Inte rnatio nal Jo urnal o f Sc ie ntific & Eng inee ring Re se arc h, Vo lume 3, Issue 2, February -2012 1
ISS N 2229-5518
CNS Inflammation and Epileptogenesis
Francis I. Baffour
Abs tract – An association betw een acute CNS inf ections and seizures is reported in the literature. The etiology of these seizures is, in some cases, precipitated by epileptogenic pathogenesis of the inf ectious agent in the brain parenchyma. A diff erent etiology of seizures as a consequence of inf ections, independent of the inf ectious agent, relates to the inf lammatory reaction. Thirdly, the f ebrile response to inf ec tions may predispose the CNS to seizures due to the metabolic aberrations of f ever. Febrile states have also been implicated in so me cases of mesial te mporal lob e epilepsy. Current
attempts to control seizures w ith anti-inf lammatory treatments have show n contrasting result, theref ore specif ic agents need to be developed and tested.
Inde x Te rms – CNS Inf ections, Epileptogenesis, Fever, Immune Modulators, Inf lammation, Mesial Te mporal Lobe Ep ilepsy , Seizures
—————————— ——————————
he central ner vous system (CNS) is consider ed an immuno-pr ivileged site because of the pr esence of a blood–brain barr ier , a lack of a conventional lymphatic drainage, and an appar ently low traffic of monocytes and lymphocytes. It is, however , becoming clear that immune and inflammatory r eactions do occur in the CNS. The pathogenesis of infections which pr ecipitate these inflammatory r eactions, along w ith the physiological r esponse to infections have both been implicated in the genesis and pr omotion of seizur es (Vezzani and Gr anata,
2010). Epilepsy is a disabling neur ological disor der char acterized by r ecurr ing, unpr ovoked seizur es. It affects about 1% of the population of all ages and often r equir es lifelong medication. In about 30% of affected individuals, epilepsy is r efr actory to phar macological tr eatment, and sur gical r emoval of the epileptic focus is suitable only for a minor ity of such patients (Johnston, 2007). Understanding the etiology underlying the occurr ence of seizur es is necessary for devising novel therapeutic appr oaches.
Seizur es may occur as a dir ect consequence of the activity of pathogens in the CNS. Pathological and physiological aberr ations in CNS functions that can lead to epileptogenesis may r esult fr om one of the follow ing: mor phological changes dur ing life cycle of the pathogen that affect normal signaling in brain par enchyma; or the effect of the pathogen on the cir culation of nutrients and other essential factor s necessary for normal CNS functions.
———— ——— ——— ——— ———
Neurocystice rcosis
Cysticer ci often live asymptomat ically in the host for years, evading the immune system. Once in the CNS, the parasite goes thr ough a ser ies of stages as it develops in a fluid- filled cyst. The matur e vesicular lesions ar e also asymptomatic, and initially, ar e capable of suppr essing the host inflammatory r esponse. The lesions later become inflamed as the host’s r esponse gradually clear s the parasite. This phase is fr equently complicated by seizur es. In some instances, the lesions w ill r esolve completely. How ever in other cases, r esolution is incomplete, leaving a r esidual calcified lesion, w hich may be associated w ith chr onic epilepsy (White, 2010).
Cerebral malaria
The parasite P. falciparum is the causative or ganism of cer ebral malar ia. 70% of malaria cases occur in sub-Sahar an Afr ica, wher e childr en ar e most commonly affected (Guerra et al, 2005), such that malar ia may account for 40% of pediatric admissions to some hospitals, 10% of which may be due to cer ebral malar ia (Foster et al, 2005). The incidence of cer ebral malaria in malaria-endemic ar eas of sub- Saharan Africa is 1.12 cases per /1000 childr en per year , w ith a mortality of 18.6% (Krishna and Newton, 1998).
The sequestration of infected and non-infected erythr ocytes within cer ebr al vessels pr edisposes the CNS to epileptogenic lesions. Adher ence of P. falciparum in blood vessels r educes the micr o-vascular flow , and leads to hypoper fusion of the CNS. In addition, the pr esence of parasites inside erythr ocytes decr eases the ability of the cell to deform to facilitate its passage thr ough the microvasculatur e. Sequestration might happen as a consequence of cyto-adher ence of infected erythr ocytes to endothelial cells via P. falciparum der ived pr oteins on the infected erythr ocyte sur face attaching to ligands in the venules (Idr o et al., 2005).
IJSER © 2012 http :// www.ijser.org
Inte rnatio nal Jo urnal o f Sc ie ntific & Eng inee ring Re se arc h, Vo lume 3, Issue 2, February -2012 2
ISS N 2229-5518
In addition to the examples above, the literatur e r eports other epileptogenic infectious agents. These include: Echinococcus, Angiostrongylus cantonensis, Toxacara sp. Schistosoma sp. and var ious HHV species (Newton and Wagner , 2009).
A differ ent etiology of seizur es as a consequence of infections r elates to the inflammatory r eaction, independent of the infectious agent. Tw o main possibilities may explain the initiation of inflammator y r esponses within the CNS (Vezzani and Granata, 2010): the immune inflammatory pr ocess is initiated within the CNS, and the infiltration of blood-bor ne immune cells or cir culating inflammatory mediators is a consequent r esponse to this intrinsic event; or the CNS is the tar get of an inflammatory r esponse that or iginates within per ipheral lymphoid tissues. The evolving theory of inflammation-dr iven epileptogenesis as a r esult of pr o-inflammatory cytokines in the CNS has been descr ibed w ith data fr om animal models and clinical obser vation.
Evidence from human specimens
An incr eased IL-1α expr ession in micr oglia cells has been documented by immunohistochemistry in brain specim ens obtained fr om patients w ho have under gone temporal lobectomy as tr eatment for mesial temporal lobe epilepsy (MTLE) (Sheng et al., 1994). A subsequent study described the pr esence of r eactive astrocytes in hippocampal specimens sur gically r emoved fr om patients with MTLE who had a history of febrile convulsions. The expr ession of NF-κB, a transcr iption factor that r egulates expr ession of inflammation genes, specifically in lesioned hippocampi indicates the activation of an inflammatory cascade in brain par enchyma (Cr espel et al., 2002). The studies in human epilepsy specimens fr om MTLE patients support the existence of an inflammatory state sustained epileptic focus in the brain par enchymal cells.
Evidence from Animal Models
Studies conducted in exper imental animal models have identified several cytokines including, IL-1β, IL-1Ra, TNF- α, IL-6, as cr itical factors in sensitizing the CNS to seizur es (Scantlebury and Heida, 2010). Elevated levels of IL-1β in CSF have been shown to r educe the seizur e thr eshold in rats, making them mor e susceptible to seizur es after minimal induction by LPS-mediated pathways. Conversely, IL-1β r eceptor -deficient r ats ar e r esistant to seizur es induced by LPS (Heida et al., 2004). The r ole of TNF-α in seizur e sensitization have been investigated in rat models
of pneumococcal meningitis. By administer ing TNF-α- converting enzyme inhibitor s to these r ats, CSF levels of TNF-α w er e mar kedly decr eased, along w ith seizur es associated with the bacterial infection (Meli et al., 2004). Finally, in transgenic mice pr imed to be in a chr onic CNS inflammatory state by over-expr essing IL-6 or TNF-α, ther e was an obser ved incr ease in sensitivity to seizur es induced by glutamater gic agonist, along with a loss of GABA neur ons in the hippocampus (Smaland et al., 2003). These data suggest that pr o-inflammatory cytokines in the CNS contr ibute to the pr opensity of seizur e development.
A thir d etiology of seizur es in the setting of CNS infections stems fr om the acute metabolic disturbance associated w ith febr ile states. A febrile seizur e is a convulsion in a child tr igger ed by a fever . These convulsions occur w ithout any brain or spinal cord infection or other ner vous system cause. Febr ile seizur es ar e usually tr igger ed by fever s fr om ear infections, r oseola infantum, upper r espiratory infections and meningitis (Johnston, 2007).
Clinical Studies
Febr ile seizur es occur in ~5% of childr en <5 years old who develop an acute infection. A fundamental question about febr ile seizur es that r emains unansw er ed is why they develop in some childr en with a febr ile illness but not others. It does not appear to be the magnitude of fever which plays a r ole. Childr en with a lower fever at the time of seizur e have an incr eased r isk for subsequent convulsion with another febrile illness, perhaps because they have a low er thr eshold for seizur es in the fir st place (El-Rhadhi,
1998). While fever is defined as a temper atur e of at least
38.4°C, some clinical studies of febrile seizur es have accepted temperatur e values as low as 38°C, as high enough to pr ecipitate centr al nervous system (CNS) dysfunction (Al-Eissa, 1995). Clinically, it is difficult to determine the exact temper atur e at seizur e onset, and, sometimes, febr ile seizur es can occur as the pr esenting sign of febrile illness.
Although most febr ile seizur es do no harm and tw o-thir ds of initial cases have no witnessed r ecurr ence, in r ar e cases they ar e the first evidence of important epilepsy syndromes. Furthermor e, febr ile seizur es in ear ly childhood have been implicated in the development of epilepsy w ith mesial temporal scler osis in later life (Cendes,
2004).
Obser vations by Falconer , based on a ser ies of 100 patients who had sur gery for intractable tempor al lobe epilepsy
IJSER © 2012 http :// www.ijser.org
Inte rnatio nal Jo urnal o f Sc ie ntific & Eng inee ring Re se arc h, Vo lume 3, Issue 2, February -2012 3
ISS N 2229-5518
(TLE) in the 1960s show ed that many of these patients had mesial tempor al scler osis as an under lying pathol ogy. Additionally, a significant pr oportion of the patient population had episodes of pr olonged febr ile seizur es in their ear ly childhood (30% in the MTS gr oup compar ed with 6% in the gr oup w ithout MTS) (Falconer et al., 1964). With this data, a causal r elationship was suggested betw een pr olonged febr ile seizur es and MTS.
Animal Model of Febrile Seizures
To study the etiology of febr ile seizur es, animal models have been developed — an inflammatory dose of the bacter ial endotoxin, lipopolysaccharide (LPS), which evokes an immune r eaction and a fever r esponse (about 1-
1.5°C) is coupled with what is normally a sub -convulsant
dose of kainic acid (KA), in immatur e rats (P14) (Heida et
al., 2004). This model mimics the most essential featur es of febr ile seizur es: the immune r esponse and the fever . Hyperthermia (>38.3°C) can decr ease gamma -aminobutyr ic acid A (GABAA) r eceptor -mediated inhibition to a gr eater extent than it decr eases excitation, w hich may shift the balance towar ds excitation and contr ibute to seiz ur e generation (Leung et al., 2007). While this phenomenon has only been studied in hyperthermic models, it is pr oposed to
occur when temperatur e is incr eased fr om physiologic fever (Scantlebur y and Heida, 2010).
Evidence from Animal Models
Ther e is now ev idence fr om animal models of febrile seizur es that supports a link between pr olonged febrile seizur es and the development of epilepsy. In one study, pr olonged (30 min) hyperthermic seizur es maintained in rats (P10), by exposur e to heated dry air , r esulted r ecurr ent seizur es in 35% of animals in adulthood. The seizur es w er e br ief, non-convulsive and of focal limbic or igin. The adult rats with r ecurr ent seizur es also had mild to moderate learning and memory deficits (Dube et al., 2009). Although, this model does not captur e the main featur es of the most sever e form of MTLE, which include secondarily generalized seizur es and overt MTS, it has pr ovided some insights into the mechanisms that lead to r ecurr ent seizur es follow ing pr olonged febr ile seizur es.
Pr olonged hyperthermic seizur es r esults in an ear ly upr egulation in alpha-amino-3- hydr oxyl-5-methyl-4- isoxazole-pr opionate (AMPA) r eceptors lacking the GluR2 subunit in the hippocampus (Richichi et al., 2008). This type of AMPA r eceptor is calcium permeable, an d, when activated, can lead to pathological incr eases in intracellular calcium. This affects many pr ocesses leading to the
abnormal function or expr ession of specific ion channels or r eceptor s r esulting in the cr eation of seizur e generating networ ks (Chen et al., 2001).
Another known metabolic disturbance of fever in childr en is an incr ease in the r espirator y rate. It is unclear how ever , whether this effect leads to a r espir atory alkalosis. This is of impor tance since alkalosis incr eases neur onal excitability (Reid et al., 2009). Animal models of febr ile seizur es exhibit r espiratory alkalosis, and administration of CO2 br ings an abr upt end to the seizur es (Schuchmann et al., 2006). Ther e is still uncertainty as to whether alkalosis plays a r emar kable r ole in the etiology of febr ile seizur es in childr en.
Anti-COX Agents
The use of anti-inflammatory dr ugs to tr eat epilepsy has shown var ied r esults. Anti-inflammatory dr ugs w ith anti- COX activities have been shown to either r educe or exacerbate seizur es induced by kainic acid. For example, aspir in decr eases kainite seizur es, and has been shown to modulate the anticonvulsant activity of sodium valpr oate (Bing et al, 2000). How ever , in differ ent studies, pr etr eatment of r ats w ith indomethacin, aspir in, nimesulide, or selective COX-2 inhibitor s augmented kainate-induced seizur es (Baik et al, 1999). These appar ent dual effects of COX inhibitor s likely depend on their specific actions on the basal production of the various pr ostaglandins, and on the differ ent profiles of pr ostaglandins pr oduced dur ing seizur es in the various exper imental models (Baik et al, 1999). Post-seizur e administration of COX-2 inhibitors pr otects hippocampal neur ons fr om damage induced by seizur es and enhances functional r ecovery of cognitive functions after kainate seizur es (Benveniste et al, 2004). The above evidence supports a r elevant r ole of pr ostaglandins in the mechanisms of neur onal hyper excitablity.
IV Immunoglobulin
Intravenous administr ation of high doses of immunoglobulins has an anticonvulsive effect. IV inj ections of human globulin-N pr otected cats against gener alized seizur es induced by electr ical stimulation of the amygdala (Hir ayama et al, 1986). Immunoglobulins have been shown to inhibit NF-κB activation induced by TNF-α in endothelial cells and macr ophages (Ichiyama et al, 2004). The evidence ther efor e suggests that their anticonvulsant effect may be at least in part mediated by anti- inflammatory mechanisms.
IJSER © 2012 http :// www.ijser.org
Inte rnatio nal Jo urnal o f Sc ie ntific & Eng inee ring Re se arc h, Vo lume 3, Issue 2, February -2012 4
ISS N 2229-5518
Steroids
The use of ster oids in v arious forms is common for mor e sever e, tr eatment -r esistant forms of childhood epilepsy. Adr enocorticotr ophic hormone (ACTH) — a peptide that r eleases endogenous ster oids — has been used successfully as a tr eatment for infantile spasms, a sever e for m of childhood epilepsy that is r esistant to conventional anti- epileptic dr ugs (Avishai-Eliner et al, 2002). The success of ACTH have been shown empirically and confir med in randomized contr olled tr ials. Consequently, ACTH r emains a mainstay of ther apy for this condition (Bar am and Hatalski, 1998).
Ther e is pr eclinical evidence that suggests brain inflammation pr omotes neur onal hyper excitablity and seizur es. The r ole played by cytokines in epileptogenesis, and the effect of metabolic derangements associated w ith fever on the CNS have been investigated in some detail. Although these studies highlight sever al novel mechanisms that might be utilized as therapeutic tar gets, clinical evidence has shown inconsistent r esults. Data fr om animal models suggest that tar geting specific cytokines might be warranted. Sever al important questions r emain, such as: how acute febrile seizur es become prolonged, or why epilepsy occurs in some and not all childr en with febrile epilepsy; whether inflammator y mechanisms ar e important at all stages of epileptogenesis and epilepsy; whether patients all have a similar degr ee of inflammation; whether various epilepsy etiologies ar e associated w ith inflammation that can be tar geted therapeutically. A clinical tr ial of an IL-1β synthesis inhibitor was initiated in
2010 because this mechanism could contribute to differ ent types of seizur es. If this and other tr ials ar e successful, assessment of which patients ar e r esponding to anti- inflammatory ther apy might become possible and, hence, the under lying epilepsy pathologies in which inflammation is important might be determined.
The Author wishes to thank Dr . Solomon Moshé of the Department of Neuroscience at t he Albert Einstein College of Medicine, Br onx, NY, USA.
[1] Al-Eissa Y.A. (1995). Fe brile se izures: rate and risk fac to rs o f rec urre nce . J Child Ne uro l. 10 (4):315-9
[2] Av ishai-Eline r, S., Baram, T.Z., and Brunso n, K.L. (2002).
ACTH Tre atme nt o f Infanti le S pasms: Mec hanisms o f its
Effec ts in Mo dulatio n o f Ne uro nal Exc itability . Int Re v
Ne uro bio l. 49 :185-197
[3] Baik, E.J., Kim, E.J., and Lee , S.H. (1999). Cyc loo xyge nase -2 se lec tive inhibito rs agg rav ate kainic ac id induc e d se izure and ne uro nal ce ll de ath in the hippoc ampus. Brain Re s.
[4] Baram, T. Z. and Hatalski, C. G. (1998). Ne uro pe ptide - me diate d e xc itability : a key trigge ring mec hanism fo r se izure ge ne ratio n in the deve lo ping brain. Tre nds Ne uro sci.
[5] Be nve niste , M., Bo ug h, K., and Wilhe m, J. (2004).
Cycloo xyge nase inhibitio n afte r piloc arpine -induc e d status
e pile ptic us. Prog # 679.11. S an Diego: Socie ty o f
Ne uro sc ie nce .
[6] Bing, G., Jhoo , W.K., and Kim, H.C. (2000). Phe nido ne pre ve nts kainite induce d ne uro to xic ity v ia antio xidant mec hanisms. Brain Re s 874:15–23
[8] Ce ndes, F. (2004). Fe brile se izure s and me sial te mpo ral sc le rosis. Curr Opin Ne uro l. 17 (2):161-4
[9] Che n, K., Aradi, I., Tho n, N., Eg hbal-Ahmadi, M., Baram, T.Z., and So ltesz I. (2001). Pe rsiste ntly mo difie d h-c hanne ls afte r co mple x fe brile se izure s co nve rt the se izure -induc e d e nhance me nt o f inhibitio n to hy pe re xc itability . Nat Me d.
[10] Crespe l, A., Co ube s, P., Ro usse t, M.C., Brana, C., Ro ug ie r, A., Ro ndo uin, G., Boc kae rt, J., Baldy -Mo ulinie r, M., and Le rne r-Nato li, M. (2002). Inflammato ry re ac tio ns in human me dial te mpo ral lo be e pile psy with hippo c ampal sc le ro sis. Brain Re s. 952 (2):159-69
[11] Dubé, C.M., Bre wste r, A.L., and Baram, T.Z. (2009). Fe brile se izures: mec hanisms and re latio nship to e pile psy . Brain Dev . 31 (5):366-71
[12] El-Radhi, A.S . (1998). Lowe r deg ree o f feve r at the ini tial fe brile co nv ulsio n is assoc iate d with inc re ase d risk o f subse que nt co nv ulsio ns. Eur J Pae diatr Ne uro l. 2 (2):91-6
[13] Falco ne r, M.A., Se rafe tinide s, E.A., and Co rse llis, J.A. (1964).
Etio logy and Pathoge ne is o f Te mpo ral Lo be Epile psy . Arc h
Ne uro l. 10:233-48
[14] Foste r, D., Marsh, K., and Waruiru, C. (2005). Indic ato rs o f life -thre ate ning malaria i n Afric an c hildre n. N Eng l J Me d.
[15] Gue rra, C.A., Hay, S .L., My int, H.Y., Noo r, A.M., and S now, R.W. (2005). The g lo bal distributio n o f c linic al e piso des o f Pla smodium falcipa rum malaria. Nature . 434:214–217
IJSER © 2012 http :// www.ijser.org
Inte rnatio nal Jo urnal o f Sc ie ntific & Eng inee ring Re se arc h, Vo lume 3, Issue 2, February -2012 5
ISS N 2229-5518
[16] He ida, J.G., Bo issé, L., and Pittman, Q.J. (2004).
Lipo po ly sacc haride -induc e d fe brile co nv ulsio ns in the rat:
sho rt-te rm se que lae . Epile psia. 45 (11):1317-29
[17] Hiray ama, H., Kurimo to, T., and Wada, S . (1986).
Antie pile ptic e ffec ts o f g lo bulin-N, an intac t human immunog lo bulin and its tissue distributio n in kindle d c ats. Int J Clin Pharmaco l The r To xico l 24:109–22
[18] Ic hiy ama, T., Haseg awa, M., and Ue no , Y. (2004).
Intrav e no us immunog lo bulin inhibits NF-kappaB ac tiv atio n and affe c ts Fcg amma rece pto r e xpre ssio n in mo nocy tes/mac ro phage s. Nauny n Sc hmie de be rgs Arc h Pharmac o l 369:428–33
[19] Idro , R., Je nkins, N.E., and Newto n, C.R. (2005).
Pathoge nesis, c linic al fe ature s, and ne uro log ic al o utco me o f ce re bral malaria. Lanc e t Ne uro l. 4 (12):827-40
[20] Jo hnsto n, M.V ., (2007). Se izures in c hildhoo d. In: Ne lso n Te xtboo k o f Pe diatric s. Klieg man RM, Be hrman RE, Je nso n HB, S tanto n BF, (e ds.), 18th e d. Philade lphia, Pa: S aunde rs Elsev ie r; 2007:c hap 593
[21] Krishna, S . and Ne wto n, C.R. (1998). Seve re falc iparum malaria i n c hildre n: c urre nt unde rstanding o f patho phy sio logy and suppo rtiv e tre atme nt. Pharmaco l The r.
[22] Le ung , L.S., Liu, X., Qu, L., and Wu, C. (2007). Hy pe rthe rmia dec re ases GABAe rg ic sy naptic transmissio n in hippo c ampal ne uro ns o f immature rats. Ne uro bio l Dis. 27 (3):320-7
[23] Me li, D.N., Loe ffle r, J.M., Baumann, P., Ne umann, U., Buhl, T., Le ppe rt, D., and Le ib, S.L. (2004). In pne umococcal me ning itis a nove l wate r-so luble inhibito r o f matrix me tallo pro te inases and TNF-al pha co nve rting e nzy me atte nuate s se izures and injury o f the ce re bral co rte x. J Ne uro immuno l. 151 (1-2):6-11
[24] Re id, A.Y., Galic , M.A., Teskey , G. C., and Pittman, Q.J. (2009). Fe brile se izure s: c urre nt v ie ws and investig atio ns. Can J Ne uro l Sc i. 36 (6):679-86
[25] Ric hic hi, C., Bre wste r, A.L., Be nde r, R.A., S imeo ne, T.A., Zha, Q., Yin, H.Z., We iss J.H., and Baram, T.Z. (2008). Mec hanisms o f se izure -induc e d ' transc riptio nal c hanne lo pathy' o f hy pe rpo larizatio n-ac tiv ate d cyclic nuc leo tide g ate d (HCN) c hanne ls. Ne uro bio l Dis. 29 (2):297-
305
[26] S maland, H., Huitro n-Re se ndiz, S., Masliah, E., Criado, J., He nrikse n, S.J., and Campbe ll, I.L. (2003). Pro fo und inc re ase in se nsitiv ity to g lutamate rg ic - but no t c ho line rg ic ago nist- induc e d se izures in transge nic mice with astrocy te pro duc tio no f IL-6. J Ne urosc i Res. 73 (2):176-87
[27] Sc antlebury , M.H. and He ida, J.G. (2010). Fe brile se izures and te mpo ral lo be e pile ptog e nesis. Epile psy Re s. 89 (1):27-33
[28] Sc huc hmann, S., Sc hmitz, D., Rive ra, C., V anhatalo , S., S alme n, B., Mac kie, K., S ipilä, S .T., Voipio, J., and Kaila, K. (2006). Expe rime ntal fe brile se izure s are prec ipitate d by a hy pe rthe rmia-induce d re spirato ry alkalosis. Nat Me d.
[29] S he ng , J.G., Boo p, F.A., Mrak, R.E., and Griffin, W.S . (1994) Inc re ase d ne uro nal be ta-amy lo id prec urso r pro te in e xpre ssio n in human te mpo ral lo be e pile psy : associatio n with inte rle ukin-1 alpha i mmuno re ac tiv ity . J Ne uroc he m. 63 (5):1872-9
[30] Vezzani, A. and Granata, T. (2005) Brain inflammatio n in e pile psy: e xpe rime ntal and c linic al e v ide nce . Epile psia. 46 (11):1724-43
[31] Wag ne r, R.J. and Ne wto n, O.R. (2009). Do he lminths c ause e pile psy? Parasite Immuno l 31(11):697-705
[32] White , A.C. Jr. (2010). Why are the re se izures in ne urocystice rco sis?: is it in the ge nes? J Infe c t Dis. 202 (8):1152-3
[33] S tudy o f V X-765 in subjec ts with tre atme nt-re sistant partial e pile psy. Re trieve d Dece mbe r 2010, http://c linic altrials.gov/c t2/sho w/NCT01048255
.
IJSER © 2012 http :// www.ijser.org