- Email: [email protected]
Usefulness of Plasma Haloperidol Levels for Monitoring Clinical Efficacy and Side Effects in Alzheimer Patients With Psychosis and Behavioral Dyscontrol
Usefulness of Plasma Haloperidol Levels for Monitoring Clinical Efficacy and Side Effects in Alzheimer Patients With Psychosis and Behavioral Dyscontrol Gregory H. Pelton, M.D., Devangere P. Devanand, M.D. Karen Bell, M.D., Karen Marder, M.D., M.P.H. Kristin Marston, M.A., Xinhua Liu, Ph.D. Thomas B. Cooper, M.A.
Objective: This study investigated whether measurement of plasma levels may be useful in monitoring clinical efficacy and side effects during oral haloperidol (HL) treatment of psychosis and behavioral dyscontrol in patients with Alzheimer disease (AD). Methods: After a single-blind placebo period of 1 week, 71 outpatients with AD were randomized to a 6-week, double-blind, placebo-controlled trial of HL 2 mg–3 mg/ day (standard dose), HL 0.5 mg–0.75 mg/day (low dose), or placebo, with plasma levels for HL drawn at the end of 6 weeks. Results: Of the 40 patients who received active HL for 6 weeks, 35 had plasma levels drawn. Plasma levels were all below the lower limit of the postulated therapeutic range in schizophrenia. Nonetheless, HL plasma level significantly correlated with clinical efficacy as measured by reduction in Brief Psychiatric Rating Scale Total score, Psychosis factor, and Hostile-Suspiciousness factor, the Behavioral Syndromes Scale for Dementia Psychomotor Agitation scale, and with the severity of extrapyramidal side effects (EPS). Oral dose did not significantly correlate with any of these efficacy or side-effect measures. Plasma levels significantly correlated with HL dose. When both HL dose and HL plasma level were included as independent variables in linear-regression analyses, only HL plasma level was a significant predictor of efficacy and EPS. Conclusion: Measurement of HL plasma levels may have potential usefulness as an adjunct in monitoring treatment response to oral HL in AD patients with psychosis or disruptive behavior. (Am J Geriatr Psychiatry 2003; 11:186–193)
Received April 17, 2001; revised June 24, July 30, 2002; accepted July 31, 2002. From the Department of Biological Psychiatry (GHP,DPD,XL,KM), the Department of Analytical Psychopharmacology (TBC), and the Memory Disorders Clinic (GHP,DPD,KB,KM), New York State Psychiatric Institute; The Taub Center for Alzheimer’s Disease Research, New York (KB,KM); the Gertrude H. Sergievsky Center at Columbia University, New York (KB,KM); and the Department of Psychiatry (GHP,DPD,XL,TBC) and the Department of Neurology (KB,KM), College of Physicians and Surgeons of Columbia University, New York. Address correspondence to Dr. Gregory Pelton, New York State Psychiatric Institute, 1051 Riverside Drive, Unit 126, New York, NY 10032. e-mail: [email protected] Copyright 䉷 2003 American Association for Geriatric Psychiatry
186
Am J Geriatr Psychiatry 11:2, March-April 2003
Pelton et al.
I
n patients with dementia who develop psychosis and behavioral dyscontrol, efficacy has been established for haloperidol (HL),1 risperidone,2,3 and olanzapine.4 Part of the challenge in establishing efficacy for these agents includes their relatively narrow therapeutic oral dose range, with low doses providing the optimal tradeoff between efficacy and side effects. In young schizophrenic patients, some investigators have found a linear or curvilinear relationship between plasma levels of antipsychotics (primarily HL), clinical response, and side effects. Most studies show a therapeutic concentration range for serum HL levels of approximately 5 ng/ml to 15 ng/ml at steady state for optimal clinical benefit,5–9 with few dissenting reports.10,11 Plasma-level measurement may be more useful in AD patients for several reasons. First, most AD patients are antipsychotic-naı¨ve when they first develop psychosis or disruptive behavior. Hence, there is less potential for the induction of enzymes in the intestinal wall and/ or liver, both of which can alter HL plasma levels.12 Second, since very low doses of HL (0.5 mg/day to 3 mg/ day) are used in AD patients, a more linear pharmacokinetic relationship between HL oral dose and HL plasma level may exist. Consequently, the curvilinear relationship between HL plasma level and response reported at higher dose levels in young schizophrenic patients, may not apply to AD patients. Rather, the HL plasma levels in AD patients may be limited to the ascending limb (low plasma level) section of the dose– response curve. Third, very low HL plasma levels will not result in the saturation of the D2 receptor system seen with higher dose treatment; that is, a pharmacodynamic difference in these patients is possible. Few studies have examined the relationship between HL plasma level and clinical response in elderly patients with dementia. Lacro et al.13 reported on plasma levels in 32 psychotic elderly patients, age 45– 83 years, diagnosed with schizophrenia (n⳱23) and AD (n⳱9). As a group, age positively correlated with the ratio of HL plasma level to HL oral dose needed for clinical stability (r⳱0.446; p⬍0.05). However, in the AD subgroup, this correlation was lost, which may have been due to small sample size. Dysken et al.,14 on the other hand, reported that in a sample of 29 AD inpatients with behavioral problems (mean age, 76 years; standard deviation [SD]: 8 years), who received a fixed oral dosage of HL (0.5mg, 1.0 mg, or 2.0 mg) every 12 hours for 3 weeks, 55% of patients showed a good re-
Am J Geriatr Psychiatry 11:2, March-April 2003
sponse to low-dose HL, but there was no significant linear or curvilinear relationship between HL plasma levels and clinical response. In an initial study, we reported on 19 elderly patients with AD and psychosis or behavioral disturbances, who received low-dose HL (0.5 mg/day to 5 mg/day) for 6–8 weeks. Oral dose strongly correlated with HL plasma level (r⳱0.82). There were suggestions that, compared with oral dose, HL plasma level correlated more strongly with clinical improvement (Brief Psychiatric Rating Scale [BPRS] scores) and extrapyramidal signs (EPS).15 In a study comparing the efficacy and side effects of two doses of oral HL and placebo in the treatment of psychosis and disruptive behavior in patients with AD,1 HL plasma levels were assayed (10 of the subjects who received HL were included in the Devanand et al. 1992 report15). We hypothesized that HL plasma level would be associated more strongly with treatment response than oral dose and that HL plasma level would also correlate strongly with severity of EPS, suggesting potential clinical usefulness in monitoring HL plasma levels.
METHODS All subjects were outpatients at a memory disorders clinic that is part of an AD research center. Subjects were required to meet DSM-III-R criteria for dementia and the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer’s Disease and Related Disorders Association criteria for probable AD.16 Diagnostic work-up included neurologic and psychiatric evaluation, neuropsychological testing, laboratory tests, and a computerized tomographic (CT) or magnetic resonance imaging (MRI) scan of the brain. Exclusion criteria were medication or alcohol dependence, stroke (clinical evidence of cortical stroke or a lesion ⱖ2 cm in diameter on any CT or MRI slice), and a history or clinical evidence of other causes of dementia, including head trauma, Parkinson disease, Huntington disease, or multiple sclerosis. Patients who met inclusion criteria for either psychosis or behavioral dyscontrol, as previously described,1 were eligible for the study. At the initial visit, global cognitive functioning was assessed by the MiniMental State Exam (Folstein 30-item MMSE);17 severity of dementia was assessed with the Clinical Dementia
187
Plasma Haloperidol Levels Rating Scale (CDR), and the modified Blessed Functional Activities Scale was given to the informant to assess functional impairment. The New York State Psychiatric Institute review board approved the research protocol. We obtained informed consent from the patient and a surrogate advocate, usually a family member. The consent procedures followed specific New York State regulations concerning research involving patients who do or do not have the capacity to consent. Family members served as informants and were required to have had contact with the patient at least once a week on average during the 3 months before the patient entered the study. Figure 1 summarizes the study design. In the initial, 1-week, single-blind phase, all patients received placebo. At the end of this week, the patients who still met entry criteria entered a 6-week random-assignment, double-blind, parallel-group, placebo-controlled comparison of three treatment conditions: standard dose HL (2 mg–3 mg/day), low-dose HL (0.5 mg–0.75 mg/day), and placebo.1 The subsequent crossover phase is not shown, as it is not relevant to this plasma-level study. During the 6-week trial, after the first week, the daily dose was raised from two capsules (HL 0.5 mg or 2 mg, or placebo) to three capsules (HL 0.75 mg –3 mg, or placebo). If side effects (e.g., EPS) were limiting, the dose was maintained at two capsules daily. At the end of the 6-week trial, 40 patients had received oral HL, with 35 of them having plasma levels drawn and assayed. Eleven of 18 patients on standard dose were taking 3 mg/day, and 7 patients were taking 2 mg/day. In the low-dose group, 14 of 17 patients were taking 0.75 mg/day, and three patients were taking 0.5 mg/day. All patients were on a stable dose for 5 weeks before evaluation of efficacy, side effects, and HL plasma levels. FIGURE 1.
Design for the placebo-controlled study of haloperidol in patients with AD
Pre-Entry 1 week
Phase A 6 weeks
Cell 1:
Placebo
Low Dose
Cell 2:
Placebo
Standard Dose
Cell 3:
Placebo
Placebo
Note: Low-dose haloperidol (HL): 0.50 mg/day–0.75 mg/day; standard-dose HL: 2 mg/day–3 mg/day.
188
All patients were free of psychotropic medications for at least 1 week before the initial 1-week placebo phase. Concomitant psychotropic medications were not prescribed during the study. Anticholinergic medications were not permitted to treat EPS, which were addressed by lowering the study medication dose. The primary outcome measures of efficacy were the Total BPRS score18 and BPRS Psychosis factor (hallucinatory behavior and unusual thought content items only; range: 0–12) and Hostile-Suspiciousness factor (hostility, suspiciousness, and uncooperativeness; range: 0–18), the Behavioral Syndromes Scale for Dementia (BSSD) item scores for Psychomotor Agitation (range: 1–6) and Physical Aggression (range: 1–6). Somatic side effects were assessed by the TreatmentEmergent Symptoms Scale (total score),19 EPS by the modified Targeting Abnormal Kinetic Effects (TAKE) scale (total score),20 and tardive dyskinesia by the Rockland Tardive Dyskinesia Scale.21 Blood for ascertaining HL levels was drawn at the 7-week time point (6 weeks in the randomized trial), with levels available on 52 of the 60 patients who completed the study (35 of 40 patients on HL, 17 of 20 patients on placebo). HL blood levels were drawn during the day, at approximately 12 hours after their last dose (range: 10–16 hours) given at bedtime the previous evening. If a patient was taking the medication twice daily, the morning dose was held until after blood was drawn. The clinical raters remained blind to the plasma-level results. A radioimmunoassay procedure (Janssen Pharmaceutica, Titusville, NJ) was used to assay plasma levels of HL. The results of the radioimmunoassay for plasma HL were strongly correlated with those from samples assayed by gas chromatography (r⳱0.96; slope: 0.98; intercept: 0.10 ng/ml). The sensitivity of the radioimmunoassay procedure enabled quantification of 100 pg/ml with a relative SD of less than 6%. Reduced HL, which would have been at extremely low concentrations, given the low oral doses used, was not assayed by this method. The SAS program for statistical analysis was used, with multivariate analyses of variance (MANOVAs) used to examine overall efficacy of oral HL (n⳱40) versus placebo (n⳱20), followed by univariate analyses.1 To confirm the results obtained with percent-change scores, analysis of covariance (ANCOVA) was conducted on the outcome measures at the end of the study, using the corresponding measures at the start of
Am J Geriatr Psychiatry 11:2, March-April 2003
Pelton et al. the study as covariates. Separate regression analyses were then performed on either BPRS Total, BPRS Psychosis sub-item total, or Hostile-Suspiciousness factor sub-item totals, with BSSD Aggression and Psychomotor Agitation, with HL dose (n⳱40), HL plasma levels (n⳱35), age, and sex as independent factors. The optimal HL plasma concentration that separated the responders from the nonresponders was calculated by constructing receiver operating characteristic (ROC) curves. Follow-up one-way analyses of variance, chisquare analyses, and t-tests were performed as indicated.
RESULTS All 71 AD outpatients in the study met entry criteria for behavioral dyscontrol, and 51 patients (71.8%) met entry criteria for psychosis. The subjects’ mean age was 71.82 (range: 50–86 years, SD: 9.63); the mean duration of illness was 5.01 years (range: 1–16 years; SD: 3.07); the mean number of years of education was 9.47 (range: 0–19 years; SD: 4.42), and 64.8% were female. The ethnic breakdown was 56.8% White, 29.7% Latino, 10.8% African American, and 2.7% other. A history of previous psychiatric disorders was uncommon (8.1%), and 37.8% of the subjects had a family history of dementia. Informants were mainly spouses (40.5%) or adult children (54.1%), and most (67.6%) lived with the patients. Fortythree patients (65.6%) had a CDR score of 1 or 2 (mild or moderate dementia), and 28 (34.4%) had a score of 3 (severe dementia). The mean MMSE score (0–30) was 9.8 (range: 1–24; SD: 5.7), and the mean Blessed scale score (0: no functional impairment; 17: severe functional impairment) was 8.1 (range: 2–16; SD: 3.29). The mean BPRS Total score was 59.2 (range: 35–83; SD: 9.35). One-way analyses of variance and chi-square analyses indicated that patients randomly assigned to the three cells did not differ in age, gender, BPRS Total score, presence or absence of psychosis, and severity of dementia (CDR and MMSE scores). Only 12 (16.9%) of the 71 patients had taken psychotropic medications during the month before study entry; three (4.2%) of these had taken low doses of antipsychotic medication on an as-needed basis. Sixty of the 71 patients completed the 6-week trial.1 Out of these 60 patients, 52 received the blood draw
Am J Geriatr Psychiatry 11:2, March-April 2003
and assay. No significant demographic or clinical differences existed between the 35 subjects who received the medication (low or high dose) and the 17 who were assigned to the placebo group. Efficacy of Oral-Dose HL Versus Placebo As previously reported,1 across the three treatment conditions, MANOVA on the efficacy measures (25% reduction in target-factor scores) revealed a main effect of treatment group (F[2,57]⳱3.23; p⬍0.05). In post-hoc univariate analyses and in two-group comparisons, standard-dose HL was significantly superior to placebo on BPRS Psychosis factor scores (t[38]⳱2.36; p⬍0.03), and BSSD Psychomotor Agitation (t[38]⳱2.18; p⬍0.04), but not for BPRS Hostile-Suspiciousness factor scores (t[38]⳱0.65; p⬍0.6). EPS tended to be greater with the standard dose than with placebo (t[38]⳱1.82; p⬍0.1). Low-dose HL did not differ from placebo on any measures of efficacy or side effects. To confirm the findings obtained with percent-change scores, the same between-group comparisons were conducted with use of ANCOVA for end-of-study dependent measures with the corresponding start-of-study measures as covariates. The results were similar to those obtained with percentchange scores. Correlation Between HL Dose and HL Plasma Level There was a strong correlation between oral dose of HL and plasma level (r[35]⳱0.69; p⬍0.001). HL Dose and HL Plasma Level: Associations With Clinical Outcome HL oral dose did not predict treatment response (Table 1). HL plasma level significantly correlated with change (improvement) in BPRS Total score, change in BPRS Psychosis factor, change in BPRS Hostile-Suspiciousness factor, change in global BSSD Aggression factor score, and change in BSSD Psychomotor Agitation (Table 1). In regression analysis, when both HL oral dose and HL plasma levels were included as two independent variables in the same analyses, with each BPRS or BSSD measure as the outcome variable (separate analyses), HL plasma level, but not HL oral dose, predicted change (improvement) in BPRS total score (F[2,32]⳱4.00; p⬍0.03), change in BPRS Psychosis factor (F[2,32]⳱6.16; p⬍0.005), change in BPRS Hostile-Suspiciousness factor
189
Plasma Haloperidol Levels
HL Plasma Level: Response Rate The optimal HL concentration that separated the responders from nonresponders was calculated by constructing ROC curves using a ⱖ20% decrease in the BPRS Total score to indicate response (ⱖ20% decrease in BPRS is standard in schizophrenia) and the continuous measure of HL plasma level. The ROC curve was plotted as the cumulative response rate (Y-axis) versus the cumulative nonresponse rate (X-axis) over an ordered list of plasma concentrations. The point of maximum sensitivity, or cut-point, revealing the greatest difference between response rate and nonresponse rate identified an optimal HL plasma range between 1.45 ng/ ml and 1.65 ng/ml as the lower limit of response (Figure 2). ROC analysis was unable to identify an upper limit concentration for HL blood level, presumably because of restricted range at higher levels. With a criterion of 20% reduction in BPRS Total score, BPRS Psychosis factor, BPRS Hostile-Suspiciousness factor, BSSD Global Aggression factor score, and BSSD Psychomotor Agitation as the outcome variable, each in separate analyses, HL plasma levels ⱖ1.5 ng/ml versus HL plasma levels ⬍1.5 ng/ml were analyzed by contingency-table analyses. Four (44%) of the nine patients with HL plasma level ⱖ1.5 ng/ml met response criteria on the BPRS Total score, versus 4 (15%) of 26 with HL plasma levels ⬍1.5 ng/ml (Fisher’s exact test [df 1], p⬍0.1; odds ratio (OR): TABLE 1.
4.4; 95% confidence interval [CI]: 0.81–23.89). Seven of nine patients (78%) with HL plasma level ⱖ1.5 ng/ ml met response criteria on the BPRS Psychosis factor, versus 8 of 26 (31%) with HL plasma levels ⬍1.5 ng/ml (Fisher’s exact test [df 1], p⬍0.02; OR: 7.85; 95% CI: 1.33–46.62). Seven of nine patients (78%) with HL plasma levels ⱖ1.5 ng/ml met response criteria on the BPRS Hostile-Suspiciousness factor, versus 14 of 26 (54%) with HL plasma levels ⬍1.5 ng/ml (Fisher’s exact test [df 1], p⬍0.2; OR: 3.00; 95% CI: 0.52–17.27). Four FIGURE 2.
Receiver operating characteristics (ROC) curve for response (using BPRS 20% change) for haloperidol blood levels ranging from 0.13 ng/ ml to 3.8 ng/ml
100
80
Sensitivity
(F[2,32]⳱5.07; p⬍0.01), change in BSSD Global Aggression factor (F[2,32]⳱4.32; p⬍0.03), and change in BSSD Psychomotor Agitation factor (F[2,32]⳱4.30; p⬍0.03).
1.45–1.65 ng/ml 60
40
20
0 0.00
0.20
0.40
0.60
0.80
1.00
1–Specificity
Correlations between oral dose or blood level and change in efficacy and side-effect measures during 6 weeks of haloperidol (HL) treatment in patients with AD Haloperidol Oral Dose (nⴔ40) a
BPRS Total score BPRS Hostility scorea BPRS Psychosis scorea BSSD Physical Aggressiona,d Psychomotor Agitationa,d Extrapyramidal symptomsa MMSEa
Haloperidol Blood Level (nⴔ35)
rb
p⬍
rc
p⬍
0.20 0.18 0.35
0.2 0.5 0.03
0.42 0.43 0.52
0.01 0.01 0.01
0.09 0.32 ⳮ0.32 0.05
0.6 0.05 0.05 0.8
0.38 0.46 ⳮ0.53 0.13
0.3 0.01 0.01 0.5
Note: AD: Alzheimer disease; BPRS: Brief Psychiatric Rating Scale; BSSD: Behavioral Syndromes Scale for Dementia; MMSE: Mini-Mental State Exam. a Change pre–post. b Spearman correlation coefficient. c Pearson correlation coefficient. d Item score (range: 1–6) on the Behavioral Syndrome Scale for Dementia.
190
Am J Geriatr Psychiatry 11:2, March-April 2003
Pelton et al. of the nine patients (44%) with HL plasma level ⱖ1.5 ng/ml met response criteria on the BSSD Aggression score, versus 7 of 26 (27%) with HL plasma levels ⬍1.5 ng/ml (Fisher’s exact test [df 1], p⬍0.3; OR: 2.17; 95% CI: 0.45–10.49). Seven of the nine patients (78%) with HL plasma level ⱖ1.5 ng/ml met response criteria on BSSD Psychomotor Agitation, versus 8 of 26 (31%) with HL plasma levels ⬍1.5 ng/ml (Fisher’s exact test [df 1], p⬍0.02; OR: 7.88; 95% CI: 1.33–46.63). We further confirmed that the HL plasma level identified by the ROC curve was an appropriately identified lower-limit concentration for clinical response by reanalyzing the data with contingency tables and an HL level of 1 ng/ml (data not shown); we found that the associations were weaker on all outcome measures. Because of a limited number of patients’ having a plasma level ⱖ2 ng/ml (n⳱5), dichotomous classification and analysis could not be done with the 2 ng/ml cut-point for classification. HL Dose and HL Plasma Level: Association With Side Effects For measures of side effects, we calculated difference scores (pre–post) in total EPS scores and 30-item MMSE. In a regression analysis, HL oral dose did not predict change in EPS or MMSE change. In a regression analysis, HL plasma level did not correlate with change in MMSE, but it did significantly correlate with increase in EPS. In a regression analysis, where both HL oral dose and HL plasma levels were included as independent variables, HL plasma level, but not HL oral dose, predicted change (worsening) of EPS over the 6-week trial (F[2,34]⳱7.40; p⬍0.005). Neither HL oral dose nor HL plasma level predicted change in MMSE over the 6-week trial.
DISCUSSION HL plasma levels, but not HL oral dose, significantly correlated with clinical efficacy as measured by reduction in BPRS Total and BPRS Psychosis factor scores, suggesting potential clinical usefulness for measurement of plasma HL levels in monitoring clinical response in AD patients with psychosis. We found no correlation between HL oral dose and increase in EPS, but there was a significant correlation between HL plasma level and increase in EPS. Since no
Am J Geriatr Psychiatry 11:2, March-April 2003
patient had EPS at the beginning of the study (all TAKE total scores were ⱕ4), it is doubtful that these were patients with increased antipsychotic sensitivity due to another neurodegenerative disease like Parkinson disease or Lewy-body disease. In fact, these results are consistent with the available data on EPS and tardive dyskinesia in aging schizophrenic patients and other elderly subjects receiving antipsychotic medications.22 These data with low-dose HL support the view that AD patients have a greater sensitivity to antipsychotics then younger schizophrenic patients. There was a strong correlation between HL oral dose and plasma level in the low-dose oral HL range (0.5 mg/day to 3 mg/day) used in this study. One possible pharmacokinetic explanation is that the use of very low doses of HL does not induce liver enzymes to a measurable extent. Another possibility is that because nearly all patients were antipsychotic-naı¨ve, there were no major changes in intestinal absorption that are known to be altered in schizophrenic patients receiving long-term antipsychotic treatment.23 Efficacy was seen at very low plasma HL levels (0.1 ng/ml to 3.8 ng/ml), as also observed in another study of patients with dementia (0.26 ng/ml to 2.01 ng/ml),14 a finding that contrasts with the putative therapeutic window of 5 ng/ml to 15 ng/ml in schizophrenic patients.5–7,9,24 In fact, every AD patient who responded to HL had a plasma level below 5 ng/ml. One pharmacodynamic explanation is loss of D2 receptors in aging. Thus, AD patients may only need a very low dose of HL to occupy a sufficiently large proportion of D2 receptors. Contrary to this explanation and the majority of clinical trials in schizophrenia, a recent PET imaging study by Kapur et al.25 and an optimal dose-finding study by Zhang-Wong et al.3 do not support a pharmacodynamic difference between drug-naı¨ve schizophrenic patients and AD patients. Using PET imaging, Kapur et al.25 found that first-break schizophrenic patients only need 65% of their D2 receptors to be occupied to obtain a clinical response, 72% occupancy to show hyperprolactinemia, and ⬎78% D2 occupancy for clinically significant EPS to emerge. They suggested that the therapeutic window is rather narrow: from 65% to 72%–78% D2 occupancy in first-break schizophrenic patients. The window corresponded to approximately 2 mg/day–3 mg/day of HL for a given patient.25 A similar therapeutic window was reported by Zhang-Wong et al. in first-break schizophrenic patients.3 The efficacy and side effect data in this study showed that HL 2 mg/day–
191
Plasma Haloperidol Levels 3 mg/day was superior to HL 0.5 mg/day–0.75 mg/day, although a subgroup on the higher dose developed moderate-to-severe EPS,1 confirming findings obtained in an earlier pilot study from our group.26 With findings contrary to ours, Dysken et al.14 treated 29 AD inpatients with behavior problems with fixed oral doses of HL (0.5 mg, 1.0 mg, or 2.0 mg) every 12 hours for 3 weeks. Using the BEHAVE-AD rating scale, they found no significant linear or curvilinear relationships between either percent change in BEHAVEAD in relation to HL plasma level, or emergence of EPS in relation to plasma HL levels. These differences may be due to several factors. First, their subjects were all inpatients, and 27/29 (97%) had previously been treated with an antipsychotic. Our patients were all outpatients, the majority drug-naı¨ve, with only 1/35 (4%) having previous exposure to an antipsychotic. Second, they found no significant change in EPS over the course of their study. In contrast, we found an increase in EPS that correlated significantly with HL blood levels. They note that the main reason for their finding was the presence of residual side effects from previous antipsychotic treatments. We, on the other hand, had no patient starting the study with significant EPS at baseline. Thus, the severity of illness, previous medication exposure, and baseline neurological status may explain the differences between these two studies.
Limitations to our study included a narrow dose range, fixed oral doses of HL, and the relatively small number of subjects. These factors may have diminished the likelihood of seeing an oral dose–efficacy and oral dose–side effect correlation. This study shows that HL plasma level is a strong predictor of outcome and emergence of EPS, with oral dose showing less robust associations. Clinically, these results suggest that measurement of HL plasma levels may be warranted if clinical response is not obtained on a low dose of HL in an AD patient or excessive EPS emerges at a lower-than-expected dose of HL. The data indirectly suggest that in AD outpatients who are neuroleptic-naı¨ve, an HL plasma-level range of 1.5 ng/ml– 4 ng/ml may be optimal in determining whether efficacy can be achieved, while limiting EPS in the AD patient. A comparable narrow therapeutic window may also apply to other antipsychotics, both typical and atypical.2,27 These results warrant further investigation of antipsychotic dosing schedules, plasma levels, and tolerability in AD patients with psychosis and behavioral dyscontrol. The PI (GHP) has financial support for work with Pfizer-Esai, a research fellowship through the AAGP, and NARSAD. This work was also supported in part by NIMH Grant MH55735.
References 1. Devanand DP, Marder K, Michaels KS, et al: A randomized, placebo-controlled, dose-comparison trial of haloperidol for psychosis and disruptive behaviors in Alzheimer’s disease. Am J Psychiatry 1998; 155:1512–1520 2. Katz IR, Jeste DV, Mintzer JE, et al: Comparison of risperidone and placebo for psychosis and behavioral disturbances associated with dementia: a randomized, double-blind trial: Risperidone Study Group. J Clin Psychiatry 1999; 60:107–115 3. Zhang-Wong J, Zipursky RB, Beiser M, Bean G: Optimal haloperidol dosage in first-episode psychosis. Can J Psychiatry 1999; 44:164–167 4. Street JS, Clark WS, Gannon KS, et al: Olanzapine treatment of psychotic and behavioral symptoms in patients with Alzheimer disease in nursing care facilities: a double-blind, randomized, placebo-controlled trial: The HGEU Study Group. Arch Gen Psychiatry 2000; 57:968–976 5. Kirch DG, Bigelow LB, Korpi ER, et al: Serum haloperidol concentration and clinical response in schizophrenia. Schizophr Bull 1988; 14:283–289 6. Perry PJ, Pfohl BM, Kelly MW: The relationship of haloperidol concentrations to therapeutic response. J Clin Psychopharmacol 1988; 8:38–43 7. Volavka J, Cooper TB: Review of haloperidol blood level and clinical response: looking through the window. J Clin Psychopharmacol 1987; 7:25–30
192
8. Van Putten T, Marder SR, Mintz J, et al: Haloperidol plasma levels and clinical response: a therapeutic window relationship. Am J Psychiatry 1992; 149:500–505 9. Palao DJ, Arauxo A, Brunet M, et al: Haloperidol: therapeutic window in schizophrenia. J Clin Psychopharmacol 1994; 14:303–310 10. Doddi S, Rifkin A, Karajgi B, et al: Blood levels of haloperidol and clinical outcome in schizophrenia. J Clin Psychopharmacol 1994; 14:187–195 11. Curry SH: Commentary: therapeutic monitoring of antipsychotic drugs. J Clin Psychopharmacol 1997; 17:124–126 12. Cheng YF, Paalzow LK, Bondesson U, et al: Pharmacokinetics of haloperidol in psychotic patients. Psychopharmacology (Berlin) 1987; 91:410–414 13. Lacro JP, Kuczenski R, Roznoski M, et al: Serum haloperidol levels in older psychotic patients. Am J Geriatr Psychiatry 1996; 4:229– 236 14. Dysken MW, Johnson SB, Holdon L, et al: Haloperidol concentrations in patients with Alzheimer’s dementia. Am J Geriatr Psychiatry 1994; 2:125–133 15. Devanand DP, Cooper T, Sackeim HA, et al: Low-dose oral haloperidol and blood levels in Alzheimer’s disease: a preliminary study. Psychopharmacol Bull 1992; 28:169–173 16. McKhann G, Drachman D, Folstein M, et al: Clinical diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA Work Group
Am J Geriatr Psychiatry 11:2, March-April 2003
Pelton et al. under the auspices of the Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology 1984; 34:939–944 17. Folstein M, Folstein S, McHugh P: Mini-Mental State: a practical method of grading the cognitive state of patients for clinicians. J Psychiatr Res 1975; 12:189–198 18. Overall JE, Gorham DR: The Brief Psychiatric Rating Scale. Psychol Rep 1962; 10:799–812 19. Guy WE (ed.): EDCU Assessment Manual for Psychopharmacology: Publication ADM 76-338, U.S. Department of Health, Education, and Welfare, 1976 20. Wocjik JD, Gelenburg AJ, La Brie RA, et al: Prevalence of tardive dyskinesia in an outpatient population. Compr Psychiatry 1980; 21:370–380 21. Simpson GM, Lee JH, Zoubok B, et al: A rating scale for tardive dyskinesia. Psychopharmacology (Berlin) 1979; 64:171–179 22. Jeste DV, Caligiuri MP, Paulsen JS, et al: Risk of tardive dyskinesia
Am J Geriatr Psychiatry 11:2, March-April 2003
in older patients. a prospective longitudinal study of 266 outpatients. Arch Gen Psychiatry 1995; 52:756–765 23. Sakalis G, Curry SH, Mould GP, et al: Physiologic and clinical effects of chlorpromazine and their relationship to plasma level. Clin Pharmacol Ther 1972; 13:931–946 24. Van Putten T, Marder SR, Mintz J, et al: Haloperidol plasma levels and clinical response: a therapeutic window relationship. Psychopharmacol Bull 1988; 24:172–175 25. Kapur S, Zipursky R, Jones C, et al: Relationship between dopamine D2 occupancy, clinical response, and side effects: a doubleblind PET study of first-episode schizophrenia. Am J Psychiatry 2000; 157:514–520 26. Devanand DP, Sackeim H, Brown R, et al: A pilot study of haloperidol treatment of psychosis and behavioral disturbance in Alzheimer’s disease. Arch Neurol 1989; 46:854–857 27. 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
193
Objective: This study investigated whether measurement of plasma levels may be useful in monitoring clinical efficacy and side effects during oral haloperidol (HL) treatment of psychosis and behavioral dyscontrol in patients with Alzheimer disease (AD). Methods: After a single-blind placebo period of 1 week, 71 outpatients with AD were randomized to a 6-week, double-blind, placebo-controlled trial of HL 2 mg–3 mg/ day (standard dose), HL 0.5 mg–0.75 mg/day (low dose), or placebo, with plasma levels for HL drawn at the end of 6 weeks. Results: Of the 40 patients who received active HL for 6 weeks, 35 had plasma levels drawn. Plasma levels were all below the lower limit of the postulated therapeutic range in schizophrenia. Nonetheless, HL plasma level significantly correlated with clinical efficacy as measured by reduction in Brief Psychiatric Rating Scale Total score, Psychosis factor, and Hostile-Suspiciousness factor, the Behavioral Syndromes Scale for Dementia Psychomotor Agitation scale, and with the severity of extrapyramidal side effects (EPS). Oral dose did not significantly correlate with any of these efficacy or side-effect measures. Plasma levels significantly correlated with HL dose. When both HL dose and HL plasma level were included as independent variables in linear-regression analyses, only HL plasma level was a significant predictor of efficacy and EPS. Conclusion: Measurement of HL plasma levels may have potential usefulness as an adjunct in monitoring treatment response to oral HL in AD patients with psychosis or disruptive behavior. (Am J Geriatr Psychiatry 2003; 11:186–193)
Received April 17, 2001; revised June 24, July 30, 2002; accepted July 31, 2002. From the Department of Biological Psychiatry (GHP,DPD,XL,KM), the Department of Analytical Psychopharmacology (TBC), and the Memory Disorders Clinic (GHP,DPD,KB,KM), New York State Psychiatric Institute; The Taub Center for Alzheimer’s Disease Research, New York (KB,KM); the Gertrude H. Sergievsky Center at Columbia University, New York (KB,KM); and the Department of Psychiatry (GHP,DPD,XL,TBC) and the Department of Neurology (KB,KM), College of Physicians and Surgeons of Columbia University, New York. Address correspondence to Dr. Gregory Pelton, New York State Psychiatric Institute, 1051 Riverside Drive, Unit 126, New York, NY 10032. e-mail: [email protected] Copyright 䉷 2003 American Association for Geriatric Psychiatry
186
Am J Geriatr Psychiatry 11:2, March-April 2003
Pelton et al.
I
n patients with dementia who develop psychosis and behavioral dyscontrol, efficacy has been established for haloperidol (HL),1 risperidone,2,3 and olanzapine.4 Part of the challenge in establishing efficacy for these agents includes their relatively narrow therapeutic oral dose range, with low doses providing the optimal tradeoff between efficacy and side effects. In young schizophrenic patients, some investigators have found a linear or curvilinear relationship between plasma levels of antipsychotics (primarily HL), clinical response, and side effects. Most studies show a therapeutic concentration range for serum HL levels of approximately 5 ng/ml to 15 ng/ml at steady state for optimal clinical benefit,5–9 with few dissenting reports.10,11 Plasma-level measurement may be more useful in AD patients for several reasons. First, most AD patients are antipsychotic-naı¨ve when they first develop psychosis or disruptive behavior. Hence, there is less potential for the induction of enzymes in the intestinal wall and/ or liver, both of which can alter HL plasma levels.12 Second, since very low doses of HL (0.5 mg/day to 3 mg/ day) are used in AD patients, a more linear pharmacokinetic relationship between HL oral dose and HL plasma level may exist. Consequently, the curvilinear relationship between HL plasma level and response reported at higher dose levels in young schizophrenic patients, may not apply to AD patients. Rather, the HL plasma levels in AD patients may be limited to the ascending limb (low plasma level) section of the dose– response curve. Third, very low HL plasma levels will not result in the saturation of the D2 receptor system seen with higher dose treatment; that is, a pharmacodynamic difference in these patients is possible. Few studies have examined the relationship between HL plasma level and clinical response in elderly patients with dementia. Lacro et al.13 reported on plasma levels in 32 psychotic elderly patients, age 45– 83 years, diagnosed with schizophrenia (n⳱23) and AD (n⳱9). As a group, age positively correlated with the ratio of HL plasma level to HL oral dose needed for clinical stability (r⳱0.446; p⬍0.05). However, in the AD subgroup, this correlation was lost, which may have been due to small sample size. Dysken et al.,14 on the other hand, reported that in a sample of 29 AD inpatients with behavioral problems (mean age, 76 years; standard deviation [SD]: 8 years), who received a fixed oral dosage of HL (0.5mg, 1.0 mg, or 2.0 mg) every 12 hours for 3 weeks, 55% of patients showed a good re-
Am J Geriatr Psychiatry 11:2, March-April 2003
sponse to low-dose HL, but there was no significant linear or curvilinear relationship between HL plasma levels and clinical response. In an initial study, we reported on 19 elderly patients with AD and psychosis or behavioral disturbances, who received low-dose HL (0.5 mg/day to 5 mg/day) for 6–8 weeks. Oral dose strongly correlated with HL plasma level (r⳱0.82). There were suggestions that, compared with oral dose, HL plasma level correlated more strongly with clinical improvement (Brief Psychiatric Rating Scale [BPRS] scores) and extrapyramidal signs (EPS).15 In a study comparing the efficacy and side effects of two doses of oral HL and placebo in the treatment of psychosis and disruptive behavior in patients with AD,1 HL plasma levels were assayed (10 of the subjects who received HL were included in the Devanand et al. 1992 report15). We hypothesized that HL plasma level would be associated more strongly with treatment response than oral dose and that HL plasma level would also correlate strongly with severity of EPS, suggesting potential clinical usefulness in monitoring HL plasma levels.
METHODS All subjects were outpatients at a memory disorders clinic that is part of an AD research center. Subjects were required to meet DSM-III-R criteria for dementia and the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer’s Disease and Related Disorders Association criteria for probable AD.16 Diagnostic work-up included neurologic and psychiatric evaluation, neuropsychological testing, laboratory tests, and a computerized tomographic (CT) or magnetic resonance imaging (MRI) scan of the brain. Exclusion criteria were medication or alcohol dependence, stroke (clinical evidence of cortical stroke or a lesion ⱖ2 cm in diameter on any CT or MRI slice), and a history or clinical evidence of other causes of dementia, including head trauma, Parkinson disease, Huntington disease, or multiple sclerosis. Patients who met inclusion criteria for either psychosis or behavioral dyscontrol, as previously described,1 were eligible for the study. At the initial visit, global cognitive functioning was assessed by the MiniMental State Exam (Folstein 30-item MMSE);17 severity of dementia was assessed with the Clinical Dementia
187
Plasma Haloperidol Levels Rating Scale (CDR), and the modified Blessed Functional Activities Scale was given to the informant to assess functional impairment. The New York State Psychiatric Institute review board approved the research protocol. We obtained informed consent from the patient and a surrogate advocate, usually a family member. The consent procedures followed specific New York State regulations concerning research involving patients who do or do not have the capacity to consent. Family members served as informants and were required to have had contact with the patient at least once a week on average during the 3 months before the patient entered the study. Figure 1 summarizes the study design. In the initial, 1-week, single-blind phase, all patients received placebo. At the end of this week, the patients who still met entry criteria entered a 6-week random-assignment, double-blind, parallel-group, placebo-controlled comparison of three treatment conditions: standard dose HL (2 mg–3 mg/day), low-dose HL (0.5 mg–0.75 mg/day), and placebo.1 The subsequent crossover phase is not shown, as it is not relevant to this plasma-level study. During the 6-week trial, after the first week, the daily dose was raised from two capsules (HL 0.5 mg or 2 mg, or placebo) to three capsules (HL 0.75 mg –3 mg, or placebo). If side effects (e.g., EPS) were limiting, the dose was maintained at two capsules daily. At the end of the 6-week trial, 40 patients had received oral HL, with 35 of them having plasma levels drawn and assayed. Eleven of 18 patients on standard dose were taking 3 mg/day, and 7 patients were taking 2 mg/day. In the low-dose group, 14 of 17 patients were taking 0.75 mg/day, and three patients were taking 0.5 mg/day. All patients were on a stable dose for 5 weeks before evaluation of efficacy, side effects, and HL plasma levels. FIGURE 1.
Design for the placebo-controlled study of haloperidol in patients with AD
Pre-Entry 1 week
Phase A 6 weeks
Cell 1:
Placebo
Low Dose
Cell 2:
Placebo
Standard Dose
Cell 3:
Placebo
Placebo
Note: Low-dose haloperidol (HL): 0.50 mg/day–0.75 mg/day; standard-dose HL: 2 mg/day–3 mg/day.
188
All patients were free of psychotropic medications for at least 1 week before the initial 1-week placebo phase. Concomitant psychotropic medications were not prescribed during the study. Anticholinergic medications were not permitted to treat EPS, which were addressed by lowering the study medication dose. The primary outcome measures of efficacy were the Total BPRS score18 and BPRS Psychosis factor (hallucinatory behavior and unusual thought content items only; range: 0–12) and Hostile-Suspiciousness factor (hostility, suspiciousness, and uncooperativeness; range: 0–18), the Behavioral Syndromes Scale for Dementia (BSSD) item scores for Psychomotor Agitation (range: 1–6) and Physical Aggression (range: 1–6). Somatic side effects were assessed by the TreatmentEmergent Symptoms Scale (total score),19 EPS by the modified Targeting Abnormal Kinetic Effects (TAKE) scale (total score),20 and tardive dyskinesia by the Rockland Tardive Dyskinesia Scale.21 Blood for ascertaining HL levels was drawn at the 7-week time point (6 weeks in the randomized trial), with levels available on 52 of the 60 patients who completed the study (35 of 40 patients on HL, 17 of 20 patients on placebo). HL blood levels were drawn during the day, at approximately 12 hours after their last dose (range: 10–16 hours) given at bedtime the previous evening. If a patient was taking the medication twice daily, the morning dose was held until after blood was drawn. The clinical raters remained blind to the plasma-level results. A radioimmunoassay procedure (Janssen Pharmaceutica, Titusville, NJ) was used to assay plasma levels of HL. The results of the radioimmunoassay for plasma HL were strongly correlated with those from samples assayed by gas chromatography (r⳱0.96; slope: 0.98; intercept: 0.10 ng/ml). The sensitivity of the radioimmunoassay procedure enabled quantification of 100 pg/ml with a relative SD of less than 6%. Reduced HL, which would have been at extremely low concentrations, given the low oral doses used, was not assayed by this method. The SAS program for statistical analysis was used, with multivariate analyses of variance (MANOVAs) used to examine overall efficacy of oral HL (n⳱40) versus placebo (n⳱20), followed by univariate analyses.1 To confirm the results obtained with percent-change scores, analysis of covariance (ANCOVA) was conducted on the outcome measures at the end of the study, using the corresponding measures at the start of
Am J Geriatr Psychiatry 11:2, March-April 2003
Pelton et al. the study as covariates. Separate regression analyses were then performed on either BPRS Total, BPRS Psychosis sub-item total, or Hostile-Suspiciousness factor sub-item totals, with BSSD Aggression and Psychomotor Agitation, with HL dose (n⳱40), HL plasma levels (n⳱35), age, and sex as independent factors. The optimal HL plasma concentration that separated the responders from the nonresponders was calculated by constructing receiver operating characteristic (ROC) curves. Follow-up one-way analyses of variance, chisquare analyses, and t-tests were performed as indicated.
RESULTS All 71 AD outpatients in the study met entry criteria for behavioral dyscontrol, and 51 patients (71.8%) met entry criteria for psychosis. The subjects’ mean age was 71.82 (range: 50–86 years, SD: 9.63); the mean duration of illness was 5.01 years (range: 1–16 years; SD: 3.07); the mean number of years of education was 9.47 (range: 0–19 years; SD: 4.42), and 64.8% were female. The ethnic breakdown was 56.8% White, 29.7% Latino, 10.8% African American, and 2.7% other. A history of previous psychiatric disorders was uncommon (8.1%), and 37.8% of the subjects had a family history of dementia. Informants were mainly spouses (40.5%) or adult children (54.1%), and most (67.6%) lived with the patients. Fortythree patients (65.6%) had a CDR score of 1 or 2 (mild or moderate dementia), and 28 (34.4%) had a score of 3 (severe dementia). The mean MMSE score (0–30) was 9.8 (range: 1–24; SD: 5.7), and the mean Blessed scale score (0: no functional impairment; 17: severe functional impairment) was 8.1 (range: 2–16; SD: 3.29). The mean BPRS Total score was 59.2 (range: 35–83; SD: 9.35). One-way analyses of variance and chi-square analyses indicated that patients randomly assigned to the three cells did not differ in age, gender, BPRS Total score, presence or absence of psychosis, and severity of dementia (CDR and MMSE scores). Only 12 (16.9%) of the 71 patients had taken psychotropic medications during the month before study entry; three (4.2%) of these had taken low doses of antipsychotic medication on an as-needed basis. Sixty of the 71 patients completed the 6-week trial.1 Out of these 60 patients, 52 received the blood draw
Am J Geriatr Psychiatry 11:2, March-April 2003
and assay. No significant demographic or clinical differences existed between the 35 subjects who received the medication (low or high dose) and the 17 who were assigned to the placebo group. Efficacy of Oral-Dose HL Versus Placebo As previously reported,1 across the three treatment conditions, MANOVA on the efficacy measures (25% reduction in target-factor scores) revealed a main effect of treatment group (F[2,57]⳱3.23; p⬍0.05). In post-hoc univariate analyses and in two-group comparisons, standard-dose HL was significantly superior to placebo on BPRS Psychosis factor scores (t[38]⳱2.36; p⬍0.03), and BSSD Psychomotor Agitation (t[38]⳱2.18; p⬍0.04), but not for BPRS Hostile-Suspiciousness factor scores (t[38]⳱0.65; p⬍0.6). EPS tended to be greater with the standard dose than with placebo (t[38]⳱1.82; p⬍0.1). Low-dose HL did not differ from placebo on any measures of efficacy or side effects. To confirm the findings obtained with percent-change scores, the same between-group comparisons were conducted with use of ANCOVA for end-of-study dependent measures with the corresponding start-of-study measures as covariates. The results were similar to those obtained with percentchange scores. Correlation Between HL Dose and HL Plasma Level There was a strong correlation between oral dose of HL and plasma level (r[35]⳱0.69; p⬍0.001). HL Dose and HL Plasma Level: Associations With Clinical Outcome HL oral dose did not predict treatment response (Table 1). HL plasma level significantly correlated with change (improvement) in BPRS Total score, change in BPRS Psychosis factor, change in BPRS Hostile-Suspiciousness factor, change in global BSSD Aggression factor score, and change in BSSD Psychomotor Agitation (Table 1). In regression analysis, when both HL oral dose and HL plasma levels were included as two independent variables in the same analyses, with each BPRS or BSSD measure as the outcome variable (separate analyses), HL plasma level, but not HL oral dose, predicted change (improvement) in BPRS total score (F[2,32]⳱4.00; p⬍0.03), change in BPRS Psychosis factor (F[2,32]⳱6.16; p⬍0.005), change in BPRS Hostile-Suspiciousness factor
189
Plasma Haloperidol Levels
HL Plasma Level: Response Rate The optimal HL concentration that separated the responders from nonresponders was calculated by constructing ROC curves using a ⱖ20% decrease in the BPRS Total score to indicate response (ⱖ20% decrease in BPRS is standard in schizophrenia) and the continuous measure of HL plasma level. The ROC curve was plotted as the cumulative response rate (Y-axis) versus the cumulative nonresponse rate (X-axis) over an ordered list of plasma concentrations. The point of maximum sensitivity, or cut-point, revealing the greatest difference between response rate and nonresponse rate identified an optimal HL plasma range between 1.45 ng/ ml and 1.65 ng/ml as the lower limit of response (Figure 2). ROC analysis was unable to identify an upper limit concentration for HL blood level, presumably because of restricted range at higher levels. With a criterion of 20% reduction in BPRS Total score, BPRS Psychosis factor, BPRS Hostile-Suspiciousness factor, BSSD Global Aggression factor score, and BSSD Psychomotor Agitation as the outcome variable, each in separate analyses, HL plasma levels ⱖ1.5 ng/ml versus HL plasma levels ⬍1.5 ng/ml were analyzed by contingency-table analyses. Four (44%) of the nine patients with HL plasma level ⱖ1.5 ng/ml met response criteria on the BPRS Total score, versus 4 (15%) of 26 with HL plasma levels ⬍1.5 ng/ml (Fisher’s exact test [df 1], p⬍0.1; odds ratio (OR): TABLE 1.
4.4; 95% confidence interval [CI]: 0.81–23.89). Seven of nine patients (78%) with HL plasma level ⱖ1.5 ng/ ml met response criteria on the BPRS Psychosis factor, versus 8 of 26 (31%) with HL plasma levels ⬍1.5 ng/ml (Fisher’s exact test [df 1], p⬍0.02; OR: 7.85; 95% CI: 1.33–46.62). Seven of nine patients (78%) with HL plasma levels ⱖ1.5 ng/ml met response criteria on the BPRS Hostile-Suspiciousness factor, versus 14 of 26 (54%) with HL plasma levels ⬍1.5 ng/ml (Fisher’s exact test [df 1], p⬍0.2; OR: 3.00; 95% CI: 0.52–17.27). Four FIGURE 2.
Receiver operating characteristics (ROC) curve for response (using BPRS 20% change) for haloperidol blood levels ranging from 0.13 ng/ ml to 3.8 ng/ml
100
80
Sensitivity
(F[2,32]⳱5.07; p⬍0.01), change in BSSD Global Aggression factor (F[2,32]⳱4.32; p⬍0.03), and change in BSSD Psychomotor Agitation factor (F[2,32]⳱4.30; p⬍0.03).
1.45–1.65 ng/ml 60
40
20
0 0.00
0.20
0.40
0.60
0.80
1.00
1–Specificity
Correlations between oral dose or blood level and change in efficacy and side-effect measures during 6 weeks of haloperidol (HL) treatment in patients with AD Haloperidol Oral Dose (nⴔ40) a
BPRS Total score BPRS Hostility scorea BPRS Psychosis scorea BSSD Physical Aggressiona,d Psychomotor Agitationa,d Extrapyramidal symptomsa MMSEa
Haloperidol Blood Level (nⴔ35)
rb
p⬍
rc
p⬍
0.20 0.18 0.35
0.2 0.5 0.03
0.42 0.43 0.52
0.01 0.01 0.01
0.09 0.32 ⳮ0.32 0.05
0.6 0.05 0.05 0.8
0.38 0.46 ⳮ0.53 0.13
0.3 0.01 0.01 0.5
Note: AD: Alzheimer disease; BPRS: Brief Psychiatric Rating Scale; BSSD: Behavioral Syndromes Scale for Dementia; MMSE: Mini-Mental State Exam. a Change pre–post. b Spearman correlation coefficient. c Pearson correlation coefficient. d Item score (range: 1–6) on the Behavioral Syndrome Scale for Dementia.
190
Am J Geriatr Psychiatry 11:2, March-April 2003
Pelton et al. of the nine patients (44%) with HL plasma level ⱖ1.5 ng/ml met response criteria on the BSSD Aggression score, versus 7 of 26 (27%) with HL plasma levels ⬍1.5 ng/ml (Fisher’s exact test [df 1], p⬍0.3; OR: 2.17; 95% CI: 0.45–10.49). Seven of the nine patients (78%) with HL plasma level ⱖ1.5 ng/ml met response criteria on BSSD Psychomotor Agitation, versus 8 of 26 (31%) with HL plasma levels ⬍1.5 ng/ml (Fisher’s exact test [df 1], p⬍0.02; OR: 7.88; 95% CI: 1.33–46.63). We further confirmed that the HL plasma level identified by the ROC curve was an appropriately identified lower-limit concentration for clinical response by reanalyzing the data with contingency tables and an HL level of 1 ng/ml (data not shown); we found that the associations were weaker on all outcome measures. Because of a limited number of patients’ having a plasma level ⱖ2 ng/ml (n⳱5), dichotomous classification and analysis could not be done with the 2 ng/ml cut-point for classification. HL Dose and HL Plasma Level: Association With Side Effects For measures of side effects, we calculated difference scores (pre–post) in total EPS scores and 30-item MMSE. In a regression analysis, HL oral dose did not predict change in EPS or MMSE change. In a regression analysis, HL plasma level did not correlate with change in MMSE, but it did significantly correlate with increase in EPS. In a regression analysis, where both HL oral dose and HL plasma levels were included as independent variables, HL plasma level, but not HL oral dose, predicted change (worsening) of EPS over the 6-week trial (F[2,34]⳱7.40; p⬍0.005). Neither HL oral dose nor HL plasma level predicted change in MMSE over the 6-week trial.
DISCUSSION HL plasma levels, but not HL oral dose, significantly correlated with clinical efficacy as measured by reduction in BPRS Total and BPRS Psychosis factor scores, suggesting potential clinical usefulness for measurement of plasma HL levels in monitoring clinical response in AD patients with psychosis. We found no correlation between HL oral dose and increase in EPS, but there was a significant correlation between HL plasma level and increase in EPS. Since no
Am J Geriatr Psychiatry 11:2, March-April 2003
patient had EPS at the beginning of the study (all TAKE total scores were ⱕ4), it is doubtful that these were patients with increased antipsychotic sensitivity due to another neurodegenerative disease like Parkinson disease or Lewy-body disease. In fact, these results are consistent with the available data on EPS and tardive dyskinesia in aging schizophrenic patients and other elderly subjects receiving antipsychotic medications.22 These data with low-dose HL support the view that AD patients have a greater sensitivity to antipsychotics then younger schizophrenic patients. There was a strong correlation between HL oral dose and plasma level in the low-dose oral HL range (0.5 mg/day to 3 mg/day) used in this study. One possible pharmacokinetic explanation is that the use of very low doses of HL does not induce liver enzymes to a measurable extent. Another possibility is that because nearly all patients were antipsychotic-naı¨ve, there were no major changes in intestinal absorption that are known to be altered in schizophrenic patients receiving long-term antipsychotic treatment.23 Efficacy was seen at very low plasma HL levels (0.1 ng/ml to 3.8 ng/ml), as also observed in another study of patients with dementia (0.26 ng/ml to 2.01 ng/ml),14 a finding that contrasts with the putative therapeutic window of 5 ng/ml to 15 ng/ml in schizophrenic patients.5–7,9,24 In fact, every AD patient who responded to HL had a plasma level below 5 ng/ml. One pharmacodynamic explanation is loss of D2 receptors in aging. Thus, AD patients may only need a very low dose of HL to occupy a sufficiently large proportion of D2 receptors. Contrary to this explanation and the majority of clinical trials in schizophrenia, a recent PET imaging study by Kapur et al.25 and an optimal dose-finding study by Zhang-Wong et al.3 do not support a pharmacodynamic difference between drug-naı¨ve schizophrenic patients and AD patients. Using PET imaging, Kapur et al.25 found that first-break schizophrenic patients only need 65% of their D2 receptors to be occupied to obtain a clinical response, 72% occupancy to show hyperprolactinemia, and ⬎78% D2 occupancy for clinically significant EPS to emerge. They suggested that the therapeutic window is rather narrow: from 65% to 72%–78% D2 occupancy in first-break schizophrenic patients. The window corresponded to approximately 2 mg/day–3 mg/day of HL for a given patient.25 A similar therapeutic window was reported by Zhang-Wong et al. in first-break schizophrenic patients.3 The efficacy and side effect data in this study showed that HL 2 mg/day–
191
Plasma Haloperidol Levels 3 mg/day was superior to HL 0.5 mg/day–0.75 mg/day, although a subgroup on the higher dose developed moderate-to-severe EPS,1 confirming findings obtained in an earlier pilot study from our group.26 With findings contrary to ours, Dysken et al.14 treated 29 AD inpatients with behavior problems with fixed oral doses of HL (0.5 mg, 1.0 mg, or 2.0 mg) every 12 hours for 3 weeks. Using the BEHAVE-AD rating scale, they found no significant linear or curvilinear relationships between either percent change in BEHAVEAD in relation to HL plasma level, or emergence of EPS in relation to plasma HL levels. These differences may be due to several factors. First, their subjects were all inpatients, and 27/29 (97%) had previously been treated with an antipsychotic. Our patients were all outpatients, the majority drug-naı¨ve, with only 1/35 (4%) having previous exposure to an antipsychotic. Second, they found no significant change in EPS over the course of their study. In contrast, we found an increase in EPS that correlated significantly with HL blood levels. They note that the main reason for their finding was the presence of residual side effects from previous antipsychotic treatments. We, on the other hand, had no patient starting the study with significant EPS at baseline. Thus, the severity of illness, previous medication exposure, and baseline neurological status may explain the differences between these two studies.
Limitations to our study included a narrow dose range, fixed oral doses of HL, and the relatively small number of subjects. These factors may have diminished the likelihood of seeing an oral dose–efficacy and oral dose–side effect correlation. This study shows that HL plasma level is a strong predictor of outcome and emergence of EPS, with oral dose showing less robust associations. Clinically, these results suggest that measurement of HL plasma levels may be warranted if clinical response is not obtained on a low dose of HL in an AD patient or excessive EPS emerges at a lower-than-expected dose of HL. The data indirectly suggest that in AD outpatients who are neuroleptic-naı¨ve, an HL plasma-level range of 1.5 ng/ml– 4 ng/ml may be optimal in determining whether efficacy can be achieved, while limiting EPS in the AD patient. A comparable narrow therapeutic window may also apply to other antipsychotics, both typical and atypical.2,27 These results warrant further investigation of antipsychotic dosing schedules, plasma levels, and tolerability in AD patients with psychosis and behavioral dyscontrol. The PI (GHP) has financial support for work with Pfizer-Esai, a research fellowship through the AAGP, and NARSAD. This work was also supported in part by NIMH Grant MH55735.
References 1. Devanand DP, Marder K, Michaels KS, et al: A randomized, placebo-controlled, dose-comparison trial of haloperidol for psychosis and disruptive behaviors in Alzheimer’s disease. Am J Psychiatry 1998; 155:1512–1520 2. Katz IR, Jeste DV, Mintzer JE, et al: Comparison of risperidone and placebo for psychosis and behavioral disturbances associated with dementia: a randomized, double-blind trial: Risperidone Study Group. J Clin Psychiatry 1999; 60:107–115 3. Zhang-Wong J, Zipursky RB, Beiser M, Bean G: Optimal haloperidol dosage in first-episode psychosis. Can J Psychiatry 1999; 44:164–167 4. Street JS, Clark WS, Gannon KS, et al: Olanzapine treatment of psychotic and behavioral symptoms in patients with Alzheimer disease in nursing care facilities: a double-blind, randomized, placebo-controlled trial: The HGEU Study Group. Arch Gen Psychiatry 2000; 57:968–976 5. Kirch DG, Bigelow LB, Korpi ER, et al: Serum haloperidol concentration and clinical response in schizophrenia. Schizophr Bull 1988; 14:283–289 6. Perry PJ, Pfohl BM, Kelly MW: The relationship of haloperidol concentrations to therapeutic response. J Clin Psychopharmacol 1988; 8:38–43 7. Volavka J, Cooper TB: Review of haloperidol blood level and clinical response: looking through the window. J Clin Psychopharmacol 1987; 7:25–30
192
8. Van Putten T, Marder SR, Mintz J, et al: Haloperidol plasma levels and clinical response: a therapeutic window relationship. Am J Psychiatry 1992; 149:500–505 9. Palao DJ, Arauxo A, Brunet M, et al: Haloperidol: therapeutic window in schizophrenia. J Clin Psychopharmacol 1994; 14:303–310 10. Doddi S, Rifkin A, Karajgi B, et al: Blood levels of haloperidol and clinical outcome in schizophrenia. J Clin Psychopharmacol 1994; 14:187–195 11. Curry SH: Commentary: therapeutic monitoring of antipsychotic drugs. J Clin Psychopharmacol 1997; 17:124–126 12. Cheng YF, Paalzow LK, Bondesson U, et al: Pharmacokinetics of haloperidol in psychotic patients. Psychopharmacology (Berlin) 1987; 91:410–414 13. Lacro JP, Kuczenski R, Roznoski M, et al: Serum haloperidol levels in older psychotic patients. Am J Geriatr Psychiatry 1996; 4:229– 236 14. Dysken MW, Johnson SB, Holdon L, et al: Haloperidol concentrations in patients with Alzheimer’s dementia. Am J Geriatr Psychiatry 1994; 2:125–133 15. Devanand DP, Cooper T, Sackeim HA, et al: Low-dose oral haloperidol and blood levels in Alzheimer’s disease: a preliminary study. Psychopharmacol Bull 1992; 28:169–173 16. McKhann G, Drachman D, Folstein M, et al: Clinical diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA Work Group
Am J Geriatr Psychiatry 11:2, March-April 2003
Pelton et al. under the auspices of the Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology 1984; 34:939–944 17. Folstein M, Folstein S, McHugh P: Mini-Mental State: a practical method of grading the cognitive state of patients for clinicians. J Psychiatr Res 1975; 12:189–198 18. Overall JE, Gorham DR: The Brief Psychiatric Rating Scale. Psychol Rep 1962; 10:799–812 19. Guy WE (ed.): EDCU Assessment Manual for Psychopharmacology: Publication ADM 76-338, U.S. Department of Health, Education, and Welfare, 1976 20. Wocjik JD, Gelenburg AJ, La Brie RA, et al: Prevalence of tardive dyskinesia in an outpatient population. Compr Psychiatry 1980; 21:370–380 21. Simpson GM, Lee JH, Zoubok B, et al: A rating scale for tardive dyskinesia. Psychopharmacology (Berlin) 1979; 64:171–179 22. Jeste DV, Caligiuri MP, Paulsen JS, et al: Risk of tardive dyskinesia
Am J Geriatr Psychiatry 11:2, March-April 2003
in older patients. a prospective longitudinal study of 266 outpatients. Arch Gen Psychiatry 1995; 52:756–765 23. Sakalis G, Curry SH, Mould GP, et al: Physiologic and clinical effects of chlorpromazine and their relationship to plasma level. Clin Pharmacol Ther 1972; 13:931–946 24. Van Putten T, Marder SR, Mintz J, et al: Haloperidol plasma levels and clinical response: a therapeutic window relationship. Psychopharmacol Bull 1988; 24:172–175 25. Kapur S, Zipursky R, Jones C, et al: Relationship between dopamine D2 occupancy, clinical response, and side effects: a doubleblind PET study of first-episode schizophrenia. Am J Psychiatry 2000; 157:514–520 26. Devanand DP, Sackeim H, Brown R, et al: A pilot study of haloperidol treatment of psychosis and behavioral disturbance in Alzheimer’s disease. Arch Neurol 1989; 46:854–857 27. 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
193