JOURNAL OF CHILD AND ADOLESCENT PSYCHOPHARMACOLOGY Volume 12, Number 4, 2002 © Mary Ann Liebert, Inc. Pp. 331–336
Case Report Glutamatergic Changes with Treatment in Attention Deficit Hyperactivity Disorder: A Preliminary Case Series Normand Carrey, M.D.,1 Frank P. MacMaster, M.Sc.,2 Sandra J. Sparkes, B.Sc.,3 Shakeela C. Khan, Ph.D.,1 and Vivek Kusumakar, M.D.1
ABSTRACT Magnetic resonance spectroscopy, a noninvasive neuroimaging method, is a technique with the potential to measure in vivo neurochemical changes to different medication treatments. Symptoms of attention deficit hyperactivity disorder (ADHD) improved in two children treated with methylphenidate and two children treated with atomoxetine, for whom pre- and posttreatment proton magnetic resonance spectroscopy examinations were obtained to assess the relation between the neurochemical profiles in the striatum and prefrontal cortex among symptom severity and response to treatment. In the striatum, a striking decrease in the glutamate/creatine ratio (mean change 56.1%) was observed between 14 and 18 weeks of therapy in all four children with ADHD. In the prefrontal cortex, however, changes in the glutamate/creatine ratio were noted only in subjects receiving atomoxetine, not in those receiving methylphenidate. These data suggest that in vivo magnetic resonance spectroscopy measurement has the potential to assess response to psychopharmacological treatment in children with ADHD.
INTRODUCTION
W
ELL-TITRATED TREATMENT
with psychostimulants is effective for control of core symptoms of attention deficit hyperactivity disorder (ADHD; Vitiello 2001). Psychostimulants act through dopamine transporter reuptake inhibition (methylphenidate) or converting the dopamine transporter into a channel, making more dopamine available in the synaptic cleft (d-amphetamine; Solanto 1998). Newer medications, such as atomoxetine, have also shown promise for the treatment of ADHD (Michel-
son et al. 2001). Atomoxetine, by inhibiting the noradrenaline transporter, may reduce ADHD symptoms by a different mechanism of action from stimulants, as other neuroamines have also been implicated in the pathophysiology of ADHD (Pliszka et al. 1996). Attempts to find neuroanatomical and neurochemical loci for ADHD have been advanced by neuroimaging techniques. Magnetic resonance spectroscopy (MRS) allows the direct and noninvasive measurement of neurochemical compounds such as N-acetyl-aspartate (NAA) and glutamate/glutamine, with a minor contribu-
Departments of 1Psychiatry, 2Institute for Biodiagnostics (Atlantic), National Research Council and Anatomy & Neurobiology, and 3Psychology, Dalhousie University, Halifax, Nova Scotia, Canada.
331
332
CARREY ET AL.
tion by gamma aminobutyric acid (GABA) (collectively referred to as the Glx resonance), choline, creatine, and myoinositol. Glutamate is an excitatory amino acid and the most abundant neurotransmitter in the brain (Shulman 2001). Because glutamate is also the major neurotransmitter in the corticostriatal afferents to the basal ganglia including subthalamic outputs back to the internal pallidum which in turn feed back to the thalamus and cortex, this neurotransmitter may be altered in ADHD. We report on four children with ADHD who had pre- and posttreatment proton magnetic resonance spectroscopic (1H-MRS) scans in the prefrontal cortex and striatum after effective treatment with methylphenidate (n = 2) or atomoxetine (n = 2).
METHODS Ethical requirements All subjects participating in the MRS protocol were approved by the local hospital Research Ethics Board of the IWK Health Centre, Halifax, Nova Scotia, and separate additional consent was obtained for the two atomoxetine participants from the corporate sponsor because these two patients were part of an ongoing open study on atomoxetine.
Case 2 The results of the K-SADS at baseline were compatible with DSM-IV diagnosis of ADHD combined subtype and oppositional defiant disorder with a baseline CGI rating of severely ill. Patient “CD” (8.2 years old now) agreed to have a baseline MRS scan prior to treatment with atomoxetine. She was started on 5 mg of atomoxetine twice a day, gradually increased to 20 mg twice a day corresponding to a dosage of 1.8 mg/kg. On atomoxetine, CD’s mood improved significantly. Furthermore, she was not as volatile and unpredictable as she was prior to treatment. After a trial of 16 weeks, both mother and teacher agreed that, although her behavior had improved significantly, her ability to focus had not improved since the beginning of the trial. Her trial was therefore discontinued. An MRS scan was repeated while on 20 mg of atomoxetine twice a day, with the patient taking 20 mg in the morning, 1 hour before the scan. At the time of her second scan, the results of CD’s repeat KSADS showed that she no longer met criteria for ADHD or oppositional defiant disorder. Despite her lack of academic progress, her CGI at end of trial was much improved (mildly ill) because of the improvement in her demeanor. Case 3
Case 1 Patient “AB’s” results on the Kiddie Schedule for Affective Disorders and Schizophrenia (KSADS; Orvaschel and Puig-Antich 1994), when she was not on methylphenidate, at the time of her initial MRS scan (9.8 years old) was compatible with a diagnosis of the combined subtype of ADHD by Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV; American Psychiatric Association 1994) criteria and a Clinical Global Index (CGI; Guy 1976) rating of severely ill. When the K-SADS was repeated at the time of her second MRS scan, 16 weeks later while on methylphenidate, AB did not meet criteria for ADHD. Her CGI at the time of her second scan was rated as minimally ill and much improved on the medication. AB took an oral dose of 10 mg of methylphenidate 1.5 hours before the MRS scan.
The result of the K-SADS completed at baseline was consistent with a diagnosis of the inattentive subtype of ADHD, with a CGI of moderately ill. Patient “EF” (11.8 years old) consented to a baseline MRS scan. He was started on 7.5 mg of atomoxetine twice a day and increased to 30 mg twice a day over 5 weeks. A repeat K-SADS administered 18 weeks later showed that he no longer met criteria for ADHD inattentive subtype and that his CGI was normal (not ill) and very much improved on medication. He consented to a repeat MRS scan while on his atomoxetine medication, receiving 30 mg 1 hour before the scan. EF’s father stated that his schoolwork had greatly improved, he was requiring less resource help for reading (from two sessions a week to one), and he was spontaneously reading books (something he had never done before the medication).
PEDIATRIC ADHD: GLUTAMATE CHANGES
333
Case 4 One of the authors (NC) reassessed patient “GH” when he was 7.4 years old for complaints of not listening at home or at school, hyperactivity, rudeness, and aggressive behavior. The results of the K-SADS at baseline revealed diagnoses of the combined subtype of ADHD, oppositional defiant disorder, and a CGI of severely ill. After he consented to a baseline MRS scan, he was retried on methylphenidate at a lower dosage of 5 mg every morning and 2.5 mg at lunch. When reassessed 14 weeks later, GH no longer met criteria for ADHD or oppositional defiant disorder, and he consented to a repeat MRS scan while on methylphenidate, 5 mg, 3.5 hours before his scan. CGI at that point in time was mildly ill and much improved on methylphenidate. Long echo MRS scans were performed using a 1.5-tesla MRI scanner (Siemens Magnetom Vision Scanner, Erlangen, Germany) using the PRESS technique (point resolved spectroscopy) on all patients at baseline (drug free) and repeated between 14 and 18 weeks of treatment. Two voxels, a 7-cc volume of interest localized to the left striatum (caudate and putamen) and a 4 cc in the right prefrontal cortex were acquired (see Fig. 1). Spectral acquisitions parameters were as follows: echo time = 135 msec, repetition time = 1.5 seconds, 256 averages for a total acquisition time of 10 minutes. The area under each of the resonances is proportional to the specific concentration of the specific compound. Peaks were fit using the analysis software integrated into the Siemens MRS package. Ratios of Glx/Cr (glutamate, glutamine, GABA/creatine) were used as the index measurement in the manner accepted in the literature (e.g., see Goff et al. 2002). 1H-MRS cerebral metabolite concentrations and ADHD symptoms before and after treatment were compared. All four subjects demonstrated reduction in the Glx/Cr ratio (mean change 56.1%) in the left striatum concurrent with a reduction in ADHD symptom severity (see Figs. 2 and 3). Other metabolite ratios, such as NAA/Cr and Choline/Cr changed little (increases of 8.8% and 15.3%, respectively). In the right prefrontal cortex, the Glx/Cr ratio remained stable in the subjects receiving
FIG. 1. Voxel placement in the right prefrontal cortex (A) and the left striatum (B).
methylphenidate (increase of 2.5%, n = 2) while dropping considerably in the atomoxetine group (47.5%, n = 2) (see Fig. 4). The remaining metabolic ratios in the prefrontal cortex did not differ greatly between treatment modalities. DISCUSSION The Glx/Cr ratio as a reflection of glutamate activity in the striatum underwent a dramatic decrease in relation to both drug treatments and correlated with improvement in CGI scores and K-SADS diagnoses. Glutamate has been implicated in a number of neurospychiatric conditions as well as in pediatric psychiatric illness (Rosenberg and Keshavan 1998). Glutamatergic pathways project from the prefrontal cortex to the basal ganglia (Taber and Fibiger 1993, 1995), areas that have been associated with ADHD. The decrease in Glx/Cr in the striatum of these subjects with ADHD after successful treatment demonstrated the utility of in vivo 1H-MRS for validating proposed mechanisms of actions for psychotropic med-
334
A
CARREY ET AL.
B
FIG. 2. Striatal spectra from a 9-year-old girl with attention deficit hyperactivity disorder (case 1 or AB) before (A) and after (B) methylphenidate treatment. NAA = N-acetyl-aspartate; ppm = part per million.
ications in children with ADHD. A decrease in this ratio was found in children with obsessive-compulsive disorder successfully treated with paroxetine (Rosenberg et al. 2000). At the core of the change in glutamate may be the interaction with striatal dopamine.
Verma and Moghaddam (1998) found that during basal conditions, metabotropic glutamate receptor activation facilitates striatal dopamine release, but during conditions of hyperstimulation, activation of metabotropic receptors reduces excess dopamine release.
FIG. 3. Striatal Glx/creatine ratios before and after treatment. ADHD = attention deficit hyperactivity disorder; Glx = glutamate/glutamine/gamma aminobutyric acid; MPH = methylphenidate. AB = case 1; CD = case 2; EF = case 3; GH = case 4.
PEDIATRIC ADHD: GLUTAMATE CHANGES
335
FIG. 4. Prefrontal Glx/creatine ratio before and after treatment. ADHD = attention deficit hyperactivity disorder; Glx = glutamate/glutamine/gamma aminobutyric acid; MPH = methylphenidate. AB = case 1; CD = case 2; EF = case 3; GH = case 4.
Dopaminergic and glutamatergic heteroreceptors have been found on glutamatergic and dopaminergic terminals, respectively, in the caudate-putamen (Groves et al. 1995). Dysregulation of reciprocal dopaminergic/glutamatergic modulation may be at the root of the deficits in ADHD and may also explain treatment response to methylphenidate. Atomoxetine may achieve a similar result in reducing ADHD symptoms but through a different pathway. In that noradrenergic fibers project extensively to the cortex, blockade of the noradrenaline transporter by atomoxetine may decrease cortical levels of glutamate, which in turn may lead to further secondary downstream decreases in striatal glutamate. Pliszka et al. (1996) proposed that ADHD may be associated with deficits in inhibitory frontostriatal connections predominantly driven by noradrenergic neurons’ effect on lower striatal structures, which in turn are driven by dopaminergic neurons. If this hypothesis is validated, then it could explain the therapeutic action of atomoxetine. Although the changes in the Glx/Cr ratio are consistent with a dopamine-glutamatergic pathway abnormality in ADHD, a potential confound of this study is the use of long echo times and the ability to resolve glutamate at 1.5 tesla. The long echo times were chosen for their robustness in this difficult-to-scan population. The longer scan times required for short
echo (quantitative acquisitions, >30 minutes) precluded their use in the study. Also, the glutamatergic resonance at 1.5 tesla consists of overlapping spectra from glutamate, glutamine, and GABA. Changes in these compounds cannot be distinguished from glutamate. This potential confound may be overcome with higher magnetic field strengths or advanced spectral editing techniques. Another potential confound was the relatively lower dosages of methylphenidate used in this study, which on the other hand could potentially increase the differences we already found. Metabolite differences in the subjects, not due to the drug effects, over the time span from first to second scan (14 to 18 weeks) could confound the findings of drug efficacy, but signal changes are usually of much smaller magnitude and take place over longer periods of time. This case series demonstrated how 1H-MRS may be utilized for the noninvasive evaluation of the therapeutic effect of a compound on neurochemistry. Controlled studies in patients with ADHD are needed to elucidate better the underlying mechanisms of this illness.
ACKNOWLEDGMENTS The Theodore and Vada Stanley Foundation and the Nova Scotia Health Research Founda-
336
CARREY ET AL.
tion supported this work. The two case studies on atomoxetine were part of an ongoing contract study funded by Eli Lilly Canada awarded to the first author.
REFERENCES American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 4th ed. (DSM-IV). Washington (DC), American Psychiatric Association, 1994. Goff DC, Hennen J, Lyoo IK, Tsai G, Wald LL, Evins AE, Yurgelun-Todd DA, Renshaw PF: Modulation of brain and serum glutamatergic concentrations following a switch from conventional neuroleptics to olanzapine. Biol Psychiatry 51:493–497, 2002. Groves P, Garcia Munoz M, Linder J, Manley M, Martone M, Young S: Elements of intrinsic organization and information processing in the neostriatum. In: Models of Information Processing in the Basal Ganglia. Edited by Houk J, Davis J, Beiser D. Cambridge (Massachusetts), MIT Press, 1995. Guy W: ECDEU Assessment Manual for Psychopharmacology, rev. Bethesda (Maryland), US Department of Health, Education, and Welfare, 1976, pp. 31–96. Michelson D, Faries D, Wernicke J, Kelsey D, Kendrick K, Sallee FR, Spencer T: Atomoxetine in the treatment of children and adolescents with attention-deficit/hyperactivity disorder: A randomized, placebo-controlled, dose-response study. Pediatrics 108:E83, 2001. Orvaschel H, Puig-Antich J: Schedule for Affective Disorders and Schizophrenia for School-Age Children: Epidemiological Version. Fort Lauderdale (Florida), Nova University, Center for Psychological Study, 1994. Pliszka SR, McCracken JT, Mass JW: Catecholamines in attention-deficit hyperactivity disorder: Current perspectives. J Am Acad Child Adolesc Psychiatry 35:264–272, 1996.
Rosenberg DR, Keshavan MS: A.E. Bennett Research Award. Toward a neurodevelopmental model of obsessive-compulsive disorder. Biol Psychiatry 43:623– 640, 1998. Rosenberg DR, MacMaster FP, Keshavan MS, Dimond K, Stewart C, Moore GJ: Striatal glutamate and treatment response in pediatric obsessive-compulsive disorder. J Am Acad Child and Adoles Psych 39:1096–1103, 2000. Shulman RG: Functional imaging studies: Linking mind and basic neuroscience. Am J Psychiatry 158:11–20, 2001. Solanto MV: Neuropsychopharmacological mechanisms of stimulant drug action in attention-deficit hyperactivity disorder: A review and integration. Behav Brain Res 94:127–152, 1998. Taber MT, Fibiger HC: Electrical stimulation of the medial prefrontal cortex increases dopamine release in the striatum. Neuropsychopharmacology 9:271–275, 1993. Taber MT, Fibiger HC: Electrical stimulation of the prefrontal cortex increases dopamine release in the nucleus accumbens of the rat: Modulation by metabotropic glutamate receptors. J Neurosci 15:3896– 3904, 1995. Verma A, Moghaddam B: Regulation of striatal dopamine release by metabotropic glutamate receptors. Synapse 28:220–226, 1998. Vitiello B: Methylphenidate in the treatment of children with attention-deficit hyperactivity disorder. CMAJ 165:1505–1506, 2001.
Address reprint requests to: Normand Carrey, M.D. Maritime Psychiatry IWK Health Centre 5850 University Avenue Halifax, Nova Scotia, Canada B3J 3G9 E-mail:
[email protected]
