REVIEW ARTICLE
`
`CNS Drugs 2008; 22 (9): 739-760
`1172-7047/08/0009-0739/$48.00/0
`
`© 2008 Adis Data Information BV. All rights reserved.
`
`CNS Adverse Events Associated with
`Antiepileptic Drugs
`Gina M. Kennedy and Samden D. Lhatoo
`Department of Neurology, Institute of Clinical Neurosciences, Frenchay Hospital,
`Bristol, England
`
`Contents
`
`Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 739
`1. Comparisons between Conventional and Newer Antiepileptic Drugs (AEDs) . . . . . . . . . . . . . . . . . . . 740
`2. Conventional AEDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 743
`2.1 Benzodiazepines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 744
`2.2 Carbamazepine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 744
`2.3 Ethosuximide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 745
`2.4 Phenobarbital (Phenobarbitone) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 746
`2.5 Phenytoin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 746
`2.6 Primidone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 747
`2.7 Valproate (Valproic Acid) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 747
`3. Newer AEDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 748
`3.1
`Felbamate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 748
`3.2 Gabapentin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 748
`3.3
`Lamotrigine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 749
`3.4
`Levetiracetam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 750
`3.5 Oxcarbazepine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 751
`3.6
`Pregabalin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 752
`3.7
`Tiagabine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 752
`3.8
`Topiramate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 752
`3.9
`Vigabatrin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 753
`3.10 Zonisamide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 754
`4. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 754
`
`Abstract
`
`A variety of newer antiepileptic drugs (AEDs) are now available for treating
`patients with epilepsy in addition to the ‘conventional’ drugs that have been
`available throughout a large part of the last century. Since these drugs act to
`suppress the pathological neuronal hyperexcitability that constitutes the final
`substrate in many seizure disorders, it is not surprising that they are prone to
`causing adverse reactions that affect the CNS.
`Information on adverse effects of the older AEDs has been mainly observation-
`al. Equally, whilst the newer drugs have been more systematically studied, their
`long-term adverse effects are not clearly known. This is illustrated by the
`relatively late emergence of the knowledge of visual field constriction in the case
`of vigabatrin, which only became known after several hundred thousand patient-
`years of use. However, older drugs continue to be studied and there has been more
`
`Page 1 of 22
`
`JAZZ EXHIBIT 2012
`Ranbaxy Inc. (Petitioner) v. Jazz Pharms. Ireland Ltd. (Patent Owner)
`Case IPR2016-00024
`
`

`
`740
`
`Kennedy & Lhatoo
`
`recent comment on the possible effect of valproate (valproic acid) on cognition
`following exposure to this drug in utero.
`With most AEDs, there are mainly dose-related adverse effects that could be
`considered generic, such as sedation, drowsiness, incoordination, nausea and
`fatigue. Careful dose titration with small initial doses can reduce the likelihood of
`these adverse effects occurring. Adverse effects such as paraesthesiae are more
`commonly reported with drugs such as topiramate and zonisamide that have
`carbonic anhydrase activity. Weight loss and anorexia can also be peculiar to
`these drugs. Neuropsychiatric adverse effects are reported with a variety of AEDs
`and may not be dose related. Some drugs, such as carbamazepine when used to
`treat primary generalized epilepsy, can exacerbate certain seizure types. Rare
`adverse effects such as hyperammonaemia with valproate are drug specific. There
`are relatively very few head-to-head comparisons of AEDs and limited informa-
`tion is available in this regard.
`In this review, we discuss the available literature and provide a comprehensive
`summary of adverse drug reactions of AEDs affecting the CNS.
`
`tion before each drug is further analysed individual-
`The last two decades have seen an exponential
`increase in antiepileptic drug (AED) development.
`ly in sections 2 and 3.
`The pharmaceutical ideal of an efficacious drug with
`The recently completed SANAD study, a ran-
`a minimum of adverse effects remains a relative
`domized, unblinded, controlled trial, published its
`concept and epilepsy treatment strategies usually
`findings in two papers, one examining the use of
`balance pursuit of seizure freedom with an accept-
`carbamazepine, gabapentin,
`lamotrigine, oxcar-
`able threshold for tolerating adverse effects. Despite
`bazepine and topiramate in the treatment of partial
`this, there is a surprising paucity of recorded adverse
`epilepsy (Arm A),[2] the other examining the use of
`drug reactions (ADRs) in a standardized format.[1]
`valproate (valproic acid), lamotrigine and topira-
`ADRs can occur in a variety of organ systems and mate in generalized and unclassifiable epilepsy
`often involve multiple systems. This review concen-
`(Arm B).[3] 1721 and 716 patients, respectively,
`trates on the effects of AEDs on the CNS and were recruited into the two arms. Although this
`provides a review of the available literature on this
`study was primarily designed to examine drug effec-
`subject. Drug efficacy is not discussed. The litera-
`tiveness and not to specifically compare ADRs with
`ture was reviewed using an internet-based PubMed
`different drug treatments, one of the primary out-
`search, using search terms that included the names
`comes was time to treatment failure. Treatment fail-
`of individual drugs and the terms ‘side effects’,
`ure resulting in drug withdrawal is usually a conse-
`‘adverse reactions’ and ‘central nervous system’.
`quence of adverse effects when it occurs early in the
`Individual study articles and review articles were
`course of treatment, and a consequence of poor
`further cross-referenced to widen the search.
`efficacy when it occurs late; therefore, this study
`provides at least some systematic and comparative
`information on ADRs with a limited number of
`AEDs. The clinical setting was generally one of
`patients being initiated on drug therapy and, there-
`fore, one in which patients were receiving mono-
`therapy. The median number of days (25th–75th
`centiles) to treatment failure in Arm A was 84
`(26–215) for unacceptable adverse events and 313
`
`To help gain an understanding of the comparative
`effectiveness and tolerability of both conventional
`and newer AEDs, an overview of the most current
`study, the SANAD (UK Standard And New An-
`tiepileptic Drugs) study,[2-4] is provided in this sec-
`
`1. Comparisons between Conventional
`and Newer Antiepileptic Drugs (AEDs)
`
`© 2008 Adis Data Information BV. All rights reserved.
`
`CNS Drugs 2008; 22 (9)
`
`Page 2 of 22
`
`

`
`CNS Adverse Events Associated with Antiepileptic Drugs
`
`(152-642) for inadequate seizure control. Car'barna-
`zepine (n = 102; 27%) and topiramate (n = 101;
`27%) therapy were associated with the greatest
`numbers of patients reporting unacceptable adverse
`events compared with gabapentin (n = 57; 15.2%),
`lamotrigine (n = 60; 15.9%) and oxcarbazepine
`(n = 49; 23.3%). However, treatment with car'barna-
`zepine (n = 43; 11.4%) and topiramate (rt = 55;
`14.7%) was also associated with the lowest number
`of patients discontinuing treatment due to inade-
`quate seizure control, i.e. these AEDs can be inter-
`preted as being more effective than lamotrigine
`(n = 60; 15.9%) and gabapentin (n = 99; 26.3%).
`When unacceptable adverse events and inadequate
`seizure control were considered as a combined rea-
`
`son for treatment termination, there were a greater
`number of patients discontinuing treatment with
`topiramate (n = 28; 7.4%), gabapentin (n = 32;
`8.5%) or lamotrigine (rt = 11; 7.9%) than carbama-
`
`zepine (rt = 20; 5.3%) or oxcarbazepine (n = 11;
`5.2%). This highlights the subjective greater impor-
`tance of efficacy compared with adverse effect
`profiles of AEDs for the majority of patients.
`
`Cumulative incidence analysis showed that car-
`barnazepine (lamotrigine : carbamazepine hazard
`ratio [HR; 95% CI] 0.62 [0.46, 0.83]) was most
`frequently associated with treatment failure for un-
`acceptable adverse events, whereas gabapentin was
`least likely to result in this kind of treatment failure
`(gabapentin : carbarnazepine HR [95% CI] 0.60
`[0.44, 0.8l]) [figure 1]. As far as the estimates for
`the proportion of patients with treatment failure
`events were concerned, lamotrigine was 10-11%
`better with respect to treatment withdrawal because
`of adverse events, and statistically different at all
`timepoints between 1 and 6 years. This estimate has
`been criticized as being biased against carbarnaze-
`pine because of faster dose titration and higher dos-
`b
`
`1.0
`
`.°an
`
`.°
`
`0 Topirarnate
`I Carbarnazepine
`0 Lam0trig'ne
`El Gabapentin
`
`1.0
`
`I Carbarmzepine
`0 Topirarnate
`A Oxcarbazepine
`O Larnotriwte
`D Gabepentit
`
`Grays test statistic = 21 .84 (3); p < 0.0001
`Log-rank rem statistic = 20.24 (3); p = 0.0002
`
`Gray's test statistic = 6.70 (4); p = 0.153
`Log-rank test statistic = 5.00 (4); p = 0.279
`
`
`
`
`
`Cumulativelncldenceprobability
`
`Topiunate
`
`1.04 (0.75, 1.44)
`1.66 (124, 224)
`1.36 (0.90, 2.05)
`
`1.60 (120, 2.15)
`1.21 (0.a1,1.a1)
`
`Fig. 1. Cumulative incidence oi unacceptable adverse events 0! antiepileptic dnrgs tor the entire treatment period (a) and alter June 2001
`(b). Data in the table below the figures are hazard ratios (HRS, 95% Cls), where HR >1 indicat that treatment lailure occurs more rapidy
`on ttug cormared with baseline (reproduced irorn Maison et al.,l" with permission from Elsevier).
`
`0ZIBN;IsDutulrI1‘OrlncIlbtIBV.AIllgt‘lsresaved.
`
`Page 3 of 22
`
`

`
`48 (36)
`
`14 (8)
`21 (9)
`38 (32)
`20 (12)
`14 (10)
`16 (7)
`17 (5)
`9 (6)
`13 (6)
`12 (4)
`
`9 (6)
`9 (9)
`
`18 (10)
`20 (15)
`13 (4)
`22 (19)
`23 (15)
`17 (9)
`22 (13)
`21 (14)
`19 (11)
`9 (6)
`
`24 (12)
`16 (15)
`
`20 (13)
`21 (13)
`17 (15)
`13 (10)
`15 (9)
`11 (7)
`17 (12)
`15 (9)
`19 (13)
`12 (7)
`
`14 (9)
`8 (4)
`
`7 (5)
`9 (6)
`20 (16)
`13 (8)
`13 (12)
`7 (5)
`3 (1)
`8 (5)
`9 (6)
`2 (1)
`
`8 (6)
`8 (6)
`
`29 (24)
`17 (11)
`17 (8)
`26 (19)
`15 (8)
`37 (31)
`17 (8)
`18 (12)
`16 (12)
`24 (19)
`
`9 (3)
`22 (19)
`
`88 (60)
`88 (54)
`105 (75)
`94 (68)
`80 (54)
`88 (59)
`76 (39)
`71 (46)
`76 (48)
`59 (37)
`
`64 (36)
`63 (53)
`
`742
`
`Kennedy & Lhatoo
`
`Table I. Frequency of clinically important adverse events of antiepileptic drugs in patients with partial epilepsy (reproduced from Marson et
`al.,[2] with permission from Elsevier)
`
`Carbamazepine Gabapentin
`378
`377
`183 (48)
`178 (47)
`
`Lamotrigine
`378
`169 (45)
`
`Oxcarbazepine
`210
`100 (48)
`
`Topiramate
`378
`200 (53)
`
`Total
`1721
`830 (48)
`
`46 (34)
`
`31 (17)
`
`22 (16)
`
`43 (33)
`
`190 (136)
`
`No. of patients/adverse effects
`No. of patients randomized
`Total number (%) of patients
`with at least one adverse event
`Tiredness/drowsiness/fatigue/
`lethargya
`Depressiona
`Headachea
`Allergic rasha
`Memory problemsa
`Dizziness/vertigoa
`Other psychiatrica
`Worsening of seizuresa
`Other neurologicala
`Other generala
`Behaviour/personality change/
`aggressiona
`Ataxiaa
`Confusion/difficulty thinking/
`disorientationa
`Anxiety/agitation/nervousnessa
`52 (41)
`15 (12)
`7 (6)
`8 (5)
`15 (11)
`7 (7)
`Weight lossa
`42 (33)
`29 (27)
`3 (1)
`4 (2)
`4 (2)
`2 (1)
`Diplopiaa
`34 (17)
`6 (3)
`8 (6)
`4 (2)
`11 (4)
`5 (2)
`Nauseaa
`44 (32)
`4 (4)
`15 (13)
`9 (6)
`7 (3)
`9 (6)
`Weight gaina
`34 (24)
`5 (4)
`1 (0)
`4 (1)
`15 (12)
`9 (7)
`Accidental injurya
`41 (20)
`8 (3)
`3 (1)
`12 (8)
`11 (6)
`7 (2)
`Pins and needles/dysaesthesiaea
`38 (27)
`26 (24)
`0 (0)
`3 (1)
`5 (1)
`4 (1)
`Sleep disturbancea
`31 (24)
`9 (8)
`4 (2)
`9 (8)
`4 (4)
`5 (2)
`Other eventsa,b
`480 (316)
`103 (64)
`46 (38)
`110 (70)
`113 (73)
`108 (71)
`a Data presented are number of patients with the adverse event by intention-to-treat analysis (per-protocol analysis inside brackets).
`b Other cardiac or vascular; other skin and appendages; abdominal pain, dyspepsia; other gastrointestinal; other visual disturbance;
`other renal tract or genital; diarrhoea; tremor; aches and pains; constipation; infection; mouth or gum problem; other respiratory or
`pulmonary; ischaemic heart disease or myocardial infarct; other haematological; other musculoskeletal; vomiting; impotence or libido
`problems; alopecia; word-finding difficulty; status epilepticus; stroke-infarction; diabetes mellitus; hearing problem or tinnitus;
`hypertension; anorexia; bruising; flu-like symptoms; haemorrhage; malignancy; shortness of breath; vaginal bleeding; arthritis;
`eczema; peptic ulceration; asthma; other hepatobiliary; urinary retention; abnormal liver function tests; anaemia; childbirth; myalgia;
`other endocrine; psoriasis; upper respiratory tract infection; catarrh; sinusitis; rhinorrhoea; urinary tract infection; faints;
`hallucinations; hepatitis; pancreatitis; psychosis; transient ischaemic attack; tachycardia; thyroid disease; venous thrombosis (sorted
`by descending total frequency).
`
`reported the lowest number of adverse events, while
`age levels in the study population than are seen in
`the topiramate group reported the greatest number of
`routine clinical practice.[4] In addition, the study
`adverse events (table I).
`design did not emphasize the use of slow-release
`carbamazepine, thought to be a factor in avoidance
`In Arm B of the SANAD study, which studied
`of adverse events when compared with usage of mainly patients with generalized and unclassifiable
`standard carbamazepine. Around 50% of study pa-
`epilepsy, the median number of days (25th–75th
`tients reported adverse events at some point, with
`centiles) to treatment failure because of unaccept-
`only small differences seen between drugs. For the
`able adverse events was 90 (28–245) and because
`intention-to-treat population, the lamotrigine group
`of inadequate seizure control was 234 (136–481).
`
`© 2008 Adis Data Information BV. All rights reserved.
`
`CNS Drugs 2008; 22 (9)
`
`Page 4 of 22
`
`

`
`CNS Adverse Events Associated with Antiepileptic Drugs
`
`743
`
`Cumulative incidence analysis of treatment failure
`for unacceptable adverse events showed that lamo-
`trigine was least likely and topiramate most likely to
`cause treatment-limiting adverse events. Topiramate
`was significantly inferior to both valproate (topira-
`mate : valproate [HR; 95% CI] 1.55 [1.07, 2.26])
`and lamotrigine (topiramate : lamotrigine [HR; 95%
`CI] 2.15 [1.41, 3.30]). The reported adverse events
`are listed in table II. Thirty-six percent of patients
`receiving valproate reported adverse events com-
`pared with 45% of patients receiving topiramate.
`
`CNS adverse events were the most common reason
`for treatment failure, except in the case of lamo-
`trigine, for which drug rash was the most common
`reason.
`
`2. Conventional AEDs
`
`The ‘conventional’ AEDs refer, in a somewhat
`arbitrary fashion, to those AEDs in use prior to the
`advent of the ‘newer’ AEDs, typified by drugs such
`as lamotrigine. Conventional AEDs still in use are
`
`Table II. Frequency of clinically important adverse events of antiepileptic drugs in patients with generalized and unclassifiable epilepsy
`(reproduced from Marson et al.,[3] with permission from Elsevier)
`
`No. of patients/adverse effects
`
`Lamotrigine
`
`Topiramate
`
`239
`88 (37)
`
`239
`107 (45)
`
`Valproate
`(valproic acid)
`238
`85 (36)
`
`Total
`
`716
`280 (39)
`
`No. of patients randomized
`Total number (%) of patients with at least one
`adverse event
`Tiredness/drowsiness/fatigue/lethargya
`58 (41)
`18 (12)
`25 (20)
`15 (9)
`Other psychiatrica
`34 (26)
`8 (7)
`19 (15)
`7 (4)
`Weight gaina
`32 (23)
`17 (16)
`7 (2)
`8 (5)
`Behaviour/personality change/aggressiona
`30 (26)
`4 (4)
`20 (18)
`6 (4)
`Worsening of seizuresa
`30 (18)
`7 (3)
`13 (9)
`10 (6)
`Accidental injurya
`20 (12)
`4 (2)
`5 (3)
`11 (7)
`Other neurologicala
`21 (12)
`10 (5)
`7 (4)
`4 (3)
`Headachea
`18 (12)
`5 (4)
`7 (4)
`6 (4)
`Memory problemsa
`17 (12)
`3 (0)
`12 (10)
`2 (2)
`Weight lossa
`17 (12)
`0 (0)
`14 (12)
`3 (0)
`Allergic rasha
`16 (13)
`2 (0)
`1 (1)
`13 (12)
`Tremora
`13 (8)
`8 (6)
`1 (0)
`4 (2)
`Depressiona
`13 (10)
`3 (3)
`9 (6)
`1 (1)
`Confusion/difficulty thinking/disorientationa
`13 (11)
`3 (2)
`7 (7)
`3 (2)
`Dizziness/vertigoa
`10 (6)
`1 (1)
`6 (3)
`3 (2)
`Anxiety/agitation/nervousnessa
`10 (9)
`1 (1)
`2 (2)
`7 (6)
`Nauseaa
`10 (8)
`4 (3)
`2 (1)
`4 (4)
`Other renal tract/genitala
`11 (7)
`3 (2)
`4 (2)
`4 (3)
`Pins and needles/dysaesthesiaea
`10 (6)
`2 (0)
`8 (6)
`0 (0)
`Ataxiaa
`9 (7)
`2 (2)
`3 (2)
`4 (3)
`Other skin and appendagesa
`11 (8)
`5 (3)
`5 (4)
`1 (1)
`Mouth/gum problemsa
`6 (5)
`3 (3)
`2 (1)
`1 (1)
`Sleep disturbancesa
`8 (7)
`1 (1)
`4 (3)
`3 (3)
`Otherb
`106 (71)
`36 (25)
`40 (25)
`30 (21)
`a Data presented are number of patients with the adverse event by intention-to-treat analysis (per-protocol analysis inside brackets).
`b Sorted by descending total frequency: abdominal pain, dyspepsia; alopecia; other general; other visual disturbance; word-finding
`difficulty; vomiting; aches and pains; other gastrointestinal; other musculoskeletal; other respiratory or pulmonary; diarrhoea;
`psychosis; anorexia; bruising; constipation; diplopia; renal or bladder stones; influenza-like symptoms; hallucinations; infection;
`vaginal bleeding; arthritis; asthma; chest infection; childbirth; faints; hypertension; ischaemic heart disease or myocardial infarct;
`other cardiac or vascular; other haematological; psoriasis; shortness of breath; status epilepticus; urinary tract infection; urinary
`retention.
`
`© 2008 Adis Data Information BV. All rights reserved.
`
`CNS Drugs 2008; 22 (9)
`
`Page 5 of 22
`
`

`
`744
`
`Kennedy & Lhatoo
`
`2.1 Benzodiazepines
`
`tive and behavioural adverse effects between
`carbamazepine, phenytoin, phenobarbital (pheno-
`clobazam monotherapy and carbamazepine or
`barbitone), valproate, primidone, ethosuximide and
`phenytoin monotherapy in childhood epilepsy.[8]
`the benzodiazepines. The most commonly reported
`CNS ADRs of these conventional AEDs are listed in Development of tachyphylaxis still means that
`table III.
`clobazam is best used intermittently, most usefully
`as a temporary rescue treatment for seizure clusters.
`Clonazepam is useful as a broad spectrum ad-
`junctive AED.[9] However, sedation and tachyphy-
`Benzodiazepines are approved for the treatment
`laxis limit use of this agent. Clobazam and clonaze-
`of status epilepticus in light of their rapid action and
`pam have similar adverse effect profiles, with
`known efficacy. There have been few placebo-con-
`clobazam perhaps being less sedative (table III).
`trolled trials of diazepam, a drug that has been in use Mood and behavioural disturbances are more likely
`since 1968 and which is still considered to be a first- with clonazepam in individuals with learning diffi-
`line agent in status epilepticus. Sedation is a com-
`culties.[10] Seizure exacerbation as a result of treat-
`mon adverse effect of diazepam, and clobazam was ment with clonazepam has been reported in observa-
`created in 1972 in an effort to minimize the sedative
`tional[11] and animal[12] studies. Clonazepam has
`effects of benzodiazepine therapy while retaining
`also been known to precipitate seizures once it is
`its anxiolytic qualities. Antiepileptic effects of withdrawn[13,14] and has even been implicated in
`clobazam were later reported in 1978. These effects
`precipitation of status epilepticus on initiation or
`are achieved by its action on the benzodiazepine withdrawal of treatment in Lennox-Gastaut syn-
`binding site on the GABAA receptor complex.[5]
`drome.[15-17] Other withdrawal symptoms include
`As expected, drowsiness and dizziness appear to
`insomnia, confusion, anxiety, agitation and severe
`be the most common adverse effects of clobazam,
`catatonia,[18] emphasizing the need for gradual drug
`although in a summary of 70 double-blind studies, withdrawal.
`the incidence of sedation (25.8%) with this drug was
`nearly half that for diazepam (45.5%).[6] Dizziness
`was also less frequently reported (7% vs 12%, re-
`spectively). Clobazam has minimal effects on cogni-
`tive function compared with clonazepam;[7] indeed
`in one randomized, controlled study there were no
`significant differences in the occurrence of cogni-
`
`2.2 Carbamazepine
`
`Carbamazepine is a sodium channel antagonist[19]
`that is effective in controlling partial seizures with
`or without secondary generalization. As an hepatic
`enzyme inducer, care should be taken when pre-
`
`Table III. CNS adverse effects of conventional antiepileptic drugs
`
`Drug
`Carbamazepine
`
`Valproate (valproic acid)
`
`Phenytoin
`
`Phenobarbital (phenobarbitone)
`
`Primidone
`
`Ethosuximide
`Clobazam
`Clonazepam
`
`Adverse effects
`Dizziness, nausea, headache, drowsiness, diplopia, ataxia, incoordination, orofacial dyskinesia,
`acute dyskinesia, asterixis, oculogyric crisis
`Tremor, drowsiness, nausea, incoordination, ataxia, nystagmus, aggression,
`hyperammonaemia
`Nystagmus, ataxia, nausea, depression, drowsiness, headache, paradoxical seizures,
`aggression
`Fatigue, lethargy, tiredness, depression, aggression
`Children: irritability, hyperkinesia, distractibility
`Fatigue, lethargy, tiredness, depression, psychosis, nystagmus, ataxia, nausea
`Children: Irritability, hyperkinesia
`Agitation, drowsiness, headache, lethargy, nausea, poor memory, dizziness
`Fatigue, drowsiness, dizziness, ataxia, irritability, aggression, psychosis
`Fatigue, drowsiness, dizziness, ataxia, irritability, sedation, psychosis
`Children: hyperkinesia, aggression
`
`© 2008 Adis Data Information BV. All rights reserved.
`
`CNS Drugs 2008; 22 (9)
`
`Page 6 of 22
`
`

`
`CNS Adverse Events Associated with Antiepileptic Drugs
`
`745
`
`scribing this agent along with other lipid-soluble
`drugs that may undergo accelerated clearance as a
`result. Even so, carbamazepine has a relatively more
`benign adverse effect profile than its counterpart
`phenytoin.[20,21] A minority of patients receiving car-
`bamazepine experience treatment-limiting ADRs
`that result in cessation of its use, such as the devel-
`opment of a rash, which occurs in 10% of pa-
`tients.[22] Carbamazepine has numerous CNS ad-
`verse effects (see table III), but slow upward titra-
`tion of the drug can help to minimize these effects.
`Use of controlled-release tablets helps prevent ma-
`jor fluctuations in drug concentrations and, seconda-
`rily, occurrence of adverse effects.
`In 1988, the Collaborative Group for Epidemiol-
`ogy of Epilepsy found ADRs in 155 patients to be
`less frequent with carbamazepine (15.5%) compared
`with phenytoin (36%), valproate (31.5%), benzo-
`diazepines (20%) and barbiturates (17%),[20] al-
`though these results may have been subject to study
`bias. In one prospective, double-blind study of 622
`patients, carbamazepine was found to have better
`toxicity scores than phenytoin, phenobarbital and
`primidone.[21] The cognitive adverse effect profile of
`carbamazepine was also better. In children, carba-
`mazepine was superior to barbiturates with respect
`to neuropsychiatric ADRs.[23] Suicidal ideation was
`reported in 47% of children receiving barbiturates
`and 4% of those taking carbamazepine. The tricyclic
`structure of carbamazepine may contribute to the
`lower incidence of depression in patients taking this
`agent compared with the other conventional AEDs.
`Behavioural improvement was noted in 17.8% of
`children receiving carbamazepine monotherapy
`compared with deterioration in behaviour in 1.7% of
`children not receiving this agent as monotherapy.[24]
`Less well known ADRs of carbamazepine have
`been described in the literature, such as psychosis
`associated with polydipsia, which resolved follow-
`ing correction of the resultant hyponatraemia.[25]
`Other psychiatric problems such as insomnia, rest-
`lessness, agitation, anxiety and frank mania have
`been reported.[26,27] Acute dyskinesia, oculogyric
`crisis and asterixis have also been described.[28,29]
`
`It is important to be aware that carbamazepine
`has also been known to exacerbate absence and
`myoclonic seizures in patients with primary genera-
`lized epilepsy.[30,31] Paradoxically, carbamazepine
`has also been associated with the development of
`new types of seizure (generalized and absence) in
`children with presumed partial complex epilepsy.[32]
`Seizure exacerbation was correlated with new EEG
`changes, including generalized spike-and-wave epi-
`leptiform discharges.
`
`2.3 Ethosuximide
`
`Ethosuximide is used almost exclusively for the
`treatment of absence seizures. It has been in clinical
`use since 1958 and acts to reduce voltage-gated
`calcium channel conductance in thalamic neu-
`rons.[33]
`There is a relative dearth of placebo-controlled
`trials of ethosuximide, but in early clinical trials the
`overall incidence of adverse effects ranged from
`26% to 46%, although study methodologies va-
`ried.[34] Commonly known CNS adverse effects of
`ethosuximide are listed in table III. The most fre-
`quently reported adverse effects are nausea, vomit-
`ing, drowsiness, tiredness, loss of appetite and dizzi-
`ness.[35-37] Ethosuximide predominantly affects the
`gastrointestinal system, with neurological effects
`being the next most common. This agent can also
`have a contrasting stimulant effect, resulting in
`mood and behavioural changes (irritability, hyperac-
`tivity, aggression) and insomnia, particularly in pa-
`tients with learning difficulties.[38] Episodes of psy-
`chosis have been reported following improved
`seizure control with ethosuximide, possibly reflect-
`ing ‘forced-normalization’ behaviour;[39] however,
`this phenomenon is not by any means peculiar to this
`drug.
`Withdrawal seizures are known to occur in pa-
`tients taking ethosuximide and, although rare, this
`agent can also exacerbate myoclonic and absence
`seizures in some individuals.[39,40] Whether this is a
`direct drug effect or a manifestation of toxicity is
`unknown. Ethosuximide is ineffective in controlling
`generalized seizures that may emerge in patients
`with absence epilepsy.[41] Again, it is difficult to
`
`© 2008 Adis Data Information BV. All rights reserved.
`
`CNS Drugs 2008; 22 (9)
`
`Page 7 of 22
`
`

`
`746
`
`Kennedy & Lhatoo
`
`differentiate an aggravating effect from the progres-
`sive nature of the epilepsy syndrome. Parkinsonian-
`like symptoms have been known to develop after
`long-term treatment with ethosuximide.[42]
`
`2.4 Phenobarbital (Phenobarbitone)
`
`(11%) and valproate (5%), mainly due to drowsiness
`and lethargy on initial dosages of 60 mg/day. A
`study using a lower starting dosage (32 mg/day) of
`phenobarbital reported less sedative effects, indicat-
`ing a dose-dependent effect.[54] As with the Veterans
`Administration Cooperative Study,[21,54] no signif-
`icant neuropsychological differences were found be-
`tween phenobarbital and other conventional AEDs
`in three of five studies reviewed in 2004.[55] One
`randomized, double-blind, triple crossover study
`found a decrement in performance on only one
`psychometric measure (digit symbol) in the pheno-
`barbital-treated group compared with carbamaze-
`pine and phenytoin.[56]
`Phenobarbital toxicity can progress to coma and
`death.[57] Treatment includes timely elimination by
`forced alkaline diuresis whilst supporting vital
`organs in an intensive care environment. Intoxica-
`tion with phenobarbital can paradoxically lead to
`seizure exacerbation, particularly in children with
`seizure disorders such as atypical benign rolandic
`epilepsy or epileptic negative myoclonus.[58]
`
`Phenobarbital is thought to mediate its effect by
`binding to the chloride channel complex on the
`GABAA receptor, with resultant prolongation of
`inhibitory post-synaptic potentials,[5] and is effective
`for treating partial and generalized seizures. In prac-
`tice, clinical use of phenobarbital in developed
`countries has become limited by its cognitive and
`behavioural adverse effect profile, although in re-
`source-poor countries it is still a very important
`AED.[43-45] Relatively poor correlations between
`blood concentrations and toxicity may be seen,
`which makes management difficult.[46]
`The most frequently reported CNS adverse ef-
`fects of phenobarbital are listed in table III. Pheno-
`barbital can produce fatigue, lethargy and drowsi-
`ness in adults and paradoxical insomnia, hyperactiv-
`ity and aggression in children.[47] In a double-blind,
`crossover study of phenobarbital and valproate in
`children with epilepsy, there was a high withdrawal
`Phenytoin is structurally related to the barbitu-
`rate due to behavioural problems in the phenobarbi-
`rates and exerts its antiepileptic effect via sodium
`tal-treated group.[48] Sedation is particularly com-
`channel antagonism.[19] An increase in circulating
`mon at the onset of treatment although this effect
`free phenytoin can sometimes produce signs of
`can be minimized by starting at a lower dose. Al-
`phenytoin toxicity without large changes in the total
`though tolerance to sedative effects may develop,
`concentration of phenytoin.[59] Metabolism of
`effects on memory, concentration and intelligence
`phenytoin, and therefore phenytoin concentrations,
`have been found to persist, particularly in chil-
`can be affected by drugs that share the same hepatic
`dren.[49,50] Phenobarbital has even been reported to microsomal system, such as carbamazepine, val-
`have long-term deleterious effects on verbal intelli-
`proate, ethosuximide, corticosteroids and warfarin.
`gence after fetal exposure in the womb.[51]
`Because of zero-order pharmacokinetics, plasma
`A randomized trial compared the efficacy and
`concentrations of phenytoin exceeding the drug’s
`narrow therapeutic window (40–80 μmol/L) can
`tolerability of 3 years’ monotherapy with phenobar-
`bital, phenytoin, carbamazepine and valproate in
`easily be reached with small increases in dose. At
`patients with newly-diagnosed epilepsy.[52] As with
`supratherapeutic levels, nystagmus and ataxia can
`other comparative studies,[53] no significant differ-
`appear to progress to lethargy, drowsiness and con-
`fusion, followed by coma in extreme cases.[60] In
`ence in efficacy was found between phenobarbital
`and other AEDs, but there were significant differ-
`some patients, phenytoin toxicity can cause a para-
`doxical liability to seizure activity.[61] More unusual
`ences in tolerability. The greatest proportion of pa-
`tients (22%) withdrew from phenobarbital treatment
`symptoms of phenytoin toxicity include involunta-
`ry movements, ophthalmoplegia[62] or peripheral
`compared with phenytoin (3%), carbamazepine
`
`2.5 Phenytoin
`
`© 2008 Adis Data Information BV. All rights reserved.
`
`CNS Drugs 2008; 22 (9)
`
`Page 8 of 22
`
`

`
`CNS Adverse Events Associated with Antiepileptic Drugs
`
`747
`
`neuropathy.[63] Reduced colour discrimination with
`phenytoin overdose has also been reported.[64]
`Prompt dose reduction usually results in symptom
`resolution, and blood concentrations are helpful in
`toxicity monitoring, although concentrations at
`which various clinical symptoms and signs manifest
`can vary in individual patients. Commonly reported
`CNS adverse effects of phenytoin are listed in table
`III.
`Phenytoin can affect cognition[65] althoug