• Zonisamide;
  • Antiepilepsy drug;
  • Epilepsy;
  • Partial-onset seizures;
  • Randomized controlled trial


  1. Top of page
  2. Abstract
  6. Acknowledgments

Summary: Purpose: This study was designed to evaluate efficacy and safety of zonisamide (ZNS) as adjunctive treatment for patients with refractory partial seizures.

Methods: This randomized, double-blind, placebo-controlled study was conducted at four epilepsy treatment centers. It included a baseline phase (8 to 12 weeks) and a double-blind treatment phase (12 weeks). Initially, patients randomized to ZNS treatment were given a 7-mg/kg/d dosage. When investigators found that adverse effects could be reduced by gradually introducing ZNS, patients were allowed to begin treatment at lower doses (100 mg or ∼1.5 mg/kg/d) titrated over several weeks to a maximum of 400 to 600 mg/d. Primary and secondary efficacy measures were the median percentage reduction from baseline in seizure frequency and the proportion of patients achieving a ≥50% reduction from baseline (responder rate). Patient and physician global assessments also served as indicators of efficacy. Safety was assessed primarily by treatment-emergent adverse events.

Results: ZNS-treated patients had a 28.9% reduction in seizure frequency, which differed significantly from the 4.7% increase in placebo-treated patients. The responder rate for ZNS-treated patients was 26.9%, compared with 16.2% for placebo-treated patients. At study's end, 66.2% of ZNS-treated patients and 12.3% of placebo-treated patients considered their condition improved; similarly, physicians assessed 63.6% of ZNS-treated patients and 10.8% of placebo-treated patients as improved. The most frequently reported adverse events with ZNS treatment included somnolence, irritability, dizziness, nausea, and fatigue.

Conclusions: As adjunctive treatment, ZNS was generally well tolerated and significantly improved seizure control among patients with refractory partial seizures.

Zonisamide (ZNS, 1,2-benzisoxazole-3-methane-sulfonamide) is a relatively new antiepileptic drug (AED) with multiple mechanisms of action, including blockage of voltage-sensitive sodium channels (1) and voltage-dependent T-type calcium channels (2), blockage of potassium-evoked glutamate response (3), reduction of glutamate-mediated synaptic excitation (4), facilitation of dopaminergic (5) and serotonergic (6) neurotransmission, scavenging of hydroxyl and nitric oxide radicals (7,8), increasing γ-aminobutyric acid (GABA) release from the hippocampus (9,10), and weak inhibition of carbonic anhydrase (11). Such mechanisms suggest that ZNS is a broad-spectrum treatment for epilepsy. Early pilot and open-label studies conducted in the United States characterized ZNS as effective (12,13). Studies conducted in Europe and Japan have confirmed the safety and efficacy of ZNS at dosages of 400 to 500 mg, administered once or twice daily (14,15). In addition, these studies have shown ZNS to be effective at the recommended starting dosage (100 mg/d). The second of two placebo-controlled clinical trials in the United States has shown that ZNS is both effective and well tolerated as an adjunctive treatment for refractory partial seizures (16).

This report presents data from the first of two placebo-controlled trials conducted in the United States. The study was a randomized, multicenter, double-blind trial examining the efficacy, safety, and tolerability of ZNS as an adjunctive treatment for refractory partial seizures at dosages of 400 to 600 mg/d.


  1. Top of page
  2. Abstract
  6. Acknowledgments

Study design

This was a randomized, double-blind, placebo-controlled, parallel-group, multicenter study. Epilepsy centers in four locations in the United States (Miami, Florida; Ann Arbor, Michigan; Gainesville, Florida; and Boston, Massachusetts) enrolled patients. Patients were recruited over a period of 29 months. All patients enrolled in this study provided written informed consent.

The trial consisted of an 8- to 12-week baseline phase, followed by a 12-week double-blind treatment phase. During the baseline phase, patients received their customary AEDs. The length of the baseline phase depended on seizure frequency. Those patients who had ≥15 seizures in the first 4 weeks or ≥30 seizures in the first 8 weeks entered the double-blind phase after 8 weeks; those who had fewer seizures continued on their usual AEDs for ≥4 weeks before entry into the double-blind phase. At the end of the baseline phase, patients were randomly assigned to receive placebo or ZNS treatment in addition to their usual AEDs.

Clinic visits were scheduled monthly during the baseline phase, weekly during the first 2 weeks of the treatment phase, and monthly thereafter. In addition, follow-up assessments were performed for each patient 6 weeks after the double-blind phase. At each clinic visit, patient seizure records were reviewed, and seizure frequency was recorded. Clinicians verified that each patient was taking the study medication or placebo and other prescribed AEDs as directed. Once treatment with ZNS or placebo began, blood samples to measure plasma ZNS levels were collected. Each clinic visit also included physical and neurologic examinations, as well as clinical laboratory tests. Electrocardiograms (ECGs) and EEGs were performed at the beginning of the baseline phase and at the beginning and end of the treatment phase. Finally, patient and physician global assessments were conducted at the end of the treatment phase (or earlier if a patient prematurely discontinued). Patients and physicians assessed response to ZNS by rating the patient's condition (relative to baseline) as having marked improvement, slight improvement, no change, or worsening.

During the treatment phase, dosages of concurrent AEDs were, in general, held constant. However, patients in both treatment groups required changes in concomitant AED dosages. In the placebo group, concomitant AED dosages were adjusted for 10 patients; six had their dosages increased, and four had their dosages decreased. In the ZNS group, concomitant AED dosages were changed for 23 patients; four had their dosages increased, and 19 had their dosages decreased. Dosage changes were small in magnitude, typically <20%.

Inclusion criteria

Patients were eligible for enrollment if they were aged between 17 and 65 years, had an unequivocal history of partial seizures that were refractory to current AED therapy; had a history of at least four complex partial seizures per month; and had no more than eight generalized tonic, clonic, or tonic–clonic seizures per month. In addition, patients had to be receiving at least one, but no more than two, of the following AEDs: phenytoin (PHT), carbamazepine (CBZ), phenobarbital (PB), or primidone (PRM. All patients were required to be in good physical health and capable of counting seizures. Female patients otherwise eligible for enrollment had to have undergone menopause or surgical sterilization, or be practicing an approved method of birth control.

Exclusion criteria

Patients excluded from the study were those with a history or evidence of progressive encephalopathy or a progressive structural lesion in the CNS; progressive ophthalmologic disease; or clinically significant cardiac, hematologic, hepatic, or renal disease. In addition, patients were excluded if they had significant mental retardation; any condition that might interfere with the absorption, distribution, or clearance of ZNS; sulfonamide sensitivity; a history of alcohol or medication abuse; glucose-6-phosphate dehydrogenase deficiency; or a history of hemolytic anemia or acute intermittent porphyria. Patients with abnormal clinical laboratory values not related to current AED therapy were excluded from the study. Patients who had received treatment with an experimental drug within 30 days before admission, those who had previously received ZNS, and those who had received acetazolamide as an anticonvulsant within the previous year were excluded from study participation.


At the time of randomization, each patient who qualified to receive double-blind treatment was assigned a randomization number and given ZNS or placebo accordingly. Randomization codes were generated by the study sponsor. Random allocation of patients to their treatment groups was concealed via the use of numbered containers, and investigators were blinded to the treatment assignment for each patient. ZNS and placebo were supplied as capsules that were identical in appearance and packaged in individually labeled bottles. Each study medication bottle label identified the investigator, patient, and protocol number, date of dispensing, and when to administer the medication. Each ZNS capsule contained 100 mg of medication. At every clinic visit, patients were given enough medication to last until their next visit.


Initially, study medication was given at a dosage of ∼7 mg/kg/d (i.e., 400 mg/d for patients weighing ≤60 kg, 500 mg/d for patients weighing 60 to 71 kg, and 600 mg/d for patients weighing ≥71 kg). The medication was given in two equally divided oral doses every 12 h. Dosages were adjusted as necessary, based on plasma ZNS levels, to obtain therapeutic plasma levels between 20 and 30 μg/ml. Dosage changes were recommended by a nonblinded observer with access to plasma-level results. The nonblinded observer also made random changes in the dosage of placebo to maintain the blind. When it became apparent that introducing the study drug gradually could reduce side effects, the study protocol was altered to allow treatment to begin at a lower dosage. Thereafter, treatment began at a dosage of 100 mg/d (or 1.5 mg/kg/d) for the first week and was increased to 200 mg/d (or 3 mg/kg/d) at week 2 and 400 mg/d (or 6 mg/kg/d) at week 3, until the targeted therapeutic dosage was reached. After this 4-week (28-day) dose-introduction period, investigators could make changes in the study drug dosages based on clinical judgment. For most patients, the final daily dose of study medication was 400 to 600 mg. The nonblinded observer ensured that ZNS dosages did not exceed 20 mg/kg/d and that ZNS plasma levels did not exceed 40 μg/ml.

Seizure data

Each patient kept a daily record of the number of seizures experienced, along with a description of each seizure and its duration. Records were started at screening and were maintained throughout the study. Investigators carefully reviewed the records at each clinic visit.

Efficacy criteria

The primary efficacy outcome variable was the median percentage reduction in frequency in patients with all partial seizures (i.e., simple and complex partial seizures), complex partial seizures only, and all seizure types (including generalized seizures), between weeks 5 and 12 of the treatment phase, relative to baseline seizure frequency. Findings for the ZNS-treated group were compared with those for the placebo-treated group. The secondary efficacy variable was the responder rate, which was defined as the proportion of patients experiencing a ≥50% reduction (as compared with baseline seizure frequency) in seizure frequency during ZNS treatment. Patient and physician global assessments also were included in the efficacy analysis.

Safety data

Safety and tolerability were assessed based primarily on the occurrence of treatment-emergent adverse events, defined as adverse events that occurred during the double-blind phase and were not present during the baseline phase, or that were present at baseline and worsened in severity during the double-blind phase. Adverse events were recorded using Coding Symbols for a Thesaurus of Adverse Reaction Terms (COSTART) terminology and tabulated by body system, maximal severity, medication attributability, and dosage for each treatment group.

Serious adverse events, defined as any events presenting a significant threat to the patient's well-being, also were recorded. Serious adverse events included any events that were fatal, life threatening, permanently or significantly disabling, required hospital admission, prolonged hospitalization, or resulted in long-term outpatient treatment. Serious adverse events also included congenital anomalies, cancer, and drug overdose.

Safety analyses also took into account the results of clinical and laboratory tests performed after each clinic visit. These included hematologic tests (i.e., hemoglobin, hematocrit, red blood cell count, white blood cell count, and platelet count), blood chemistry (i.e., serum enzyme levels, glucose, total protein, calcium, phosphorus, albumin, creatinine, blood urea nitrogen, uric acid, sodium, potassium, and chloride), and complete urinalysis with microscopic examination. Vital signs, such as pulse, blood pressure, temperature, respiratory rate, and body weight, also were measured at each clinic visit.

Statistical methods

Baseline characteristics were summarized by using descriptive statistics for continuous variables and numbers and percentages for categorical variables. Seizure-frequency data for each patient were normalized for a 28-day period. Seizure frequencies were compared between treatment groups by using a two-way analysis of variance (ANOVA) on rank-transformed data, with main effects of treatment, center, and treatment-by-center interaction. A two-way ANOVA, with main effects of treatment, center, and treatment-by-center interaction, was used to assess the comparability of the groups for the remaining continuous variables. Comparability of the groups with regard to gender, age group, and race was assessed by using the Cochran–Mantel–Haenszel (CMH) test with study centers as the strata.

Efficacy and safety analysis involved an intent-to-treat (ITT) population. The ITT population included all patients who received at least one dose of study medication. A last-observation-carried-forward approach was used for the seizure-frequency and responder-rate analyses. For the primary efficacy parameter, the median percentage reductions in seizure frequency in the ZNS and placebo groups were compared by using a two-way ANOVA model on rank-transformed data. The model included the main effects of treatment, center, and treatment-by-center interaction.

For the secondary efficacy variable, the proportion of patients with a ≥50% reduction in seizure frequency (responder rate), the CMH test was used to compare the effectiveness of therapy in the two treatment groups while controlling for center. The Breslow–Day test was used to assess treatment-by-center interaction. Physician and patient global assessments were analyzed by using the CMH test comparing global evaluation category and treatment while controlling for center.

Fisher's exact test was used to evaluate treatment differences in the occurrence of adverse events. Principal analyses focused on treatment-emergent adverse events. Changes from baseline in clinical laboratory measurements were summarized descriptively, and changes from baseline were compared between treatment groups by using the Kruskal–Wallis test. All statistical tests were performed at the 0.05 level of significance by using two-tailed tests.

Discontinuation procedures

Patients could withdraw before completion of the trial if they withdrew consent or if the investigator considered discontinuation to be in the patient's best interest. Reasons for mandatory discontinuation included hazardous exacerbation of seizure activity, serious or intolerable side effects, significant bone marrow suppression, pregnancy, or significant protocol violations. For each patient who withdrew, the reason for withdrawal was indicated on the patient's case-report form, and the patient was scheduled for a clinic visit on the last day of the study, as well as a 6-week follow-up assessment.


  1. Top of page
  2. Abstract
  6. Acknowledgments

Patient characteristics and disposition

Patient demographics and baseline seizure characteristics are summarized in Table 1. Overall, treatment groups were comparable with regard to age, race, weight, and baseline seizure frequency. Significantly more men than women were in the ZNS treatment group than in the placebo group; the potential impact of this imbalance is assessed in terms of gender-specific differences in efficacy, as described later in this article.

Table 1. Patient demographic characteristics
  1. ZNS, zonisamide.

  2. aSignificant difference observed between treatment groups (p < 0.05).

  3. bOther included patients of Hispanic heritage.

  4. cAge at seizure onset was unavailable for one patient in the placebo group and one patient in the ZNS group.

Sex, no. (%)an = 74n = 78
 Male43 (58.1)58 (74.4)
 Female31 (41.9)20 (25.6)
Race, no. (%)n = 74n = 78
 Caucasian64 (86.5)68 (87.2)
 African American5 (6.8)4 (5.1)
 Otherb5 (6.8)6 (7.7)
Age (yr)n = 74n = 78
 Mean ± SD36.4 ± 11.335.6 ± 12.1
Weight (kg)n = 74n = 78
 Mean ± SD72.7 ± 16.174.6 ± 15.7
Age at seizure onset (yr)cn = 73n = 77
 Mean ± SD16.5 ± 10.515.9 ± 12.5
Baseline seizure frequency (seizures/mo)  
 All partial  
     Median (range)9.6 (2.0–186.7)9.1 (1.3–201.0)
 Complex partial  
     Median (range)7.8 (0.3–119.2)8.0 (0.7–201.0)
 All seizures  
     Median (range)10.6 (2.0–190.7)9.1 (1.3–201.0)

Figure 1 shows patient disposition after randomization in the Consolidated Standards of Reporting Trials (CONSORT) format (17). One hundred fifty-two patients were enrolled during the baseline phase of the study, and 152 were randomized. One hundred thirty-one patients, 64 (82.1%) in the ZNS group and 67 (90.5%) in the placebo group, completed the trial.


Figure 1. Patient disposition.

Download figure to PowerPoint

Twenty patients received ZNS at a starting dosage of 7 mg/kg/d; 58 patients had their dosage of ZNS titrated up over a 3-week period. The final dosage of ZNS was <400 mg/d for 27 patients (34.6%); the other 51 patients (65.3%) received ≥400 mg/d. Among patients taking placebo, most (55 patients, 74.3%) were taking five or more capsules per day.

Efficacy findings

Reduction in seizure frequency

With regard to all partial seizures, patients treated with ZNS showed a 28.9% reduction from baseline in seizure frequency (Table 2). In contrast, the frequency of all partial seizures increased by 4.7% in placebo-treated patients. The difference between ZNS-treated and placebo-treated patients was significant (p = 0.0009), and no Treatment × Center interaction was identified.

Table 2. Percentage reduction in seizure frequency and responder rates for the placebo and ZNS groups
 Seizure type
 All partialComplex partialAll seizures
  1. ZNS, zonisamide.

Reduction in seizure frequency
 Placebon = 74n = 72n = 74
  −4.7%0.5% −6.6%
 ZNSn = 78n = 78n = 78
 p Value0.00090.00070.0005
Responder rate
 Placebon = 74n = 72n = 74
 ZNSn = 78n = 78n = 78
 p Value0.11410.01590.0796

With regard to complex partial seizures, patients taking ZNS experienced a significantly greater reduction (27.4%) in seizure frequency, compared with a 0.5% reduction for placebo-treated patients (Table 2; p = 0.0007). No Treatment × Center interaction for the percentage reduction in complex partial seizure frequency was noted.

ZNS had a similar significant impact on the frequency of seizures of all types. A 25.5% reduction in seizure frequency was observed in ZNS-treated patients, compared with a 6.6% increase in placebo-treated patients (Table 2; p = 0.0005).

Responder rates and median therapeutic dosage level

With regard to complex partial seizures, the proportion of responders was significantly higher in the ZNS treatment group (24 of 78, 30.8%) than in the placebo group (10 of 72, 13.9%; p = 0.0159). Although responder rates were numerically higher for all partial seizures or seizures of all types among patients treated with ZNS than among those who received placebo, the differences were not significant (p = 0.1141 for all partial seizures; p = 0.0796 for all seizures; Table 2). Notably, however, the number of ZNS-treated patients (compared with placebo-treated patients) experiencing a ≥25% reduction in seizure frequency from week 5 to week 12 was significantly greater for all partial seizures (p = 0.0133), complex partial seizures (p = 0.0110), and all seizures (p = 0.0117).

The final median dosage of ZNS for patients who achieved a ≥50% reduction from baseline in the frequency of all partial seizures was 500 mg/d, or 6.4 mg/kg/d. The corresponding final median plasma level of ZNS was 14.2 μg/ml. Among patients who did not achieve a ≥50% reduction from baseline seizure frequency, the final median dosage was 400 mg/d, or 6.1 mg/kg/d, and the corresponding final median plasma level of ZNS was 13.9 μg/ml. For all partial seizures, responder rates were similar for patient receiving higher (i.e., >200 mg/d) ZNS dosages and those receiving lower (i.e., ≤200 mg/d) ZNS dosages. Minimal differences were found in median ZNS plasma levels between patients who achieved a ≥50% reduction in seizure frequency and those who did not (data not shown).

Gender analysis

Because of the aforementioned gender imbalance between treatment groups, the influence of gender on seizure frequency during weeks 5 through 12 was assessed. The median percentage reduction from baseline in all partial seizures for male subjects in the ZNS group was 23.3%, compared with a 6.1% increase in the placebo group. The median percentage reduction from baseline in all partial seizures for female subjects in the ZNS group was 34.8%, relative to a 3.3% increase in the placebo group. Responder rates were similar between male and female subjects, and when controlling for gender differences, the responder rates for complex partial seizures were significantly higher in the ZNS group than in the placebo group (p = 0.0188). No significant interaction was observed between gender and treatment group that might influence the percentage reduction in seizure frequency.

Global assessments

Global assessment data were available for 77 patients in the ZNS group and all 74 patients in the placebo group. Physicians' global assessments rated 49 (63.6%) patients in the ZNS group but only eight (10.8%) patients in the placebo group as improved (slightly or markedly) by treatment. Similarly, 51 (66.2%) patients in the ZNS group considered their condition to be improved with treatment, compared with only nine (12.3%) patients in the placebo group. Furthermore, physicians rated only 28 (36.4%) patients in the ZNS group, but 66 (89.2%) in the placebo group, as unchanged or worse, and only 26 (33.8%) in the ZNS group but 64 (87.7%) in the placebo group considered their condition to be unchanged or worse with treatment. For both the patient and the physician ratings, the treatment differences in the distribution of global evaluation of the patient's condition were significant (p < 0.001).

Safety results

Adverse events

Data from all 152 patients who entered the double-blind phase of the study were included in the analysis. Overall, treatment-emergent adverse events occurred with significantly greater incidence in the ZNS group compared with the placebo group (p < 0.05). In both treatment groups, adverse events were generally mild (17 of 78, 21.8% ZNS; 20 of 74, 27.0% placebo) or moderate (30 of 78, 38.5% ZNS; 17 of 74, 23.0% placebo) in severity. The most frequently reported adverse events associated with ZNS were somnolence, irritability, dizziness, nausea, and fatigue.

As noted previously in Fig. 1, 12 patients in the ZNS group and one patient in the placebo group withdrew because of adverse events. Most of the adverse events reported by patients who withdrew from ZNS treatment were considered by investigators to be either possibly related (29 of 78, 37.2%) or definitely related (39 of 78, 50.0%) to ZNS.

No clear relation was evident between dose or plasma ZNS level and adverse events. More than half of the patients who tolerated ZNS dosages of 5.1 to 7.5 mg/kg/d through week 12 experienced treatment-emergent adverse events at that dosage. Treatment-emergent adverse events in patients taking these higher dosages occurred with greater frequency than in those receiving lower dosages (0.1 to 5.0 mg/kg/d). In addition, it was discovered during the trial that a slow titration of the drug was associated with fewer adverse events than was a faster titration.

Serious adverse events

Thirteen patients (eight in the ZNS group and five in the placebo group) experienced one or more serious or potentially serious adverse events, including one death. Among patients taking placebo, one patient died on day 43 of the study, apparently of suffocation after a seizure. Four placebo-treated patients experienced other potentially serious adverse events (i.e., neoplasm, suicidal behavior, thrombosis, and psychosis). The suicidal behavior was judged as possibly related to the study drug, whereas the other serious adverse events that occurred within the placebo group were judged as unrelated to study drug. Eight ZNS-treated patients experienced 11 potentially serious adverse events (i.e., depression, cholestatic jaundice, cerebrovascular accident, schizophreniform behavior, insomnia, anxiety, tachycardia, cholecystitis, suicidal behavior, and intentional overdose). In ZNS-treated patients, depression, insomnia, and anxiety were judged as possibly related to study drug; all other serious or potentially serious adverse events were judged unrelated to the study drug, or the relation to the study drug was unknown.

Clinical laboratory values

All clinical laboratory results were analyzed to identify changes of potential clinical importance. Although significant differences were observed in hematocrit, glucose, phosphorus, creatinine, sodium, and chloride levels when values for the ZNS group were compared with those for the placebo group, investigators considered none of the differences to be of clinical importance. In addition, no clinically important increases or decreases in laboratory values (relative to known normal ranges for any value measured) were observed (data not shown).

EEG and ECG results

EEG results were consistent with the diagnosis of partial seizures, and most were unchanged relative to baseline. Three ZNS-treated patients showed deterioration in EEG patterns during the double-blind treatment phase. However, study drug attributability was not specified for these patients. Review of ECG results showed no clinically significant changes from baseline that could be attributed to ZNS.

Physical and neurologic examinations

Among ZNS-treated patients, most physical and neurologic abnormalities observed during the treatment phase were present during the baseline phase. Physical abnormalities possibly associated with ZNS were previously discussed in the Adverse events section. Patient body weight tended to decrease during ZNS therapy, with a mean decrease of 3.03 kg during the double-blind period, compared with a mean 0.37-kg increase in body weight in the placebo group during the double-blind period. However, these differences in body-weight change between the two groups were not evaluated for statistical significance.

With regard to neurologic abnormalities observed during the treatment phase, those emerging with ZNS treatment were similar in number and nature to those emerging with placebo treatment. Examples of neurologic abnormalities observed during treatment included nystagmus, tremor, and impaired gait.

Protocol deviations

Five ZNS-treated patients had been receiving a concurrent AED (chlorazepate; CZP), indicated as an adjunctive therapy for the management of partial seizures on study entrance. Additionally, several patients were taking valproic acid (VPA) as a concurrent AED. Although eight patients in the ZNS group and four patients in the placebo group did not meet the entrance criterion of four seizures per month, these patients were included in the study.


  1. Top of page
  2. Abstract
  6. Acknowledgments

Results from this study confirm that ZNS improves seizure control in adults with partial seizures. As observed in the second pivotal United States, placebo-controlled study across all seizure categories, the decrease in seizure frequency seen with ZNS was significant relative to the slight increases or decreases seen with placebo (16). The second United States study showed a greater seizure frequency reduction in comparison with this study; however, this inconsistency is most likely related to protocol differences (i.e., length of time patients had been receiving treatment).

More than one fourth of patients qualified as responders to ZNS, compared with roughly one sixth of patients who responded to placebo treatment. Additionally, for complex partial seizures, responder rates were significantly greater in the ZNS group than in the placebo group. Although responder rates did not significantly differ between the ZNS and placebo groups for all partial seizures and all seizures, a greater numeric trend toward higher responder rates was noted in the ZNS group than in the placebo group, and p values for these categories (0.1141 for all partial seizures and 0.0796 for all seizures) were low, approaching statistical significance.

As further evidence of efficacy, most (66.2%) patients who received ZNS assessed their condition as markedly or slightly improved, compared with only 12.3% of placebo-treated patients. Physician ratings of the ZNS and placebo groups were similar to patient ratings.

These efficacy findings are consistent with those published earlier (12–14,16). In an open-label, historical-controlled study, for all partial seizures, a ZNS dosage of 500 mg/d was observed to yield a 52% reduction in seizure frequency and a responder rate of 41% (13). A European trial that used the same protocol as that in this study found that ZNS administered at a dosage of ∼7 mg/kg/d produced a 40.5% decrease in the frequency of all partial seizures, compared with a 9.0% increase for placebo-treated patients (14). The second pivotal clinical trial conducted in the United States found that ZNS at a dosage of 400 mg/d reduced the frequency of all partial seizures by 38% and yielded a responder rate of 42% (16).

Although findings from the second United States pivotal study hint at correlations among ZNS dosage, ZNS plasma levels, and response to treatment (16), the current study provides no evidence to support such relations. However, this study was not specifically designed or powered to detect dose–response relations, and a more detailed study is warranted.

In general, ZNS treatment was well tolerated, with adverse effects most often involving functions of the CNS. In evaluating the data, one important consideration is that all patients were receiving, in addition to ZNS or placebo, other AEDs that were capable of producing a variety of adverse events. The most frequently reported adverse events associated with ZNS treatment included dizziness, irritability, nausea, fatigue, somnolence, and ataxia. These adverse events are common among patients receiving AEDs.

Although treatment-emergent adverse events seemed to be evenly distributed across mild, moderate, and severe classifications, the risk for severe adverse events was clearly associated with initiation of treatment at 7 mg/kg/d. A gradual introduction of ZNS over a 2- to 4-week period improved tolerability and reduced the occurrence of adverse events. Furthermore, another study that followed up patients over a period of 20 weeks showed that adverse effects reported in the first few weeks of ZNS treatment subside over time (16).

Notably, the mean body weight of ZNS-treated patients tended to decrease during the study, as the body weight of placebo-treated patients increased slightly. This trend toward weight loss in ZNS-treated patients could have advantages in overweight patients but would require monitoring in patients at risk for unwanted weight loss.

Overall, this study showed that ZNS is well tolerated in most patients and effective as adjunctive treatment of refractory partial seizures.


  1. Top of page
  2. Abstract
  6. Acknowledgments

Acknowledgment:  This study was supported by Dainippon Pharmaceutical Company and Elan Pharmaceuticals, Inc.


  1. Top of page
  2. Abstract
  6. Acknowledgments
  • 1
    Schauf CL. Zonisamide enhances slow sodium inactivation inMyxicola. Brain Res 1987;413: 1858.
  • 2
    Suzuki S, Kawakami K, Nishimura S, et al. Zonisamide blocks T-type calcium channel in cultured neurons of rat cerebral cortex. Epilepsy Res 1992;12: 217.
  • 3
    Okada M, Kawata Y, Mizuno K, et al. Interaction between Ca2+, K+, carbamazepine and zonisamide on hippocampal extracellular glutamate monitored with a microdialysis electrode. Br J Pharmacol 1998;124: 127785.
  • 4
    Zhu W, Rogawski A. Zonisamide depresses excitatory synaptic transmission by a presynaptic action [ Abstract ]. Epilepsia 1999;40: 245.
  • 5
    Okada M, Kaneko S, Hirano T, et al. Effects of zonisamide on dopaminergic system. Epilepsy Res 1995;22: 193205.
  • 6
    Okada M, Hirano T, Kawata Y, et al. Biphasic effects of zonisamide on serotonergic system in rat hippocampus. Epilepsy Res 1999;34: 18797.
  • 7
    Komatsu M, Okamura Y, Hiramatsu M. Free radical scavenging activity of zonisamide and its inhibitory effect on lipid peroxide formation in iron-induced epileptogenic foci of rats. Neurosciences 1995;21: 239.
  • 8
    Mori A, Noda Y, Packer L. The anticonvulsant zonisamide scavenges free radicals. Epilepsy Res 1998;30: 1538.
  • 9
    Kawai M, Endoh A, Kinno I, et al. Effect of zonisamide on release of aspartic acid and gamma-aminobutyric acid from the hippocampal slices of E1 mice. Neurosciences 1994;20: 1159.
  • 10
    Endoh A, Kinno I, Kawai M, et al. Effect of zonisamide on neurotransmitter in mouse brain. Neurosciences 1994;20: P1736.
  • 11
    Masuda Y, Karasawa T. Inhibitory effect of zonisamide on human carbonic anhydrase in vitro. Arzneimittelforschung 1993;43: 4168.
  • 12
    Sackellares JC, Donofrio PD, Wagner JG, et al. Pilot study of zonisamide (1,2-benzisoxazole-3-methanesulfonamide) in patients with refractory partial seizures. Epilepsia 1985;26: 20611.
  • 13
    Leppik IE, Willmore LJ, Homan RW, et al. Efficacy and safety of zonisamide: results of a multicenter study. Epilepsy Res 1993;14: 16573.
  • 14
    Schmidt D, Jacob R, Loiseau P, et al. Zonisamide for add-on treatment of refractory partial epilepsy: a European double-blind trial. Epilepsy Res 1993;15: 6773.
  • 15
    Yagi K, Seino M. Methodological requirements for clinical trials in refractory epilepsies: our experience with zonisamide. Prog Neuropsychopharmacol Biol Psychiatry 1992;16: 7985.
  • 16
    Faught E, Ayala R, Montouris GG, et al. Randomized controlled trial of zonisamide for the treatment of refractory partial-onset seizures. Neurology 2001;57: 17749.
  • 17
    Begg C, Cho M, Eastwood S, et al. Improving the quality of reporting of randomized controlled trials: the CONSORT statement. JAMA 1996;276: 6379.