- Email: [email protected]
A Randomized, Double-Blind, Crossover Comparison of Risperidone and Haloperidol in Korean Dementia Patients With Behavioral Disturbances
A Randomized, Double-Blind, Crossover Comparison of Risperidone and Haloperidol in Korean Dementia Patients With Behavioral Disturbances Guk-Hee Suh, M.D., Ph.D., Hyun Gyun Son, M.D., Ph.D. Young-Su Ju, M.D., M.P.H., Ph.D., Kyeong Hyeong Jcho, M.D., Ph.D. Byeong Kil Yeon, M.D., Ph.D., Young Min Shin, M.D., Ph.D. Baik Seok Kee, M.D., Ph.D., Sung-Ku Choi, M.D.
Objective: Behavioral disturbances in dementia are extremely prominent and distressful, and often result in serious physical, social, and economic consequences. The authors compared the efficacy and tolerability of risperidone and haloperidol in the treatment of behavioral and psychological symptoms of dementia (BPSD) in institutionalized elderly Korean patients with Alzheimer disease, vascular dementia, or mixed dementia. Methods: This was an 18-week double-blind, crossover study involving 120 patients who were randomly assigned to receive flexible doses (0.5–1.5 mg/day) of risperidone or haloperidol. BPSD were assessed using the Korean version of the Behavioral Pathology in Alzheimer’s Disease Rating Scale (BEHAVE-AD-K), the Korean version of the Cohen-Mansfield Agitation Inventory (CMAI-K), and the Clinical Global Impression of Change scale (CGI-C). Safety and tolerability assessments included the Extrapyramidal Symptom Rating Scale and the incidence of adverse events. Results: Both risperidone and haloperidol were efficacious in alleviating BPSD. However, when receiving risperidone, patients showed significantly greater improvement than when receiving haloperidol in the total and subscale scores of the BEHAVE-ADK, the total and subscale scores of the CMAI-K, and the scores on the CGI-C scale. Also, risperidone had an additional benefit on aggressiveness and anxieties/phobias. The risk of antipsychotic-induced parkinsonism throughout this study was significantly lower with risperidone than with haloperidol. Conclusion: Risperidone had a favorable efficacy and tolerability profile compared with haloperidol in the treatment of BPSD in this patient population. (Am J Geriatr Psychiatry 2004; 12:509–516)
Received October 18, 2002; revised March 13, May 22, 2003; accepted June 17, 2003. From the Department of Psychiatry, Hallym University Medical Center, Hangang Sacred Heart Hospital, Seoul, Korea (GHS,HGS), the Department of Occupational & Environmental Medicine, Hallym University Sacred Heart Hospital, Seoul, Korea (YSJ), the Department of Psychiatry, Kangnam Hospital Public Corporation, Seoul, Korea (KHJ,YMS), the Department of Psychiatry, Hallym University Medical Center, Kangdong Sacred Heart Hospital, Seoul, Korea (BKY), the Department of Psychiatry, Pil Dong Hospital, Chung Ang University, College of Medicine, Seoul, Korea (BSK), and the Medical Department, Janssen Korea, Seoul, Korea (SKC). Send correspondence to Associate Professor Guk-Hee Suh, Department of Psychiatry, Hallym University Medical Center, Hangang Sacred Heart Hospital, 94-200 Yungdungpo-Dong, Seoul 150-030, Korea. e-mail: [email protected] 䉷 2004 American Association for Geriatric Psychiatry
Am J Geriatr Psychiatry 12:5, September-October 2004
509
Comparison of Risperidone and Haloperidol
D
ementia is a chronic illness with serious physical, social, and economic consequences.1 The prevalence of dementia among Koreans age 65 years and older is 7%–10%, and this is increasing because of an upsurge in the number and proportion of people in the population who are age 75 years or older.1 Behavioral and psychological symptoms of dementia (BPSD) are very prominent clinically and are often the source of significant distress to caregivers, leading them to seek professional help. This can result in premature institutionalization, with increased costs of care and significant reduction of quality of life, not only for the patients,2 but also for their families and caregivers.3 Despite the serious consequences of these symptoms, there is a lack of clinical trials for the treatment of BPSD. Antipsychotics are generally acknowledged to be more effective than a variety of other pharmacological agents in the treatment of patients with BPSD.4 Data support a modest efficacy for conventional antipsychotics,5,6 but the use of these drugs in dementia patients has been restricted by their potential for serious side effects. Theoretically, novel antipsychotics have a lower potential to cause extrapyramidal symptoms (EPS) than conventional neuroleptics. The results of short-term studies indicate that atypical antipsychotics, especially risperidone, are generally effective and well tolerated in the treatment of patients with BPSD,7–10 but there is a lack of efficacy data in non-Caucasian populations. This study was designed to compare the efficacy and tolerability of risperidone and haloperidol in the treatment of BPSD in Korean patients with dementia.
METHODS Patient Population All 280 dementia patients residing at the Seoul City Jung-Kye Welfare Center for the Elderly (a semi-hospitalized long-term care institution in Seoul) were screened for eligibility in this study, which was conducted from June to December 2001. Patients were under the full-time medical surveillance of psychiatrists, medical doctors, and nurses and were looked after by professional caregivers. The professional caregivers assisted the dementia patients with drug
510
intake and the study protocol and provided information required at assessment interviews. Eligibility criteria included an age of 65 years or over, a diagnosis of dementia of the Alzheimer type with behavioral disturbance, vascular dementia with behavioral disturbance, or a combination of the two, according to the criteria of the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition (DSMIV).11 Eligible patients had a score of 4 or higher on the Functional Assessment Staging Test (FAST),12 a total score of 8 or higher on the Korean version of the Behavioral Pathology in Alzheimer’s Disease Rating Scale13 (BEHAVE-AD-K; Suh et al., data on file), and a score of more than 3 on any two items of the Korean version of Cohen-Mansfield Agitation Inventory (CMAI-K)14 (Suh et al., data on file). Exclusion criteria included other conditions that diminish cognitive function (e.g., Lewy-body dementia, hypothyroidism), other psychiatric disorders that might contribute to the psychotic symptoms (e.g., schizophrenia, delusional disorder), clinically relevant organic or neurologic disease, unstable medical conditions (e.g., poorly controlled hypertension, angina, or diabetes), abnormal electrocardiograms as diagnosed by a cardiologist or laboratory tests, a history of allergic reaction to antipsychotic treatment, and a history of neuroleptic malignant syndrome. The study was conducted in accordance with Good Clinical Practices and the Declaration of Helsinki. The study protocol was reviewed and approved by the independent Ethics Committee of Hallym University. Each patient or his or her legal guardian provided written informed consent for participation after the procedure had been fully explained. Design and Procedures This was a randomized, double-blind, 18-week crossover study that included four periods: a 1-week washout period during which all psychotropic medications were discontinued; an 8-week double-blind active treatment period (Phase I), a second 1-week washout period, and a final 8-week crossover activetreatment period (Phase II). All consenting patients were randomly assigned to either the risperidonefirst group or the haloperidol-first group according to a predefined randomization code. At each visit after the Phase I baseline evaluation (Week 0), eligible patients were evaluated with the
Am J Geriatr Psychiatry 12:5, September-October 2004
Suh et al. BEHAVE-AD-K, CMAI-K, and Clinical Global Impression of change (CGI-C) scale.15 The 25-item BEHAVE-AD scale is specifically designed to assess the severity of BPSD and provides seven subscales: paranoid and delusional ideation, hallucinations, activity disturbances, aggressiveness, diurnal rhythm disturbances, affective disorders, and anxieties and phobias.13 The sum of these subscales provides the total BEHAVE-AD score. The Psychosis subscale is the sum of the Paranoid and Delusional Ideation and Hallucination subscales. The 29-item CMAI scale is specifically designed to assess the frequency of manifestations of agitation-related behaviors in elderly persons with dementia and provides a Physical (e.g., hitting) and Verbal (e.g., screaming) Aggression scale, which can be combined to give a Total Aggression score, as well as a Physical (e.g., wandering) and Verbal (e.g., grunting) Non-Aggression scale.14 Also, the CGI-C scale is designed to assess global symptoms and patient’s overall condition and reflects the relevance of changes in behavioral ratings. The CGI-C is a 7-point scale ranging from 1: much better, through 4: unchanged, to 7: much worse.15 At baseline, all patients were scored as 4 (unchanged). All reported adverse events were recorded, and the severity of EPS such as parkinsonism, akathisia, dystonia, and dyskinesia was assessed by use of the Extrapyramidal Symptom Rating Scale (ESRS).16 After randomization and the initiation of double-blind treatment, patients were assessed weekly during the first 4 weeks of the study and then every 2 weeks (twice) until the end of the final (8th) week. Vital signs, such as blood pressure, pulse rate, and respiration rate were checked routinely and monitored every other day by day-duty nurses. Standard 12lead electrocardiograms and laboratory tests were done at the initial screening period and interim washout period. Study Drug The study drugs, haloperidol (Myung In. Pharm. Co. Ltd., Seoul, Korea) and risperidone (Janssen Korea, Seoul, Korea), were blinded in non-transparent capsules (0.5 mg/tablet) and packaged individually with the patient-identification label. The randomization code was provided by a statistician who was not involved in the clinical study. The study drug was dispensed twice: before the start of Phase I and before
Am J Geriatr Psychiatry 12:5, September-October 2004
the start of Phase II. Before the new drug was distributed, the remaining drug from Phase I was collected. Two psychiatrists, who were blinded to the study drug and not involved in the rating, prescribed 0.5 mg h.s. of either drug at Week 0. If needed, the drug dosage could be adjusted by increments of 0.5 mg, once every 4 or more days when one of two psychiatrists visited the institution to treat the patients. The target dose for this study was 1 mg/day, but the dose could be stepped up to 1.5 mg/day if symptoms were poorly controlled. The dosage could also be titrated downward at any time and by any amount in the event of clinically significant treatment-emergent adverse events or if, in the prescriber’s judgment, a decrease was warranted. The dosing regimen and intervals between dose titrations were individualized for each patient. Concomitant use of antipsychotic drugs, antidepressants, and mood stabilizers was not permitted. Lorazepam was permitted if limited to 4 days/week for the first 4 weeks of active treatment. Statistical Analysis A sample size of 98 patients per drug was selected in order to detect a 4-point difference in the total score on the BEHAVE-AD-K with 80% power and a twotailed level of significance of 0.05. The sample size was then increased to 120 patients to account for a 20% rate of dropout. In order to increase the power of the study and reduce the sample size, a two-phase crossover design was applied. Primary analyses were planned for all patients who underwent randomization and assessment at least once during drug treatment, but because of the low number of patients who discontinued the study early, the final analyses consisted of data from patients who completed drug treatment per protocol. The merged dataset for each drug (which included the data for all patients who received the drug during Phase I and all those who received it during Phase II) was used for analysis. Baseline measures were taken at Week 0 for those who received the drug during Phase I, and Week 9 for those who received the drug during Phase II; endpoint measures were taken at Week 8 for those who received the drug during Phase I, and Week 18 for those who received the drug during Phase II. We applied univariate analyses to calculate basic statistics and to test the normality (either continuous or
511
Comparison of Risperidone and Haloperidol binary distributed) of all variables. We used longitudinal data-analysis methods to control autocorrelation and to conduct multivariate analyses, as the study design was ultimately longitudinal. In the cases of continuous dependent variables (BEHAVE-AD-K [Delusion, Psychosis {Delusion Ⳮ Hallucination}, Activity Disturbance, Aggressive Behavior, Anxieties and Phobias, Total Score]; CMAI-K [Aggressive Behavior, Physically Non-Aggressive Behavior, Verbally Agitated Behavior, Total score]; ESRS [parkinsonism, total score]; and CGI-C), we applied the mixed model. In the cases of binary dependent variables (BEHAVEAD-K [Hallucination, Diurnal Rhythm Disturbances, Affective Disturbances] and ESRS [dystonia, dyskinetic movement]), we applied Generalized Estimating Equations (GEEs). Multivariate analyses, using the mixed model and GEEs where appropriate, were conducted using the following independent variables: drug (risperidone or haloperidol), weeks of treatment (0, 1, 2, 3, 4, 6, or 8 weeks) and individual indicator (to adjust individual variability) as covariates. Within-group comparisons (baseline to endpoint) were analyzed with paired t-tests. Statistical significance was assumed at p ⱕ0.05. The statistical package SAS (Version 6.12) was used for all analyses.
The two groups were similar with respect to baseline demographics and disease characteristics. Furthermore, there were no statistically significant differences at the two baselines, before crossover (Week 0), and after crossover (Week 9) for any variables on the BEHAVE-AD-K, CMAI-K, CGI-C scale, and ESRS between the patients who received risperidone during Phase I and those who received it during Phase II, and between patients who received haloperidol during Phase I and in Phase II. Efficacy The mean baseline and endpoint scores on the BEHAVE-AD-K, CMAI-K, and CGI-C scale are listed in Table 2. At the endpoint of treatment, when compared with baseline by paired t-tests, risperidone resulted in significant improvement in scores on the CMAI-K Total scale (t[112]⳱3.03; p⳱0.003) and the subscales of Aggressive Behavior (t[113]⳱8.44; p ⬍0.0001) and Physical Non-Aggressive behavior (t[113]⳱7.34; p ⬍0.0001). Significant improvements were also observed for the BEHAVE-AD-K Total score (t[113]⳱12.7; p ⬍0.0001) and all of its subscales FIGURE 1.
Schematic Presentation of Flow in this Crossover Comparative Clinical Study of Risperidone and Haloperidol
RESULTS Of the 120 randomized patients, 114 (95%) completed the study and were included in the final data analysis. The reasons for patient discontinuation are shown in the study flow diagram (Figure 1). The mean (standard deviation [SD]) age of patients was 80.9 (8.2) years, and 82% were female. Alzheimer disease (AD) was the most common diagnosis, occurring in 65.8% of the patients, followed by vascular dementia (28.3%) and mixed dementia (5.8%). The mean score on the Korean version of the Mini Mental State Exam (MMSE-K)17 at the baseline evaluation was 9.6 (6.6), which was not statistically different from the mean score on the MMSE-K at the end of study. Of all the patients, 66% were at FAST Stage 6A or higher. The mean duration of AD was 5.4 years, that of vascular dementia was 5.7 years, and that of mixed dementia was 4.3 years (Table 1). During the last week of treatment, the mean daily dose of haloperidol was 0.83 (0.35) mg and that of risperidone was 0.80 (0.32) mg.
512
Patients screened (N=280)
Patients enrolled and randomly assigned (N=120) Phase I Risperidone (N=60)
Haloperidol (N=60)
Discontinued: Seizure (N=1) Nausea (N=1)
Discontinued: Somnolence (N=1)
Haloperidol (N=58)
Risperidone (N=59)
Discontinued: Somnolence (N=2)
Discontinued: Nausea (N=1)
Completed trial (N=56)
Completed trial (N=58)
Phase II
Am J Geriatr Psychiatry 12:5, September-October 2004
Suh et al. (t ⱖ3.43; p ⬍0.001). Patients significantly improved as observed on the CGI-C scale (t[113]⳱3.61; p⳱0.0008). Haloperidol, when compared with baseline, resulted in significant improvement in the BEHAVE-AD-K Total score (t[113]⳱5.10; p ⬍0.0001) and its subscales of Psychosis (t[113]⳱2.96; p⳱0.004), Activity Disturbances (t[112]⳱5.42; p ⬍0.0001), Aggressiveness (t[113]⳱5.58; p ⬍0.0001), and Affective Disturbances (t[113]⳱2.99; p⳱0.003), as well as the TABLE 1.
CMAI-K Total (t[112]⳱14.2; p ⬍0.0001) and subscale of Aggressive Behavior (t[113]⳱7.41; p ⬍0.0001). Risperidone (N⳱114) was superior to haloperidol (N⳱114) on a variety of efficacy variables, as assessed by the mixed model for the BEHAVE-AD-K Total score (numerator degrees of freedom [NDF]⳱1; denominator degrees of freedom (DDF)⳱1,398; Type III F⳱8.58; p⳱0.004) and for the BEHAVE-AD-K subscales: Aggressiveness (NDF: 1, DDF: 1,397; Type III
Demographics and Baseline Assessment Scores
Characteristic
RIS-to-HAL Group Nⴔ60
HAL-to-RIS Group Nⴔ60
Both Nⴔ120
13/47 80.4 (5.6) 65–96
11/49 81.5 (7.7) 65–97
24/96 80.9 (8.2) 65–97
42 (70%) 15 (25%) 3 (5%)
37 (61.7%) 19 (31.7%) 4 (6.6%)
79 (65.8%) 34 (28.3%) 7 (5.8%)
19.3 (7.7) 86.3 (20.2) 10.2 (20.3) 9.7 (7.5)
19.2 (8.8) 89.4 (21.4) 8.8 (16.0) 9.4 (5.2)
19.2 (8.2) 87.8 (20.7) 9.6 (18.1) 9.6 (6.6)
19 41
22 38
41 79
Sex, M/F Age, years, mean (SD) range Diagnosis (DSM-IV), N (%) Alzheimer disease Vascular dementia Mixed dementiaa Mean baseline scores, total (SD) BEHAVE-AD-K CMAI-K ESRS MMSE-K FAST stage 4–5 6a–7e
Note: RIS-to-HAL group: risperidone during Phase I, haloperidol during Phase II; HAL-to-RIS group: haloperidol during Phase I, risperidone during Phase II; SD: standard deviation; DSM-IV: Diagnostic and Statistical Manual of Mental Disorders, 4th Edition; BEHAVE-AD-K: Korean version of the Behavioral Pathology in Alzheimer’s Disease Rating Scale; CMAI-K: Korean version of the Cohen-Mansfield Agitation Inventory; ESRS: Extrapyramidal Symptom Rating Scale; MMSE-K: Korean version of the Mini-Mental State Exam; FAST: Functional Assessment Staging Test. a These diagnoses were based on the results of neuroimaging.
TABLE 2.
Differences in Efficacy Between Risperidone and Haloperidol in the Treatment of Patients With BPSD According to the Changes in the Scores on the BEHAVE-AD-K, CMAI-K, and CGI-C Risperidone Nⴔ114
Haloperidol Nⴔ114
Differencea
Assessment
Baseline
Endpoint
Baseline
Endpoint
p
BEHAVE-AD-K Total score Psychosis Activity Disturbances Aggressiveness Diurnal Rhythm Disturbances Affective Disturbance Anxieties and Phobias CMAI-K Total score Aggressive Behavior Physical Non-Aggressive Behavior Verbally Agitated Behavior CGI-C
19.2 (0.7) 8.3 (0.4) 3.8 (0.2) 3.1 (0.2) 0.9 (0.1) 0.7 (0.1) 2.5 (0.2) 89.3 (1.9) 22.2 (0.6) 17.2 (0.6) 15.1 (0.5) 4
12.0 (0.5) 4.6 (0.3) 2.7 (0.2) 2.0 (0.1) 0.4 (0.1) 0.2 (0.1) 2.2 (0.1) 75.1 (1.7) 18.2 (0.5) 14.8 (0.5) 14.0 (0.4) 3.9 (0.1)
19.2 (0.8) 8.0 (0.4) 3.8 (0.2) 3.2 (0.2) 0.7 (0.1) 0.5 (0.1) 2.3 (0.2) 89.4 (2.0) 23.6 (0.7) 17.8 (0.5) 15.7 (0.5) 4
14.5 (0.6) 6.0 (0.3) 3.0 (0.2) 2.3 (0.1) 0.5 (0.1) 0.3 (0.1) 2.4 (0.1) 83.5 (1.8) 20.3 (0.5) 16.8 (0.5) 15.2 (0.4) 4.2 (0.1)
0.004 0.582 0.858 0.002 0.038 0.248 ⬍0.0001 ⬍0.0001 0.001 0.024 0.002 0.001
Note: Values are mean (standard error). These results are from the merged data for each drug. BPSD: Behavioral and Psychological Symptoms of Dementia; BEHAVE-AD-K: Korean version of the Behavioral Pathology in Alzheimer’s Disease Rating Scale; CMAI-K: Korean version of the Cohen-Mansfield Agitation Inventory; CGI-C: Clinical Global Impression of Change. a p-value showing difference between risperidone group and haloperidol group as assessed by multivariate analyses, using a mixed model or Generalized Estimating Equations, using the independent variables: drug (risperidone or haloperidol), weeks of treatment (0, 1, 2, 3, 4, 6, or 8 weeks) and individual indicator (to adjust individual variability) as covariates.
Am J Geriatr Psychiatry 12:5, September-October 2004
513
Comparison of Risperidone and Haloperidol F⳱10.2; p⳱0.002) and Anxieties and Phobias (NDF: 1; DDF: 1,398; Type III F⳱14.75; p⳱0.0001). Risperidone was also superior to haloperidol as assessed by the mixed model for the CMAI-K Total score (NDF: 1; DDF: 1,398; Type III F⳱16.26; p⳱0.0001) and for the CMAI-K subscales Aggressive Behavior (NDF: 1; DDF: 1,395; Type III F⳱11.0; p⳱0.001), Physically Non-Aggressive Behavior (NDF: 1; DDF: 1,397; Type III F⳱3.96; p⳱0.024) and Verbally Agitated Behavior (NDF: 1; DDF: 1,397; Type III F⳱8.09; p⳱0.002). Statistically significant differences favoring risperidone were also observed for CGI-C score (mixed model, NDF: 1; DDF: 1,397; Type III F⳱39.05; p⳱0.0001) and Diurnal Rhythm Disturbance (GEE: N⳱114; z ⳱ ⳮ2.075; p⳱0.038; Table 2, Figure 2). Drug Safety and Tolerability There were no statistically significant differences between the risperidone-first group and the haloperidol-first group in baseline scores on the ESRS or its three subscales of parkinsonism: expressive automatic movements (bradykinesia, rigidity, gait and posture, tremor, akathisia, sialorrhea, postural instability), dystonia, or dyskinetic movements. When compared with baseline, risperidone was not associFIGURE 2.
Mean Total Scores on the Korean version of the Behavioral Pathology in Alzheimer’s Disease Rating Scale (BEHAVE-AD-K) for Dementia Patients Receiving Risperidone and Haloperidol
20
Risperidone
Mean Total Score on BEHAVE-AD-K
Haloperidol 18
16
14
12
10 0
1
2
3
4
5
Time (weeks)
514
ated with a significant increase in EPS over the study period, as determined by the total and three subscale scores on the ESRS. In contrast, haloperidol was associated with a significant worsening of EPS over the course of the study, as revealed by a significant increase in the total and parkinsonism subscale scores of the ESRS when compared with baseline by the mixed model (NDF: 1; DDF: 698; Type III F⳱4.846; p⳱0.0362; and NDF: 1; DDF: 696; Type III F⳱225.17; p⳱0.0001, respectively). When the differences between baseline and endpoint of risperidone and haloperidol were compared directly by the mixed model, both the total ESRS and the parkinsonism subscale were significantly increased with haloperidol treatment (NDF: 1; DDF: 1,193; Type III F⳱159; p⳱0.0001; NDF: 1; DDF: 1,194; Type III F⳱239; p⳱0.0001; Table 3, Figure 3). Furthermore, compared with risperidone by means of GEEs, somnolence (N⳱114; z ⳱ –6.63; p⳱0.0001), insomnia (N⳱114; z ⳱ –3.77; p⳱0.0001), and sialorrhea (N⳱114; z ⳱ –4.04; p⳱0.0001) were reported significantly more often when patients received haloperidol. Most of the patients tolerated the study drugs. There was no significant change in vital signs over the course of the study. A total of six patients discontinued the study early because of adverse events: seizure (N⳱1) and nausea (N⳱2) in the patients treated with risperidone, and somnolence (N⳱3) in the patients treated with haloperidol. The seizure was not considered drug-related.
6
7
8
DISCUSSION Both risperidone and haloperidol were efficacious in alleviating BPSD. With both study drugs, significant reductions were noted in the total scores on the BEHAVE-AD-K and its subscale scores of psychosis, activity, and affective disturbances, as well as in the total scores on the CMAI-K and its aggressive behavior subscale scores. However, when receiving risperidone, patients showed significantly greater improvement than when receiving haloperidol on the CMAIK Total and all subscale scores. Risperidone had an additional benefit on BEHAVE-AD-K Total, and Aggressiveness, Diurnal Rhythm Disturbances, and Anxieties and Phobias subscales that was not observed with haloperidol. Furthermore, a statistically
Am J Geriatr Psychiatry 12:5, September-October 2004
Suh et al. TABLE 3.
Differences in Tolerability Between Risperidone and Haloperidol in the Treatment of Patients With BPSD According to the Changes in Scores on the ESRS Risperidone Nⴔ114
Haloperidol Nⴔ114
Difference
Assessment
Baseline
Endpoint
p*
Baseline
Endpoint
p**
p***
ESRS Total score Parkinsonism Dystonia Dyskinetic movement
8.7 (0.9) 6.1 (1.3) 0.8 (0.5) 1.6 (0.4)
13.5 (1.6) 9.6 (1.1) 1.8 (0.5) 2.1 (0.4)
0.6606 0.1564 0.7498 0.2338
8.8 (1.5) 6.4 (1.0) 0.7 (0.3) 1.6 (0.4)
22.6 (2.1) 16.8 (1.4) 3.2 (0.8) 2.5 (0.5)
0.0362 ⬍0.0001 0.0007 0.1627
0.0001 0.0001 0.6503 0.4144
Note: Values are mean (standard error). BPSD: Behavioral and Psychological Symptoms of Dementia; ESRS: Extrapyramidal Symptoms Rating Scale. These results are from the merged data for each drug. *p-value shows difference between baseline and endpoint in risperidone group (paired t-test). **p-value shows difference between baseline and endpoint in haloperidol group (paired t-test). ***p-value shows difference between risperidone group and haloperidol group as assessed by multivariate analyses, using a mixed model or Generalized Estimating Equations using the independent variables Drug (risperidone or haloperidol), Weeks of treatment (0, 1, 2, 3, 4, 6, or 8 weeks), and individual indicator (to adjust individual variability) as covariates.
significant difference was observed on the CGI-C between risperidone- and haloperidol-treated patients in favor of risperidone. Although the degree of change observed on the CGI assessment scale was small in both treatment groups, the difference between both groups was significant and indicates that patients receiving risperidone improved more than patients receiving haloperidol. The lack of significant change in clinical
FIGURE 3.
Mean Scores on the Parkinsonism Subscale of the Extrapyramidal Symptom Rating Scale for Dementia Patients Receiving Risperidone and Haloperidol
20
Mean Score
15
10
5 Risperidone Haloperidol 0 0
1
2
3
4
5
6
7
Time (weeks)
Am J Geriatr Psychiatry 12:5, September-October 2004
8
impression is difficult to interpret in light of significant improvements in behavioral symptoms (as assessed on the BEHAVE-AD and CMAI). The slight increase in CGI-C score in the haloperidol group would mean that patients tended to worsen during treatment, which may be related to the adverse events observed in this group. This study applied several scales to assess changes in BPSD. It is generally agreed that no single scale will fulfill all the criteria for a good assessment tool; therefore, we used three well-validated tools (BEHAVE-AD, CMAI, and CGI-C)18–20 that had also been applied previously in placebo-controlled clinical trials examining behavioral and psychological symptoms of dementia.7–10 The combination of these scales allows for measuring change in both severity and frequency of psychological and behavioral symptoms, as well as change in patients’ overall clinical status. During the last week of treatment, the mean daily doses of haloperidol (0.83 [0.35] mg) and risperidone (0.80 [0.32] mg) were similar. Given the efficacy of risperidone observed in this study, the optimal effective doses in elderly Korean patients may be lower than the average dose of 1 mg/day that was reported in previous controlled studies.7–10 The crossover design is a sensitive method to determine the efficacy of new drugs because it eliminates between-patient variability. The crossover design requires a much smaller number of patients for a similar statistical power because patients act as their own controls, which is a particularly important advantage when the type or severity of dementia varies widely in the patients recruited.21–23 However, if the response in the first period carries into the second
515
Comparison of Risperidone and Haloperidol period (carryover effect) or if time factors cannot be kept constant in a lengthy crossover (time effects), the statistical power may be jeopardized. In this study, a 1-week washout period between Phase I and Phase II was strictly adhered to. As a result, prevalence of BPSD at Phase I and Phase II baseline evaluations showed small dissimilarities, but there was no statistically significant difference between the baseline prevalence of BPSD before and after the crossover phase. In summary, risperidone has a better efficacy profile than haloperidol in the treatment of BPSD in a population of elderly Korean patients with AD, vas-
cular dementia, or mixed dementia. Both drugs can be implemented with relative safety at a dose of less than 1 mg/day. Risperidone was associated with fewer EPS than haloperidol and was therefore better tolerated. In a patient population that is particularly vulnerable to the side effects of antipsychotic drugs, risperidone would be preferred over haloperidol. The authors thank Dr. Ajit Shah, from Imperial College School of Medicine, London, UK, for sharing his expertise on BPSD and giving us valuable advice. This investigator-initiated study (RIS-KOR-27) was financially supported by Janssen Korea, Seoul, Korea.
References 1. Suh GH, Shah A: A review of epidemiological transition in dementia: cross-national comparisons of the indices related to Alzheimer’s disease and vascular dementia. Acta Psychiatr Scand 2001; 104:4–11 2. Finkel SI, Costa e Silva J, Cohen G, et al: Behavioral and psychological signs and symptoms of dementia: a consensus statement on current knowledge and implications for research and treatment. Int Psychogeriatr 1996; 9:65–68 3. Donaldson C, Tarrier N, Burns A: The impact of the symptoms of dementia on caregivers. Br J Psychiatry 1997; 170:62–68 4. American Psychiatric Association: Practice Guidelines for the Treatment of Patients With Alzheimer’s Disease and Other Dementias of Late Life. Am J Psychiatry 1997; 154:1–39 5. Raskin DE: Antipsychotic medication and the elderly. J Clin Psychiatry 1985; 46:36–40 6. Schneider LS, Pollock VE, Lyness SA: A meta-analysis of controlled trials of neuroleptic treatment in dementia. J Am Geriatr Soc 1990; 38:553–563 7. Katz I, Jeste D, Mintzer J, et al: Comparison of risperidone and placebo for psychosis and behavioral disturbances associated with dementia: a randomized, double-blind trial. J Clin Psychiatry 1999; 60:107–115 8. De Deyn PP, Rabheru K, Rasmussen A, et al: A randomized trial of risperidone, placebo, and haloperidol for behavioral symptoms of dementia. Neurology 1999; 53:946–955 9. Chan W, Lam LC, Choy CN, et al: A double-blind, randomised comparison of risperidone and haloperidol in the treatment of behavioral and psychological symptoms in Chinese dementia patients. Int J Geriatr Psychiatry 2001; 16:1156–1162 10. Brodaty H, Ames D, Snowdon J, et al: A randomized, placebocontrolled trial of risperidone for the treatment of aggression, agitation, and psychosis of dementia. J Clin Psychiatry 2003; 64:134–143 11. American Psychiatric Association: Diagnostic and Statistical Man-
516
ual of Mental Disorders, 4th Edition. Washington, DC, American Psychiatric Association, 1997 12. Reisberg B: Functional assessment staging (FAST). Psychopharmacol Bull 1988; 24:653–659 13. Reisberg B, Borensteen J, Sabbs S, et al: Behavioral symptoms in Alzheimer’s disease: phenomenology and treatment. J Clin Psychiatry 1987; 48(suppl):9–15 14. Cohen-Mansfield J, Marx MS, Rosenthal AS: A description of agitation in a nursing home. J Gerontol 1989; 44:M77–M84 15. Guy W (ed): Clinical Global Impression (CGI-C), in ECDEU Assessment Manual for Psychopharmacology, Revised. Rockville, MD, National Institute of Mental Health, U.S. Dept. of Health, Education, and Welfare, 1976, Publication ADM 76-338 16. Chouinard G, Ross-Chouinard A, Annable L, et al: The Extrapyramidal Rating Scale. Can J Neurol Sci 1980; 7:233 17. Kwon YC, Park J-H: Korean version of Mini-Mental State Exam (MMSE-K), part I: development of the test for the elderly. J Korean Neuropsychiatr Assoc 1989; 28:125–135 18. Oremus M, Perrault A, Demers L, et al: Review of outcome measurement instruments in Alzheimer’s disease drug trials: psychometric properties of global scales. J Geriatr Psychiatry Neurol 2000; 13:197–205 19. Finkel SI, Lyons JS, Anderson RL: Reliability and validity of the Cohen-Mansfield Agitation Inventory in institutionalized elderly. Int J Geriatr Psychiatry 1992; 7:487–490 20. Ferris SH, Mackell JA: Behavioral outcomes in clinical trials for Alzheimer disease. Alzheimer Dis Assoc Disord 1997; 11(suppl4):S10–S15 21. Cleophas TJ, Zwinderman AH: Crossover studies with continuous variables: power analysis. Am J Ther 2002; 9:69–73 22. Feingold A, Oliveto A, Schottenfeld R, et al: Utility of crossover designs in clinical trials: efficacy of desipramine vs. placebo in opioid-dependent cocaine abusers. Am J Addict 2002; 11:111– 123 23. Richens A: Proof-of-efficacy trials: crossover versus parallelgroup. Epilepsy Res 2001; 45:43–47
Am J Geriatr Psychiatry 12:5, September-October 2004
Objective: Behavioral disturbances in dementia are extremely prominent and distressful, and often result in serious physical, social, and economic consequences. The authors compared the efficacy and tolerability of risperidone and haloperidol in the treatment of behavioral and psychological symptoms of dementia (BPSD) in institutionalized elderly Korean patients with Alzheimer disease, vascular dementia, or mixed dementia. Methods: This was an 18-week double-blind, crossover study involving 120 patients who were randomly assigned to receive flexible doses (0.5–1.5 mg/day) of risperidone or haloperidol. BPSD were assessed using the Korean version of the Behavioral Pathology in Alzheimer’s Disease Rating Scale (BEHAVE-AD-K), the Korean version of the Cohen-Mansfield Agitation Inventory (CMAI-K), and the Clinical Global Impression of Change scale (CGI-C). Safety and tolerability assessments included the Extrapyramidal Symptom Rating Scale and the incidence of adverse events. Results: Both risperidone and haloperidol were efficacious in alleviating BPSD. However, when receiving risperidone, patients showed significantly greater improvement than when receiving haloperidol in the total and subscale scores of the BEHAVE-ADK, the total and subscale scores of the CMAI-K, and the scores on the CGI-C scale. Also, risperidone had an additional benefit on aggressiveness and anxieties/phobias. The risk of antipsychotic-induced parkinsonism throughout this study was significantly lower with risperidone than with haloperidol. Conclusion: Risperidone had a favorable efficacy and tolerability profile compared with haloperidol in the treatment of BPSD in this patient population. (Am J Geriatr Psychiatry 2004; 12:509–516)
Received October 18, 2002; revised March 13, May 22, 2003; accepted June 17, 2003. From the Department of Psychiatry, Hallym University Medical Center, Hangang Sacred Heart Hospital, Seoul, Korea (GHS,HGS), the Department of Occupational & Environmental Medicine, Hallym University Sacred Heart Hospital, Seoul, Korea (YSJ), the Department of Psychiatry, Kangnam Hospital Public Corporation, Seoul, Korea (KHJ,YMS), the Department of Psychiatry, Hallym University Medical Center, Kangdong Sacred Heart Hospital, Seoul, Korea (BKY), the Department of Psychiatry, Pil Dong Hospital, Chung Ang University, College of Medicine, Seoul, Korea (BSK), and the Medical Department, Janssen Korea, Seoul, Korea (SKC). Send correspondence to Associate Professor Guk-Hee Suh, Department of Psychiatry, Hallym University Medical Center, Hangang Sacred Heart Hospital, 94-200 Yungdungpo-Dong, Seoul 150-030, Korea. e-mail: [email protected] 䉷 2004 American Association for Geriatric Psychiatry
Am J Geriatr Psychiatry 12:5, September-October 2004
509
Comparison of Risperidone and Haloperidol
D
ementia is a chronic illness with serious physical, social, and economic consequences.1 The prevalence of dementia among Koreans age 65 years and older is 7%–10%, and this is increasing because of an upsurge in the number and proportion of people in the population who are age 75 years or older.1 Behavioral and psychological symptoms of dementia (BPSD) are very prominent clinically and are often the source of significant distress to caregivers, leading them to seek professional help. This can result in premature institutionalization, with increased costs of care and significant reduction of quality of life, not only for the patients,2 but also for their families and caregivers.3 Despite the serious consequences of these symptoms, there is a lack of clinical trials for the treatment of BPSD. Antipsychotics are generally acknowledged to be more effective than a variety of other pharmacological agents in the treatment of patients with BPSD.4 Data support a modest efficacy for conventional antipsychotics,5,6 but the use of these drugs in dementia patients has been restricted by their potential for serious side effects. Theoretically, novel antipsychotics have a lower potential to cause extrapyramidal symptoms (EPS) than conventional neuroleptics. The results of short-term studies indicate that atypical antipsychotics, especially risperidone, are generally effective and well tolerated in the treatment of patients with BPSD,7–10 but there is a lack of efficacy data in non-Caucasian populations. This study was designed to compare the efficacy and tolerability of risperidone and haloperidol in the treatment of BPSD in Korean patients with dementia.
METHODS Patient Population All 280 dementia patients residing at the Seoul City Jung-Kye Welfare Center for the Elderly (a semi-hospitalized long-term care institution in Seoul) were screened for eligibility in this study, which was conducted from June to December 2001. Patients were under the full-time medical surveillance of psychiatrists, medical doctors, and nurses and were looked after by professional caregivers. The professional caregivers assisted the dementia patients with drug
510
intake and the study protocol and provided information required at assessment interviews. Eligibility criteria included an age of 65 years or over, a diagnosis of dementia of the Alzheimer type with behavioral disturbance, vascular dementia with behavioral disturbance, or a combination of the two, according to the criteria of the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition (DSMIV).11 Eligible patients had a score of 4 or higher on the Functional Assessment Staging Test (FAST),12 a total score of 8 or higher on the Korean version of the Behavioral Pathology in Alzheimer’s Disease Rating Scale13 (BEHAVE-AD-K; Suh et al., data on file), and a score of more than 3 on any two items of the Korean version of Cohen-Mansfield Agitation Inventory (CMAI-K)14 (Suh et al., data on file). Exclusion criteria included other conditions that diminish cognitive function (e.g., Lewy-body dementia, hypothyroidism), other psychiatric disorders that might contribute to the psychotic symptoms (e.g., schizophrenia, delusional disorder), clinically relevant organic or neurologic disease, unstable medical conditions (e.g., poorly controlled hypertension, angina, or diabetes), abnormal electrocardiograms as diagnosed by a cardiologist or laboratory tests, a history of allergic reaction to antipsychotic treatment, and a history of neuroleptic malignant syndrome. The study was conducted in accordance with Good Clinical Practices and the Declaration of Helsinki. The study protocol was reviewed and approved by the independent Ethics Committee of Hallym University. Each patient or his or her legal guardian provided written informed consent for participation after the procedure had been fully explained. Design and Procedures This was a randomized, double-blind, 18-week crossover study that included four periods: a 1-week washout period during which all psychotropic medications were discontinued; an 8-week double-blind active treatment period (Phase I), a second 1-week washout period, and a final 8-week crossover activetreatment period (Phase II). All consenting patients were randomly assigned to either the risperidonefirst group or the haloperidol-first group according to a predefined randomization code. At each visit after the Phase I baseline evaluation (Week 0), eligible patients were evaluated with the
Am J Geriatr Psychiatry 12:5, September-October 2004
Suh et al. BEHAVE-AD-K, CMAI-K, and Clinical Global Impression of change (CGI-C) scale.15 The 25-item BEHAVE-AD scale is specifically designed to assess the severity of BPSD and provides seven subscales: paranoid and delusional ideation, hallucinations, activity disturbances, aggressiveness, diurnal rhythm disturbances, affective disorders, and anxieties and phobias.13 The sum of these subscales provides the total BEHAVE-AD score. The Psychosis subscale is the sum of the Paranoid and Delusional Ideation and Hallucination subscales. The 29-item CMAI scale is specifically designed to assess the frequency of manifestations of agitation-related behaviors in elderly persons with dementia and provides a Physical (e.g., hitting) and Verbal (e.g., screaming) Aggression scale, which can be combined to give a Total Aggression score, as well as a Physical (e.g., wandering) and Verbal (e.g., grunting) Non-Aggression scale.14 Also, the CGI-C scale is designed to assess global symptoms and patient’s overall condition and reflects the relevance of changes in behavioral ratings. The CGI-C is a 7-point scale ranging from 1: much better, through 4: unchanged, to 7: much worse.15 At baseline, all patients were scored as 4 (unchanged). All reported adverse events were recorded, and the severity of EPS such as parkinsonism, akathisia, dystonia, and dyskinesia was assessed by use of the Extrapyramidal Symptom Rating Scale (ESRS).16 After randomization and the initiation of double-blind treatment, patients were assessed weekly during the first 4 weeks of the study and then every 2 weeks (twice) until the end of the final (8th) week. Vital signs, such as blood pressure, pulse rate, and respiration rate were checked routinely and monitored every other day by day-duty nurses. Standard 12lead electrocardiograms and laboratory tests were done at the initial screening period and interim washout period. Study Drug The study drugs, haloperidol (Myung In. Pharm. Co. Ltd., Seoul, Korea) and risperidone (Janssen Korea, Seoul, Korea), were blinded in non-transparent capsules (0.5 mg/tablet) and packaged individually with the patient-identification label. The randomization code was provided by a statistician who was not involved in the clinical study. The study drug was dispensed twice: before the start of Phase I and before
Am J Geriatr Psychiatry 12:5, September-October 2004
the start of Phase II. Before the new drug was distributed, the remaining drug from Phase I was collected. Two psychiatrists, who were blinded to the study drug and not involved in the rating, prescribed 0.5 mg h.s. of either drug at Week 0. If needed, the drug dosage could be adjusted by increments of 0.5 mg, once every 4 or more days when one of two psychiatrists visited the institution to treat the patients. The target dose for this study was 1 mg/day, but the dose could be stepped up to 1.5 mg/day if symptoms were poorly controlled. The dosage could also be titrated downward at any time and by any amount in the event of clinically significant treatment-emergent adverse events or if, in the prescriber’s judgment, a decrease was warranted. The dosing regimen and intervals between dose titrations were individualized for each patient. Concomitant use of antipsychotic drugs, antidepressants, and mood stabilizers was not permitted. Lorazepam was permitted if limited to 4 days/week for the first 4 weeks of active treatment. Statistical Analysis A sample size of 98 patients per drug was selected in order to detect a 4-point difference in the total score on the BEHAVE-AD-K with 80% power and a twotailed level of significance of 0.05. The sample size was then increased to 120 patients to account for a 20% rate of dropout. In order to increase the power of the study and reduce the sample size, a two-phase crossover design was applied. Primary analyses were planned for all patients who underwent randomization and assessment at least once during drug treatment, but because of the low number of patients who discontinued the study early, the final analyses consisted of data from patients who completed drug treatment per protocol. The merged dataset for each drug (which included the data for all patients who received the drug during Phase I and all those who received it during Phase II) was used for analysis. Baseline measures were taken at Week 0 for those who received the drug during Phase I, and Week 9 for those who received the drug during Phase II; endpoint measures were taken at Week 8 for those who received the drug during Phase I, and Week 18 for those who received the drug during Phase II. We applied univariate analyses to calculate basic statistics and to test the normality (either continuous or
511
Comparison of Risperidone and Haloperidol binary distributed) of all variables. We used longitudinal data-analysis methods to control autocorrelation and to conduct multivariate analyses, as the study design was ultimately longitudinal. In the cases of continuous dependent variables (BEHAVE-AD-K [Delusion, Psychosis {Delusion Ⳮ Hallucination}, Activity Disturbance, Aggressive Behavior, Anxieties and Phobias, Total Score]; CMAI-K [Aggressive Behavior, Physically Non-Aggressive Behavior, Verbally Agitated Behavior, Total score]; ESRS [parkinsonism, total score]; and CGI-C), we applied the mixed model. In the cases of binary dependent variables (BEHAVEAD-K [Hallucination, Diurnal Rhythm Disturbances, Affective Disturbances] and ESRS [dystonia, dyskinetic movement]), we applied Generalized Estimating Equations (GEEs). Multivariate analyses, using the mixed model and GEEs where appropriate, were conducted using the following independent variables: drug (risperidone or haloperidol), weeks of treatment (0, 1, 2, 3, 4, 6, or 8 weeks) and individual indicator (to adjust individual variability) as covariates. Within-group comparisons (baseline to endpoint) were analyzed with paired t-tests. Statistical significance was assumed at p ⱕ0.05. The statistical package SAS (Version 6.12) was used for all analyses.
The two groups were similar with respect to baseline demographics and disease characteristics. Furthermore, there were no statistically significant differences at the two baselines, before crossover (Week 0), and after crossover (Week 9) for any variables on the BEHAVE-AD-K, CMAI-K, CGI-C scale, and ESRS between the patients who received risperidone during Phase I and those who received it during Phase II, and between patients who received haloperidol during Phase I and in Phase II. Efficacy The mean baseline and endpoint scores on the BEHAVE-AD-K, CMAI-K, and CGI-C scale are listed in Table 2. At the endpoint of treatment, when compared with baseline by paired t-tests, risperidone resulted in significant improvement in scores on the CMAI-K Total scale (t[112]⳱3.03; p⳱0.003) and the subscales of Aggressive Behavior (t[113]⳱8.44; p ⬍0.0001) and Physical Non-Aggressive behavior (t[113]⳱7.34; p ⬍0.0001). Significant improvements were also observed for the BEHAVE-AD-K Total score (t[113]⳱12.7; p ⬍0.0001) and all of its subscales FIGURE 1.
Schematic Presentation of Flow in this Crossover Comparative Clinical Study of Risperidone and Haloperidol
RESULTS Of the 120 randomized patients, 114 (95%) completed the study and were included in the final data analysis. The reasons for patient discontinuation are shown in the study flow diagram (Figure 1). The mean (standard deviation [SD]) age of patients was 80.9 (8.2) years, and 82% were female. Alzheimer disease (AD) was the most common diagnosis, occurring in 65.8% of the patients, followed by vascular dementia (28.3%) and mixed dementia (5.8%). The mean score on the Korean version of the Mini Mental State Exam (MMSE-K)17 at the baseline evaluation was 9.6 (6.6), which was not statistically different from the mean score on the MMSE-K at the end of study. Of all the patients, 66% were at FAST Stage 6A or higher. The mean duration of AD was 5.4 years, that of vascular dementia was 5.7 years, and that of mixed dementia was 4.3 years (Table 1). During the last week of treatment, the mean daily dose of haloperidol was 0.83 (0.35) mg and that of risperidone was 0.80 (0.32) mg.
512
Patients screened (N=280)
Patients enrolled and randomly assigned (N=120) Phase I Risperidone (N=60)
Haloperidol (N=60)
Discontinued: Seizure (N=1) Nausea (N=1)
Discontinued: Somnolence (N=1)
Haloperidol (N=58)
Risperidone (N=59)
Discontinued: Somnolence (N=2)
Discontinued: Nausea (N=1)
Completed trial (N=56)
Completed trial (N=58)
Phase II
Am J Geriatr Psychiatry 12:5, September-October 2004
Suh et al. (t ⱖ3.43; p ⬍0.001). Patients significantly improved as observed on the CGI-C scale (t[113]⳱3.61; p⳱0.0008). Haloperidol, when compared with baseline, resulted in significant improvement in the BEHAVE-AD-K Total score (t[113]⳱5.10; p ⬍0.0001) and its subscales of Psychosis (t[113]⳱2.96; p⳱0.004), Activity Disturbances (t[112]⳱5.42; p ⬍0.0001), Aggressiveness (t[113]⳱5.58; p ⬍0.0001), and Affective Disturbances (t[113]⳱2.99; p⳱0.003), as well as the TABLE 1.
CMAI-K Total (t[112]⳱14.2; p ⬍0.0001) and subscale of Aggressive Behavior (t[113]⳱7.41; p ⬍0.0001). Risperidone (N⳱114) was superior to haloperidol (N⳱114) on a variety of efficacy variables, as assessed by the mixed model for the BEHAVE-AD-K Total score (numerator degrees of freedom [NDF]⳱1; denominator degrees of freedom (DDF)⳱1,398; Type III F⳱8.58; p⳱0.004) and for the BEHAVE-AD-K subscales: Aggressiveness (NDF: 1, DDF: 1,397; Type III
Demographics and Baseline Assessment Scores
Characteristic
RIS-to-HAL Group Nⴔ60
HAL-to-RIS Group Nⴔ60
Both Nⴔ120
13/47 80.4 (5.6) 65–96
11/49 81.5 (7.7) 65–97
24/96 80.9 (8.2) 65–97
42 (70%) 15 (25%) 3 (5%)
37 (61.7%) 19 (31.7%) 4 (6.6%)
79 (65.8%) 34 (28.3%) 7 (5.8%)
19.3 (7.7) 86.3 (20.2) 10.2 (20.3) 9.7 (7.5)
19.2 (8.8) 89.4 (21.4) 8.8 (16.0) 9.4 (5.2)
19.2 (8.2) 87.8 (20.7) 9.6 (18.1) 9.6 (6.6)
19 41
22 38
41 79
Sex, M/F Age, years, mean (SD) range Diagnosis (DSM-IV), N (%) Alzheimer disease Vascular dementia Mixed dementiaa Mean baseline scores, total (SD) BEHAVE-AD-K CMAI-K ESRS MMSE-K FAST stage 4–5 6a–7e
Note: RIS-to-HAL group: risperidone during Phase I, haloperidol during Phase II; HAL-to-RIS group: haloperidol during Phase I, risperidone during Phase II; SD: standard deviation; DSM-IV: Diagnostic and Statistical Manual of Mental Disorders, 4th Edition; BEHAVE-AD-K: Korean version of the Behavioral Pathology in Alzheimer’s Disease Rating Scale; CMAI-K: Korean version of the Cohen-Mansfield Agitation Inventory; ESRS: Extrapyramidal Symptom Rating Scale; MMSE-K: Korean version of the Mini-Mental State Exam; FAST: Functional Assessment Staging Test. a These diagnoses were based on the results of neuroimaging.
TABLE 2.
Differences in Efficacy Between Risperidone and Haloperidol in the Treatment of Patients With BPSD According to the Changes in the Scores on the BEHAVE-AD-K, CMAI-K, and CGI-C Risperidone Nⴔ114
Haloperidol Nⴔ114
Differencea
Assessment
Baseline
Endpoint
Baseline
Endpoint
p
BEHAVE-AD-K Total score Psychosis Activity Disturbances Aggressiveness Diurnal Rhythm Disturbances Affective Disturbance Anxieties and Phobias CMAI-K Total score Aggressive Behavior Physical Non-Aggressive Behavior Verbally Agitated Behavior CGI-C
19.2 (0.7) 8.3 (0.4) 3.8 (0.2) 3.1 (0.2) 0.9 (0.1) 0.7 (0.1) 2.5 (0.2) 89.3 (1.9) 22.2 (0.6) 17.2 (0.6) 15.1 (0.5) 4
12.0 (0.5) 4.6 (0.3) 2.7 (0.2) 2.0 (0.1) 0.4 (0.1) 0.2 (0.1) 2.2 (0.1) 75.1 (1.7) 18.2 (0.5) 14.8 (0.5) 14.0 (0.4) 3.9 (0.1)
19.2 (0.8) 8.0 (0.4) 3.8 (0.2) 3.2 (0.2) 0.7 (0.1) 0.5 (0.1) 2.3 (0.2) 89.4 (2.0) 23.6 (0.7) 17.8 (0.5) 15.7 (0.5) 4
14.5 (0.6) 6.0 (0.3) 3.0 (0.2) 2.3 (0.1) 0.5 (0.1) 0.3 (0.1) 2.4 (0.1) 83.5 (1.8) 20.3 (0.5) 16.8 (0.5) 15.2 (0.4) 4.2 (0.1)
0.004 0.582 0.858 0.002 0.038 0.248 ⬍0.0001 ⬍0.0001 0.001 0.024 0.002 0.001
Note: Values are mean (standard error). These results are from the merged data for each drug. BPSD: Behavioral and Psychological Symptoms of Dementia; BEHAVE-AD-K: Korean version of the Behavioral Pathology in Alzheimer’s Disease Rating Scale; CMAI-K: Korean version of the Cohen-Mansfield Agitation Inventory; CGI-C: Clinical Global Impression of Change. a p-value showing difference between risperidone group and haloperidol group as assessed by multivariate analyses, using a mixed model or Generalized Estimating Equations, using the independent variables: drug (risperidone or haloperidol), weeks of treatment (0, 1, 2, 3, 4, 6, or 8 weeks) and individual indicator (to adjust individual variability) as covariates.
Am J Geriatr Psychiatry 12:5, September-October 2004
513
Comparison of Risperidone and Haloperidol F⳱10.2; p⳱0.002) and Anxieties and Phobias (NDF: 1; DDF: 1,398; Type III F⳱14.75; p⳱0.0001). Risperidone was also superior to haloperidol as assessed by the mixed model for the CMAI-K Total score (NDF: 1; DDF: 1,398; Type III F⳱16.26; p⳱0.0001) and for the CMAI-K subscales Aggressive Behavior (NDF: 1; DDF: 1,395; Type III F⳱11.0; p⳱0.001), Physically Non-Aggressive Behavior (NDF: 1; DDF: 1,397; Type III F⳱3.96; p⳱0.024) and Verbally Agitated Behavior (NDF: 1; DDF: 1,397; Type III F⳱8.09; p⳱0.002). Statistically significant differences favoring risperidone were also observed for CGI-C score (mixed model, NDF: 1; DDF: 1,397; Type III F⳱39.05; p⳱0.0001) and Diurnal Rhythm Disturbance (GEE: N⳱114; z ⳱ ⳮ2.075; p⳱0.038; Table 2, Figure 2). Drug Safety and Tolerability There were no statistically significant differences between the risperidone-first group and the haloperidol-first group in baseline scores on the ESRS or its three subscales of parkinsonism: expressive automatic movements (bradykinesia, rigidity, gait and posture, tremor, akathisia, sialorrhea, postural instability), dystonia, or dyskinetic movements. When compared with baseline, risperidone was not associFIGURE 2.
Mean Total Scores on the Korean version of the Behavioral Pathology in Alzheimer’s Disease Rating Scale (BEHAVE-AD-K) for Dementia Patients Receiving Risperidone and Haloperidol
20
Risperidone
Mean Total Score on BEHAVE-AD-K
Haloperidol 18
16
14
12
10 0
1
2
3
4
5
Time (weeks)
514
ated with a significant increase in EPS over the study period, as determined by the total and three subscale scores on the ESRS. In contrast, haloperidol was associated with a significant worsening of EPS over the course of the study, as revealed by a significant increase in the total and parkinsonism subscale scores of the ESRS when compared with baseline by the mixed model (NDF: 1; DDF: 698; Type III F⳱4.846; p⳱0.0362; and NDF: 1; DDF: 696; Type III F⳱225.17; p⳱0.0001, respectively). When the differences between baseline and endpoint of risperidone and haloperidol were compared directly by the mixed model, both the total ESRS and the parkinsonism subscale were significantly increased with haloperidol treatment (NDF: 1; DDF: 1,193; Type III F⳱159; p⳱0.0001; NDF: 1; DDF: 1,194; Type III F⳱239; p⳱0.0001; Table 3, Figure 3). Furthermore, compared with risperidone by means of GEEs, somnolence (N⳱114; z ⳱ –6.63; p⳱0.0001), insomnia (N⳱114; z ⳱ –3.77; p⳱0.0001), and sialorrhea (N⳱114; z ⳱ –4.04; p⳱0.0001) were reported significantly more often when patients received haloperidol. Most of the patients tolerated the study drugs. There was no significant change in vital signs over the course of the study. A total of six patients discontinued the study early because of adverse events: seizure (N⳱1) and nausea (N⳱2) in the patients treated with risperidone, and somnolence (N⳱3) in the patients treated with haloperidol. The seizure was not considered drug-related.
6
7
8
DISCUSSION Both risperidone and haloperidol were efficacious in alleviating BPSD. With both study drugs, significant reductions were noted in the total scores on the BEHAVE-AD-K and its subscale scores of psychosis, activity, and affective disturbances, as well as in the total scores on the CMAI-K and its aggressive behavior subscale scores. However, when receiving risperidone, patients showed significantly greater improvement than when receiving haloperidol on the CMAIK Total and all subscale scores. Risperidone had an additional benefit on BEHAVE-AD-K Total, and Aggressiveness, Diurnal Rhythm Disturbances, and Anxieties and Phobias subscales that was not observed with haloperidol. Furthermore, a statistically
Am J Geriatr Psychiatry 12:5, September-October 2004
Suh et al. TABLE 3.
Differences in Tolerability Between Risperidone and Haloperidol in the Treatment of Patients With BPSD According to the Changes in Scores on the ESRS Risperidone Nⴔ114
Haloperidol Nⴔ114
Difference
Assessment
Baseline
Endpoint
p*
Baseline
Endpoint
p**
p***
ESRS Total score Parkinsonism Dystonia Dyskinetic movement
8.7 (0.9) 6.1 (1.3) 0.8 (0.5) 1.6 (0.4)
13.5 (1.6) 9.6 (1.1) 1.8 (0.5) 2.1 (0.4)
0.6606 0.1564 0.7498 0.2338
8.8 (1.5) 6.4 (1.0) 0.7 (0.3) 1.6 (0.4)
22.6 (2.1) 16.8 (1.4) 3.2 (0.8) 2.5 (0.5)
0.0362 ⬍0.0001 0.0007 0.1627
0.0001 0.0001 0.6503 0.4144
Note: Values are mean (standard error). BPSD: Behavioral and Psychological Symptoms of Dementia; ESRS: Extrapyramidal Symptoms Rating Scale. These results are from the merged data for each drug. *p-value shows difference between baseline and endpoint in risperidone group (paired t-test). **p-value shows difference between baseline and endpoint in haloperidol group (paired t-test). ***p-value shows difference between risperidone group and haloperidol group as assessed by multivariate analyses, using a mixed model or Generalized Estimating Equations using the independent variables Drug (risperidone or haloperidol), Weeks of treatment (0, 1, 2, 3, 4, 6, or 8 weeks), and individual indicator (to adjust individual variability) as covariates.
significant difference was observed on the CGI-C between risperidone- and haloperidol-treated patients in favor of risperidone. Although the degree of change observed on the CGI assessment scale was small in both treatment groups, the difference between both groups was significant and indicates that patients receiving risperidone improved more than patients receiving haloperidol. The lack of significant change in clinical
FIGURE 3.
Mean Scores on the Parkinsonism Subscale of the Extrapyramidal Symptom Rating Scale for Dementia Patients Receiving Risperidone and Haloperidol
20
Mean Score
15
10
5 Risperidone Haloperidol 0 0
1
2
3
4
5
6
7
Time (weeks)
Am J Geriatr Psychiatry 12:5, September-October 2004
8
impression is difficult to interpret in light of significant improvements in behavioral symptoms (as assessed on the BEHAVE-AD and CMAI). The slight increase in CGI-C score in the haloperidol group would mean that patients tended to worsen during treatment, which may be related to the adverse events observed in this group. This study applied several scales to assess changes in BPSD. It is generally agreed that no single scale will fulfill all the criteria for a good assessment tool; therefore, we used three well-validated tools (BEHAVE-AD, CMAI, and CGI-C)18–20 that had also been applied previously in placebo-controlled clinical trials examining behavioral and psychological symptoms of dementia.7–10 The combination of these scales allows for measuring change in both severity and frequency of psychological and behavioral symptoms, as well as change in patients’ overall clinical status. During the last week of treatment, the mean daily doses of haloperidol (0.83 [0.35] mg) and risperidone (0.80 [0.32] mg) were similar. Given the efficacy of risperidone observed in this study, the optimal effective doses in elderly Korean patients may be lower than the average dose of 1 mg/day that was reported in previous controlled studies.7–10 The crossover design is a sensitive method to determine the efficacy of new drugs because it eliminates between-patient variability. The crossover design requires a much smaller number of patients for a similar statistical power because patients act as their own controls, which is a particularly important advantage when the type or severity of dementia varies widely in the patients recruited.21–23 However, if the response in the first period carries into the second
515
Comparison of Risperidone and Haloperidol period (carryover effect) or if time factors cannot be kept constant in a lengthy crossover (time effects), the statistical power may be jeopardized. In this study, a 1-week washout period between Phase I and Phase II was strictly adhered to. As a result, prevalence of BPSD at Phase I and Phase II baseline evaluations showed small dissimilarities, but there was no statistically significant difference between the baseline prevalence of BPSD before and after the crossover phase. In summary, risperidone has a better efficacy profile than haloperidol in the treatment of BPSD in a population of elderly Korean patients with AD, vas-
cular dementia, or mixed dementia. Both drugs can be implemented with relative safety at a dose of less than 1 mg/day. Risperidone was associated with fewer EPS than haloperidol and was therefore better tolerated. In a patient population that is particularly vulnerable to the side effects of antipsychotic drugs, risperidone would be preferred over haloperidol. The authors thank Dr. Ajit Shah, from Imperial College School of Medicine, London, UK, for sharing his expertise on BPSD and giving us valuable advice. This investigator-initiated study (RIS-KOR-27) was financially supported by Janssen Korea, Seoul, Korea.
References 1. Suh GH, Shah A: A review of epidemiological transition in dementia: cross-national comparisons of the indices related to Alzheimer’s disease and vascular dementia. Acta Psychiatr Scand 2001; 104:4–11 2. Finkel SI, Costa e Silva J, Cohen G, et al: Behavioral and psychological signs and symptoms of dementia: a consensus statement on current knowledge and implications for research and treatment. Int Psychogeriatr 1996; 9:65–68 3. Donaldson C, Tarrier N, Burns A: The impact of the symptoms of dementia on caregivers. Br J Psychiatry 1997; 170:62–68 4. American Psychiatric Association: Practice Guidelines for the Treatment of Patients With Alzheimer’s Disease and Other Dementias of Late Life. Am J Psychiatry 1997; 154:1–39 5. Raskin DE: Antipsychotic medication and the elderly. J Clin Psychiatry 1985; 46:36–40 6. Schneider LS, Pollock VE, Lyness SA: A meta-analysis of controlled trials of neuroleptic treatment in dementia. J Am Geriatr Soc 1990; 38:553–563 7. Katz I, Jeste D, Mintzer J, et al: Comparison of risperidone and placebo for psychosis and behavioral disturbances associated with dementia: a randomized, double-blind trial. J Clin Psychiatry 1999; 60:107–115 8. De Deyn PP, Rabheru K, Rasmussen A, et al: A randomized trial of risperidone, placebo, and haloperidol for behavioral symptoms of dementia. Neurology 1999; 53:946–955 9. Chan W, Lam LC, Choy CN, et al: A double-blind, randomised comparison of risperidone and haloperidol in the treatment of behavioral and psychological symptoms in Chinese dementia patients. Int J Geriatr Psychiatry 2001; 16:1156–1162 10. Brodaty H, Ames D, Snowdon J, et al: A randomized, placebocontrolled trial of risperidone for the treatment of aggression, agitation, and psychosis of dementia. J Clin Psychiatry 2003; 64:134–143 11. American Psychiatric Association: Diagnostic and Statistical Man-
516
ual of Mental Disorders, 4th Edition. Washington, DC, American Psychiatric Association, 1997 12. Reisberg B: Functional assessment staging (FAST). Psychopharmacol Bull 1988; 24:653–659 13. Reisberg B, Borensteen J, Sabbs S, et al: Behavioral symptoms in Alzheimer’s disease: phenomenology and treatment. J Clin Psychiatry 1987; 48(suppl):9–15 14. Cohen-Mansfield J, Marx MS, Rosenthal AS: A description of agitation in a nursing home. J Gerontol 1989; 44:M77–M84 15. Guy W (ed): Clinical Global Impression (CGI-C), in ECDEU Assessment Manual for Psychopharmacology, Revised. Rockville, MD, National Institute of Mental Health, U.S. Dept. of Health, Education, and Welfare, 1976, Publication ADM 76-338 16. Chouinard G, Ross-Chouinard A, Annable L, et al: The Extrapyramidal Rating Scale. Can J Neurol Sci 1980; 7:233 17. Kwon YC, Park J-H: Korean version of Mini-Mental State Exam (MMSE-K), part I: development of the test for the elderly. J Korean Neuropsychiatr Assoc 1989; 28:125–135 18. Oremus M, Perrault A, Demers L, et al: Review of outcome measurement instruments in Alzheimer’s disease drug trials: psychometric properties of global scales. J Geriatr Psychiatry Neurol 2000; 13:197–205 19. Finkel SI, Lyons JS, Anderson RL: Reliability and validity of the Cohen-Mansfield Agitation Inventory in institutionalized elderly. Int J Geriatr Psychiatry 1992; 7:487–490 20. Ferris SH, Mackell JA: Behavioral outcomes in clinical trials for Alzheimer disease. Alzheimer Dis Assoc Disord 1997; 11(suppl4):S10–S15 21. Cleophas TJ, Zwinderman AH: Crossover studies with continuous variables: power analysis. Am J Ther 2002; 9:69–73 22. Feingold A, Oliveto A, Schottenfeld R, et al: Utility of crossover designs in clinical trials: efficacy of desipramine vs. placebo in opioid-dependent cocaine abusers. Am J Addict 2002; 11:111– 123 23. Richens A: Proof-of-efficacy trials: crossover versus parallelgroup. Epilepsy Res 2001; 45:43–47
Am J Geriatr Psychiatry 12:5, September-October 2004