Central Benzodiazepine Receptor Distribution After Subcortical Hemorrhage Evaluated by Means of [123I]Iomazenil and SPECT
Background and Purpose [123I]Iomazenil is a single-photon emission computed tomography (SPECT) tracer that selectively binds to central benzodiazepine receptor in the neuron membrane. With this ligand, we studied the central benzodiazepine receptor distribution in the cortex remote from subcortical hematoma in intracerebral hemorrhage patients.
Methods Four patients with unilateral putaminal hemorrhage and one patient with right thalamic hemorrhage were studied (mean±1 SD age, 50.0±8.8 years). The initial volume of hematoma ranged from 4.3 to 31.0 mL (mean±1 SD, 17.5±12.3 mL). SPECT images obtained 3 hours after intravenous administration of [123I]iomazenil (167 MBq/750 ng) were analyzed. In three patients, perfusion was evaluated with [123I]IMP. On SPECT images, the radioactivity ratio of the ipsilateral to the contralateral cerebral cortex (I/C ratio) or of the contralateral to the ipsilateral cerebellar hemisphere (C/I ratio) was measured.
Results The I/C ratio for iomazenil was significantly decreased compared with unity in the temporal lobe (0.84±0.08, P<.01) and the parietal lobe (0.87±0.10, P<.05), but the C/I ratio in the cerebellum (1.00±0.03) was not. The C/I ratio for perfusion in the cerebellar cortex (0.83±0.04, P<.01) was significantly decreased compared with that in normal subjects.
Conclusions Central benzodiazepine receptor–[123I]iomazenil binding was decreased in the ipsilateral cerebral cortex remote from the subcortical hematoma. This preliminary result may facilitate further study of the potential damage of the cortical neurons remote from subcortical hematoma.
In patients with putaminal or thalamic hemorrhage, a disorder of higher cortical function is sometimes observed.1 2 Small putaminal hemorrhage produces a syndrome of hemimotor and hemisensory defects. With increasing volume of hematoma, abnormalities of cortical functions such as aphasia and hemianopsia become manifest. In these patients, CT scan accurately demonstrates the lesion involved. The location and extent of a lesion as visualized on CT scan are correlated with the types of deficits of higher cortical functions.3 However, the cerebral cortex responsible for such functions and the cortex remote from the hematoma usually present a normal appearance on CT scan.
Recently, the SPECT tracer I-123 Ro 16-0154 (ethyl-5,6-dihydro-7-iodo-5-methyl-6-oxo-4H-imidazo[1,5-a][1,4]-benzodiazepine-3-carboxylate; iomazenil) was developed for central BZD receptor imaging.4 The central BZD receptor is found exclusively in the membrane of neurons. In humans, [123I]iomazenil accumulates at the cerebral and cerebellar cortices but not at the basal ganglia, thalamus, and brain stem.5 Histopathologic study indicates that decreased BZD receptor density is associated with a reduction in neuron density.6 We recently reported a reduction in [123I]iomazenil uptake in the peri-infarct area with CT normodensity.7 The [123I]iomazenil uptake may thus be a sensitive marker of neuronal injury in the brain.
To investigate the potential neuronal damage in the cerebral cortex remote from subcortical hemorrhage, we evaluated the distribution of central BZD receptor binding in brain hemorrhage patients with [123I]iomazenil and SPECT.
Subjects and Methods
Five patients (four men and one woman) with unilateral intracerebral hemorrhage in the territory of perforating arteries were examined. Table 1⇓ summarizes the clinical data for each patient as to the location of the hematoma demonstrated on the initial CT scan, the level of consciousness on admission, the neurological deficits at the time of iomazenil study, and the type of therapy. The patients ranged in age from 35 to 57 years (mean±1 SD, 50.0±8.8 years). They were all hypertensive before the present insult. Four patients showed putaminal hemorrhage and one thalamic hemorrhage on the initial CT scan. All of the patients showed hemiplegia or hemiparesis on the side contralateral to the brain lesion. In three patients (patients 1, 2, and 3), associated cortical signs of aphasia or hemianopsia were present. A ventricular shunt was placed to resolve hydrocephalus in the acute phase in patient 1. Patients 2 and 3 received stereotactic aspiration of the hematoma. Written informed consent was received from all patients before study initiation.
Brain CT Scan
The initial CT scan was carried out within 12 hours after onset. Ten axial images with 10-mm slice thickness and 10-mm center-to-center spacing were obtained parallel to the OM line. The intraparenchymal hematoma volume was estimated according to the method of Broderick et al.8 The assessment of morphological abnormalities of the brain (such as the degree of brain swelling in the acute phase, extension of the CT low-density area shown on CT, widening of cortical sulci, and hydrocephalic change in the chronic phase) was conducted independently by two neuroradiologists (J.H. and E.S.).
All patients except patient 4 underwent four-vessel conventional cerebral angiography to detect underlying pathological vessels. In addition, the presence of steno-occlusive lesions due to arteriosclerosis of the cervical and intracranial major arteries was determined. The study was performed within 1 week after onset.
Patients received approximately 167 MBq of [123I]iomazenil by intravenous bolus injection of 1.5 mL of solution into the cubital vein. The specific activity was approximately 222 MBq/μg at administration. The mass dose of iomazenil injected was 0.75 μg. Dosimetry estimates for [123I]iomazenil were within the dose range acceptable for clinical use.9 Before the study, the patients were given oral potassium iodide to prevent contamination of the thyroid by administered radioactivity.
SPECT Procedure for [123I]Iomazenil
The iomazenil study was carried out on day 7 (with day 0 being the day of onset) in patient 4, day 17 in patient 5, day 42 in patient 3, day 72 in patient 2, and 19 months after onset in patient 1. A ring-type single-photon emission computed tomograph (Headtome SET-080, Shimadzu Co) was used to measure the distribution of radioactivity in the brain. The scanner simultaneously produces 31 tomographic axial slices. A low-energy all-purpose collimator was used for data acquisition. Data were recorded on a 64×64 matrix. A Butterworth filter, with cutoff of 0.45 Ny and order of 3, and a Ramp filter were used for image reconstruction. Reconstructed images were corrected for tissue absorption with an attenuation coefficient of 0.065 cm−1. The in-plane and axial spatial resolutions of the scanner were 14 mm and 22 mm with FWHM, respectively. SPECT imaging was started 180 minutes after injection, with a data acquisition time of 12 minutes. The plane of iomazenil imaging was set parallel to the OM line. All imaging data were transferred to a conventional UNIX workstation system.
Evaluation of Central BZD Receptor Binding of [123I]Iomazenil
As has been demonstrated in a displacement trial in normal volunteers, more than 85% of cortical radioactivity at 6 hours after injection is associated with specific binding to central BZD receptors.5 In the present study, [123I]iomazenil images obtained at 180 minutes after injection were used for the analysis of central BZD receptor distribution. The image counts for the frontal lobe, temporal lobe, parietal lobe, occipital lobe, basal ganglia, thalamus, and cerebellar cortices of the affected hemisphere, as well as the mirror regions in the nonaffected hemisphere, were read using 16-mm-diameter circular ROIs. For the frontal lobe, three ROIs were located on the superior, middle, and inferior frontal gyrus, respectively, at OM+40 mm. For the temporal lobe, five ROIs were located on the lateral temporal cortex at OM+40 mm. For the parietal lobe, three ROIs were placed on the inferior (two ROIs) and superior (one ROI) parietal lobule at OM+70 mm. For the occipital lobe, three ROIs were placed on the occipital cuneus, superior occipital gyrus, and middle occipital gyrus at OM+40 mm. For the cerebellum, three ROIs were placed on the cerebellar cortex. The mean count ratio of the affected (ipsilateral) to nonaffected (contralateral) cerebral hemisphere, defined herein as the I/C ratio, was calculated. The count ratio of the contralateral to ipsilateral cerebellar hemisphere, defined as the C/I ratio, was also calculated. The mean cortical I/C ratio was defined as the average of the I/C ratio of the frontal, temporal, occipital, and parietal lobe for all patients. The statistical significance of the difference from the unity for the regional I/C ratio and mean cortical I/C ratio was examined by paired t test and one-sample Wilcoxon test because of the small sample size. The relationship of the mean cortical I/C ratio and hematoma volume was also analyzed.
Evaluation of CBF
Patients 3, 4, and 5 underwent the perfusion study on days 35, 13, and 24, respectively. Blood flow imaging was performed using [123I]IMP, and from 111 to 167 MBq of [123I]IMP was administered per injection. Perfusion images were obtained with the same scanner used in the iomazenil study along with a low-energy all-purpose collimator. The slice was adjusted to the OM line. The IMP studies were performed according to the autoradiographic method to quantify CBF.10 The I/C ratio of the CBF images was calculated for the frontal lobe, temporal lobe, parietal lobe, occipital lobe, basal ganglia, and thalamus, and the C/I ratio for the cerebellar cortex. These values were compared with the right-to-left ratios of the CBF of age- and sex-matched normal subjects (15 men; mean age, 48.5±7.2 years). The Wilcoxon–Mann-Whitney test for the small sample size was used for the statistical analysis.
CT and Angiography
Table 2⇓ summarizes the morphological findings demonstrated on the CT scans. The mean volume of the hematoma demonstrated on the initial CT scan within 12 hours after onset was 17.5±12.3 mL, and the hematoma volume ranged from 4.3 to 31.0 mL. There was no evidence on the CT scans of lesions in the cortical area other than a linear low-density lesion corresponding to the site of insertion of the shunt tube through the right frontal cortex in one patient (patient 1). Cerebral angiography disclosed irregular arterial walls but no steno-occlusive lesions in the cervical and intracranial major arteries.
Distribution of [123I]Iomazenil
The initial CT, the CT at the chronic phase, and the [123I]iomazenil image for patients 2 and 5 are shown in Figs 1⇓ and 2⇓, respectively. In all of the patients, visual inspection revealed focal decrease of [123I]iomazenil uptake in the ipsilateral cerebral cortex compared with the corresponding contralateral region. The I/C and C/I ratios of [123I]iomazenil are summarized in Table 3⇓. The temporal lobe (P<.01) and parietal lobe (P<.05) showed significant reduction compared with unity. The mean cortical I/C ratio was also significantly decreased (P<.05). There was no asymmetry of [123I]iomazenil uptake in the cerebellum. In Fig 3⇓, the mean cortical I/C ratio in each patient is shown plotted against the initial volume of the hematoma. The mean cortical I/C ratio was significantly inversely correlated with the initial hematoma volume (y=0.98−0.0053x; r=−.94, P<.02).
Clinical Symptoms and [123I]Iomazenil Uptake
Three of the five patients (patients 1, 2, and 3) showed continuous cortical signs. The cortical dysfunction was observed on admission, continued until the iomazenil study, and persisted during the follow-up period of 6 months. The initial hematoma volume, determined by measurements made from the CT scan, was 13.2 mL or more in these patients. In patient 1, left homonymous hemianopsia was found, associated with an I/C ratio of 0.75 for the right occipital lobe, 0.79 for the right temporal lobe, and 0.80 for the right parietal lobe. In patient 2, global aphasia was seen, accompanied by a local decrease in the left temporal lobe (I/C ratio, 0.75) and parietal lobe (I/C ratio, 0.75). In patient 3, residual motor aphasia was noted. In this patient, [123I]iomazenil uptake was reduced in the right Broca’s area (I/C ratio, 0.78).
In patients 4 and 5, no cortical signs were evident. The regional cortical values for the I/C ratio for iomazenil were all ≥0.87 (temporal lobe) in patient 4 and ≥0.86 (temporal lobe) in patient 5.
Table 3⇑ also summarizes the mean C/I ratio of CBF for the cerebellum and the mean I/C ratios for supratentorial structures for the patients and for the normal subjects (right-to-left ratios for supratentorial structures and left-to-right ratio for the cerebellum). A significant difference was found in the cerebellum, the thalamus, and the temporal lobe (P≤.05). Fig 4⇓ illustrates the CT, blood flow images, and [123I]iomazenil images at the level of the basal ganglia (top) and cerebellum (bottom) in patient 3.
In an experimental model of middle cerebral artery occlusion in the baboon, Sette et al11 demonstrated a decrease in the binding of [11C]flumazenil to central BZD receptor in the infarction and CT-normodense peri-infarct area. In human cerebral infarction, reduction in the binding of [123I]iomazenil to central BZD receptor extends to the CT-normodense area surrounding the infarction.7 These studies indicated that central BZD receptor binding is a sensitive marker of ischemic damage of the brain. The decreased [123I]iomazenil uptake in our patients may indicate that there is a CT-negative region of neuronal injury in the ipsilateral cortex, predominantly in the ipsilateral temporal and parietal lobes.
Compression of cortical gray matter due to subcortical hematoma and surrounding edema and subsequent ischemic injury12 is a possible mechanism of pathogenesis in reduction of [123I]iomazenil–central BZD receptor binding. There is, however, no report describing the pathological evidence of such neuronal damage in the cortex remote from subcortical hematoma. Another possible mechanism is transneuronal degeneration.13 A hematoma seated in the basal ganglia and thalamus primarily destroys neurons projecting to cortical neurons and neurons receiving fibers from cortical neurons. Loss of these neurons may induce a remote neuronal injury.
Three patients with aphasia or anopsia showed an I/C ratio of ≤0.78 in the cortical area. On the other hand, in two patients without cortical symptoms, the cortical I/C ratio was >0.85. Although the degree of neuronal damage was variable in our patients, it seems that there is a critical threshold of I/C ratio at around 0.80 to 0.85, below which the neurological deficits are manifest.
Impairment of hemodynamics and metabolism in areas of the structurally normal brain distant from subcortical lesions has been demonstrated by a number of investigators.14 15 These phenomena have been found in the ipsilateral and contralateral cerebral hemisphere, as well as the contralateral cerebellar hemisphere. They have been ascribed to a decrease in neuronal activity caused by interruption of the input from afferent fiber pathways, a phenomenon known as diaschisis. In the present study, although the number of patients was very limited, we observed a widespread hypoperfusion in the ipsilateral cerebral cortex and in the contralateral cerebellum. The hypoperfusion was not due to insufficiency of perfusion because no steno-occlusive vascular lesions were confirmed by angiography. Within the hypoperfused cerebral cortex, [123I]iomazenil uptake was decreased in the ipsilateral temporal and parietal lobes. Therefore, hypoperfusion in these areas is not entirely due to functional deactivation but is also partly due to the adaptation of CBF to neuronal damage.
Crossed cerebellar diaschisis16 17 was also found in our patients. However, this was not associated with a reduction of [123I]iomazenil uptake. A similar phenomenon has also been observed in patients with unilateral cerebral infarction.
There are several methodological limitations in the present study. The recently developed methods of quantitative evaluation of BZD receptor density and affinity5 were not used in the present study. One may suspect that a decrease in [123I]iomazenil uptake in the ipsilateral cortex would result from reduced delivery of the ligand because the CBF was also reduced in that area. However, in the cerebellum, decreased blood flow did not necessarily result in reduced [123I]iomazenil uptake. This is one piece of evidence that [123I]iomazenil uptake is not entirely a function of delivery.
Because of the limited spatial resolution of the scanner used, radioactivity in the cortex might be affected by that of the basal ganglia. However, the mean radioactivity accumulated at the ipsilateral basal ganglia was only 15% lower than that accumulated at the contralateral basal ganglia. In all of the patients, only one of the five temporal ROIs was located 20 mm away from the hematoma lesion. This ROI count is underestimated to some extent because of reduced iomazenil uptake in the ipsilateral basal ganglia. All other ROIs were more than 28 mm (twice the FWHM of the in-plane resolution) from the lesion. On the basis of these analyses, we considered that reduced iomazenil uptake in the ipsilateral cortex did not result from a methodological error due to the limited spatial resolution.
In conclusion, central BZD receptor distribution was studied using [123I]iomazenil and SPECT in patients with unilateral subcortical hemorrhage. The central BZD receptor–[123I]iomazenil binding was decreased in the cerebral cortex remote from the subcortical hematoma, and the magnitude of the decline was correlated with the initial volume of hematoma. Profound decrease in central BZD receptor–[123I]iomazenil binding was associated with the presence of aphasia and hemianopsia. Although the pathogenesis is not clear, this finding may indicate neuronal damage that neither CT nor pathological study has yet uncovered. To test the validity of this observation and elucidate the pathogenesis, further studies should be performed in a larger patient sample.
Selected Abbreviations and Acronyms
|C/I ratio||=||contralateral to ipsilateral count ratio|
|CBF||=||cerebral blood flow|
|FWHM||=||full width at half maximum|
|I/C ratio||=||ipsilateral to contralateral count ratio|
|ROI||=||region of interest|
|SPECT||=||single-photon emission computed tomography|
The present study was supported by a grant-in-aid from the Ministry of Health and Welfare, Japan. The authors acknowledge the encouraging comments of Dr Hitoshi Fukasawa and Dr Yasuji Yoshida of the Department of Pathology, Akita Research Institute of Brain and Blood Vessels, Akita, Japan. We also appreciate the advice on statistical analysis of Dr Yuhko Miura. We thank Nihon-Mediphysics Corp for providing the [123I]iomazenil.
- Received April 24, 1995.
- Revision received July 31, 1995.
- Accepted August 24, 1995.
- Copyright © 1995 by American Heart Association
Fisher CM. Clinical syndromes in cerebral hemorrhage. In: Fields WS, ed. Pathogenesis and Treatment of Cerebrovascular Disease. Springfield, Ill: Charles C Thomas Publishing; 1961:318-338.
Alexander MP, Naeser MA, Palumbo CL. Correlations of subcortical CT lesion sites and aphasia profile. Brain. 1987;110:961-991.
Beer HF, Blauenstein PA, Hasler PH, Delaloye B, Riccabona G, Bangerl I, Hunkeler W, Bonetti EP, Pieri L, Richards JG, Schubiger PA. In vitro and in vivo evaluation of Iodine-123 Ro 16-0154: a new imaging agent for SPECT investigations of benzodiazepine receptors. J Nucl Med. 1990;31:1007-1014.
Abi-Dargham A, Laruelle M, Seibyl J, Ratter Z, Baldwin RM, Zoghbi SS, Zea-Ponce Y, Bremner JD, Hyde TM, Charney DS, Hoffer PB, Innis RB. SPECT measurement of benzodiazepine receptors in human brain with iodine-123-iomazenil: kinetic and equilibrium paradigms. J Nucl Med. 1994;35:228-238.
Benavides J, Serrano A, Duval D, Bourdiol F, Toulmond S, Scatton B. Autoradiographic detection and quantitation of traumatic brain lesions in the rat brain by using site radioligands. In: Kriegstein J, Oberpichler H, eds. Pharmacology of Cerebral Ischemia. Stuttgart, Germany: Wissenschaftliche Verlagsgesellschaft mbH; 1989:39-45.
Hatazawa J, Satoh T, Shimosegawa E, Okudera T, Inugami A, Ogawa T, Fujita H, Noguchi K, Kanno I, Miura S, Murakami M, Iida H, Miura Y, Uemura K. Evaluation of cerebral infarction with iodine-123-iomazenil SPECT. J Nucl Med. In press.
Broderick JP, Brott TG, Duldner JE, Tomsick T, Huster G. Volume of intracerebral hemorrhage: a powerful and easy-to-use predictor of 30-day mortality. Stroke. 1993;24:987-993.
Dey HM, Seibyl JP, Stubbs JB, Zoghbi SS, Baldwin RM, Smith EO, Zubal IG, Zea-Ponce Y, Olson C, Charney DS, Hoffer PB, Innis RB. Human biodistribution and dosimetry of the SPECT benzodiazepine receptor radioligand iodine-123-iomazenil. J Nucl Med. 1994;35:399-404.
Iida H, Itoh H, Nakazawa M, Hatazawa J, Nishimura H, Onishi Y, Uemura K. Quantitative mapping of regional cerebral blood flow using iodine-123-IMP and SPECT. J Nucl Med. 1995;35:2019-2030.
Sette G, Baron JC, Young AR, Miyazawa H, Tillet I, Barre L, Travere JM, Derlon JM, MacKenzie ET. In vivo mapping of brain benzodiazepine receptor changes by positron emission tomography after focal ischemia in the anesthetized baboon. Stroke. 1993;24:2046-2058.
Miller JD, Adams JH. The pathophysiology of raised intracranial pressure. In: Adams JH, Duchen LW, eds. Greenfield’s Neuropathology. 5th ed. Baltimore, Md: Edward Arnold Publishers; 1992:69-105.
Duchen LW. General pathology of neurons and neuroglia. In: Adams JH, Duchen LW, eds. Greenfield’s Neuropathology. 5th ed. Baltimore, Md: Edward Arnold Publishers; 1992:1-68.
Feeny DM, Baron JC. Diaschisis. Stroke. 1986;15:817-830.
Baron JC, Bousser MG, Comar D, Castaigne P. Crossed cerebellar diaschisis in human supratentorial brain infarction. Ann Neurol. 1980;8:128. Abstract.
Pantano P, Baron JC, Bousser MG, Derousne C, Comar D. Crossed cerebellar diaschisis: further studies. Brain. 1986;109:677-694.