg and occasionally approached zero, as compared to 30 to 40 mm Hg in normal tissue. Nevertheless, none of the patients in that group subsequently developed cerebral infarcts on computed tomography, possibly because of a protective effect from anesthesia and the short duration of ischemia. Hence, deriving tissue viability tPO2 thresholds from these values is not possible. Recently, tPO2 was measured invasively in periinfarct areas in patients with severe MCA syndrome and >50% infarction of the MCA territory on CT, and the authors reported that tPO2 values reaching <10.5 mm Hg predicted a malignant course with 100% specificity and 94% sensitivity,12 and in such patients the tPO2 varies with cerebral perfusion pressure, indicating impaired cerebral autoregulation.13
|
A Brief History of the Nitroimidazoles |
|---|
After the discovery that 5-nitroimidazole (azomycin) was active against anaerobic infections2 many nitroimidazoles such as metronidazole were generated. However, the unique characteristics of nitroimidazoles in hypoxic environments lead researchers to investigate other applications. Oncologists recognized that misonidazole increases the response of hypoxic tumors to radiation or chemical therapy.14 Subsequently, labeled nitroimidazoles were developed to allow the visualization of hypoxic tissue in vivo using nuclear medicine techniques.14,15 These have been labeled with 14C, 125I, and 3H for use with autoradiography, 99mTc and 123I for use with single photon emission computed tomography (SPECT),16–18 and 18F for use with PET. Fluorine-18-Fluoromisonidazole (18F-FMISO) is currently the most widely used tracer in a variety of applications (see below). Other PET hypoxia tracers have also been developed to address issues relating to 18F-FMISO, including 18F-Fluoroerythronitroimidazole (18F-FETNIM), 18F-Fluoroazomycinarabinofuranoside (18F-FAZA), and Cu(II)-Diacetyl-bis[N4methylthiosemicarbazone] (Cu-ATSM).2,19,20 The latter, though not a nitroimidazole, has favorable pharmacokinetic properties in nonstroke studies,21–23 and its trapping in hypoxic tissue appears similar to nitroimidazoles,24 although dissenting views exist.22
|
Pharmacokinetics of Nitroimidazoles |
|---|
Good penetration of 18F-FMISO and 18F-FAZA into the normal brain has been reported in mice 3 hours after IV injection of the tracers, with blood/brain ratios around 0.7 and 1.3, respectively.25 Results regarding 18F-FAZA also suggest that this compound has faster clearance kinetics in normal brain tissue than 18F-FMISO. With regard to metabolism, 18F-FMISO studies in man26 indicate that 92% to 96% of circulating 18F was in the form of the parent compound after 90 minutes, and 85% at 2 to 3 hours.27 Of note, in vivo metabolism of 18F-FMISO seems to be more rapid in mice.28
To measure the rate of nitroimidazole retention in the target tissue and hence the severity of hypoxia within it, a 2-tissue and a 4-tissue compartmental model have been proposed for nitroimidazole imaging in oncology studies.26,27,29,30 Regarding stroke, an irreversible 2-tissue compartment model for 18F-FMISO in the rat stroke model appeared to be reasonably accurate and practicable in representing the irreversible binding properties of the tracer.31 Further work is, however, needed with respect to the most appropriate number of tissue compartments to use, and deriving a rate constant that reflects hypoxic trapping.
|
Nonstroke Clinical Studies |
|---|
Vinjamuri et al32 have reported on the successful detection of hypoxia related to traumatic brain injury in humans using 123I-IAZA SPECT. Fluoromisonidazole was shown to be useful in the detection of anaerobic odontogenic infections and in imaging of tuberculosis-associated hypoxia with high sensitivity in inaccessible sites.33 Nevertheless, nitroimidazoles have been most widely used in the investigation of tumor hypoxia.
Valk et al34 were the first to report on imaging malignant brain tumors in humans using 18F-FMISO PET. Significant retention of the tracer at late time points was demonstrated in 2 of 3 patients, with tumor-to-plasma ratios of 1.1, 1.49, and 0.7, respectively. All 3 patients showed BBB disruption at the tumor site, as assessed with Rubidium-82 imaging. Bruehlmeier et al26 studied 11 patients with various types of brain tumors using 18F-FMISO combined with 15O-H2O PET to investigate relationship to tumor perfusion. At late scan time points, increased FMISO uptake was observed in 8/11 tumors (tracer distribution volumes >1 using a 3-compartment model). Pixel-by-pixel analysis yielded an early positive correlation between 18F-FMISO and 15O-H2O that disappeared beyond 90 minutes, indicating perfusion-independent FMISO trapping beyond this time point. Accordingly, analysis of late images (150 to 170 minutes) showed that the presence or absence of tracer trapping was associated with various patterns of hypo- and hyper-perfusion, which may be explained by a number of factors: (1) an imbalance of oxygen supply and demand, maintaining a low tPO2 despite adequate perfusion, particularly because tracer uptake is seen in the tumor periphery which is the most active and proliferating part; (2) malignant tumors often exhibit high vascularity and flow but low effective tissue perfusion; (3) in tumor tissue nitroimidazole tracers may bind to tumor macromolecules independent of tissue hypoxia; and (4) FMISO may be sensitive to mild degrees of hypoxia.
Using SPECT, 123I-IAZA has also been used in patients with brain tumors. In one study,35 none of 11 glioblastomas showed significant hypoxia (tumor to normal tissue ratio <1.1). In another study,16 again none of 4 glioblastomas showed hypoxia, yet 1 of 2 brain metastases from small cell lung carcinoma showed increased uptake of the tracer (ratio=3.1). This highlights the significance of tumor tissue type in the sensitivity of the various nitroimidazoles to hypoxia. As with the above-described FMISO PET study, the latter study demonstrated that 5/14 tumors showed increased IAZA uptake despite normal or increased perfusion.
To confirm that 18F-FMISO imaging depicts only viable tumor tissue, Rajendran et al36 aimed to investigate the relationship of 18F-FMISO trapping to glucose metabolism as assessed by 18F-FDG in glioblastoma multiforme. Increased FMISO trapping was detected in all patients (ratios 1.2) but the correlation to FDG uptake was weak (r=0.38), independent of tumor size. This variable relationship, also reported in animals,37 indicates that interpretation of such combined studies is difficult, probably in part because of uncertainties in the biochemical behavior of FDG in tumor tissue.
One intriguing finding with nitroimidazole agents, with potential implications for stroke, is their trapping in some tumors despite normal or even increased perfusion. Were it in normal brain, this observation would point to trapping mechanisms other than tissue hypoxia for these agents, calling into question their validity as hypoxia tracers in stroke. However, several arguments support the idea that this trapping indeed reflects hypoxia, mainly (1) nitroimidazole trapping in animal models of tumors does correspond to very low tPO2; (2) the probable existence in highly proliferative parts of some tumors of an imbalance of oxygen supply and demand maintaining a low tPO2 despite normal or even increased perfusion relative to normal tissue, and overall reduced oxygen consumption38; and (3) the possibility of a poorly effective tissue perfusion, attributable to tumor-related physical barriers to oxygen diffusion. Thus, although tumoral tissue cannot be directly compared to the normal but acutely ischemic brain, nitroimidazole trapping in tumors may still indicate impending cell death and tissue necrosis from severe hypoxia, akin to a "slow" penumbra. Alternative interpretations to severe hypoxia to explain nitroimdazoles binding in tumors should however be considered, firstly that they may bind to tumor-specific macromolecules, mimicking hypoxia, and secondly that they may also be sensitive to mild degrees of hypoxia. The latter, if true, would have implications for stroke as it would suggest these tracers may reflect not just the penumbra but also the oligemia.39,40 More work is needed to clarify these issues.
|
Footnotes |
|---|
M.T. and R.R.M. contributed equally to this work.
|
Supplemental References |
|---|
1. Olive PL, Banath JP, Aquino-Parsons C. Measuring hypoxia in solid tumors—is there a gold standard? Acta Oncol. 2001; 40: 917–923.[CrossRef][Medline] [Order article via Infotrieve]
2. Ballinger JR. Imaging hypoxia in tumors. Semin Nucl Med. 2001; 31: 321–329.[CrossRef][Medline] [Order article via Infotrieve]
3. Evans SM, Judy KD, Dunphy I, Jenkins WT, Hwang WT, Nelson PT, Lustig RA, Jenkins K, Magarelli DP, Hahn SM, Collins RA, Grady MS, Koch CJ. Hypoxia is important in the biology and aggression of human glial brain tumors. Clin Cancer Res. 2004; 10: 8177–8184.
4. Liu S, Shi H, Liu W, Furuichi T, Timmins GS, Liu KJ. Interstitial po2 in ischemic penumbra and core are differentially affected following transient focal cerebral ischemia in rats. J Cereb Blood Flow Metab. 2004; 24: 343–349.[Medline] [Order article via Infotrieve]
5. Hou H, Grinberg OY, Grinberg SA, Demidenko E, Swartz HM. Cerebral tissue oxygenation in reversible focal ischemia in rats: Multi-site epr oximetry measurements. Physiol Meas. 2005; 26: 131–141.[CrossRef][Medline] [Order article via Infotrieve]
6. Tobias JD. Cerebral oxygenation monitoring: Near-infrared spectroscopy. Expert Rev Med Devices. 2006; 3: 235–243.[CrossRef][Medline] [Order article via Infotrieve]
7. Crockard HA, Symon L, Branston NM, Juhasz J. Changes in regional cortical tissue oxygen tension and cerebral blood flow during temporary middle cerebral artery occlusion in baboons. J Neurol Sci. 1976; 27: 29–44.[CrossRef][Medline] [Order article via Infotrieve]
8. Farrar JK. Tissue po2 threshold of ischemic cell damage following mca occlusion in cats. J Cereb Blood Flow Metab. 1991; 11: S553.
9. Harris RJ, Richards PG, Symon L, Habib AH, Rosenstein J. Ph, k+, and po2 of the extracellular space during ischemia of primate cerebral cortex. J Cereb Blood Flow Metab. 1987; 7: 599–604.[Medline] [Order article via Infotrieve]
10. Hoffman WE, Charbel FT, Gonzalez-Portillo G, Ausman JI. Measurement of ischemia by changes in tissue oxygen, carbon dioxide, and ph. Surg Neurol. 1999; 51: 654–658.[CrossRef][Medline] [Order article via Infotrieve]
11. Hoffman WE, Charbel FT, Edelman G. Brain tissue oxygen, carbon dioxide, and ph in neurosurgical patients at risk for ischemia. Anesth Analg. 1996; 82: 582–586.[Abstract]
12. Dohmen C, Bosche B, Graf R, Staub F, Kracht L, Sobesky J, Neveling M, Brinker G, Heiss WD. Prediction of malignant course in mca infarction by pet and microdialysis. Stroke. 2003; 34: 2152–2158.
13. Dohmen C, Bosche B, Graf R, Reithmeier T, Ernestus RI, Brinker G, Sobesky J, Heiss WD. Identification and clinical impact of impaired cerebrovascular autoregulation in patients with malignant middle cerebral artery infarction. Stroke. 2007; 38: 56–61.
14. Nunn A, Linder K, Strauss HW. Nitroimidazoles and imaging hypoxia. Eur J Nucl Med. 1995; 22: 265–280.[CrossRef][Medline] [Order article via Infotrieve]
15. Chapman JD, Franko AJ, Sharplin J. A marker for hypoxic cells in tumors with potential clinical applicability. Br J Cancer. 1981; 43: 546–550.[Medline] [Order article via Infotrieve]
16. Groshar D, McEwan AJ, 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.
17. Stypinski D, McQuarrie SA, Wiebe LI, Tam YK, Mercer JR, McEwan AJ. Dosimetry estimations for 123i-iaza in healthy volunteers. J Nucl Med. 2001; 42: 1418–1423.
18. Parliament MB, Chapman JD, Urtasun RC, McEwan AJ, Golberg L, Mercer JR, Mannan RH, Wiebe LI. Noninvasive assessment of human tumor hypoxia with 123i-iodoazomycin arabinoside: Preliminary report of a clinical study. Br J Cancer. 1992; 65: 90–95.[Medline] [Order article via Infotrieve]
19. Sorger D, Patt M, Kumar P, Wiebe LI, Barthel H, Seese A, Dannenberg C, Tannapfel A, Kluge R, Sabri O. [18f]fluoroazomycinarabinofuranoside (18faza) and [18f]fluoromisonidazole (18fmiso): A comparative study of their selective uptake in hypoxic cells and pet imaging in experimental rat tumors. Nucl Med Biol. 2003; 30: 317–326.[CrossRef][Medline] [Order article via Infotrieve]
20. Gronroos T, Eskola O, Lehtio K, Minn H, Marjamäki P, Bergman J, Haaparanta M, Forsback S, Solin O. Pharmacokinetics of [18f]fetnim: A potential marker for pet. J Nucl Med. 2001; 42: 1397–1404.
21. Takahashi N, Fujibayashi Y, Yonekura Y, Welch MJ, Waki A, Tsuchida T, Sadato N, Sugimoto K, Itoh H. Evaluation of 62cu labeled diacetyl-bis(n4-methylthiosemicarbazone) as a hypoxic tissue tracer in patients with lung cancer. Ann Nucl Med. 2000; 14: 323–328.[Medline] [Order article via Infotrieve]
22. Yuan H, Schroeder T, Bowsher JE, Hedlund LW, Wong T, Dewhirst MW. Intertumoral differences in hypoxia selectivity of the pet imaging agent 64cu(ii)-diacetyl-bis(n4-methylthiosemicarbazone). J Nucl Med. 2006; 47: 989–998.
23. Obata A, Yoshimi E, Waki A, Lewis JS, Oyama N, Welch MJ, Saji H, Yonekura Y, Fujibayashi Y. Retention mechanism of hypoxia selective nuclear imaging/radiotherapeutic agent cu-diacetyl-bis(n4-methylthiosemicarbazone) (cu-atsm) in tumor cells. Ann Nucl Med. 2001; 15: 499–504.[Medline] [Order article via Infotrieve]
24. Padhani AR, Krohn KA, Lewis JS, Alber M. Imaging oxygenation of human tumors. Eur Radiol. 2006.
25. Piert M, Machulla HJ, Picchio M, Reischl G, Ziegler S, Kumar P, Wester HJ, Beck R, McEwan AJ, Wiebe LI, Schwaiger M. Hypoxia-specific tumor imaging with 18f-fluoroazomycin arabinoside. J Nucl Med. 2005; 46: 106–113.
26. Bruehlmeier M, Roelcke U, Schubiger PA, Ametamey SM. Assessment of hypoxia and perfusion in human brain tumors using pet with 18f-fluoromisonidazole and 15o-h2o. J Nucl Med. 2004; 45: 1851–1859.
27. Casciari JJ, Graham MM, Rasey JS. A modeling approach for quantifying tumor hypoxia with [f-18]fluoromisonidazole pet time-activity data. Med Phys. 1995; 22: 1127–1139.[CrossRef][Medline] [Order article via Infotrieve]
28. Rasey JS, Hofstrand PD, Chin LK, Tewson TJ. Characterization of [18f]fluoroetanidazole, a new radiopharmaceutical for detecting tumor hypoxia. J Nucl Med. 1999; 40: 1072–1079.
29. Thorwarth D, Eschmann SM, Paulsen F, Alber M. A kinetic model for dynamic [18f]-fmiso pet data to analyze tumor hypoxia. Phys Med Biol. 2005; 50: 2209–2224.[CrossRef][Medline] [Order article via Infotrieve]
30. Heiss WD, Herholz K. Brain receptor imaging. J Nucl Med. 2006; 47: 302–312.
31. Takasawa M, Beech JS, Fryer TD, Hong YT, Hughes JL, Igase K, Jones PS, Smith R, Aigbirhio FI, Menon DK, Clark JC, Baron JC. Imaging of brain hypoxia in permanent and temporary middle cerebral artery occlusion in the rat using 18f-fluoromisonidazole and positron emission tomography: A pilot study. J Cereb Blood Flow Metab. 2007; 27: 679–689.[Medline] [Order article via Infotrieve]
32. Vinjamuri S, O’Driscoll K, Maltby P, McEwan AJ, Wiebe LI, Critchley M. Identification of hypoxic regions in traumatic brain injury. Clin Nucl Med. 1999; 24: 891–892.[CrossRef][Medline] [Order article via Infotrieve]
33. Liu RS, Yeh SH, Chang CP, et al. Detection of odontogenic infections by f-18 fluoromisonidazole (fmiso). J Nucl Med. 1994; 35: 113P.
34. Valk PE, Mathis CA, Prados MD, Gilbert JC, Budinger TF. Hypoxia in human gliomas: Demonstration by pet with fluorine-18-fluoromisonidazole. J Nucl Med. 1992; 33: 2133–2137.
35. Urtasun RC, Parliament MB, McEwan AJ, Mercer JR, Mannan RH, Wiebe LI, Morin C, Chapman JD. Measurement of hypoxia in human tumors by noninvasive spect imaging of iodoazomycin arabinoside. Br J Cancer Suppl. 1996; 27: S209–212.[Medline] [Order article via Infotrieve]
36. Rajendran JG, Mankoff DA, O’Sullivan F, Peterson LM, Schwartz DL, Conrad EU, Spence AM, Muzi M, Farwell DG, Krohn KA. Hypoxia and glucose metabolism in malignant tumors: Evaluation by [18f]fluoromisonidazole and [18f]fluorodeoxyglucose positron emission tomography imaging. Clin Cancer Res. 2004; 10: 2245–2252.
37. Sorensen M, Horsman MR, Cumming P, Munk OL, Keiding S. Effect of intratumoral heterogeneity in oxygenation status on fmiso pet, autoradiography, and electrode po2 measurements in murine tumors. Int J Radiat Oncol Biol Phys. 2005; 62: 854–861.[CrossRef][Medline] [Order article via Infotrieve]
38. Herholz K, Herscovitch P, Heiss W-D. Neuropet. Positron emission tomography in neuroscience and clinical neurology. Heidelberg: Springer-Verlag; 2004.
39. Baron JC. Perfusion thresholds in human cerebral ischemia: Historical perspective and therapeutic implications. Cerebrovasc Dis. 2001; 11 Suppl 1: 2–8.[Medline] [Order article via Infotrieve]
40. Heiss WD. Flow thresholds of functional and morphological damage of brain tissue. Stroke. 1983; 14: 329–331.
This article has been cited by other articles:
 |
|
 |
|
  J. A. Alawneh, R. R. Moustafa, and J.-C. Baron Hemodynamic Factors and Perfusion Abnormalities in Early Neurological Deterioration Stroke, June 1, 2009; 40(6): e443 - e450. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||