Antiepileptic Drugs, 5th Edition

General Principles


Antiepileptic Drug Monotherapy in Adults: Selection and Use in New-Onset Epilepsy

Richard H. Mattson MD

Professor of Neurology and Director of Medical Studies, Department of Neurology, Yale University School of Medicine and Veterans Administration, West Haven, Connecticut

The new onset of epileptic seizures usually leads to initiation of antiepileptic drug (AED) treatment to prevent recurrence. Although this chapter is directed at treatment of adults, a number of basic principles involved in this process are common to all patients. These principles include how to select, how to initiate, and how to maintain AED treatment.

Before beginning treatment, it is necessary to decide whether to treat at all or withhold therapy. After a single (first) seizure, there is uncertainty that any further seizures will occur. The probabilities vary with many factors and range from 20% to 70% (1). Ultimately, the patient, family, and physician need to balance the relative risk to that individual of another seizure against the potential for harm occurring as a consequence of AED treatment. Finally, if treatment has been associated with a remission of seizures, a decision must be made as to when and if it is advisable to discontinue treatment and, if so, how it should be done. The issues are very similar to those surrounding initiation of therapy and depend on the risks and benefits of a recurrent seizure compared with the adverse effects of AED therapy.

If more than two unprovoked seizures have occurred, the probabilities of recurrence exceed 70% (2), and most often treatment is recommended. The selection of an initial AED from among the more than a dozen widely available requires a careful review of all the properties of the various drugs and the individual needs of the patient. After selection has been made, the treatment, start-up, and maintenance must be planned and implemented.


When first seen, the patient with adult-onset epilepsy presents with a history of some type of seizure (3). This is most commonly of tonic-clonic type with or without partial seizures or, less frequently, absence and myoclonic seizures. After evaluation is complete, a diagnosis of an epilepsy type or syndrome is made whenever possible (4). Localization-related (partial) epilepsy is the most common type and is of symptomatic or cryptogenic etiology. Less often, generalized idiopathic epilepsy appears in adult life. When possible, treatment is based on epilepsy syndrome because seizure types in a particular syndrome are likely to respond to the same AED. For example, valproate (VPA) is effective for tonic-clonic, absence, and myoclonic seizures often occurring in juvenile myoclonic epilepsy, whereas carbamazepine (CBZ) is especially effective for partial and tonic-clonic seizures characteristic of localization-related epilepsy. However, the epilepsy syndrome frequently is not identifiable, particularly at onset, so clinical trials often do not clearly distinguish the type. In addition, efficacy trials use seizure control as a primary outcome. For this reason, the focus of AED selection often emphasizes seizure rather than epilepsy type.


Selection of a drug for epilepsy treatment depends on multiple characteristics (Table 7.1).

A drug, of course, must be effective in preventing recurrent seizures. Total control is the usual goal, but may not be possible with monotherapy or even combined drug treatment. Efficacy may be limited to a specific seizure type for some compounds, whereas others have a broad spectrum of activity. The degree of efficacy or potency is important, but


some patients have easily controlled tonic-clonic seizures and the most potent drug may not be necessary, especially if the drug with greatest potency has poorer tolerability (5, 6,7). For patients not fully controlled, measures such as reduction in number or rate of attacks or change in severity of seizures are important outcomes (8, 9, 10).


▪ Efficacy

▪ Safety

▪ Tolerability

▪ Pharmacokinetic properties

▪ Formulations

▪ Expense

Adverse effects usually are divided into those of tolerability and safety. The latter is of particular importance. Serious medical hazard usually leads to avoidance or discontinuation of AEDs. Occasionally, a drug with a risk of systemic side effects of potentially life-threatening type may be justifiable for severe refractory epilepsy. Thus, a place exists for the use of felbamate (FLB) in selected patients (11).

Although safety is of greatest concern, tolerability is a more common problem in AED use. Some drugs, such as the barbiturates, produce sedation or some cognitive difficulty at high dose that compromises the quality of life. Virtually all AEDs can produce unwanted, annoying, and sometimes almost incapacitating side effects. These usually are correctable by reduction of dose, but this may then allow recurrence of seizures.

Pharmacokinetic properties include absorption, distribution to different body parts, biotransformation, and elimination. These determine how much and how often an AED must be given to maintain the most desirable pharmacodynamic effects. Hepatic metabolism is a major mode of clearance of all the older and a number of the newer AEDs. Many drugs share the same biotransforming enzyme systems, resulting in frequent interactions that can make management more complex.


FIGURE 7.1. Partial seizure group remaining in study. Carbamazepine (CBZ) and phenytoin (PHT) were more successful than phenobarbital (PB) or primidone (PRM).

Formulations of some drugs have rapid absorption, with peak blood levels often producing transient side effects. This can be avoided by using more delayed- or extended-release formulations. At times, sprinkle, rectal, or parenteral formulations are needed or desirable, and are not available for all products.

Finally, expense is an issue. Some of the population have insurance or other sources of economic support for the cost of AEDs, but this is not true for many individuals. Although national health plans provide medication in some countries, in others the cost of the newer AEDs runs approximately three or more times that of the older, established compounds. In the United States, for example, cost for each of the new AEDs may range from $3,000 to $6,000 per year, compared with $100 to $1,500 for older AEDs.


The relative advantages of each AED are based on the criteria described previously. A single outcome measure that allows comparison is the life-table analysis. The length of time a patient continues on drug is an indication of acceptable efficacy, safety, and tolerability. A large, blinded, randomized trial has been conducted comparing CBZ, phenobarbital (PB), phenytoin (PHT), or primidone (PRM) monotherapy in treatment of partial epilepsy in adults (12). Life-table analysis of patients with partial and tonic-clonic seizures revealed an outcome significantly less successful for PRM The primary reason was poor tolerability during initiation of therapy. For patients with partial seizures, CBZ and PHT were significantly more successful than either PB or PRM, in this case owing to differences in efficacy and tolerability (Figure 7.1). Later trials comparing CBZ and


VPA failed to detect a difference in retention for tonic-clonic, partial, or all seizures combined (13, 14, 15).

The first active-control monotherapy trial of the new AEDs showed that FLB was more successful than low-dose VPA (16). Brodie and colleagues (17) compared CBZ with lamotrigine (LTG) as monotherapy in patients with new-onset epilepsy and overall found fewer withdrawals from the study in patients on LTG because of better tolerability (Figure 7.2). Steiner and coinvestigators (18) found equal retention in a similar trial comparing LTG and PHT. More recently, three independent clinical monotherapy trials in patients with partial epilepsy compared oxcarbazepine (OXC) with CBZ, PHT, or VPA, and in each case equal or superior retention was achieved with OXC (19, 20, 21). No efficacy differences were found in these trials, but OXC was better tolerated. In a somewhat more unusual design, topiramate (TPM) was compared in patients selected for treatment with CBZ or VPA according to the decision by the physician as to the optimal drug for therapy. Then the patient was randomized to high- versus low-dose TPM (200 or 100 mg) or continued on the original AED. Retention was not significantly different between CBZ, VPA, and either dose of TPM (22). In another large, double-blind study, patients with new-onset epilepsy were randomized to gabapentin (GBP) or LTG, and no differences in retention were found between the two drugs (23). Two studies indicate outcome for vigabatrin (VGB) therapy is comparable with CBZ in terms of retention (24,25). VGB showed better tolerability but less efficacy against seizures. Another trial showed two doses of GBP (900 and 1,200 mg/day) were more successful than a low dose (300 mg/day) and produced results equivalent to those obtained in an unblinded comparison group treated with 600 mg CBZ (26).


FIGURE 7.2. Continuation on carbamazepine (CBZ) or lamotrigine (LTG) monotherapy after double-blind randomization in adults with new-onset epilepsy. (From Brodie MJ, Richens A, Yeun AW. Double-blind comparison of lamotrigine and carbamazepine in newly diagnosed epilepsy. Lancet 1995;345:476-479, with permission.)

It is evident that most studies of new AEDs have involved CBZ, although some have compared a study drug or drugs with PHT or VPA, and in these trials no significant differences have been detected. Only one trial has compared success between two or more of the new AEDs (23). LTG has shown somewhat better retention than CBZ in an elderly population of patients with new-onset epilepsy (27). The other new AEDs, levetiracetam (LEV), tiagabine (TGB), and zonisamide (ZNS), have not been sufficiently tested in randomized monotherapy studies of new-onset adult epilepsy to evaluate outcome. However, add-on trials have clearly demonstrated efficacy for each drug.

In summary, comparatively few differences are found in overall retention among either new or older AEDs in comparative trials using life-table analyses. Overall, 60% to 80% of patients with new-onset epilepsy remain on the drug to which they were randomized. Success is less if an intent-to-treat analysis is used and includes patients leaving the study for apparent non-study-related reasons. Outcome is somewhat better in studies including generalized epilepsy, in contrast to those limiting intake to partial epilepsy (12, 13, 14, 15). The close similarity in success for the AEDs using life-table analyses makes it necessary to turn to alternative criteria for making treatment selection.




Comparisons of efficacy are difficult because different outcome measures are used in clinical trials and results are not always comparable (10) (Table 7.2). For initial treatment, the most important goal is complete seizure control for a period of time (e.g., 6, 12, 24 months). Another frequently used measure is time to first (or nth) seizure. Time and percentage of patients entering remission provide similar information expressed in a different way. Because complete control is not possible for many patients, other measures of efficacy have been used. The number of seizures that occur in a test period can be compared between a study drug and another, or with placebo. Often this number is compared with a lead-in baseline test period. The reduction (or increase) in seizures is a measure of efficacy of the treatment. When the observation period is not the same for all subjects, the seizure number per period of time (week, month) can be used. The change in seizures commonly is expressed as a reduction relative to a baseline number. Outcome most often is expressed as the percentage of patients achieving a 50% or greater reduction. This outcome measure is used most often for studies involving patients with difficult-to-control epilepsy rather than new or early-onset disease, in which case total control often is achieved. Seizure severity may change even for a specific seizure type. For example, a tonic-clonic seizure may be less severe if unaccompanied by tongue biting, incontinence, or postictal confusion, or preceded by a lengthy aura. Several instruments have been developed to express the severity in numeric terms (8,9).

Compounding the problem of assessing efficacy of AEDs is the limitation of many studies in defining the inclusion and analysis of specific seizure types (3) or the epilepsy classification (4).

Tonic-Clonic Seizures

Evidence based on prospective, randomized trials suggests little, if any, difference in efficacy for treatment of tonic-clonic seizures. Most studies indicate tonic-clonic seizures are more fully controlled than those of partial type (5, 6, 7,12, 13, 14). Although modest differences are found in different studies, most standard and older AEDs successfully control approximately 60% of patients for a year of follow-up from the beginning of what should be adequate dosing. Because some patients enter remission after one or more seizures while on therapy, approximately 70% to 80% will come under control (14,15).


▪ Complete seizure control

▪ Time to first (nth) seizure

▪ Number (%) entering remission

▪ Seizure number (rate)

▪ Seizure severity score

▪ Surrogate markers (electroencephalogram)

Some studies (28, 29, 30, 31, 32) suggest a higher percentage of patients with tonic-clonic seizures associated with idiopathic generalized epilepsy syndromes can be well controlled than those with secondarily generalized tonic-clonic seizures associated with localization-related epilepsy. Prospective, blinded large trials have not been conducted to address this issue. Retrospective analysis of the study of PHT compared to VPA revealed those patients with probable idiopathic generalized epilepsy had better control with VPA therapy (31). Because accurate diagnosis often is difficult, many trials do not restrict entry to a single epilepsy type (12,13,15,20,21).

In most studies, the trial design does not possess enough power to test for differences between drugs for tonic-clonic seizures. The problem arises in part from the fact that these seizures occur relatively infrequently compared with partial seizures. Consequently, a long duration of study is required to detect outcomes. In addition, tonic-clonic seizures are successfully controlled in a high percentage of cases, so detection of a significant difference becomes more difficult. Despite these limitations, all the older standard AEDs except ethosuximide (ESM) appear to be effective. Based on consensus evidence and nonrandomized trials, VPA is thought to be somewhat more effective than CBZ, PB, or PHT (33).

Partial Seizures

All the available AEDs except ESM are believed to be effective for partial seizures, although less so than for tonic-clonic seizures. The older multicenter studies (12, 13, 14, 15) reveal better control of tonic-clonic than partial seizures over time. Some studies suggest that in more difficult control problems, CBZ and perhaps PHT are slightly more effective than the barbiturates or VPA. In the first Department of Veterans Affairs Cooperative (VA COOP) Study (12), CBZ provided better complete control of partial seizures than the barbiturates (PB and PRM). In the second VA COOP study (13), CBZ was more effective than VPA using time to first seizure (Figure 7.3), seizure score or seizure rate, but no difference was found for complete control for 1 year of treatment. Other investigators have not found a difference in efficacy between these two drugs. The apparent difference in efficacy may be explained by several factors. The VA patient population represented a more difficult-to-control group because only approximately half had never been previously treated. The numbers of subjects in the VA trial were very large, allowing detection of small differences. Finally, outcome measures were different in the various studies.




FIGURE 7.3. Time to first seizure, carbamazepine (CBZ) (top line) versus valproate (VPA), Department of Veterans Affairs Cooperative Study #264. (From Mattson RH, Cramer JA, Collins JF, et al. A comparison of valproate with carbamazepine for the treatment of complex partial seizures and secondarily generalized tonic/clonic seizures in adults. N Engl J Med 1992;327:765-771, with permission.) Difference p < .05.

Mixed Partial and Tonic-Clonic Seizures

Patients with partial (localization-related) epilepsies often experience both partial as well as generalized seizures, so it is appropriate to combine all groups together in assessing the outcome of treatment. The results parallel the findings for specific seizure types, with success being intermediate between patients with only partial or only tonic-clonic seizures because tonic-clonic seizures are more easily controlled than those of partial type. In the VA COOP Studies, approximately 40% to 50% could be completely controlled after treatment with one of the standard older AEDs (5,10,11). The study of Heller et al. also noted better control of tonic clonic than partial seizures (14).

Absence Seizures

Although absence seizures almost always have their onset before adult life, occasionally they are not recognized or are misdiagnosed earlier. Many early open studies established efficacy initially with use of trimethadione and other diones, and later with ESM and related analogs (34). VPA was found to have comparable efficacy with ESM in controlled trials (35). VPA also caused a dramatic decrease in the surrogate end point of spike-wave discharges on the electroencephalogram (36). Increasing evidence has indicated that LTG (37, 38,39, 40) is effective as both add-on therapy and monotherapy. TPM, ZNS, and perhaps LEV also may be effective (41,42). Controlled, comparative trials have not established the relative efficacy of these newer AEDs.

For treatment of absence seizures, only ESM and VPA possess known long-term efficacy. Benzodiazepines are very useful acutely, but tolerance develops over time. Absence seizures have been worsened or unmasked especially with the use of CBZ (43,44), but PHT also has been found to have an adverse effect at times. Recent reports indicate that AEDs that increase brain γ-aminobutyric acid (GABA; i.e., GBP, TGB and VGB) also may have an aggravating effect (44).

Myoclonic Seizures

No controlled trials have been done to compare the relative efficacy of the available AEDs in treatment of myoclonic seizures, but extensive experience has shown VPA to be highly effective (45, 46, 47, 48, 49, 50). The benzodiazepines, including clonazepam, diazepam, and nitrazepam, are all effective and used especially for resistant seizures. Tolerance commonly develops after some months, limiting their value. Increasing evidence indicates TPM and ZNS are effective as well. LTG has been reported both to help and to aggravate myoclonic seizures (38,51,52)

Tonic and Atonic Seizures

Tonic or atonic seizures do not commonly arise in adulthood and usually are found in patients with refractory, extratemporal partial epilepsy in adults or in secondary generalized epilepsies with onset in childhood. Most AEDs have some effect in lessening severity. Empirical trial often


is needed to find the most helpful drug. This is a condition that may well justify the use of FLB (9,53).



Safety problems are infrequent, but may occur with use of all the AEDs, with the possible exception of GBP. Idiosyncratic hypersensitivity reactions with rash are seen in 5% to 10% after start-up, and usually occur within the first 6 months of therapy (12, 13, 14, 15.). Rarely, these proceed to more serious toxicity, including Stevens-Johnson syndrome, epidermal toxic necrolysis, hepatitis, or aplastic anemia. Only VPA among the older drugs does not cause these hypersensitivity reactions. Among the new AEDs, LTG has been associated most frequently with idiosyncratic hypersensitivity reactions that sometimes have progressed to Stevens-Johnson syndrome or toxic epidermal necrolysis syndrome. Slow titration of the dose, especially when coadministered with VPA, minimizes the risk (54). In monotherapy trials, using a gradual increase in dosage, the frequency of rash was comparable with that with CBZ (17,23,27). OXC and ZNS also have been found to cause rash on occasion, but GBP, LEV, TGB, and TPM have an incidence comparable with placebo in clinical trials, a considerable safety advantage for these drugs.

Although CBZ was associated with aplastic anemia early after its introduction, this problem has proven to be very rare. On the other hand, CBZ commonly is associated with clinically unimportant leukopenia (12), hyponatremia, and, rarely, cardiac arrhythmias (13,55,56). All the older AEDs have been associated with rare idiosyncratic hepatitis, vasculitis, and multiorgan failure. These immunologically mediated problems are not seen with VPA use, but rare metabolic hepatic failure, pancreatitis, and thrombocytopenia all have been reported. PHT also can cause neuropathy and cerebellar degeneration.

All the older AEDs have been associated with some degree of decreased bone calcium and pathologic fractures. Although thought to be due in part to enzyme induction, the same changes are associated with use of VPA, a nonenzyme inducer (57).

The barbiturates also can produce chronic connective tissue disorders such as Dupuytren's contracture (58).

FLB has caused clinically significant aplastic anemia and hepatic failure, markedly limiting use of the drug (11). VGB can cause retinal damage and concentric visual field loss (59). Although it is available in most developed countries except the United States, future use of VGB likely will be reserved for those whose poorly controlled seizures justify the risk.


All AEDs can produce somnolence and mental slowing if dosage is too high or increased too rapidly. LTG and FLB are least likely to have this effect in long-term use. CBZ most often produces visual disturbance or dizziness; PHT, ataxia and mental slowing or sedation; PB, sedation and affective/behavioral/cognitive change; and VPA, tremor and, at times, sedation. PRM causes dizziness and somnolence. All these older AEDs, including ESM, can cause gastrointestinal side effects, especially at start-up. PB is least likely of the older drugs to cause dizziness or gastrointestinal complaints (12,60) (Table 7.3).

































Veterans Administration cooperative study. (From Steiner TJ, Dellaportas CI, Findley LJ, et al. lamotrigine monotherapy in newly diagnosed untreated epilepsy: a double-blind, randomized comparison with phenytoin. Epilepsia 1999;40:601-607, with permission.) N= 240.

Many of the new AEDs show superior tolerability in terms of central nervous system complaints (61, 62, 63, 64, 65, 66). LTG, GBP, OXC, and VGB were better tolerated than CBZ in comparative testing (17,19,24,26,27). FLB, TPM, and ZNS sometimes are associated with gastrointestinal symptoms and, not infrequently, weight loss. This may be considered either a negative or positive effect, depending on the individual. TPM and, to some degree, TGB and ZNS appear to produce sedation and cognitive difficulties if given too rapidly. In addition to these complaints, FLB and, occasionally, LTG have been associated with headache and insomnia.

A major problem with many AEDs is an unwanted effect on cognition, mood, or behavior. Among the older AEDs, neuropsychological testing has consistently revealed deficits associated with use of PB (67,68). No consistent differences have been found among the other AEDs when comparable doses have been given (67). In VA COOP Study #264, a behavioral toxicity battery showed no differences from baseline to a retest after 6 months of treatment with CBZ or VPA, and no differences were found between the two drugs. The only AED effect detected was a failure to find a practice effect that was observed in a control group (69).


All of the older drugs are cleared at least partly by hepatic metabolism followed by renal elimination of inactive metabolites. As a consequence, drug interactions are complex and can result in changes in blood levels. PHT, PB, and PRM induce clearance of CBZ and VPA, whereas many drugs inhibit metabolism of CBZ and cause levels to rise,


leading to side effects. Except for VPA, these AEDs may induce the clearance of other drugs, such as warfarin sodium, cyclosporine, and oral contraceptives, leading to less-than-expected effectiveness of the drugs and potential serious clinical outcomes. Knowledge and anticipation of these problems can obviate them, but for the occasional or new prescriber of AEDs, the subtleties of patient care with the older AEDs can make management difficult and, at times, dangerous.

The rate of clearance affects frequency of dosing. Most of the older drugs can be administered in twice-daily divided doses when used as monotherapy. CBZ and VPA may have shorter half-lives and if control requires high doses, variation between peak and trough levels can lead to insufficient control or transient excess dose side effects. To maintain reasonably constant blood/brain levels, three or four times daily regimens may be required; these often lead to decreased compliance and suboptimal clinical effect (70).

Most of the new AEDs have pharmacokinetic properties that make management easier (71). GBP, LEV, and VGB are renally eliminated. Although LTG, TPM, and ZNS undergo conjugation in the liver and their clearance can be increased or decreased by other drugs, they are not enzyme inducers and do not effect clearance of other compounds, with the exception of some lowering of sex hormone levels by TPM. OXC and TGB are primarily metabolized by the liver, but have minimal effects on the metabolism of other drugs, again with the exception of some decrease in sex hormone levels associated with OXC use. Both are, to some extent, inducible by drugs such as PHT and PB. Protein binding is not clinically significant for any of the new AEDs except TGB. The half-life of many of the new AEDs is sufficiently long to allow twice-daily dosing, with the exception of GBP, TGB, and VGB. Although these AEDs have half-lives of approximately 6 hours, they have a mechanism of GABA action that may cause a prolonged effect beyond what might be expected from blood level concentrations (72). GBP also has limited absorption because of saturable active L-amino transport. To achieve maximal levels for each dose, divided administration is advised.


Most patients are able to take the AEDs orally, in pill or capsule form. However, some cause gastrointestinal irritation and some individuals, especially children or the mentally handicapped, will not or cannot take the pills. At times, chewable, liquid, or sprinkle formulations are desirable. For drugs with a short half-life or rapid absorption causing blood level peaks and poor tolerability, a delayed-release product can allow prolonged absorption, leading to more constant concentrations, improving efficacy as well as tolerability. When oral intake is not possible, a parenteral formulation allows administration of the drug if rapid clinical effect is required.

CBZ is available in regular tablets, chewable tablets [Tegretol 100 mg (Novartis, Summit, NJ)], and slow-release [Tegretol-XR, Retard, or Carbatrol (Athena Neurosciences, South San Francisco, CA)], but no parenteral formulation. After autoinduction and especially if coadministered with enzyme-inducing drugs like PB or PHT, peak-and-trough effects may occur unless multiple daily doses are administered. In these cases, the slow-release formulations are particularly useful clinically.

The other older standard AEDs (PB, PHT, VPA) have an extensive variety of formulations compared with the newer AEDs. PB has a slow clearance, so the amount given daily is only approximately 10% of total body stores. The rate of absorption is not important because peak effects are minimal. It is available in tablets as well as an elixir for oral administration. It also is available in a liquid in propylene glycol solution or as the sodium salt soluble in water. Both are suitable for parenteral administration. PHT is available as a tablet, a suspension, a delayed preparation (Dilantin; Pfizer, NY), and parenteral formulations. PHT is highly insoluble and is dissolved in propylene glycol at a pH of 11. The solution is highly irritating to tissue and must be given carefully and slowly by vein. A water-soluble prodrug, fosphenytoin (Cerebyx; Pfizer, NY), can safely be given intravenously (i.v.) or by intramuscular injection. VPA is available as a syrup, sprinkles, capsules, and delayed-absorption or extended-release tablets. A water-soluble parenteral formulation [Depacon (Abbott Laboratories, Abbott Park, IL), Epilim; Sanofi, NY] is very well tolerated when given i.v. at rapid rates.

The new AEDs have a limited number of formulations, although LTG and TPM are available as chewable tablets and sprinkles. No parenteral formulations have yet been marketed for these products, although many are water soluble and should not pose major difficulties for the development of parenteral formulations. The monohydroxy metabolite of OXC (MHD), has the potential for parenteral administration.


Most of these older AEDs (Table 7.4) have been available for 40 years or more, in contrast to the newer AEDs, which have been extensively used for less than a decade. As a consequence, much more information has accumulated about optimal selection and use of these older drugs. These were reviewed in detail in the last edition of this book (73). We have a better awareness of their adverse effects, especially those that do not become evident until after long-term use. Examples include recognition of neuropathy with use of PHT, connective tissue disorders with use of the barbiturates, and teratogenesis with almost all the older AEDs. Even new AEDs may have such adverse effects not recognized in


early use. VGB has proven to cause retinal dysfunction, but this was not recognized for a decade and a half despite special scrutiny of visual function in these patients (57).







Very effective for partial and tonic-clonic seizures
Minimal long-term sedative, cognitive,behavioral adverse effects

More frequent transient side effects during initiation of treatment; rash; complex pharmacokinetics; no parenteral formulation
Leukopenia, hyponatremia

A drug of first choice for partial epilepsies


Very effective for treatment of absence seizures

Effective only for absence seizures
Frequent gastrointestinal side effects

A drug of first choice for treatment of absence seizures


Very effective for tonic-clonic and effective for partial seizures
Available from multiple routes of administration

Adverse sedative, cognitive, effective, behavioral effects; rash; chronic connective tissue effects; complex pharmacokinetics

No longer drug of first choice, but effective, relatively safe and very inexpensive


Very effective for partial and tonic-clonic seizures
Parenteral formulation available

Cosmetic side effects; pharmacokinetics are complex; rash; chronic neuropathy, cerebellar ataxia
Saturation kinetics

A drug of first choice for partial epilepsies potent enzyme inducer; excellent for rapid initiation of treatment


Very effective for tonic-clonic seizures; effective for partial seizures

Side effects common during initiation of therapy; other adverse effects same as phenobarbital

Not a drug of first choice, but when used alone is effective and minimally toxic; metabolized to phenobarbital


Broad-spectrum efficacy for all seizure types;
Intravenous formulation available

Weight gain; teratogenicity, tremor, alopecia; rare pancreatitis, hepatitis, bleeding disorder

First choice for idiopathic epilepsy; an alternative drug of first choice for partial epilepsy; excellent for rapid parenteral therapy


CBZ was developed by Geigy Ltd. in the 1950s along with a number of related tricyclic compounds and first reported to be of use in both epilepsy and trigeminal neuralgia in the 1960s (56). Its use especially in the localization-related (partial) epilepsies with tonic-clonic and partial seizures has grown steadily over the years, to the point where it is widely regarded as one of, if not the drug of choice for these epilepsy types (33).

CBZ has been studied extensively for efficacy in a variety of designs, and no other AED has ever been known to possess greater efficacy. Some studies have found CBZ to have greater efficacy in treatment of partial (and sometimes tonic-clonic) seizures than the barbiturates, GBP, VPA, and VGB (10,11,22, 23, 24). No efficacy differences have been found between CBZ and PHT despite many comparative trials. In any given patient, however, one may prove more effective than the other. The efficacy of CBZ compared with LTG or TPM has not been defined because the trials were not designed with sufficient numbers of patients with partial seizures to detect differences that might be clinically significant. In contrast to efficacy for partial and secondarily generalized seizures, considerable observational reports indicate CBZ does not help and may worsen or unmask absence and myoclonic seizures (43,44,51).

CBZ causes safety problems similar to those with other older AEDs. The most common problem is a hypersensitivity rash that occurs in approximately 10% of patients. The onset usually occurs within the first days or weeks of initiation of treatment. The rash may be mild and subside spontaneously or with lowering or temporarily discontinuing the drug. Rarely, this progresses to Stevens-Johnson syndrome or toxic epidermal necrolysis with multisystem involvement and serious morbidity as well as death. A more widely perceived risk of toxicity is the occurrence of agranulocytosis or aplastic anemia. This complication is quite rare and is estimated to occur only in 1 person in 100,000 or 200,000 people exposed to the drug. On the other hand, leukopenia of mild degree is a common accompaniment but of no clinical significance (except as a source of anxiety for the treating physician). Hyponatremia, usually of modest degree, also may occur. This rarely is of clinical importance unless some other factor leads to sodium loss such as use of some diuretics, lowsodium diet, or hemodilution from excessive solute-free fluid administration. Infrequently, cardiac arrhythmias have been reported during use of CBZ (13,57). Some gastrointestinal distress may occur at drug initiation but is not usually a persistent or significant safety problem.

Central nervous system adverse effects are dose and blood level related. In addition, they are especially prominent during initiation before tolerance and autoinduction


have developed. Dizziness, blurred vision, and diplopia are most frequent, although as levels increase sedation can be a complaint. The problems may occur intermittently during peak concentrations of drug, but can be minimized by use of more frequent dosing of smaller quantities or use of slow-release formulations. Cognitive effects are minimal in most individuals, including both volunteer subjects and patients. In contrast to some AEDs that may aggravate depression, CBZ often is used in treatment of mood disorders. In longterm use, CBZ usually is free of adverse effects at doses that produce control in most patients.

The pharmacokinetic properties of CBZ are somewhat complex. Metabolized to an active metabolite, the 10,11-epoxide, CBZ is primarily oxidized by the hepatic cytochrome P450 (CYP) 3a isoenzyme system and undergoes autoinduction. As a consequence, blood concentrations decrease over days or weeks at a constant dose. CBZ can induce the metabolism and increase the clearance of a number of drugs as well as sex hormones, decreasing the amount available for efficacy. The metabolism of CBZ also is inducible by PHT and the barbiturates in particular. The resulting increased clearance my decrease the half-life to as short as 6 hours, necessitating frequent dosing to avoid wide swings in blood/brain concentrations and consequent loss in efficacy or excess effects. When used as monotherapy, the half-life of CBZ is closer to 9 to 12 hours, which often allows twice-daily or at most three-times-daily dosing. Inhibition by other medications can cause clearance to slow, levels to rise, and adverse effects to appear. Commonly used medications such as the macrolide antibiotics are just one example (56). CBZ is available in a variety of oral formulations, including slow-release capsules, but owing to marked insolubility is not available for parenteral use. CBZ is among the least expensive AEDs.

In summary, after almost four decades of use, CBZ has become the gold standard for treatment of partial and secondarily generalized tonic-clonic seizures on the basis of unexcelled efficacy, long-term safety, and modest cost. With more experience, some newer AEDs may become preferable because of better tolerability, safety, and pharmacokinetic properties.


ESM is an older standard AED with limited use in adults. It was introduced in the 1950s for the treatment of absence and similar seizures. It was safer than and largely replaced the diones, which had more adverse effects. No efficacy was found for treatment of tonic-clonic or partial seizures, limiting its use to add-on or joint therapy unless the patient had only childhood absence epilepsy. It has remained a drug of choice for this syndrome (34). When VPA became widely available in the 1970s and demonstrated efficacy for broad-spectrum seizure control as well as equal efficacy in treatment of absence seizures, the indication for ESM use declined. However, it was found that some patients who did not respond to either AED had a better response to the combination. Some patients who had absence in addition to tonic-clonic seizures but had intolerable or toxic adverse effects from VPA could be treated with ESM and another AED in combination. ESM is available only in oral formulation and is not a primary drug choice for absence status epilepticus. A variety of adverse effects have been reported with use of ESM. Because the drug often has been combined with other drugs such as PB or PHT, it sometimes is difficult clearly to define the effect of ESM alone. Safety issues include the rare reports of Stevens-Johnson syndrome, aplastic anemia, and hepatic and renal failure. However, consistent reports suggest the most common problems associated with use of ESM are nausea, abdominal pain, and, at times, weight loss.

In summary, ESM has a small but valuable role as a first-line AED for absence seizures. It is very rare for these to appear in new-onset adult epilepsy in the absence of other seizure types, in which case VPA or even LTG or TPM are more likely choices.


PB, the oldest AED in general use, remains the most widely administered AED in the world. Introduced into use by Hauptman in 1912, PB was one of a group of barbiturates synthesized and studied by Bayer, the German chemical/pharmaceutical company, in the 1800s. Although used initially as a sedative, its dramatic effectiveness in epilepsy treatment became evident. Widespread use occurred only gradually, in part as a consequence of issues such as World War I.

Although current seizure terminology was not used in early reports, efficacy for treatment for partial and tonic-clonic seizures seemed well established. In one of the only “placebo”-controlled monotherapy trials in epilepsy, Sommerfeld-Ziskin and Ziskin (74) followed epilepsy patients after randomization to PB or placebo (diet) and found a marked decrease of seizures in the PB group after a year or more of follow-up. More recently, controlled studies of PB compared with other AEDs fail to reveal significant differences in efficacy except for the VA COOP Study #118, which indicated CBZ was somewhat more effective in preventing partial seizures (12). PB is not considered helpful in treating absence seizures, although animal models would have predicted broad-spectrum efficacy. PB has compared less favorably than most AEDs because of the occurrence of adverse effects. Safety issues are similar to those with many older AEDs, with rash appearing in 5% to 10% of patients and, rarely, much more serious problems, including Stevens-Johnson syndrome, aplastic anemia, and hepatitis. Chronic use also has been associated with connective tissue disorders, including Dupuytren's contracture, frozen shoulder, and other related conditions (59). The most common


problems associated with use of PB are those related to the central nervous system. Not surprisingly, in view of the fact that PB was first developed as a sedative, sleepiness is a common problem, especially at high doses. This is by no means invariable, and some patients tolerate quite high blood levels without complaint. Cognitive compromise not only has been a subjective complaint, but neuropsychological test batteries have consistently demonstrated some compromises compared with control subjects or patients on other standard AEDs (67,68). Behavior problems may be caused or aggravated especially in children with mental handicaps. The slow clearance and long half-life make once-daily dosing appropriate.

In summary, PB has proven to be as effective in controlling tonic-clonic seizures as any other AED, can be taken once daily, is available in parenteral formulations, and is very inexpensive. A greater prevalence of chronic adverse effects, especially of the psychological type, compared with many other AEDs has made PB a second-line AED choice except when cost is a primary consideration.


PHT was studied by Tracy Putnam in an effort to find a chemical compound possessing an antiepileptic action with efficacy comparable with or better than that of PB, but without the sedative properties. In pioneering work using a cat electroshock model he had developed, he screened a number of hydantoins supplied by Parke-Davis. Diphenylhydantoin (PHT) was found to have the desired characteristics, and together with Houston Merritt, Putnam tested the drug in patients with epilepsy with considerable success. Introduced in 1938, PHT continues to be widely used and remains the most frequently prescribed AED in the United States. Clinical comparative trials have shown PHT treatment of patients with partial and tonic-clonic seizures to have equally successful retention compared with other AEDs (12,14,28, 29, 30, 31,33). No study has found any other AED to have greater efficacy than PHT in treatment of partial and tonic-clonic seizures. In the VA COOP study #118, a significantly greater number of patients taking PHT entered into complete control at 1 year of follow-up compared with those on PB or PRM (75). On the other hand, PHT has no efficacy in control of absence, myoclonic, or atonic seizures and at times may aggravate these attacks (44,51).

Serious adverse effects are comparable with those of the other older AEDs. Rash occurs in approximately 5% to 10% of patients exposed to the drug. Rare occurrences of Stevens-Johnson syndrome, toxic epidermal necrolysis, hepatic failure, and aplastic anemia have all been reported, albeit rarely. Sedation can be experienced, especially as the dose rises. Mild mental slowing also may be noted, although carefully conducted neuropsychological test batteries do not reveal significant differences between PHT, CBZ, and VPA if doses and blood levels are in the usual range for treatment (67). Commonly, incoordination and ataxia appear as blood levels rise. Nystagmus often but not always parallels these side effects. Mental slowing and stupor appear with increasing overdose. Long-term side effects include hirsutism, which may constitute a cosmetic problem in children or women, particularly when affecting the face. Gingival hyperplasia is especially a problem in children or in adults with poor dental hygiene. Although the problem is much less evident in adults, it may occur in some patients despite scrupulous dental care. More important is the occasional development of peripheral neuropathy or cerebellar degeneration in some patients after long-term PHT therapy.

The pharmacokinetics of PHT are complex. Clearance occurs by oxidation to an inactive dihydroxy metabolite before renal excretion. The drug is metabolized primarily by the CYP4 hepatic isoenzyme system. This enzyme is saturable, resulting in rate-limiting clearance and an increasing half-life as blood concentration rises. This change at times may lead to unexpected increases or decreases in blood and brain levels with associated changes in clinical effect unless dosage is closely managed. On the other hand, the relatively long half-life makes once- or twice-daily dosing practical and enhances compliance. PHT is a potent inducer of hepatic enzymes and increases clearance of many AEDs, hormones, and medications administered for other medical problems. Awareness is necessary and appropriate increases in other medications need to be considered if PHT is coadministered.

PHT is available in an extensive number of formulations, facilitating use when rapid parenteral administration is indicated or when oral administration by tablet or capsule is not possible. It is well tolerated and is especially suitable when rapid start of treatment is indicated.

In summary, PHT is one of the oldest available AEDs and is unsurpassed in efficacy for controlling partial and tonic-clonic seizures. It also is available in many formulations and is useful for rapid loading of drug when prompt control of seizures is indicated. Its long half-life allows infrequent administration, and it is quite inexpensive. It is well tolerated at usual doses for most patients. Unfortunately, balancing these advantages are a number of undesirable characteristics. The pharmacokinetics are complex and can make dosing difficult. Adverse effects of a cosmetic type can be a problem for some patients. Of more concern is the development of long-term complications such as neuropathy or cerebellar degeneration. The introduction of many new AEDs with fewer adverse effects and more favorable pharmacokinetic properties suggests that, increasingly, PHT will be selected as one of a few drugs suitable for rapid start-up and for patients with difficult-to-control seizures.


PRM was introduced in 1954 and in open trials was found to improve seizure control especially for patients with partial epilepsy. PRM also was used for treatment of juvenile


myoclonic epilepsy with good success, particularly before the introduction of VPA (50).

Biotransformation of PRM to PB made it unclear if PRM was more than a prodrug. Considerable evidence from animal studies as well as patient trials has accumulated to indicate that when PRM is used as monotherapy, insufficient PB is found in the blood to account for all the clinical effect. PRM is primarily metabolized to PB and is cleared in approximately 8 to 12 hours, making twice-daily or three-times-daily dosing advisable to avoid peak effects. The metabolically derived PB has the same pharmacokinetic properties as PB that is used as monotherapy.

Although its adverse effects are similar to those of PB, PRM is associated with frequent dizziness, sedation, and gastrointestinal disturbance unless initiation and titration are carried out at very low doses (25 to 50 mg/day) and increased only as tolerated. The difficulty with start-up resulted in poorer retention in the VA COOP Study (12). However, once past that period, PRM was comparable in retention with CBZ, PB, and PHT (12). By some measures, both CBZ and PHT showed somewhat greater efficacy than PRM or PB (12,75).

In summary, PRM therapy is difficult to initiate and shares the long-term adverse effects of PB. It no longer is considered a first-line treatment for epilepsy, but for patients who have been successfully treated with this AED, it is reasonable to continue unless chronic adverse effects become evident. It is a first-line drug for tremor, so for patients with seizures and tremor it can treat both problems.


VPA, the newest of the older drugs, was one of the first, if not the first, AED synthesized. Valproic acid, an oily compound, was used as a solvent, and it was in this context that Meurier et al. discovered its antiepileptic property serendipitously in 1963. Use soon spread to other countries from France, and its value for treatment of a wide variety of seizure types and epilepsy syndromes soon was evident. In contrast to CBZ, PB, and PHT, the primary AEDs available at the time, VPA was highly effective against absence and myoclonic seizures as well as tonic-clonic seizures. Indeed, efficacy against most seizure types made VPA the first true broad-spectrum agent. Despite extensive experience and a consensus that VPA is in general the drug of choice for treatment of the generalized epilepsies, the only controlled, comparative clinical trials were conducted for absence seizures decades ago. In those studies, VPA and ESM were equally effective. More recent trials have found VPA to be comparably effective in preventing partial and secondarily generalized tonic-clonic seizures (13, 14, 15). CBZ is somewhat more effective for partial seizure control using some outcome measures, but as with most of the AEDs, the differences in efficacy for treatment of partial (localizationrelated) epilepsies are modest (13). VPA is especially useful in patients with coexistent migraine headache or bipolar disease.

Serious toxicity is relatively uncommon. VPA is much less likely to cause hypersensitivity rash and related problems such as Stevens-Johnson syndrome than other AEDs having an aromatic ring structure (i.e., CBZ, PB, PRM, LTG, and PHT). Other systemic toxicities can be important. Hepatic failure with potentially fatal outcome has occurred primarily in children younger than 2 years of age and taking coadministered enzyme-inducing AEDs. For adolescent patients or in adults on monotherapy, the risk is very low. Bleeding disorders occur very infrequently and are due to disturbance of several factors. A somewhat dose-related decreased platelet count usually is not of clinical importance. Rare but potentially very serious pancreatitis also has been associated with VPA. Thinning of the hair, especially at high dosage, can occur but is transient. Gastrointestinal side effects were quite common at start-up, especially with the valproic acid formulation. Delayed- or slow-release formulations are much better tolerated. The most common adverse effect is weight gain, which is reported in some series to occur in approximately half the patients (13). This adverse effect is initially primarily a cosmetic issue, but over long term might predispose the patient to the many medical comorbidities of overweight (diabetes mellitus, hypertension, elevated lipids, and arteriosclerosis). However, such a risk is conceptual and no data are available to confirm it. An association between polycystic ovary syndrome and VPA use has been reported, but the frequency is debated. Another issue of importance to women of childbearing years is the teratogenic potential of VPA. Like most other older AEDs, VPA is associated with approximately a doubling in the incidence of congenital malformations, but VPA also is specifically associated with a 1% to 2% incidence of spina bifida, a particular concern to women of childbearing age.

Neurologic adverse effects are minimal at usual dosages. As dose and blood levels increase, an action tremor commonly appears, along with some sedation, but these are reversible. Neuropsychological testing reveals minimal compromise, and results are similar to those obtained with use of CBZ (67, 68, 69).

The pharmacokinetics of VPA are somewhat complex The drug is metabolized by oxidation and may produce active metabolites. If used together with an enzyme-inducing drug such as PHT, the clearance is increased and the half-life may be only 6 to 8 hours. This may lead to use of multiple doses daily unless delayed- or slow-release formulations are used. The absence of enzyme-inducing properties makes VPA easier to use with other inducible drugs or hormones than CBZ, PB, or PHT. VPA is available in virtually every desired formulation, including a water-soluble form for rapid parenteral administration.

In summary, despite a number of adverse effects, VPA was the first truly broad-spectrum AED that had efficacy


for all the major seizure types. It remains the first drug for consideration in treatment of the idiopathic generalized epilepsies and associated absence, myoclonic, and tonic-clonic seizures, and is a reasonable alternative for all adult-onset epilepsies. Compared with the newer AEDs, it is relatively inexpensive. It is very well tolerated parenterally and is very suitable for patients requiring rapid initiation of treatment.


Limitations in Information about New Antiepileptic Drugs

Our knowledge of AEDs when they initially are made available for marketing usually is quite limited, especially since the introduction of PHT. Government regulations and economic considerations have significantly influenced the type of trials and consequently the information available when compounds are released for clinical use. The trial designs have significant limitations in providing knowledge of when and how to treat epilepsy with new drugs (Table 7.5).

Formal evaluation of the new AEDs often has been done only in placebo-controlled studies, in which the drugs were used as add-on therapy in patients with partial seizures, with or without secondarily generalized tonic-clonic seizures, refractory to optimal standard care. Some drugs have been investigated in other refractory epilepsy types such as Lennox-Gastaut syndrome, and some have been used as acute treatment against placebo in a medication withdrawal epilepsy surgery evaluation protocol. Some of the studies used a dose-ranging protocol to identify an optimally safe and effective dose. In some monotherapy studies, one group of patients is treated with one dose that is too low to expect optimal effect. Such designs provide evidence of efficacy for regulatory purposes and licensing, but do not provide clinically useful information to indicate how a new drug compares with one already in use. Knowledge about the long-term efficacy and safety of these new drugs also is limited because the blinded, controlled treatment phase in most trials lasts only 3 to 4 months.


Focus on patients with refractory epilepsy

Add-on design confounds identification of therapeutic and adverse effects attributable to new antiepileptic drug

Most studies are of partial seizures

Little or no data on:


1.    Efficacy/safety in other seizure/epilepsy types

2.    Efficacy/safety as monotherapy

3.    Long-term efficacy/safety

4.    Optimal doses

5.    Rare adverse events

The rationale for evaluating the new AEDs in various selected populations of patients with epilepsy is to allow therapeutic benefits to be readily detected and quantified in a short period. Sufficient power for statistical analysis is critical for government approval and licensing. Furthermore, the use of investigational drugs is more justifiable in patients whose seizures are poorly controlled and for whom the risk-to-benefit ratio of undefined potential toxicity is acceptable. Consequently, the efficacy studies usually indicate a modest degree of usefulness in very selected seizure and epilepsy types, and may give a limited view of the spectrum of efficacy of a new AED. For example, LTG seemed to have limited efficacy in early trials in which it was compared with placebo as an add-on therapy for patients with uncontrolled partial seizures. However, despite its modest efficacy as add-on therapy, LTG later was found to be as effective as CBZ or PHT in monotherapy trials involving patients with new-onset epilepsy (17,18).

The limited nature of the premarketing clinical trials also makes it difficult to assess the spectrum of efficacy of a new AED. For reasons previously stated, these studies primarily enroll patients with refractory partial seizures. Few trials are conducted in patients who have idiopathic generalized epilepsy with tonic-clonic, absence, or myoclonic seizures, because these individuals' seizures are more easily controlled with standard AEDs such as VPA and ESM. Based on data submitted to the U.S. Food and Drug Administration (FDA), LTG and TPM initially were approved in the United States only as adjunct therapy for partial seizures. Yet, increasing experience suggests that these drugs are effective for multiple seizure and epilepsy types (37, 38, 39, 40, 41, 42). The dissociation between approved indication and common usage was most apparent for VPA. Although widely given for almost all seizure and epilepsy types, VPA was FDA approved only for the treatment of absence seizures for 20 years.

Because the premarketing clinical trials usually are conducted in patients with highly resistant seizures who are receiving combinations of AEDs, the true efficacy of the investigational drug alone is difficult to determine. One reason for this is that the observed effects may be due to pharmacodynamic or pharmacokinetic drug-drug interactions (71). For example, blood levels of LTG vary depending on the drug(s) with which it is coadministered. CBZ and PHT enhance LTG clearance, whereas VPA has an inhibitory effect. Therefore, LTG blood levels may be severalfold higher in a patient concomitantly receiving VPA plus LTG than LTG plus PHT, even when the LTG dose is kept constant. As a result, the efficacy of LTG may appear to be quite different under these two circumstances.

The uncertainty over optimal doses of investigational drugs is based on the design of initial studies. This also confounds assessment of how these agents should be administered when they become commercially available. For example, long-term clinical experience indicates that doses of GBP and LTG used in early clinical trials were


suboptimal. In contrast, the doses used for initial TPM testing were much higher than necessary and may have contributed to more frequent reporting of poor tolerability. In its early trials, TPM was administered at doses up to 1,000 mg/day, whereas later analyses suggest that many patients obtained a good response and experienced fewer side effects when treated with TPM at 400 mg/day. As monotherapy, TPM doses as low as 100 mg/day were as effective as usual doses of CBZ or VPA in patients with new-onset epilepsy (22).

Adverse effects and tolerability of new AEDs are equally difficult to assess on the basis of data provided by the premarketing trials. Adverse effects reported in those studies may be the result of the investigational drug being evaluated, but they more likely occurred as a consequence of polytherapy. For example, dizziness was reported by 38% of patients in the add-on trials combining LTG with CBZ or PHT, but by only 8% of patients in studies of LTG monotherapy. These data indicate that LTG is well tolerated as monotherapy, and most of its neurotoxic effects can be attributed to combination treatment.

The relatively small number of patients entered into early clinical trials also precludes detection of rare or uncommon toxicities. This problem is highlighted by the late recognition of the risk of aplastic anemia and hepatic failure associated with FLB use (11).


Drug (Brand Name)




Felbamate (Felbatol)

Broad spectrum of efficacy, including Lennox-Gastaut syndrome; alerting

Rare fatal aplastic anemia and hepatitis; headache; insomnia; weight loss; drug interactions

Use limited owing to risks; inhibits metabolism of PB, PHT, and VPA: induces metabolism of CBZ

Gabapentin (Neurontin)

Effective in partial and tonic-clonic seizures; well tolerated and very safe; no known interactions

Limited absorption; short half-life; moderately limited spectrum of efficacy

Mechanism of action unknown but may enhance GABA synthesis and calcium channel function

Lamotrigine (Lamictal)

Broad spectrum of efficacy; sense of well-being; may be effective in Lennox-Gastaut syndrome

Hypersensitivity reactions occasionally severe; metabolism inducible by CBZ, PB, and PH, and inhibited by VPA

Extensive experience; excellent overall efficacy/tolerability, but requires slow dose titration

Levetiracetam (Keppra)

Effective for partial seizures; possible broad spectrum; well tolerated; no interactions

Long-term safety unknown

Delayed action in animal models, suggesting unique mechanism of action

Oxcarbazepine (Trileptal)

Very effective for partial and tonic-clonic seizures; extensive experience

Rash, hyponatremia

Pharmacodynamics similar to CBZ, but better pharmacokinetics

Tiagabine (Gabatril)

Effective in partial and tonic-clonic seizures

Undergoes hepatic metabolism; short half-life; affected by enzyme induction

Unique mechanism of action; blocks GABA reuptake

Topiramate (Topamax)

Very effective in partial and tonic-clonic seizures; may have broad spectrum of efficacy; weight loss

Possible cognitive or behavioral problems on initiation; risk of renal calculi; paresthesias; weight loss

Unique compound related to sugars; renal elimination; long half-life

Vigabatrin (Sabril)

Quite effective in partial and tonic-clonic seizures; infantile spasms; long duration of action; well tolerated

Visual field loss; may not be marketed in the USA; uncommon psychiatric symptoms (psychoses); weight gain

Unique mechanism of action; irreversibly inhibits GABA transaminase; serum levels not closely related to efficacy

Zonisamide (Zonegran)

Effective for partial and tonic-clonic seizures; perhaps broad spectrum

Renal calculi; sedation; dizziness

Related to sulfa drugs

CBZ, carbamazepine; PB, phenobarbital; PHT, phenytoin; VPA, valproate; GABA, γ-aminobutyric acid.

The AEDs described as “new” vary considerably, from those that have been extensively used worldwide for 10 years, such as OXC, to the very new AED, LEV, which had very limited exposure when first brought to market.

Each of the new AEDs has advantages and disadvantages, which are summarized in Table 7.6 and described in greater detail in the following sections. Information available concerning many of these drugs is much less extensive than that for the older standard agents. For some, such as GBP, LTG, OXC, and ZNS, however, considerable use has already occurred, and the characteristics of these drugs are increasingly well known. For LEV, TGB, and, to some degree, TPM, knowledge about efficacy and adverse effects, as well as optimal usage, still is rapidly evolving (61, 62, 63, 64, 65, 66).



FLB (Felbatol; Wallace Laboratories, Cranbury, NJ) was marketed in 1994 after add-on studies and clinical trials investigating its use as monotherapy. These studies revealed


that FLB had some efficacy in patients with partial seizures, although its potency seemed to be modest. Investigation in one controlled monotherapy trial in patients with partial-onset seizures randomized patients to FLB or low-dose VPA (76). The study design allowed demonstration of efficacy in the investigational agent without exposing control patients to unacceptable risks—that is, low-dose VPA was expected to prevent convulsive seizures and status epilepticus, but to be insufficient for controlling partial seizures. Using this “pseudoplacebo:” design, the investigators demonstrated a significant difference in efficacy. The superiority of FLB over VPA was not proven because VPA was administered in a suboptimal dose.

In a second monotherapy trial, patients who had stopped taking AEDs for presurgical recording of seizures were randomized to FLB or placebo in addition to the anticonvulsant regimen at the end of the surgical evaluation (77). In this setting, placebo treatment was thought to be justifiable because AED therapy already was discontinued for an appropriate medical reason, and an additional seizure might provide even more useful information for a decision about possible surgical treatment. In this trial, FLB again demonstrated monotherapy efficacy in seizure control, albeit for a short period, compared with placebo. These new monotherapy clinical trial decisions have been used in testing subsequent new AEDs. The efficacy of FLB also has been demonstrated in the refractory population of patients with Lennox-Gastaut syndrome. As a result, FLB has been approved for use either as adjunct therapy or monotherapy in both adults with partial epilepsy and children with Lennox-Gastaut syndrome.

Unfortunately, FLB was associated with frequent adverse effects during its clinical evaluation, especially in patients to whom it was administered as adjunctive therapy and the dosage rapidly escalated. FLB has complex pharmacokinetic interactions with other agents. It decreases CBZ and increases PHT and VPA levels, which confounded assessment of its specific pharmacodynamic effects. Headache occasionally was reported by FLB-treated patients, and led some patients to discontinue treatment. Gastrointestinal distress also was reported, and weight loss was relatively common. However, because many patients had gained weight during prior CBZ or VPA therapy, this “adverse effect” was, at times, a welcome benefit. Similarly, insomnia (or alerting) associated with FLB proved to be of considerable value to the families of many infants and children previously treated with multiple sedating AEDs for epilepsy syndromes, such as Lennox-Gastaut syndrome. Even when seizure improvement was modest, the overall alerting effect was a significant positive benefit.

Of greatest importance in FLB's safety profile was the later realization of the risk of serious idiosyncratic reactions. Aplastic anemia and liver failure developed in some patients, and approximately one-third of these individuals died. Eventually, the probability of these serious events was estimated to be 1:2,000 to 1:5,000 exposed to FLB (11). Consequently, the manufacturer and the FDA advised that the drug be used only when the benefit warranted the significant risk. Clearly, selected patients fulfill these criteria, although FLB should be used only after treatment with the standard and other new AEDs has failed. It can be argued that a trial and failure with FLB should be considered before surgery, which itself carries a 1% to 3% risk of catastrophic outcome.

FLB should be used as monotherapy whenever possible because of the complex pharmacokinetic interactions and adverse effects encountered when it is coadministered with other AEDs. It should be continued only when a significant clinical benefit is achieved. Monitoring of blood cell counts and liver function tests is advised, although there is no proof that early recognition of aplastic anemia or hepatic failure, followed by discontinuation of FLB, prevents a catastrophic outcome.

In summary, FLB is an effective broad-spectrum AED whose potential toxicity limits use to those patients failing other drugs.


Many placebo-controlled trials showed modest efficacy in treatment of partial and tonic-clonic seizures when GBP (Neurontin; Pfizer, NY) was used as an add-on to standard AED therapy (63,64). The dosage evaluated ranged from 600 to 1,800 mg/day, and these studies showed a dose-response pattern for seizure control. Based on the testing of the lower doses, the predicted effectiveness of GBP is probably less than subsequently demonstrated in studies in which doses up to 4,800 mg/day were found to be safe and effective. Open-label experience from extensive studies and other clinical trials established the long-term efficacy and safety of GBP in patients with refractory epilepsy, and indicated that such patients may benefit from and tolerate GBP doses higher than those used in the controlled clinical investigations.

GBP has been tested as monotherapy in a surgical evaluation model, in which patients were given GBP 300 or 3,600 mg/day after their third seizure. The higher dose was significantly more effective than the lower dose, providing protection against a subsequent seizure during an 8-day follow-up (78). A crossover study of monotherapy in incompletely controlled patients with partial seizures showed that in approximately 20% of patients, intractable seizures could be successfully managed with GBP monotherapy (600, 1,200, or 2,400 mg/day). However, the seizure exacerbation led many patients in this refractory epilepsy study to discontinue study participation, and no dose-response effect was identified among those who continued monotherapy (79).

A European dose-ranging GBP monotherapy trial of patients with new-onset epilepsy, however, demonstrated


superior efficacy among patients who received GBP 900 or 1,200 mg/day, compared with 300 mg/day. It was found that the overall success rate, measured by retention in the study, by both higher doses was comparable with that achieved in an unblinded parallel group that received CBZ 600 mg/day. In this study, CBZ treatment failed primarily because of adverse effects, whereas GBP provided less seizure control but better tolerability (26).

A large, multicenter, double-blind monotherapy trial of patients with new-onset epilepsy compared GBP with LTG, and no differences in retention or seizure control could be detected over a 24-week period of observation (23)

On the other hand, GBP has been evaluated for the treatment of refractory generalized seizures in a doubleblind, parallel-control trial using 1,200 mg/day and there was no benefit compared with placebo for the treatment of generalized tonic-clonic seizures, myoclonic seizures, or absence seizures (80). Anecdotal reports have suggested that, occasionally, absence seizures may increase with the use of GBP (44).

The side effect profile of GBP is very favorable. No serious systemic safety problems have arisen despite a very large patient exposure both for epilepsy and pain therapy. Indeed, it would appear that GBP is one of the safest drugs used in the entire field of medicine. However, experience based on extensive use has been available for less than a decade, and some chronic problems might eventually become evident. Other systemic adverse effects are uncommon. Gastrointestinal symptoms and weight gain occasionally are reported. Hypersensitivity reaction rash is very unusual, and seems to occur no more frequently than with the use of placebo.

Central nervous system tolerability also is excellent in most adults. Fatigue, dizziness, and ataxia, side effects commonly associated with start-up of most AEDs, have been reported early in drug administration, and seemed to be dose related. However, the frequency and severity of these complaints are relatively low. Approximately 9% of patients who received GBP in a controlled monotherapy trial withdrew because of drug-related adverse effects, a low number, not much greater than might be expected from placebo (23).

Cognitive/neuropsychological batteries have detected no impairment of performance with GBP use, whereas some psychological test results and quality-of-life measures improved after patients were crossed over from other AEDs to GBP monotherapy.

GBP also has a favorable pharmacokinetic profile. It is completely eliminated renally, and does not undergo hepatic metabolism. Consequently, in contrast to the older AEDs, there are no drug interactions to consider in patients receiving GBP. The elimination half-life of GBP is approximately 6 hours, which suggests that multiple dosing is advisable. However, there is some evidence that a more prolonged effect may occur, perhaps because of increased brain levels of GABA (72). Therefore, the duration of GBP's anticonvulsant activity may be more prolonged than might be expected based on the blood concentrations.

GBP is absorbed into the systemic circulation and enters the central nervous system by active L-amino transport. This system is saturable, and may limit the amount of drug entering the circulation from the gastrointestinal tract or the brain from the blood in any given period. The renal elimination of GBP results in higher blood concentrations and slower elimination in patients with decreased renal function. Changes in elimination are directly related to creatinine clearance. Changes in dosage can be modified according to this expected clearance, but clinical response probably is more important. Blood level determination also may be of assistance.

In summary, GBP has been shown to be an effective AED for control of partial and secondarily generalized tonic-clonic seizures and has outstanding safety. The degree of efficacy has been difficult to assess. Add-on and initial monotherapy trials often used low dosages and were conducted in refractory patients. The small improvement in control suggested modest efficacy. However, in subsequent open trials, much higher dosages were well tolerated, and these should be considered before GBP therapy is considered a failure. Although the outcomes from available monotherapy trials do not allow comparison of efficacy with older drugs like CBZ or PHT, it may be argued that its favorable side effect profile and pharmacokinetic properties make it a treatment of first choice with selected patients having partial or tonic-clonic seizures, reserving more traditional but less well tolerated drugs such as CBZ, PHT, and VPA for patients in whom GBP monotherapy fails.


LTG (Lamictal; Glazo Smith Kline, Research Triangle, NC) was one of a number of antifolate compounds developed by Wellcome, Ltd. that proved to possess broad-spectrum anticonvulsant properties in animal models and was subsequently brought to clinical trials. Dose-related efficacy has been found in multiple studies in which LTG was given as an add-on with other AEDs in patients with refractory partial epilepsy. Approximately 20% more patients receiving LTG 400 mg/day, compared with placebo, experienced a ≥50% reduction in seizures. Although these results suggest modest efficacy, it must be remembered that the drug was tested in a group of patients with refractory seizures. The trials often were conducted as add-on therapy to enzyme-inducing AEDs, so the effective blood levels of LTG were considerably less than had it been given as monotherapy.

Clinical trial experience also suggests that, like VPA, LTG has a broad spectrum of antiepileptic efficacy. In monotherapy studies, enrolling patients with new-onset epilepsy of all types, LTG was found to be as effective as CBZ or PHT and better tolerated (15,16). In an active-control study, patients with partial seizures previously


treated with CBZ or PHT were crossed to either LTG therapy (250mg twice daily) or low-dose VPA (500mg twice daily) (81). Significantly more patients were successfully maintained on LTG than VPA monotherapy. LTG also has efficacy in the treatment of absence and myoclonic seizures, as well as multiple seizures associated with the Lennox-Gastaut syndrome (37, 38, 39, 40).

Systemic events have been infrequent in patients receiving LTG, except for idiosyncratic rash. Some gastrointestinal complaints may be reported on start-up, but otherwise the drug is well tolerated. Hypersensitivity reactions have occurred in approximately 10% of patients, but the incidence can be much higher. The probability of rash or more serious reactions such as Stevens-Johnson syndrome seems to be related to the rate of administration and corresponding blood levels (54). Therefore, the dose of LTG should be titrated slowly. This is particularly important when it is coadministered with VPA, which inhibits LTG clearance and causes a rapid increase in LTG blood concentrations. The usual adult dosage is 150 to 600 mg/day.

LTG's adverse effect profile was particularly difficult to assess in the premarketing clinical trials. Among patients with partial seizures who received LTG as add-on therapy to CBZ, neurotoxic side effects, including dizziness, diplopia, and ataxia, were common and often limited LTG use unless the CBZ dose was reduced. In marked contrast, however, the use of LTG monotherapy was only infrequently associated with neurologic side effects, thus confirming that the events that occurred in the add-on trials were attributable to the combination of drugs, rather than to LTG alone. LTG seems to have no adverse effects on cognition, and may have positive effects on mood and behavior (82).

LTG is metabolized in the liver, by glucuronidation, before renal elimination (71). Its half-life is approximately 24 hours when it is used as monotherapy or together with noninteracting drugs. Metabolism is induced by CBZ, PHT, and the barbiturates, and the half-life of LTG is reduced to approximately 12 hours when it is administered concomitantly with these enzyme inducers. As noted previously, VPA inhibits the metabolism of LTG, resulting in a doubling or tripling of the half-life. Although the metabolism of LTG is affected by other older AEDs, it is not an enzyme inducer and is minimally protein bound. Consequently, coadministered drugs are not affected by its use.

The long half-life allows once- or twice-daily dosing, which enhances compliance (70).

In summary, LTG has demonstrated good efficacy as a broad-spectrum AED, possesses favorable pharmacokinetic properties as monotherapy, has excellent long-term tolerability, is relatively nonsedating, and is antidepressant. The only important limitation is a hypersensitivity reaction that is comparable to that seen with the older AEDs (i.e., CBZ, PB, PHT), but can be serious. This adverse effect notwithstanding, LTG has evolved to be a first-line choice for epilepsy therapy.


LEV (Keppra; UCB Pharmaceuticals, New Smyrna, GA) is one of the newest AEDs, so less is known about its efficacy and adverse effects. In experimental animal models, LEV has demonstrated unique properties. It has no effect on acute seizures produced by Metrazol or electroshock, but is highly effective against genetic animal or kindled models of epilepsy. These experimental studies failed to reveal mechanisms of action similar to those of other AEDs (83).

Efficacy has been demonstrated in add-on trials of partial and secondarily generalized seizures. A combination of three multicenter, double-blind, placebo-controlled, parallel-group studies proved LEV at 1,000, 2,000, or 3,000 mg/day to be statistically significantly better than placebo for all seizures combined, as well as subgroups of simple, complex, or secondarily generalized tonic-clonic seizures (84). The spectrum of activity has not been fully explored, but spike-wave patterns were significantly suppressed in genetic mouse models and photic sensitivity in a small open clinical epilepsy trial. These results suggest efficacy against generalized seizures such as absence and myoclonic seizures (42).

Pharmacokinetic studies indicate fairly prompt and complete absorption and distribution. Elimination is renal. Interaction studies have shown no effect on the metabolism of other compounds, nor the converse.

Adverse effects have been few in current clinical trials and no safety problems have arisen, although numbers of patients exposed to LEV remain relatively small, so rare idiosyncratic reactions can easily go undetected at this phase in evaluation (85). Dosage has ranged widely from 600 to 4,800 mg/day and more, if needed and tolerated.

At present, the drug is available only for oral administration.

In summary, LEV's clinical spectrum, extent of efficacy, optimal dosing, and so forth, have yet to be well defined. However, the safety and pharmacokinetic properties of LEV are especially promising.


OXC (Trileptal; Novartis) is a tricyclic AED closely related to CBZ and developed in the 1960s by Geigy Limited. Although a new drug to the United States and a number of other countries, it has been widely used in parts of Europe since its introduction in Denmark in the early 1990s. It is marketed extensively throughout the world, and considerable information is available concerning its indications and safety.

OXC has been studied in comparative active-control clinical trials and demonstrated efficacy equal to the standard drugs CBZ, PHT, and VPA (17, 18, 19) in new-onset epilepsy. When administered in usual and tolerated dosages, effects are the same as those of CBZ. Occasionally, patients have responded better to OXC on an individual basis. OXC


also has been used successfully as monotherapy for the treatment of uncontrolled partial and secondarily generalized tonic-clonic seizures (86, 87, 88). In a high-dosage (2,400 mg/day) versus low-dosage (300 mg/day) OXC trial, the high dosage was markedly more effective in preventing exit from the study than the low dosage (“pseudoplacebo”). These studies clearly established monotherapy efficacy for regulatory purposes. OXC is not effective against absence or myoclonic seizures, and like its close relative CBZ, it may at times cause aggravation of these seizure types. The usual adult dosages are 900 to 2,400 mg/day.

The adverse effects of OXC are very similar to those of CBZ. Visual disturbance and occasional sedation or gastrointestinal complaints may accompany high doses, but OXC is less likely to cause these symptoms with acute start-up. Idiosyncratic rash and related problems are comparable with those with CBZ, and a cross-sensitivity reaction occurs in approximately 25% of patients who have rash from earlier administration of CBZ (89). Hyponatremia is seen as well, although rarely of clinically significant degree. On the other hand, leukopenia, often associated with CBZ therapy, is not noted as commonly with OXC.

Although the clinical profile of efficacy and adverse effects is quite similar for OXC and CBZ, the pharmacokinetics are importantly distinctive. OXC is reduced, converting the keto to a hydroxyl group, producing monohydroxy OXC (MHD). This is the pharmacologically active metabolite. Although OXC has minimal effects on metabolism of other AEDs, it does sometimes cause elevation of PHT levels when given in high doses. Some reduction in oral contraceptive effects also has been reported. Unlike CBZ, there is no autoinduction.

OXC is available orally, in tablet form. A parenteral formulation of the water-soluble metabolite, MHD, which is the pharmacologically active metabolite, is undergoing development. This would provide a much-needed parenteral formulation for this important family of AEDs.

In summary, OXC, although closely related to CBZ, has several distinct advantages. It is equally as effective as CBZ and other standard AEDs in the treatment of partial and secondarily generalized seizures and overall is better tolerated, with fewer adverse effects. OXC's pharmacokinetics are more favorable, with fewer interactions than the standard AEDs. Clearly, this should be a drug of first choice for monotherapy and treatment of localization-related epilepsy, although like all the new AEDs, it is considerably more expensive than the older drugs.


TGB (Gabaitril; Sanofi, France) was specifically designed and synthesized to inhibit GABA reuptake and prevent seizures, unlike most AEDs, which were discovered by serendipity. Compared with some of the more extensively used new AEDs such as GBP, LTG, and VGB, TGB has been evaluated in relatively few studies. In add-on, placebo-controlled studies, TGB was associated with statistically significant efficacy (90,91). The results of a monotherapy trial suggested effectiveness by some measures, but no significant difference was found between low- and high-dose TGB, probably because efficacy was present at the low dose (92). Insufficient clinical information is available to predict the spectrum of efficacy for TGB. The usual dosage is 24 to 56 mg/day.

TGB usually is well tolerated. Treatment initiation and high dosages (>56 mg/day) occasionally elicit the usual AED-related central nervous system complaints of sedation, dizziness, tremor, and, less often, confusion. TGB also has been associated with some other cognitive complaints, although neuropsychological batteries conducted in some TGB clinical trials have failed to link its use to any significant disturbances (93). No serious safety problems have arisen.

TGB's pharmacologic properties are less favorable than most of the other new AEDs. TGB is metabolized in the liver and has a half-life of approximately 6 hours, although its clearance is even more rapid when it is administered with enzyme-inducing drugs. Nevertheless, TGB has been effective in clinical trials when administered in a twice-daily or three-times-daily regimen. TGB also is highly protein bound, although the clinical significance of any drug displacement is minimal in the absence of a meaningful blood level monitoring technique.

In summary, the role of TGB as an AED is still uncertain, but its use in its approved indication as adjunctive treatment seems to be appropriate. TGB is safe, reasonably well tolerated, and easy to administer as an adjunct; therefore, it may be an appropriate therapeutic selection in patients whose seizures do not respond adequately to standard AEDs.


TPM (Topamaz; Ortho-McNeil Pharmaceutical, Raritan, NJ), originally developed by Johnson and Johnson as an oral hypoglycemic, proved instead to be an effective AED. Clinical trials using TPM as an add-on drug for the treatment of partial and secondarily generalized tonic-clonic seizures revealed a ≥50% reduction in seizures in 40% to 50% of patients receiving TPM 200 to 400 mg/day. Little increased benefit was observed in doses up to 1,000 mg/day. A monotherapy study of crossover to 1,000 or 100 mg/day of TPM in patients with refractory partial seizure demonstrated a statistically significant advantage for the higher dose and also established efficacy as monotherapy (94). In newly diagnosed patients with epilepsy, both 100 and 200 mg/day dosages of TPM achieved success comparable with CBZ or VPA, and the 100 mg/day group experienced the best tolerability (22).

TPM also has been tested in patients with other seizure types, and preliminary evidence suggests its efficacy for


treating generalized seizures, and that it may have a broad spectrum of efficacy similar to FLB and LTG (41,95). However, there is too little information to define clearly the spectrum and potency of TPM in clinical use compared with other AEDs.

TPM has especially favorable pharmacokinetic characteristics. It is well absorbed, water soluble, not significantly protein bound, eliminated primarily by the kidneys, and has a half-life of approximately 24 hours. When given with enzyme-inducing drugs such as CBZ, PHT, or barbiturates, oxidation and glucuronidation of TPM are enhanced, and clearance is approximately 40% more rapid than when it is administered as monotherapy or coadministered with non-enzyme-inducing drugs (71).

Although trials indicated considerable potency of TPM, they also revealed multiple side effects. Some of the reported adverse events were characteristic of other AEDs, such as fatigue, gastrointestinal upset, and dizziness, and many of those were mild and self-limiting. However, a subpopulation of approximately 15% of patients experienced psychological disturbances, including impaired thinking or irritability. The exact terms for this thinking disturbance are not always easy to translate from the standardized form and terminology used. Neuropsychological testing indicates modest but clear verbal memory difficulties in some patients (96). Nonetheless, these complaints caused some TPM-treated patients to withdraw from clinical trials. Although neurologic, and especially psychological, adverse effects seemed to be relatively prominent in the TPM clinical trials, many patients received high doses (800 or 1,000 mg/day) that now are recognized to contribute to side effects without increased efficacy, and dose escalation was too rapid to allow tolerance to develop.

However, other systemic side effects were infrequent. The incidence of idiosyncratic rash was no greater than that among placebo-treated patients, and in contrast to the experience with GBP, VGB, CBZ, and VPA, weight gain was not a problem, and some patients even lost weight. Renal calculi developed in approximately 1% to 3% of patients, but most of the stones passed without adverse sequelae. This side effect emphasizes the need for adequate fluid intake by patients receiving TPM. Paresthesias are a common side effect (25% to 33% of patients on monotherapy), usually affecting the arms, but are not usually a cause for discontinuing drug. Some reports indicate TPM may cause acute-angle glaucoma, but the frequency of this problem is unclear (97).

In summary, TPM seems to be a potent and generally safe new AED for the treatment of partial and secondarily generalized seizures. Some meta-analyses of add-on trials suggest it is the most potent of the new AEDs (63). Increasing evidence indicates TPM has broad-spectrum efficacy. The fact that adverse effects occurred more frequently with TPM than with other AEDs in initial trials may, in part, be attributable to very high doses given too quickly. Weight loss may be an advantage.


In placebo-controlled clinical trials, VGB (Sabril; Bridgewater, NJ) has demonstrated clear efficacy in the treatment of partial and secondarily generalized seizures, in which 40% to 50% of patients experienced a ≥50% reduction in seizures. In a monotherapy study, comparable success rates were achieved in patients treated with VGB and CBZ; VGB was better tolerated but less effective in controlling partial seizures (24,25). Uncontrolled trials suggest that absence and myoclonic seizures are not helped by VGB, and may even increase (44).

Safety has developed as a serious concern. Visual field defects have developed in approximately 30% of patients treated with VGB. Others report the problem less frequently, but it often is asymptomatic and not readily detected (57,98). Otherwise, VGB has a favorable profile. Its use has been associated with some sedation and weight gain, but serious systemic toxicity has not been reported (63, 64, 65, 66). Information in the European literature indicates that in a small percentage of VGB-treated patients psychoses develop, particularly depression, which resolved with drug discontinuation.

The pharmacodynamic properties of VGB are especially favorable. Because of the irreversible inhibition of GABA transaminase, the effect of VGB on brain GABA levels persists long after the drug has been eliminated from the body. Although VGB is renally excreted, with a half-life of approximately 6 hours, GABA levels in the brain remain increased for ≥24 to 48 hours (72). Therefore, twice-daily and probably even once-daily administration is a reasonable treatment schedule. VGB has no known interactions with other drugs, except for slight lowering of PHT levels.

In summary, VGB is an effective drug for the treatment of partial and secondarily generalized seizures and shows promise in infantile spasms. The occurrence of visual compromise will probably limit use to those whose epilepsy severity justifies the risk.


ZNS (Zonegran; Elan Pharmaceuticals) is a sulfa compound that was first developed by Dainippon Pharmaceutical Corporation in Japan and studied initially through license to Parke-Davis. Although the drug appeared promising, the occurrence of renal calculi led to discontinuation of trials in the United States, but they were continued in Japan and the drug was marketed and used extensively both there and in a number of other countries. The drug has been found to be statistically significantly more effective than placebo in add-on trials for partial and secondarily generalized tonic-clonic seizures. The number of patients achieving 50% or greater reduction in seizures averaged approximately 33% or a little more on dosages of 400 to 800 mg/day (65,66,99). This would place it at an intermediate


level of efficacy, between GBP/LTG and VGB/TPM. Randomized, controlled monotherapy trials have not been reported despite extensive use in open trials. Some evidence of broad-spectrum efficacy is seen both in animal models and in limited trials, which have included the generalized epilepsies, with absence and myoclonic seizures. Specifically, Baltic myoclonic epilepsy has been aided by ZNS therapy (100,101).

Serious adverse effects have been infrequent. Approximately 3% of patients have had renal calculi, but these almost always have passed without invasive intervention (100). Other safety issues have been very uncommon. Some cross-reactivity has been seen in individuals allergic to sulfa drugs. Some somnolence, dizziness, and mental slowing can be seen during start-up, depending on rate of escalation and total dose. These are minimized with slow administration and limiting the dosage to necessary amounts to achieve control.

ZNS is renally excreted after conjugation, and has a half-life of 2 to 3 days as monotherapy, or 1 to 1.5 days for patients on enzyme-inducing drugs. ZNS is available in tablet form, with no parenteral formulation being available.

In summary, ZNS, like OXC, is a new compound to the U.S. market. However, extensive experience throughout the world over the past decade has allowed awareness of efficacy and adverse effects, although further trials of monotherapy and other epilepsy types are indicated (Table 7.6). It appears to have broad-spectrum AED potential.


When a decision has been made to initiate AED therapy for prevention of recurrent seizures, selection must be made from the many drugs available. No one drug of choice can be defined for any seizure or epilepsy type. Most adult-onset epilepsy is symptomatic or cryptogenic (etiology not determined) of partial (localization-related) type with partial or tonic-clonic seizures. The patient, family, and significant others, with advice and recommendations from the physician and other professionals, must decide which constellation of drug characteristics (as summarized in Tables 7.1 and 7.2) is most appropriate for each individual patient. However, CBZ has been tested against other drugs most often, has proven as effective as any other, and was somewhat more potent for treating partial seizures than some AEDs (i.e., GBP, PB, VGB, VPA). PHT has equal efficacy to CBZ but may have cumulative chronic adverse effects, as do PB and VPA. On the basis of systematic review, the Scottish Intercollegiate Guidelines Network (SIGN) considered CBZ the best selection overall (33). These guidelines were developed before the comparative trials of the newer AEDs in which both LTG and OXC were found to have better tolerability than the older drugs. One study also indicated TPM compared favorably with CBZ and VPA. Another trial has shown GBP and LTG to be comparatively successful as monotherapy. Consequently, many reasonable options now exist.

The idiopathic generalized epilepsies with tonic-clonic, myoclonic, or absence seizures most often arise in childhood or early adolescence, but both juvenile myoclonic epilepsy and grand mal on awakening may first come to attention in adult life. For many years, VPA has been the consensus drug of choice and was recommended in the SIGN Guidelines, although no randomized, comparative trials with level I evidence have been conducted (33). The introduction of LTG and TPM with broad-spectrum antiseizure properties now offers reasonable alternatives to VPA when anticipated adverse effects of VPA are undesirable. LEV and ZNS also show potential, but evidence is insufficient to make definite recommendations.

Treatment Initiation

Other factors may dictate the most appropriate AED to use in therapy. Patients presenting with serial seizures often need prompt termination of attacks. Although this may be done with a benzodiazepine, some longer-acting AED therapy is required. This is easily accomplished with the use of parenteral PHT, VPA, or PB. For patients needing to achieve control quickly but not requiring parenteral administration, PB, PHT, GBP, or VPA can be brought to effective concentrations within a day or two with oral loading. In contrast, CBZ, LTG, TGB, TPM, and ZNS are associated with poor tolerability or increased risk of idiosyncratic rash with rapid titration. OXC and LEV appear to be intermediate in tolerability for rapid initiation.

Regardless of the adverse effects at initiation, many subside in time despite increasing dose and blood levels (Figure 7.4). The patient should be encouraged to allow some time for his or her body to adjust to the drug and not abandon treatment with the onset of side effects of nonserious type.

As a consequence of these issues of tolerability, specific rates of administration are recommended for each drug. Some general principles of starting treatment and rate of administration are shown in Table 7.7. The rate should be modified on an individual basis. If tolerance is poor, the dosage should be withheld or reduced for several days or weeks. If tolerance is good, more rapid escalation provides quicker protection against recurrent seizures. After increasing the dosage to what might be recommended for an individual's age and weight, further changes can be made on the basis of clinical response. These changes should take into account the pharmacokinetic properties of the drug (102,103). On maintenance dosages of PHT, a steady state would not be reached for approximately 7 to 10 days and, for PB, for several weeks. In general, the clinician can assume a steady state will be reached no sooner than five half-lives of the drug. For CBZ, the blood levels achieved by


the first week decrease on a steady dose owing to autoinduction. The dosage subsequently is increased as needed to obtain seizure control. If CBZ is discontinued for any reason, deinduction occurs within a few days and restarting at the prior dosage results in much higher blood levels, often causing adverse effects. These metabolic changes may be especially important when CBZ is discontinued and then restarted in a presurgical/diagnostic epilepsy monitoring unit evaluation or if a patient is unable to take oral medication for whatever reason (104).


FIGURE 7.4. Functional tolerance to early side effects. Adverse events versus time on antiepileptic drug therapy, Department of Veterans Affairs Cooperative Study, 118. (From Mattson RH, Cramer JA, Collins JF. Early tolerance to antiepileptic drug side effects: a controlled trial of 247 patients. In: Koella WP, et al, eds. Tolerance to beneficial and/or adverse effects of antiepileptic drugs. New York: Raven Press, 1986:149-156, with permission.)

If rash appears, the drug should be withheld. Approximately half the instances of rash clear spontaneously and do not recur. A small percentage, however, may progress to multisystem involvement, including Stevens-Johnson syndrome, and treatment in someone who has had rash must be monitored closely, with prompt cessation of drug with any worsening. For patients with prior exposure to an AED that caused a rash, there is increased likelihood of cross-reactivity with older AEDs and probably LTG (82). Criteria for drug selection in these individuals should include a low risk of hypersensitivity. Such drugs include GBP, LEV, TGB, TPM, and VPA.


▪ Discuss plan with patient and family

▪ Use a “test” dose at bedtime

▪ If side effects, delay next dose

▪ If side effects recur, reduce dose

▪ Increase dose as tolerated

When one of the older AEDs is selected, a decision also may need to be made whether to prescribe a generic or brand-name, originator product (105,106). With new-onset epilepsy, control often is achieved without using high doses of medication (4,5,10, 11, 12, 13). Because the outcome is a clinical one, aided by AED blood levels, a properly manufactured generic should suffice in most cases. In fact, VA COOP Study #118 used generic PHT in the trial without difficulty. However, in a number of circumstances a generic product (or variable suppliers of the product) is not recommended. When seizures are more difficult to control, dosages need to be increased near or to the point of poor tolerability. In such situations, small fluctuations in bioavailability may be clinically quite important. CBZ and PHT are two AEDs that may have little room for variability. Small changes in AED levels may lead to side effects or loss of control. For products that have a short half-life, such as CBZ or VPA, extended-release formulations may prevent peak/trough effects that also can lead to adverse effects or breakthrough seizures. Valproic acid often causes more gastrointestinal side effects than a brand-name delayed-absorption VPA formulation. Other brand-name products may be important for some patients. Thus, a sprinkle, chewable, or liquid form is useful for those who cannot or will not swallow a tablet or capsule. Brand-name fosphenytoin is a much better tolerated parenteral formulation than generic PHT.

Older versus New Antiepileptic Drugs

When selecting an AED for initiating therapy (or as alternative therapy), it is unclear what place the new AEDs


should occupy. No evidence exists that the newer compounds possess greater efficacy than the older ones, but pharmacokinetic properties are improved and, in some, tolerability and safety appear to be better. On the other hand, they usually are much more expensive. Consequently, unless there is a particular reason to select one of the newer drugs for starting therapy, CBZ, PB, PHT, and VPA are most appropriate for partial and secondarily generalized seizures associated with symptomatic epilepsy. ESM is indicated for pure absence seizures, and VPA for seizures of tonic-clonic, absence, and myoclonic type found in the idiopathic generalized epilepsies. When adverse effects or pharmacokinetic characteristics of the older drugs are undesirable for an individual patient, it is reasonable to select one of the newer AEDs.


Any problem after initiation of therapy is reason to schedule a visit, urgently if some systemic problem is suspected. Minor issues such as dose changes often can be dealt with by telephone (but documentation of the discussion should be made in the records). If logistically possible, a return visit within a few weeks or a month may be useful even if no problems have arisen. This presents further opportunity to provide patient education (preferably with a significant other as well) so the patient can knowledgeably collaborate in management of what quite possibly will be a long-term condition. Although much information may have been discussed at an initial visit, many facts may not have been heard, were misinterpreted, or were subsequently contradicted by readings or comments from family and friends.

Antiepileptic Drug Monitoring

Monitoring the concentration of AEDs in the blood has proven to be a valuable supplement to patient care, especially for compounds with complex pharmacokinetic properties like CBZ or PHT. There are specific times and situations when determination of blood levels is most useful (Table 7.8). On the other hand, routine monitoring was not found to yield better outcomes in patients whose physicians had such information compared with patients managed only by clinical response (107). Experience and open studies have indicated that there are blood AED concentrations most often associated with control and freedom from adverse effects, but these are based on population statistics, and do not apply to each individual patient. Consequently, most guidelines suggest concentrations of AEDs for effectiveness and freedom from side effects, yet many individuals obtain good control on low, or even “subtherapeutic” concentrations, and others may require and tolerate levels much higher. This principle applies to most, if not all, AEDs, and emphasizes that clinical outcome is the primary measure determining drug dosage. The desirable levels of drug with use of the new AEDs are not established, but there is no reason to believe they will not be as useful as has been proven true for the older drugs (108, 109, 110). The initial target concentrations of 2 to 5 µg/mL for drugs like GBP, LTG, and TPM probably were too low, and clinical experience suggests that maximal control while maintaining reasonable tolerability may not be realized until concentrations are in the 10- to 20-µg/mL range.


▪ After starting drug at steady state

▪ When adding or subtracting an interacting drug

▪ When side effects are occurring

▪ When seizures break through

▪ Periodically to assess compliance

Safety Monitoring

In addition to obtaining blood AED levels, some testing for safety monitoring may be advisable. Initial evaluation of the cause of the epilepsy often will have included complete blood counts, liver function tests, blood urea nitrogen, and serum glucose and electrolytes. These values can serve as a baseline before initiating treatment. The utility of repeating these determinations on any regular basis is controversial (111). If the safety risks are relatively high, as with use of FLB, the manufacturer recommends regular monitoring. For rare idiosyncratic aplastic anemia, hepatitis, nephritis, or pancreatitis, it is unlikely regular testing is cost effective or clinically helpful. Rather, history and physical examination should provide evidence of systemic adverse effects to be further investigated with blood testing (112).


Optimal drug selection, fine tuning of dosing based on pharmacokinetic properties, avoidance of adverse effects, and selection of the most suitable formulation all are of no value in seizure control if the patient does not take the medication as prescribed. The reasons for failure to take the AED are multiple and include denial of illness, insufficient education by medical personnel, subtle or expected adverse effects, complexity of regimen, and simple forgetfulness (112). Using electronic monitoring methods, Cramer et al. (70) found compliance fell off considerably with three- or four-times-daily prescribing, and whenever possible twiceor once-daily administration should be used. For drugs with a rapid absorption and short half-life, a slow-release formulation may allow such dosing and avoid peak/trough effects while enhancing compliance. Containers holding doses for each day or week as well as fitting intake to a fixed daily activity (e.g., meals, washing) also may be useful. Asking the patient when he or she takes the medication may reveal


vagueness warranting further education. AED blood levels at clinic visits that are in the target range and refilling prescriptions at appropriate times are some indication of compliance, but do not mean medication is taken as prescribed. Patients may skip and later double-up on doses, a practice that may lead either to seizures or adverse effects despite the appearance of compliance. If breakthrough seizures occur, obtaining blood AED levels as soon as possible helps to determine if the drug or the patient has failed to maintain seizure control and possibly avert unnecessarily changing to another AED.


  1. Berg AT, Shinnar S. The risk of seizure recurrence following a first unprovoked seizure: a quantitative review. Neurology1991; 41:965-972.
  2. Hauser WA, Rich SS, Lee JR, et al. Risk of recurrent seizures after two unprovoked seizures. N Engl J Med1998;338:429-434.
  3. Commission on Classification and Terminology of the International League Against Epilepsy. Proposal for revised clinical and electroencephalographic classification of epileptic seizures. Epilepsia1981;22:489-501.
  4. Commission on Classification and Terminology of the International League Against Epilepsy. Proposal for revised classification of epilepsies and epileptic syndromes. Epilepsia1989;30: 389-399.
  5. Elwes R, Johnson AL, Shorvon SD, et al. The prognosis for seizure control in newly controlled epilepsy. N Engl J Med1984;311:944-947.
  6. Kwan P, Brodie MJ. Effectiveness of the first antiepileptic drug. Epilepsia2001;42:1255-1261.
  7. Mattson RH, Cramer JA, Collins JF, and the Department of Veterans Affairs Epilepsy Cooperative Studies No. 118 and No. 264 Groups. Prognosis for total control of complex partial and secondarily generalized tonic-clonic seizures. Neurology1996; 47:68-76.
  8. Cramer JA, Mattson RH. Quantitative approaches to seizure severity. In: Meinardi H, Cramer JA, Baker GA, et al., eds. New York: Plenum Press, 55-71.
  9. Baker GA, Camfield C, Camfield P, et al. ILAE Commission report: Commission on Outcome Measures in Epilepsy 1994-1997. Epilepsia1998;39:213-231.
  10. Mattson RH. Monotherapy trials: endpoints. Epilepsy Res2001; 45:1109-1117.
  11. Pellock JM, Brodie MJ. Felbamate: 1997 update. Epilepsia1997;38:1261-1264.
  12. Mattson RH, et al. Comparison of carbamazepine, phenobarbital, phenytoin, and primidone in partial and secondarily generalized tonic/clonic seizures. N Engl J Med1985;313:145-151.
  13. Mattson RH, Cramer JA, Collins JF, et al. A comparison of valproate with carbamazepine for the treatment of complex partial seizures and secondarily generalized tonic/clonic seizures in adults. N Engl J Med1992;327:765-771
  14. Heller AJ, Chesterman P, Elwes RD, et al. Phenobarbitone, phenytoin, carbamazepine or sodium valproate for newly diagnosed adult epilepsy: a randomized comparative monotherapy trial. J Neurol Neurosurg Psychiatry1995;58:44-50.
  15. Richens A, Davidson DL, Cartlidge NE, et al. A multicentre comparative trial of sodium valproate and carbamazepine in adult onset epilepsy. J Neurol Neurosurg Psychiatry1994;57: 682-687.
  16. Sachdeo R, Kramer PLD, Rosenberg A, et al. Felbamate monotherapy for partial onset seizures. Ann Neurol1992;32: 386-392.
  17. Brodie MJ, Richens A, Yeun AW. Double-blind comparison of lamotrigine and carbamazepine in newly diagnosed epilepsy. Lancet1995;345:476-479.
  18. Steiner TJ, Dellaportas CI, Findley LJ, et al. lamotrigine monotherapy in newly diagnosed untreated epilepsy: a doubleblind, randomized comparison with phenytoin. Epilepsia1999; 40:601-607.
  19. Dam M, Ekberg R, Loyning Y, et al. A double-blind study comparing oxcarbazepine and carbamazepine in patients with newly diagnosed, previously untreated epilepsy.Epilepsy Res1989;3:70-76.
  20. Christe W, Gunter K, Vigonius U, et al. A double-blind controlled clinical trial: oxcarbazepine versus sodium valproate in adults with newly diagnosed epilepsy. Epilepsy Res1997;26: 451-460.
  21. Bill PA, Vigonius U, Pohlmann H, et al. A double-blind controlled clinical trial of oxcarbazepine versus phenytoin in adults with previously untreated epilepsy. Epilepsy Res1997;27:195-204.
  22. Privitera MD, Brodie MJ, Mattson RH, et al. Topiramate, carbamazepine and valproate monotherapy: a double-blind comparison in the spectrum of newly diagnosed epilepsy (in press).
  23. Brodie MJ, Chadwick DW, Anhut H, et al., and the Gabapentin Study Group 945-212. Gabapentin versus lamotrigine: a double-blind comparison in newly diagnosed epilepsy.Epilepsia (in press).
  24. Kalvianien R, Aikia M, Saukkonen AM, et al. Vigabatrin vs carbamazepine monotherapy in patients with newly diagnosed epilepsy. Arch Neurol1995;52:989-996.
  25. Chadwick DW, and the Vigabatrin European Monotherapy Study Group. Safety and efficacy of vigabatrin and carbamazepine in newly diagnosed epilepsy: a multicentre, randomized double-blind study. Lancet1999;354:13-19.
  26. Chadwick DW, Anhut H, Greiner MJ, et al. A double-blind trial of gabapentin monotherapy for newly diagnosed partial seizures. Neurology1998;51:1282-1288.
  27. Brodie MJ, Overstall PW, Giorgi L, et al. Multicentre, doubleblind, randomized comparison between lamotrigine and carbamazepine in elderly patients with newly diagnosed epilepsy. Epilepsy Res1999;37:81-87.
  28. Callaghan N, Kenny RA, O'Neill B, et al. A prospective study between carbamazepine, phenytoin and sodium valproate as monotherapy in previously untreated and recently diagnosed patients with epilepsy. J Neurol Neurosurg Psychiatry1985;48: 639-644.
  29. Turnbull DM, Rawlins MD, Weightman D, et al. A comparison of phenytoin and valproate in previously untreated adult epileptic patients. J Neurol Neurosurg Psychiatry1982;45:55-59.
  30. Ramsay RE, Wilder BJ, Berger JR, et al. A double-blind study comparing carbamazepine with phenytoin as initial seizure therapy in adults. Neurology1983;33:904-910.
  31. Wilder BJ, Ramsay E, Murphy JV, et al. Comparison of valproic acid and phenytoin in newly diagnosed tonic-clonic seizures. Neurology1983;33:1474-1476.
  32. Mattson RH, Cramer JA. Crossover from polytherapy to monotherapy in primary generalized epilepsy. Am J Med1988; 84:23-28.
  33. Scottish Intercollegiate Guidelines Network (SIGN). Diagnosis and management of epilepsy in adults.Edinburgh: SIGN, 1997.
  34. Sherwin AL. ESM: clinical use. In: Levy RH, Mattson RH, Meldrum BS, eds. Antiepileptic drugs,4th ed. New York: Raven Press, 1995:667-673.
  35. Sato S, White BG, Penry JK, et al. Valproic versus ESM in the treatment of absence seizures. Neurology1982;32:157-153.



  1. Villareal HJ, Wilder BJ, Willmore LJ, et al. Effect of valproic acid on spike-wave discharges in patients with absence seizures. Neurology1978;28:886-891.
  2. Buoni S, Grosso S, Fois A. Lamotrigine in typical absence epilepsy. Brain Dev1999;21:303-306.
  3. Gericke CA, Picard F, de Saint-Martin A, et al. Efficacy of lamotrigine in idiopathic generalized epilepsy syndromes: a video-EEG-controlled, open study. Epileptic Disord1999;1:159-165.
  4. Frank LM, Enlow T, Holmes GL, et al. Lamictal (lamotrigine) monotherapy for typical absences in children. Epilepsia2000; 41:357-359.
  5. Panayiotopoulos CP. Treatment of typical absence seizures and related epileptic syndromes. Paediatr Drugs2001;3:379-403.
  6. Wheless JW. Use of topiramate in childhood generalized seizure disorders. J Child Neurol2000; 15[Suppl 1]:S7-S13.
  7. French JA. The role of the new antiepileptic drugs. Am J Manag Care2001;7[Suppl]:S209-S214.
  8. Shields WD, Saslow E. Myoclonic, atonic and absence seizures following institution of carbamazepine therapy in children. Neurology1983;33:1487-1489.
  9. Perucca E, Gram L, Avanzini G, et al. Antiepileptic drugs as a cause of worsening seizures. Epilepsia1998;39:5-17.
  10. Convanis A, Gupta AK, Jeavons PM. Sodium valproate: monotherapy and polytherapy. Epilepsia1982;23:693-720.
  11. Delgado-Escueta AV, Enrile-Bascal F. Juvenile myoclonic epilepsy of Janz. Neurology1984;34:285-294.
  12. Bourgeois B, Beaumanoir A, Blasjev B, et al. Monotherapy with valproate in primary generalized epilepsies. Epilepsia1987;28 [Suppl 2]:S8-S11.
  13. Penry JK, Dean JC, Riela AR. Juvenile myoclonic epilepsy: long term response to therapy. Epilepsia1989;30[Suppl 14]: 519-523.
  14. Panayiotopoulos CP, Obeid T, Tahan AR. Juvenile myoclonic epilepsy: a 5-year prospective study. Epilepsia1994;35: 285-296.
  15. Janz D, Durner M. Juvenile myoclonic epilepsy. In: Engel J, Pedley TA, eds. A textbook of epilepsy.Philadelphia: Lippincott-Raven, 1997:2389-2400.
  16. Genton P, Gelisse P, Thomas MD, et al. Do carbamazepine and phenytoin aggravate juvenile myoclonic epilepsy? Neurology2000;55:1106-1109.
  17. Biraben A, Allain H, Scarabin JM, et al. Exacerbation of juvenile myoclonic epilepsy with lamotrigine. Neurology2000;55: 1758.
  18. The Felbamate Study Group in Lennox-Gastaut Syndrome. Efficacy of felbamate in childhood epileptic encephalopathy (Lennox-Gastaut syndrome). N Engl J Med1993;328:29-33.
  19. Messenheimer J, Mullens EL, Giorgi I, et al. Safety review of adult clinical trial experience with lamotrigine. Drug Saf1998; 18:281-296.
  20. Kenneback G, Bergfeldt L, Vallin H, et al. Electrophysiologic effects and clinical hazards of carbamazepine treatment for neurologic disorders in patients with abnormalities of the cardiac conduction system. Am Heart J1991;121:1421-1429.
  21. Mattson RH. Carbamazepine. In: Engel J, Pedley TA, eds. A textbook of epilepsy.Philadelphia: Lippincott-Raven, 1997.
  22. Sato Y, Kondo I, Ishida S, et al. Decreased bone mass and increased bone turnover with valproate therapy in adults with epilepsy. Neurology2001;57:445-449.
  23. Mattson RH, Cramer JA, McCutchen CM, and the VA Cooperative Epilepsy Study Group. Barbiturate related connective tissue disorders Arch Intern Med1989;149:911-914.
  24. Eke T, Talbot JF, Lawden MC. Severe persistent visual field constriction associated with vigabatrin. BMJ1997;314:180-181.
  25. Mattson RH, Cramer JA, Collins JF. Early tolerance to antiepileptic drug side effects: a controlled trial of 247 patients. In: Koella WP, et al., eds. Tolerance to beneficial andor adverse effects of antiepileptic drugs.New York: Raven Press, 1986:149-156.
  26. Bialer M, et al. Progress report on new antiepileptic drugs: a summary of the Fourth Eilat Conference (EILAT IV). Epilepsy Res1999;34:1-41.
  27. Bialer M, Johannessen SI, Kupferberg HJ, et al. Progress report on new antiepileptic drugs: a summary of the Fifth Eilat Conference (EILAT V). Epilepsy Res2001;43:11-58.
  28. Marson AG, et al. New antiepileptic drugs: a systematic review of their efficacy and tolerability. BMJ1996;313:1169-1174.
  29. Cramer JA, et al. New antiepileptic drugs: comparison of key clinical trials. Epilepsia1999;40:590-600.
  30. Marson AG, Hutton JL, Leach JP, et al. Levetiracetam, oxcarbazepine, remacemide and zonisamide for drug resistant localization-related epilepsy: a systematic review.Epilepsy Res2001; 46:259-270.
  31. Cramer JA, Ben-Menachem EB, French J. Review of treatment options for refractory epilepsy and vagal nerve stimulation. Epilepsy Res2001;47:17-25.
  32. Meador KJ, Loring DW, Hun K, et al. Comparative cognitive effects of anticonvulsants. Neurology1990;40:391-394.
  33. Vining EPG, Mellits FD, Dorsen MM, et al. Psychologic and behavioral effects of antiepileptic drugs in children: a doubleblind comparison between phenobarbital and valproic acid. Pediatrics1987;80[Suppl 1]:165-174.
  34. Prevey ML, Delaney RC, Cramer JA, et al. The effect of valproate on cognitive functioning: a comparison with carbamazepine. Arch Neurol1996;53:1008-1016.
  35. Cramer JA, Mattson RH, et al. How often is medication taken as prescribed? A novel assessment technique. JAMA1989;261: 3273-3277.
  36. Perucca E. The clinical pharmacokinetics of the new antiepileptic drugs. Epilepsia1999;40[Suppl 9]:S7-S13.
  37. Kocsis JD, Mattson RH. GABA levels in the brain: a target for new antiepileptic drugs. Neuroscientist1996;2:326-334.
  38. Mattson RH. Selection of antiepileptic drug therapy. In: Levy RH, Mattson RH, Meldrum B, eds. Antiepileptic drugs,4th ed. New York: Raven Press, 1995:123-135.
  39. Sommerfeld-Ziskin E, Ziskin E. Effect of phenobarbital on the mentality of epileptic patients. Arch Neurol Psychiatry1940;43: 70-79.
  40. Mattson RH, Cramer JA, and the VA Cooperative Study Group. Seizure remission after active epilepsy. Epilepsia1990; 31:648(abstr).
  41. Faught E, Sachdeo RC, Remler MP, et al. Felbamate monotherapy for partial onset seizures: an active control trial. Neurology1993;41:1785-1789.
  42. Bourgeois B, Leppik IE, Sackellares JC, et al. Felbamate: a double-blind controlled trial in patients undergoing presurgical evaluation in partial seizures. Neurology1993;43:693-696.
  43. Bergey GK, Morris HH, Rosenfeld W, et al., and the U.S. Gabapentin Study Group 88/89. Gabapentin monotherapy: I. an 8-day, double-blind, dose-controlled, multicenter study in hospitalized patients with refractory complex partial and secondarily generalized seizures. Neurology1997;49:739-745.
  44. Beydoun A, Fischer J, Labar DR, et al., and the U.S. Gabapentin Study Group 82/83. Gabapentin monotherapy: II. a 26-week, double-blind, dose-controlled, multicenter study of conversion from polytherapy in outpatients with refractory complex partial or secondarily generalized seizures. Neurology1997;49:746-752.
  45. Chadwick DW, Liederman DB, Sauerman W, et al. Gabapentin in generalized seizures. Epilepsy Res1996;25:191-197.
  46. Gillam F, Vasquez B, Sackellares JC, et al. An active-control trial of lamotrigine monotherapy for partial seizures. Neurology1998;51:1018-1025.



  1. Edwards KR, Sackellares JC, Vuong A, et al. Lamotrigine monotherapy improves depressive symptoms in epilepsy: a double-blind comparison with valproate. Epilepsy Behav2001;2: 28-36.
  2. Klitgard H, Matagne A, Gobert J, et al. Evidence for a unique profile of levetiracetam in rodent models of seizures and epilepsy. Eur J Pharmacol 1998;353:191-206.
  3. Privitera M. Efficacy of levetiracetam: a review of three pivotal trials. Epilepsia2001;42[Suppl 4]:S31-S35.
  4. French J, Edrich P, Cramer JA. A systematic review of the safety profile of levetiracetam: a new antiepileptic drug. Epilepsy Res2001;47:77-90.
  5. Beydoun A, Sachdeo RC, Rosenfeld WE, et al. Oxcarbazepine monotherapy of partial-onset seizures: a multicenter, doubleblind, clinical trial. Neurology2000;54:2245-2251.
  6. Sachdeo R, Beydoun A, Schacter S, et al. Oxcarbazepine (Trileptal) as monotherapy in patients with partial seizures. Neurology2001;57:864-871.
  7. Schachter SC, Vasquez B, Fisher RS, et al. Oxcarbazepine: a double-blind, randomized, placebo-control, monotherapy trial for partial seizures. Neurology1999;52:732-737.
  8. Beran RG. Cross-reactive skin eruptions with both carbamazepine and oxcarbazepine. Epilepsia1993;34:163-165.
  9. Kalviainen R, Brodie MJ, Duncan J, et al. A double-blind, placebo-controlled trial of TGB given three-times daily as add-on therapy for refractory partial seizures: Northern European TGB Study Group. Epilepsy Res1998;30:31-40.
  10. Uthman BM, Rowan AJ, Ahmann PA, et al. TGB for complex partial seizures: a randomized, add-on, dose-response trial. Arch Neurol1998;55:56-62.
  11. Schacter SC. TGB monotherapy in the treatment of partial epilepsy. Epilepsia1995;36[Suppl 6]:S2-S6.
  12. Dodrill CB, Arnett JL, Sommerville KW, et al. Cognitive and quality of life effects of differing dosages of TGB in epilepsy. Neurology1997;48:1025-1031.
  13. Sachdeo RC, Reife RA, Lim P, et al. Topiramate monotherapy for partial onset seizures. Epilepsia1997;38:294-300.
  14. Biton V, et al. A randomized, placebo-controlled study of topiramate in primary generalized tonic-clonic seizures. Neurology1999;52:1330-1337.
  15. Aldenkamp AP, Baker G, Mulder OG. et al. A multicenter clinical study to evaluate the effect on cognitive function of topiramate compared with valproate as add-on therapy to carbamazepine in patients with partial-onset seizures. Epilepsia2000; 41:1167-1178.
  16. Banta JT, Hoffman K, Budenz DL, et al. Presumed topiramate-induced bilateral acute angle-closure glaucoma. Am J Ophthalmol2001;132:112-114.
  17. Krauss GI, Johnson MA, Miller NR. Vigabatrin associated retinal cone system dysfunction: electroretinogram and ophthalmologic findings. Neurology1998;50:614-618.
  18. Faught E, Ayala R, Montouris GG, et al. Randomized controlled trial of zonisamide for the treatment of refractory partial-onset seizures. Neurology2001;57:1774-1779.
  19. Henry TR, Leppik IE, Gumnit RJ, et al. Progressive myoclonic epilepsies treated with zonisamide. Neurology1988;38: 928-931.
  20. Kyllerman M, Ben-Menachem E. Zonisamide for progressive myoclonus epilepsy. Epilepsy Res1998;29:109-114.
  21. Perucca E, Dulac O, Shorvon S, et al. Harnessing the clinical potential of antiepileptic drug therapy: dosage optimization. CNS Drugs2001; 15:609-621.
  22. Porter RJ. How to use antiepileptic drugs. In: Levy RH, Mattson RH, Meldrum BS, eds. Antiepieptic drugs,4th ed. New York: Raven Press, 1995:137-148.
  23. Spencer SS, Packey DJ. Antiepileptic drug management before and after epilepsy surgery. In: Levy RH, Mattson RH, Meldrum BS, eds. Antiepileptic drugs,4th ed. New York: Raven Press, 1995:189-200.
  24. Crawford P, Hall WW, Chapell B, et al. Generic prescribing for epilepsy: is it safe? Seizure1996;5:1-5.
  25. Nuwer MR, Browne, TR, Dodson WE, et al. Generic substitutions for antiepileptic drugs. Neurology1990;40:1647-1651.
  26. Jannuzzi G, Cian P, Fattore C, et al. A multicenter randomized controlled trial on the clinical impact of therapeutic monitoring in patients with newly diagnosed epilepsy: the Italian TDM Study Group in Epilepsy. Epilepsia2000;41:222-230.
  27. Perucca E. Is there a role for therapeutic drug monitoring of new anticonvulsants? Clin Pharmacokinet 2000;38:191-204.
  28. Tomson T, Johannessen SI. Therapeutic monitoring of the new antiepileptic drugs. Eur J Clin Pharmacol2000;55:697-705.
  29. Mattson RH. Antiepileptic drug monitoring: a reappraisal. Epilepsia1995;36[Suppl 5]:S22-S29.
  30. Pellock JM, Willmore LJ. A rational guide to routine blood monitoring in patients receiving antiepileptic drugs. Neurology1991;41:961-964.
  31. Cramer JA, Mattson R. Compliance with antiepileptic drug therapy. In: Levy RH, Mattson RH, Meldrum BS, eds. Antiepileptic drugs,4th ed. New York: Raven Press, 1995: 149-159.