- Email: [email protected]
Visualisation of changes in regional cerebral blood flow (rCBF) produced by ketamine using long TE gradient-echo sequences: Preliminary results
Magnetic
Pergamon
Resonance Imaging, Vol. 13, No. 4, pp. 549-553, 1995 Copyright 0 1995 Elsevier Science Ltd Printed in the USA. All rights reserved 0730-725X195 $9.50 + .OO
0730-725X(95)00010-0
l Original Contribution VISUALISATION OF CHANGES IN REGIONAL CEREBRAL BLOOD FLOW (rCBF) PRODUCED BY KETAMINE USING LONG TE GRADIENT-ECHO SEQUENCES: PRELIMINARY RESULTS N.G.
BURDETT,* D.K. MENON,? T.A. CARPENTER,* J.G. JONES,? AND L.D. HALL* *Herchel Smith Laboratory for Medicinal Chemistry, Robinson Way, Cambridge CB2 2PZ, UK and tDepartment of Anaesthesia, University of Cambridge School for Clinical Medicine, Addenbrookes Hospital, Cambridge CB2 ZQQ, UK
Autoradiographic studies have shown that low dose ketamine produces increases in regional glucose utilisation and blood flow in the hippocampus, cerebral cortex, and olfactory lobe in the rat brain, probably due to antagonism at the NMDA receptor. Functional MRI using deoxyhaemoglobin contrast can be used to study changes in regional cerebral blood flow (rCBF). Long TE gradient-echo sequences were used to study changes in rCBF produced by low dose ketamine in rats anaesthetised with nitrous oxide, supplemented with either halothane (HAL) or fentanyl/fluanisone/midazolam (FFM) combination. Images from rats in the FFM group showed a lO-14% increase in signal intensity in the hippocampus, cerebral cortex, and olfactory lobe following either a single bolus or a low dose infusion of ketamine (p < .OS). These changes were significantly reduced in the HAL group (p < &OS). Halothane is known to attenuate the changes in regional glucose utilisation produced by the noncompetitive NMDA antagonist dizocilpine (MK-801), and its effects on ketamine-induced changes in rCBF seen in this study may be due to a similar effect. The potential use of functional MRI in studying the effect of pharmacological interventions on rCBF is discussed. Keywords: MRI; Functional imaging; Regional cerebral blood flow (rCBF); antagonists.
INTRODUCTION
l/28/94;
(NMDA)
pine (MK-801).6 All previous studies have used autoradiographic techniques,4-6 which are cumbersome when dynamic changes are being studied. The present study shows that functional MRI with GRE provides a convenient means of noninvasively monitoring changes in rCBF produced by ketamine in a rat model. Since halothane has previously been shown to attenuate rCBF changes produced by dizocilpine,’ the hypothesis that ketamine induced changes in rCBF would be attenuated by halothane was also tested.
The changes in signal intensity seen with neuronal activation in brain magnetic resonance (MR) images obtained using long TE gradient-echo sequences (GRE) have been well documented.’ These are believed to be due to reductions in the regional concentrations of deoxyhaemoglobin’ caused by increases in regional cerebral blood flow (rCBF) that exceed the increase in local oxygen requirements.3 In the rat, low doses of ketamine, a clinical anaesthetic agent, have been shown to produce increases in regional cerebral glucose utilisation in the hippocampus, cerebral cortex and olfactory bulb4p5; these changes in glucose utilisation are accompanied by by increases in rCBF.5 These effects on rCBF and glucose uptake produced by ketamine are thought to be due to effects at the N-methyl-D-aspartate (NMDA) receptor, since similar effects on rCBF are seen with the noncompetitive NMDA receptor antagonist dizocilRECEIVED
N-methyl-u-aspartate
METHODS These studies were carried out under a project licence approved by the Home Office of the UK Government (PPL 80/00509: NMR studies of anaesthetic effects in the brain). Eight Sprague-Dawley rats with a mean (range) weight of 375 (250-500)
g were studied; they
were maintained
lighting and thermal
in a controlled
Address correspondence to N.G. Burdett.
AccEPTED~/~/~~. 549
550
Magnetic Resonance Imaging 0 Volume 13,
environment and allowed free accessto food and water until the experiment began. Anaesthesia was induced in all eight animals using halothane in oxygen. In four animals (the HAL group), anaesthesia was maintained using 1% halothane in nitrous oxide and oxygen mixture with inspired oxygen concentration of 30%. In the remaining four animals (the HH group), anaesthesia was maintained with nitrous oxide and oxygen supplemented with an intravenous Hypnorm@/Hypnovel@ combination (consisting of fentanyll20 mg/kg, fluanisone 3.75 mg/kg, and midazolam 1.9 mg/kg). Body temperature was maintained at 36-38°C using a warming blanket. Six of those animals (three in each group) were intubated and artificially ventilated to control arterial carbon dioxide levels, and four of the ventilated animals (two in each group) underwent femoral arterial cannulation to provide continuous monitoring of blood pressure. Venous catheters were placed in all animals, either in a tail or femoral vein. Following induction of anaesthesia and placement of cannulae, the animals were transferred to a specially designed scanning pod, which incorporated an enveloping transmit/receive quadrature head coil with an internal diameter of 3.8 cm, and then placed in the magnet. Imaging was performed using gradient-echo sequence with an echo time (TE) of 60 ms and a repetition time (TR) of 1500 ms with two signal averages, using a 2.35 T, 30-cm horizontal bore magnet coupled to a Bruker Biospec II console. A pulse angle of 30” was used in most of these experiments except in the initial study in each group, where pulse angles of 50” were used. However, the pulse angle was not changed during the course of a given experiment. Single, 1.5-mm slice, coronal 128 x 256 images (FOV = 5 cm) were oriented using pilot scans, and two baseline images were acquired to assessreproducibility. Special efforts were made to identify major blood vessels in these heavily T,* -weighted images, and their location was noted so as to localise changes in signal intensity produced by change in vessel calibre, orientation or deoxyhaemoglobin content during activation. Following this, ketamine (Ketalar; Parke Davis) was delivered IV either as a 10 mg/kg bolus (two animals in each group), or a pharmacokinetically modelled low dose infusion (two animals in each group), designed to achieve and maintain plasma ketamine levels produced by a bolus dose of 5 mg/kg,4 and imaging was recommenced. The infusion protocol was as follows: 5 mg/kg over a 1min period followed by a variable rate infusion using an Ohmeda 9000 syringe pump (BOC Healthcare, UK). The animals then received the following percentages of the bolus infusion rate over the following period: 14.1%, O-l min; 6.8%, l-3 min; 5.8%, 3-5 min; 2.0%, 5-10min; 1.2%, lo-15 min; 0.7%, 15-20min.
Number
4, 1995
At 20 min the infusion was stopped; subsequent decrements in plasma ketamine concentrations over the succeeding 5 min have been calculated at less than 4%. This infusion regime thus provided for 20 min of imaging time at relatively stable blood ketamine concentrations. A single postketamine image was acquired in the animals that received the 10 mg bolus of ketamine, while two images were acquired in the animals that received the low dose infusion. In the latter animals imaging was repeated 20 min after the ketamine infusion was discontinued. In all instances no changes were made in ventilation parameters or anaesthetic levels once imaging was commenced. Signal intensities were calculated from regions of interest on the images using software written by Dr. N. J. Herod (of the Herchel Smith Laboratory for Medicinal Chemistry). The cortex was defined manually using spline curves. The ROI was chosen so as to avoid areas of partial volume, obvious vessels and the brain edge (Fig. 1). Percentage changes in signal intensities in the two groups were compared using Student’s t-test. Changes in signal intensities were also demonstrated by subtraction images. Since these preliminary experiments involved relatively low resolution images, we felt that demarcation of various anatomical areas was dif-
Fig. 1. Coronal image of rat brain showing a typical cortical ROI selected for analysis.
Ketamine-inducedchanges in rCBF
ficult , and likely to be prone to error. Quantitative measurements are therefore only presented for a cortical ROI. RESULTS Administration of ketamine resulted in a rise in mean arterial blood pressure in all animals where it was monitored, typically by 15-20 mmHg. Since satisfactory images were not obtained in one of the rats anaesthetised with halothane which received a bolus dose of ketamine, the results presented are from four rats in the FFM group and three rats in the HAL group. There was no significant difference between the two baseline images in either group of animals (Fig. 2A). However, following a bolus dose of ketamine, significant increases in signal intensity were seen in the hippocampus, cerebral cortex, and olfactory bulb (Fig. 2B). Marked increases in signal intensity were also seen in the posterior part of the eye, possibly due to increases in retinal blood flow. There were no significant increases in signal intensity in the rest of the brain or in the extracranial soft tissues. Similar changes in regional signal intensity were also seen when the rats were given a low dose ketamine infusion, but the changes were sustained rather than transient. Withdrawal of ketamine resulted in a reduction in signal intensity. These changes
(A)
0
N.G.
BURDETT
ET AL.
551
were less consistent and markedly reduced in magnitude when basal anaesthesia was maintained with halothane. The attenuation of ketamine-induced changes in signal intensity were seen in several areas including the hippocampus and the cortex. Changes in signal intensity in a region from within the cortex are shown in Fig. 3; these demonstrate clearly the difference in response to ketamine as a function of the agents used to maintain basal anaesthesia. Difference images of before and after the ketamine injection (Fig. 1B) show that the patterns of increase in signal intensity correspond to previous autoradiographic studies of changes in rCBF with ketamine.5 DISCUSSION The increases in regional signal intensity measured in images obtained using long TE gradient-echo images are believed to reflect increases in changes in deoxyhaemoglobin contrast produced by increases in regional blood flow. The magnitude of these changes was somewhat larger than might be expected at 2.35 T, but it is unlikely that they are artifactual for several reasons. First, their diffuse regional distribution did not correlate in location or pattern with major vessels identified on the baseline TT -weighted image, making it unlikely that the changes arose primarily within large vessels.
09
Fig. 2. Coronal images of rat brain: (A) Difference between two baseline images during anaesthesia with nitrous oxideand oxygen supplemented with fentanyl/fluanisone/midazolam. (B) Difference between baseline image and that obtained after low dose ketamine.
552
Magnetic Resonance Imaging 0 Volume 13, Number 4, 1995
Hypnorm
I Midazolam
I N,O
*#,
Halothane % change in SI over baseline (mean?SE)
Fig. 3. Changes in meanf SEM signalintensity in cortical ROIs similarto the oneshownin Fig. 1 (expressed asa percentage
of baseline)during two baselinescansand after administrationof low doseketaminein animalsanaesthetised with nitrous oxide and oxygen supplementedwith either halothane(0) or fentanyl/fluanisone/midazolam (0). The numbersin brackets representthe number of animalsincluded in the analysisin the FFM and HAL groups,respectively. *p < .05, postketamine vs. baseline;**p < 305, postketaminevs. baseline;#p < .005, halothanevs. fentanyl/fluanisone/midazolam.
Second, the low flip angles and long TR used would tend to reduce inflow effects from tissues or CSF that were outside the imaged slice. Third, the distribution of signal change agrees well with autoradiographic studies of ketamine induced increases in regional CBF and glucose uptake in the literature.4,5 Finally, the lack of significant changes in signal intensity in the bulk of the brain and muscle suggests that these changes are due to increased local flow and not to systemic haemodynamic changes caused by the ketamine. The results of the present study raise three important points. First, they show that the increases in regional glucose utilisation produced by ketamine (which have been demonstrated previously in several studies4,5) are accompanied by increases in rCBF. This suggests that ketamine does not disrupt the normal close relationship between cerebral metabolism and blood flow. In this respect, the present results confirm the results of Cavazzuti et al. ,5 who used [ 14C]iodoantipyrine to study changes in rCBF with ketamine. Second, this work suggests that halothane attenuates the rCBF changes produced by ketamine. The fact that halothane has similar effects on the rCBF effects produced by dizocilpine’ is open to one of two interpretations. One possibility is that halothane nonspecifically disrupts normal regional flow metabolism coupling, and hence, in the presence of halothane, the neuronal activation produced by ketamine is unaccompanied by commensurate changes in rCBF. However, if this were the case, an increase in local oxygen utilization in the face of an unchanged oxygen delivery would result in increases in regional deoxyhaemoglobin levels and hence reductions in regional signal intensity. No such effect was observed. The other possibility is
that the attenuation of ketamine and dizocilpine induced changes in rCBF and glucose utilisation by halothane represents a specific reversal of their effect at the NMDA receptor. It is important to note that the spatial pattern of modification by halothane of ketamineinduced changes in deoxyhaemoglobin contrast was not identical to its effects on regional changes in regional glucose utilization produced by dizocilpine. For example, Kurumaji and McCulloch’ reported that while halothane markedly reduced dizocilpine induced increase in hippocampal glucose utilisation, it had little effect in the cortex, where dizocilpine continued to result in reduced metabolism. However, our results show that halothane attenuated ketamine induced changes in deoxyhaemoglobin contrast (and, by inference, rCBF) both in the hippocampus and in the cortex (Fig. 2B). The reason for this discrepancy is not known, but may arise either from differences in methodology or differential regional effects of of the two agents on rCBF and glucose utilisation. Alternatively, the effects on cortical signal changes may be an independent effect of halothane. Finally, there is the more general question of the role of functional MRI in the study of cerebrovascular pharmacology. These experiments demonstrate that it is possible to detect changes in signal intensity in heavily 7’; -weighted MR images as a result of pharmacological interventions, and allow the comparison of different combinations of drugs. However, caution must be exercised interpreting these results. The MRI sequences used are prone to both motion artifact and “out of slice” effects; furthermore, changes in the signal from large vessels may also produce artifacts. Although careful experimental design may overcome many of these
Ketamine-inducedchangesin rCBF
problems, the use of deoxyhaemoglobin contrast for functional MR imaging suffers from several disadvantages. First, it can only detect changes in rCBF and cannot, at present, be used to measure absolute levels of blood flow at any given instant. Second, the quantitative measurement of changes in rCBF using long TE GRE will require either the quantification of susceptibility changes in blood at 2.35 T as a function of oxygen saturation, or calibration against some other technique such as hydrogen clearance or autoradiography. Even if such calibration were possible, the technique will depend on the maintenance of normal flow metabolism relationships, which may be variably disrupted by pharmacological agents that have specific vascular effects. These restrictions on the utility of deoxyhaemoglobin contrast suggest that other techniques of functional MRI, such as rapid imaging with intravenous gadopentate dimeglumine’ or the method of arterial spin inversion,’ may be more useful in some situations. Acknowledgments-Grateful thanks are dueto Nikki Sibsonfor her help with these experiments. This work was supported by a munificent benefaction from Dr Herchel Smith (LDH, TAC) and by a research studentship (NGB). REFERENCES 1. Kwong , K.K. ; Belliveau, J.W. ; Chesler, D. A, ; Goldberg, I.E.; Weisskoff, R.M.; Poncelet, B.P.; Kennedy, D.N.; Hoppel, B.E.; Cohen, M.S.; Turner, R.; Cheng, H.-M.; Brady, T. J.; Rosen, B.R. Dynamic magnetic resonance im-
2.
3.
4. 5.
6.
0
N.G.
BURDETT ET AL.
553
aging of human brain activity during primary sensory stimulation. Proc. Natl. Acad. Sci. 89:5675-5679; 1992. Thulborn, K.R. ; Waterton, J. C. ; Matthews, P.M.; Radda, G.K. Oxygenation dependence of the transverse relaxation time of water protons in whole blood at high field. Biochim. Biophys. Acta 714:265-270; 1982. Ogawa, S.; Lee, T.; Nayak, AS.; Glynn, P. Oxygenationsensitive contrast in magnetic resonance image of rodent brain at high magnetic fields. Magn. Reson. Med. 14:6878; 1990. Davis, D.W.; Mans,A.M.; Biebuyck, J.F.; Hawkins, R.A. The influence of ketamine on regional brain glucose use. Anesthesiology 69:199-205; 1988. Cavazzuti, M.; Porro, C.A.; Biral, G.P.; Benassi, C.; Barbieri, G.C. Ketamine effects on local cerebral blood flow and metabolism in the rat. J. Cereb. Blood Flow Metab. 7:806-811; 1987. Kurumaji, A.; Nehls, D.G.; Park, C.K.; McCulloch, J. Effects of NMDA antagonists, MK-801 and CPP on cerebral glucose use. Brain Res. 496:268-284; 1989. Kurumaji, A.; McCulloch, J. Effects of MK-801 upon local cerebral glucose utilisation in conscious rats and rats anaesthetised with halothane. J. Cereb. Blood Flow Metab. 9:786-794; 1989. Belliveau, J.W.; Kennedy, D.N.; McKinstry, R.C.; Buchbinder, B.R.; Weisskoff, R.M.; Cohen, M.S.; Vevea, J.M.; Brady, T.J.; Rosen, B.R. Functional mapping of the human visual cortex by magnetic resonance imaging. Science 254:716-719; 1991. Roberts, D.A.; Detre, J.A.; Bolinger, L.; Insko, E.K.; Leigh, J.S. Quantitative magnetic resonance imaging of human brain perfusion at 1.5 T using steady-state inversion of arterial water. Proc. Natl. Acad. Sci. USA 91: 33-37; 1994.
Pergamon
Resonance Imaging, Vol. 13, No. 4, pp. 549-553, 1995 Copyright 0 1995 Elsevier Science Ltd Printed in the USA. All rights reserved 0730-725X195 $9.50 + .OO
0730-725X(95)00010-0
l Original Contribution VISUALISATION OF CHANGES IN REGIONAL CEREBRAL BLOOD FLOW (rCBF) PRODUCED BY KETAMINE USING LONG TE GRADIENT-ECHO SEQUENCES: PRELIMINARY RESULTS N.G.
BURDETT,* D.K. MENON,? T.A. CARPENTER,* J.G. JONES,? AND L.D. HALL* *Herchel Smith Laboratory for Medicinal Chemistry, Robinson Way, Cambridge CB2 2PZ, UK and tDepartment of Anaesthesia, University of Cambridge School for Clinical Medicine, Addenbrookes Hospital, Cambridge CB2 ZQQ, UK
Autoradiographic studies have shown that low dose ketamine produces increases in regional glucose utilisation and blood flow in the hippocampus, cerebral cortex, and olfactory lobe in the rat brain, probably due to antagonism at the NMDA receptor. Functional MRI using deoxyhaemoglobin contrast can be used to study changes in regional cerebral blood flow (rCBF). Long TE gradient-echo sequences were used to study changes in rCBF produced by low dose ketamine in rats anaesthetised with nitrous oxide, supplemented with either halothane (HAL) or fentanyl/fluanisone/midazolam (FFM) combination. Images from rats in the FFM group showed a lO-14% increase in signal intensity in the hippocampus, cerebral cortex, and olfactory lobe following either a single bolus or a low dose infusion of ketamine (p < .OS). These changes were significantly reduced in the HAL group (p < &OS). Halothane is known to attenuate the changes in regional glucose utilisation produced by the noncompetitive NMDA antagonist dizocilpine (MK-801), and its effects on ketamine-induced changes in rCBF seen in this study may be due to a similar effect. The potential use of functional MRI in studying the effect of pharmacological interventions on rCBF is discussed. Keywords: MRI; Functional imaging; Regional cerebral blood flow (rCBF); antagonists.
INTRODUCTION
l/28/94;
(NMDA)
pine (MK-801).6 All previous studies have used autoradiographic techniques,4-6 which are cumbersome when dynamic changes are being studied. The present study shows that functional MRI with GRE provides a convenient means of noninvasively monitoring changes in rCBF produced by ketamine in a rat model. Since halothane has previously been shown to attenuate rCBF changes produced by dizocilpine,’ the hypothesis that ketamine induced changes in rCBF would be attenuated by halothane was also tested.
The changes in signal intensity seen with neuronal activation in brain magnetic resonance (MR) images obtained using long TE gradient-echo sequences (GRE) have been well documented.’ These are believed to be due to reductions in the regional concentrations of deoxyhaemoglobin’ caused by increases in regional cerebral blood flow (rCBF) that exceed the increase in local oxygen requirements.3 In the rat, low doses of ketamine, a clinical anaesthetic agent, have been shown to produce increases in regional cerebral glucose utilisation in the hippocampus, cerebral cortex and olfactory bulb4p5; these changes in glucose utilisation are accompanied by by increases in rCBF.5 These effects on rCBF and glucose uptake produced by ketamine are thought to be due to effects at the N-methyl-D-aspartate (NMDA) receptor, since similar effects on rCBF are seen with the noncompetitive NMDA receptor antagonist dizocilRECEIVED
N-methyl-u-aspartate
METHODS These studies were carried out under a project licence approved by the Home Office of the UK Government (PPL 80/00509: NMR studies of anaesthetic effects in the brain). Eight Sprague-Dawley rats with a mean (range) weight of 375 (250-500)
g were studied; they
were maintained
lighting and thermal
in a controlled
Address correspondence to N.G. Burdett.
AccEPTED~/~/~~. 549
550
Magnetic Resonance Imaging 0 Volume 13,
environment and allowed free accessto food and water until the experiment began. Anaesthesia was induced in all eight animals using halothane in oxygen. In four animals (the HAL group), anaesthesia was maintained using 1% halothane in nitrous oxide and oxygen mixture with inspired oxygen concentration of 30%. In the remaining four animals (the HH group), anaesthesia was maintained with nitrous oxide and oxygen supplemented with an intravenous Hypnorm@/Hypnovel@ combination (consisting of fentanyll20 mg/kg, fluanisone 3.75 mg/kg, and midazolam 1.9 mg/kg). Body temperature was maintained at 36-38°C using a warming blanket. Six of those animals (three in each group) were intubated and artificially ventilated to control arterial carbon dioxide levels, and four of the ventilated animals (two in each group) underwent femoral arterial cannulation to provide continuous monitoring of blood pressure. Venous catheters were placed in all animals, either in a tail or femoral vein. Following induction of anaesthesia and placement of cannulae, the animals were transferred to a specially designed scanning pod, which incorporated an enveloping transmit/receive quadrature head coil with an internal diameter of 3.8 cm, and then placed in the magnet. Imaging was performed using gradient-echo sequence with an echo time (TE) of 60 ms and a repetition time (TR) of 1500 ms with two signal averages, using a 2.35 T, 30-cm horizontal bore magnet coupled to a Bruker Biospec II console. A pulse angle of 30” was used in most of these experiments except in the initial study in each group, where pulse angles of 50” were used. However, the pulse angle was not changed during the course of a given experiment. Single, 1.5-mm slice, coronal 128 x 256 images (FOV = 5 cm) were oriented using pilot scans, and two baseline images were acquired to assessreproducibility. Special efforts were made to identify major blood vessels in these heavily T,* -weighted images, and their location was noted so as to localise changes in signal intensity produced by change in vessel calibre, orientation or deoxyhaemoglobin content during activation. Following this, ketamine (Ketalar; Parke Davis) was delivered IV either as a 10 mg/kg bolus (two animals in each group), or a pharmacokinetically modelled low dose infusion (two animals in each group), designed to achieve and maintain plasma ketamine levels produced by a bolus dose of 5 mg/kg,4 and imaging was recommenced. The infusion protocol was as follows: 5 mg/kg over a 1min period followed by a variable rate infusion using an Ohmeda 9000 syringe pump (BOC Healthcare, UK). The animals then received the following percentages of the bolus infusion rate over the following period: 14.1%, O-l min; 6.8%, l-3 min; 5.8%, 3-5 min; 2.0%, 5-10min; 1.2%, lo-15 min; 0.7%, 15-20min.
Number
4, 1995
At 20 min the infusion was stopped; subsequent decrements in plasma ketamine concentrations over the succeeding 5 min have been calculated at less than 4%. This infusion regime thus provided for 20 min of imaging time at relatively stable blood ketamine concentrations. A single postketamine image was acquired in the animals that received the 10 mg bolus of ketamine, while two images were acquired in the animals that received the low dose infusion. In the latter animals imaging was repeated 20 min after the ketamine infusion was discontinued. In all instances no changes were made in ventilation parameters or anaesthetic levels once imaging was commenced. Signal intensities were calculated from regions of interest on the images using software written by Dr. N. J. Herod (of the Herchel Smith Laboratory for Medicinal Chemistry). The cortex was defined manually using spline curves. The ROI was chosen so as to avoid areas of partial volume, obvious vessels and the brain edge (Fig. 1). Percentage changes in signal intensities in the two groups were compared using Student’s t-test. Changes in signal intensities were also demonstrated by subtraction images. Since these preliminary experiments involved relatively low resolution images, we felt that demarcation of various anatomical areas was dif-
Fig. 1. Coronal image of rat brain showing a typical cortical ROI selected for analysis.
Ketamine-inducedchanges in rCBF
ficult , and likely to be prone to error. Quantitative measurements are therefore only presented for a cortical ROI. RESULTS Administration of ketamine resulted in a rise in mean arterial blood pressure in all animals where it was monitored, typically by 15-20 mmHg. Since satisfactory images were not obtained in one of the rats anaesthetised with halothane which received a bolus dose of ketamine, the results presented are from four rats in the FFM group and three rats in the HAL group. There was no significant difference between the two baseline images in either group of animals (Fig. 2A). However, following a bolus dose of ketamine, significant increases in signal intensity were seen in the hippocampus, cerebral cortex, and olfactory bulb (Fig. 2B). Marked increases in signal intensity were also seen in the posterior part of the eye, possibly due to increases in retinal blood flow. There were no significant increases in signal intensity in the rest of the brain or in the extracranial soft tissues. Similar changes in regional signal intensity were also seen when the rats were given a low dose ketamine infusion, but the changes were sustained rather than transient. Withdrawal of ketamine resulted in a reduction in signal intensity. These changes
(A)
0
N.G.
BURDETT
ET AL.
551
were less consistent and markedly reduced in magnitude when basal anaesthesia was maintained with halothane. The attenuation of ketamine-induced changes in signal intensity were seen in several areas including the hippocampus and the cortex. Changes in signal intensity in a region from within the cortex are shown in Fig. 3; these demonstrate clearly the difference in response to ketamine as a function of the agents used to maintain basal anaesthesia. Difference images of before and after the ketamine injection (Fig. 1B) show that the patterns of increase in signal intensity correspond to previous autoradiographic studies of changes in rCBF with ketamine.5 DISCUSSION The increases in regional signal intensity measured in images obtained using long TE gradient-echo images are believed to reflect increases in changes in deoxyhaemoglobin contrast produced by increases in regional blood flow. The magnitude of these changes was somewhat larger than might be expected at 2.35 T, but it is unlikely that they are artifactual for several reasons. First, their diffuse regional distribution did not correlate in location or pattern with major vessels identified on the baseline TT -weighted image, making it unlikely that the changes arose primarily within large vessels.
09
Fig. 2. Coronal images of rat brain: (A) Difference between two baseline images during anaesthesia with nitrous oxideand oxygen supplemented with fentanyl/fluanisone/midazolam. (B) Difference between baseline image and that obtained after low dose ketamine.
552
Magnetic Resonance Imaging 0 Volume 13, Number 4, 1995
Hypnorm
I Midazolam
I N,O
*#,
Halothane % change in SI over baseline (mean?SE)
Fig. 3. Changes in meanf SEM signalintensity in cortical ROIs similarto the oneshownin Fig. 1 (expressed asa percentage
of baseline)during two baselinescansand after administrationof low doseketaminein animalsanaesthetised with nitrous oxide and oxygen supplementedwith either halothane(0) or fentanyl/fluanisone/midazolam (0). The numbersin brackets representthe number of animalsincluded in the analysisin the FFM and HAL groups,respectively. *p < .05, postketamine vs. baseline;**p < 305, postketaminevs. baseline;#p < .005, halothanevs. fentanyl/fluanisone/midazolam.
Second, the low flip angles and long TR used would tend to reduce inflow effects from tissues or CSF that were outside the imaged slice. Third, the distribution of signal change agrees well with autoradiographic studies of ketamine induced increases in regional CBF and glucose uptake in the literature.4,5 Finally, the lack of significant changes in signal intensity in the bulk of the brain and muscle suggests that these changes are due to increased local flow and not to systemic haemodynamic changes caused by the ketamine. The results of the present study raise three important points. First, they show that the increases in regional glucose utilisation produced by ketamine (which have been demonstrated previously in several studies4,5) are accompanied by increases in rCBF. This suggests that ketamine does not disrupt the normal close relationship between cerebral metabolism and blood flow. In this respect, the present results confirm the results of Cavazzuti et al. ,5 who used [ 14C]iodoantipyrine to study changes in rCBF with ketamine. Second, this work suggests that halothane attenuates the rCBF changes produced by ketamine. The fact that halothane has similar effects on the rCBF effects produced by dizocilpine’ is open to one of two interpretations. One possibility is that halothane nonspecifically disrupts normal regional flow metabolism coupling, and hence, in the presence of halothane, the neuronal activation produced by ketamine is unaccompanied by commensurate changes in rCBF. However, if this were the case, an increase in local oxygen utilization in the face of an unchanged oxygen delivery would result in increases in regional deoxyhaemoglobin levels and hence reductions in regional signal intensity. No such effect was observed. The other possibility is
that the attenuation of ketamine and dizocilpine induced changes in rCBF and glucose utilisation by halothane represents a specific reversal of their effect at the NMDA receptor. It is important to note that the spatial pattern of modification by halothane of ketamineinduced changes in deoxyhaemoglobin contrast was not identical to its effects on regional changes in regional glucose utilization produced by dizocilpine. For example, Kurumaji and McCulloch’ reported that while halothane markedly reduced dizocilpine induced increase in hippocampal glucose utilisation, it had little effect in the cortex, where dizocilpine continued to result in reduced metabolism. However, our results show that halothane attenuated ketamine induced changes in deoxyhaemoglobin contrast (and, by inference, rCBF) both in the hippocampus and in the cortex (Fig. 2B). The reason for this discrepancy is not known, but may arise either from differences in methodology or differential regional effects of of the two agents on rCBF and glucose utilisation. Alternatively, the effects on cortical signal changes may be an independent effect of halothane. Finally, there is the more general question of the role of functional MRI in the study of cerebrovascular pharmacology. These experiments demonstrate that it is possible to detect changes in signal intensity in heavily 7’; -weighted MR images as a result of pharmacological interventions, and allow the comparison of different combinations of drugs. However, caution must be exercised interpreting these results. The MRI sequences used are prone to both motion artifact and “out of slice” effects; furthermore, changes in the signal from large vessels may also produce artifacts. Although careful experimental design may overcome many of these
Ketamine-inducedchangesin rCBF
problems, the use of deoxyhaemoglobin contrast for functional MR imaging suffers from several disadvantages. First, it can only detect changes in rCBF and cannot, at present, be used to measure absolute levels of blood flow at any given instant. Second, the quantitative measurement of changes in rCBF using long TE GRE will require either the quantification of susceptibility changes in blood at 2.35 T as a function of oxygen saturation, or calibration against some other technique such as hydrogen clearance or autoradiography. Even if such calibration were possible, the technique will depend on the maintenance of normal flow metabolism relationships, which may be variably disrupted by pharmacological agents that have specific vascular effects. These restrictions on the utility of deoxyhaemoglobin contrast suggest that other techniques of functional MRI, such as rapid imaging with intravenous gadopentate dimeglumine’ or the method of arterial spin inversion,’ may be more useful in some situations. Acknowledgments-Grateful thanks are dueto Nikki Sibsonfor her help with these experiments. This work was supported by a munificent benefaction from Dr Herchel Smith (LDH, TAC) and by a research studentship (NGB). REFERENCES 1. Kwong , K.K. ; Belliveau, J.W. ; Chesler, D. A, ; Goldberg, I.E.; Weisskoff, R.M.; Poncelet, B.P.; Kennedy, D.N.; Hoppel, B.E.; Cohen, M.S.; Turner, R.; Cheng, H.-M.; Brady, T. J.; Rosen, B.R. Dynamic magnetic resonance im-
2.
3.
4. 5.
6.
0
N.G.
BURDETT ET AL.
553
aging of human brain activity during primary sensory stimulation. Proc. Natl. Acad. Sci. 89:5675-5679; 1992. Thulborn, K.R. ; Waterton, J. C. ; Matthews, P.M.; Radda, G.K. Oxygenation dependence of the transverse relaxation time of water protons in whole blood at high field. Biochim. Biophys. Acta 714:265-270; 1982. Ogawa, S.; Lee, T.; Nayak, AS.; Glynn, P. Oxygenationsensitive contrast in magnetic resonance image of rodent brain at high magnetic fields. Magn. Reson. Med. 14:6878; 1990. Davis, D.W.; Mans,A.M.; Biebuyck, J.F.; Hawkins, R.A. The influence of ketamine on regional brain glucose use. Anesthesiology 69:199-205; 1988. Cavazzuti, M.; Porro, C.A.; Biral, G.P.; Benassi, C.; Barbieri, G.C. Ketamine effects on local cerebral blood flow and metabolism in the rat. J. Cereb. Blood Flow Metab. 7:806-811; 1987. Kurumaji, A.; Nehls, D.G.; Park, C.K.; McCulloch, J. Effects of NMDA antagonists, MK-801 and CPP on cerebral glucose use. Brain Res. 496:268-284; 1989. Kurumaji, A.; McCulloch, J. Effects of MK-801 upon local cerebral glucose utilisation in conscious rats and rats anaesthetised with halothane. J. Cereb. Blood Flow Metab. 9:786-794; 1989. Belliveau, J.W.; Kennedy, D.N.; McKinstry, R.C.; Buchbinder, B.R.; Weisskoff, R.M.; Cohen, M.S.; Vevea, J.M.; Brady, T.J.; Rosen, B.R. Functional mapping of the human visual cortex by magnetic resonance imaging. Science 254:716-719; 1991. Roberts, D.A.; Detre, J.A.; Bolinger, L.; Insko, E.K.; Leigh, J.S. Quantitative magnetic resonance imaging of human brain perfusion at 1.5 T using steady-state inversion of arterial water. Proc. Natl. Acad. Sci. USA 91: 33-37; 1994.