ORIGINAL ARTICLE Annals of Nuclear Medicine Vol. 14, No. 3, 159-163, 2000 Hemodynamic and metabolic state of hyperfixation with 99mTc-HMPAO brain SPECT in subacute stroke Ihnho CHO,*,** Kohei HAYASHIDA,* Satoshi IMAKITA,* Norihiko KUME* and Kazuki FUKUCHI* *Department of Radiology, National Cardiovascular Center **Department of Nuclear Medicine, Yeungnam University Hospital By means of positron emission tomography (PET), we investigated the hemodynamic and metabolic state of the hyperfixation identified as the increased accumulation with 99mTc-d,l-hexamethylpropyleneamine oxime (HMPAO) by single photon emission computed tomography (SPECT) in patients with subacute stroke. We studied four patients with subacute stroke having hyperfixed areas evaluated with CBF, CMRO2, OEF and CBV by PET. The hyperfixation rate with 99mTc-HMPAO was obtained by comparing the surplus rate with standardized CBF. The OEF and CMRO2 values in the hyperfixed areas of 4 patients were significantly lower than those in normal 5 controls (p < 0.01 ), but CBF and CBV were almost the same in patients and normal controls, but the hyperfixation rate of 0.30 +- 0.15 in 4 patients correlated well with CBV (r = 0.97, y = 11.75x + 0.42; p < 0.05). Hyperfixation with 99mTc-HMPAO in the infarct area revealing a mismatch between CMRO2 and CBF meant relative luxury perfusion. The hyperfixation rate determined by 99mTc-HMPAO brain SPECT correlated with CBV in the PET study. We can conclude that one of the main factors which caused hyperfixation was vasodilatation as well as the blood brain barrier disruption and the neovascularization. Key words: hyperfixation, cerebral infarction, Tc-99m HMPAO, brain SPECT, PET INTRODUCTION THE DISTRIBUTION of 99mTc-d,l-hexamethylpropyleneamine oxime (HMPAO) with single photon emission computed tomography (SPECT) was proportional to that of the regional cerebral blood flow (CBF),1 and the increased activity with 99mTc-HMPAO brain SPECT in subacute stroke was understandable as hyperemia.2 Nevertheless, it has been reported that this increased activity in subacute stroke might be due to hyperfixation with 99mTc-HMPAO manifested as a higher count rate over the infarct area than that of the opposite side.3 The asymmetry index of the hyperfixation in part of the infarct region in patients with subacute stroke showed excess of 9% to 30% with 133Xe.3,4 Received December 6, 1999, revision accepted February 10, 2000. For reprint contact: Kohei Hayashida, M.D., Department of Radiology, National Cardiovascular Center, 5-7-1 Fujishirodai, Suita, Osaka 565-8565, JAPAN. E-mail: khysd@hsp.ncvc.go,jp We measured CBF, the cerebral metabolic rate of oxygen consumption (CMRO2) and the oxygen extraction fraction (OEF) in four patients with subacute stroke by means of 15O gas steady-state positron emission tomography (PET) and compared it with that of the hyperfixed areas in 99mTc-HMPAO brain SPECT. MATERIALS AND METHODS Patients and Controls We studied 4 patients (3 males, 1 female; mean age, 63 +- 18 years; range, 36-78 yr) who showed increased accumulation in the infarct region on 99mTc-HMPAO brain SPECT images. Two patients had atrial fibrillation due to cardiac embolism. The other two were diagnosed as having artery-to-artery embolic stroke with intraluminal angiographical filling defects suggesting emboli. The clinical and neuroradiological features of the patient are summarized in Table 1. Five healthy volunteers (3 males, 2 females; mean age, 48 +- 16 years; range, 28-73 yr) underwent PET studies to determine normal values. The volunteers were free of cerebrovascular risk factors and neurologic symptoms or signs. No focal brain lesions were detected by CT and Magnetic Resonance Imaging (MRI). The ethics committee of our hospital approved this study and written informed consent to participate was given by all subjects. PET and SPECT examinations The PET study was followed by SPECT within 24 hours, then by MRI within 6 days. The CBF, OEF, CMRO2 and CBV values were measured with a Headtome IV PET scanner (Shimadzu, Kyoto, Japan) at a spatial resolution of 4.5 mm at FWHM and under 15O-labeled gas inhalation.5 Emission scans were corrected for the effect of tissue attenuation by comparison with corresponding transmission scans with an external 68Ge-68Ga ring source. Separate scans were obtained during continuous inhalation of 15O-labeled carbon dioxide and 15O-oxygen to measure CBF and OEF, respectively. A third scan measured CBV after the inhalation of 15O-labeled carbon monoxide. During the scans, serial blood samples were obtained through a fine gauge brachial artery catheter to measure arterial isotope activity and arterial oxygen content (O2con). The CMRO2 value was calculated as CBF x OEF x O2con. Transaxial slices were 6 mm thick. SPECT studies were performed by means of a ring-type gamma camera (Headtome SET-070; Shimadzu, Kyoto, Japan) with a 7.2-mm FWHM. Patients were intravenously injected with 740 MBq of 99mTc-HMPAO while resting in the supine position with their eyes covered in a dimly lit room. The post-injection interval was 10 min. Image data from 30-min acquisitions were collected into a 128 x 128 matrix with a high resolution collimator. All data were corrected for attenuation of 0.11 cm-1, and tomographic data were reconstructed with a filtered backprojection algorithm so that 5-mm thick transaxial slices were obtained. MRI were obtained with a 1.5 Tesla superconducting magnet (Siemens, Magnetom H15). The intensity of contrast enhancement in the infarcted areas after the administration of Gd-DTPA was graded as: none, faint, mild, moderate and marked.6 Tc-99m HMPAO brain SPECT and PET images were analyzed with MRI images for anatomical guidance. The regions of interest (ROIs) were placed manually over areas of increased accumulation in the infarct area and the contralateral cerebellum in 99mTc-HMPAO brain SPECT images. We also drew ROIs at the areas of PET images corresponding to the same areas in the SPECT images to obtain CBF, CMRO2, OEF and CBV values. The hyperfixation rate (HFR) was calculated as the surplus rate compared with CBF after standardization with the cerebellum shown as follows. (Where, Chf = mean counts in the hyperfixed area with 99mTc-HMPAO brain SPECT, Ccb = mean counts in the cerebellum with 99mTc-HMPAO brain SPECT, CBFhf = CBF in the hyperfixed area with PET, CBFcb = CBF in the cerebellum with PET.) All data are expressed as means +- SD. We compared all of the PET parameters for the patients and normal controls by the unpaired t-test and determined the relationships between the hyperfixation rate and the PET parameters by means of linear-regression analysis and the Pearson correlation coefficient. A two-tailed probability of < 0.05 was considered statistically significant. RESULTS The CBF, CBV, OEF and CMRO2 values obtained from the hyperfixed area of infarcts in four patients were 34.02 +- 6.30 ml/100g/min, 3.95 +- 1.76 ml/100g, 0.20 +- 0.03 and 0.99 +- 0.26 ml/100g/min, respectively, and the values obtained from 5 controls were 40.44 +- 5.01 ml/100g/min, 2.66 +- 0.71 ml/100g, 0.48 +- 0.09 and 3.28 +- 0.59 ml/100 g/min, respectively. The OEF and CMRO2 in the hyperfixed area of infarcts in four patients were significantly lower than those of normal controls (p < 0.01), whereas CBF and CBV values did not significantly differ from those of controls (Table 2 and Fig. 1). The CBF and CMRO2 ratios of the hyperfixed area of infarct to the normal brain were 0.84 +- 0.16 and 0.30 +- 0.08, respectively. MRI revealed moderate or marked Gd-DTPA enhancement at all hyperfixed areas of infarcts. The hyperfixation rate in four patients was 0.30 +- 0.15, which significantly correlated with CBV (r = 0.97, y = 11.75x + 0.42; p < 0.05), but not with CBF (r = 0.21 ), OEF (r = 0.61) or CMRO2 (r = 0.66, Fig. 2). DISCUSSION The hyperfixation in our 4 patients at the site of the infarct during the subacute phase from 7 to 28 days showed the findings of relative luxury perfusion and its surplus rate correlated well with CBV. We proposed comparing the surplus rate with CBF after standardization with the cerebellum, which might be a more accurate means of quantification in patients with a hyperfixed area. During the subacute phase, 99mTc-HMPAO hyperfixation appeared in regions of infarct due to reperfusion hyperemia, increased retention possibly related to increased blood brain barrier permeability, or retention in macrophages and leukocytes.3,4 Forty-eight percent of patients with subacute stroke had the hyperfixation by 99mTc-HMPAO of which the count exceeded that of the opposite side.3 The CBF in focal hyperfixation of the infarct was about 80% of that of healthy brain tissue.4 The hyperfixation by 99mTc-HMPAO in subacute infarct was reported as an almost equivalent to or increased in terms of CBF compared with unaffected areas where CMRO2 was decreased.7 The values of CMRO2 in our study were more depressed than those of CBF in areas of subacute infarct with focal hyperfixation, indicating relative luxury perfusion.8,9 We could suggested that one of the main factors which caused hyperfixation was vasodilatation in addition to the blood brain barrier disruption and the neovascularization as previously reported. During tracer transport, the unidirectional extraction fraction (E), which is the proportion of tracer extracted from the blood into tissue, depends on the capillary surface area (S) and the capillary permeability for the tracer (P), which is called the capillary permeability-surface product (PS). E increases with increasing S and P.10 The CBV displayed the grade of vasodilation, vascular space and capillary surface area and disruption of the blood brain barrier caused high capillary permeability of the tracer. The vasodilation indicated by the high CBV value and the disruption of the blood brain barrier that was proven by Gd-DTPA enhancement on MRI caused high PS. The high unidirectional extraction fraction caused by high PS in the hyperfixed area might therefore be the cause of the high retention fraction of 99mTc-HMPAO, and it was not correlated with CBF. The grade of hyperfixation in the present study was related to the vascular surface area, which was depicted with CBV, and which was proven by the significant correlation between the hyperfixation rate and CBV in the hyperfixed area. Shintai et al. also considered that hyperfixation is related to the increased HMPAO extraction caused by hyperpermeability across the blood-brain barrier.11 The present study revealed relative luxury perfusion and disruption of the blood brain barrier in hyperfixed areas of 99mTc-HMPAO brain SPECT images. In conclusion, the present study showed the hyperfixation by 99mTc-HMPAO between 7 and 28 days from the onset of stroke in the infarct area indicated relative luxury perfusion the rate of which closely correlated well with CBV. Proposed that one of the main factors which caused hyperfixation was vasodilatation as well as the blood brain barrier disruption and the neovascularization. REFERENCES 1. Podreka I, Suess E, Goldenberg G, Steiner M, Brucke T, Muller C, et al. Initial experience with technetium-99m HM-PAO brain SPECT. J Nucl Med 28: 1657-1666, 1987. 2. Moretti JL, Defer G, Cinotti L, Cesaro P, Degos JD, Vigneron N, et al. "Luxury perfusion" with 99mTc-HMPAO and 123I-IMP image during the subacute phase of stroke. Eur J Nucl Med 16: 17-22, 1990. 3. Sperling B, Lassen NA. Hyperfixation of HMPAO in subacute ischemic stroke leading to spuriously high estimates of cerebral blood flow by SPECT. Stroke 24: 193-194, 1993. 4. Sperling B, Lassen NA. Unreliability of CBF-SPECT in subacute ischemic stroke. Stroke 24: 1099-1100, 1993. 5. Frackowiak RSJ, Lenzi GL, Jones T, Heather JD. Quantitative measurement of regional cerebral blood flow and oxygen metabolism in man using 15O and positron emission tomography: Theory, procedure, and normal values. J Comput Assist Tomogr 4: 727-736, 1980. 6. Imakita S, Nishimura T, Yamada N, Naito H, Takamiya M, Yamada Y, et al. Magnetic resonance imaging of cerebral infarction: Time course of Gd-DTPA enhancement and CT comparison. Neuroradiology 30: 372-378, 1988. 7. Shishido F, Uemura K, Inugami A, Ogawa T, Fujita H, Shimosegawa E, et al. Discrepant 99mTc-ECD images of CBF in patients with subacute cerebral infarction: A comparison of CBF, CMRO2 and 99mTc-HMPAO imaging. Ann Nucl Med 9: 161-166, 1995. 8. Baron JC, Pathophysiology of acute cerebral ischemia: PET studies in humans. Cardiovasc Dis 1 (suppl 1): 22-31, 1991 9. Lassen NA. The luxury perfusion syndrome and its possible relation to acute metabolic acidosis localised within the brain. Lancet 19: 1113-1115, 1966. 10. Phelps ME, Mazziotta JC, Schelbert HR. Positron emission tomography and autoradiography; principles and applications for the brain and heart. New York; Raven press, pp. 304-307, 1986. 11. Shintani S, Tsuruoka S, Shiigai T. Paradoxical hyperfixation of hexamethylpropyleneamine oxime in cerebral infarction. Am J Neuroradiol 18: 1905-1908, 1997.