- Email: [email protected]
Pharmacokinetics of lithium in rat brain regions by spectroscopic imaging
Magnetic Resonance Imaging 23 (2005) 859 – 863
Pharmacokinetics of lithium in rat brain regions by spectroscopic imaging Subbaraya Ramaprasada,T, Elzbieta Rippa, Jiaxiong Pia,b, Melvin Lyonc a
Department of Radiology, University of Nebraska Medical Center, Omaha, NE 68198, USA Department of Computer Science, University of Nebraska at Omaha, Omaha, NE 68132, USA c Department of Biological Sciences, University of Southern California, Los Angeles, CA 90089, USA Received 1 October 2004; accepted 6 July 2005 b
Abstract Lithium (Li) and its salts have been demonstrated to be the most effective drug in both acute and prophylactic treatment of bipolar disorder. The exact molecular mechanisms and particular target regions accounting for its mood-stabilizing effect remain unknown. Knowledge of Li distribution and its regional pharmacokinetic properties in the living brain is of value in localizing its action in the brain. Pharmacokinetic measurements in different anatomical regions of the human brain are not yet available. Limited pharmacokinetic measurements in rat brain subvolumes have been performed using atomic absorption technique. However, a noninvasive way of estimating the pharmacokinetics in different regions of the brain where the drug exerts its beneficial effects would allow such methods to be used in the study of patients undergoing Li therapy. Earlier 7Li MR studies on rat brain regions have provided preliminary pharmacokinetic information from the whole brain. Using 7Li MR spectroscopic imaging (SI) technology, Li distribution in brain regions of the rat at therapeutic dosages has been recently demonstrated by us. Here we report feasibility of local pharmacokinetic measurements on brain regions obtained by magnetic resonance SI technology. Our results suggest that Li is most active in a region stretching from the anterior cingulate cortex and striatum to the caudal midbrain, with greatest activity including the preoptic area and hypothalamic region. Some activity was seen in prefrontal cortex, but only minimal amounts in the region of the cerebellum and metencephalic brainstem. D 2005 Elsevier Inc. All rights reserved. Keywords: Regional pharmacokinetics; Spectroscopic imaging (SI); Dual tuned RF coil; Bipolar disorder, Acute dosing; Rat brain
1. Introduction Lithium (Li) and its salts have been used as effective drugs in both acute and prophylactic treatment of mania and manic depressive illness [1]. Lithium is a centrally acting drug with significant beneficial effects on the individual’s behavior. It is important to find out if the brain regions of relevance to depression or mania are able to achieve the required Li concentration that will lead to desired biological effect. This result may largely depend upon the pharmacokinetics in that particular brain region. The regional specificity in Li brain distribution could underlie important steps in its action. Thus, studies to estimate Li disposition in different brain regions are of value in exploring the mechanism of behavioral effects. For example, subchronic treatment with Li enhances the nerve growth factor in distinct brain areas that control emotion and motor activity T Corresponding author. Tel.: +1 402 559 6990; fax: +1 402 559 1011. E-mail address: [email protected] (S. Ramaprasad). 0730-725X/$ – see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.mri.2005.07.007
that are of relevance to pathogenesis of depression [2]. Early studies on Li concentration in various brain regions [3–10] and pharmacokinetic measurements on rat whole brain, as well as select regions [11,12], made use of atomic absorption (AA) technique. Such AA studies have reported larger concentrations of Li in brain areas such as hypothalamus, striatum, cerebellum and diencephalons. The AA methods are destructive and hence provide results at one time point per subject. To follow more accurately the variation in local Li concentration with time in a single subject, a reliable, reproducible and accurate in vivo method for monitoring brain Li levels is necessary. Noninvasive MR methods such as spectroscopic imaging (SI) modality is very attractive as several regions of interest can be studied in a single experimental set up. With this view in mind, we are developing the 7Li MRSI method to determine the concentration and pharmacokinetic properties of Li in specific brain regions using a rat model. 7 Li MR technique has the ability to measure brain Li concentrations noninvasively (see Refs. [13,14] and related
860
S. Ramaprasad et al. / Magnetic Resonance Imaging 23 (2005) 859 – 863
references therein). Magnetic resonance techniques have been employed to study whole-brain Li in humans [15,16]. More recently, preliminary pharmacokinetic measurements on the whole brain in a single living animal have been demonstrated using localized Li magnetic resonance spectroscopic techniques [17]. The greatest advantage of MR techniques is that they do not involve ionizing radiation, and studies can be performed in a noninvasive manner. Additionally, methods like this will significantly reduce the number of animals to be studied as compared to destructive AA methodology. The availability of many high field animal imagers has provided investigators with a powerful technological tool to study concentration and pharmacokinetics of less sensitive nucleus such as Li in tissue volumes [18]. Recently, preliminary data on Li pharmacokinetics in rat brain have been presented by us [19], and in this study, we investigate in detail the pharmacokinetics of Li in rat brain regions using a 7-T animal imager. 2. Methods 2.1. Animal handling and preparation Male Sprague–Dawley rats weighing 250 to 400 g were housed under standard conditions on a 12-h light/ dark cycle and had free access to laboratory chow and tap water. The studies were performed according to the approved protocol from the Institutional animal welfare committee. An adjustment period of at least 1 week was allowed before the SI studies were performed. For MR studies, the rats were anesthetized in a compact plexiglass chamber by delivering 2.0% isoflurane with a mix of N2O and oxygen in 3:1 ratio. The anesthetized rats were placed on an animal holder in supine position. The holder was then placed inside the dual tuned RF coil unit inside the magnet bore such that the brain was close to the isocenter. The anesthetic was delivered through the nasal cone and the respiration rate was monitored throughout the study. An isoflurane level of 1% was used to keep the animals immobile. 2.2. Li administration Lithium was administered intraperitoneally as its chloride salt at a dose of 10 mEq/kg. The dose of 10 mEq/kg of LiCl was chosen based on the earlier studies on rats by AA studies where the dose ranged from 5 to 10 mEq [8 –10]. Lithium phantoms (2– 4 mM) were also prepared as external standards for calibrating the in vivo signals. 2.3. RF coil The dual tuned (1H-7Li) Litz coil built by Doty was used in these studies. The two-coil configuration is such that the inner coil tunes to 1H at 300.4 MHz and the outer coil to 116.6 MHz. The probe has a 68-mm inner diameter clear bore, an 80-mm window length, an outer diameter of
112 mm and a module length of ~210 mm, with RF coil centered within the length. 2.4. 1H and 7Li MR studies All MR studies were performed on the Bruker 7T animal imager operating at 300.0 MHz for 1H and 116.6 MHz for 7 Li nucleus. Before performing the Li MR studies, the magnetic field in the volume of interest was shimmed until the water proton line width of less than 0.4 ppm was achieved. For this purpose, both the manual and auto shim procedures were used as appropriate. Typically, the animal remained immobile from 6 to 7 h at a stretch during data acquisition, and a series of Li SI data were collected every hour using 2D SI technique in FID mode. The 1H scout images were acquired using spin echo technique. The acquisition parameters were number of averages of 1, TR of 1 s, echo time (TE) of 10 to 20 ms, FOV of 64 mm, a slice thickness of 4 mm and an acquisition matrix of 256 256. The data acquisition time was approximately 4 min. The acquisition parameters for Li SI study were a 528 sinc pulse of 578 As, TR of 1.0 s; number of acquisitions 16; slice thickness of 4 mm; and FOV of 64 mm for sagittal slices. The spectral width was 4000 Hz and the data acquisition size was 1024 points. The total acquisition time for a matrix size of 1616 was approximately 1 h. The T 1 measurements on rat brain (n = 4) and phantoms of known concentration were measured using the stimulated echo amplitude modulation sequence with variable mixing times (TM). The T 1’s on the phantoms were closely similar to the rat brain values. 2.4.1. Analysis of SI data All data processing and analysis were performed on a Silicon graphics work station using the Bruker Paravision software (version 2.1.1). The raw SI data were apodized with a 30-Hz line broadening in the time domain. The spectrum from each voxel was processed in the absolute phase mode, and the intensities were calibrated against data from a phantom voxel of known concentration (2– 4 mM). Both the phantom and rat data were acquired using an identical set of MR parameters and the voxel volumes maintained the same. 2.4.2. Lithium quantitation Lithium quantitation of various voxels was performed using the phantom replacement technique [20]. The Li intensity measured from each voxel was calibrated against the data from a phantom of known Li concentration. The average T 1’s for the brain and the phantom were used in the quantitative measurements of Li concentration. The mean value of the brain T 1 at this field strength was 2.2F0.3 s. The phantom T 1’s were adjusted to be the values expressed as milliequivalent per kilogram wet weight of the brain tissue that was sampled by MR
S. Ramaprasad et al. / Magnetic Resonance Imaging 23 (2005) 859 – 863
861
method. The equation pertaining to this calibration method is given below. ½Brain ¼ ½PhantomTS ðBrainÞ=S ðPhantomÞ where [Phantom] and [Brain] are Li concentrations of the phantom and the brain tissue sampled by MR. The S(Brain) and S(Phantom) are the MR signal intensities corrected for partial saturation using the T 1 values. Regional brain Li concentrations were obtained using the MR-derived voxel intensities and are expressed as milliequivalent per kilogram wet weight of the brain tissue. 2.4.3. Analysis of Li pharmacokinetics in brain The pharmacokinetic parameters were determined using the PK solution software [21], which relies on the use of noncompartmental methods of analysis for the estimation of pharmacokinetic parameters. The data were collected over a period of ~ 6 h during which time the entire process of absorption, distribution and elimination was not complete; hence, all the pharmacokinetic parameters were not accessible from the analysis. The pharmacokinetic profiles obtained for various voxels permitted the determination of the parameters such as the absorption half-life (t 1/2ab), the time at which maximum Li concentration occurs (T max).
3. Results and discussion The brain Li concentration from 15 voxels (n = 6 rats) were determined using the SI data collected at 90 min post Li administration. The mean observed value was 1.0F 0.5 mEq/kg with a maximum value of 2.2 mEq/kg and a minimum value of zero (represented by a void in the SI image). This is in the range of values observed for rats treated with similar dose and measured at 2 h post single dose administration [4,6,10,22,23]. 3.1. Spectroscopic imaging The SI localization method was first checked by using phantoms of different volume with Li solutions of known concentration. Following the acquisition of 1H scout image for anatomical referencing, spectroscopic images of rat brains (n =6) were obtained at various times following a single intraperitoneal dose administration of 10 mEq LiCl. Typically, the SI data were collected at 90, 152, 223, 292 and 356 min post Li administration. Because the rat brain remains virtually motionless during the entire period of data accumulation lasting over 6 to 7 h, accurate registration of the Li data on the anatomical image is straightforward and reliable, and does not need any further mathematical treatment. The sagittal 7Li SI spectral data so obtained are shown in Fig. 1 as an overlay on the corresponding 1 H anatomical image. The various brain voxels were identified with rat brain anatomical structures using a stereotaxic atlas [24]. An examination of the Li intensity in the brain regions indicates nonuniform Li distribution. The
Fig. 1. Lithium SI data represented as a map overlaid on the 1H anatomical image of the rat head. The thick yellow line tracks the brain shape as seen from the 1H image. Note that Li is distributed unevenly at this dose and is lowest in the cerebellum and met encephalic brainstem area of the brain and noticeably higher elsewhere.
brain areas such as cerebellum and metencephalic brain stem do not show any significant Li concentration. This was observed to be the case during the entire 6 to 7 h of data collection for all the rats studied. A detailed analysis of the signal intensities via integral values for the signals from each voxel showed a variation by a factor of 2 to 6. In general, the results, as shown in Fig. 1, indicate highest Li activity in septal/striatal and thalamic/midbrain regions including high activity in hypothalamus and preoptic regions. Less activity was found in prefrontal cortex and in the cerebellum and metencephalic brainstem regions. These results are in good agreement with the general findings in other studies, including those using the AA method [8,9,11]. 3.2. Pharmacokinetic measurements The feasibility studies of making pharmacokinetic measurements on small brain regions were performed on a group of six rats. The three representative voxels chosen for the study are shown in Fig. 2A, which displays Li signals from different regions of the rat head and brain. These three regions were chosen as they lie entirely within the brain, and there is little or no contamination from extra cranial Li. In these studies, the pharmacokinetic measurements were performed on Li signals from brain voxels of 64 Al (444 mm3). The pharmacokinetic profiles were constructed using data over 6 h for the three representative brain volumes 1 to 3 shown in Fig. 2A. At the spatial resolution that was achieved in these studies, the partial volume effects are minimal and, together with the averaging of data from six independent studies, provides a highly accurate depiction of Li pharmacokinetics in
862
S. Ramaprasad et al. / Magnetic Resonance Imaging 23 (2005) 859 – 863
Fig. 2. (A) The 7Li SI data overlaid on the anatomical reference image of a rat brain region. The voxel volume sampled was ~64 Al. The brain regions 1 to 3 considered for pharmacokinetic analysis, which lie completely inside the brain. (B) The various voxels represent the following brain structures: 1, olfactory bulb and related structures; 2, medial prefrontal cortex and caudate nucleus; 3, dorsal hippocampus and thalamus. Experimental Li intensity data are represented by symbols, and the solid lines that connect the points represent the PK profiles.
different brain regions. At roughly 90F10 min of an intraperitoneal administration, quantifiable Li signal could be observed. No appreciable changes in Li concentration were seen over the next 5 to 6 h as seen from the PK profiles shown in Fig. 2B. These results show a pattern of absorption and distribution similar to that observed on whole brain of Li-treated rats. However, the rates of uptake in each of the regions are faster than those observed in the whole-brain study [17]. The T max values observed here are significantly lower than those observed by others [8,11,12]. These changes may, in large part, be due to different dosages and different strains of rats used in the studies. The general results demonstrate that there are regional pharmacokinetic differences from different brain regions, which were also demonstrated by AA studies [8,11,12]. In these studies, we selected several brain regions and some representative head tissue regions for a comparative study. The average pharmacokinetic parameters obtained for each of the three regions of the Table 1 Structures enclosed by the three representative voxels inside the brain and the half-life of Li absorption measured from SI imaging studies Voxel number
Brain structures
t 1/2ab (min)
1
Olfactory bulb and related structures Medial prefrontal cortex and rostral caudate Fornix, septal region, anterior thalamus and posterior caudate
23.5
91
17.7
356
27.5
225
2 3
T max (min)
brain from six different studies are summarized in Table 1. The results from the tissue studies showed an average t 1/2ab of 55 min, which was similar to the value from the composite data representing the whole brain. In some studies that extended up to 5 h of continuous data collection, we noticed that local pharmacokinetics showed complete absorption and elimination during this period, whereas in many other regions, it was still increasing. In Fig. 2B, the PK profiles for regions 1 to 3 show a plateau region from 90 to 356 min. From the above, it is clear that the t 1/2ab values for various brain regions are substantially shorter than the values reported for the whole brain. These values are also substantially different than those reported by an earlier AA study, which, however, used male Wistar rats [10]. The values reported by another group [11] are closer to those reported here. The observed discrepancies may be attributable to different dosage levels, animal strains or sampling regions in the various studies. Furthermore, the smaller brain regions for which data are available by AA studies cannot yet be duplicated under the current spatial resolution of SI studies. In order to compare the results from the current study with the existing whole-brain data, we constructed composite PK profiles using data from all voxels that represent the brain. The composite pharmacokinetics data constructed for the entire brain was used to obtain t 1/2ab values. The average value of 54 min obtained in this study is compared reasonably with the value of 70 min obtained in the whole brain from an earlier study [12].
S. Ramaprasad et al. / Magnetic Resonance Imaging 23 (2005) 859 – 863
Our results show that the whole-brain PK values such as t 1/2ab obtained here agree reasonably with the earlier AA studies. 4. Conclusions This constitutes the first demonstration of regional pharmacokinetic measures on brain Li in a mammalian model using SI technique. The feasibility of performing SI studies at a high spatial resolution was tested on more than 10 rats. Pharmacokinetic measurements performed at similar time points post Li administration confirm that brain Li spectroscopic images can be generated over time at 7-T field strength after an acute dose of 10 mEq/kg Li. A nonuniform distribution is seen in the brain regions following an acute dose administration. The results also indicate that the pharmacokinetic parameters are different for different brain regions.
[7]
[8] [9]
[10]
[11]
[12]
[13] [14] [15]
Acknowledgments This research was supported in part by a grant from National Alliance for Research on Schizophrenia and Depression (NARSAD). References [1] Post RM. Alternatives to lithium for bipolar affective illness. In: Tasman A, Goldfinger CA, Kaufman CA, editors. Reviews of psychiatry, vol. 9. Washington (DC)7 American Psychiatric Press; 1990. p. 170. [2] Hellweg R, Lang UE, Nagel M, Baumgartner A. Subchronic treatment with lithium increases nerve growth factor content in distinct brain regions of adult rats. Mol Psychiatry 2002;7:604 – 8. [3] Ho AK, Gershon S, Pinckney L. The effects of acute and prolonged lithium treatment on the distribution of electrolytes potassium and sodium. Arch Int Pharmacodyn Ther 1970;186:54 – 65. [4] Bond PA, Brooks BA, Judd A. The distribution of lithium sodium and magnesium in rat brain and plasma after various periods of administration of lithium in the diet. Br J Pharmacol 1975;53:235 – 9. [5] Edelfors S. Distribution of sodium potassium and lithium in the brain of lithium treated rats. Acta Pharm Toxicol 1975;37:387 – 92. [6] McGovern AJ, Makanjoula R, Arbuthnott GW, Loudon JB, Glen AI. Lithium neurotoxicity. I. The concentration of lithium in dopaminergic
[16]
[17]
[18] [19]
[20] [21] [22]
[23]
[24]
863
systems of rat brain determined by flameless atomic absorption spectrophotometry. Acta Pharm Toxicol 1978;42:259 – 63. Plenge P. Lithium effects on rat brain glucose metabolism in vivo. Effects after administration of lithium by various routes. Psychopharmacology 1982;77:348 – 55. Ghoshdastidar D, Dutta RN, Poddar MK. In vivo distribution of lithium in plasma and brain. Indian J Exp Biol 1989;27:950 – 4. Rios C, Mendez RG. Determination of lithium in rat brain regions and synaptosomes by graphite furnace atomic absorption spectrometry. J Pharmacol Methods 1990;24:327 – 32. Ebadi MS, Simmons VJ, Hendrickson MJ, Lacy PS. Pharmacokinetics of lithium and its regional distribution in rat brain. Eur J Pharmacol 1974;27:324 – 9. Mukherjee BP, Bailey PT, Pradhan SN. Temporal and regional differences in brain concentration of lithium in rat. Psychopharmacology 1976;48:119 – 26. Mendez RG, Rios C, Jung H, Altagracia M. Pharmacokinetics of lithium in brain regions and plasma. Proc West Pharmacol Soc 1991;34:347 – 9. Ramaprasad S, Komoroski RA. NMR imaging and localized spectroscopy of lithium. Lithium 1994;5:127 – 38. Komoroski RA. Applications of 7Li NMR in biomedicine. Magn Reson Imaging 2000;8:103 – 16. Renshaw PF, Wicklund S. In vivo measurement of lithium in humans by nuclear magnetic resonance spectroscopy. Biol Psychiatry 1988;23:465 – 75. Komoroski RA, Newton JEO, Walker E, Cardwell D, Jagannathan NR, Ramaprasad S, et al. In vivo NMR spectroscopy of lithium-7 in humans. Magn Reson Med 1990;15:347 – 56. Ramaprasad S, Newton JEO, Cardwell D, Fowler AH, Komoroski RA. In vivo 7Li NMR imaging and localized spectroscopy of rat brain. Magn Reson Med 1992;25:308 – 18. Ramaprasad S. Lithium spectroscopic imaging of rat brain at therapeutic doses. Magn Reson Imaging 2004;22:727 – 34. Ramaprasad S, Ripp E, Lyon M. Regional pharmacokinetic measurements of lithium in rat brain by MRSI. 21st Annual meeting of ESMRMB. 2004. p. 25. Soher BJ, van Zijl PCM, Duyn JH, Barker PB. Quantitative proton imaging of the human brain. Magn Reson Med 1996;35:356 – 63. PK Solutions 2.0, Noncompartmental pharmacokinetics data analysis. Colorado7 Summit research Services; 2001. Ebara T, Smith DF. Lithium levels in blood platelets, serum, red blood cells and brain regions in rats given acute or chronic lithium salt treatments. J Psychiatr Res 1979;15:183 – 8. Morrison JM, Pritchard HD, Braude MC, D’Aguanno W. Plasma and brain lithium levels after lithium carbonate and lithium chloride administration by different routes in rats. Proc Soc Exp Biol Med 1971;137:889 – 92. Paxinos G, Watson C. The rat brain stereotaxic coordinates. 2nd ed. New York7 Academic Press; 1986.
Pharmacokinetics of lithium in rat brain regions by spectroscopic imaging Subbaraya Ramaprasada,T, Elzbieta Rippa, Jiaxiong Pia,b, Melvin Lyonc a
Department of Radiology, University of Nebraska Medical Center, Omaha, NE 68198, USA Department of Computer Science, University of Nebraska at Omaha, Omaha, NE 68132, USA c Department of Biological Sciences, University of Southern California, Los Angeles, CA 90089, USA Received 1 October 2004; accepted 6 July 2005 b
Abstract Lithium (Li) and its salts have been demonstrated to be the most effective drug in both acute and prophylactic treatment of bipolar disorder. The exact molecular mechanisms and particular target regions accounting for its mood-stabilizing effect remain unknown. Knowledge of Li distribution and its regional pharmacokinetic properties in the living brain is of value in localizing its action in the brain. Pharmacokinetic measurements in different anatomical regions of the human brain are not yet available. Limited pharmacokinetic measurements in rat brain subvolumes have been performed using atomic absorption technique. However, a noninvasive way of estimating the pharmacokinetics in different regions of the brain where the drug exerts its beneficial effects would allow such methods to be used in the study of patients undergoing Li therapy. Earlier 7Li MR studies on rat brain regions have provided preliminary pharmacokinetic information from the whole brain. Using 7Li MR spectroscopic imaging (SI) technology, Li distribution in brain regions of the rat at therapeutic dosages has been recently demonstrated by us. Here we report feasibility of local pharmacokinetic measurements on brain regions obtained by magnetic resonance SI technology. Our results suggest that Li is most active in a region stretching from the anterior cingulate cortex and striatum to the caudal midbrain, with greatest activity including the preoptic area and hypothalamic region. Some activity was seen in prefrontal cortex, but only minimal amounts in the region of the cerebellum and metencephalic brainstem. D 2005 Elsevier Inc. All rights reserved. Keywords: Regional pharmacokinetics; Spectroscopic imaging (SI); Dual tuned RF coil; Bipolar disorder, Acute dosing; Rat brain
1. Introduction Lithium (Li) and its salts have been used as effective drugs in both acute and prophylactic treatment of mania and manic depressive illness [1]. Lithium is a centrally acting drug with significant beneficial effects on the individual’s behavior. It is important to find out if the brain regions of relevance to depression or mania are able to achieve the required Li concentration that will lead to desired biological effect. This result may largely depend upon the pharmacokinetics in that particular brain region. The regional specificity in Li brain distribution could underlie important steps in its action. Thus, studies to estimate Li disposition in different brain regions are of value in exploring the mechanism of behavioral effects. For example, subchronic treatment with Li enhances the nerve growth factor in distinct brain areas that control emotion and motor activity T Corresponding author. Tel.: +1 402 559 6990; fax: +1 402 559 1011. E-mail address: [email protected] (S. Ramaprasad). 0730-725X/$ – see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.mri.2005.07.007
that are of relevance to pathogenesis of depression [2]. Early studies on Li concentration in various brain regions [3–10] and pharmacokinetic measurements on rat whole brain, as well as select regions [11,12], made use of atomic absorption (AA) technique. Such AA studies have reported larger concentrations of Li in brain areas such as hypothalamus, striatum, cerebellum and diencephalons. The AA methods are destructive and hence provide results at one time point per subject. To follow more accurately the variation in local Li concentration with time in a single subject, a reliable, reproducible and accurate in vivo method for monitoring brain Li levels is necessary. Noninvasive MR methods such as spectroscopic imaging (SI) modality is very attractive as several regions of interest can be studied in a single experimental set up. With this view in mind, we are developing the 7Li MRSI method to determine the concentration and pharmacokinetic properties of Li in specific brain regions using a rat model. 7 Li MR technique has the ability to measure brain Li concentrations noninvasively (see Refs. [13,14] and related
860
S. Ramaprasad et al. / Magnetic Resonance Imaging 23 (2005) 859 – 863
references therein). Magnetic resonance techniques have been employed to study whole-brain Li in humans [15,16]. More recently, preliminary pharmacokinetic measurements on the whole brain in a single living animal have been demonstrated using localized Li magnetic resonance spectroscopic techniques [17]. The greatest advantage of MR techniques is that they do not involve ionizing radiation, and studies can be performed in a noninvasive manner. Additionally, methods like this will significantly reduce the number of animals to be studied as compared to destructive AA methodology. The availability of many high field animal imagers has provided investigators with a powerful technological tool to study concentration and pharmacokinetics of less sensitive nucleus such as Li in tissue volumes [18]. Recently, preliminary data on Li pharmacokinetics in rat brain have been presented by us [19], and in this study, we investigate in detail the pharmacokinetics of Li in rat brain regions using a 7-T animal imager. 2. Methods 2.1. Animal handling and preparation Male Sprague–Dawley rats weighing 250 to 400 g were housed under standard conditions on a 12-h light/ dark cycle and had free access to laboratory chow and tap water. The studies were performed according to the approved protocol from the Institutional animal welfare committee. An adjustment period of at least 1 week was allowed before the SI studies were performed. For MR studies, the rats were anesthetized in a compact plexiglass chamber by delivering 2.0% isoflurane with a mix of N2O and oxygen in 3:1 ratio. The anesthetized rats were placed on an animal holder in supine position. The holder was then placed inside the dual tuned RF coil unit inside the magnet bore such that the brain was close to the isocenter. The anesthetic was delivered through the nasal cone and the respiration rate was monitored throughout the study. An isoflurane level of 1% was used to keep the animals immobile. 2.2. Li administration Lithium was administered intraperitoneally as its chloride salt at a dose of 10 mEq/kg. The dose of 10 mEq/kg of LiCl was chosen based on the earlier studies on rats by AA studies where the dose ranged from 5 to 10 mEq [8 –10]. Lithium phantoms (2– 4 mM) were also prepared as external standards for calibrating the in vivo signals. 2.3. RF coil The dual tuned (1H-7Li) Litz coil built by Doty was used in these studies. The two-coil configuration is such that the inner coil tunes to 1H at 300.4 MHz and the outer coil to 116.6 MHz. The probe has a 68-mm inner diameter clear bore, an 80-mm window length, an outer diameter of
112 mm and a module length of ~210 mm, with RF coil centered within the length. 2.4. 1H and 7Li MR studies All MR studies were performed on the Bruker 7T animal imager operating at 300.0 MHz for 1H and 116.6 MHz for 7 Li nucleus. Before performing the Li MR studies, the magnetic field in the volume of interest was shimmed until the water proton line width of less than 0.4 ppm was achieved. For this purpose, both the manual and auto shim procedures were used as appropriate. Typically, the animal remained immobile from 6 to 7 h at a stretch during data acquisition, and a series of Li SI data were collected every hour using 2D SI technique in FID mode. The 1H scout images were acquired using spin echo technique. The acquisition parameters were number of averages of 1, TR of 1 s, echo time (TE) of 10 to 20 ms, FOV of 64 mm, a slice thickness of 4 mm and an acquisition matrix of 256 256. The data acquisition time was approximately 4 min. The acquisition parameters for Li SI study were a 528 sinc pulse of 578 As, TR of 1.0 s; number of acquisitions 16; slice thickness of 4 mm; and FOV of 64 mm for sagittal slices. The spectral width was 4000 Hz and the data acquisition size was 1024 points. The total acquisition time for a matrix size of 1616 was approximately 1 h. The T 1 measurements on rat brain (n = 4) and phantoms of known concentration were measured using the stimulated echo amplitude modulation sequence with variable mixing times (TM). The T 1’s on the phantoms were closely similar to the rat brain values. 2.4.1. Analysis of SI data All data processing and analysis were performed on a Silicon graphics work station using the Bruker Paravision software (version 2.1.1). The raw SI data were apodized with a 30-Hz line broadening in the time domain. The spectrum from each voxel was processed in the absolute phase mode, and the intensities were calibrated against data from a phantom voxel of known concentration (2– 4 mM). Both the phantom and rat data were acquired using an identical set of MR parameters and the voxel volumes maintained the same. 2.4.2. Lithium quantitation Lithium quantitation of various voxels was performed using the phantom replacement technique [20]. The Li intensity measured from each voxel was calibrated against the data from a phantom of known Li concentration. The average T 1’s for the brain and the phantom were used in the quantitative measurements of Li concentration. The mean value of the brain T 1 at this field strength was 2.2F0.3 s. The phantom T 1’s were adjusted to be the values expressed as milliequivalent per kilogram wet weight of the brain tissue that was sampled by MR
S. Ramaprasad et al. / Magnetic Resonance Imaging 23 (2005) 859 – 863
861
method. The equation pertaining to this calibration method is given below. ½Brain ¼ ½PhantomTS ðBrainÞ=S ðPhantomÞ where [Phantom] and [Brain] are Li concentrations of the phantom and the brain tissue sampled by MR. The S(Brain) and S(Phantom) are the MR signal intensities corrected for partial saturation using the T 1 values. Regional brain Li concentrations were obtained using the MR-derived voxel intensities and are expressed as milliequivalent per kilogram wet weight of the brain tissue. 2.4.3. Analysis of Li pharmacokinetics in brain The pharmacokinetic parameters were determined using the PK solution software [21], which relies on the use of noncompartmental methods of analysis for the estimation of pharmacokinetic parameters. The data were collected over a period of ~ 6 h during which time the entire process of absorption, distribution and elimination was not complete; hence, all the pharmacokinetic parameters were not accessible from the analysis. The pharmacokinetic profiles obtained for various voxels permitted the determination of the parameters such as the absorption half-life (t 1/2ab), the time at which maximum Li concentration occurs (T max).
3. Results and discussion The brain Li concentration from 15 voxels (n = 6 rats) were determined using the SI data collected at 90 min post Li administration. The mean observed value was 1.0F 0.5 mEq/kg with a maximum value of 2.2 mEq/kg and a minimum value of zero (represented by a void in the SI image). This is in the range of values observed for rats treated with similar dose and measured at 2 h post single dose administration [4,6,10,22,23]. 3.1. Spectroscopic imaging The SI localization method was first checked by using phantoms of different volume with Li solutions of known concentration. Following the acquisition of 1H scout image for anatomical referencing, spectroscopic images of rat brains (n =6) were obtained at various times following a single intraperitoneal dose administration of 10 mEq LiCl. Typically, the SI data were collected at 90, 152, 223, 292 and 356 min post Li administration. Because the rat brain remains virtually motionless during the entire period of data accumulation lasting over 6 to 7 h, accurate registration of the Li data on the anatomical image is straightforward and reliable, and does not need any further mathematical treatment. The sagittal 7Li SI spectral data so obtained are shown in Fig. 1 as an overlay on the corresponding 1 H anatomical image. The various brain voxels were identified with rat brain anatomical structures using a stereotaxic atlas [24]. An examination of the Li intensity in the brain regions indicates nonuniform Li distribution. The
Fig. 1. Lithium SI data represented as a map overlaid on the 1H anatomical image of the rat head. The thick yellow line tracks the brain shape as seen from the 1H image. Note that Li is distributed unevenly at this dose and is lowest in the cerebellum and met encephalic brainstem area of the brain and noticeably higher elsewhere.
brain areas such as cerebellum and metencephalic brain stem do not show any significant Li concentration. This was observed to be the case during the entire 6 to 7 h of data collection for all the rats studied. A detailed analysis of the signal intensities via integral values for the signals from each voxel showed a variation by a factor of 2 to 6. In general, the results, as shown in Fig. 1, indicate highest Li activity in septal/striatal and thalamic/midbrain regions including high activity in hypothalamus and preoptic regions. Less activity was found in prefrontal cortex and in the cerebellum and metencephalic brainstem regions. These results are in good agreement with the general findings in other studies, including those using the AA method [8,9,11]. 3.2. Pharmacokinetic measurements The feasibility studies of making pharmacokinetic measurements on small brain regions were performed on a group of six rats. The three representative voxels chosen for the study are shown in Fig. 2A, which displays Li signals from different regions of the rat head and brain. These three regions were chosen as they lie entirely within the brain, and there is little or no contamination from extra cranial Li. In these studies, the pharmacokinetic measurements were performed on Li signals from brain voxels of 64 Al (444 mm3). The pharmacokinetic profiles were constructed using data over 6 h for the three representative brain volumes 1 to 3 shown in Fig. 2A. At the spatial resolution that was achieved in these studies, the partial volume effects are minimal and, together with the averaging of data from six independent studies, provides a highly accurate depiction of Li pharmacokinetics in
862
S. Ramaprasad et al. / Magnetic Resonance Imaging 23 (2005) 859 – 863
Fig. 2. (A) The 7Li SI data overlaid on the anatomical reference image of a rat brain region. The voxel volume sampled was ~64 Al. The brain regions 1 to 3 considered for pharmacokinetic analysis, which lie completely inside the brain. (B) The various voxels represent the following brain structures: 1, olfactory bulb and related structures; 2, medial prefrontal cortex and caudate nucleus; 3, dorsal hippocampus and thalamus. Experimental Li intensity data are represented by symbols, and the solid lines that connect the points represent the PK profiles.
different brain regions. At roughly 90F10 min of an intraperitoneal administration, quantifiable Li signal could be observed. No appreciable changes in Li concentration were seen over the next 5 to 6 h as seen from the PK profiles shown in Fig. 2B. These results show a pattern of absorption and distribution similar to that observed on whole brain of Li-treated rats. However, the rates of uptake in each of the regions are faster than those observed in the whole-brain study [17]. The T max values observed here are significantly lower than those observed by others [8,11,12]. These changes may, in large part, be due to different dosages and different strains of rats used in the studies. The general results demonstrate that there are regional pharmacokinetic differences from different brain regions, which were also demonstrated by AA studies [8,11,12]. In these studies, we selected several brain regions and some representative head tissue regions for a comparative study. The average pharmacokinetic parameters obtained for each of the three regions of the Table 1 Structures enclosed by the three representative voxels inside the brain and the half-life of Li absorption measured from SI imaging studies Voxel number
Brain structures
t 1/2ab (min)
1
Olfactory bulb and related structures Medial prefrontal cortex and rostral caudate Fornix, septal region, anterior thalamus and posterior caudate
23.5
91
17.7
356
27.5
225
2 3
T max (min)
brain from six different studies are summarized in Table 1. The results from the tissue studies showed an average t 1/2ab of 55 min, which was similar to the value from the composite data representing the whole brain. In some studies that extended up to 5 h of continuous data collection, we noticed that local pharmacokinetics showed complete absorption and elimination during this period, whereas in many other regions, it was still increasing. In Fig. 2B, the PK profiles for regions 1 to 3 show a plateau region from 90 to 356 min. From the above, it is clear that the t 1/2ab values for various brain regions are substantially shorter than the values reported for the whole brain. These values are also substantially different than those reported by an earlier AA study, which, however, used male Wistar rats [10]. The values reported by another group [11] are closer to those reported here. The observed discrepancies may be attributable to different dosage levels, animal strains or sampling regions in the various studies. Furthermore, the smaller brain regions for which data are available by AA studies cannot yet be duplicated under the current spatial resolution of SI studies. In order to compare the results from the current study with the existing whole-brain data, we constructed composite PK profiles using data from all voxels that represent the brain. The composite pharmacokinetics data constructed for the entire brain was used to obtain t 1/2ab values. The average value of 54 min obtained in this study is compared reasonably with the value of 70 min obtained in the whole brain from an earlier study [12].
S. Ramaprasad et al. / Magnetic Resonance Imaging 23 (2005) 859 – 863
Our results show that the whole-brain PK values such as t 1/2ab obtained here agree reasonably with the earlier AA studies. 4. Conclusions This constitutes the first demonstration of regional pharmacokinetic measures on brain Li in a mammalian model using SI technique. The feasibility of performing SI studies at a high spatial resolution was tested on more than 10 rats. Pharmacokinetic measurements performed at similar time points post Li administration confirm that brain Li spectroscopic images can be generated over time at 7-T field strength after an acute dose of 10 mEq/kg Li. A nonuniform distribution is seen in the brain regions following an acute dose administration. The results also indicate that the pharmacokinetic parameters are different for different brain regions.
[7]
[8] [9]
[10]
[11]
[12]
[13] [14] [15]
Acknowledgments This research was supported in part by a grant from National Alliance for Research on Schizophrenia and Depression (NARSAD). References [1] Post RM. Alternatives to lithium for bipolar affective illness. In: Tasman A, Goldfinger CA, Kaufman CA, editors. Reviews of psychiatry, vol. 9. Washington (DC)7 American Psychiatric Press; 1990. p. 170. [2] Hellweg R, Lang UE, Nagel M, Baumgartner A. Subchronic treatment with lithium increases nerve growth factor content in distinct brain regions of adult rats. Mol Psychiatry 2002;7:604 – 8. [3] Ho AK, Gershon S, Pinckney L. The effects of acute and prolonged lithium treatment on the distribution of electrolytes potassium and sodium. Arch Int Pharmacodyn Ther 1970;186:54 – 65. [4] Bond PA, Brooks BA, Judd A. The distribution of lithium sodium and magnesium in rat brain and plasma after various periods of administration of lithium in the diet. Br J Pharmacol 1975;53:235 – 9. [5] Edelfors S. Distribution of sodium potassium and lithium in the brain of lithium treated rats. Acta Pharm Toxicol 1975;37:387 – 92. [6] McGovern AJ, Makanjoula R, Arbuthnott GW, Loudon JB, Glen AI. Lithium neurotoxicity. I. The concentration of lithium in dopaminergic
[16]
[17]
[18] [19]
[20] [21] [22]
[23]
[24]
863
systems of rat brain determined by flameless atomic absorption spectrophotometry. Acta Pharm Toxicol 1978;42:259 – 63. Plenge P. Lithium effects on rat brain glucose metabolism in vivo. Effects after administration of lithium by various routes. Psychopharmacology 1982;77:348 – 55. Ghoshdastidar D, Dutta RN, Poddar MK. In vivo distribution of lithium in plasma and brain. Indian J Exp Biol 1989;27:950 – 4. Rios C, Mendez RG. Determination of lithium in rat brain regions and synaptosomes by graphite furnace atomic absorption spectrometry. J Pharmacol Methods 1990;24:327 – 32. Ebadi MS, Simmons VJ, Hendrickson MJ, Lacy PS. Pharmacokinetics of lithium and its regional distribution in rat brain. Eur J Pharmacol 1974;27:324 – 9. Mukherjee BP, Bailey PT, Pradhan SN. Temporal and regional differences in brain concentration of lithium in rat. Psychopharmacology 1976;48:119 – 26. Mendez RG, Rios C, Jung H, Altagracia M. Pharmacokinetics of lithium in brain regions and plasma. Proc West Pharmacol Soc 1991;34:347 – 9. Ramaprasad S, Komoroski RA. NMR imaging and localized spectroscopy of lithium. Lithium 1994;5:127 – 38. Komoroski RA. Applications of 7Li NMR in biomedicine. Magn Reson Imaging 2000;8:103 – 16. Renshaw PF, Wicklund S. In vivo measurement of lithium in humans by nuclear magnetic resonance spectroscopy. Biol Psychiatry 1988;23:465 – 75. Komoroski RA, Newton JEO, Walker E, Cardwell D, Jagannathan NR, Ramaprasad S, et al. In vivo NMR spectroscopy of lithium-7 in humans. Magn Reson Med 1990;15:347 – 56. Ramaprasad S, Newton JEO, Cardwell D, Fowler AH, Komoroski RA. In vivo 7Li NMR imaging and localized spectroscopy of rat brain. Magn Reson Med 1992;25:308 – 18. Ramaprasad S. Lithium spectroscopic imaging of rat brain at therapeutic doses. Magn Reson Imaging 2004;22:727 – 34. Ramaprasad S, Ripp E, Lyon M. Regional pharmacokinetic measurements of lithium in rat brain by MRSI. 21st Annual meeting of ESMRMB. 2004. p. 25. Soher BJ, van Zijl PCM, Duyn JH, Barker PB. Quantitative proton imaging of the human brain. Magn Reson Med 1996;35:356 – 63. PK Solutions 2.0, Noncompartmental pharmacokinetics data analysis. Colorado7 Summit research Services; 2001. Ebara T, Smith DF. Lithium levels in blood platelets, serum, red blood cells and brain regions in rats given acute or chronic lithium salt treatments. J Psychiatr Res 1979;15:183 – 8. Morrison JM, Pritchard HD, Braude MC, D’Aguanno W. Plasma and brain lithium levels after lithium carbonate and lithium chloride administration by different routes in rats. Proc Soc Exp Biol Med 1971;137:889 – 92. Paxinos G, Watson C. The rat brain stereotaxic coordinates. 2nd ed. New York7 Academic Press; 1986.