- Email: [email protected]
Anti-kindling effect of slow repetitive transcranial magnetic stimulation in rats
Neuroscience Letters 351 (2003) 9–12 www.elsevier.com/locate/neulet
Anti-kindling effect of slow repetitive transcranial magnetic stimulation in rats David J. Anschela,b,*, Alvaro Pascual-Leonea, Gregory L. Holmesb a
Laboratory for Magnetic Brain Stimulation, Beth Israel Deaconess Medical Center, and Department of Neurology, Harvard Medical School, 330 Brookline Avenue, Boston, MA 02215, USA b Department of Neurology, Harvard Medical School, Children’s Hospital, Boston, MA, USA Received 20 January 2003; received in revised form 14 July 2003; accepted 15 July 2003
Abstract The cerebrospinal fluid (CSF) of animals exposed to electroconvulsive shock (ECS) has anticonvulsant properties when injected into naive animals. The present study investigated whether the CSF of humans exposed to 1 or 10 Hz repetitive transcranial magnetic stimulation (rTMS) has similar properties. Using a 4 day rat flurothyl kindling seizure model we found that the kindling rate was significantly decreased by intraventricular injection of CSF from depressed patients exposed to 1 Hz rTMS. The CSF from patients that underwent 10 Hz rTMS showed a trend toward an increased kindling rate. These results support the similarity of ECS and rTMS and suggest that 1 Hz and 10 Hz rTMS produce distinct physiologic changes. q 2003 Elsevier Ireland Ltd. All rights reserved. Keywords: Transcranial magnetic stimulation; Repetitive transcranial magnetic stimulation; Cerebrospinal fluid; Endogenous anticonvulsant; Kindling
Repetitive transcranial magnetic stimulation (rTMS) is being used in the evaluation of higher cognitive functions and modulation of neuroplasticity [10]. Additionally rTMS may have a therapeutic utility in neurobehavioral disorders such as major depressive disorder [4], schizophrenia [5], or drug-resistant epilepsy [16,17]. However, the physiology of these effects remains unclear. One possibility is that the behavioral effects of rTMS are due to the rTMS-induced modulation of cortical excitability (for review see Ref. [12]). These modulatory effects can last beyond the duration of the rTMS train and appear to partly depend upon the frequency of stimulation [11]. In most cases 1 Hz rTMS will lead to a lasting decrease in cortical excitability [12], while 10 Hz rTMS will increase cortical excitability [9]. Alternatively, though not mutually exclusive, the effects of rTMS might be related to the release of chemical substances with a more diffuse, rather than brain sitespecific effect on brain function. For example, animals exposed to electroconvulsive shock (ECS) have been shown * Corresponding author. Stanford Comprehensive Epilepsy Center, Department of Neurology – Room A343, Stanford, CA 94305-5235, USA. Tel.: þ 1-650-725-6648; fax: þ1-650-498-6326. E-mail address: [email protected] (D.J. Anschel).
to release into their cerebrospinal fluid (CSF) an opiate-like substance that has anticonvulsant properties [13,18]. It is conceivable that the suggested anti-epileptic properties of rTMS [16,17] might be related to a similar mechanism. In the present study we used an animal model of kindling to ascertain whether the CSF of patients exposed to rTMS at 1 or 10 Hz has any anti-kindling or anticonvulsant properties. Male Sprague –Dawley rats (270 –350 g; Charles River Laboratories) were used for all studies. They were housed individually in standard cages and had access to food and water ad libitum. Animals were treated in accordance with the guidelines set by the National Institutes of Health and Children’s Hospital Boston for the humane treatment of animals. Human CSF was obtained from patients with medication-resistant depression participating in a protocol studying the antidepressant effects of left or right prefrontal rTMS at a stimulation frequency of 1 or 10 Hz. Informed consent was obtained prior to obtaining human CSF. Participation in this study was not part of the treatment trial of depression with TMS. All depressed patients had an additional competency evaluation performed by an independent psychiatrist before being allowed to participate in
0304-3940/03/$ - see front matter q 2003 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/S0304-3940(03)00937-6
10
D.J. Anschel et al. / Neuroscience Letters 351 (2003) 9–12
the present study which was approved by the institutional review board at Beth Israel Deaconess Medical Center. A single rat was placed in a sealed 12.6 l plastic container. Flurothyl (Aldrich Chemical Co.: 2,2,2-Trifluoroethyl ether, 99%) was infused into the container on to a piece of filter paper at the rate of 0.1 ml/min using an infusion pump (Baxter AS40A). The typical seizure produced by this technique begins with myoclonic jerks, then there will be twisting of the neck and body, followed by what appears to be a recovery, then running, and finally clonic and ultimately tonic seizure activity. A timer was begun at the start of flurothyl infusion and the time was recorded at two points: (1) when the rat started to have myoclonic jerks; and (2) when there was a generalized tonic seizure. The infusion was then immediately stopped and the animal removed from the container. This process was repeated once per day for up to four consecutive days. There were four experimental groups of rats. The first group of rats was prepared using a sterile technique and pentobarbital anesthesia (50 mg/kg). The animals underwent stereotactic surgical implantation of a guide screw [8] directly over the left lateral ventricle (bregma 2 0.80 mm; lateral 1.5 mm). Following surgery, the animals were allowed to rest undisturbed for 4 –7 days. Each rat was anesthetized with isoflurane (Baxter) and injected with 13 ml of artificial CSF (aCSF) (Harvard Apparatus) through the screw guide [8] into the left lateral ventricle. The guide screw dummy cannula was then replaced and the rat was placed back in his cage. Ten minutes after recovery from the anesthesia (approximately 15 min post injection), the seizure threshold was measured as described above. This group had five rats. The second group of animals had the same surgical implant procedure as those in group 1; these rats were then injected with normal human CSF instead of aCSF. Seizure threshold was then measured as described above. The normal human CSF was obtained after lumbar punctures done on neurological patients for medically indicated reasons. In all cases the protein and glucose levels as well as cell counts were within normal limits. None of these patients were taking psychotropic or anticonvulsant medications. Informed consent was obtained from these patients prior to utilizing their CSF in the present study. Three human subjects were CSF donors in this group. The first was a 25-year-old woman with suspected multiple sclerosis. The second was a 49-year-old man with a severe headache eventually diagnosed as migraine. The third was a 73-yearold woman with a new headache and memory problems in whom meningoencephalitis or subarachnoid hemorrhage were ruled out. This group of animals initially had nine rats. Three of these rats could not be tested beyond day 2 because of problems with the experimental apparatus so that on days 3 and 4 post injection this group had six rats. The third group of animals had the same surgical implant procedure as those in group 1; these three rats were then injected with human CSF taken from a patient who had
undergone eight daily sessions of 1 Hz rTMS for depression. This was a 41-year-old male with a 7 year-long history of medication-resistant depression. The patient had undergone withdrawal of all medications and a 2 week wash-out period prior to the rTMS. Each daily rTMS session consisted of stimulation to the left dorsolateral prefrontal cortex with a continuous train of 1 Hz rTMS of 26.6 min duration (1600 stimuli) at an intensity of 90% of the patient’s motor threshold. The CSF was removed just prior to the 9th rTMS session, which was during the second week of the treatment course and 24 h after the 8th rTMS session. Seizure threshold in the rats was then measured as described above. The fourth group of animals had the same surgical implant procedure as those in group 1; these two rats were then injected with human CSF taken from a patient who had been treated with eight sessions of 10 Hz rTMS for depression. This was a 21-year-old male with a 7 year – long history of medication-resistant depression. The patient had undergone withdrawal of all medication and a 2 week wash-out period prior to the rTMS. Each daily rTMS session consisted of stimulation to the right dorsolateral prefrontal cortex with 20 trains of 10 Hz rTMS. Each train lasted 8 s (80 stimuli) and the inter-train interval was 22 s. A total of 1600 stimuli were delivered in 20 min. The stimulation intensity was 90% of the patient’s resting motor threshold. The CSF was removed just prior to the 9th rTMS session, which was during the second week of the treatment course and 24 h after the 8th rTMS session. Seizure threshold in the rats was then measured as described above. Following the experiment, the injection site was confirmed by histological analysis in randomly sampled rats. Analysis of variance (ANOVA) with repeated measures was used to calculate the relationship of seizure threshold between animals injected with control CSF, and those injected with aCSF, post 1 Hz TMS CSF, and post 10 Hz TMS CSF. A series of planned two-tailed t-test were then used post-hoc. Results were considered statistically significant when P # 0:05. The latencies for the myoclonic and tonic seizures are shown in Figs. 1 and 2, respectively. Significant ‘group £ time’ interactions for myoclonic seizure onset are shown in Fig. 3. The decrease in seizure threshold over time (4 days) was highly significant (P , 0:001) for all groups, for both myoclonic and tonic seizures, demonstrating a kindling effect. The aCSF group was not significantly different from the controls for either myoclonic or tonic seizures. Rats injected with CSF taken from the patient who had undergone 10 Hz rTMS showed a statistically nonsignificant decrease in myoclonic seizure threshold on day 1 compared to control (Fig. 1). There was a significant ‘group £ time’ interaction for the 10 Hz TMS group versus the control when measuring onset of myoclonus (F ¼ 11:15, P , 0:001) (Fig. 3). This may be explained by the differences in myoclonic seizure threshold, particu-
D.J. Anschel et al. / Neuroscience Letters 351 (2003) 9–12
Fig. 1. Daily myoclonic seizure onset in response to flurothyl. All groups were compared to the control CSF over all 4 days by ANOVA with repeated measures. The 1 Hz TMS group was significantly different (F ¼ 10:15, P ¼ 0:006). A day-by-day analysis using an unpaired, two-tailed t-test showed a statistically significant difference between control and 1 Hz TMS groups on day 1 (P ¼ 0:016).
larly on days 1 and 4. On day 1 the myoclonic seizure threshold was decreased as discussed above. On day 4 the myoclonic seizure threshold in the 10 Hz TMS group was actually increased compared with control CSF. This may be due to a rebound effect. In future studies with a larger sample size this should become clearer. Our main finding is that seizure onset (both myoclonic and tonic) over the 4 day experiment was significantly delayed in the 1 Hz TMS group. In the 1 Hz TMS group myoclonic seizure latencies were longer than in the control CSF group over the 4 days of testing (F ¼ 10:15, P ¼ 0:006) (Fig. 1) and the ‘group £ time’ interaction was significant (F ¼ 8:52, P , 0:001) (Fig. 3). This later effect is most likely due entirely to the drop in myoclonic seizure threshold from day 1 to day 2 versus control. Additionally, on day 1 the 1 Hz TMS group differed from the control CSF group (P ¼ 0:016) in onset of myoclonus (Fig. 1). For the
Fig. 2. Daily tonic seizure onset in response to flurothyl. All groups were compared to the control CSF over all 4 days by ANOVA with repeated measures. The 1 Hz TMS group was significantly different from control (F ¼ 4:75, P ¼ 0:045).
11
Fig. 3. Groups with significant ‘group £ time’ interactions during daily myoclonic seizure in response to flurothyl. All groups were compared to the control CSF over all 4 days by ANOVA with repeated measures. The 10 Hz TMS and 1 Hz TMS groups were significantly different from control (F ¼ 11:15, P , 0:001 and F ¼ 8:52, P , 0:001, respectively). The effect in the 10 Hz group may be explained by the differences in myoclonic seizure threshold, particularly on days 1 and 4. On day 1 the myoclonic seizure threshold was decreased. On day 4 the myoclonic seizure threshold in the 10 Hz TMS group was actually increased compared with control CSF. This may be due to a rebound effect. For the 1 Hz group this effect is most likely due entirely to the drop in myoclonic seizure threshold from day 1 to day 2 versus control.
tonic seizures, the 1 Hz TMS group also showed longer latencies than controls (F ¼ 4:75, P ¼ 0:045) over the 4 days of testing (Fig. 2). These findings are consistent with the phenomenon of ‘quenching’, where 1 Hz electrical stimulation has been shown to inhibit the development and expression of amygdala kindled seizures in the rat [19]. RTMS has been shown to decrease susceptibility to pentylenetetrazol-induced seizures in rats [1] and to improve intractable epilepsy in humans [16]. Low frequency rTMS has been shown to lead to a suppression of cortical excitability in the majority of subjects [12]. The fact that the effect was relatively short-lived in the present study is consistent with a recent study showing that the antiseizure effect of rTMS is transient [17]. There have been multiple theories to explain the effects of slow rTMS, but the present study is the first to suggest that there might be an anticonvulsant substance in the CSF following rTMS. This may be an opioid as has been suggested by animal ECS experiments [6,18], an endogenous benzodiazepine [2], or a thyrotropin-releasing hormone (TRH) related peptide [14]. It has been demonstrated that ECS in rats induces the synthesis of TRH [7,14]. Two previous studies (which did not measure TRH) have suggested that 5 Hz [3] and 10 Hz [15] rTMS of the dorsolateral prefrontal cortex can cause an increase in thyroid stimulating hormone (TSH). Additionally, intrathecal administration of TRH can induce remissions of major depression [14], perhaps providing a link to the antidepressant effects of rTMS. TRH has also been shown to have anticonvulsant properties [7]. Other potential
12
D.J. Anschel et al. / Neuroscience Letters 351 (2003) 9–12
seizure threshold modifying changes, which might occur in the CSF, would be changes in monoamines or electrolytes. The day 1 anti-seizure and overall anti-kindling effects seen in the 1 Hz TMS group more robustly affect the myoclonic than the tonic seizures in this experimental model. This suggests that the underlying physiological basis of these two seizure types might be different. Our findings must be interpreted with caution. Since this was a pilot study involving removal of CSF from humans the number of human subjects was limited. It will be important for these results to be replicated. In addition, adding a control group consisting of depressed patients with TMS would be useful for future studies. The results of this study suggest that human CSF taken from subjects treated with 1 Hz rTMS may have an anticonvulsive property and CSF taken from subjects treated with 10 Hz rTMS may lower seizure threshold. Although preliminary, this study is an important step toward a better understanding of the mechanisms of action of rTMS.
Acknowledgements We would like to thank Dr Mark Thall for his assistance with patient evaluations. This study was supported in part by the National Institutes of Mental Health (A.P.L.) (MH57980) and the National Institute of Neurological Disorders and Stroke (G.L.H.) (NS27084).
References [1] N. Akamatsu, Y. Fueta, Y. Endo, K. Matsunaga, T. Uozumi, S. Tsuji, Decreased susceptibility to pentylenetetrazol-induced seizures after low-frequency transcranial magnetic stimulation, Neurosci. Lett. 310 (2001) 153–156. [2] R. Avallone, M.L. Zeneroli, I. Venturini, L. Corsi, P. Schrier, M. Kleinschnitz, C. Ferrarese, F. Farina, C. Baraldi, N. Pecora, M. Frigo, M. Baraldi, Endogenous benzodiazepine-like compounds and diazepam binding inhibitor in serum of patients with liver cirrhosis with and without overt encephalopathy, Gut 42 (1998) 861 –867. [3] M.S. George, E.M. Wassermann, W.A. Williams, J. Steppel, A. Pascual-Leone, P. Basser, M. Hallett, R.M. Post, Changes in mood and hormone levels after rapid-rate transcranial magnetic stimulation of the prefrontal cortex, J. Neuropsychiatry Clin. Neurosci. 8 (1996) 172–180.
[4] G. Hasey, Transcranial magnetic stimulation in the treatment of mood disorder: a review and comparison with electroconvulsive therapy, Can. J. Psychiatry 46 (2001) 720 –727. [5] R.E. Hoffman, N.N. Boutros, R.M. Berman, E. Roessler, A. Belger, J.H. Krystal, D.S. Charney, Transcranial magnetic stimulation of left temporoparietal cortex in three patients reporting hallucinated ‘voices’, Biol. Psychiatry 46 (1999) 130– 132. [6] L. Isaac, J. Swanger, Alteration of electroconvulsive threshold by cerebrospinal fluid from cats tolerant to electroconvulsive shock, Life Sci. 33 (1983) 2301–2304. [7] M.J. Kubek, D. Liang, K.E. Byrd, A.J. Domb, Prolonged seizure suppression by a single implantable polymeric-TRH microdisk preparation, Brain Res. 809 (1998) 189– 197. [8] S. Lal, M. Lacoix, P. Tofilon, G.N. Fuller, R. Sawaya, F.F. Lang, An implantable guide-screw system for brain tumor studies in small animals, J. Neurosurg. 92 (2000) 326–333. [9] F. Maeda, J.P. Keenan, J.M. Tormos, H. Topka, A. Pascual-Leone, Interindividual variability of the modulatory effects of repetitive transcranial magnetic stimulation, Exp. Brain Res. 133 (2000) 425 –430. [10] A. Pascual-Leone, F. Tarazona, J. Keenan, J.M. Tormos, R. Hamilton, M.D. Catala, Transcranial magnetic stimulation and neuroplasticity, Neuropsychologia 37 (1999) 207– 217. [11] A. Pascual-Leone, J.M. Tormos, J.P. Keenan, F. Tarazona, C. Canete, M.D. Catala, Study and modulation of human cortical excitability with transcranial magnetic stimulation, J. Clin. Neurophysiol. 15 (1998) 333 –343. [12] J.R. Romero, D.J. Anschel, R. Sparing, M. Gangitano, A. PascualLeone, Subthreshold low frequency repetitive transcranial magnetic stimulation selectively decreases facilitation in the motor cortex, Clin. Neurophysiol. 113 (2002) 101–107. [13] H.A. Sackeim, The anticonvulsant hypothesis of the mechanism of action of ECT: current status, J. ECT 15 (1999) 5–26. [14] A. Sattin, The role of TRH and related peptides in the mechanism of action of ECT, J. ECT 76 (1999) 76 –92. [15] M.P. Szuba, J.P. O’Reardon, A.S. Rai, J. Snyder-Kastenberg, J.D. Amsterdam, D.R. Gettes, E. Wassermann, D.L. Evans, Acute mood and thyroid stimulating hormone effects of transcranial magnetic stimulation in major depression, Biol. Psychiatry 50 (2001) 22–27. [16] F. Tergau, U. Naumann, W. Paulus, B.J. Steinhoff, Low-frequency repetitive transcranial magnetic stimulation improves intractable epilepsy, Lancet 353 (1999) 2209. [17] W.H. Theodore, M.D. Hunter, R. Chen, F. Vega-Bermudez, B. Boroojerdi, P. Reeves-Tyer, K. Werhahn, K.R. Kelley, L. Cohen, Transcranial magnetic stimulation for the treatment of seizures, Neurology 59 (2002) 560–562. [18] F.C. Tortella, J.B. Long, Endogenous anticonvulsant substance in rat cerebrospinal fluid after a generalized seizure, Science 228 (1985) 1106–1108. [19] S.R. Weiss, X.L. Li, J.B. Rosen, H. Li, T. Heynen, R.M. Post, Quenching: inhibition of development and expression of amygdala kindled seizures with low frequency stimulation, NeuroReport 6 (1995) 2171–2176.
Anti-kindling effect of slow repetitive transcranial magnetic stimulation in rats David J. Anschela,b,*, Alvaro Pascual-Leonea, Gregory L. Holmesb a
Laboratory for Magnetic Brain Stimulation, Beth Israel Deaconess Medical Center, and Department of Neurology, Harvard Medical School, 330 Brookline Avenue, Boston, MA 02215, USA b Department of Neurology, Harvard Medical School, Children’s Hospital, Boston, MA, USA Received 20 January 2003; received in revised form 14 July 2003; accepted 15 July 2003
Abstract The cerebrospinal fluid (CSF) of animals exposed to electroconvulsive shock (ECS) has anticonvulsant properties when injected into naive animals. The present study investigated whether the CSF of humans exposed to 1 or 10 Hz repetitive transcranial magnetic stimulation (rTMS) has similar properties. Using a 4 day rat flurothyl kindling seizure model we found that the kindling rate was significantly decreased by intraventricular injection of CSF from depressed patients exposed to 1 Hz rTMS. The CSF from patients that underwent 10 Hz rTMS showed a trend toward an increased kindling rate. These results support the similarity of ECS and rTMS and suggest that 1 Hz and 10 Hz rTMS produce distinct physiologic changes. q 2003 Elsevier Ireland Ltd. All rights reserved. Keywords: Transcranial magnetic stimulation; Repetitive transcranial magnetic stimulation; Cerebrospinal fluid; Endogenous anticonvulsant; Kindling
Repetitive transcranial magnetic stimulation (rTMS) is being used in the evaluation of higher cognitive functions and modulation of neuroplasticity [10]. Additionally rTMS may have a therapeutic utility in neurobehavioral disorders such as major depressive disorder [4], schizophrenia [5], or drug-resistant epilepsy [16,17]. However, the physiology of these effects remains unclear. One possibility is that the behavioral effects of rTMS are due to the rTMS-induced modulation of cortical excitability (for review see Ref. [12]). These modulatory effects can last beyond the duration of the rTMS train and appear to partly depend upon the frequency of stimulation [11]. In most cases 1 Hz rTMS will lead to a lasting decrease in cortical excitability [12], while 10 Hz rTMS will increase cortical excitability [9]. Alternatively, though not mutually exclusive, the effects of rTMS might be related to the release of chemical substances with a more diffuse, rather than brain sitespecific effect on brain function. For example, animals exposed to electroconvulsive shock (ECS) have been shown * Corresponding author. Stanford Comprehensive Epilepsy Center, Department of Neurology – Room A343, Stanford, CA 94305-5235, USA. Tel.: þ 1-650-725-6648; fax: þ1-650-498-6326. E-mail address: [email protected] (D.J. Anschel).
to release into their cerebrospinal fluid (CSF) an opiate-like substance that has anticonvulsant properties [13,18]. It is conceivable that the suggested anti-epileptic properties of rTMS [16,17] might be related to a similar mechanism. In the present study we used an animal model of kindling to ascertain whether the CSF of patients exposed to rTMS at 1 or 10 Hz has any anti-kindling or anticonvulsant properties. Male Sprague –Dawley rats (270 –350 g; Charles River Laboratories) were used for all studies. They were housed individually in standard cages and had access to food and water ad libitum. Animals were treated in accordance with the guidelines set by the National Institutes of Health and Children’s Hospital Boston for the humane treatment of animals. Human CSF was obtained from patients with medication-resistant depression participating in a protocol studying the antidepressant effects of left or right prefrontal rTMS at a stimulation frequency of 1 or 10 Hz. Informed consent was obtained prior to obtaining human CSF. Participation in this study was not part of the treatment trial of depression with TMS. All depressed patients had an additional competency evaluation performed by an independent psychiatrist before being allowed to participate in
0304-3940/03/$ - see front matter q 2003 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/S0304-3940(03)00937-6
10
D.J. Anschel et al. / Neuroscience Letters 351 (2003) 9–12
the present study which was approved by the institutional review board at Beth Israel Deaconess Medical Center. A single rat was placed in a sealed 12.6 l plastic container. Flurothyl (Aldrich Chemical Co.: 2,2,2-Trifluoroethyl ether, 99%) was infused into the container on to a piece of filter paper at the rate of 0.1 ml/min using an infusion pump (Baxter AS40A). The typical seizure produced by this technique begins with myoclonic jerks, then there will be twisting of the neck and body, followed by what appears to be a recovery, then running, and finally clonic and ultimately tonic seizure activity. A timer was begun at the start of flurothyl infusion and the time was recorded at two points: (1) when the rat started to have myoclonic jerks; and (2) when there was a generalized tonic seizure. The infusion was then immediately stopped and the animal removed from the container. This process was repeated once per day for up to four consecutive days. There were four experimental groups of rats. The first group of rats was prepared using a sterile technique and pentobarbital anesthesia (50 mg/kg). The animals underwent stereotactic surgical implantation of a guide screw [8] directly over the left lateral ventricle (bregma 2 0.80 mm; lateral 1.5 mm). Following surgery, the animals were allowed to rest undisturbed for 4 –7 days. Each rat was anesthetized with isoflurane (Baxter) and injected with 13 ml of artificial CSF (aCSF) (Harvard Apparatus) through the screw guide [8] into the left lateral ventricle. The guide screw dummy cannula was then replaced and the rat was placed back in his cage. Ten minutes after recovery from the anesthesia (approximately 15 min post injection), the seizure threshold was measured as described above. This group had five rats. The second group of animals had the same surgical implant procedure as those in group 1; these rats were then injected with normal human CSF instead of aCSF. Seizure threshold was then measured as described above. The normal human CSF was obtained after lumbar punctures done on neurological patients for medically indicated reasons. In all cases the protein and glucose levels as well as cell counts were within normal limits. None of these patients were taking psychotropic or anticonvulsant medications. Informed consent was obtained from these patients prior to utilizing their CSF in the present study. Three human subjects were CSF donors in this group. The first was a 25-year-old woman with suspected multiple sclerosis. The second was a 49-year-old man with a severe headache eventually diagnosed as migraine. The third was a 73-yearold woman with a new headache and memory problems in whom meningoencephalitis or subarachnoid hemorrhage were ruled out. This group of animals initially had nine rats. Three of these rats could not be tested beyond day 2 because of problems with the experimental apparatus so that on days 3 and 4 post injection this group had six rats. The third group of animals had the same surgical implant procedure as those in group 1; these three rats were then injected with human CSF taken from a patient who had
undergone eight daily sessions of 1 Hz rTMS for depression. This was a 41-year-old male with a 7 year-long history of medication-resistant depression. The patient had undergone withdrawal of all medications and a 2 week wash-out period prior to the rTMS. Each daily rTMS session consisted of stimulation to the left dorsolateral prefrontal cortex with a continuous train of 1 Hz rTMS of 26.6 min duration (1600 stimuli) at an intensity of 90% of the patient’s motor threshold. The CSF was removed just prior to the 9th rTMS session, which was during the second week of the treatment course and 24 h after the 8th rTMS session. Seizure threshold in the rats was then measured as described above. The fourth group of animals had the same surgical implant procedure as those in group 1; these two rats were then injected with human CSF taken from a patient who had been treated with eight sessions of 10 Hz rTMS for depression. This was a 21-year-old male with a 7 year – long history of medication-resistant depression. The patient had undergone withdrawal of all medication and a 2 week wash-out period prior to the rTMS. Each daily rTMS session consisted of stimulation to the right dorsolateral prefrontal cortex with 20 trains of 10 Hz rTMS. Each train lasted 8 s (80 stimuli) and the inter-train interval was 22 s. A total of 1600 stimuli were delivered in 20 min. The stimulation intensity was 90% of the patient’s resting motor threshold. The CSF was removed just prior to the 9th rTMS session, which was during the second week of the treatment course and 24 h after the 8th rTMS session. Seizure threshold in the rats was then measured as described above. Following the experiment, the injection site was confirmed by histological analysis in randomly sampled rats. Analysis of variance (ANOVA) with repeated measures was used to calculate the relationship of seizure threshold between animals injected with control CSF, and those injected with aCSF, post 1 Hz TMS CSF, and post 10 Hz TMS CSF. A series of planned two-tailed t-test were then used post-hoc. Results were considered statistically significant when P # 0:05. The latencies for the myoclonic and tonic seizures are shown in Figs. 1 and 2, respectively. Significant ‘group £ time’ interactions for myoclonic seizure onset are shown in Fig. 3. The decrease in seizure threshold over time (4 days) was highly significant (P , 0:001) for all groups, for both myoclonic and tonic seizures, demonstrating a kindling effect. The aCSF group was not significantly different from the controls for either myoclonic or tonic seizures. Rats injected with CSF taken from the patient who had undergone 10 Hz rTMS showed a statistically nonsignificant decrease in myoclonic seizure threshold on day 1 compared to control (Fig. 1). There was a significant ‘group £ time’ interaction for the 10 Hz TMS group versus the control when measuring onset of myoclonus (F ¼ 11:15, P , 0:001) (Fig. 3). This may be explained by the differences in myoclonic seizure threshold, particu-
D.J. Anschel et al. / Neuroscience Letters 351 (2003) 9–12
Fig. 1. Daily myoclonic seizure onset in response to flurothyl. All groups were compared to the control CSF over all 4 days by ANOVA with repeated measures. The 1 Hz TMS group was significantly different (F ¼ 10:15, P ¼ 0:006). A day-by-day analysis using an unpaired, two-tailed t-test showed a statistically significant difference between control and 1 Hz TMS groups on day 1 (P ¼ 0:016).
larly on days 1 and 4. On day 1 the myoclonic seizure threshold was decreased as discussed above. On day 4 the myoclonic seizure threshold in the 10 Hz TMS group was actually increased compared with control CSF. This may be due to a rebound effect. In future studies with a larger sample size this should become clearer. Our main finding is that seizure onset (both myoclonic and tonic) over the 4 day experiment was significantly delayed in the 1 Hz TMS group. In the 1 Hz TMS group myoclonic seizure latencies were longer than in the control CSF group over the 4 days of testing (F ¼ 10:15, P ¼ 0:006) (Fig. 1) and the ‘group £ time’ interaction was significant (F ¼ 8:52, P , 0:001) (Fig. 3). This later effect is most likely due entirely to the drop in myoclonic seizure threshold from day 1 to day 2 versus control. Additionally, on day 1 the 1 Hz TMS group differed from the control CSF group (P ¼ 0:016) in onset of myoclonus (Fig. 1). For the
Fig. 2. Daily tonic seizure onset in response to flurothyl. All groups were compared to the control CSF over all 4 days by ANOVA with repeated measures. The 1 Hz TMS group was significantly different from control (F ¼ 4:75, P ¼ 0:045).
11
Fig. 3. Groups with significant ‘group £ time’ interactions during daily myoclonic seizure in response to flurothyl. All groups were compared to the control CSF over all 4 days by ANOVA with repeated measures. The 10 Hz TMS and 1 Hz TMS groups were significantly different from control (F ¼ 11:15, P , 0:001 and F ¼ 8:52, P , 0:001, respectively). The effect in the 10 Hz group may be explained by the differences in myoclonic seizure threshold, particularly on days 1 and 4. On day 1 the myoclonic seizure threshold was decreased. On day 4 the myoclonic seizure threshold in the 10 Hz TMS group was actually increased compared with control CSF. This may be due to a rebound effect. For the 1 Hz group this effect is most likely due entirely to the drop in myoclonic seizure threshold from day 1 to day 2 versus control.
tonic seizures, the 1 Hz TMS group also showed longer latencies than controls (F ¼ 4:75, P ¼ 0:045) over the 4 days of testing (Fig. 2). These findings are consistent with the phenomenon of ‘quenching’, where 1 Hz electrical stimulation has been shown to inhibit the development and expression of amygdala kindled seizures in the rat [19]. RTMS has been shown to decrease susceptibility to pentylenetetrazol-induced seizures in rats [1] and to improve intractable epilepsy in humans [16]. Low frequency rTMS has been shown to lead to a suppression of cortical excitability in the majority of subjects [12]. The fact that the effect was relatively short-lived in the present study is consistent with a recent study showing that the antiseizure effect of rTMS is transient [17]. There have been multiple theories to explain the effects of slow rTMS, but the present study is the first to suggest that there might be an anticonvulsant substance in the CSF following rTMS. This may be an opioid as has been suggested by animal ECS experiments [6,18], an endogenous benzodiazepine [2], or a thyrotropin-releasing hormone (TRH) related peptide [14]. It has been demonstrated that ECS in rats induces the synthesis of TRH [7,14]. Two previous studies (which did not measure TRH) have suggested that 5 Hz [3] and 10 Hz [15] rTMS of the dorsolateral prefrontal cortex can cause an increase in thyroid stimulating hormone (TSH). Additionally, intrathecal administration of TRH can induce remissions of major depression [14], perhaps providing a link to the antidepressant effects of rTMS. TRH has also been shown to have anticonvulsant properties [7]. Other potential
12
D.J. Anschel et al. / Neuroscience Letters 351 (2003) 9–12
seizure threshold modifying changes, which might occur in the CSF, would be changes in monoamines or electrolytes. The day 1 anti-seizure and overall anti-kindling effects seen in the 1 Hz TMS group more robustly affect the myoclonic than the tonic seizures in this experimental model. This suggests that the underlying physiological basis of these two seizure types might be different. Our findings must be interpreted with caution. Since this was a pilot study involving removal of CSF from humans the number of human subjects was limited. It will be important for these results to be replicated. In addition, adding a control group consisting of depressed patients with TMS would be useful for future studies. The results of this study suggest that human CSF taken from subjects treated with 1 Hz rTMS may have an anticonvulsive property and CSF taken from subjects treated with 10 Hz rTMS may lower seizure threshold. Although preliminary, this study is an important step toward a better understanding of the mechanisms of action of rTMS.
Acknowledgements We would like to thank Dr Mark Thall for his assistance with patient evaluations. This study was supported in part by the National Institutes of Mental Health (A.P.L.) (MH57980) and the National Institute of Neurological Disorders and Stroke (G.L.H.) (NS27084).
References [1] N. Akamatsu, Y. Fueta, Y. Endo, K. Matsunaga, T. Uozumi, S. Tsuji, Decreased susceptibility to pentylenetetrazol-induced seizures after low-frequency transcranial magnetic stimulation, Neurosci. Lett. 310 (2001) 153–156. [2] R. Avallone, M.L. Zeneroli, I. Venturini, L. Corsi, P. Schrier, M. Kleinschnitz, C. Ferrarese, F. Farina, C. Baraldi, N. Pecora, M. Frigo, M. Baraldi, Endogenous benzodiazepine-like compounds and diazepam binding inhibitor in serum of patients with liver cirrhosis with and without overt encephalopathy, Gut 42 (1998) 861 –867. [3] M.S. George, E.M. Wassermann, W.A. Williams, J. Steppel, A. Pascual-Leone, P. Basser, M. Hallett, R.M. Post, Changes in mood and hormone levels after rapid-rate transcranial magnetic stimulation of the prefrontal cortex, J. Neuropsychiatry Clin. Neurosci. 8 (1996) 172–180.
[4] G. Hasey, Transcranial magnetic stimulation in the treatment of mood disorder: a review and comparison with electroconvulsive therapy, Can. J. Psychiatry 46 (2001) 720 –727. [5] R.E. Hoffman, N.N. Boutros, R.M. Berman, E. Roessler, A. Belger, J.H. Krystal, D.S. Charney, Transcranial magnetic stimulation of left temporoparietal cortex in three patients reporting hallucinated ‘voices’, Biol. Psychiatry 46 (1999) 130– 132. [6] L. Isaac, J. Swanger, Alteration of electroconvulsive threshold by cerebrospinal fluid from cats tolerant to electroconvulsive shock, Life Sci. 33 (1983) 2301–2304. [7] M.J. Kubek, D. Liang, K.E. Byrd, A.J. Domb, Prolonged seizure suppression by a single implantable polymeric-TRH microdisk preparation, Brain Res. 809 (1998) 189– 197. [8] S. Lal, M. Lacoix, P. Tofilon, G.N. Fuller, R. Sawaya, F.F. Lang, An implantable guide-screw system for brain tumor studies in small animals, J. Neurosurg. 92 (2000) 326–333. [9] F. Maeda, J.P. Keenan, J.M. Tormos, H. Topka, A. Pascual-Leone, Interindividual variability of the modulatory effects of repetitive transcranial magnetic stimulation, Exp. Brain Res. 133 (2000) 425 –430. [10] A. Pascual-Leone, F. Tarazona, J. Keenan, J.M. Tormos, R. Hamilton, M.D. Catala, Transcranial magnetic stimulation and neuroplasticity, Neuropsychologia 37 (1999) 207– 217. [11] A. Pascual-Leone, J.M. Tormos, J.P. Keenan, F. Tarazona, C. Canete, M.D. Catala, Study and modulation of human cortical excitability with transcranial magnetic stimulation, J. Clin. Neurophysiol. 15 (1998) 333 –343. [12] J.R. Romero, D.J. Anschel, R. Sparing, M. Gangitano, A. PascualLeone, Subthreshold low frequency repetitive transcranial magnetic stimulation selectively decreases facilitation in the motor cortex, Clin. Neurophysiol. 113 (2002) 101–107. [13] H.A. Sackeim, The anticonvulsant hypothesis of the mechanism of action of ECT: current status, J. ECT 15 (1999) 5–26. [14] A. Sattin, The role of TRH and related peptides in the mechanism of action of ECT, J. ECT 76 (1999) 76 –92. [15] M.P. Szuba, J.P. O’Reardon, A.S. Rai, J. Snyder-Kastenberg, J.D. Amsterdam, D.R. Gettes, E. Wassermann, D.L. Evans, Acute mood and thyroid stimulating hormone effects of transcranial magnetic stimulation in major depression, Biol. Psychiatry 50 (2001) 22–27. [16] F. Tergau, U. Naumann, W. Paulus, B.J. Steinhoff, Low-frequency repetitive transcranial magnetic stimulation improves intractable epilepsy, Lancet 353 (1999) 2209. [17] W.H. Theodore, M.D. Hunter, R. Chen, F. Vega-Bermudez, B. Boroojerdi, P. Reeves-Tyer, K. Werhahn, K.R. Kelley, L. Cohen, Transcranial magnetic stimulation for the treatment of seizures, Neurology 59 (2002) 560–562. [18] F.C. Tortella, J.B. Long, Endogenous anticonvulsant substance in rat cerebrospinal fluid after a generalized seizure, Science 228 (1985) 1106–1108. [19] S.R. Weiss, X.L. Li, J.B. Rosen, H. Li, T. Heynen, R.M. Post, Quenching: inhibition of development and expression of amygdala kindled seizures with low frequency stimulation, NeuroReport 6 (1995) 2171–2176.