[1] 10w Zonisamide Improves Wearing-Off in Parkinson's Disease: A Randomized, Double-Blind Study

[1] 249w Background: Previously, we reported 50 mg/d zonisamide improved wearing-off without increasing dyskinesia in patients with Parkinson's disease (PD). Methods: To determine the efficacy of zonisamide for treatment of "off" time in PD patients, we conducted a multicenter, randomized, double-blind, parallel-group, placebo-controlled study in Japan. Patients with PD and wearing-off received placebo for 4 weeks and then were treated for 12 weeks with zonisamide 25 or 50 mg/d or placebo, in addition to their previous therapy. The primary endpoint was the change from baseline in daily "off" time as determined by patients' diaries at the final assessment. Secondary endpoints included changes from baseline in the total scores of the Unified Parkinson's Disease Rating Scale Parts I, II, III, and IV, the dyskinesia duration, and PDQ-39 score. Results: Of 422 patients enrolled, 389 (131 for placebo, 130 for zonisamide 25 mg, and 128 for zonisamide 50 mg) were randomized, and 354 (120, 119, and 115, respectively) completed the study. The "off" time significantly reduced by 0.719 6 0.179 h for zonisamide, 50 mg compared with placebo (0.011 6 0.173 h, P 5 0.005). Although the incidence of somnolence was higher for zonisamide (3.1% for zonisamide 25 mg, 6.3% for zonisamide 50 mg) than for placebo (2.3%), the incidences of the other adverse events, including dyskinesia or hallucination, for both zonisamide treatments were comparable to those for placebo. Conclusion:The study provides evidence that confirms the efficacy of zonisamide 50 mg/d for reduction in "off" time in PD patients with wearing-off phenomena. V

[1] 168w A total of 422 patients were enrolled in the study, and 33 discontinued the therapy during the run-in period. Thus, 389 patients were randomized and proceeded to the treatment period. All of the 389 were included in the safety analysis, and 375 (129 for placebo, 125 for zonisamide 25 mg, 121 for zonisamide 50 mg, respectively) in the efficacy analysis (FAS), and 14 were excluded from FAS because of missing baseline or post-baseline "off" time data (Fig. 1). During the treatment period, 35 discontinued the study treatment, and the remaining 354 completed the study. The most common reason for discontinuation was adverse events, especially for zonisamide 50 mg. Adverse events leading study discontinuation were chest discomfort, pulmonary embolism, and disseminated intravascular coagulation/endotoxin shock caused by bacterial enterocolitis for placebo; abnormal liver function test (occurring before randomization), and drug eruption for zonisamide 25 mg; hallucination, upper abdominal dysesthesia, somnolence, paresthesia of mouth, infectious peritonitis caused by gastric ulcer perforation, dysesthesia/decreased appetite, and decreased granulocyte count for zonisamide 50 mg.

[2] 102w Baseline characteristics of FAS are summarized in Table 1, and of all randomized subjects in Supplemental Data table. The patients had a mean age of 63.6 y, mean PD duration of 8.5 y, mean wearing-off duration of 2.7 y, and disease acuity as indicated by the mean value on the modified Hoehn and Yahr scale of 2.32 ("on" state) and 3.30 ("off" state). The mean levodopa (L-dopa) equivalent daily dose was 487.9 mg, and mean L-dopa daily dose was 374.1 mg. The patients receiving any antiparkinson drugs other than L-dopa accounted for 88.5% for dopamine agonists, 46.7% for selegiline (MAO-B inhibitor), and

[3] 167w Change in the "off" time (least square mean 6 standard error) at week 12 (the primary efficacy endpoint) was 20.011 6 0.173 h/d for placebo, 20.436 6 0.176 h/ d for zonisamide 25 mg, and 20.719 6 0.179 h/d for zonisamide 50 mg, showing a significantly longer reduction for zonisamide 50 mg than for placebo (difference, 20.709 h/d [95% confident interval, 21.198, 20.219]; P 5 0.005) (Fig. 2). The results of post hoc sensitive analysis using MMRM also showed that zonisamide 50 mg was superior to placebo (20.725 6 0.184 vs. 20.034 6 0.179 h/d, P 5 0.007). The "off" time for zonisamide 50 mg reduced significantly at or after week 4, whereas that for placebo did not. The responder rate was 36.8% (46/125 patients) for zonisamide 25 mg, and 40.5% (49/121 patients) for zonisamide 50 mg; these rates were significantly higher than that for placebo (20.9%, 27/129 patients) (P 5 0.006 and P < 0.001). Sleep time was not significantly changed in all of the treatments.

[4] 146w Change in the UPDRS II (off) total score at week 12 was 20.9 6 0.3 for placebo, 21.1 6 0.3 for zonisamide 25 mg, and 22.0 6 0.3 for zonisamide 50 mg (P 5 0.021 vs. placebo). Change in the UPDRS Part III total score at week 12 was 22.3 6 0.5 for placebo, 22.0 6 0.5 for zonisamide 25 mg, and 23.8 6 0.5 for zonisamide 50 mg. No significant differences were found between the zonisamide treatments and placebo with respect to changes from baseline in UPDRS Part I, II (on), and IV and PDQ-39 scores. For tremor, the reduction in item 21 (action or postural tremor of hands) score was greater for zonisamide 50 mg than for placebo, although no notable differences in the change in scores of items 16 (tremor) and 20 (tremor at rest) were found between the zonisamide treatments and placebo.

[5] 73w Change in the dyskinesia duration (mean 6 standard deviation [SD]) at week 12 was 20.027 6 0.945 h/d for placebo, 0.197 6 1.200 h/d (P 5 0.103) for zonisamide 25 mg, and 0.138 6 0.907 h/d (P 5 0.235) for zonisamide 50 mg. The reduction did not increase dose dependently, although the duration increased for the zonisamide treatments compared with placebo. Further analysis showed that zonisamide did not increase troublesome dyskinesia (Table 2).

[6] 89w The incidence of adverse events was 49.6% (65/131 patients) for placebo, 57.7% (75/130 patients) for zonisamide 25 mg, and 60.9% (78/128 patients) for zonisamide 50 mg. Adverse events with an incidence of at least 3% are presented in Table 3. The incidences of hallucination and dyskinesia for the zonisamide treatments were similar to those for placebo (hallucination; 0.8% for placebo, 1.5% for zonisamide 25 mg, 1.6% for zonisamide 50 mg: dyskinesia; 7.6%, 6.9%, 7.0%, respectively). Only somnolence and constipation occurred more frequently for the zonisamide treatments than for placebo.

[1] 288w The current study showed that 50 mg zonisamide significantly reduced "off" time and improved "off" level of activity of daily life (UPDRS II) in patients with PD. Our previous randomized, double-blind study (phase 2b/3) 5 showed that zonisamide reduced "off" time (reduction; 0.253 h/d [P 5 0.866] for zonisamide 25 mg and 1.395 h/d [P 5 0.005] for zonisamide 50 mg vs. 0.178 h/d for placebo). In the current study, the efficacy of zonisamide on wearing-off in patients with PD has been confirmed, but the mean reduction was 0.7 h/d, not so long. For dopamine agonists, 1-to 2-h/d prolongation of "on" time was reported, 23 but the most study patients were treated with only L-dopa. Concomitant rate of a dopamine agonist is quite higher in Japan than in Western countries (approximately 90% vs. approximately 50%). A Japanese study showed that the "off" time reduction was greater for rotigotine than for placebo by 1.4 h/ d. 24 A Japanese study on entacapone showed that the "on" time prolongation was 1.4 h/d, but strong placebo effects indicated that the difference from placebo was only 0.9 h/d. 25 Our study included the placebo run-in period to minimize placebo effects; the reduction for placebo was only 0.011 h/d, indeed. Approximately 90% of the study patients had been treated with a dopamine agonist, approximately 50% a MAO-B inhibitor, and approximately 40% a catechol-O-methyltransferase inhibitor. This means that the study patients had insufficiently responded to the treatment consisting of 3.2 antiparkinson drugs (mean), including L-dopa. Although the mean reduction of "off" time is not so long, zonisamide significantly improved the duration of off time with improvement of "off" level (UPDRS II), so that we think this study showed clinical efficacy of zonisamide on wearing-off.

[2] 106w The use of LOCF, which was used for the analysis of "off" time, to deal with missing data has been widely criticized. 26,27 Therefore, a post-hoc sensitive analysis was performed, using the MMRM method. This method can lead to estimators with comparatively small bias, and control type I error rates at a nominal level in a situation in which the missing mechanism on the available data is missing completely at random or missing at random. 28 The sensitive analysis results also showed that zonisamide 50 mg was superior to placebo for reduction in "off" time, which was consistent with the results of primary analysis using LOCF.

[3] 77w In the phase 2b/3 study, 5 the "off" time reduction for zonisamide 50 mg was 1.4 h/d, and the difference from placebo was 1.2 h/d. The reduction in the present study was,smaller. In the period between the phase 2b/3 and present studies in Japan, pramipexole, ropinirole, entacapone, and zonisamide 25 mg have been becoming available. Such better treatment circumstance, ie, variety of available antiparkinson drugs, is one of reasons for the smaller reduction in the current study.

[4] 109w The study also showed that zonisamide therapy is associated with a very low incidence of specific adverse events, including dyskinesia and hallucination over a 12week period, which is similar to the results in the phase 2b/3 study. This suggests that zonisamide is effective for advanced PD patients receiving many antiparkinson drugs with rare occurrence of dyskinesia and hallucination, which are not avoidable side effects of dopaminergic drugs. Although the exact mechanism of the effects of zonisamide on PD has not yet been elucidated, the fact that zonisamide has both dopaminergic and nondopaminergic effects may have contributed to the efficacy on PD, with a low incidence of such adverse events.

[5] 144w In the current study we could not show significant benefit on UPDRS Part III assessment, as demonstrated in our two previous studies. The reason for this may be that the subjects of the present study were typical wearing-off patients; that is, they responded well to L-dopa and additional standard drugs (the level of "on" was good), but the duration of effect of those drugs was shortened. Because the level of "on" state at the baseline was very mild, a significant improvement by zonisamide could not be detected. Indeed, the UPDRS Part III score at baseline in the present study was lower than or similar to the scores at the final assessment in the previous studies (baseline UPDRS Part III total score [mean 6 SD]: the phase 2b/3 study, 23.61 6 12.20; the phase 3 study, 21.95 6 9.49; the current study, 17.52 6 11.28).

[6] 12w This study showed that zonisamide significantly improved wearing-off without increasing troublesome dyskinesia.

[1] 23w The study patients had PD that met Ward and Gibb's diagnostic criteria 17 for the disease. The major inclusion criteria were the following:

[2] 108w 1. Patients who provided written voluntary consent to participate in the study 2. Men or women aged 20 through 74 y 3. Patients who were treated with any combination drugs of levodopa and dopa decarboxylase inhibitor for at least 6 consecutive months (182 consecutive d) and whose condition had responded to levodopa during the first few years of levodopa therapy 4. Patients whose regimen of levodopa and other antiparkinson drugs remained unchanged from 2 weeks before the run-in period and through the end of the run-in period 5. Patients whose mean daily "off" time was at least 2 h for the last 7 d of the run-in period

[3] 7w The major exclusion criteria were the following:

[4] 69w 1. Patients with epilepsy or Parkinson's syndrome excluding PD 2. Patients receiving a levodopa monotherapy 3. Patients with any histories of surgery for PD within 6 months before screening, or any histories of treatment with zonisamide or experiences of participation in any other clinical studies of zonisamide 4. Patients with any severe psychiatric symptom or sign, cognitive impairment (ie, Mini-Mental State Examination score of 21), or serious diseases 5.

[5] 48w Patients with any histories of serious hepatic disease, malignant syndrome, or drug allergy, or any existing or previous malignant tumor within 5 y before screening 6. Women who were or might be pregnant, who did not practice effective contraception and were of childbearing potential, or who were breastfeeding

[6] 87w Randomization was performed according to a computer-generated schedule with a 1:1:1 ratio, using permuted-block design (the fixed block size; 6). The investigators did not know the block size. The concealment of allocation sequence and the blindness was ensured by the double dummy method. Indistinguishability of the study drugs was confirmed before the study initiation and when unblinding. Key codes excluding one patient kept blinded from the study initiation through database lock. His emergency key code was unblinded because of an adverse event after his data were locked.

[7] 133w Clinical assessment including the Unified Parkinson's Disease Rating Scale (UPDRS) 18 was conducted with the patients in the "on" state every 4 weeks. The investigators (raters) received specific training for the UPDRS assessment provided by the sponsor. Quality of life was assessed by the patients, using the Parkinson's Disease Questionnaire-39 (PDQ-39) 19 at baseline and the end of the study. The patients were instructed to record "on" and "off" state ("on" means the medication was effective [easy to move], "off" means not), and the occurrence of dyskinesia every 30 min while awake on patient diaries for the last 7 d before each visit (excluding screening visit). As safety measurements, vital signs and laboratory tests were evaluated every visit and 12-lead resting electrocardiography at screening, baseline, and week 12. Adverse events were also evaluated.

[8] 68w The study was conducted in accordance with ethical principles that have their origin in the Declaration of Helsinki, with Good Clinical Practice, and with local regulations. The protocol was reviewed and approved by Institutional Review Board of each study site before the study initiation. All of the patients were fully informed of the objectives, procedures, and possible benefits and risks of the trial and provided written voluntary consent.

[9] 371w The efficacy analyses were performed on the full analysis set (FAS). The FAS included all of the patients who were randomized, took at least 1 dose of the study drug, and had at least 1 post-baseline "off" time datum during the treatment period. The primary efficacy endpoint was the change from baseline in the daily "off" time at week 12. "Off" time was calculated using the diary information for the last 7 d before each visit (excluding screening visit). When only less-than-5-d "off" time data were available, the datum at the visit was handled as missing. Missing values were inputted by the last observation carried forward (LOCF) method. An analysis of covariance model with treatment group as factor, and baseline value as covariate was used to compare the zonisamide treatments (25 mg/d and 50 mg/d) with placebo. A fixed sequence test was used to control the overall type I error at 5% for the primary endpoint. If zonisamide 50 mg was superior to placebo, then 25 mg was to be tested to compare with placebo. Other efficacy endpoints (excluding UPDRS and PDQ-39) were analyzed in the same way, using the fixed sequence testing procedure. For analysis on UPDRS and PDQ-39, Dunnett's test was used because the dose dependency was not clear in the previous studies or no data were available at designing the study. However, these tests could not adjust the multiplicity between multiple variables other than the primary. The mixed model for repeated measurements (MMRM) [20][21][22] was used to evaluate the change from baseline in the off time at week 12, which is the primary endpoint, for the sensitivity analysis. The model included treatment group (placebo, zonisamide 25 mg, and zonisamide 50 mg; as a categorical variable), visit (weeks 4, 8, and 12; as a categorical variable), baseline "off" time, and the treatment-by-visit interaction. An unstructured covariance matrix was used for the within-subject correlation, and the Kenward-Rogers approximation was used to calculate the denominator degree of freedom. As a post-hoc analysis after treatment unblinding, a Z O N I S A M I D E I M P R O V E S W E A R I N G -O F F Movement Disorders, Vol. 30, No. 10, 2015

[10] 20w responder was defined as a patient with at least 20% reduction in "off" time from the baseline at week 12.

[11] 24w Comparison between the zonisamide treatments and placebo in the responder rate was performed by using a logistic regression model with treatment and baseline value.

[12] 79w The planned sample size of 114 patients per group was based on the results of the previous phase 2b/3 study. The sample size was intended to provide 85% power to detect a difference of 1.0 (standard deviation [SD] 5 2.5) h in the mean change in the daily "off" time for zonisamide 50 mg compared with placebo at the two-sided 5% level of significance. Allowing for anticipated dropouts, the target number of patients was 120 per group (total, 360).

[1] 35w In Parkinson's disease (PD), long-term levodopa therapy may cause motor complications such as wearing-off and dyskinesia. 1 Although several options exist, unmet needs for patients, such as improving wearing-off without increasing dyskinesia, continue to occur.

[2] 115w Zonisamide (1,2-benzisoxazole-3-methanesulfonamide) 2 was launched in Japan in 1989 as an antiepileptic and has subsequently been approved in countries throughout the world. When zonisamide was administered to a Japanese patient with PD who had developed a seizure in the clinical setting, zonisamide was effective in improving the symptoms of PD as well as the seizure disorder. 3 We then conducted three double-blind, placebo-controlled studies (phase 2, 2b/3, and 3 studies), and these studies showed that the adjunctive administration of zonisamide was efficacious and safe for patients treated with a combination of levodopa and other antiparkinson drugs. 4,5 Zonisamide showed beneficial effects on motor ------------------------------------------------------------*Correspondence to: Miho Murata; 4-1-1, Ogawa-Higashi, Kodaira, Tokyo, 187-8551, Japan, E-mail: mihom@ncnp.go.jp

[3] 12w Funding agencies: This study was supported by Sumitomo Dainippon Pharma Co., Ltd.

[4] 44w Relevant conflicts of interest/financial disclosures: Miho Murata receives a consulting fee from Sumitomo Dainippon Pharma Co., Ltd. Kazuko Hasegawa has nothing to report, Ichiro Kanazawa has nothing to report. Junichi Fukasaka, Kenji Kochi, and Rieko Shimazu are employees of Sumitomo Dainippon Pharma Co., Ltd.

[5] 16w Full financial disclosures and author roles may be found in the online version of this article.

[6] 42w symptoms of PD with a very low incidence of adverse events, including dyskinesia and hallucination. As a result, 25 mg/d zonisamide was approved in 2009 in Japan as an adjunctive treatment in patients with PD whose condition responded insufficiently to levodopa treatment.

[7] 155w Zonisamide has multiple functions, including inhibition of sodium channels, 6 T-type calcium channels, 7,8 monoamine oxidase-B (MAO-B) activity, 9,10 and striatal d 1 -receptor-associated gamma aminobutyric acid-ergic transmission, 11 and activation of dopamine synthesis 6,9 and dopamine release. 9,12 Although the exact mechanism of zonisamide on PD has not yet been elucidated, both the dopaminergic and nondopaminergic functions of zonisamide may contribute to the effects on PD with a low incidence of dyskinesia and hallucination. Zonisamide has been reported to have neuroprotective effects both in vitro 13,14 and in vivo. 15,16 Zonisamide has a very long half-life (25 mg; t 1/2 90 h); therefore, its beneficial effects on wearingoff phenomena is expected. Indeed, the phase 2b/3 study also showed that zonisamide 50 mg reduced "off" time. 5 In the present study, we sought to determine the efficacy of zonisamide for treatment of "off" time in PD patients with the wearing-off phenomena by comparing with placebo.

[8] 156w The study was conducted in a multicenter, randomized, double-blind, placebo-controlled, parallel-group manner. It consisted of a 4-week placebo run-in period and a 12-week double-blind treatment period. Patients were required to visit the study site 5 times (including screening visit) every 4 weeks. Patients received placebo orally once daily during the run-in period in a single-blind manner to minimize placebo effects. Immediately before the double-blind treatment period (baseline, week 0), eligibility was evaluated, and the eligible patients were randomly assigned to one of three treatment groups: the placebo, zonisamide 25mg, and zonisamide 50-mg groups. They received the study drug orally once daily during the treatment period. The dosage and regimen of patients' ongoing antiparkinson drugs and other drugs that may affect PD symptoms remained unchanged from 2 weeks before the start of the run-in period to the end of the treatment period. The study was conducted from August 2010 to December 2011 at 65 hospitals throughout Japan.

[9] 45w . Mean number of concomitant antiparkinson drugs was 3.2, including L-dopa. At baseline, the mean "off" time was 6.37 h/d, and UPDRS Part III total score was 17.5. No significant differences were found in the baseline characteristics among the treatment groups in any analysis set.