DRUG RESISTANT EPILEPSY Review Submitted for Partial Fulfillment of Master Degree in Neuropsychiatry By Ahmed Mohammed Fahmy Alharoun Supervised By Prof. / Mohammed Yasser Metwally Professor of neuropsychiatry Faculty of medicine - Ain Shams University Dr. / Ahmed Abdel-Moniem Gaber Assistant professor of neuropsychiatry Faculty of medicine - Ain Shams University Dr. / Haitham Hamdy Salem Lecturer of neuropsychiatry Faculty of medicine - Ain Shams University Faculty of medicine Ain Shams University 2010 www.yassermetwally.com
Transcript
DRUG RESISTANT EPILEPSY
Review Submitted for Partial Fulfillment ofMaster Degree in Neuropsychiatry
By
Ahmed Mohammed Fahmy Alharoun
Supervised By
Prof. / Mohammed Yasser MetwallyProfessor of neuropsychiatry
Faculty of medicine - Ain Shams University
Dr. / Ahmed Abdel-Moniem GaberAssistant professor of neuropsychiatry
Faculty of medicine - Ain Shams University
Dr. / Haitham Hamdy SalemLecturer of neuropsychiatry
Gross pathology 53 Malformations of cortical development 54 Do seizures start within the lesion or the
perilesional region? 79 Is epileptogenesis primarily a result of circuit
abnormalities or cellular/molecular defects? 85 Expression of drug transporters in pathological
lesions associated with refractory epilepsy 92
Management 94 Anti-epileptic drugs 98 Epilepsy surgery 98 Ketogenic diet 106 Vagus nerve stimulation 108 Treatments under investigation 110
Discussion 113
Summary 123
Recommendations 126
References 127
Arabic Summary ---
Introduction & Aim of the Work
1
INTRODUCTION
In many patients with epilepsy, seizures are well-controlled with currently available anti-epileptic drugs.However, seizures persist in a considerable proportion of thesepatients (Remy and Beck, 2006). It is estimated that 20-25% ofthe epileptic patients fail to achieve good control with thedifferent anti-epileptic drug treatments, developing refractoryepilepsy (Lazarowski et al., 2007).
It is not known why and how epilepsy becomes drugresistant in some patients while others with seemingly identicalseizure types and epilepsy syndromes can achieve seizurecontrol with medication (Schmidt and Loscher, 2005).
The causes of drug resistant epilepsy are numerous, manydue to abnormalities in brain maturation, severe brain injurieswith resultant irreversible changes to cerebral neuroglialorganization and inhibitory neuron function, kindlingphenomenon, seizure-induced disturbances of oxygen supply,as well as acquired (or hereditary) changes in transporterproteins of the blood-brain barrier which function in the effluxof anti-epileptic drugs from the brain (Awasthi et al, 2005).
Several hypotheses have been formulated to explain thepathogenesis of drug resistance in epilepsy:
Introduction & Aim of the Work
2
1. The drug fails to reach the neuronal target (transporterhypothesis). This includes the cerebrovascular over-expression of multidrug transporter proteins.
2. The drug fails to act at the neuronal target (targethypothesis). (Marchi et al., 2010)
Regarding transporter hypothesis, analysis of multidrugtransporter expression in human and experimental epilepsy hasfavored the concept that increased levels of these transportersare present in epileptic tissue, thus lowering intraparenchymaldrug concentrations and rendering several anticonvulsant drugsless effective (Remy et al., 2003a).
The abnormal parenchyma cells present in theepileptogenic tissues from different refractory epilepsysyndromes, such as dysembryoplastic neuroepithelial tumors,focal cortical dysplasia and hippocampal sclerosis, couldexpress multidrug transporters P-gp or MRP1 constitutively(Sisodiya et al., 2002).
According to the target hypothesis, pharmaco-resistanceoccurs when target sites are structurally and/or functionallymodified in such a way that they become less sensitive to anti-epileptic drugs. A large number of targets for anti-epilepticdrugs have been identified in the brain, many of which undergomolecular changes during chronic epilepsy. So far, a reducedsensitivity of drug targets to anti-epileptic drugs in chronichuman and experimental epilepsy has been suggested for thevoltage-gated Na+ channel and the GABAA receptor.
Introduction & Aim of the Work
3
Cumulative evidence suggests that there is reduced pharmaco-sensitivity to Na+ channel-acting drugs in patients withintractable temporal lobe epilepsy (Remy et al., 2003a).
Introduction & Aim of the Work
4
At the molecular level, altered drug targets may arise dueto the transcriptional regulation of ion channel subunit genes. Ithas become clear over the recent years that seizures causecoordinated and cell-specific transcriptional changes that resultin either an up- or down-regulation of families of ion channelmRNAs. This, in turn, causes alterations in the subunitcomposition or density of ion channels, resulting in alteredintrinsic and synaptic neuronal properties (Beck, 2007).
Introduction & Aim of the Work
5
AIM OF THE WORK
Our aim is to explore the pathological mechanismsunderlying drug resistance in epilepsy and to search reviewstrying to explain how and why drug resistance occurs in somepatients with epilepsy to reach better management of thesepatients.
Definition, Incidence & Risk Factors
6
DEFINITION, INCIDENCE ANDRISK FACTORS
Although many new antiepileptic drugs have beendeveloped in the past decade, epilepsy remains resistant to drugtherapy in about one-third of patients. Approximately 20% ofpatients with primary generalized epilepsy and up to 60% ofpatients who have focal epilepsy develop drug resistance duringthe course of their condition, which for many is lifelong (Siegel,2004)
Considering that epilepsy is one of the most commonchronic neurologic disorders, drug-resistant epilepsy is a majorpublic health problem. The consequences of drug-resistantepilepsy can be quite severe, including mortality rates that are 4to 7 times higher in people with drug-resistant seizures. It is notknown why and how epilepsy becomes drug resistant in somepatients while others with seemingly identical seizure types andepilepsy syndromes can achieve seizure control withmedication (Schmidt and Loscher, 2005).
People with pharmaco-resistant epilepsy are about 2 to10 times more likely to die compared with the generalpopulation. The risk is inversely linked to seizure control.“Sudden unexpected death in epilepsy” is the most frequent
Definition, Incidence & Risk Factors
7
type of death in patients with pharmaco-resistant epilepsy. Thiscategory excludes deaths from trauma or drowning. Case-control studies have shown that the risk of sudden unexpecteddeath is closely and inversely associated with seizure control;the rate is significantly higher in patients who have a higherfrequency of convulsive seizures. In addition, freedom fromseizures, achieved after successful epilepsy surgery, reduces therisk of death from all causes (Pati and Alexopoulos, 2010).
Despite advances in anti-epileptic drug therapy andepilepsy surgery in recent years, intractable epilepsy remains amajor clinical problem. An important characteristic ofmedically intractable (pharmaco-resistant) epilepsy is that mostpatients with refractory epilepsy are resistant to several, if notall anti-epileptic drugs, even though these drugs act by differentmechanisms (Kwan and Brodie, 2000).
During treatment with a variety of different anti-epilepticdrugs, as many as 20–40% of newly treated patients withepilepsy will not enter long-term remission for several yearsand despite medical management with modern Anti-epilepticdrugs , a number of these patients continue to have drug-resistant epilepsy with frequent debilitating seizures (Schmidtand Loscher, 2005).
Definition, Incidence & Risk Factors
8
The exact fraction of epilepsy patients who areconsidered refractory varies in the literature, mostly because thecriteria for classification as pharmaco-resistant have varied. Asubstantial proportion (about 30%) of epilepsy patients do notrespond to any of two to three first line Anti-epileptic drugs ,despite administration in an optimally monitored regimen(Remy and Beck, 2006).
Although no single accepted definition exists of drugresistant epilepsy, different definitions may be appropriate,depending on the type of seizure and epilepsy syndrome and thepurpose for which the definition is used. Definitions usuallyinclude the number of anti-epileptic drug failures and theminimal remission or seizure frequency during a specifiedduration of therapy (Schmidt and Loscher, 2005).
The different definitions of refractoriness emergedepending on the context. All are based on the 3 maincomponents of intractability: number of Anti-epileptic drugspreviously taken, frequency of seizures and duration of non-controlled epilepsy (Beleza, 2009).
In investigational studies, criteria of refractorinessinclude: (1) absence of response to 2 Anti-epileptic drugstolerated at reasonable doses; (2) minimum frequency of
Definition, Incidence & Risk Factors
9
seizures (e.g. 1 seizure per month) to be considered refractoryor the duration of minimum remission (e.g. 6-12 months) to bequalified as nonrefractory, and (3) duration of 1 year to 1decade of non-controlled epilepsy. Depending on the criteriaapplied, the frequency of refractory epilepsy varies from 10 to37.5% (Kwan and Brodie, 2000).
A flexible scale of refractoriness has been developed forclinical use and classifies epilepsy as potential (no seizurefreedom with Anti-epileptic drugs taken less than 1 year andpredictive factors for refractoriness), probable (no seizurefreedom more than 1 year with at least 2 Anti-epileptic drugs )or definitely refractory (catastrophic epilepsy or no freedom ofseizure for more than 1 year after 5 years of treatment with atleast 3 Anti-epileptic drugs ) depending on the duration ofepilepsy and medical treatment, seizure control and number ofAnti-epileptic drugs used. A sub-classification of refractorinessas acceptable or inacceptable was also included, taking intoaccount the patients’ impression of the impact of epilepsy(seizures, co-morbidity and adverse effects of Anti-epilepticdrugs ) on their quality of life (e.g. a patient with acceptablerefractory epilepsy presenting infrequent nocturnal seizuresmay become definitely refractory with the occurrence of diurnalgeneralized tonic-clonic seizures affecting employment,education and driving) (Starreveld and Guberman, 2006).
Definition, Incidence & Risk Factors
10
It is mandatory to exclude false refractoriness related tonon-epileptic seizures, inadequate anti-epileptic drugs,noncompliance and seizure-precipitating factors. Video-EEGmonitoring is an essential tool in this process, aiming toperform a differential diagnosis of paroxysmal events and acorrect classification of seizures and epileptic syndromes. Non-epileptic events more frequently found include cardiovascularsyncopes sleep diseases and psychogenic events (Beleza, 2009).
False pharmaco-resistance may not be easilyrecognizable, and this possibility needs to be investigated in anypatient presenting with difficult-to-control seizures. Up to 30%of patients referred to clinics with a diagnosis of pharmaco-resistant epilepsy may have been misdiagnosed, and many canbe helped by optimizing their treatment. Causes of falsepharmaco-resistant epilepsy include misdiagnosis of epilepsy(i.e. patients with psychogenic non-epileptic seizuresmisdiagnosed and inappropriately treated with multipleantiepileptic drugs) , misdiagnosis of epilepsy type leading toinappropriate drug selection (i.e. misdiagnosis of temporal lobeseizures for absence seizures, or vice versa), Inappropriateassessment of response or lack of response (i.e. druginteractions leading to increased side effects and decreasedtolerability), Inappropriate dosage and inappropriate patientbehavior (i.e. poor compliance) (Pati and Alexopoulos, 2010).
Definition, Incidence & Risk Factors
11
Based on partly prospective clinical observations in aseries of patients with newly diagnosed epilepsy, those who didnot achieve complete seizure control for 12 consecutive monthswith the first two or three Anti-epileptic drugs were given thepredictive diagnosis of refractory or drug-resistant epilepsy(Kwan and Brodie, 2000).
In general, many experts would agree that whenever apatient does not become seizure free for 12 months during long-term treatment with several suitable Anti-epileptic drugs atmaximal tolerated doses, the epilepsy can be broadly classifiedas drug-resistant, pharmaco-resistant, or medically refractory(Schmidt and Loscher, 2005).
Refractory epilepsy is established when there isinadequate seizure control despite using potentially effectiveAnti-epileptic drugs at tolerable levels for 1-2 years, andexcluding non-epileptic events and poor compliance (Beleza,2009).
Epidemiologic studies suggest three different patterns ofdrug resistance in epilepsy: de novo, progressive, and waxing-and-waning.
De novo drug resistanceIn some patients, resistance is present from the time of
onset of the very first seizure, before any antiepileptic drug is
Definition, Incidence & Risk Factors
12
even started. One landmark study showed that patients withnewly diagnosed epilepsy for whom the first drug wasineffective had only an 11% probability of future success,compared with 41% to 55% in patients who had had to stoptaking the drug because of intolerable side effects oridiosyncratic reactions. Most patients for whom the first drugfails will be resistant to most and often all antiepileptic drugs(Kwan and Brodie, 2000).
Progressive drug resistanceIn some patients, epilepsy is initially controlled but then
gradually becomes refractory. This pattern may be seen, forexample, in childhood epilepsies or in patients withhippocampal sclerosis (Berg et al., 2006).
Waxing and waning resistanceIn some patients, epilepsy has a waxing and waning
pattern: i.e., it alternates between a remitting (pharmaco-responsive) and relapsing (pharmaco-resistant) course. Patientsthought to have drug-resistant epilepsy may become seizure-free when other drugs are tried. Changes in drug bioavailability,local concentration of the drug in the brain, receptor changes,the development of tolerance, and interactions with newmedications may be implicated, though the exact mechanism isnot understood (Loscher and Schmidt, 2006).
Definition, Incidence & Risk Factors
13
Conceptually, the variable response to antiepileptic drugscan be attributed to factors related to the disease, the patient,and the drugs, or to other unknown factors. These factors arenot mutually exclusive and may be either constitutive oracquired during the course of the disease. Factors related to the disease: These factors include
etiology, epilepsy progression resulting in persistent changesof the epileptogenic network, and alterations of drug targetsor drug uptake into the brain.
Factors related to the drugs: Several drug-related factorshave been implicated, such as the development of tolerance,lack of antiepileptogenic (disease-modifying) actions tointerrupt the ongoing process of epileptogenesis rather thanonly suppressing seizures, and paucity of drugs with specificmechanisms of action tailored to difficult-to-controlepilepsies.
Patient characteristics: Variability in response (efficacyand adverse effects) to each antiepileptic drug can be due tointerindividual differences in any of four interrelatedfundamental factors: DNA, RNA, proteins, or metabolites.Age-related changes in pharmacokinetic andpharmacodynamic variables may contribute to age-dependentpharmaco-resistance. Least studied are environmental factorsthat may play a role in the development or expression ofpharmaco-resistance (Pati and Alexopoulos, 2010).
Definition, Incidence & Risk Factors
14
In epilepsy, 3 prognostic groups are generallyconsidered: (1) spontaneous remission (20–30%) as seen inbenign epilepsy with centrotemporal spikes or childhoodabsences; (2) remission on Anti-epileptic drugs (20–30%) asoccurs in most focal epilepsy and myoclonic juvenile epilepsysyndromes; (3) persistent seizures under Anti-epileptic drugs(30–40%) among which refractory epilepsy is included (Kwanand Sander, 2004).
The pathogenesis of refractoriness is multifactorial andvariable and could include genetic and environmental factors.The underlying syndrome or causative pathology is a majorfactor in determining drug response. Certain structuralabnormalities, such as hippocampal sclerosis, appear to beparticularly pharmaco-resistant. At the molecular level, changesin the neuronal network and the composition or functioning ofneurotransmitter receptors also may play a role (Kwan andBrodie, 2005).
A younger age at onset of epilepsy predictsrefractoriness. Seizures in the immature brain of a child mayresult in nonpruning of neurons and contribute to high numbersof gap junctions, which leads to abnormal connectivity, thehyper connected cortex (Ko and Holmes, 1999).
Definition, Incidence & Risk Factors
15
High seizure frequency (more than 1 seizure per month)occurring soon after the diagnosis of epilepsy either before orafter treatment onset correlates with refractoriness in the shortterm (2–4 years) and long term (30–35 years) (Berg et al.,2001).
In focal epilepsy, hippocampal sclerosis, corticaldysplasia and hemorrhage are associated with refractoriness.Depression has recently also been associated with lack ofresponse to Anti-epileptic drugs . Neurobiological processesthat underpin depression may interact with those producingseizures to increase the extent of brain dysfunction and therebythe likelihood of developing pharmacoresistant epilepsy (Hitiriset al., 2007).
Electroencephalography is useful for predictingrefractoriness. The quantity of interictal spikes is predictive ofseverity in temporal lobe epilepsy. Oligospikers, patients withtemporal lobe epilepsy with less than 1 spike per hour, correlatewith less severe epilepsy. In addition, some studies describe theassociation between multifocal spikes and intractability (beleza2009).
Factors that have been associated with treatment-resistantepilepsy include:
• Early onset of seizures
Definition, Incidence & Risk Factors
16
• Long history of poor seizure control• Having more than one type of seizure• Remote symptomatic etiology (eg, patients with a history of
brain infection or head trauma)• Certain structural abnormalities (eg, cortical dysplasia)• Certain abnormalities on electroencephalography (EEG)• Cognitive disability• History of status epilepticus
(Pati and Alexopoulos, 2010)
Pharmacoresistance
17
PHARMACORESISTANCE
Transporter hypothesis P-glycoprotein as a member of a transporter
superfamily Substrates and inhibitors of p-glycoprotein Other efflux transporters in the brain Cerebral expression of p-glycoprotein Control of p-glycoprotein expression Mechanisms of over-expression Are antiepileptic drugs substrates of drug
transporters?
Target hypothesis Voltage-gated Na+ channels Other types of voltage-gated channels Neurotransmitter systems: GABA Genetic control of anti-epileptic drug targets
Pharmacoresistance
18
In human epilepsy, one of the most common neurologicaldiseases, the development of resistance to anticonvulsanttherapy is a crucial clinical problem, but the mechanisms ofdrug resistance have remained elusive (Remy et al., 2003a).
In the presence of adequate, carefully monitored serumanti-epileptic drug levels, drugs have to traverse the blood brainbarrier. Consequently, one scenario to explainpharmacoresistance could be that sufficient intraparenchymalanti-epileptic drug concentrations are not attained, even in thepresence of adequate anti-epileptic drug serum levels. Such aphenomenon could arise via an enhanced function of multidrugtransporters that control intraparenchymal anti-epileptic drugconcentrations. Following permeation into the CNSparenchyma, drugs have to bind to one or more targetmolecules to exert their desired action. Thus,pharmacoresistance may also be caused by a modification ofone or more drug target molecules. These modifications wouldthen cause a reduced efficacy of a given anti-epileptic drug atthe target (Remy and Beck, 2006).
Consequently, two main hypotheses have been proposedto account for Pharmacoresistant epilepsy.
The first hypothesis contends that pharmaco-resistancearises because Anti-epileptic drugs do not gain access to theirsites of action in the brain. This phenomenon is thought to be
Pharmacoresistance
19
caused by over-expression of drug efflux transporters at theblood brain barrier that limit anti-epileptic drug access to thebrain. Furthermore, this can also occur in glial and neuronalmembranes, potentially reducing drug efficacy by restrictingaccess to intracellular target sites. Because of the centralimportance of multidrug transporters, this hypothesis has beendesignated “transporter hypothesis”.The target hypothesis, on the other hand, contends that targetreceptor sites are somehow altered in the epileptic brain so thatthey are much less sensitive to the anticonvulsant effects ofsystemically administered drugs (Beck, 2007).
Transporter hypothesis
An important characteristic of pharmacoresistant epilepsyis that most patients with refractory epilepsy are resistant tomost, and often all, anti-epileptic drugs . As a consequence,patients not controlled on monotherapy with the first anti-epileptic drug have a chance of only about 10% or lower to becontrolled by other anti-epileptic drugs, even when using anti-epileptic drugs that act by diverse mechanisms. This point tononspecific and possibly adaptive mechanisms, such asdecreased drug uptake into the brain by seizure-induced over-expression of multidrug transporters in the blood brain barrier(Loscher and Potschka, 2002).
Pharmacoresistance
20
For drugs to enter the brain, they must traverse eitherblood brain barrier or the blood CSF barrier. Because of theseanatomical barriers, entry of drugs into the brain is restricted(Pardridge, 1999).
The restrictive nature of the brain microvessel endothelialcells that form the blood brain barrier is due in part to theformation of tight junctions between the cells and to the lack oftransendothelial pathways such as transcellular channels orfenestrations. Consequently, brain capillaries restrict thepenetration of hydrophilic, polar, large, or protein-boundcompounds, whereas nonpolar, highly lipid-soluble drugspenetrate easily through the blood brain barrier by passivediffusion. With respect to the blood CSF barrier, for a drug toenter the CSF it must pass through the choroid plexus. Becausecapillary endothelial cells of the choroid plexus are fenestratedand lack tight junctions, the permeation barrier within thechoroid plexus exists at the level of the epithelial cells liningthe surface. These epithelial cells are mainly joined by tightjunctions, which restrict entry of water-soluble molecules(Spector, 2000).
However, apart from passive diffusion, drugs may alsoenter and leave the brain by carrier-mediated transportprocesses. In this respect, the recent finding of multidrugtransporters of the ATP-binding cassette superfamily, such as
Pharmacoresistance
21
P-glycoprotein (P-gp) and multidrug resistance-associatedprotein (MRP), in the endothelial cells of the blood brain barrieris of particular interest, since these outwardly directed activeefflux mechanisms appear to act as an active defensemechanism, limiting brain accumulation of many lipophilicdrugs. Furthermore, both P-gp and MRP are expressed inchoroid plexus epithelial cells that form the blood CSF barrier(Loscher and Potschka, 2002).
The primary function of these proteins is to pumplipophilic drugs and other xenobiotics out of cells and therebyprevent the accumulation of potentially toxic substances. Indoing so, however, these proteins may also decrease theefficacy of pharmacological agents by limiting their access totarget tissues in the brain (Loscher, 2007).
A number of drug transporter genes and their proteins areover-expressed in the blood brain barrier of individuals withrefractory epilepsy. This has been demonstrated in tissues takenfrom epileptic foci at the time of resective surgery. Specificdata indicate that there is a 130% increase in the expression ofMDR1, the gene encoding for P-gp, a 180% increase in MRP5,and a 225% increase in MRP2 in brain capillary endothelialcells isolated from epileptic individuals in comparison withnon-epileptic controls. Additional data indicate that P-gp,
Pharmacoresistance
22
MRP1, and MRP2 are also over-expressed in glial and/orneuronal cells of patients with intractable epilepsy(Dombrowski et al., 2001).
Based on the assumption that penetration of drugs fromblood into brain and CSF depends mainly on the drugs’ lipidsolubility, drugs required to act within the brain, such as anti-epileptic drugs, have generally been made lipophilic. Anti-epileptic drugs penetrate into the CSF by simple diffusion andLipid solubility plays the major role in determining thedifference in rate of entry of anti-epileptic drugs into the brain.However, one anti-epileptic drug, valproate, did not fit into thisscheme. Valproate is almost completely ionized at physiologicPH. However, valproate penetrated into the CSF and brain veryrapidly. Indeed, valproate was the first anti-epileptic drug forwhich an active transport in the blood CSF barrier and bloodbrain barrier has been proposed (Loscher and Potschka, 2002).
P-glycoprotein as a member of a transportersuperfamily
P-gp is believed to confer the multidrug resistancephenotype by reducing intracellular drug accumulation throughits function as an active efflux pump. It belongs to a highlyconserved protein superfamily, the ATP-binding cassette
Pharmacoresistance
23
proteins, which has more than 100 members and can be foundin all kinds of organisms (Kwan and Brodie, 2005).
With few exceptions, they function as active pumps atthe cell membrane through hydrolysis of ATP to drive the fluxof their substrates against the concentration gradient. Thesubstrate range for these proteins is diverse and includes drugs,nutrients, amino acids, sugars, peptides, pigments, and metals(Silverman, 1999).
At least 48 human ATP-binding cassette (ABC) geneshave been identified and grouped under subfamilies ABCA(ABC1), ABCB (MDR/TAP), ABCC (MRP/CFTR), ABCD(ALD), ABCE (OABP) and ABCF (GCN20), and ABCG (White)(Dean et al., 2001).
P-gp is encoded by a small gene family, comprising twogenes in humans, designated MDR1 (systematic name ABCB1)and MDR2 (ABCB4), located near each other on chromosome7q21.1. It is encoded by three genes in rodents, mdr1a, mdr1b,and mdr2. Human MDR1 and rodent mdr1a and 1b encode thedrug transporter associated with multidrug resistance, whereasMDR2 and mdr2 do not confer the multidrug-resistancephenotype. The rodent mdr1a and mdr1b together are believedto cover the same tissue distribution and function as the singlehuman MDR1 P-gp (Kwan and Brodie, 2005).
Pharmacoresistance
24
Human P-gp is an integral membrane protein with 1,280amino acids. Similar to other ATP-binding cassette transporters,P-gp comprises two homologous halves, each consisting of onetransmembrane domain containing six segments, and onenucleotide-binding domain or ATP-binding cassette unit. Theexact three-dimensional structure and mechanism of action ofP-gp are still unclear (McKeegan et al., 2003).
Fig. (1) Typical structure of the ATP-binding cassettetransporters (Lazarowski et al., 2007).
Substrates and inhibitors of p-glycoproteinUnlike most energy-dependent pumps, P-gp is highly
promiscuous. Hundreds of structurally and chemically unrelatedcompounds, as diverse as anthracyclines (doxorubicin),alkaloids (vincristine), specific peptides (cyclosporine A),steroid hormones (hydrocortisone), local anesthetics(dibucaine), and dye molecules (rhodamine 123), have beenidentified as substrates. How P-gp recognizes such a diverse
Pharmacoresistance
25
range of compounds is largely unexplained (Kwan and Brodie,2005).
Ever since the role of P-gp was recognized in drugresistance, inhibitors to circumvent its functions have beensought. This has yielded another long list of compounds thatincrease the sensitivity of cells in vitro to cytotoxic substratedrugs. The mechanisms of action of these inhibitors ormodulators are largely unknown, but many of them aresubstrates for P-gp, suggesting that they act by competitiveinhibition (Avendano and Menendez, 2002).
Table (1) P-gp substrates (Kwan and Brodie, 2005).
Anticancer drugs Other cytotoxic agents SteroidsDoxorubicin Colchicine AldosteroneDaunorubicin Emetine DexamethasoneVinblastine PuromycinVincristine MiscellaneousEtoposide Peptides DigoxinTeniposide Leupeptin Protease inhibitorsPaclitaxel Pepstatin A LoperamideDocetaxel Gramicidin D Rhodamine 123Actinomycin D Nonactin 99mTc-SESTAMIBIMitoxantrone Triton X-100
Table (2) P-gp inhibitors (Kwan and Brodie, 2005).Calciumchannelblockers
Assessing the role of P-gp in drug resistance of CNSdisorders such as epilepsy is further complicated by thepresence of other efflux transporters in the brain, which arerelatively less well characterized. Among them, the best studiedis the multidrug resistance protein [MRP, also called multidrugresistance-related or -associated protein] family with at leastnine members identified in humans (Borst et al, 2000).
Although also capable of conferring multidrug resistance,MRPs differ from (and overlap with) P-gp in substratespecificity, structure, tissue distribution, and possiblephysiologic functions (Seelig et al., 2000).
With respect to the role of MRPs in blood brain barrierfunction, differences in the cellular location of MRPs have to beconsidered. Whereas MRP2 is located in apical cell membranes,which is the appropriate position for a protective role, otherMRPs, including MRP1, MRP3, and MRP5, are locatedbasolaterally, so that over-expression of the latter MRPs inbrain capillary endothelial cells would not reduce entry of drugsinto the brain (Dombrowski et al., 2001).
Pharmacoresistance
27
MRP2
Using cDNA arrays, a 225% increase in MRP2 geneexpression was found in brain capillary endothelial cells frompatients with drug resistant temporal lobe epilepsy (TLE),suggesting that MRP2 expression changes may play animportant role in resistance to anti-epileptic drugs bydecreasing the permeability of anti-epileptic drugs across theblood brain barrier (Dombrowski et al., 2001).
Both inhibition of MRP2 and lack of MRP2 result in asignificant increase of drug levels in the brain. The localizationof MRP2 to the luminal surface of the brain capillaryendothelium, and the wide spectrum of drugs accepted assubstrates by MRP2 implicate that this transporter may be asimportant as P-gp in blood brain barrier function. MRP2 and P-gp have an overlapping substrate spectrum which is knownfrom substrate recognition studies on P-gp and MRPs (Loscherand Potschka, 2002).
In addition to the MRP2 gene, the genes encoding forMRP3 and MRP5 were found to be significantly over-expressedin patients with refractory epilepsy, whereas expression of theMRP1 gene was not significantly altered. Of these MRP genes,the largest over-expression in brain capillary endothelial cellswas seen for the MRP2 gene (Dombrowski et al., 2001).
Pharmacoresistance
28
Cerebral expression of p-glycoprotein
P-gp is expressed on the apical side of the choroid plexusepithelia, contributing to the blood CSF barrier. At the bloodbrain barrier level, ongoing debate exists as to the precisesubcellular localization (luminal vs. antiluminal) of P-gp in thecerebral capillary endothelial cells. Most of the publishedstudies do suggest that P-gp is principally expressed at theluminal (apical) membrane of the capillary endothelial cells inthe brain. However, some authors have localized P-gp toneighboring astrocytic foot processes on the antiluminal(basolateral) side of brain microvasculature. Expression of P-gpat locations other than the apical membrane of capillaryendothelium suggests that the function of P-gp may extendbeyond the blood brain barrier (Kwan and Brodie, 2005).
Yet, there is some evidence that under pathologicalconditions, such as epilepsy, P-gp in astrocyte foot processesmay be involved in blood brain barrier function. In the rodentbrain, the mdr1a P-gp isoform is predominantly expressed inbrain microvessel endothelial cells of the blood brain barrier,whereas the mdr1b P-gp isoform is preferentially expressed inastrocytes (Decleves et al., 2000).
In the normal human brain, P-gp is highly expressed incapillary endothelial cells, but cannot be detected by routine
Pharmacoresistance
29
immunohistochemistry in brain parenchyma, i.e., astrocytes orneurons. One explanation for the apparent difference inastrocytic P-gp expression could be that P-gp in normal humanastrocytes is below the detection level of the assays used,because under pathological conditions, such as epilepsy, P-gpbecomes detectable in human astrocytes (Sisodiya et al., 2002).
Control of p-glycoprotein expression
A-Genetic controlThe nature of ATP-binding cassette transporter genes
indicates that their expression can be induced in previouslynon-expressive cells. Consequently, over-expression of theseproteins could be observed in blood brain barrier-related cellsand brain parenchyma cells, including neurons from clinicalend experimental studies (Lazarowski et al., 2007).
Tishler et al. (1995) were the first to report that brainexpression of MDR1 gene, which encodes the multidrugtransporter P-gp in humans, is markedly increased in themajority of patients with medically intractable partial (mostlytemporal lobe) epilepsy. In line with enhanced MDR1expression in epileptogenic brain tissue, immunohistochemistryfor P-gp showed increased staining in capillary endotheliumand astrocytes.
Pharmacoresistance
30
By using gene arrays to study mRNAs of multidrugtransporters in endothelial cells isolated from surgicallyresected epileptic foci of patients with pharmacoresistant partialepilepsy, Janigro and colleagues’ determined increasedexpression of MDR1 and the gene encoding MRP2, indicatingthat over-expression of both P-gp and MRP2 in the blood brainbarrier may be involved in resistance to anti-epileptic drugs(Dombrowski et al., 2001).
MDR1 is highly polymorphic. Sequence analysescovering the entire gene in 24 healthy white volunteersidentified 15 mutations, with a relatively commonpolymorphism, C3435T in exon 26. MDR1 3435 TT genotypewas associated with lower P-gp expression in enterocytes thanthose with CT or CC genotypes. Subjects with CC genotypewere found to have higher P-gp function than those with TTgenotype (Hoffmeyer et al., 2000).
Considerable ethnic variation is found in the frequency ofthe MDR1 C3435T genotype, such that 65% to 83% of Africanblacks express the TT genotype, but only around 25% of whitepeople do so, with the frequencies among Asians lyingsomewhere in between (Ameyaw et al., 2001).
B-Environmental controlExpression of P-gp is highly inducible by environmental
factors. MDR1 expression has been found to be induced by heat
Pharmacoresistance
31
shock, arsenite, partial hepatectomy, growth factors, sodiumbutyrate, protein kinase C agonists, and even its substrates andinhibitors (Kwan and Brodie, 2005).
Mechanisms of over-expression
An important question is whether the over-expression ofefflux transporters in epileptic brain tissue is constitutive oracquired/induced, or both mechanisms may be at play. Aconstitutive over-expression could occur as a result of geneticpredisposition, or it could be intrinsic to the development of thespecific pathology. It also is conceivable that over-expression isacquired, such as induction by recurrent seizures or even theanti-epileptic drugs intended to prevent them. Currentpreliminary evidence suggests that both situations could betaking place (Kwan and Brodie, 2005).
Acquired (induced) over-expressionAn open question is whether the over-expression of P-gp
and MRPs in epileptogenic brain tissue of patients withpharmacoresistant epilepsy is a consequence of epilepsy,uncontrolled seizures, and chronic treatment with anti-epilepticdrugs, or combinations of these factors. Becausepharmacoresistant patients have the same extent of neurotoxicside effects under anti-epileptic drug treatment as patients whoare controlled by anti-epileptic drugs, the over-expression of
Pharmacoresistance
32
drug transporters in pharmacoresistant patients is most likelyrestricted to the epileptic focus or circuit (Loscher andPotschka, 2002).
This is substantiated by the finding that over-expressionof P-gp and MRP1 was found in epileptogenic tissue but notadjacent normal tissue. In this respect, it is also interesting tonote that in patients in whom the epileptic focus has beenresected during epilepsy surgery -resulting in seizure controlunder treatment with anti-epileptic drugs - seizures may recurafter cessation of anti-epileptic drug treatment and becomepharmacoresistant again, suggesting that a “secondary focus”has become activated and drug-resistant (Sisodiya et al., 2002).
For ethical and methodologic reasons, it is difficult toconduct longitudinal studies to examine the possibility ofinduction of transporters by epileptogenesis, seizures, or anti-epileptic drug treatment in humans. Most data in this regardhave come from animal studies (Kwan and Brodie, 2005).
Using monoclonal antibody staining, there was observedregional-specific over-expression of P-gp in the hilus of thedentate gyrus and CA3 region of the hippocampus in rats 1week after pilocarpine-induced status epilepticus. Mostimportantly, there was no comparable staining of hippocampal
Pharmacoresistance
33
neurons in non-seizing controls, indicating that the upregulationof P-gp was a consequence of the experimentally inducedseizure activity (Loscher, 2007).
Results observed in some experiments indicated thatMDR1 over-expression depends on the seizure-stressfrequency. Also, this expression exhibits a selective sequentialpattern in term of the type of cells affected: as the frequency ofthe induced-seizures increased, more cells, that is, endothelialcells, astrocytes, and surrounding neurons, became MDR-1positive (Lazarowski et al., 2007).
In rats, kainate-induced seizures have been found totransiently over-express P-gp in astroglia and, less marked,capillary endothelial cells in the hippocampus, indicating thatseizures rather than epilepsy are responsible for over-expressionof drug transporters. This could explain that one of the majorpredictors of pharmacoresistance is high seizure frequency priorto initiation of treatment (Zhang et al., 1999).
Constitutive over-expressionOver-expression of efflux transporters may be
constitutive or intrinsic to the lesion itself. This is supported bythe observation that in the case of malformations of corticaldevelopment, P-gp was upregulated in postmortem archival
Pharmacoresistance
34
tissues from patients who had died before experiencing seizuresor exposure to anti-epileptic drugs (Kwan and Brodie, 2005).
Furthermore, when determining P-gp and MRP1expression in three common causes of refractory epilepsy,namely dysembryoplastic neuroepithelial tumors, focal corticaldysplasia, and hippocampal sclerosis, and comparing theexpression in the abnormal epileptogenic tissue with P-gp andMRP1 expression in histologically normal adjacent tissue, therewas over-expression of both P-gp and MRP1 in reactiveastrocytes in the epileptogenic tissue in all three conditions, andMRP1 over-expression in dysplastic neurons in focal corticaldysplasia (Sisodiya et al., 2002).
Increased expression of P-gp, as well as MRP, has beenassociated with refractory epilepsy in tuberous sclerosis.Increased levels of P-gp were found in astrocytes and bloodvessel endothelium in hippocampal samples obtained frompatients with refractory mesial temporal lobe epilepsycompared with control samples (Robey et al., 2008).
Other mechanismsIn addition to intrinsic or acquired over-expression of
multidrug transporters in the blood brain barrier or blood CSFbarrier of patients with epilepsy, functional polymorphisms of
Pharmacoresistance
35
these transporters may play a role in pharmacoresistance (Kerbet al., 2001).
Furthermore, over-expression and functionalpolymorphisms of multidrug transporters in patients withpharmacoresistant epilepsy need not necessarily be restricted tothe brain, but could also occur in other tissues, such as the smallintestine, where P-gp is thought to form a barrier againstentrance of drugs from the intestinal lumen into thebloodstream, thereby limiting their oral bioavailability (Loscherand Potschka, 2002).
In view of data indicating that the endothelial barrierfunction of the blood brain barrier is transiently disruptedduring seizures, over-expression of multidrug transporters inglial endfeet covering the blood vessels may represent a“second barrier” under these conditions (Loscher andPotschka, 2002).
Over-expressed multidrug transporters lower theextracellular concentration of anti-epileptic drugs in thevicinity of the epileptogenic pathology and thereby render theepilepsy caused by these pathologies resistant to anti-epilepticdrug treatment (Sisodiya et al., 2002).
P-gp over-expression can be induced locally bymicroinjection of glutamate. Intracerebral microinjections of
Pharmacoresistance
36
glutamate at nanomolar levels were sufficient to locallyincrease P-gp expression without seizure activity. P-gpexpression in more distal regions of the brain was notsignificantly affected. These particular data suggest thatmolecular factors inducing P-gp over-expression in the bloodbrain barrier, even in the absence of seizures, couldprecondition the refractory phenotype for some epilepticsyndromes (Bauer et al., 2008).
Are antiepileptic drugs substrates of drugtransporters?
According to the drug transporter hypothesis, theexpression of drug efflux transporters in the brain capillaries ofnormal brain should not restrict the penetration of anti-epilepticdrugs to any significant extent. However, in epileptic brain,where multidrug transporters are over-expressed in braincapillaries as well as in the astrocytic foot processessurrounding these capillaries, brain penetration of anti-epilepticdrugs should be reduced in a regional-specific fashion.Moreover, over-expression of drug efflux transporters inneurons and glia of the brain parenchyma would further limitthe effectiveness of anti-epileptic drugs by restricting access tointracellular target sites. To verify this hypothesis, it first needsto be determined if anti-epileptic drugs are substrates for
Pharmacoresistance
37
multidrug transporters in the blood brain barrier. Recent datafrom multiple laboratories support this claim (Loscher, 2007).
Indeed, there is increasing evidence that various majoranti-epileptic drugs are substrates for one or more of theseefflux carriers. At least three strategies are used in this respect.One is to evaluate whether the brain penetration of anti-epileptic drugs can be affected by P-gp or MRP inhibitors; asecond is to use cell lines that over-express P-gp or MRPs; anda third is to study drug penetration into the brain of mdr or mrpknockout mice (Potschka et al., 2001).
The only P-gp inhibitors that have been clinicallyevaluated in combination with anti-epileptic drugs in patientswith epilepsy are calcium channel blockers such as verapamil,nifedipine, or diltiazem. Verapamil and diltiazem increased theplasma concentrations of carbamazepine (probably byinhibiting its CYP3A4-mediated metabolism) and causedunacceptable neurotoxicity, but encouraging clinicalobservations were reported for add-on treatment with nifedipinein patients with refractory partial seizures (Loscher andSchmidt, 1994).
However, because calcium channel antagonists exertanticonvulsant activity of their own and inhibit CYP3A4, it isnot possible to judge whether the favorable effect of
Pharmacoresistance
38
combinations of nifedipine and anti-epileptic drugs was due toinhibition of P-gp, inhibition of CYP3A4, or blockade ofcalcium channels (Loscher and Potschka, 2002).
Using microdialysis in freely moving rats, it was foundthat brain penetration of systemically administered phenytoinnearly doubled 0.5- 1.5 hours after a unilateral intracerebralinjection of the P-gp transport inhibitor, verapamil, incomparison with the contralateral, noninjected hemisphere(Potschka and Loscher, 2001).
Similarly, phenytoin brain concentrations aresignificantly increased in mdr1 knockout mice that lack thegene encoding for P-gp in comparison with wild-type controls.Conversely, over-expression of P-gp, produced as a result ofkainate-induced status epilepticus, significantly decreases theblood brain barrier permeability of phenytoin in thehippocampus in comparison with non-epileptic control rats(Rizzi et al., 2002).
Taken together, these results demonstrate that themultidrug transporter, P-gp, regulates phenytoin entry into thebrain.
Additional data have shown that a number of other anti-epileptic drugs, including carbamazepine, mild malformation ofcortical development, lamotrigine, gabapentin, and topiramate,are also substrates for P-gp, MRPs, or both. Hence, it appears
Pharmacoresistance
39
that an over-expression of P-gp could potentially affect theefficacy of a variety of different anti-epileptic drugs (Loscherand Potschka, 2005).
Absence of MRP2 in the blood brain barrier led toincreased penetration of phenytoin into the brain andsignificantly enhanced anticonvulsant activity compared withrats with intact MRP2 function. Similar results were obtainedwhen phenytoin was combined with probenecid to inhibitMRP2. Significant increase of drug penetration into the brainby probenecid has previously been reported for the major anti-epileptic drugs valproate and carbamazepine and has beenattributed to inhibition of MRP2 in the blood brain barrier(Loscher and Potschka, 2002).
In contrast to phenytoin, phenobarbital’s braindistribution or anticonvulsant activity were not affected byprobenecid or lack of MRP2 in the blood brain barrier,indicating that not all anti-epileptic drugs are substrates for thistransporter. However, Phenobarbital is a substrate for P-gp sothat both MRP2 and P-gp act in concert to restrict the brainpenetration of anti-epileptic drugs (Dombrowski et al., 2001).
More recent animal studies revealed that the bidirectionalmovement of valproate across the blood brain barrier (andpossibly also blood CSF barrier) is mediated jointly by passive
Pharmacoresistance
40
diffusion and carrier-mediated transport. The uptake ofvalproate from blood to brain is facilitated by a medium-chainfatty acid transporter, which accounts for two-thirds of thebarrier permeability, whereas the mechanisms governing theefflux of valproate from the brain involve a probenecid-sensitive, active transport system at the brain capillaryendothelium (Shen, 1999).
Valproate was the first anti-epileptic drug for whichblood CSF barrier and blood brain barrier transport by aprobenecid sensitive carrier, most likely MRP, has beenreported (Potschka et al., 2001).
Valproate is a substrate for MRPs in brain capillaryendothelial cells, which raises the possibility that MRPs mayserve as the efflux transporters of valproate and explains thepreviously described effects of probenecid on brain and CSFlevels of valproate, because probenecid is an inhibitor of MRP1and MRP2 (Borst et al., 2000).
In mdr1 knockout mice, the brain/plasma concentrationratio for carbamazepine was found to be significantly higher inknockout mice than in wild-type controls. Furthermore,significant increases in brain levels were found for topiramate,lamotrigine, and gabapentin in mdr1 knockout mice, althoughno significant differences to controls were seen for
Pharmacoresistance
41
phenobarbital, phenytoin, valproate, and vigabatrin (Sills andKwan, 2001).
However, use of knockout mice is limited in the study ofdrug resistance because of the redundancy of the transporters:another transport protein may take over the function of one thathas been knocked out (Schinkel, 1999).
With respect to the use of cell lines to study anti-epilepticdrug transport, it was found that intracellular phenytoin levelsin a MDR1-expressing neuroectodermal cell line were only one-fourth that in MDR1- negative cells, suggesting that P-gpsignificantly contributes to cell export of phenytoin (Tishler etal., 1995).
In human colon carcinoma cells, phenobarbital and, to amuch lesser extent, phenytoin were found to up-regulate P-gp, aphenomenon described for several substrates of P-gp (Schuetzet al., 1996).
Even though findings appear to support a role formultidrug transporters in pharmacoresistant epilepsy, there are anumber of conceptual questions that remain enigmatic. Firstly,epileptic seizures are known to result in a disruption of theblood brain barrier, which would be expected to result in betteraccess of anti-epileptic drugs to brain parenchyma despite theupregulation of multidrug transporters. Secondly, patients are in
Pharmacoresistance
42
many cases treated with anti-epileptic drugs until CNS sideeffects develop. This seems to indicate that relevant CNSconcentrations of anti-epileptic drugs are reached despitetransporter upregulation, yet, these patients are resistant totreatment. This apparent discrepancy could potentially arise vialocal upregulation of drug transporters that only affects anti-epileptic drug concentrations at the epileptic focus (Remy andBeck, 2006).
Target hypothesisIt is widely accepted that the efficacy of an anti-epileptic
drug is determined by its ability to cross blood brain barrier andbind to intraparenchymal target sites. According to the targethypothesis, pharmacoresistance occurs when target sites arestructurally and/or functionally modified in such a way thatthey become less sensitive to anti-epileptic drugs (Beck, 2007).
This means that molecular drug targets may undergo agenetic or functional modification after which they are nolonger sensitive to their ligands (Remy et al., 2003a).
In general terms, a drug target is defined as one thatproduces a clinically relevant response to a therapeutic dose ofa drug, in this case an anti-epileptic drug. A particularly well-investigated drug target of many first-line anticonvulsants is thevoltage-gated Na+ channel. A large number of targets for anti-
Pharmacoresistance
43
epileptic drugs have been identified in the brain, many ofwhich undergo molecular changes during chronic epilepsy. Sofar, a reduced sensitivity of drug targets to anti-epileptic drugsin chronic human epilepsy has been suggested for the voltage-gated Na+ channel and the GABAA receptor (Beck, 2007).
Based on the specific targets involved in anti-epilepticdrug mechanisms, anti-epileptic drugs can be dividedmechanistically into drugs acting by (a) modulation of voltage-gated ion channels (including sodium, calcium, and potassiumchannels); (b) enhancement of synaptic inhibition [e.g., bypotentiating inhibition mediated by γ -aminobutyric acid(GABA)]; and (c) inhibition of synaptic excitation (e.g., byblockade of glutamate receptors). The fact that several anti-epileptic drugs act by more than one of these mechanisms isthought to explain their broad spectrum of clinical efficacy(Schmidt and Loscher, 2005).
Loss of anti-epileptic drug-sensitivity is less pronouncedin CA1 neurons than in dentate granule neurons in experimentalepilepsy. Thus, these results suggest that target mechanisms ofdrug resistance are cell type and anti-epileptic drug specific(Schaub et al., 2007).
Pharmacoresistance
44
Voltage-gated Na+ channelsVoltage gated Na+ channels are composed of one of
several different pore-forming α subunits that form a complexwith additional accessory subunits. They are closed at restingmembrane potential but open rapidly upon depolarization. Theresultant Na+ inward currents subsequently decrease rapidlytoward baseline levels as the Na+ channels undergo inactivationduring prolonged depolarization. After inactivation, Na+
channels require hyperpolarization to return to the resting state.The transition between these functional states is rapid andoccurs on a time scale of milliseconds, enabling Na+ channels tosustain fast action potentials (Remy et al., 2003a).
The target hypothesis is based primarily on studies withcarbamazepine on voltage-gated sodium channels inhippocampal neurons. The primary mechanism of this majoranti-epileptic drug is well established and thought to be relatedto its action on voltage-gated Na+ channels (Remy et al.,2003b).
Carbamazepine is known to inhibit voltage-dependentNa+ currents via two classes of mechanisms. It modestly blocksNa+ channels in their resting state at hyperpolarized membranepotentials, but the blocking effects are enhanced when theresting membrane potential is depolarized. In addition tovoltage dependent block, carbamazepine inhibits Na+ currentsin an activity- or use-dependent manner; that is, blocking
Pharmacoresistance
45
effects are more pronounced when the cell membrane isrepetitively depolarized at high frequencies. Use- and voltage-dependent block of Na+ channels by carbamazepine generally isassumed to result from preferential binding of these drugs to theinactivated state of the channel (Remy et al., 2003a).
In epileptic brain tissue, use dependent block of Na+
channel activity is absent. Similar findings have been observedin human brain tissue obtained from carbamazepine -responsiveand carbamazepine -resistant patients. In carbamazepine -responsive patients, use-dependent blockade of Na+ channelactivity is observed. In contrast, use-dependent effects areabsent in carbamazepine-resistant patients. Taken together,these findings suggest that a clinical loss of carbamazepineefficacy in epilepsy patients is associated with a reduction in thedrug sensitivity of Na+ channels in the brain (Beck, 2007).
In CA1 neurons, the effects of carbamazepine on thesteady-state inactivation properties of inward Na+ current weretransiently reduced in the kindling model of epilepsy. Incontrast to these comparatively modest and transient effects, acomplete and long-lasting loss of use-dependent blockingeffects of carbamazepine was found in the pilocarpine model ofepilepsy in hippocampal dentate granule cells, as well as inepilepsy patients with carbamazepine-resistant temporal lobeepilepsy (Remy et al., 2003a).
Pharmacoresistance
46
This dramatic loss of a major mechanism of action ofcarbamazepine did not extend to other anti-epileptic drugsknown to affect inward Na+ current. Following pilocarpineinduced status epilepticus; the use-dependent effects ofphenytoin were reduced, but not completely lost, while theeffects of lamotrigine were completely unchanged. It is atpresent unclear why use-dependent block by carbamazepineand phenytoin is lost or reduced, whereas use dependent blockby lamotrigine remains intact in experimental epilepsy. This isan intriguing question because it has been suggested that allthree drugs bind to the same site on Na+ channels (Remy andBeck, 2006).
What mechanisms can account for an altered sensitivityof Na+ channels in epileptic tissue? One possibility may be thatthe subunit composition of these channels is altered, such thatthe expression of anti-epileptic drug -insensitive subunits orsubunit combinations is promoted. Indeed, numerous changesin Na+ channel subunit expression have been observed in bothhuman and experimental epilepsy. In this respect, thedownregulation of accessory Na+ channel β1 and β2 subunitsfollowing experimentally induced status epilepticus appears tobe a consistent finding (Ellerkmann et al., 2003).
Pharmacoresistance
47
Other types of voltage-gated channels
Ca2+ channels can be subdivided into two groups: highthreshold Ca2+ currents and a group of low threshold currents(also termed T-type Ca2+ currents). T-type channels arecritically important in controlling the excitability of thepostsynaptic compartment of neurons, both in normal andepileptic neurons. A number of anti-epileptic drugs has beenshown to inhibit high threshold Ca2+ channels in native neuronsat high therapeutic concentrations. Some anti-epileptic drugspotently inhibit low threshold T-type Ca2+ channels, which arenot expressed presynaptically. The effects of anti-epilepticdrugs on the T-type Ca2+ channel subunits, as well as in nativeneurons, are diverse (Remy and Beck, 2006).
Neurotransmitter systems: GABA
GABA is the predominant inhibitory neurotransmitter inthe adult brain and plays a critical role in the regulation ofexcitability of neuronal networks (Mody and Pearce, 2004).
GABA binding to GABAA receptors causes opening ofthe receptor which is permeable to Cl_ and to a lesser extent toHCO3. In the presence of a normal adult transmembraneous Cl_
gradient, this results in expression of an inhibitory post-synapticcurrent that hyperpolarizes the post-synaptic neuronal
Pharmacoresistance
48
membrane. Direct modulators of GABAA receptors includebenzodiazepines and barbiturates. Specifically, GABAreceptors on isolated hippocampal neurons from epileptic brainwere less responsive to zolpidem, a benzodiazepine site agonist,in comparison to nonepileptic controls. These results suggestthat alterations in GABAA receptor activity may be anothertarget mechanism of pharmacoresistance to certainanticonvulsant drugs (Remy and Beck, 2006).
GABAA receptors mediate the majority of fast inhibitoryneurotransmission in the brain. GABAA receptors can beassembled from seven distinct subunit families defined bysequence similarity: alpha, beta, gamma, delta, pi, theta, andrho. Most GABAA receptor subtypes in the brain are believed tobe composed of alpha, beta, and gamma subunits. The role ofthe other subunits, which have a very limited expression patternin the brain, remains to be determined (Schmidt and Loscher,2005).
The major GABAA receptor subtype (60% of all GABAA
receptors) is assembled from the subunits α1β2γ2, with only afew brain regions lacking this receptor. This receptor subtypemediates to a large extent the anticonvulsant action ofbenzodiazepines, whereas, for instance, α4- or α6-containingsubunit assemblies are insensitive to benzodiazepines and other
Pharmacoresistance
49
benzodiazepine site agonists such as zolpidem. Thus anychange in the subunit composition of GABAA receptors canhave dramatic consequences for the anticonvulsant efficacy ofbenzodiazepines and possibly other anti-epileptic drugs that actvia the GABAA receptor (Schmidt and Loscher, 2005).
In normal dentate granule cells, GABAA receptors areinsensitive to zinc, which is released from mossy fibers andfunctions as a negative allosteric modulator of GABAA
receptors. This zinc insensitivity of normal GABAA receptors isa result of high levels of expression of the α1 subunit in thesecells. In epileptic rats, expression of the α1 subunit decreases,and expression of α4 and δ subunits increases, leading to anassembly of GABAA receptors that are strikingly zinc sensitive.In addition to the enhanced zinc sensitivity, GABAA receptorsfrom the epileptic hippocampus lose their sensitivity toaugmentation by zolpidem (Coulter, 2000).
Regarding GABAA receptor agonists, reduced activity ofsuch substances has been described in a chronic model ofepilepsy. In the pilocarpine model of epilepsy, GABAA
receptors of dentate granule cells show a reduced sensitivity todrugs acting on the benzodiazepine receptor site 1 (Cohen etal., 2003).
Pharmacoresistance
50
Combined molecular and functional studies indicate thata transcriptionally mediated switch in the alpha subunitcomposition of GABAA receptors occurs in epileptic animals, inparticular a decrease of α1 subunits and an increase of α4subunits. These findings correlate well with the observedchanges in benzodiazepine receptor pharmacology (Remy andBeck, 2006).
The specific changes in drug targets described above arean attractive concept to explain pharmacoresistance. It isimportant to realize, however, that not only changes in drugtargets themselves, but also changes in other molecules thataffect their function may have important consequences for anti-epileptic drug efficacy. This idea is exemplified by recentfindings regarding the role of GABA in epilepsy. GABA mayon occasion act as an excitatory neurotransmitter in theimmature brain (Remy and Beck, 2006).
A depolarizing action of GABAA receptor activationarises because of an altered chloride homeostasis, resulting in achanged chloride gradient across the neuronal membrane. Thealtered chloride reversal potential then results in a net outwardflux of Cl_ through the GABAA receptor ionophore, causingdepolarization of the neuron (Mody and Pearce, 2004).
Pharmacoresistance
51
Interestingly, in addition to the developing brain,depolarizing GABA responses appear to be a feature of someneurons in the epileptic brain (Cohen et al., 2002).
Augmenting such depolarizing GABA-mediatedpotentials by application of GABA agonists is likely tofacilitate action potential generation to excitatory input andthereby would increase neuronal excitability instead ofdecreasing it (Gulledge and Stuart, 2003).
Whether depolarizing GABA responses really play a rolein pharmacoresistance to GABA mimetic drugs remains to beseen. These considerations do, however, illustrate the need toconsider changes in drug targets within the more general settingof a chronically epileptic brain (Remy and Beck, 2006).
Genetic control of anti-epileptic drug targets
It has become clear over the recent years that seizurescause coordinated and cell-specific transcriptional changes thatresult in either an up- or down-regulation of families of ionchannel mRNAs. This, in turn, causes alterations in the subunitcomposition or density of ion channels, resulting in alteredintrinsic and synaptic neuronal properties (Beck, 2007).
Pharmacoresistance
52
It is important to note that gene polymorphisms relevantfor pharmacoresistance may occur both in promoter regions aswell as in introns and exons. Gene polymorphisms within thecoding regions of such genes would result in a difference in ionchannel or transporter proteins that precedes the onset ofepilepsy. Polymorphisms in promoter regions, which affect thetranscription of such genes, may affect activity-dependenttranscriptional regulation of these genes by seizures. Thisprovides a potential mechanism for the acquisition of apharmacoresistant phenotype during epileptogenesis inpharmacoresistant—as opposed to Pharmacoresponsive-patients(Remy and Beck, 2006).
Firstly, genetic polymorphisms in ion channel drugtargets have been identified that are associated with clinicaldrug response. For instance, a polymorphism in the SCN1Agene seems to correlate with the maximal doses of the anti-epileptic drugs carbamazepine and phenytoin used clinically inindividual patients (Tate et al., 2005).
In addition, mutations in the gene encoding the accessorybeta subunit of voltage-gated Na+ channels are associated withgeneralized epilepsy with febrile seizures and incorporation ofmutated accessory subunits into the channel complex appears togive rise to channels that are less phenytoin sensitive (Schaubet al., 2007).
Pharmacoresistance
53
Are transcriptional regulatory processes the only onesimportant in seizure-induced plasticity and altered targetsensitivity? This appears unlikely, primarily because ofexperiments performed by Heinemann and colleagues overseveral years. These researchers have demonstrated in vitro thatseizure activity induced by lowering extracellular magnesiumlevels or blocking potassium channels evolves from an anti-epileptic drug responsive to an anti-epileptic drug resistant formwithin 30–120 min. This transition is likely too fast to bemediated by a transcriptional change in ion channel expression.Alternatively, ion channel subunits may be modified by redoxmodulation or phosphorylation, a set of mechanisms that maybe invoked much more rapidly than transcriptional changes(Beck, 2007).
Gross Pathology
54
GROSS PATHOLOGY
Malformations of cortical development Focal cortical dysplasia Periventricular nodular heterotopias Tuberous sclerosis complex Glioneuronal tumors
Do seizures start within the lesion or theperilesional region?
Is epileptogenesis primarily a result ofcircuit abnormalities or cellular/moleculardefects?
Expression of drug transporters inpathological lesions associated withrefractory epilepsy
Gross Pathology
55
Malformations of cortical developmentDuring normal cortical development, immature cortical
neurons and glia are primarily generated in the germinalventricular zone (proliferative stage), although someGABAergic cortical interneurons have also recently beenshown to originate from the subcortical ganglionic eminences.While the different stages of cortical development overlaptemporally, neurons generated during the proliferative stageproceed to migrate either radially or tangentially to their finallocation in the cortex (neuronal migration stage) and thendevelop mature dendrites and axons and form synapticconnections (cortical organization stage) (Wong, 2008).
Malformations of cortical development refer tomalformations arising from various aetiologies and presentingwith diverse characteristics. In the clinical practice, theclassification system of Kuzniecky and Barkovich has gainedwide acceptance. According to this classification,malformations of cortical development fall into the followingcategories: (i) malformations caused by abnormal neuronal andglial proliferation (e.g. hemimegalencephaly, focal corticaldysplasia), (ii) malformations caused by abnormal neuronalmigration (e.g. heterotopia, lissencephaly) and (iii)malformations caused by abnormal cortical organization (e.g.polymicrogyria, focal cortical dysplasia without balloon cells)(Kuzniecky and Barkovich, 2001).
Gross Pathology
56
Malformations of cortical development are increasinglyrecognized as causes of epilepsy. The clinical significance andimpact of malformations of cortical development are especiallyhigh, because malformations of cortical development arefrequently associated with pharmacoresistant epilepsy that isrefractory to available seizure medications (Wong, 2008).
Malformations of cortical development are oftenassociated with severe epilepsy and developmental delay.Hemimegalencephaly, an enlarged dysplastic hemisphere, canpresent as early onset severe epileptic encephalopathy or aspartial epilepsy. In focal cortical dysplasia, MRI shows focalcortical thickening and simplified gyration. Tuberous sclerosis(TS) is a multisystemic disorder primarily involving thenervous system; 60% of patients having epilepsy, with 50%having infantile spasms. Bilateral periventricular nodularheterotopia consists of confluent and symmetric nodules of greymatter along the lateral ventricles. X-linked bilateralperiventricular nodular heterotopia presents with epilepsy infemales and prenatal lethality in most males. Schizencephaly(cleft brain) has a wide anatomo-clinical spectrum, includingpartial epilepsy in most patients. Polymicrogyria (excessivenumber of small and prominent convolutions) has a widespectrum of clinical manifestations ranging from early onsetepileptic encephalopathy to selective impairment of cognitivefunctions (Guerrini et al., 2003).
Gross Pathology
57
The histopathological spectrum of malformations ofcortical development is large ranging from prominent to onlyminute changes. Whereas hemimegalencephaly,polymicrogyria, nodular or band heterotopias can be reliablydiagnosed in vivo, focal cortical dysplasias often escapeimaging techniques (MRI) and may considerably vary in theirsize and localization (Blumcke et al., 2009).
About 40% of children with drug resistant epilepsyharbour a cortical malformation and up to 50% of the pediatricepilepsy surgery operations are carried out in children withmalformations of cortical development (Guerrini et al., 2003).
Gross Pathology
58
Table (3) Classification of cortical malformations (Barkovich et al., 2001)I. Malformations due to abnormal neuronal and glial proliferation or apoptosisA. Decreased proliferation/increased apoptosis: microcephalies1. Microcephaly with normal to thin cortex2. Microlissencephaly (extreme microcephaly with thick cortex)3. Microcephaly with polymicrogyria/cortical dysplasiaB. Increased proliferation/decreased apoptosis (normal cell types): megalencephaliesC. Abnormal proliferation (abnormal cell types)1. Non-neoplastica. Cortical hamartomas of tuberous sclerosisb. Cortical dysplasia with balloon cellsc. Hemimegalencephaly2. Neoplastic (associated with disordered cortex)a. DNT (dysembryoplastic neuroepithelial tumor)b. Gangliogliomac. GangliocytomaII. Malformations due to abnormal neuronal migrationA. Lissencephaly/subcortical band heterotopia spectrumB. Cobblestone complex1. Congenital muscular dystrophy syndromes2. Syndromes with no involvement of muscleC. Heterotopia1. Subependymal (periventricular)2. Subcortical (other than band heterotopia)3. Marginal glioneuronalIII. Malformations due to abnormal cortical organization(including late neuronal migration)A. Polymicrogyria and schizencephaly1. Bilateral polymicrogyria syndromes2. Schizencephaly (polymicrogyria with clefts)3. Polymicrogyria with other brain malformations or abnormalities4. Polymicrogyria or schizencephaly as part of multiplecongenital anomaly/mental retardation syndromesB. Cortical dysplasia without balloon cellsC. MicrodysgenesisIV. Malformations of cortical development, not otherwise classifiedA. Malformations secondary to inborn errors of metabolism1. Mitochondrial and pyruvate metabolic disorders2. Peroxisomal disordersB. Other unclassified malformations1. Sublobar dysplasia2. Others
Gross Pathology
59
Focal cortical dysplasiaFocal cortical dysplasia is the single most important
cause of focal intractable epilepsy in childhood. It ishistopathologically proved in at least 20% of patients in adultepilepsy surgery series and in almost 50% of childrenundergoing surgical therapy for epilepsy (Luders and Schuele,2006).
In focal cortical dysplasia, histological abnormalities arerestricted to one lobe or to a segment of a few centimeters.Extensive examination of brains with focal lesions may,however, show widespread minor dysplastic changes (Guerriniet al., 2003).
The true prevalence of focal cortical dysplasia in patientswith epilepsy is unknown; recent epidemiological studiesidentified malformations of cortical development in up to 25%of all children with symptomatic epilepsy (Fujiwara andShigematsu, 2004).
Focal cortical dysplasia as the cause of focal epilepsy canbe roughly estimated at 5–10% in developed countries.However, the prevalence of focal cortical dysplasia in focalepilepsy may in fact be lower than 5%, when referring toepilepsy patients on a global basis. In less-developed countries,
Gross Pathology
60
infectious diseases such as neurocysticercosis may play a moreprominent role in epilepsy aetiology (Bast et al., 2006).
Several classification schemes of cortical malformationshave been proposed based on imaging characteristics, genetics,and neuropathology. The histopathological classificationsystems proposed by Palmini and Luders and Tassi et al. arethe two prime systems that are currently used. Palmini andLuders proposed the most frequently used histologicalclassification. It divides focal cortical dysplasia into three majorsubtypes: mild malformation of cortical development, focalcortical dysplasia type I, and focal cortical dysplasia type II.Two further subcategories are recognized within each of thetypes: mild malformation of cortical development type I, mildmalformation of cortical development type II, focal corticaldysplasia type Ia, focal cortical dysplasia type Ib, focal corticaldysplasia type IIa, and focal cortical dysplasia type IIb (Krseket al., 2008).
Gross Pathology
61
Table (4) Histopathological classification of focal corticaldysplasia (Bast et al., 2006).Palmini and Luders, 2004 Tassi et al., 2002
Architectural dysplasia:Abnormal cortical laminationand ectopic neurones in whitematter.
Cytoarchitectural dysplasia:Giant neurofilament-enrichedneurones and altered corticallamination.
Taylor-type dysplasia:Giant dysmorphic neurones andballoon cells with laminardisruption.
Gross Pathology
62
Focal cortical dysplasia exhibits a variety of features thatin MRI studies are not always all present in combination, norpathognomonic in isolation: (i) local cortical thickening (oftenin combination with cortical hyperintensity), (ii) blurring of thegrey-matter to white-matter surface, (iii) signal changes in theunderlying white matter, usually with an increased signal onT2-weighted images and occasionally with a decreased signalon T1-weighted images (Kuzniecky and Barkovich, 2001).
Fig. (2) MRI of four patients with focal cortical dysplasia. Typicalfeatures are (i) focal cortical thickening (patient 1 and 3), (ii)blurring of the grey matter to white-matter surface (patient 1–3), (iii)signal changes in the underlying white matter, usually with anincreased signal on T2-weighted images (patient 1–4) (Bast et al.,2006).
Gross Pathology
63
Reports from epilepsy surgery programmes may not beused as the gold standard to evaluate MRI sensitivity because ofan obvious bias: if no focal cortical dysplasia is detected inMRI, the patient is less likely to be offered an operative therapythan in the case of a focal lesion. Thus, the varying rates offocal cortical dysplasia detected may be attributed to the diversequality and precision of presurgical MRI investigations (Bast etal., 2006).
Up to 50% of patients with refractory cryptogenicepilepsy (i.e. no MRI lesion) undergoing surgical treatment areshown to have an focal cortical dysplasia: Bautista et al. (2003)reported of a total of 21 patients with normal MRI whounderwent resective surgery for intractable epilepsy inCleveland, OH between 1997 and 2000. In 1 ⁄ 9 temporal and 9 ⁄12 extratemporal lobe resections, histopathological findingsunmasked an focal cortical dysplasia.
Incomplete myelination may result in negative MRIfindings. In infants, where myelination is incomplete, MRI mayfail to reveal an focal cortical dysplasia, because of the lack ofvisual differentiation between white and grey matter (Bast etal., 2006).
Gross Pathology
64
Fig. (3) Effect of myelination on MRI sensitivity. Top: Normal T2-weighted MRI of a 7-month-old child suffering from left temporallobe epilepsy. Bottom: Same patient at 19 months of age. MRI issuspect for focal cortical dysplasia with blurred grey–white mattersurface and prolonged T2-signal (Bast et al., 2006).
Focal cortical dysplasia usually presents with intractablepartial epilepsy, starting at a variable age, but generally beforethe end of adolescence. Since lesions may be located anywherein the brain, any type of focal seizure can be observed and focalstatus epilepticus has been frequently reported. However,infantile spasms may be the first manifestation (Guerrini et al.,2003).
Gross Pathology
65
Epilepsy due to focal cortical dysplasia commonly beginsin the first few years of life and may occur shortly after birth.The histopathological type without balloon cells (type 2A) isrelated to a very early onset compared with focal corticaldysplasia with balloon cells (type 2B). Single cases presentedwith epilepsy onset after the fourth decade in life in lateadulthood (Fauser et al., 2004).
The following epilepsy risk factors were reported in thepersonal history of 55 patients operated for focal corticaldysplasia: positive family history of epilepsy in 18%, febrileseizures in 16%, status epilepticus in 11%, trauma in 16%, CNSinfection in 11% and perinatal complications in 4%. Therefore,one should consider the possibility of an focal corticaldysplasia, even in the presence of other obvious factors ofepileptogenesis (Bautista et al., 2003).
In focal cortical dysplasia patients, who have beenoperated on in early childhood, drawbacks in psychomotordevelopment were observed in up to 70–80% (Francione et al.,2003). It is presumed that the size of lesion, the localization(especially in the case of temporal localization) and even thehistopathological subtype play a major role in the manifestationand grade of developmental delay (Lawson et al., 2005).
Gross Pathology
66
The term “dual pathology” describes the coincidence ofextrahippocampal temporal lesions and Ammon’s hornsclerosis. According to this definition, the majority ofassociated lesions are malformations of cortical development:Salanova et al. investigated 37 patients operated for dualpathology. Heterotopia and cortical dysplasia were the mostcommon findings. In contrast, over one-third of cases of focalcortical dysplasia are associated with a hippocampal sclerosis.Patients with dual pathology showed a tendency for earlierepilepsy onset and longer epilepsy duration compared withpatients presenting with plain focal cortical dysplasia (Bast etal., 2006).
The presence of cortical dysplasia in patients whounderwent temporal lobectomy was first described by Taylor etal. (1971). Since then, an association of hippocampal sclerosiswith macroscopic or microscopic cortical dysplasia in thetemporal lobe has been reported as a quite common pathologywith dysplastic features found in the temporal neocortex in 10-50% of patients with hippocampal sclerosis (Fauser et al.2006).
Patients with hippocampal sclerosis and associatedcortical dysplasia are difficult to distinguish from the patientswith isolated hippocampal sclerosis on the basis of generalclinical features, MRI findings, or ictal clinical semiology. The
Gross Pathology
67
clinical significance of temporal pole MRI abnormalities intemporal lobe epilepsy patients with hippocampal sclerosis isstill unclear (Marusic et al., 2007).
Mild ipsilateral anterior temporal changes can be seen onMRI of a substantial number of patients with hippocampalsclerosis and represent by some authors an abnormal persistentimmature appearance, including an abnormality of myelin ormyelination (Mitchell et al. 2003).
MRI volumetric and PET studies have found groupdifferences between patients with isolated hippocampalsclerosis and hippocampal sclerosis associated with corticaldysplasia. The presence of bilateral temporal lobe atrophy issuggestive of a more widespread (bilateral) temporal lobeinvolvement in patients with hippocampal sclerosis and corticaldysplasia and in patients with isolated hippocampal sclerosis,the most prominent hypometabolism was in the anterior andmesial temporal lobe, whereas in dual pathology, it was in thelateral temporal lobe. However, these group differences maynot distinguish associated cortical dysplasia preoperatively inindividual patient (Diehl et al. 2004).
In presurgical evaluation, the presence of dual pathologymust be taken into consideration because of the commonassociation between focal cortical dysplasia and hippocampal
Gross Pathology
68
sclerosis. There were patients that on the ground of MRIinvestigation were thought to have focal cortical dysplasia andpostoperatively after histopathological evaluation presentedhippocampal sclerosis. Others that were diagnosed with plainhippocampal sclerosis per MRI showed a dual pathology in thehistopathological examination performed postoperatively (Tassiet al., 2002).
Reports on the postoperative outcome of patients withdual pathology are controversial. Early studies reported thatpatients with hippocampal sclerosis and associated microscopiccortical dysplasia have a higher risk for seizure recurrencesafter epilepsy surgery as compared with patients with onlyhippocampal sclerosis. More recent investigations, however,demonstrates that these patients can have a very favorableoutcome provided that both pathologies were removed.Therefore, the distinction between this group of patients andthose with isolated hippocampal sclerosis is important and mayassist in the presurgical diagnosis and improve thepostoperative seizure outcome (Marusic et al., 2007).
In epilepsy caused by focal cortical dysplasia, surgicalresection is an important treatment modality. The postoperativerate of focal cortical dysplasia patients rendered seizure-freevaries from about 50% to approximately 65% in major patientcollectives (Bast et al., 2006).
Gross Pathology
69
Various histopathological subtypes were shown to have adiverse postoperative prognosis: the proportion of patientsrendered seizure-free as a result of surgical treatment wassignificantly lower in focal cortical dysplasia type 2 (especiallytype 2a) compared with milder forms of focal cortical dysplasia(mild cortical dysplasia/ focal cortical dysplasia type 1) (Fauseret al., 2004).
The frequently encountered resistance to treatment ismainly attributed to the following two factors: Intrinsic epileptogenicity: It is presumed, that the affected
tissue in focal cortical dysplasia is itself highly epileptogenic,which is a distinct characteristic when compared with non-dysplastic lesions.
Multi-drug-transporter: An activation of diverse multi-drug-transporter proteins in glial cells and dysplastic neurones canbe shown in the case of focal cortical dysplasia. MDR1 wasproven to be elevated in focal cortical dysplasia tissue in anumber of studies. Furthermore, multidrug- resistanceassociated protein 1 (MRP 1) and the major vault proteinmay act as upregulated drug-transporters in focal corticaldysplasia. (Bast et al., 2006)
Gross Pathology
70
Periventricular nodular heterotopias
Heterotopias are malformations of cortical developmentcharacterized by the presence of apparently normal brain cellsin abnormal positions. Three broad categories are recognized:band heterotopia (double cortex), individual misplaced neuronsin the white matter (neuronal heterotopia) and nodules of greymatter within the white matter (nodular heterotopia) (Tassi etal., 2005). Barkovich and Kuzniecky, (2000) did not includeneuronal heterotopia among malformations of corticaldevelopment due to abnormal migration; however, in the recentclassification (Barkovich et al., 2001), conditions of abundantneurons in the white matter are again considered as being due toabnormal neuronal migration.
Nodular heterotopia are further divided into:subependymal heterotopia (subsuming periventricular nodularheterotopia), which appears on MRI as nodular subependymalmasses having the same signal intensity as cortical grey matter;and subcortical heterotopia, which appear as irregular clustersof nodules of grey matter within the white matter (Barkovich etal., 2001).
Periventricular nodular heterotopias are among the mostcommon malformations of cortical development and affectedpatients are frequently characterized by focal drug-resistant
Gross Pathology
71
epilepsy. Periventricular nodular heterotopias are made up ofround nodular masses of normal neurons and glial cells with nolaminar organization, located close to the periventriculargerminal matrix, and hence called periventricular orsubependymal nodular heterotopias. For this particular locationwithin the brain and the normal features of the heterotopic cells,it has been considered the result of a primary failure of neuronalmigration (Battaglia et al., 2006).
Periventricular nodular heterotopias may present asmalformations attributable to a generalised abnormal corticaldevelopment or in focal or multifocal abnormalities.Generalised periventricular nodular heterotopias consist ofbilateral contiguous nodules creating an irregular bumpysurface lining the ventricular wall. Focal or multifocalperiventricular nodular heterotopias are considered a localisedabnormality with multiple but not contiguous nodules. Bothgeneralised and localized nodular heterotopia are attributable toabnormal neuronal migration and may be isolated or associatedwith other cortical and brain malformations (D’Orsi et al.,2004).
Bilateral and symmetrical periventricular nodularheterotopias occur mostly in female subjects; it may be familialand causally related to point mutations of the FLN1 gene. In
Gross Pathology
72
addition to these cases, familial patients with bilateralperiventricular nodular heterotopia not related to FLN1mutations, and sporadic female patients with bilateral butclearly asymmetrical periventricular nodular heterotopias havebeen reported. Unilateral periventricular nodular heterotopiasare frequently located in the posterior paratrigonal region of thelateral ventricles and may extend into the white matter toinvolve adjacent neocortical and archicortical areas (Battagliaet al., 2006).
Large periventricular nodular heterotopias extendingfrom the subependymal region to involve overlying malformedcortical areas in different lobes have been termed subcorticalheterotopias, but it is not yet clear whether periventricular andsubcortical heterotopia are separate entities or differentextensions of the same brain dysgenesis (Barkovich, 2000).
Periventricular nodular heterotopias probably result froman arrest in the migrational progress of neuroblasts from theperiventricular layer to the cortex, which usually occursmaximally between the 7th and 16th gestational weeks, alongradial glial fibres or is due to a failure of programmed cell deathof groups of neuroblasts within the periventricular germinalmatrix (Aghakhani et al., 2005).
Gross Pathology
73
When there is a primary, limited, and pure deficit inneuronal migration onset, the remaining neuroblasts maymigrate normally to form the regular six layered cortex andperiventricular nodular heterotopias only appear. When theongoing process of migration or of the later stage of neuronalmigration and cortical organisation are also impaired,subcortical heterotopia or polymicrogyria and schizencephalymay develop with periventricular nodular heterotopias (D’Orsiet al., 2004).
Subependymal or periventricular heterotopia is the mostcommonly identified type of heterotopia in clinical practice.The prevalence of periventricular nodular heterotopias inpatients with epilepsy is unknown. The associated epilepsysyndrome is variable and seizures may be generalized or focal,often suggesting mesial or neocortical temporal and parieto-occipital onset. Periventricular nodular heterotopias may eitherbe the epileptogenic source or part of a more widespreadepileptogenic network involving the hippocampus, and theoverlying or distant neocortex (Aghakhani et al., 2005).
Periventricular nodular heterotopias are considered to beassociated with developmental delay and epilepsy, but a widevariety and heterogeneity of clinical pictures are often present.Epilepsy can begin in the second or third decade of life or
Gross Pathology
74
earlier, with seizures ranging from rare to very frequent, oftenresistant to polytherapy. Most patients present with partialseizures, exceptionally with status epilepticus. Mentalretardation, usually absent or mild, can also be severe andassociated with neurological deficits and dysmorphic features(D’Orsi et al., 2004).
Tuberous sclerosis complex
Tuberous sclerosis or tuberous sclerosis complex is amultisystemic disorder involving primarily the central nervoussystem, the skin, and the kidney. A prevalence of 1:30 000 – 50000 has been reported. In the brain, the characteristic featuresare cortical tubers, subependymal nodules and giant cell tumors.Cortical tubers are more directly related to epileptogenesis.They are identified by their nodular appearance, firm texture,and variability in site, number and size. Microscopically, thetubers consist of subpial glial proliferation with orientation ofthe glial processes perpendicular to the pial surface, and anirregular neuronal lamination with giant multinucleated cellsthat are not clearly neuronal or astrocytic. The junction betweengray and white matter is indistinct and may be partlydemyelinated. These pathological changes are similar to thoseseen in focal cortical dysplasia. Cortical tubers are usually wellvisualized by MRI as enlarged gyri with atypical shape and
Gross Pathology
75
abnormal signal intensity, mainly involving the subcorticalwhite matter (Guerrini et al., 2003).
Tuberous sclerosis complex is transmitted as anautosomal dominant trait, with variable expression seen withinfamilies. Recurrence in siblings of non-affected parents hasrarely been reported. Between 50 to 75% of all cases aresporadic. Linkage studies have allowed the identification of twoloci for tuberous sclerosis complex (TSC), mapping tochromosome 9q34 (TSC1) and 16p13.3 (TSC2) (Povey et al.,1994). About 50% of the familial cases are linked to TSC1 (VanBakel et al., 1997).
Clinical assessment indicated that sporadic patients withTSC1 mutations had, on average, a milder disease than didpatients with TSC2 mutations, including a lower frequency ofseizures, moderate to severe mental retardation, fewersubependymal nodules and cortical tubers, less severe kidneyinvolvement, no retinal hamartomas, and less severe facialangiofibroma (Dabora et al., 2001).
Epileptic seizures are frequent in tuberous sclerosiscomplex. They usually begin before the age of 15, mostly in thefirst 2 years of life: 63.4% before one year, 70% before twoyears. Infantile spasms are the most common manifestation ofepilepsy in the first year of life, sometimes preceded by partialseizures (Guerrini et al., 2003).
mainly comprise gangliogliomas and dysembryoplasticneuroepithelial tumours (Blumcke et al., 2009). Bothneoplasms are rare, with an incidence of approximately 1.3% ofall brain tumours (Blumcke and Wiestler, 2002). Any lobe canbe affected, but temporal lobe locations appear to be far morefrequent for both gangliogliomas and dysembryoplasticneuroepithelial tumors (Guerrini et al., 2003).
They are frequent in children and young adults sufferingfrom pharmacologically intractable focal epilepsy. However,the differentiation between a neoplastic and dysplastic lesion isoften difficult to obtain, either using electrophysiology (EEGrecording), imaging, histopathology or molecular-geneticanalysis (Blumcke et al., 2009).
In large series of patients with surgically-treated drug-resistant epilepsy due to neoplastic lesions, gangliogliomas anddysembryoplastic neuroepithelial tumor represent the majority(50 – 75%) of histopathologically diagnosed lesions (Zentner etal., 1997).
Gangliogliomas are histologically characterised by aglioma component intermixed with an atypical neuronal organglion cell component. Atypical neuronal or ganglion cells
Gross Pathology
77
are frequently binucleate. Cell proliferation studies show thatthe tumour growth rate is slow. Dysembryoplasticneuroepithelial tumors are similar to gangliogliomas, butcytological atypia are more rare. Dysplastic neurons frequentlylie adjacent to the neoplastic lesions (Prayson, 2001).
Neuroradiological studies typically show a hypodenselesion on CT scan, with possible associated hyperdensecalcified lesions. Overlying skull can be deformed insuperficially located lesions. MRI scans show a hyperintenseT1 lesion that is usually peripherally enhanced after gadoliniumadministration. Gray and white matters are both involved. Awell demarcated, multilocular appearance is typically seen(Guerrini et al., 2003).
Dysplastic disorganization of the cortex near but separatefrom the tumour has often been observed and particularlystudied in dysembryoplastic neuroepithelial tumors (Prayson etal., 1993).The large majority of glioneuronal tumours presenteither with mild malformations of cortical development or withfocal cortical dysplasia type IA; only a few cases have beenreported to be associated with focal cortical dysplasia type II(Ferrier et al., 2006).
Whether focal cortical dysplasia occurring in associationwith a glioneuronal tumour represents a distinct entity (differentfrom isolated focal cortical dysplasias) is a matter of ongoing
Gross Pathology
78
debate. The maldevelopmental and dysembryoplastic nature ofglioneuronal tumours is likely to compromise always normalmaturation of the adjacent neocortex. However, none of thecurrent classification systems for tumours of the central nervoussystem, nor malformations of cortical development specificallyaddress or clarify this issue (Blumcke et al., 2009).
Clinical presentation is with drug resistant partialepilepsy. In a population of 89 patients with dysembryoplasticneuroepithelial tumors, partial seizures were the first clinicalsigns in 75%, while only 9% had neurological deficitsconsisting of quadranopsia. Epilepsy started at a mean age ofnine years (range 1-20 years) and proved resistant to differentantiepileptic medications. Complete surgical removal of thelesion was associated with remission of epilepsy in all patients(Guerrini et al., 2003).
The cellular mechanisms underlying epileptogenicity ofglioneuronal tumours and/or the perilesional cortical tissue arestill not clearly defined. Intrinsic epileptogenicity is supportedby electrocorticography, surgical and immunocytochemicalstudies, suggesting the presence of a hyperexcitable neuronalcomponent (Aronica et al. 2001a).
Developmental alterations compromising the balancebetween excitation and inhibition are likely to play a role in thepathogenesis of epileptic focal discharges in patients with
Gross Pathology
79
glioneuronal tumours (Aronica et al., 2007a). Recent evidencepoints to the inflammatory response as a contributing factor inthe epileptogenicity of these developmental lesions (Ravizza etal., 2006). Similarly to other brain tumours (such as gliomas),the peritumoral region may also be relevant for the generationand propagation of seizure activity (Van Breemen et al., 2007).
Long-term follow-up studies of a large series of patientsrevealed favorable outcomes for patients with supratentorialglioneuronal tumours, with only rare cases of tumourrecurrence or malignant progression to glioblastoma reportedfor gangliogliomas (Luyken et al., 2003). Limited informationis available about recurrence or malignant transformation ofdysembryoplastic neuroepithelial tumors (Maher et al., 2008).
The large majority of patients with glioneuronal tumoursbecame seizure free after surgical resection. Short duration ofepilepsy before surgery, absence of secondary generalizedseizures or status epilepticus, absence of additional pathologiesand complete resection predicted a better post-operative seizureoutcome (Luyken et al., 2003).
Thus, an early identification of glioneuronal tumoursassociated with chronic intractable epilepsy, followed by aprompt referral to epilepsy surgery centers provides the bestchance for curing epilepsy and preventing its recurrence and
Gross Pathology
80
possible malignant transformation. However, it has not beensystematically studied whether failure of post-surgical seizurerelief results from unrecognized cortical disorganization in thevicinity of the resection site (Blumcke et al., 2009).
Do seizures start within the lesion or theperilesional region?
Based on the focal, circumscribed nature of the lesions inmost of these malformations of cortical development, it wouldseem to make intrinsic sense that seizures originate within thelesions. The high success rate of “lesionectomy” duringepilepsy surgery, with many studies reporting over a 60– 75%seizure-free rate, also supports the idea that the lesions directlyproduce seizures (Aronica et al., 2001a).
However, this still leaves a substantial minority ofpatients that continue to have seizures following lesionectomy,suggesting that the epileptogenic zone was not contained withinthe lesion in those cases. Furthermore, the success oflesionectomy in eliminating seizures may have otherinterpretations: The margins of resection typically contain some“normal” perilesional tissue, which may actually be the primarysource of the seizures. Alternatively, perilesional cortex,immediately adjacent to or even distant from the lesion, maygenerate the seizures, but may be somehow dependent on the
Gross Pathology
81
lesion for epileptogenesis; removal of the lesion eliminates thisdriving force for seizure generation within the remaining cortex(Wong, 2008).
Finally, studies correlating radiographic and pathologicaldata increasingly indicate that areas of “normal”-appearingcortex on MRI in patients with other discrete regions ofmalformations of cortical development often contain subtlehistopathological abnormalities (Porter et al., 2003).
In tuberous sclerosis complexIn tuberous sclerosis complex, a variety of clinical
studies, including electrophysiological and radiographicinvestigations, suggest that cortical tubers are the primary siteof epileptogenesis. EEG often identifies both interictalepileptiform abnormalities and seizures originating from theimmediate region of a putative epileptogenic tuber on MRI.Furthermore, nuclear medicine radiographic studies, such asPET and SPECT, often point to specific tubers as being thesource of seizures (Koh et al., 2000).
Finally, surgical approaches for epilepsy specificallytargeting tubers often result in seizure freedom in at least 75%of patients strongly supporting the idea that tubers are thesource of the seizures in these cases (Weiner et al., 2006).
Gross Pathology
82
Despite the abundant clinical evidence implicating tubersas the epileptogenic foci in tuberous sclerosis complex, anumber of limitations reduce the certainty of this conclusion. Inmany cases, radiographic and electroencephalographic data donot have a high enough spatial resolution to clearly distinguishan epileptogenic source from within a tuber itself versus theadjacent perituberal region. In addition, most surgical resectionsalso include at least some margin of normal-appearing cortexsurrounding the tuber, making it difficult to rule out theperituberal region as the source of seizures and the reason forsuccess with surgery. Furthermore, some patients continue tohave seizures despite an appropriately targeted tuberectomy.Finally, most of the clinical data supporting the importance oftubers derive from series of tuberous sclerosis complex patientsthat underwent epilepsy surgery, which likely represents abiased, preselected group. Other tuberous sclerosis complexpatients, who were not deemed to be good surgical candidates,as well as patients who failed epilepsy surgery, may have othermechanisms of epileptogenesis that are not as tightly linked totubers (Wong, 2008).
Some clinical evidence suggests that the nontuber regionsof cortex could also be a source of epileptogenesis. QuantitativeMRI studies indicate that tuberous sclerosis complex patients
Gross Pathology
83
have diffusely decreased cortical grey matter volume, notspecifically related to cortical tubers (Ridler et al., 2001).
Also, diffuse cellular abnormalities have been reported insome cases of tuberous sclerosis complex brains independent oftubers, such as atypical poorly differentiated cells anddecreased neuronal counts (Roske et al., 2003).
Although the relevance of these more diffuseradiographic and histological abnormalities to epilepsy is notestablished, they at least raise the possibility that nontubercortex is abnormal and may be capable of generating seizures(Wong, 2008).
In addition, more direct clinical evidence for the role ofnontuber cortex in causing seizures is seen in rare reports oftuberous sclerosis complex patients with intractable epilepsy,who become seizure free following surgical resection ofnormal-appearing, tuber-free brain tissue (Wang et al., 2007).
In focal cortical dysplasiaClinical evaluations, including EEG, MEG, and various
imaging methods, frequently provide evidence that seizuresoriginate intrinsically from within the focal cortical dysplasiaevident on MRI (Bast et al., 2004).
Gross Pathology
84
Surgical resection of focal cortical dysplasia results inseizure freedom in at least 50% of patients, again supporting theconcept of the epileptogenic foci being contained within thefocal cortical dysplasia (Tassi et al., 2002).
There is substantial clinicopathological evidence thatseizures can also arise from regions beyond the focal corticaldysplasia, or at least outside the area of the focal corticaldysplasia that is grossly evident on MRI. A likely explanationfor surgical failures in patients with focal cortical dysplasia isthat the true area of focal cortical dysplasia may extend on themicroscopic level beyond the region of obvious abnormalityapparent on MRI (Gomez-Anson et al., 2000).
Limited by the resolution of imaging technology as theresolution of imaging methods continues to improve, brainregions that appear “normal” by current techniques mayeventually be identified as “lesional.” In fact, a retrospectivediagnosis of focal cortical dysplasia is often made bypathological analysis of brain tissue resected from patients withintractable epilepsy, who had no evidence of focal corticaldysplasia on preoperative MRI (Bautista et al., 2003).
Consistent with this idea, advanced quantitative MRItechniques have found abnormalities in grey matter volume
Gross Pathology
85
beyond the regions of focal cortical dysplasia identified byconventional visual inspection (Wong, 2008).
In gangliogliomasPerilesional mechanisms of epileptogenesis are also
strongly implicated in cases of ganglioglioma. Epileptiformdischarge patterns emanating from ganglioglioma onelectrocorticography (Ferrier et al., 2006) and the success oflesionectomy in eliminating seizures (Aronica et al., 2001a),suggest the possibility that seizures could directly start withinthe ganglioglioma itself.
However, the limited spatial resolution of this type ofclinical data, as similarly discussed above for tuberous sclerosiscomplex, makes it difficult to rule out that peritumoralmechanisms actually account for these clinical observations.There are limited electrophysiological data documentingwhether cells in ganglioglioma are electrically excitable. Thus,many studies have focused on secondary effects of the tumor onperitumoral regions as the basis for epileptogenesis inganglioglioma (Wong, 2008).
In fact, electrophysiological data from both human tissueand animal models indicate that the regions adjacent to or at theborder of gliomas have the highest potential to generateepileptiform activity (Patt et al., 2000).
Gross Pathology
86
Is epileptogenesis primarily a result of circuitabnormalities or cellular/molecular defects?
Epileptogenesis in malformations of corticaldevelopment is primarily due to circuit abnormalities or cellularand molecular defects. Seizures clearly consist of synchronouselectrical activity reverberating through complex neuronalnetworks and thus ultimately must always include abnormalitieson the circuit level. However, from a pathophysiologicalstandpoint, mechanisms of epileptogenesis could involve eitherprimary changes in circuit organization or initial cellular andmolecular defects that secondarily translate to the networklevel. On the extremes, circuit abnormalities (the epilepticcircuit) might consist of aberrant connectivity of neurons thatare otherwise completely normal in function, whereascellular/molecular mechanisms (the epileptic neuron) wouldinvolve a defect involving intrinsic neuronal function in thecontext of normally wired and fully operational circuits.Ultimately, both network and cellular/molecular abnormalitieswill stimulate epileptogenesis by upsetting the normalphysiological balance between excitation and inhibition in thebrain (Wong, 2008).
In tuberous sclerosis complexValencia et al. (2006) reported some limited
immunohistochemical evidence for anomalous GABAergic
Gross Pathology
87
inhibitory circuits within cortical tubers. With intracellularrecordings of normal-appearing neurons in nontuber tissueresected from a tuberous sclerosis complex patient, there wasevidence of neuronal hyperexcitability but no impairment ofsynaptic inhibition (Wang et al., 2007).
On the cellular level, a popular hypothesis about seizuregeneration in tuberous sclerosis complex is that an abnormalcell type, in particular the giant cell in tubers, could serve as anintrinsic “pacemaker” that initiates and drives epileptiformactivity and seizures. Contrary to the “pacemaker” hypothesis,giant cells from tuberous sclerosis complex patients wereactually found to be electrically inexcitable, with no evidence ofvoltage-activated sodium or calcium currents. Cytomegalicneurons from tuberous sclerosis complex specimens werecapable of generating action potentials, including repetitivecalcium spikes in response to stimulation, and thus havepotential for contributing to epileptic discharges, but showed noevidence of intrinsic pacemaker properties. Thus, the availablephysiological data do not support the concept that giant cells orother dysmorphic neurons within tubers are, by themselves, theprimary generators of epileptiform activity in tuberous sclerosiscomplex (Wong, 2008).
The most evidence for potential abnormalities promotingepileptogenesis in tuberous sclerosis complex arguably exists
Gross Pathology
88
on the molecular level, largely derived from single-cellpolymerase chain reaction (PCR) and microarray analysis ofcells from tubers resected from tuberous sclerosis complexpatients with intractable epilepsy. Molecular characterizationrevealed increased mRNA expression of specific glutamateNMDA receptor subunits and a decrease in specific GABAA
receptor subunits in giant cells and dysplastic neurons. Thesespecific changes in neurotransmitter expression within tuberscould have obvious effects in promoting hyperexcitability andseizures (White et al., 2001).
In focal cortical dysplasiaOn the circuit level, both immunohistochemical and
electrophysiological studies indicate that abnormal, potentiallyhyperexcitable networks exist within focal cortical dysplasiatissue specimens resected from patients with intractableepilepsy (Wong, 2008).
A decrease or abnormal organization of GABAergicinterneurons within focal cortical dysplasia was demonstratedas assayed immunocytochemically by markers - parvalbumin,calbindin, or glutamic acid decarboxlyase. Also, Intracellularrecordings from pyramidal neurons in neocortical slices fromhuman dysplastic cortex demonstrated physiological evidence
Gross Pathology
89
of decreased GABA-mediated synaptic inhibition (Calcagnottoet al., 2005).
While there is strong evidence for impaired GABAergiccircuits in focal cortical dysplasia, which could clearly promotehyperexcitability and seizures, less is known about theorganization of circuits within and around focal corticaldysplasia that actually generate the seizures. Balloon cells donot appear to receive synaptic contacts (Alonso-Nanclares etal., 2005).
Small subpopulation of cytomegalic interneurons hasbeen described in focal cortical dysplasia that exhibit intrinsicbursting behavior (Andre et al., 2007). Furthermore, basket-likeclusters of GABAergic interneurons often surround cytomegalicneurons and could serve to synchronize epileptiform activity(Alonso-Nanclares et al., 2005).
It is conceivable that a primary impairment ofGABAergic circuits could lead to disinhibition,synchronization, and hyperexcitability of otherwise normal-appearing cortical pyramidal neurons within and surroundingthe regions of focal cortical dysplasia (Calcagnotto et al.,2005).
Gross Pathology
90
Dysplastic and heterotopic pyramidal neuronsmicrodissected from human focal cortical dysplasia specimensexhibit decreased expression of specific GABAA receptorsubunits (Crino et al., 2001).
Immunocytochemical and single-cell mRNAamplification techniques have shown that specific subunits ofNMDA, AMPA, and metabotropic glutamate receptors arealtered, typically increased, in dysplastic neurons from focalcortical dysplasia (Aronica et al., 2003a). Physiologically,NMDA receptors of pyramidal neurons from focal corticaldysplasia display decreased sensitivity to magnesium inhibition,which could promote increased neuronal excitability (Andre etal., 2004).
While glutamate and GABA receptors tend to receive themost attention in mechanisms of epileptogenesis, a number ofother molecular players can also influence excitability in focalcortical dysplasia. Some evidence indicates that epileptic tissuefrom focal cortical dysplasia may recapitulate or maintainimmature properties. The immature brain tends to have adecreased seizure threshold which may, in part, be due to aparadoxical excitation due to GABA during early braindevelopment. While GABA causes hyperpolarization andinhibition of neurons in adulthood, a relatively elevatedintracellular chloride concentration in immature neurons leads
Gross Pathology
91
to depolarization and excitation by GABA in neonatal rodentand human cortex (Wong, 2008).
Astrocytes control the extracellular levels of excitatoryions and neurotransmitters, such as potassium and glutamate,which may directly affect neuronal excitability. In addition,astrocytes recently have been shown to release glutamate andother substances as intrinsic “gliotransmitters,” which candirectly stimulate neurons and participate in synaptic signalingin the so-called tripartite synapse (Haydon, 2003).
Given the prominent role of mature glial cells andglioneuronal progenitors cells in the malformations of corticaldevelopment with abnormal glioneuronal proliferation, it wouldbe logical to hypothesize that glial abnormalities might alsocontribute to epileptogenesis in these malformations of corticaldevelopment, such as tuberous sclerosis complex and focalcortical dysplasia. Histological abnormalities in glia in tubers,such as the presence of poorly differentiated giant cells withmixed glial-neuronal properties and astrocyte proliferation,suggest a possible role of glia in epileptogenesis in tuberoussclerosis complex. There are fewer data related to astrocyticregulation of glutamate and potassium in focal corticaldysplasia, although the similarities in histological features ofglia in these malformations of cortical development comparedwith tuberous sclerosis complex make it likely that analogous
Gross Pathology
92
astrocytic abnormalities occur in all these disorders (Wong,2008).
In gangliogliomasLess is known about the effects of ganglioglioma on
circuit properties and network excitability, although, asdiscussed earlier, it may be neuronal networks at the border orperilesional regions of ganglioglioma that are the mostepileptogenic (Patt et al., 2000). Evidence has been found fordecreased GABAergic interneurons in perilesional epilepticnetworks adjacent to ganglioglioma (Aronica et al., 2007b).
On the cellular and molecular level, a number ofabnormalities in glutamate and GABA receptors have beendocumented in ganglioglioma, which are analogous to tuberoussclerosis complex or focal cortical dysplasia, again suggestingthat these malformations of cortical development sharecommon pathophysiological origins and mechanisms ofepileptogenesis (Wong, 2008).
Immunocytochemical studies indicate that neuronalcomponents of ganglioglioma highly express NMDA andAMPA receptors (Aronica et al., 2001b). Single-cell mRNAanalysis detected changes primarily in metabotropic glutamatereceptors in ganglioglioma, but also demonstrated reducedexpression in GABAA receptor subunits (Samadani et al.,2007).
Gross Pathology
93
Expression of drug transporters inpathological lesions associated with refractoryepilepsy
Tishler et al. (1995) measured MDR1 expression in 19patients undergoing respective epilepsy surgery, 15 of whomreceived temporal lobectomy for a mixture of pathologies(mostly hippocampal sclerosis). MDR1 mRNA level was foundto be >10 times higher in 11 of the 19 resected samplescompared with controls (“normal” brain tissues resected duringremoval of arteriovenous malformations).
Sisodiya et al. (2002) stained both P-gp and MRP1 inastrocytic cells, but not capillary endothelium, in thehippocampus in cases of hippocampal sclerosis.
More recently, Aronica et al. (2004) performed detailedimmunostaining studies in brain sections from 16 patients withhippocampal sclerosis and found upregulation of P-gp andMRP2 in capillary endothelium.
The other group of pathologic lesions frequentlyassociated with intractable epilepsy that has been more widelyinvestigated for over-expression of efflux transporters ismalformations of cortical development. focal cortical dysplasiatissues removed from patients with refractory epilepsy showedintralesional (but not perilesional) induction of MRP1 in
Gross Pathology
94
dysplastic neurons, balloon cells, and glial processes aroundblood vessels, whereas P-gp over-expression was observedprimarily in the glial component and capillary endothelium(Sisodiya et al., 2001).
In patients with uncontrolled seizures associated withtuberous sclerosis complex, immunostaining in resected corticaltubers for P-gp and MRP1 was observed in dysplastic neurons,balloon cells, astrocytes, and microglial cells, whereas only P-gp was upregulated in blood vessels (Lazarowski et al., 2004).
In addition to hippocampal sclerosis and malformationsof cortical development, over-expression of P-gp or MRP orboth has been noted in other epileptic pathologies. Positiveimmunostaining of both transporters was observed in reactiveastrocytes within dysembryoplastic neuroepithelial tumors(Sisodiya et al., 2002).
In ganglioglioma, removed from patients with intractableepilepsy, both P-gp and MRP1were detected in neuronal cells,MRP in glial cells, and P-gp in capillary endothelium within thelesion (Aronica et al., 2003b).
Management
95
MANAGEMENT
Anti-epileptic drugs
Epilepsy surgery
Ketogenic diet
Vagus nerve stimulation
Treatments under investigation
Management
96
Although relative drug-resistant epilepsy can bediagnosed after failure of two anti-epileptic drugs , absolutedrug resistance requires failure of six anti-epileptic drugs , as asignificant minority of patients (16.6%) is rendered seizure freeby addition of newly administered anti-epileptic drugs evenafter failure of two to five antiepileptic drugs (Schmidt, 2009).
The chances of controlling epilepsy decline sharply afterfailure of the second or third antiepileptic drug trial. In fact,some clinicians would argue against trying another antiepilepticdrug in these patients, who may be candidates for surgicalprocedures that have high rates of success (Berg, 2004).
Common causes of treatment failure, such as poorcompliance or inappropriate selection of first-line antiepilepticdrugs, should be addressed early on by the treating physician.Nonadherence to the prescribed regimen is a very commoncause of uncontrolled seizures, so it is critical to maintain agood rapport with the patient and to inquire about reasons fornoncompliance (Pati and Alexopoulos, 2010).
Importantly, the prognosis for most patients with newlydiagnosed epilepsy, whether good or bad, becomes apparentwithin a few years of starting treatment. A history of a lack of asustained seizure-free period for 12 consecutive months, inspite of two or three suitable and tolerated antiepileptic drugs, isa definite red flag for clinicians and should prompt referral to aspecialist center (Kwan and Brodie, 2000).
Management
97
Evaluating patients with suspected pharmacoresistantepilepsy demands a systematic and holistic approach with equalemphasis on quality of life and psychosocial and cognitivefactors (Siegel, 2004).
The medical, social, and economic consequences ofpoorly controlled seizures can be enormous. Recurrent seizuresare associated with significant risks for death, physical injury,cognitive impairment, and psychosocial problems. Frequentseizures not only influence quality of life, morbidity, andmortality in epilepsy, but also significantly increase costs(Schmidt, 2009).
The clinical assessment should be based on the followingprinciples:
1. Review and confirm the diagnosis of epilepsy with the helpof a careful history, video-EEG, and imaging. When seizurescannot be controlled with drugs, it is important to verify thatthe events in question are indeed epileptic. Continuous video-EEG monitoring may be necessary to capture andcharacterize the clinical manifestations and correspondingEEG changes.
2. Identify the cause, type of seizure or seizures, andsyndromic classification, if any.
3. Review past and present medications, doses, efficacy, andside effects. Consider the possibility of drug interactions.
4. Choose antiepileptic drugs primarily on the basis of thetype of seizures and the individual clinical scenario: Which
Management
98
drug is likely to be most efficacious with the fewest sideeffects, and which one is appropriate for the patient’scomorbidities and concomitant medications?
5. Discuss issues such as seizure precautions, lifestylemodifications, psychosocial dysfunction, and suddenunexpected death. (Pati and Alexopoulos, 2010)
Fig. (4) Clinical approach to patients with pharmacoresistant epilepsy
(Pati and Alexopoulos, 2010).
Management
99
Anti-epileptic drugs
In chronic epilepsy (more than 5 years), the addition of anew anti-epileptic drug provided a seizure freedom of 17% anda 50–99% seizure reduction of 25%. For those who did notrespond to the first trial, a similar benefit might be expected forat least 2 more trials. At the end, 28% of the patients wereseizure free. The application of a systematic protocol to thetreatment of refractory epilepsy using a new anti-epileptic drugmight improve seizure control in a substantial proportion ofcases. The nihilistic view that intractability is inevitable ifseizure control is not obtained within a few years of the onset oftherapy is incorrect (Luciano and Shorvon, 2007).
In refractory epilepsy, it is convenient to perform asystematized management of anti-epileptic drug: (1) increaseuntil the maximum tolerable dose; (2) if no response, replacethe anti-epileptic drug, if there is a partial response, add anotheranti-epileptic drug which should be chosen based on themechanism of action of the first anti-epileptic drug (e.g.lamotrigine and valproate are synergic), its efficacy and adverseeffects (Brodie, 2005).
Epilepsy surgery
Resective surgery is based on removal of the entireepileptogenic area without causing a permanent neurological
Management
100
deficit. The localization of the epileptogenic zone in focalepilepsy is typically based on seizure semiology, interictal andictal EEG findings, as well as FDG-PET, SPECT and MRIlesions (Rosenow and Luders, 2001).
Is the patient a candidate for epilepsy surgery?
The rationale for surgical management ofpharmacoresistant focal epilepsies is to eliminate orsignificantly reduce the patient’s propensity for spontaneousseizures by removing the epileptogenic focus (Pati andAlexopoulos, 2010).
Several factors need to be considered in the course of acomprehensive and multidisciplinary specialized evaluationbefore answering the critical question of whether a patient withintractable seizures may be a candidate for respective epilepsysurgery.
1. Is the epilepsy diagnosis correct?2. Is the epilepsy focal? Have the following possibilities been
excluded: generalized or multifocal epilepsy, situational orprovoked seizures, or an epilepsy syndrome withspontaneous remission?
3. Do seizures remain poorly controlled despite adequatepharmacologic trials?
Management
101
4. If so, do the seizures or medication side effects significantlyaffect the patient’s quality of life?
5. Can an epileptogenic lesion be seen on MRI, and what isthe suspected etiology?
6. Is there converging evidence for a single epileptogenicfocus?
7. Are there abnormalities elsewhere in the brain?8. What are the chances of a good outcome in terms of seizure
control and improvement in quality of life?9. What are the risks of surgery, and how do these compare
with the risks of not having surgery?10. What are the patient’s perceptions and attitudes toward
epilepsy surgery? (Alexopoulos and Najm, 2009)
Focal epilepsy with a lesion not adjacent to the eloquentcortex and concordant with semiology, ictal EEG, interictalEEG and PET/SPECT may be removed based solely on surfaceevaluation. In the case of focal epilepsy without a lesion, alesion adjacent to an eloquent cortex or if there is noconcordance between the different zones, invasive monitoringis recommended (Rosenow and Luders, 2001).
Management
102
Fig. (5) Specialized diagnostic and treatment options for patientswith pharmacoresistant epilepsy (Pati and Alexopoulos, 2010).
Preoperative counseling is essential for the patient andhis or her family, addressing the goals, risks, and benefits of thesurgery. Treatment decisions should take into account thepossible impact of surgery on the patient’s medical andpsychosocial circumstances (risks of ongoing seizures vs
Management
103
surgical intervention; impact on the patient’s independence,employment status, emotional well-being, and psychiatric andother comorbidities) (Pati and Alexopoulos, 2010).
Curative procedures
Curative procedures include lobectomy, lesionectomy,and multilobar or hemispheric surgery (hemispherectomy).
A.Anterior temporal lobectomy and hippocampectomy
More than half of the procedures in surgical epilepsyprograms are anterior temporal lobe resections. Mesial temporallobe epilepsy associated with hippocampal sclerosis is the mostcommon form of focal epilepsy, with around 60% of thepatients having temporal resection. 60–70% of the patients arefree of seizures at 1–2 years of follow up and only 58% areseizure free at 10 years (Duncan, 2007).
B. Lesionectomy and lobectomy
Lesionectomy and lobectomy are respective approachestargeting seizure foci outside the temporal lobe (most often inthe frontal lobe, less commonly in the parietal or occipitallobes) or within the temporal lobe but outside the hippocampus(neocortical temporal lobe epilepsies). Patients with seizuresdue to structural lesions that are visible on MRI (“lesionalepilepsies,” eg, cavernous angiomas or circumscribed low grade
Management
104
tumors) may become seizure-free after limited resectionstargeting the lesion itself (lesionectomy) or extending to involvepart of a lobe or an entire lobe (lobectomy) (Pati andAlexopoulos, 2010).
Extratemporal lobe surgery for focal epilepsy accountsfor less than half of all epilepsy operations. In frontal lobeepilepsy surgery, the probability of becoming seizure free is55.7% at 1 year, 45.1% at 3 years, and 30.1% at 5 years. Thesubset of patients with favorable prognostic factors – an MRIlesion restricted to one frontal lobe, complete resection, and aregional or lateralized ictal scalp EEG pattern – show a seizurefree outcome approaching that seen after temporal lobectomy,with 50– 60% being seizure free at 3 years. Regarding etiology,patients with low-grade tumors have the best outcome (62%),followed by patients with MRI malformations of corticaldevelopment (52%) (Beleza, 2009).
On the other hand, identifying the epileptogenic focus inpatients with no visible structural abnormality on MRI(“nonlesional epilepsies”) can be challenging and usuallyrequires intracranial investigations. In this instance, the aim ofsurgery is to resect regions that are electrographically abnormal.In general, the postoperative outcome is less favorable innonlesional focal epilepsies than in lesional epilepsies(Cascino, 2004).
Management
105
C. Multilobar resections and hemispherectomy
Multilobar resections and hemispherectomy are indicatedwhen seizures arise from extensive, diffuse, or multiple regionsof a single hemisphere (Pati and Alexopoulos, 2010).
If the neurologic function supported by the abnormalhemisphere is intact, a tailored multilobar resection aims ateliminating the epileptogenic focus without creating newdeficits. If, however, the underlying hemispheric abnormality isassociated with significant contralateral hemiparesis,hemiplegia, or visual field deficits, the need to preservefunction does not limit surgery, and hemispherectomy can beconsidered. Hemispherectomy can be the procedure of choicefor young children with catastrophic epilepsies of diverseetiologies such as malformations of cortical development,Rasmussen’s encephalitis, Sturge-Weber syndrome, and remotevascular insults (Gonzalez-Martinez et al., 2005).
Palliative procedures
A.Corpus callosotomy
Corpus callosotomy (transection of the corpus callosum)is performed in a small number of patients, ie, those who havedisabling seizures that rapidly become generalized or injuriousdrop attacks and are not candidates for focal resection. Bydisconnecting the two hemispheres, this procedure aims to
Management
106
hinder the fast interhemispheric spread of seizure discharges(Pati and Alexopoulos, 2010).
Callosotomy may be complete or involve only a portionof the corpus callosum. The extent of resection has beencorrelated with favorable outcome (Tanriverdi et al., 2009).
Some investigators report a 50% or greater reduction inseizure frequency, with drop attacks and generalized tonic-clonic seizures showing the most consistent improvement. Inaddition, behavior and quality of life may also improve (Asadi-Pooya et al., 2008).
B.Multiple subpial transections
Multiple subpial transections are reserved for seizuresarising from eloquent cortex (ie, from areas that cannot beremoved without causing unacceptable neurologic deficits).Therefore, the surgeon only transects the epileptogenic cortexin a vertical manner, so as to interrupt the horizontal corticalconnections without resection. This approach is thought todisrupt the synchrony of seizure propagation while preservingphysiologic function (Pati and Alexopoulos, 2010).
A meta-analysis of small case series suggests somedecrease in seizure frequency with no or minimal neurologiccompromise in up to 60% of patients (Spencer et al., 2002).
Management
107
Complications of epilepsy surgery
Resective surgery is not without risk, but often the risk ismuch less than that posed by uncontrolled epilepsy in the longterm. Operative mortality rates vary from almost zero fortemporal lobe surgery to 2.5% for hemispherectomy. Thereported risk of permanent surgical morbidity varies by type ofsurgery from 1.1% for temporal lobe resection to about 5% forfrontal lobe resection (Chapell et al., 2003).
Ketogenic diet
Originally developed almost a century ago, the dietmimics the biochemical changes associated with starvation. It isa strict regimen, high in fat and low in carbohydrate and protein(typically in a ratio of 4:1 or 3:1 in adolescents and very youngchildren) (Pati and Alexopoulos, 2010).
Ketogenic diet is mainly used in pediatric patients (due totolerability) as second line treatment in focal nonsurgicalrefractory and generalized symptomatic epilepsy. A recentrandomized controlled trial showed a reduction in seizurefrequency more than 50% in 38% of children with drug-resistant epilepsy (Neal et al., 2008).
Such a strict regimen is difficult to implement andmaintain and requires close supervision by a dietician andphysician. In addition to the practical complexities, concerns
Management
108
also exist about the long-term effects of the diet on the child’sgrowth and overall health. For these reasons, the ketogenic dietis restricted to a small group of young patients withpharmacoresistant epilepsy and is not usually used long (Patiand Alexopoulos, 2010).
Potential adverse effects of the ketogenic diet includelethargy, weight loss, nausea and vomiting, constipation, anddiarrhea. Furthermore, the diet’s use necessitates the frequentmonitoring of complete blood count levels, electrolyte values,and liver and renal status, as additional infrequent adverseeffects can include hyperlipidemia, hypoglycemia,hypocalcemia, electrolyte imbalances, and metabolic acidosis, inaddition to cardiac and renal abnormalities (Olson, 2005).
There are few data indicating when it is appropriate toterminate the diet in patients who have a favorable response,but most clinicians wean the patient after 2 to 3 years. Reportson the use of the ketogenic diet in adults are scarce, althoughbenefit was seen in a small series. No long-term follow-up dataexist for adults, especially regarding the risk of atherosclerosis(Pati and Alexopoulos, 2010).
Management
109
Vagus nerve stimulation
Vagus nerve stimulation is a nonpharmacologicalternative for adults and for adolescents over age 12 years whohave intractable focal seizures and who are not favorablesurgical candidates. Its effectiveness in younger patients and inthose who have intractable generalized seizures is less clear(Holmes et al., 2004). It might also be effective in children withdrop attacks and Lennox-Gastaut syndrome (Beleza, 2009).
A device consisting of a pulse generator is implantedsubcutaneously in the precordium, and a lead wire is tunneledunder the skin and attached to the left vagus nerve. Thegenerator is programmed using a telemetry wand held over thedevice, with settings for current intensity (typically 1–2 mA),pulse width (250–300 μsec), frequency (30 Hz), and “dutycycle” (typically 30 seconds on stimulation, followed by 3 to 5minutes off, cycling 24 hours/day). Hence, it provides “open-loop stimulation,” ie, continuous stimulation that is notmodified in response to the patient’s EEG seizure activity.Patients or caregivers can also activate the device manually(“on demand”) at the first sign or warning of an impendingseizure by swiping a handheld magnet (Pati and Alexopoulos,2010).
Common side effects such as cough, voice alteration, andhoarseness are usually stimulation-dependent and tend to
Management
110
diminish with time. Notably, vagus nerve stimulation has noneof the cognitive side effects often encountered with increasingdoses of antiepileptic drugs. As with other implantablestimulators, some safety concerns exist in patients undergoingmagnetic resonance imaging (Pati and Alexopoulos, 2010).
At least one-third of patients who receive this treatmentshow a sustained response, defined as a 50% or greaterreduction in seizures. However, few achieve freedom fromseizures, and therefore this therapy is considered palliative andis reserved for patients who are not candidates for surgery or forwhom surgery has failed (Pati and Alexopoulos, 2010).
Management
111
Treatments under investigation
Polymers, electrical brain stimulation and prediction ofseizures may be available in the future for treating patients withrefractory epilepsy. Cell transplantation and gene therapy,although holding great promise, are still far from routineclinical use (Nilsen and Cock, 2004).
Local drug delivery
Polymers containing anti-epileptic drugs consist of 2- to3-mm microspheres that might be placed near the epileptogeniczone. Advantages include: (1) new anti-epileptic drugs couldbe used including those which do not cross the blood-brainbarrier or show systemic toxicity; (2) they may be useful whenthe epileptogenic zone is near eloquent cortex; (3) they preventnoncompliance (Kwan and Brodie, 2006).
Implanting wafers impregnated with chemotherapeuticagents into the resection cavity results in prolongation ofsurvival without an increased incidence of adverse events (Hartet al., 2008).
Targeted electrical stimulation
To modulate abnormal cortical hyperexcitability,electrical stimulation can be applied to the peripheral nervoussystem (eg, vagus nerve stimulation) or central nervous system.
Management
112
Central nervous system stimulation can be broadly divided intotwo approaches:
Direct stimulation targets presumed epileptogenic braintissue such as the neocortex or hippocampus.
Indirect stimulation targets presumed seizure-gatingnetworks such as in the cerebellum and various deep brainnuclei in the basal ganglia or thalamus (deep brainstimulation), which are believed to play a central role inmodulating the synchronization and propagation of seizureactivity (Pati and Alexopoulos, 2010).
Electrical brain stimulation is still not accepted as a routinetreatment for epilepsy, partly because there is no consensusregarding the better region to stimulate and in what type ofseizure it is most effective. The epileptogenic zone and thecentromedian or anterior nuclei of the thalamus seem to be themost effective targets for electrical stimulation. The efficacyseems to be similar to vagal nerve stimulation which has alower risk and less comorbidity. This intervention is thusunlikely to be routinely used in the future (Beleza, 2009).
Cell and gene therapies
In ex vivo gene therapy, bioengineered cells capable ofdelivering anticonvulsant compounds might be transplanted intospecific areas of the brain. On the other hand, in vivo gene
Management
113
therapy would involve delivering genes by viral vectors toinduce the localized production of antiepileptic compounds insitu. Cell transplantation is aimed at restoring the physiologicbalance of neurotransmitters. Cell transplantation (heterologousfetal cell grafts or embryonic or adult stem cells) has thepotential to form restorative synaptic connections andassimilate within existing cells and networks in the host tissue(Pati and Alexopoulos, 2010).
Discussion
114
DISCUSSION
Epilepsy is a common and devastating neurologicaldisorder. In many patients with epilepsy, seizures are well-controlled with currently available anti-epileptic drugs , but asubstantial proportion (about 30%) of patients continue to haveseizures despite carefully optimized drug treatment (Remy andBeck, 2006).
Firstly it is mandatory to exclude false refractoriness relatedto nonepileptic seizures, inadequate anti-epileptic drugs,noncompliance and seizure-precipitating factors. Video-EEGmonitoring is an essential tool in this process, aiming toperform a differential diagnosis of paroxysmal events and acorrect classification of seizures and epileptic syndromes(Beleza, 2009).
An important characteristic of medically intractable(pharmaco-resistant) epilepsy is that most patients withrefractory epilepsy are resistant to several, if not all anti-epileptic drugs , even though these drugs act by differentmechanisms (Kwan and Brodie, 2000).
Although no single accepted definition exists of drugresistant epilepsy, different definitions have been proposeddepending on the context. All are based on the 3 main
Discussion
115
components of intractability: number of anti-epileptic drugspreviously taken, frequency of seizures and duration of non-controlled epilepsy (Beleza, 2009).
In general, many experts would agree that whenever apatient does not become seizure free for 12 months during long-term treatment with several suitable anti-epileptic drugs atmaximal tolerated doses, the epilepsy can be broadly classifiedas drug-resistant, pharmaco-resistant, or medically refractory(Schmidt and Loscher, 2005).
Refractory epilepsy is established when there is inadequateseizure control despite using potentially effective anti-epilepticdrugs at tolerable levels for 1-2 years, and excluding non-epileptic events and poor compliance (Beleza, 2009).
Two main hypotheses have been proposed to account forPharmacoresistant epilepsy. The transporter hypothesis statesthat pharmaco-resistance arises because anti-epileptic drugs donot gain access to their sites of action in the brain because ofover-expression of drug efflux transporters at the blood brainbarrier that limit anti-epileptic drug access to the brain. Thetarget hypothesis, on the other hand, states that target receptorsites are somehow altered in the epileptic brain so that they aremuch less sensitive to the anticonvulsant effects of systemicallyadministered drugs (Beck, 2007).
Discussion
116
The transporter and target hypotheses are the mostcommonly cited mechanisms of refractoriness, although theycannot yet fully explain refractoriness (Beleza, 2009).
Transporter hypothesis is based on two separate but parallelassumptions. First, increased expression of drug transporters isassociated with refractory epilepsy. Second, anti-epileptic drugsare substrates of these transporters. Otherwise, over-expressionof the transporters would have no clinical relevance (Kwan andBrodie, 2005).
According to the drug transporter hypothesis, restrictedaccess of anti-epileptic drugs to the seizure focus is the resultof locally increased expression of drug transporter proteins,most notably P-gp, encoded by the ABCB1 gene (Beleza, 2009).
A number of drug transporter genes and their proteins areover-expressed in the blood brain barrier of individuals withrefractory epilepsy. This has been demonstrated in tissues takenfrom epileptic foci at the time of resective surgery(Dombrowski et al., 2001).
In various pathologies commonly associated with drug-resistant epilepsy, upregulation of P-gp has been noted in braincapillary endothelium, consistent with the putative enhancedblood brain barrier function. It has been postulated that thismight represent an adaptive phenomenon contributing to cell
Discussion
117
survival, but its relevance to drug resistance is unclear. It islikely that the specific cell-distribution patterns of these effluxtransporters serve different cellular functions, which remain tobe fully illuminated (Kwan and Brodie, 2005).
Evidence to support the assumption that increasedexpression of drug transporters is associated with refractoryepilepsy has mainly been derived from epileptic brain tissuesremoved during epilepsy surgery from patients with drugresistant epilepsy. Interpretation of findings from surgicalstudies can be difficult because of the lack of proper normalcontrols for comparison (Kwan and Brodie, 2005).
The proposed mechanism suffers from a lack of evidencethat many clinically used anti-epileptic drugs are substrates forhuman P-gp or any other known human blood brain barrierefflux transporter (Anderson and Shen, 2007).
Although an impression is emerging that several anti-epileptic drugs may be subject to active transport by P-gp orMRP, inconsistencies in experimental findings exist, likelyreflecting the methodologic differences used by differentinvestigators. Further studies using more specific and sensitivemodels are needed before conclusive identification of anti-epileptic drugs as substrates of the various transporters can bemade (Kwan and Brodie, 2005).
Discussion
118
The transporter hypothesis has also failed to receive supportfrom recent genetic studies that failed to report an associationbetween polymorphisms in the ABCB1 gene and drug resistance(Leschziner et al., 2007).
Findings from several animal models suggest that P-gpexpression could be induced by seizures or treatment withcertain anti-epileptic drugs . Whether this induction effect canbe applied to the human situation is unknown and requiresfurther investigation (Kwan and Brodie, 2005).
There is as yet no direct proof that over-expression ofmultidrug transporters is a possible cause of drug resistance inthe treatment of epilepsy. In addition, other mechanisms ofpharmacoresistance should be identified, because it is likelythat different factors underlie multidrug resistance in epilepsy(Loscher and Potschka, 2002).
According to the target hypothesis, epilepsypharmacoresistance occurs when intrinsic or acquired changesin drug targets make them less sensitive to anti-epileptic drugs .Recent studies have provided evidence of reduced sensitivity tocarbamazepine in brain tissue from patients who were clinicallyunresponsive to carbamazepine and underwent resectivesurgery. However, it is unknown whether pharmacodynamicinsensitivity in these tissues extended to anti-epileptic drugs
Discussion
119
with different mechanisms of action or even to other anti-epileptic drugs that target sodium channels (Beleza, 2009).
The acquired version of the target hypothesis proposes thatthe pharmacodynamic sensitivity of the anti-epileptic drugtarget is modified by the disease state (Lazarowski et al., 2007).There are many examples of changes in the activity of voltage-gated and neurotransmitter-activated ion channels in acquiredepilepsy models, some of which lead to reduced responsivenessto anti-epileptic drugs (Remy and Beck, 2006). However, thereis no evidence that the efficacy of anti-epileptic drugs acting ondifferent targets is similarly affected (Beleza, 2009).
It seems highly unlikely that multiple targets will all besimultaneously altered in such away as to producepharmacoresistance. This should be kept in mind whenconsidering the target hypothesis for most anti-epileptic drugs(Beck, 2007).
An alternative mechanism for explaining refractoriness wassuggested “the intrinsic disease severity”. This hypothesisclaims that there are differences in inherent epilepsy severityreflected in the frequency of seizures in the early phase ofepilepsy. Possibly, common neurobiological factors mayunderlie both epilepsy severity and drug refractoriness.Subsequently, to advance in the understanding and therapeuticmanagement of refractory epilepsy, it is crucial to identify
Discussion
120
biomarkers which define the most severe forms of epilepsy.Unfortunately, there are few studies on the contribution ofgenetics to the severity of epilepsy (beleza, 2009).
Malformations of cortical development are often associatedwith severe epilepsy and developmental delay. About 40% ofdrug-resistant epilepsies are caused by malformations ofcortical development. Classification of malformations ofcortical development is based on embryological braindevelopment, recognizing forms that result from faulty neuronalproliferation, neuronal migration and cortical organization(Guerrini et al., 2003).
Studies of different malformations of cortical developmentwith abnormal glioneuronal proliferation, such as tuberoussclerosis complex, focal cortical dysplasia, and ganglioglioma,share some interesting trends regarding the site of origin forseizures. Both the lesion and the perilesional regions have beenimplicated in causing epileptogenesis in these disorders, but,somewhat paradoxically, an accumulating amount of evidencedemonstrates the importance of perilesional cortex in producingseizures. It is likely that the relative contribution of perilesionalversus lesional mechanisms varies between different types ofmalformations of cortical development and different patientswith the same type of malformations of cortical development.Recent studies suggest that the “perilesional” region may have
Discussion
121
subtle structural and cellular abnormalities. With furtheradvances in mechanistic studies and improved resolution ofimaging techniques, future research should reveal more detailedinformation about the perilesional region and its relationship tothe lesion (Wong, 2008).
There was over-expression of both P-gp and MRP1 inreactive astrocytes in the epileptogenic tissue indysembryoplastic neuroepithelial tumors, focal corticaldysplasia and hippocampal sclerosis, and MRP1 over-expression in dysplastic neurons in focal cortical dysplasia(Sisodiya et al., 2002) .
As the clinically practical definition of “lesion” is currentlylimited to the anatomical resolution of imaging methods, futureadvances will likely result in continual expansion of thedefinition and extent of the epileptogenic “lesion” from theanatomical to the cellular and molecular levels, so that thepresent distinction between “lesion” and “perilesional” regionsmay become obsolete (Wong, 2008).
The chances of controlling epilepsy decline sharply afterfailure of the second or third antiepileptic drug trial. In fact,some clinicians would argue against trying another antiepilepticdrug in these patients, who may be candidates for surgicalprocedures that have high rates of success (Berg, 2004).
Discussion
122
Defining epilepsy refractory to medical treatment impliesconsidering surgery. The prognosis will differ according to theepilepsy syndrome and etiology involved and depending onwhether the intervention is curative or palliative. In addition,other interventions such as a ketogenic diet and the contributionof anti-epileptic drugs should not be disregarded (beleza,2009).
The ketogenic diet, which was developed almost a centuryago, controls seizure activity by a mechanism that has not yetbeen identified. The dietary changes involved are complicatedand require extensive family commitment, but they may beextremely effective in seizure reduction. The diet includes 80%to 90% of calories from fat, protein appropriate for growth, andextreme carbohydrate restriction; the diet typically is moresuccessful with younger children in whom diet is easilycontrolled by parents (Olson, 2005).
Such a strict regimen is difficult to implement and maintainand requires close supervision by a dietician and physician. Inaddition to the practical complexities, concerns also exist aboutthe long-term effects of the diet on the child’s growth andoverall health. For these reasons, the ketogenic diet is restrictedto a small group of young patients with pharmacoresistant
Discussion
123
epilepsy and is not usually used long (Pati and Alexopoulos,2010).
Epilepsy surgery can be classified as curative or palliative,depending on the goal. Curative procedures include lobectomy,lesionectomy, and multilobar or hemispheric surgery(hemispherectomy). Palliative procedures, in contrast tocurative ones, rarely eliminate seizures entirely. It is importantto determine that patients are not candidates for a curativeresective procedure before considering palliative surgicaloptions such as corpus callosotomy, multiple subpialtransections, or vagus nerve stimulation (Pati and Alexopoulos,2010).
Operative mortality rates vary from almost zero fortemporal lobe surgery to 2.5% for hemispherectomy. Thereported risk of permanent surgical morbidity varies by type ofsurgery from 1.1% for temporal lobe resection to about 5% forfrontal lobe resection (Chapell et al., 2003).
Polymers, electrical brain stimulation and prediction ofseizures may be available in the future for treating patients withrefractory epilepsy. Cell transplantation and gene therapy,although holding great promise, are still far from routineclinical use (Nilsen and Cock, 2004).
Summary
124
SUMMARY
In many patients with epilepsy, seizures are well-controlled with currently available anti-epileptic drugs , but asubstantial proportion of patients continue to have seizuresdespite carefully optimized drug treatment.
Different definitions exist for drug resistant epilepsy.Definitions usually include number of anti-epileptic drugfailures and seizure frequency in a specified duration oftherapy.
Risk factors for refractoriness include early onset ofseizures, high seizure frequency and certain structuralabnormalities such as cortical dysplasia.
Refractory epilepsy is established when there isinadequate seizure control despite using potentially effectiveanti-epileptic drugs at tolerable levels for 1-2 years, andexcluding non-epileptic events and poor compliance.
Two concepts have been put forward to explain thedevelopment of pharmacoresistance. The transporter hypothesiscontends that the expression of multidrug transporters in thebrain is augmented, leading to impaired access of anti-epilepticdrugs to CNS targets. The target hypothesis states that
Summary
125
epilepsy-related changes in the properties of the drug targetsthemselves may result in reduced drug sensitivity.
According to transporter hypothesis, multidrugtransporters are over-expressed in epileptogenic brain tissue andthese transporters include P-gp and MRPs. This limits anti-epileptic drug access to drug targets.
Regarding target hypothesis, drug targets are modifiedstructurally and/or functionally and this makes them lesssensitive to anti-epileptic drugs . Reduced sensitivity of drugtargets to anti-epileptic drugs has been suggested for thevoltage-gated Na+ channel and the GABAA receptor.
Certain structural abnormalities of the brain have beenassociated with drug resistant epilepsy. Of these abnormalities,malformations of cortical development have a prominent roleand are often associated with severe epilepsy and about 40% ofdrug-resistant epilepsies are caused by malformations ofcortical development. These malformations of corticaldevelopment include focal cortical dysplasia, periventricularnodular heterotopias, tuberous sclerosis complex,gangliogliomas and dysembryoplastic neuroepithelial tumors.
Surgery is considered when epilepsy is refractory tomedical treatment. Surgery ranges from curative to palliativedepending on the epilepsy syndrome and etiology. Curative
Summary
126
procedures include lobectomy, lesionectomy, andhemispherectomy. Palliative procedures include corpuscallosotomy, multiple subpial transections and vagus nervestimulation.
Recommendations
127
RECOMMENDATIONS
Future directions of research and experimental studiesshould address revealing more data about the following
Evidence that the multidrug transporters regulateintraparenchymal concentrations of anti-epileptic drugs .
Evidence that multidrug transporter expression and/ortransporter function is upregulated in human andexperimental epilepsy.
Regarding drug targets, evidence should be available thatdrug targets are less sensitive to a given anti-epilepticdrug in chronic epilepsy.
Evidence that genetic or pharmacological manipulationof drug transporters/drug targets affects sensitivity toanti-epileptic drugs .
In addition, data on human epilepsy patients should beobtained regarding Association of polymorphisms indrug transporter/drug target genes with clinicalpharmacoresistance.
References
128
REFERENCES
Aghakhani Y, Kinay D, Gotman J, Soualmi L, AndermannF, Olivier A and Dubeau F (2005): The role ofperiventricular nodular heterotopias in epileptogenesis.Brain; 128:641-651.
Alexopoulos AV and Najm IM: Neurosurgical management offocal epilepsies in adults. In: Panayiotopoulos CP(Editor-in-chief): Focal Epilepsies: Seizures, Syndromesand Management. Oxford, Medicinae; 2009:204-220.
Alonso-Nanclares L, Garbelli R, Sola RG, Pastor J, Tassi L,Spreafico R and DeFelipe J (2005): Microanatomy ofthe dysplastic neocortex from epileptic patients. Brain;128:158-173.
Ameyaw MM, Regateiro F, Li T, Liu X, Tariq M, MobarekA, Thornton N, Folayan GO, Githang'a J, Indalo A,Ofori-Adjei D, Price-Evans DA and McLeod HL(2001): MDR1 pharmacogenetics: Frequency of theC3435T mutation in exon 26 is significantly influencedby ethnicity. Pharmacogenetics; 11:7-21.
Anderson GD and Shen DD (2007): Where is the evidencethat p-glycoprotein limits brain uptake of antiepilepticdrug and contributes to drug resistance in epilepsy?Epilepsia; 48: 2372-2374.
Andre VM, Flores-Hernandez J, Cepeda C, Starling AJ,Nguyen S, Lobo MK, Vinters HV, Levine MS andMathern GW (2004): NMDA receptor alterations inneurons from pediatric cortical dysplasia tissue. CerebralCortex; 14:634-646.
Andre VM, Nanping W, Yamazaki I, Nguyen ST, FisherRS, Vinters HV, Mathern GW, Levine MS andCepeda C (2007): Cytomegalic interneurons: A newabnormal cell type in severe pediatric cortical dysplasia. JNeuropathol Exp Neurol; 66:491-504.
Aronica E, Boer K, Redeker S, van Rijen PC, Troost D andGorter JA (2007a): Differential expression patterns ofChloride transporters, NKCC1 and KCC2, in epilepsy-associated malformations of cortical development.Neuroscience; 145:185-196.
References
130
Aronica E, Gorter JA, Jansen GH, van Veelen CW, vanRijen PC, Leenstra S, Ramkema M, Scheffer GL,Scheper RJ and Troost D (2003b): Expression andcellular distribution of multidrug transporter proteins intwo major causes of medically intractable epilepsy: focalcortical dysplasia and glioneuronal tumors.Neuroscience; 118:417-429.
Aronica E, Gorter JA, Jansen GH, Van Veelen CW, VanRijen PC, Ramkema M, Troost D (2003a): Expressionand cell distribution of group I and group II metabotropicglutamate receptor subtypes in taylor-type focal corticaldysplasia. Epilepsia; 44:785-795.
Aronica E, Gorter JA, Ramkema M, Redeker S, Ozbas-Gerçeker F, van Vliet EA, Scheffer GL, Scheper RJ,van der Valk P, Baayen JC and Troost D (2004):Expression and cellular distribution of multidrugresistance-related proteins in the hippocampus of patientswith mesial temporal lobe epilepsy. Epilepsia; 45:441-451.
Aronica E, Leenstra S, van Veelen CW, van Rijen PC,Hulsebos TJ, Tersmette AC, Yankaya B and Troost D(2001a): Glioneuronal tumors and medically intractableepilepsy: A clinical study with long-term follow-up ofseizure outcome after surgery. Epilepsy Res; 43:179-191.
Aronica E, Redeker S, Boer K, Spliet WGM, van Rijen PC,Gorter JA and Troost D (2007b): Inhibitory networksin epilepsy-associated gangliogliomas and in theperilesional epileptic cortex. Epilepsy Res; 74:33-44.
Aronica E, Yankaya B, Jansen GH, Leenstra S, van VeelenCWM, Gorter JA and Troost D (2001b): Ionotropicand metabotropic glutamate receptor protein expressionin glioneuronal tumours from patients with intractableepilepsy. Neuropathol Appl Neurobiol; 27:223-237.
Asadi-Pooya AA, Sharan A, Nei M and Sperling MR (2008):Corpus callosotomy. Epilepsy Behav; 13:271-278.
Avendano C and Menendez JC (2002): Inhibitors ofmultidrug resistance to antitumor agents (MDR). CurrMed Chem; 9:159-193.
Awasthi S, Hallene KL, Fazio V, Singhal SS, Cucullo L,Awasthi YC, Dini G and Janigro D (2005): RLIP76, Anon-ABC transporter, and Drug Resistance in Epilepsy.BMC Neuroscience; 6:61.
Barkovich AJ and Kuzniecky RI (2000): Gray matterheterotopia. Neurology; 55:1603-1608
References
132
Barkovich AJ, Kuzniecky RI, Jackson GD, Guerrini R andDobyns WB (2001): Classification system formalformations of cortical development: update 2001.Neurology; 57: 2168-2178.
Barkovich AJ (2000): Morphologic characteristics ofsubcortical heterotopia: MR imaging study. Am JNeuroradiol; 21:290-295.
Bast T, Oezkan O, Rona S, Stippich C, Seitz A, Rupp A,Fauser S, Zentner J, Rating D and Scherg M (2004):EEG and MEG source analysis of single and averagedinterictal spikes reveals intrinsic epileptogenicity in focalcortical dysplasia. Epilepsia; 45:621-631.
Bast T, Ramantani G, Seitz A and Rating D (2006): Focalcortical dysplasia: Prevalence, clinical presentation andepilepsy in children and adults. Acta Neurol Scand;113:72-81.
Battaglia G, Chiapparini L, Franceschetti S, Freri E, TassiL, Bassanini S, Villani F, Spreafico R, D'Incerti L andGranata T (2006): Periventricular nodular heterotopia:Classification, epileptic history, and genesis of epilepticdischarges. Epilepsia; 47:86-97.
Bauer B, Hartz AM, Pekcec A, Toellner K, Miller DS andPotschka H (2008): Seizure-induced up-regulation of p-glycoprotein at the blood-brain barrier through glutamateand cyclooxygenase-2 signaling. Mol Pharmacol;73:1444-1453.
Bautista JF, Foldvary-Schaefer N, Bingaman WE andLuders HO (2003): Focal cortical dysplasia andintractable epilepsy in adults: Clinical, EEG, imaging,and surgical features. Epilepsy Res; 55:131-136.
Beck H (2007): Plasticity of Antiepileptic Drug Targets.Epilepsia; 48(S1):14-18.
Beleza P (2009): Refractory Epilepsy: A Clinically OrientedReview. Eur Neurol; 62:65-71.
Berg AT, Shinnar S, Levy SR, Testa FM, Smith-Rapaport Sand Beckerman B (2001): Early development ofintractable epilepsy in children: A prospective study.Neurology; 56:1445-1452.
Berg AT, Vickrey BG, Testa FM, Levy SR, Shinnar S,DiMario F and Smith S (2006): How long does it takefor epilepsy to become intractable? A prospectiveinvestigation. Ann Neurol; 60:73-79.
Berg AT (2004): Understanding the delay before epilepsysurgery: Who develops intractable focal epilepsy andwhen? CNS Spectr; 9:136-144.
Blumcke I and Wiestler OD (2002): Gangliogliomas: Anintriguing tumor entity associated with focal epilepsies. JNeuropathol Exp Neurol; 61:575-584.
Blumcke I, Vinters HV, Armstrong D, Aronica E, Thom Mand Spreafico R (2009): Malformations of corticaldevelopment and epilepsies: Neuropathological findingswith emphasis on focal cortical dysplasia. EpilepticDisord; 11:181-193
Borst P, Evers R, Kool M and Wijnholds J (2000): A familyof drug transporters: The multidrug resistance-associatedproteins. J Natl Cancer Inst; 92:1295-1302.
Chapell R, Reston J, Snyder D, Treadwell J, Treager S andTurkelson C (2003): Management of treatment-resistantepilepsy. Evid Rep Technol Assess; 77:1-8.
Cohen AS, Lin DD, Quirk GL and Coulter DA (2003):Dentate granule cell GABAA receptors in epileptichippocampus: Enhanced synaptic efficacy and alteredpharmacology. Eur J Neurosci; 17:1607-1616.
Cohen I, Navarro V, Clemenceau S, Baulac M and Miles R(2002): On the origin of interictal activity in humantemporal lobe epilepsy in vitro. Science; 298:1418-1421.
Coulter DA (2000): Mossy fiber zinc and temporal lobeepilepsy: Pathological association with altered “epileptic”gamma-aminobutyric acid A receptors in dentate granulecells. Epilepsia (Abstract); 41(S6):96-99.
Crino PB, Duhaime AC, Baltuch G and White R (2001):Differential expression of glutamate and GABAA
receptor subunit mRNA in cortical dysplasia. Neurology;56:906-913.
References
136
Dabora SL, Jozwiak S, Franz DN, Roberts PS, Nieto A,Chung J, Choy YS, Reeve MP, Thiele E, Egelhoff JC,Kasprzyk-Obara J, Domanska-Pakiela D andKwiatkowski DJ (2001): Mutational analysis in a cohortof 224 tuberous sclerosis patients indicates increasedseverity of TSC2, compared with TSC1, disease inmultiple organs. Am J Hum Genet; 68:64-80.
Dean M, Rzhetsky A and Allikmets R (2001): The humanATP-binding cassette (ABC) transporter superfamily.Genome Res; 11:1156-1166.
Decleves X, Regina A, Laplanche JL, Roux F, Boval B,Launay JM and Scherrmann JM (2000): Functionalexpression of p-glycoprotein and multidrug resistanceassociated protein (MRP1) in primary cultures of ratastrocytes. J Neurosci Res; 60:594-601.
Diehl B, Najm I, LaPresto E, Prayson R, Ruggieri P,Mohamed A, Ying Z, Lieber M, Babb T, Bingaman Wand Luders HO (2004): Temporal lobe volumes inpatients with hippocampal sclerosis with or withoutcortical dysplasia. Neurology; 62:1729-1735.
Dombrowski SM, Desai SY, Marroni M, Cucullo L,Goodrich K, Bingaman W, Mayberg MR, Bengez Land Janigro D (2001): Over-expression of multiple drugresistance genes in endothelial cells from patients withrefractory epilepsy. Epilepsia; 42:1501-1506.
D'Orsi G, Tinuper P, Bisulli F, Zaniboni A, Bernardi B,Rubboli G, Riva R, Michelucci R, Volpi L, TassinariCA and Baruzzi A (2004): Clinical features and longterm outcome of epilepsy in periventricular nodularheterotopia. Simple compared with plus forms. J NeurolNeurosurg Psychiatry; 75:873-878.
Fauser S, Schulze-Bonhage A, Honegger J, Carmona H,Huppertz HJ, Pantazis G, Rona S, Bast T, Strobl K,Steinhoff BJ, Korinthenberg R, Rating D, Volk B andZentner J (2004): Focal cortical dysplasias: Surgicaloutcome in 67 patients in relation to histological subtypesand dual pathology. Brain; 127:2406-2418.
Ferrier CH, Aronica E, Leijten FS, Spliet WG, van HuffelenAC, van Rijen PC and Binnie CD (2006):Electrocorticographic discharge patterns in glioneuronaltumors and focal cortical dysplasia. Epilepsia; 47:1477-1486.
Francione S, Vigliano P, Tassi L, Cardinale F, Mai R, LoRusso G and Munari C (2003): Surgery for drugresistant partial epilepsy in children with focal corticaldysplasia: Anatomical-clinical correlations andneurophysiological data in 10 patients. J NeurolNeurosurg Psychiatry; 74:1493-1501.
Fujiwara T and Shigematsu H (2004): Etiologic factors andclinical features of symptomatic epilepsy: Focus onpediatric cases. Psychiatry Clin Neurosci; 58:9-12.
Gomez-Anson B, Thom M, Moran N, Stevens J andScaravili F (2000): Imaging and radiographic-pathological correlation in histologically-proven cases offocal cortical dysplasia and other glial and neuronoglialmalformative lesions in adults. Neuroradiology; 42:157-167.
Gonzalez-Martinez JA, Gupta A, Kotagal P, Lachhwani D,Wyllie E, Luders HO and Bingaman WE (2005):Hemispherectomy for catastrophic epilepsy in infants.Epilepsia; 46:1518-1525.
Guerrini R, Sicca F and Parmeggiani L (2003): Epilepsy andmalformations of the cerebral cortex. Epileptic Disord;5(S2):9-26.
Gulledge AT and Stuart GJ (2003): Excitatory actions ofGABA in the cortex. Neuron; 37:299-309.
Hart MG, Grant R, Garside R, Rogers G, Somerville M andStein K (2008): Chemotherapeutic wafers for high gradeglioma. Cochrane Database Syst Rev; CD007294.
Haydon P (2003): Glia: Listening and talking to the synapse.Nat Review Neurosci; 2:185-193.
Hitiris N, Mohanraj R, Norrie J, Sills GJ and Brodie MJ(2007): Predictors of pharmacoresistant epilepsy.Epilepsy Res; 75: 192-196.
Hoffmeyer S, Burk O, von Richter O, Arnold HP,Brockmoller J, Johne A, Cascorbi I, Gerloff T, RootsI, Eichelbaum M and Brinkmann U (2000):Functional polymorphisms of the human multidrug-resistance gene: Multiple sequence variations andcorrelation of one allele with p-glycoprotein expressionand activity in vivo. Proc Natl Acad Sci USA; 97:3473-3478.
Holmes MD, Silbergeld DL, Drouhard D, Wilensky AJ andOjemann LM (2004): Effect of vagus nerve stimulationon adults with pharmacoresistant generalized epilepsysyndromes. Seizure; 13:340-345.
Kerb R, Hoffmeyer S and Brinkmann U (2001): ABC drugtransporters: Hereditary polymorphisms andpharmacological impact in MDR1, MRP1 and MRP2.Pharmacogenomics; 2:51-64.
Ko TS and Holmes GL (1999): EEG and clinical predictors ofmedically intractable childhood epilepsy. ClinNeurophysiol; 110:1245-1251.
Koh S, Jayakar P, Dunoyer C, Whiting SE, Resnick TJ,Alvarez LA, Morrison G, Ragheb J, Prats A, Dean P,Gilman J and Duchowny MS (2000): Epilepsy surgeryin children with tuberous sclerosis complex. Epilepsia;41:1206-1213.
Krsek P, Maton B, Korman B, Pacheco-Jacome P, JayakarP, Dunoyer C, Rey G, Morrison G, Ragheb J, VintersHV, Resnick T and Duchowny M (2008): DifferentFeatures of Histopathological Subtypes of Pediatric FocalCortical Dysplasia. Ann Neurol; 63:758-769.
Kuzniecky RI and Barkovich AJ (2001): Malformations ofcortical development and epilepsy. Brain Dev; 23:2-11.
Kwan P and Brodie MJ (2000): Early identification ofrefractory epilepsy. NEJM; 342:314-319.
Kwan P and Brodie MJ (2005): Potential Role of DrugTransporters in the Pathogenesis of Medically IntractableEpilepsy. Epilepsia; 46:224-235.
References
142
Kwan P and Brodie MJ (2006): Refractory epilepsy:Mechanisms and solutions. Expert Rev Neurother; 6:397-406.
Kwan P and Sander JW (2004): The natural history ofepilepsy: An epidemiological view. J Neurol NeurosurgPsychiatry; 75:1376-1381.
Lawson JA, Birchansky S, Pacheco E, Jayakar P, ResnickTJ, Dean P and Duchowny MS (2005): Distinctclinicopathologic subtypes of cortical dysplasia ofTaylor. Neurology; 64:55-61.
Lazarowski A, Czornyj L, Lubienieki F, Girardi E,Vazquez S and D’Giano C (2007): ABC Transportersduring Epilepsy and Mechanisms Underlying MultidrugResistance in Refractory Epilepsy. Epilepsia;48(S5):140-149
Lazarowski A, Lubieniecki F, Camarero S, Pomata H,Bartuluchi M, Sevlever G and Taratuto AL (2004):Multidrug resistance proteins in tuberous sclerosis andrefractory epilepsy. Pediatr Neurol; 102:30-36.
Leschziner GD, Andrew T, Leach JP, Chadwick D, CoffeyAJ, Balding DJ, Bentley DR, Pirmohamed M andJohnson MR (2007): Common ABCB1 polymorphismsare not associated with multidrug resistance in epilepsyusing a gene-wide tagging approach. PharmacogenetGenomics; 17:217-220.
Loscher W and Potschka H (2005): Drug resistance in braindiseases and the role of drug efflux transporters. Nat RevNeurosci; 6:591-602.
Loscher W and Potschka H (2002): Role of multidrugtransporters in pharmacoresistance to antiepileptic drugs.The Journal of Pharmacology and ExperimentalTherapeutics; 301:7-14.
Loscher W and Schmidt D (2006): Experimental and clinicalevidence for loss of effect (tolerance) during prolongedtreatment with antiepileptic drugs. Epilepsia; 47:1253-1284.
Loscher W and Schmidt D (1994): Strategies in antiepilepticdrug development: Is rational drug design superior torandom screening and structural variation? Epilepsy Res;17:95-134.
References
144
Loscher W (2007): Drug Transporters in the Epileptic Brain.Epilepsia; 48(S1):8-13
Luciano AL and Shorvon SD (2007): Results of treatmentchanges in patients with apparently drug-resistant chronicepilepsy. Ann Neurol; 62:375-381.
Luders H and Schuele SU (2006): Epilepsy surgery in patientswith malformations of cortical development. Curr OpinNeurol; 19:169-174.
Luyken C, Blumcke I, Fimmers R, Urbach H, Elger CE,Wiestler OD and Schramm J (2003): The spectrum oflong-term epilepsy-associated tumors: Long-term seizureand tumor outcome and neurosurgical aspects. Epilepsia;44:822-830.
Maher CO, White JB, Scheithauer BW and Raffel C (2008):Recurrence of dysembryoplastic neuroepithelial tumorfollowing resection. Pediatr Neurosurg; 44:333-336.
Marchi N, Gonzalez-Martinez J, Nguyen MT, Granata Tand Janigro D (2010): Transporters in drug-refractoryepilepsy: Clinical significance. Clin Pharmacol Ther;87:13-18.
Marusic P, Tomasek M, Krsek P, Krijtova H, Zarubova J,Zamecnik J, Mohapl M, Benes V, Tichy M andKomarek V (2007): Clinical characteristics in patientswith hippocampal sclerosis with or without corticaldysplasia. Epileptic Disord; 9(S1):75-82.
McKeegan KS, Borges-Walmsley MI and Walmsley AR(2003): The structure and function of drug pumps: Anupdate. Trends Microbiol; 11:21-29.
Mitchell LA, Harvey AS, Coleman LT, Mandelstam SA andJackson GD (2003): Anterior temporal changes on MRimages of children with hippocampal sclerosis: An effectof seizures on the immature brain? Am J Neuroradiol,24:1670-1677.
Mody I and Pearce RA (2004): Diversity of inhibitoryneurotransmission through GABAA receptors. TrendsNeurosci, 27:569-575.
Neal EG, Chaffe H, Schwartz RH, Lawson MS, Edwards N,Fitzsimmons G, Whitney A and Cross JH (2008): Theketogenic diet for the treatment of childhood epilepsy: Arandomized controlled trial. Lancet Neurol, 7:500-506.
Pati S and Alexopoulos AV (2010): Pharmacoresistantepilepsy: From pathogenesis to current and emergingtherapies. Cleveland Clinic Journal of Medicine; 77:457-467.
Patt S, Steenbeck J, Hochstetter A, Kraft R, Huonker R,Haueisen J, Haberland N, Ebmeier K, Hliscs R,Fiehler J, Nowak H and Kalff R (2000): Sourcelocalization and possible causes of interictal epilepticactivity in tumor-associated epilepsy. Neurobiol Dis;7:260-269.
Porter BE, Judkins AR, Clancy RR, Duhaime A, Dlugos DJand Golden JA (2003): Dysplasia: A common finding inintractable pediatric temporal lobe epilepsy. Neurology;61:365-368.
References
147
Potschka H and Loscher W (2001): In vivo evidence for p-glycoprotein-mediated transport of phenytoin at theblood-brain barrier of rats. Epilepsia; 42:1231-1240.
Potschka H, Fedrowitz M and Loscher W (2001): P-glycoprotein and multidrug resistance-associated proteinare involved in the regulation of extracellular levels ofthe major antiepileptic drug carbamazepine in the brain.Neuroreport; 12:3557-3560.
Povey S, Burley MW, Attwood J, Benham F, Hunt D,Jeremiah S J, Franklin D, Gillett G, Malas S, RobsonEB, Tippett P, Edwards J H, Kwiatkowski DJ, SuperM, Mueller R, Fryer A, Clarke A, Webb D andOsborne J (1994): Two loci for tuberous sclerosis: oneon 9q34 and one on 16p13. Ann Hum Genet; 58:107-127.
Prayson RA, Estes ML and Morris HH (1993): Coexistenceof neoplasia and cortical dysplasia in patients presentingwith seizures. Epilepsia; 34:609-615.
Prayson RA: Pathology of neoplastic lesions causing epilepsy.In: Luders HO and Comair YG (Editors): Epilepsysurgery. Second edition. Philadelphia, LippincottWilliams and Wilkins; 2001: 915-926.
References
148
Ravizza T, Boer K, Redeker S, Spliet WG, van Rijen PC,Troost D, Vezzani A and Aronica E (2006): The IL-1beta system in epilepsy-associated malformations ofcortical development. Neurobiol Dis; 24:128-143.
Remy S and Beck H (2006): Molecular and CellularMechanisms of Pharmacoresistance in Epilepsy. Brain,129:18-35.
Remy S, Gabriel S, Urban BW, Dietrich D, Lehmann TN,Elger CE, Heinemann U and Beck H (2003a): A novelmechanism underlying drug resistance in chronicepilepsy. Ann Neurol, 53:469-479.
Remy S, Urban BW, Elger CE and Beck H (2003b):Anticonvulsant pharmacology of voltage-gated Na+channels in hippocampal neurons of control andchronically epileptic rats. Eur J Neurosci; 17:2648-2658.
Ridler K, Bullmore ET, DeVries PJ, Suckling J, Barker GJ,Meara SJ, Williams SC and Bolton PF (2001):Widespread anatomical abnormalities of grey and whitematter structure in tuberous sclerosis. Psychol Med;31:1437-1441.
Rizzi M, Caccia S, Guiso G, Richichi C, Gorter JA, AronicaE, Aliprandi M, Bagnati R, Fanelli R, D'Incalci M,Samanin R and Vezzani A (2002): Limbic seizuresinduce p-glycoprotein in rodent brain: Functionalimplications for pharmacoresistance. J Neurosci;22:5833-5839.
Robey RW, Lazarowski A and Bates SE (2008): P-glycoprotein: A clinical target in drug-refractoryepilepsy? Mol Pharmacol; 73:1343-1346.
Rosenow F and Luders H (2001): Presurgical evaluation ofepilepsy. Brain; 124:1683-1700.
Roske B, Stoltenburg G, Baier PM, Konig R and Schlote W(2003): Tuberous sclerosis complex with disseminatedtelencephalic distribution of atypical cells and theirrelation to corticogenesis. Clin Neuropathol; 22:119-128.
Samadani U, Judkins A, Alkpalu A, Aronica E and CrinoPB (2007): Differential gene expression inganglioglioma. Epilepsia; 48:646-653.
Schaub C, Uebachs M and Beck H (2007): DiminishedResponse of CA1 Neurons to Antiepileptic Drugs inChronic Epilepsy. Epilepsia; 48:1339-1350.
Schinkel AH (1999): P-glycoprotein, a gatekeeper in theblood-brain barrier. Adv Drug Deliv Rev; 36:179-194.
Schmidt D and Loscher W (2005): Drug resistance inepilepsy: Putative neurobiologic and clinicalmechanisms. Epilepsia; 46:858-877.
Schmidt D (2009): Drug treatment of epilepsy: Options andlimitations. Epilepsy & Behavior; 15:56-65
Schuetz EG, Beck WT and Schuetz JD (1996): Modulatorsand substrates for p-glycoprotein and cytochrome P4503A coordinately up-regulate these proteins in humancolon carcinoma cells. Mol Pharmacol; 49:311-318.
Seelig A, Li Blatter X and Wohnsland F (2000): Substraterecognition by p-glycoprotein and the multidrugresistance-associated protein MRP1: A comparison. Int JClin Pharm Ther; 38:111-121.
Shen DD: Valproate: absorption, distribution, and excretion. In:Loscher W (Editor): Valproate. Basel, Birkhauser;1999:77-90,
References
151
Siegel AM (2004): Presurgical evaluation and surgicaltreatment of medically refractory epilepsy. NeurosurgRev; 27:1-18.
Sills GJ and Kwan P (2001): P-glycoprotein-mediatedantiepileptic drug transport: A role in refractory epilepsy?Epilepsia; 42(S7):83.
Silverman JA (1999): Multidrug-resistance transporters.Pharmaceut Biotech; 12:353-386.
Sisodiya SM, Lin WR, Harding BN, Squier MV, Keir G andThom M (2002): Drug resistance in epilepsy: Expressionof drug resistance proteins in common causes ofrefractory epilepsy. Brain; 125:22-31.
Sisodiya SM, Lin WR, Squier MV and Thom M (2001):Multidrug-resistance protein 1 in focal cortical dysplasia.Lancet; 357:42-43.
Spector R (2000): Drug transport in the mammalian centralnervous system: Multiple complex systems. A criticalanalysis and commentary. Pharmacology; 60:58-73.
Spencer SS, Schramm J, Wyler A, O'Connor M, Orbach D,Krauss G, Sperling M, Devinsky O, Elger C, Lesser R,Mulligan L and Westerveld M (2002): Multiplesubpial transection for intractable partial epilepsy: Aninternational meta-analysis. Epilepsia; 43:141-145.
Starreveld E and Guberman A (2006): The spectrum ofintractable epilepsy: A proposed scale. Adv Neurol;97:493-496.
Tanriverdi T, Olivier A, Poulin N, Andermann F andDubeau F (2009): Long-term seizure outcome aftercorpus callosotomy: A retrospective analysis of 95patients. J Neurosurg; 110:332-342.
Tassi L, Colombo N, Cossu M, Mai R, Francione S, LoRusso G, Galli C, Bramerio M, Battaglia G, GarbelliR, Meroni A and Spreafico R (2005): Electroclinical,MRI and neuropathological study of 10 patients withnodular heterotopia, with surgical outcomes. Brain;128:321-337
Tassi L, Colombo N, Garbelli R, Francione S, Lo Russo G,Mai R, Cardinale F, Cossu M, Ferrario A, Galli C,Bramerio M, Citterio A and Spreafico R (2002): Focalcortical dysplasia: Neuropathological subtypes, EEG,neuroimaging and surgical outcome. Brain; 125:1719-1732.
Tate SK, Depondt C, Sisodiya SM, Cavalleri GL, Schorge S,Soranzo N, Thom M, Sen A, Shorvon SD, Sander JW,Wood NW and Goldstein DB (2005): Geneticpredictors of the maximum doses patients receive duringclinical use of the anti-epileptic drugs carbamazepine andphenytoin. Proceedings of the National Academy ofSciences; 102:5507-5512.
Tishler DM, Weinberg KT, Hinton DR, Barbaro N, AnnettGM and Raffel C (1995): MDR1 gene expression inbrain of patients with medically intractable epilepsy.Epilepsia; 36:1-6.
Van Bakel I, Sepp T, Ward S, Yates JRW and Green AJ(1997): Mutations in the TSC2 gene: Analysis of thecomplete coding sequence using the protein truncationtest [PTT]. Hum Molec Genet; 6:1409-1414.
Van Breemen MS, Wilms EB and Vecht CJ (2007): Epilepsy inpatients with brain tumours: Epidemiology, mechanisms,and management. Lancet Neurology; 6:421-430.
Wang Y, Greenwood JSF, Calcagnotto ME, Kirsch HE,Barbaro NM and Baraban SC (2007): Neocorticalhyperexcitability in a human case of tuberous sclerosiscomplex and mice lacking neuronal expression of TSC1.Ann Neurol; 61:139-152.
References
154
Weiner HL, Carlson C, Ridgway EB, Zaroff CM, Miles D,LaJoie J and Devinsky O (2006): Epilepsy surgery inyoung children with tuberous sclerosis: Results of a novelapproach. Pediatrics; 117:1494-1502.
White R, Hua Y, Scheithauer B, Lynch DR, Henske EP andCrino PB (2001): Selective alterations in glutamate andGABA receptor subunit mRNA expression in dysplasticneurons and giant cells of cortical tubers. Ann Neurol;49:67-78.
Wong M (2008): Mechanisms of epileptogenesis in tuberoussclerosis complex and related malformations of corticaldevelopment with abnormal glioneuronal proliferation.Epilepsia; 49:8-21.
Zentner J, Hufnagel A, Wolf HK, Ostertun B, Behrens E,Campos MG, Elger CE, Wiestler OD and Schramm J(1997): Surgical treatment of neoplasms associated withmedically intractable epilepsy. Neurosurgery; 41:378-386.
Zhang L, Ong WY and Lee T (1999): Induction of p-glycoprotein expression in astrocytes followingintracerebroventricular kainate injections. Exp Brain Res;126:509-516.
