Radiation-Induced Brain Metabolic Changes in the Acute and Early Delayed Phase Detected With Quantitative Proton Magnetic Resonance Spectroscopy

Transcription

1 ORIGINAL ARTICLE Radiation-Induced Brain Metabolic Changes in the Acute and Early Delayed Phase Detected With Quantitative Proton Magnetic Resonance Spectroscopy Tatsuro Kaminaga, MD and Katsuo Shirai, MD Objective: Quantitative proton magnetic resonance spectroscopy (MRS) was performed before and after radiation therapy to estimate its usefulness for evaluating radiation-induced metabolic brain changes. Methods: Twenty patients with multiple brain metastases not having received any previous brain radiation were selected for the study. The total radiation dose varied from 40 (20 fractions) to 50 (25 fractions) Gy, with an opposition technique. MRS was performed just before irradiation, during the acute phase (n = 20, days) and in the early delayed phase (n = 15, months) after radiation. The concentration of N-acetyl-L-aspartate (NAA), choline-containing substance (Cho), and creatine/phosphocreatine (Cr) was quantified. Results: The concentration of NAA decreased (P = 0.05 versus before radiation), and the concentration of Cho increased (P = versus before radiation) during the early delayed phase. The concentration of Cr was not changed before or after radiation. Conclusions: Radiation-induced changes in brain metabolism were well detected with quantitative MRS in the early delayed phase. Quantitative MRS is a novel tool for estimating radiation-induced neurotoxicity. Key Words: magnetic resonance, spectroscopy, radiation, injurious effect (J Comput Assist Tomogr 2005;29: ) It is established that proton magnetic resonance spectroscopy (MRS) is an advantageous tool for noninvasive estimation of in vivo cerebral metabolite concentration. Recently, quantification methods for cerebral metabolite concentration using MRS were proposed 1 7 to enable more precise observation of time-serial changes in cerebral metabolite concentration in the same patient. On the other hand, radiation therapy is a useful method for reduction of metastatic brain tumors. However, irradiation also causes damage to the normal cerebral parenchyma, which is divided into three phases: acute, early, and Received for publication December 12, 2004; accepted February 22, From the Department of Radiology, Teikyo University Medical School, Tokyo, Japan. Reprints: Tatsuro Kaminaga, Department of Radiology, Teikyo University Medical School, Kaga, Itabashi-ku, Tokyo, , Japan ( kami@med.teikyo-.ac.jp). Copyright Ó 2005 by Lippincott Williams & Wilkins late delayed reaction. Acute and early delayed reactions are associated with cerebral parenchyma edema 8,9 and demyelination. 8,10 These reactions are usually transient. The late delayed reaction is irreversible and well known as delayed cerebral necrosis. 8 For imaging of radiation-induced brain metabolic changes, computed tomography, magnetic resonance imaging (MRI), and brain single photon emission tomography have been used. However, none of these methods can measure radiation-induced changes in cerebral metabolite concentration that can be detected by MRS. It is important to measure cerebral metabolite concentration for precise estimation of radiation-induced changes after therapeutic brain irradiation. As shown in previous reports, MRS is considered to be a useful tool for estimating radiation-induced brain metabolic changes. However, many reports did not quantify metabolite concentration. To our knowledge, this is the initial report concerning sequential measurement of N-acetyl-Laspartate (NAA), choline-containing substance (Cho), and creatine/phosphocreatine (Cr) concentration before and in acute and early-delayed phase. MATERIALS AND METHODS This study was performed from 1996 to 2000 under the approval of the ethics committee of Teikyo University Medical School and written informed consent was obtained before every examination. This study also conformed to the guidelines of the Helsinki Declaration, Twenty patients with multiple brain metastases were sequentially examined with MRI and MRS, just before irradiation, in the acute phase (n = 20, days) and in the early delayed phase (n = 15, month) after irradiation. The features of the patients are described in Table 1. Patients were selected to meet the following conditions. None of them had received radiation therapy before and none of them received chemotherapy during the irradiation period. Whole brain irradiation was performed using 4.0 MeV Liniac x-ray (LINAC ML20MDX, Mitsubishi electric Co. Ltd., Itami, Japan) with an opposed technique. A total of 40 (20 fractions) to 50 (25 fractions) Gy x-ray was used. In 1 case with malignant lymphoma (case 20 in Table 1), the total radiation dose using a whole brain opposed technique was 40 Gy, followed by 10 Gy local booster irradiation. MRI and MRS were performed with a 1.5 Tesla machine (Signa Horizon, General Electric Medical Systems Co. Ltd., Milwaukee, WI). T1- (TR/TE = 500/9 milliseconds) and T2- (TR/TE = 3600/84 milliseconds) weighted images were J Comput Assist Tomogr Volume 29, Number 3, May/June

2 Kaminaga and Shirai J Comput Assist Tomogr Volume 29, Number 3, May/June 2005 TABLE 1. Features of Patients Number Age/Gender Diagnosis Total Radiation Dose (Gy) Radiation Technique 1 70/Male Lung cancer 50 Whole brain opposed 2 69/Male Lung cancer 50 Whole brain opposed 3 59/Male Lung cancer 50 Whole brain opposed 4 57/Male Lung cancer 50 Whole brain opposed 5 60/Male Lung cancer 50 Whole brain opposed 6 59/Male Lung cancer 50 Whole brain opposed 7 70/Male Lung cancer 50 Whole brain opposed 8 66/Male Lung cancer 50 Whole brain opposed 9 68/Male Lung cancer 50 Whole brain opposed 10 59/Male Lung cancer 50 Whole brain opposed 11 62/Male Lung cancer 40 Whole brain opposed 12 42/Female Lung cancer 50 Whole brain opposed 13 75/Female Lung cancer 50 Whole brain opposed 14 68/Female Lung cancer 50 Whole brain opposed 15 63/Female Lung cancer 50 Whole brain opposed 16 45/Female Breast cancer 50 Whole brain opposed 17 58/Female Breast cancer 50 Whole brain opposed 18 66/Female Breast cancer 50 Whole brain opposed 19 68/Female Breast cancer 50 Whole brain opposed 20 71/Male Malignant lymphoma 50 Whole brain opposed + local booster acquired. MRS was performed using a point resolved spectroscopy sequence with a chemical shift selective water suppression pulse under the following conditions: TR/TE = 4000/30, 140, 270, 540 milliseconds, width of sampling frequency 2500 Hz, data point number The volume of regions of interest (ROI) ranged from 3.6 to 5.1 cm 3. Signals were repeated over 150 times and the results were averaged. The ROI was set in an occipital lobe cortex containing white matter (Fig. 1). No metastatic lesions or edema were observed on T1- and T2-weighted images inside these ROIs. Free induction decay was filtered by an exponential filter with line broadening of 3 Hz and a Gaussian function, and transformed by the Fourier transformation. Phase adjustment was performed according to the method of Klose. 15 Data processing was automatically performed with the SA/GE software package (General Electric Medical Systems, Milwaukee, WI). Curve fitting to the peak with a Lorentz function and measurement of each peak area were also performed automatically with the SA/GE software package. To estimate the volume ratio of cerebrospinal fluid space (CSF) inside a ROI, acquisition of a water proton signal was done under the following conditions; TR = 4000 msec, TE = 20, 30, 50, 80, 140, 270, 450, 700, 1200, 1500 milliseconds, averaging 2 times. The signal intensity curve versus echo-time of water protons were fitted to a double-exponential curve using the least square method as follows: S ¼ SðbÞ 3 Exp: fÿte=t2ðbþg þ SðcÞ 3 Exp: fÿte=t2ðcþg S: Total signal intensity S(b): Signal intensity from brain parenchyma (TE = 0) FIGURE 1. A T2-weighted axial image (TR/TE = 3600/84 milliseconds) of patient 5 (Table 1). Region of interest was placed in the occipital lobe cortex containing white matter. S(c): Signal intensity from CSF (TE = 0) T2 (b): T2 relaxation time of water protons in cerebral parenchyma T2 (c): T2 relaxation time of water protons in CSF Exp.: exponential function A cerebrospinal ratio of ROI was determined as follows: Ratio ¼ SðcÞ=SðcÞþSðbÞ: The concentration of NAA, Cho, and creatine/ phosphocreatine (Cr) were measured with 200 mmol/l of N-acetyl-L-alanine as an external standard. The B1 field was adjusted for the external standard, which was placed in front of the head. Four acquisitions (TR = 4000 milliseconds, TE = 30, 80, 140, 270 milliseconds, averaging = 20 times) were performed for the calculation of the T2 relaxation time of N- acetyl-l-alanine. The T2 relaxation times of NAA, Cho, and Cr were calculated from other acquisitions (TR = 4000 milliseconds, TE = 30, 80, 140, 270 milliseconds, averaging = 128 times). The transmitter and receiver gain and center frequency were fixed during each acquisition. The signal intensity curve versus the echo-time of these metabolites was fitted to a mono-exponential curve using the least-square method. The quantitative values of NAA, Cho, and Cr were calculated from the ratio of the peak area to that of the external standard, the estimated T2 relaxation time of each metabolite, and that of 200 mmol/l of N-acetyl-L-alanine solution as follows: 294 q 2005 Lippincott Williams & Wilkins

3 J Comput Assist Tomogr Volume 29, Number 3, May/June 2005 Radiation Changes in Brain Metabolism Shown by MRS Concentration of NAA, Cho or Cr = F 3 P 3 R 3 V 3 ½Exp: fÿte=t2ðext:þg=exp: fÿte=t2ðmetabolitesþgš F: correlation factor for the proton number. FðNAAÞ ¼2=3 FðCrÞ ¼2=3 FðChoÞ ¼2=9 P: concentration of L-alanine using the external standard (200 mmol/l) R: peak area ratios of each metabolite to external standard V: correlation factor for CSF = {S(c) + S(b)}/S(b) T2 (metabolites): T2 relaxation times of each metabolite T2 (ext.): T2 relaxation times of the external standard (L-alanine) The concentrations of each metabolite were calculated under the assumption that the tissue density of cerebral parenchyma is 1.6 g/ml. The effect of T1 relaxation was not considered, since protons within metabolites were almost fully relaxed under the condition of TR = 4000 milliseconds. Any significant difference in metabolite concentration between each phase was estimated by the paired t test. RESULTS Figure 2 represents a sequential MR spectrum change in the patients (number 5 of Table 1), and Figure 3 represents the total results. The T2 relaxation time was decided as follows: NAA milliseconds, Cho milliseconds, and Cr milliseconds. The NAA concentration decreased in the early delayed phase compared with that in the preradiation period and acute phase (P = 0.05 and 0.02). In the early delayed phase, Cho concentration increased compared with the pre-radiation period and acute phase (p = and 0.002). No obvious lactate or lipid peak ( ppm) was detected. There was no evidence of radiation-induced edema, necrosis, and demyelination spots on MRI in any patients. DISCUSSION Quantification of MRS has been achieved using various methods. Several papers proposed quantification methods using tissue water as an internal standard. We used N-acetyl- L-alanine as an external standard since the external standard material is not affected with brain radiation. Our results for the T2 relaxation times of NAA, Cho, and Cr were almost the same compared with those of the normal brain. 16 This indicated that there was no significant edema inside ROIs since the T2 relaxation time of these metabolites decreases in edematous cerebral parenchyma. 17 We use local ROIs in the occipital lobe with a PRESS technique to avoid tumorous and edematous regions and to increase the contribution of the cerebral cortex. Serious misregistration and contamination of the MR signal from skull lipids are notable problems. However, FIGURE 2. A C, Time-serial changes in the MR spectrum in patient 5 (Table 1). A, Pre-radiation period. B, Acute phase after radiation C, Early delayed phase after radiation. The Cho peak (3.2 ppm) seemed to increase in the early delayed phase. q 2005 Lippincott Williams & Wilkins 295

4 Kaminaga and Shirai J Comput Assist Tomogr Volume 29, Number 3, May/June 2005 FIGURE 3. Sequential changes in cerebral metabolites. A, N-acetyl-L-aspartate (NAA); B, choline-containing substance (Cho); C, creatine/phosphocreatine (Cr) in the pre-radiation and post-radiation periods. the signal contamination due to the skull lipid was not significant since no lipid peak was detected in our study. The possibility of misregistration is unavoidable with a localized MRS technique. The method for measurement of total brain metabolite concentration was proposed, 18 however, the MR signal from tumorous and edematous regions is not negligible in this method. Radiation therapy is frequently used for brain metastasis, and radiation-induced brain damage is a serious side effect. Radiation-induced brain damage is divided into 3 phases: an acute phase, early delayed phase, and late delayed phase. Of these, late delayed complications, which are caused by radiation necrosis several years after radiation, are the most serious, sometimes life-limiting, complication. The frequency of radiationinduced brain damage is determined by the total radiation dose, method of radiation, and duration of radiation. 8 A total dose of 40 to 50 Gy whole brain irradiation with 20 to 25 fractions has a little potential to cause radiation-induced brain damage. Whole brain irradiation with an opposed technique is not such a sophisticated method for avoiding radiationinduced brain damage. However, it is commonly used for multiple brain metastasis since the risk of metastasis elsewhere in the brain is high. Furthermore, the expected survival time is not so long in these patients. 19 The main symptoms in the acute phase reaction come from brain edema, which originates from vascular damage, such as increased permeability, 20,21 and destruction of the blood brain barrier. 22 No significant change in metabolite concentration was detected in our study, and no change in metabolite concentration has been reported in the acute phase. In the early delayed phase, the concentration of choline increased and no significant change was detected in creatine. The increase in Cho concentration was interpreted as breakdown and/or increased turnover of myelin and the cell membrane. Cho concentration varies in relation to the production 23 and degradation of Cho-containing phospholipids that are abundant in myelin and the cell membrane. 24 Myelin breakdown originating from oligodendrocyte damage was reported as radiation-induced brain damage. 25 Demyelination spot is one characteristic feature of irradiated brains in the early delayed phase. 10 The breakdown of the cell membrane and myelin due to irradiation may increase water-soluble phosphocholine. 11 This water-soluble phosphocholine contributes to the Cho signal 23 that is increased in the early delayed phase. In addition, Rubin et al examined this process directly using irradiated rabbits. 26 The NAA signal decreased in the early delayed phase after radiation. This result correlated with the literature. 13,14 Almost all NAA is thought to localize in the neurons and axons, 27,28 and the NAA signal is related to the number of axons, dendrites, and synapses, as well as that of neurons. 29 The decrease in the NAA signal is also detected at damaged neurons (such as ischemic neurons, cerebritis, the epileptic brain, and brain injury) They are interpreted as injury and/or dysfunction of neurons. Rango et al reported that the NAA signal was related to the trans-synaptic NAA concentration and NAA metabolism in neuron that is influenced by functional neuronal activity. 34 The NAA signal decrease in the early delayed phase can be transient. 13 Thus, the loss of neuron number is less responsible for this NAA signal decrease. However, quantified MRS measurement after the early delayed phase is needed to make a conclusion. The major responsible factor for radiation-induced NAA signal decrease in early delayed phase is difficult to conclude. However, both the neuron injury and dysfunction and functional neuronal inactivity were all induced by radiation in the patients after radiation therapy, and it is true that this NAA signal decrease represents neurotoxicity of radiation. The effectiveness for degradation of tumors and neurotoxicity are conflicting factors in brain irradiation. To establish suitable radiation doses and methods, both factors must be estimated correctly. MRS is a suitable method for in vivo estimation of neurotoxicity. The NAA signal indicates function of neuron and the Cho signal indicates the degree of demyelination. Demyelination is a direct expression of glial cell damage that leads to delayed cerebral necrosis q 2005 Lippincott Williams & Wilkins

5 J Comput Assist Tomogr Volume 29, Number 3, May/June 2005 Radiation Changes in Brain Metabolism Shown by MRS Thus, a quantified Cho value could predict the degree of delayed cerebral necrosis. The NAA, Cho, and Cr signals were quantified in acute and early delayed phase after whole brain irradiation in this study. Also, it is proved that MRS has a possibility to monitor radiation-induced neurotoxicity. However, some problems, such as correlating MRS findings to pathologic findings, investigating quantified MRS finding in late delayed phase, and estimating the possibility for the prediction of the radiation necrosis, are still outstanding. Further study is needed on these interesting problems. REFERENCES 1. Alger JR, Symko SC, Bizzi A, et al. Absolute quantitation of short TE brain 1H-MR spectra and spectroscopic imaging data. J Comput Assist Tomogr. 1993;17: Ernst T, Kreis R, Ross BD. Absolute quantification of water and metabolites in the human brain. Part I: Compartments and water. J Magn Reson. 1993;102: Christiansen P, Henriksen O, Stubgaard M, et al. in vivo quantification of brain metabolites by 1H-MRS using water as an internal standard. Magn Reson Imaging. 1993;11: Kreis R, Ernst T, Ross BD. Absolute quantification of water and metabolites in the human brain. Part II: Metabolites concentrations. J Magn Reson. 1993;102: Danielsen ER, Henriksen O. Absolute quantitative proton NMR spectroscopy based on the amplitude of the local water suppression pulse. Quantification of brain water and metabolites. NMR Biomed. 1994; 7: Provencher SW. Automatic quantitation of localized in vivo 1H spectra with LCModel. NMR Biomed. 2001;14: Barantin L, Le Pape A, Akoka S. A new method for absolute quantitation of MRS metabolites. Magn Reson Med. 1997;38: Sheline GE, Wara WM, Smith V. Therapeutic irradiation and brain injury. Int J Radiat Oncol Biol Phys. 1980;6: Kamarad V. Acute radiation sickness morphology of CNS syndrome. Acta Univ Palacki Olomuc Fac Med. 1989;121:7 144 (p.19). 10. Lampert PW, Davis RL. Delayed effects of radiation on the human central nervous system. Early and late delayed reaction. Neurology. 1964;14: Richards T, Budinger TF. NMR imaging and spectroscopy of the mammalian central nervous system after heavy ion radiation. Radiat Res. 1988;113: Szigety SK, Allen PS, Huyser-Wierenga D, et al. The effect of radiation on normal human CNS as detected by NMR spectroscopy. Int J Radiat Oncol Biol Phys. 1993;25: Esteve F, Rubin C, Grand S, et al. Transient metabolic changes observed with proton MR spectroscopy in normal human brain after radiation therapy. Int J Radiat Oncol Biol Phys. 1998;40: Movsas B, Li BS, Babb JS, et al. Quantifying radiation therapy-induced brain injury with whole-brain proton MR spectroscopy: initial observations. Radiology. 2001;221: Klose U. In vivo proton spectroscopy in presence of eddy currents. Magn Reson Med. 1990;14: Frahm J, Bruhn H, Gyngell ML, et al. Localized proton NMR spectroscopy in different regions of the human brain in vivo. Relaxation times and concentrations of cerebral metabolites. Magn Reson Med. 1989; 11: Kamada K, Houkin K, Hida K, et al. Localized proton spectroscopy of focal brain pathology in humans: significant effects of edema on spin-spin relaxation time. Magn Reson Med. 1994;31: Gonen O, Viswanathan AK, Catalaa I, et al. Total brain N-acetylaspartate concentration in normal, age-grouped females: quantitation with non-echo proton NMR spectroscopy. Magn Reson Med. 1998;40: Hall WA, Djalilian HR, Nussbaum ES, et al. Long-term survival with metastatic cancer to the brain. Med Oncol. 2000;17: Gobbel GT, Seilhan TM, Fike JR. Cerebrovascular response after interstitial irradiation. Radiat Res. 1992;130: Dereski MO, Chopp M, Chen Q, et al. Normal brain tissue response to photodynamic therapy: histology, vascular permeability and specific gravity. Photochem Photobiol. 1989;50: Nakata H, Yoshimine T, Murasawa A, et al. Early blood-brain barrier disruption after high-dose single-fraction irradiation in rats. Acta Neurochir (Wien). 1995;136: Miller BL, Chang L, Booth R, et al. In vivo 1H MRS choline: correlation with in vitro chemistry/histology. Life Sci. 1996;58: Hida K, Kwee IL, Nakada T. in vivo 1H and 31P NMR spectroscopy of the developing rat brain. Magn Reson Med. 1992;23: Belka C, Budach W, Kortmann RD, et al. Radiation induced CNS toxicity molecular and cellular mechanisms. Br J Cancer. 2001;85: Rubin P, Whitaker JN, Ceckler TL, et al. Myelin basic protein and magnetic resonance imaging for diagnosing radiation myelopathy. Int J Radiat Oncol Biol Phys. 1988;15: Simmons ML, Frondoza CG, Coyle JT. Immunohistological localization of N-acetyl-aspartate with monoclonal antibodies. Neuroscience. 1991; 45: Moffett JR, Namboodiri MA, Cangro CB, et al. Immunohistochemical localization of N-acetylaspartate in rat brain. Neuroreport. 1991;2: Van der Knaap MS, van der Grond J, van Rijen PC, et al. Age-dependent change in localized proton and phosphorus MR spectroscopy of the brain. Radiology. 1990;176: Uno M, Harada M, Nagahiro S. Quantitative evaluation of cerebral metabolites and cerebral blood flow in patients with carotid stenosis. Neurol Res. 2001;23: Guerrini L, Belli G, Cellerini M, et al. Proton MR spectroscopy of cerebellitis. Magn Reson Imaging. 2002;20: Cendes F, Andermann F, Preul M, et al. Lateralization of temporal lobe epilepsy based on regional metabolic abnormalities in proton magnetic resonance spectroscopic images. Ann Neurol. 1994;35: Demougeot C, Garnier P, Mossiat C, et al. N-Acetylaspartate, a marker of both cellular dysfunction and neuronal loss: its relevance to studies of acute brain injury. J Neurochem. 2001;77: Rango M, Spagnoli D, Tomei G, et al. Central nervous system transsynaptic effects of acute axonal injury: a 1H magnetic resonance spectroscopy study. Magn Reson Med. 1995;33: Van der Kogel AJ. Radiation-induced damage in the central nervous system: an interpretation of target cell responses. Br J Cancer Suppl. 1986;7: q 2005 Lippincott Williams & Wilkins 297