Efficacy, tolerability, and safety of rapid initiation of topiramate versus phenytoin in patients with new-onset epilepsy: A randomized double-blind clinical trial


  • Clinical trial registration: ClinicalTrials.gov identifier: NCT00210782. (Study identifier: CR004663) A Comparison of the Effectiveness and Safety of Topiramate and Phenytoin in Subjects with New Onset Epilepsy Requiring Rapid Initiation of Antiepileptic Drug Treatment.

Address correspondence to Eugene Ramsay, MD, Epilepsy Institute, Ochsner Baptist Medical Center, 4299 Clara Street, MacFarland Building, Suite 440, New Orleans, LA 70115. E-mail: eramsay@epiworld.com


Purpose: To evaluate topiramate (TPM) and phenytoin (PHT) monotherapy following rapid oral initiation in new-onset epilepsy.

Methods: Randomized, double-blind, 28-day trial of TPM (100 mg/day beginning on day 1) versus PHT (1,000 mg on day 1 followed by 300 mg/day maintenance dosing) in 261 patients with new-onset epilepsy. The primary end point was time to seizure, and the primary objective was to establish noninferiority of TPM to PHT in the risk of seizure.

Results: At day 28, the estimated seizure-free rate was 81.1% for TPM treatment in comparison with 90.3% for PHT treatment. Noninferiority of TPM to PHT (primary objective) could not be established [hazard ratio (HR) 2.0, 95% confidence interval (CI), 0.98 to 4.12, p = 0.366), and PHT could not be shown to be superior to TPM. A higher percentage discontinued with PHT compared to TPM for all reasons (21.1 vs. 12.8%) and due to adverse events (13.4 vs. 6.8%). The most common treatment-related adverse events in both groups were dizziness, paresthesia, and somnolence. A post hoc analysis showed that TPM was superior to PHT in time to discontinuation (retention rate) for all causes (89.4% vs. 80.3%, p = 0.047).

Conclusion: This study was inconclusive in establishing noninferiority of TPM 100 mg/day compared to a standard regimen of oral PHT in seizure risk in this population of patients with new-onset epilepsy. Given the superiority of TPM in overall retention and favorable tolerability without titration, it may nonetheless be an appropriate option in some patients with new-onset epilepsy requiring rapid treatment initiation.

Clinicians faced with a patient presenting with newly emergent unprovoked seizures must determine whether the risk of further seizures warrants the initiation of antiepileptic drug (AED) therapy. The weeks and months following a second unprovoked seizure are a period of high risk for additional seizures (Hauser et al., 1998), and risk of recurrence is increased by certain factors, such as symptomatic etiology (Hauser et al., 1998; Ramos Lizana et al., 2000; Berg, 2008). Overall, 40–50% of individuals experiencing a first unprovoked seizure will experience a recurrence within 2 years, and effective treatment can reduce this risk by as much as half (Berg, 2008).

Once an AED has been initiated, it may be desirable to rapidly achieve effective blood levels. Titration over days or weeks may not be appropriate if the imminent risk of recurrent seizures outweighs the increased rate and severity of adverse events (AEs) that are potentially associated with rapid initiation of an AED. Oral loading with levetiracetam 1,500 mg/day was found to be well-tolerated, with only 11% of patients experiencing AEs (transient irritability, imbalance, tiredness, lightheadedness) within the first few hours after loading (Koubeissi et al., 2008). Rapid initiation of gabapentin 900 mg/day compared with a 3-day titration period resulted in an increased incidence of only one (dizziness) of the four AEs most commonly associated with gabapentin (somnolence, dizziness, ataxia, fatigue) during the first week of therapy (Fisher et al., 2001). Rapid oral loading of carbamazepine 8 mg/kg per day in the emergency department appeared successful, in that 93% of patients presenting with subtherapeutic serum levels of carbamazepine achieved therapeutic levels within 3 h after loading; however, 58% of patients experienced AEs, the most common of which were drowsiness (26%) and nausea (23%) (Purcell et al., 2007). Rapid initiation of therapy without titration may not be possible with some AEDs, owing to safety or tolerability concerns. Even moderately rapid titration of lamotrigine, for example, increases the risk of allergic skin reaction, and also the risk of the rash being severe (Guberman et al., 1999; Wong et al., 1999). Whereas the previously mentioned studies were mainly performed to evaluate the safety and tolerability profile of these second-generation AEDs, there is limited information regarding the initial effectiveness of rapid initiation.

Phenytoin (PHT) has become a frequently used AED in the initiation of epilepsy treatment in the United States due, in part, to the availability of an intravenous formulation with improved pharmacokinetics (Cloyd et al., 1980). A frequent approach is to use an oral loading dose of 1,000 mg administered in three doses, 400 mg followed by two 300 mg doses at 2-h intervals, with maintenance dosing (300 mg/day) beginning 24 h after the loading dose (Wilder et al., 1973; Dilantin, 2009). Side effects with oral loading doses of PHT—which include nausea, vomiting, dizziness and drowsiness—are frequent, and have been described as probably inevitable but well tolerated and outweighed by the risk of further seizures (Osborn et al., 1987; Van Der Meyden et al., 1994; Wilder et al., 1973). There is, however, considerable interpatient variability in PHT protein binding and thus free PHT blood concentrations (Banh et al., 2002). There is also variability in the plasma concentration needed to maintain adequate seizure control (Cloyd et al., 1980; Schmidt et al., 1986; Schmidt & Haenel, 1984). Furthermore, there is little information available about how effectively seizures are controlled in the early period following rapid initiation of PHT, or, indeed, other AEDs in the treatment of new-onset epilepsy. Finally, PHT is not an ideal agent for long-term therapy as it has not been shown to significantly reduce the incidence of late-seizures and because of its toxic effects (Aronson et al., 1992; Temkin et al., 1990).

Topiramate is a structurally unique compound that is an effective anticonvulsant with a good safety profile after oral administration (Latini et al., 2008). Side effects associated with rapid introduction of TPM have been largely mild-to-moderate central nervous system (CNS) effects (Biton et al., 2001). Topiramate has demonstrated efficacy as monotherapy in patients with newly diagnosed or recently diagnosed epilepsy (Arroyo et al., 2005; Gilliam et al., 2003; Glauser et al., 2007).

The recommended dose of TPM monotherapy is 400 mg/day in two divided doses, initiated at a divided daily dose of 50 mg with titration up to 400 mg/day over 6 weeks. The controlled study of TPM as first-line monotherapy for epilepsy randomized patients to receive target doses of either 50 or 400 mg/day, and the primary end point was the comparison of time to first seizure. In this study, many patients remained seizure-free on doses lower than 400 mg/day: Approximately 58% of patients randomized to 400 mg/day achieved this maximal dose; the mean dose achieved in the trial was 275 mg/day (Arroyo et al., 2005; Ortho-McNeil Topamax PI). In addition, analysis of exits due to seizure recurrence during titration demonstrated a statistically significant separation by the time patients in the high-dose arm had achieved a daily dose of 100 mg. Another randomized, double-blind, controlled study compared TPM 100 mg and 200 mg/day as monotherapy with the investigators’ choice of carbamazepine or valproate in patients with newly diagnosed epilepsy. The TPM 100 mg/day dose did not differ from the other AEDs in time to discontinuation or time to first seizure and was associated with the lowest rate of discontinuation due to adverse events (Privitera et al., 2003).

The objective of this trial (Study identifier: CR004663) was to evaluate the efficacy, tolerability, and safety during the first month of monotherapy treatment with TPM (target dose of 100 mg/day beginning on day 1 of treatment) versus treatment with PHT sodium extended release administered according to a rapid oral initiation schedule (1,000 mg on day 1 followed by an initial maintenance dose of 300 mg/day) in patients with seizures indicative of new-onset epilepsy requiring rapid initiation of oral AED therapy.



To qualify for enrollment, patients must have been 12–65 years of age (inclusive), weighed ≥50 kg (110 lbs) and experienced 1–20 unprovoked, complex partial or primary/secondarily generalized tonic–clonic (GTC) seizures within the past 3 months. Patients must have had seizures indicative of new-onset epilepsy or epilepsy relapse, characterized by unprovoked seizures diagnosed using appropriate clinical or laboratory assessments and ≥2 lifetime unprovoked seizures or one lifetime unprovoked seizure and electroencephalography (EEG) demonstrating focal or generalized epileptiform EEG abnormalities other than slow spike-wave or benign EEG variants. Inclusion also required patients to be candidates for rapid initiation of AED therapy based on investigator assessment of seizure frequency, seizure severity, risk of seizure recurrence, or other clinical features.

Patients were excluded if they had used AEDs within 30 days prior to randomization (except for ≤5 days of oral or parenteral benzodiazepine for antiseizure management), or if they had seizures provoked by alcohol withdrawal, drug intoxication, acute meningitis or encephalitis, acute head injury, acute stroke, acute hypoxia/ischemia encephalopathy, metabolic derangement, or intracranial neoplasm. Other exclusion criteria were: absence or myoclonic seizures only, psychogenic nonepileptic seizures, current status epilepticus or history of convulsive status epilepticus, previously failed adequate trial of TPM or PHT due to lack of efficacy or AEs, previous discontinuation of any AED therapy for noncompliance, and use of prohibited concomitant medications. Patients were excluded from this trial if they had a history (within the past 6 months) of a major psychiatric disorder, a history of nephrolithiasis, active liver disease, or liver function tests greater than twice the upper limit of normal at the time of screening.

Study design

The study was performed in accordance with current International Committee on Harmonization guidelines on Good Clinical Practice and applicable regulatory requirements. An Independent Data Monitoring Committee (IDMC) was established to review partially unblinded data on an ongoing basis to ensure the safety of the patients enrolled in the trial. The IDMC could make recommendations regarding the continuation of the trial and had the authority to stop the trial if there was a clear imbalance in seizure rates between treatment groups.

This multicenter, randomized, double-blind, double-dummy, parallel-group comparative trial included a screening phase lasting for up to 7 days, a 28-day double-blind phase, and an optional open-label 12-week extension phase (Fig. S1). At visit 1, during the screening phase, the patient’s medical and medication history was reviewed, physical and neurologic examinations were performed, and vital signs and samples for clinical laboratory evaluation were obtained. Patients completed their informed consent and if all evaluations were completed and the results determined that the patients were eligible for enrollment, visit 1 could be combined with visit 2.

Patients were randomized at visit 2 to the TPM or PHT treatment group. This double-blind phase consisted of a 1-day initiation period (day 1) and a maintenance period (days 2–28). The target dose for patients randomized to TPM treatment was 100 mg/day, given as three divided doses on cay 1 (50 mg, 25 mg, and 25 mg, respectively, at 2-h intervals) and at a dose of 50 mg twice daily during the maintenance period. For patients with tolerability problems during initiation, the initiation dose could be limited after 50 or 75 mg, or the time between initiation doses could be increased. Patients who received a total initiation dose of 50, 75, or 100 mg could enter the maintenance period. The target dose for patients randomized to PHT treatment was 1,000 mg on day 1, given as 3 doses (400 mg, 300 mg, and 300 mg, respectively, at 2-h intervals) followed by a target maintenance dose of 300 mg/day given as 300 mg once daily. For patients with tolerability problems during initiation, the initiation dose could be limited after 400 or 700 mg, or the time between initiation doses could be increased. Patients who received a total initiation dose of 400, 700, or 1,000 mg could enter the maintenance period. During the first week of the maintenance period (days 2–7), investigators could adjust the TPM or PHT dose upward or downward. After week 1, the dose could be adjusted only downward by one TPM tablet and one PHT capsule. To remain in the study, patients must have been taking at least 75 mg TPM or 200 mg PHT from day 8 onward; patients unable to maintain these minimum doses were required to discontinue the study.

Patients returned for visit 3 on day 7 and for visit 4 on day 28. Patients could remain in the trial until day 28 or until they experienced a complex partial or GTC seizure, at which time visit 4 procedures were scheduled. Visit 4 procedures included blood sampling for blinded determination of serum AED level. Patients who did not enter the open-label extension phase had Visit 4/Final Visit/Early Termination procedures performed and began to taper study medication by approximately 30% of their achieved dose every 3 days with concurrent initiation of alternative therapy. A final visit (visit 4T) was conducted after completion of the conversion to alternative therapy.

Patients who completed the double-blind phase or who exited for a complex partial or GTC seizure had the opportunity to enter the open-label extension phase, during which time all patients received TPM therapy.


The primary efficacy variable was time to first complex partial or GTC seizure. Secondary efficacy variables were time to first complex partial and time to first GTC seizure. Each patient was given a diary and asked to record the date, description, and duration of any seizure experienced during the trial; information could also be completed by an observer who witnessed the seizure. Investigators reviewed the diaries, classified the seizures and entered the information into the case report form. Blood levels of AEDs were determined from serum samples. Safety was assessed by measurement of vital signs, physical examinations, brief neurologic examinations, clinical laboratory tests, and evaluation of AEs.

Statistical analyses

The safety analysis set included all randomized patients who took at least one dose of study medication. The intent-to-treat (ITT) analysis set included all randomized patients who took at least one dose of study medication and provided at least one postrandomization efficacy measurement.

The primary efficacy variable was time to first seizure (complex partial or GTC seizure). For patients who did not have a seizure by the end of the double-blind phase, times to first seizure were censored to 28 days; for patients who discontinued the double-blind phase for other reasons, times to first seizure were censored at the time of discontinuation. The primary null hypothesis of the study was that the hazard of having a first seizure in the TPM-treated group was greater than 2.275 times the hazard of having a first seizure in the PHT-treated group. A hazard ratio (HR) of 2.275 corresponds to a 20% difference in the cumulative seizure-free rate at day 28 (80% for the PHT group, 60% for the TPM group), assuming time to first seizure follows an exponential distribution. The HR was estimated using a Cox proportional hazard model with treatment group as the explanatory variable. TPM would be considered as effective as PHT if the upper limit of the 95% CI of the HR based on the Cox proportional hazard analysis was ≤2.275. Proportions of seizure-free patients over the 28-day treatment period were estimated using Kaplan-Meier method for the ITT analysis set for each treatment group. Cox proportional hazard models were used for assessment of the effect of covariates, such as gender, age, baseline weight, baseline seizure type, and duration since diagnosis of epilepsy.

All planned analyses were performed. In addition, a post hoc analysis of patient retention was carried out by obtaining Kaplan-Meier estimates for the cumulative rate of discontinuation over the 28-day double-blind treatment period. Discontinuations included experiencing a complex partial or GTC seizure, or withdrawal due to an adverse event, loss to follow-up, patient choice, or other reason. The statistical significance of any difference in study completion survival functions of the two treatment groups was assessed by a log-rank test.

All statistical tests were conducted at the two-sided 5% significance level.


A total of 261 patients were randomized to receive either treatment with TPM (n = 133) or PHT (n = 128). Of these, 259 patients (132 and 127, respectively) were included in the safety analysis set and 254 (128 and 126, respectively) in the ITT analysis set. Demographic and baseline characteristics of the safety analysis set are shown in Table 1. Approximately three-fifths (63.3%) of the patients were white and one-fourth (25.1%) were black. A greater proportion of female patients were randomized to the TPM treatment group compared with the PHT treatment group (59.8% vs. 43.3%, respectively). A summary of seizure history in this patient group is shown in Table 2; the two groups were similar with regard to seizure history. As determined by Cox proportional hazard regression model, none of the demographic or baseline disease characteristics were found to be significant predictors of seizure risk.

Table 1.   Baseline demographic characteristics (safety analysis set)
 TPM (n = 132)PHT (n = 127)Total (n = 259)
  1. aData missing from three patients in each group.

Mean age (SD), year 33.2 (14.1) 35.3 (15.5) 34.2 (14.8)
Gender, n (%)
 Male 53 (40.2) 72 (56.7)125 (48.3)
 Female 79 (59.8) 55 (43.3)134 (51.7)
Race, n (%)
 White 90 (68.2) 74 (58.3)164 (63.3)
 Black 30 (22.7) 35 (27.6) 65 (25.1)
 Asian  0 (0.0)  1 (0.8)  1 (0.4)
 Other 12 (9.1) 17 (13.4) 29 (11.2)
Mean height (SD), ina 66.1 (4.0) 67.1 (4.4) 66.6 (4.2)
Mean weight (SD), lb183.8 (57.1)180.9 (41.9)182.4 (50.2)
Table 2.   Baseline seizure history (safety analysis set)
 TPM (n = 132)PHT (n = 127)Total (n = 259)
  1. GTC, generalized tonic–clonic.

  2. aData missing from one patient in PHT group.

  3. bPrior to 3-month retrospective baseline.

  4. cWithin 3-month retrospective baseline; data missing from one patient in TPM group and one patient in PHT group.

Mean age at 1st seizure (SD), yeara29.2 (16.1)  31.8 (17.1) 30.5 (16.6)
Mean time from 1st seizure to randomization (SD), yeara3.7 (7.5)3.1 (6.5)3.4 (7.0)
Mean age at epilepsy diagnosis (SD), yeara 31.3 (15.9) 34.0 (16.4) 32.6 (16.2)
Mean time from diagnosis to randomization (SD), yeara 1.8 (5.8)1.3 (4.9) 1.5 (5.3)
Lifetime seizure history, n (%)b
 None50 (37.9)45 (35.4)95 (36.7)
 GTC51 (38.6)62 (48.8)113 (43.6)
 Complex partial38 (28.8)28 (22.0)66 (25.5)
 Simple partial9 (6.8)15 (11.8)24 (9.3)
 Absence5 (3.8)3 (2.4)8 (3.1)
 Myoclonic3 (2.3)1 (0.8)4 (1.5)
 Other1 (0.8)2 (1.6)3 (1.2)
Baseline seizure history, n (%)c
 GTC99 (75.0)101 (79.5)200 (77.2)
 Complex partial44 (33.3)38 (29.9)82 (31.7)
 Simple partial14 (10.6)13 (10.2)27 (10.4)
 Absence2 (1.5)1 (0.8)3 (1.2)
 Myoclonic5 (3.8)3 (2.4)8 (3.1)
 Other0 (0.0)1 (0.8)1 (0.4)
Mean number GTC seizures within 3-month retrospective baseline (SD) 2.1 (1.8)  1.9 (1.5)  2.0 (1.7)
Mean number complex partial seizures within 3-month retrospective baseline (SD) 5.6 (6.3)  6.7 (10.8)  6.1 (8.7)

The disposition of all patients in the randomized analysis set is shown in Table 3. A higher percentage of PHT-treated patients discontinued during the double-blind phase (21.1%) compared with TPM-treated patients (12.8%). The main reason for discontinuation was AEs; 12.5% of PHT-treated patients and 6.0% of TPM-treated patients. The majority of patients went on to enter the open-label phase of the study, 82.0% of the TPM treatment group and 76.6% of the PHT treatment group.

Table 3.   Disposition of randomized patients (randomized analysis set)
 TPM (n = 133)PHT (n = 128)
  1. aEither completed the study up to day 28 without experiencing a complex partial or GTC seizure, or remained in the double-blind phase of the study up to the point at which they experienced an exit complex partial or GTC seizure.

Completed double-blind period, n (%)a116 (87.2)101 (78.9)
Discontinued during double-blind period, n (%)17 (12.8)27 (21.1)
 Lost to follow-up2 (1.5)2 (1.6)
 Patient choice4 (3.0)2 (1.6)
 Adverse event8 (6.0)16 (12.5)
 Other3 (2.3)7 (5.5)
Entered the open-label phase109 (82.0)98 (76.6)
Did not enter the open-label phase24 (18.0)30 (23.4)

The mean number of days that patients were on double-blind study medication appeared similar in the two treatment groups, 25.6 ± 9.22 days in the TPM group and 25.1 ± 8.90 days in the PHT group (Table S1). The mean AED blood concentration was 3.6 ± 1.34 μg/mL for TPM and 8.5 ± 7.12 μg/ml for PHT. Approximately 70% of patients in the ITT analysis set were able to tolerate the full dose of study medication, with no apparent differences between the groups.

Analysis of the primary efficacy variable, time to first seizure, using a Cox proportional hazard model with treatment group as the explanatory variable, did not establish noninferiority of TPM to PHT, since the upper 95% confidence limit (CL) of the HR (TPM/PHT) was greater than 2.275, the prespecified CL (hazard ratio = 2.0, 95% CI 0.98 to 4.12) (Table S2). In addition, since the lower 95% CL was less than one, superiority of PHT to TPM monotherapy after rapid oral initiation was not established (p = 0.366). Kaplan-Meier seizure-free survival curves for TPM and PHT are shown in Fig. 1. At Day 28, the estimated seizure-free rate was 81.1% for TPM treatment in comparison with 90.3% for PHT treatment. Midway through the trial, at day 14, the estimated seizure-free rate was 92.1% for TPM treatment in comparison with 95.0% for PHT treatment. Cox proportional hazard modeling of the effect of various covariates revealed none with a significant effect on time to first seizure (Table S2). Similar results on covariates were obtained when GTC and complex partial seizures were analyzed separately (results not shown).

Figure 1.

Time to first exit seizure: ITT analysis set.

Kaplan-Meier estimates of the time to discontinuation for all causes are shown in Fig. 2. Retention rates appeared similar up to around day 8 (93.9% for TPM-treated patients and 92.9% for PHT-treated patients), at which time they began to diverge. On day 28 of the trial, retention rates were 89.4% in the TPM treatment group and 80.3% in the PHT treatment group. The retention rate-time functions of the two treatment groups over the 28-day period were significantly different in favor of the TPM group (p = 0.047, log-rank test).

Figure 2.

Time to discontinuation (all cause): safety analysis set.

The occurrence of AEs for the safety analysis set is summarized in Table 4. The two treatment groups appeared similar in terms of the proportion of patients with any AE and the proportion of patients with AEs related to study drug. AEs causing permanent discontinuation of study medication were experienced by 17 patients (13.4%) in the PHT treatment group in comparison with 9 patients (6.8%) in the TPM treatment group. One TPM patient experienced multiple GTC seizures and was considered as a study completer per protocol. One PHT patient was discontinued by the investigator for “other” reasons (unreliable at reporting side effect). Serious AEs were experienced by two patients (1.5%) in the TPM treatment group (syncope and grand mal convulsion), both of which led to withdrawal, and by five patients (3.9%) in the PHT treatment group (general cardiac failure, suicide attempt, cerebrovascular disorder and hemiparesis, and two patients with rash), leading to withdrawal in the two patients experiencing rash. No deaths occurred during this trial. AEs experienced by at least 5% of patients are shown by body system in Table 5. The most common treatment-related AEs were dizziness (13.6% TPM, 19.7% PHT), paresthesia (9.8% TPM, 0.8% PHT), and somnolence (6.8% TPM, 8.7% PHT). One notable difference between the two groups was the higher incidence of rash in PHT-treated patients versus TPM-treated patients (7.9% vs. 0.8%, respectively). The incidence of rash causing withdrawal from the study was also higher in the PHT group (6.3%) than in the TPM group (0%). Cognitive AEs led to withdrawal in two (1.5%) TPM-treated patients and no PHT-treated patients.

Table 4.   Summary of adverse events (AE) (safety analysis set)
(n = 132)
n (%)
(n = 127)
n (%)
  1. aAn event was considered related if its relationship to study medication was possible, probable or very likely.

  2. bOne patient in the TPM treatment group experienced multiple GTC seizures and was considered a study completer per protocol. One patient in the PHT treatment group was discontinued by the investigator for “other” reason (unreliable at reporting of side effect) in Table 3.

Patients with any AE86 (65.2)86 (67.7)
Patients with AE related to study druga75 (56.8)71 (55.9)
Patients with AE causing withdrawalb9 (6.8)17 (13.4)
Patients with AE related to study drug causing withdrawal6 (4.5)16 (12.6)
Patients with any serious AE2 (1.5)5 (3.9)
Patients with serious AE related to study druga2 (1.5)2 (1.6)
Patients who died0 (0.0)0 (0.0)
Table 5.   Adverse events with an incidence ≥5% in either treatment group (safety analysis set)
Body system organ class: Preferred termTPM
(n = 132)
(n = 127)
Body as a whole Fatigue12 (9.1)11 (8.7)
CNS and PNS disorders
 Dizziness26 (19.7)35 (27.6)
 Headache11 (8.3)15 (11.8)
 Paresthesia29 (22.0)5 (3.9)
Gastrointestinal system disorders Nausea9 (6.8)12 (9.4)
Psychiatric disorders
 Anorexia9 (6.8)7 (5.5)
 Confusion8 (6.1)2 (1.6)
 Difficulty with concentration/attention8 (6.1)3 (2.4)
 Nervousness7 (5.3)2 (1.6)
 Somnolence16 (12.1)18 (14.2)
Skin and appendage disorders Rash1 (0.8)10 (7.9)
Vision disorders Vision abnormal6 (4.5)9 (7.1)


This trial, comparing TPM 100 mg/day without titration and PHT 300 mg/day following an oral loading dose of 1,000 mg as monotherapy in patients with new-onset epilepsy requiring rapid oral initiation of AED treatment, did not demonstrate noninferiority of TPM to PHT in terms of time to first seizure over the first month of therapy (the primary end point) at these dosages. However, seizure-exit rates appeared similar for the two treatment groups over the first 2 weeks of the trial, casting uncertainty on the benefit of rapid initiation of PHT, especially considering the high incidence of rash leading to treatment discontinuation. In addition, the superiority of PHT to TPM monotherapy after rapid oral initiation was not established.

Study results must be interpreted in light of the dosing schedule used. A 28-day period was selected to eliminate the need for measurement of AED blood levels to monitor PHT dosing. Low maintenance doses were chosen, thought to provide comparable efficacy of both TPM and PHT. In addition, a potential limitation of the study methodology is the possibility of bias introduced by informative censoring.

In terms of retention rate, measured as time to study discontinuation for all causes, the proportion of patients remaining on therapy in the TPM treatment group was significantly superior to that in the PHT treatment group. Seizure recurrence was used as the primary end point in pivotal TPM monotherapy trials (Arroyo et al., 2005; Gilliam et al., 2003; Glauser et al., 2007), and has been used in comparative monotherapy studies in newly diagnosed patients for other AEDs (Bill et al., 1997; Steiner et al., 1999). Although the precise definitions were not identical, several important studies have used patient retention measures as a primary end point (Mattson et al., 1985, 1992; Marson et al., 2007a,b). The landmark VA Cooperative Study demonstrated that carbamazepine and PHT are more effective and better tolerated than phenobarbital or primidone in the treatment of partial and secondarily GTC seizures, by assessing treatment success in terms of the point at which the AED failed to control seizures or caused unacceptable side effects (Mattson et al., 1985). The SANAD study of the effectiveness of AEDs used time to treatment failure as the primary outcome measure (Marson et al., 2007a,b). Retention rate is a composite measure of a drug’s efficacy and safety/tolerability profiles, encompassing all possible reasons for drug discontinuation, including ineffectiveness and intolerability. Retention rates incorporate the willingness of patients to continue drug therapy, sometimes despite AEs. It has been suggested that retention rates may be the most clinically relevant parameter of an AED, and are a useful measure by which to compare the long-term performance of AEDs in clinical practice (Bootsma et al., 2008).

Overall, PHT and TPM were well tolerated in this study. However, the proportion of patients who discontinued treatment due to AEs was lower in the TPM treatment group; around twice as many patients who received PHT discontinued from the double-blind phase of the trial because of AEs. There was a large number of discontinuations due to rash among PHT-treated patients, a side effect that cannot be controlled by reducing dose and that usually necessitates an early therapy switch to another AED, with the risk of seizure exacerbation. The tolerability results indicate that TPM therapy can be initiated rapidly, without titration, when rapid initiation of treatment is warranted. The AEs associated with TPM treatment in this trial were typical of those seen in other studies of TPM where a gradual titration was employed. Rapid oral initiation of TPM is, therefore, a reasonable choice in a patient in whom TPM therapy might be considered.

In addition to the tolerability issues associated with the use of PHT observed in this and other studies (rash, CNS side effects, nausea), long-term use of PHT may be associated with peripheral neuropathy (including absent patellar or Achilles reflex, decreased conduction velocity, and sensory impairment) and impaired neuromuscular transmission (increased neuromuscular fatigability following repetitive stimulation) (So & Penry, 1981), gingival hyperplasia, coarsening of the facial features, and hirsutism (Trevisol-Bittencourt et al., 1999). Patients with new-onset epilepsy are often prescribed PHT as initial treatment in the emergency department (Huff et al., 2001), and it is common practice for this treatment to continue when patients consult a neurologist (Ortho-McNeil Neurology Consultant Surveys, 2003–2004) despite the possibility of these long-term safety concerns.

In summary, this clinical trial of rapid oral initiation of AED monotherapy in new-onset epilepsy did not establish the noninferiority of treatment with TPM compared with PHT in terms of time to first exit seizure at the dosages used, but demonstrated in a post hoc analysis the superiority of TPM treatment in terms of patient retention. Within the parameters of the dosing described in this study, treatment-limiting AEs were nearly twice as frequent with PHT treatment as with TPM treatment. The trial demonstrated that TPM may be an alternative for initiation without titration in patients for whom urgent treatment is required.


The authors acknowledge the additional investigators who enrolled patients in this trial: A Abubakr (Edison, NJ), R Armstrong (Asheville, NC), R Ayala (Tallahassee, FL), J Balmakund (St. Cloud, MN), R Beach (Syracuse, NY), M Bensalem-Owen (Lexington, KY), V Biton (Little Rock, AR), W Carlini (Medford, OR), J Cavasos (San Antonio, TX), G Connor (Tulsa, OK), J DeCerce (Jacksonville, FL), AJ Friedman (Rockville MD), BV Gallo (Miami, FL), R Gross (Rochester, NY), N Haddad (Little Rock, AR), M Harris (Suwanee, GA), R Hull (Huntsville, AL), A Isa (Maitland, FL), PW Kaplan (Baltimore, MD), P Klein (Bethesda, MD), WA Knubley (Fort Smith, AR), D Labiner (Tucson, AZ), K Liow (Wichita, KS), WA McElveen (Bradenton, FL), C O’Donovan (Winston-Salem, NC), E Passaro (St. Petersburg, FL), E Pearlman (Savannah, GA), B Philbrook (Atlanta, GA), K Rathke (Raleigh, NC), W Rosenfeld (Chesterfield, MO), S Sahai-Srivastava (Los Angeles, CA), R Schwartz (Hattiesburg, MS), J Slater (Houston, TX), M Sperling (Philadelphia, PA), M Tabbaa (Panama City, FL), AB Thomas (San Antonio TX), T Ting (Baltimore, MD), A Todorov (Northport, AL), A Towne (Richmond, VA), B Vaughn (Chapel Hill, NC).

Funding for this study was provided by Ortho-McNeil Janssen Scientific Affairs, LLC.

Authors affiliated with (or employed by) the study sponsor provided technical review and input for the manuscript. Editorial assistance was provided by Sharon Schaier, PhD, of PAREXEL MMS with funding provided by Ortho-McNeil Janssen Scientific Affairs, LLC. All authors had full access to the study data but did not request to independently analyze the data.


Eugene Ramsay has received support from, and/or has served as a paid consultant for Abbott Laboratories, Bertex, Carter-Wallace, Cephalon, Cyberonics, Eisai, GlaxoSmithKline, IVAX, Marion Merrell Dow, Novartis, Ortho-McNeil Pharmaceuticals, Ovation Pharmaceuticals, Pfizer, RW Johnson, Schwarz Pharma, SmithKline Beecham, UCB Pharma, and X-cel Pharma.

Edward Faught has received support from, and/or has served as a paid consultant for Ortho-McNeil/Johnson & Johnson.

Allan Krumholz has received support from, and/or has served as a paid consultant for Ortho-McNeil Pharmaceuticals, Ovation Pharmaceuticals, and UCB Pharma.

Dean Naritoku has received support from, and/or has served as a paid consultant for Abbott Laboratories, Eisai, GlaxoSmithKline, Ortho-McNeil Pharmaceuticals, and Schwarz Pharma/UCB.

Michael Privitera has received support from, and/or has served as a paid consultant for the American Epilepsy Society, epiFellows Foundation, GlaxoSmithKline, Janssen-Cilag, National Institutes of Health, Ortho-McNeil Pharmaceuticals, Pfizer Inc, and UCB Pharma.

Lesley Schwarzman, Lian Mao, Frank Wiegand, and Joseph Hulihan are employees of Ortho-McNeil Janssen Scientific Affairs, LLC.


We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.