The debate about optimal dosing of antipsychotic medication is still a lively one. Haase (1) introduced the concept "neuroleptic threshold dose": a dose at which patients develop minimal extrapyramidal side effects. Exceeding this dose resulted in more extrapyramidal side effects but not in a better response (2). The concept of a "therapeutic window" emerged (3). Since the 1980s, the occupancy of striatal dopamine D2 receptors and, consequently, the "neuroleptic threshold dose" or "therapeutic window" has been assessed in vivo with scintigraphic techniques (4–6). It has been suggested that a D2 receptor occupancy threshold in the range of 60%–70% is needed to obtain satisfactory antipsychotic response (7–9). A stepped increase in response was found beyond 65% D2 receptor occupancy; prolactin elevation became prominent beyond 72%, while extrapyramidal side effects were evident beyond 78% (10).
Dopaminergic neurotransmission also plays an important role in the reward system, and inhibition of this system can be accompanied by reduced motivation, drive, and spontaneity (11). Occupancy of dopamine receptors by antipsychotic medication is thought to influence a patient’s subjective experience (12–15).
We hypothesized that there would also be a "window" for the influence of antipsychotic medication on the subjective experience of patients. In an earlier study, we found that levels of D2 occupancy lower than those that cause extrapyramidal side effects may be related to negative subjective experience (12). The "subjective side effect window" in dosing antipsychotic drugs may be small. We hypothesized that "optimal occupancy of D2 receptors" would lie between 60% and 70%. We proposed that D2 binding should not exceed 70% to avoid excessive inhibition of dopamine neurotransmission, resulting in a negative impact on subjective experience. On the other hand, D2 occupancy below 60% leaves a patient in a psychotic state (too high a functional level of the dopamine system) accompanied by negative subjective well-being. This line of thought can be compared to the hypothesis about an optimal range in which D2 receptor stimulation by drugs of abuse can be perceived as reinforcing: too little may not be sufficient, but too much may have an aversive effect (16).
The subjective experience of patients treated with antipsychotic medication is related to medication compliance (17–19). Moreover, subjectively experienced side effects are more distressing than other side effects (20). Strategies to improve the subjective experience of patients receiving antipsychotic medication are needed to improve compliance and thereby ameliorate the course of schizophrenia.
The development of a new generation of antipsychotic medications could be important in this respect. Olanzapine has been found to be associated with a better subjective experience than typical antipsychotic drugs (21, 22). However, different levels of D2 occupancy have confounded previous comparisons of antipsychotic drugs (23). If the severity of negative subjective experience is related to D2 receptor occupancy, then the new antipsychotic medications might not be more beneficial for subjective experience than typical antipsychotic drugs in doses with comparable D2 receptor occupancy. Therefore, another strategy to improve the subjective experience might be treatment with low-dose typical antipsychotic drugs.
Limitations of earlier studies on subjective experience with typical and atypical antipsychotic medications were the noncomparative dosing and the lack of a randomized double-blind design. We therefore conducted a 6-week, randomized, double-blind study of fixed low-dose olanzapine or haloperidol in young patients with recent-onset schizophrenia. To study the relation between subjective experience and D2 receptor occupancy by these drugs, we assessed in vivo striatal dopamine D2 receptor occupancy with [123I]iodobenzamide ([123I]IBZM) single photon emission computed tomography (SPECT).
This prospective, randomized, double-blind study was conducted at the Adolescent Clinic of the Academic Medical Center in Amsterdam, a specialized unit for in- and outpatient treatment of young persons with recent-onset schizophrenia. Local ethics committee approvals were obtained, and subjects gave their written informed consent after a complete description of the study.
Male and female patients aged 17–28 years who met DSM-IV criteria for schizophrenia and who had been admitted to the Adolescent Clinic were eligible for recruitment. Discharge diagnoses according to DSM-IV criteria were based on longitudinal clinical assessment. Sources of information were patients, parents, and other significant others (Longitudinal Expert Assessment of Diagnosis procedure) (24). Exclusion criteria were neurological or endocrine disease, mental retardation, the use of adjunctive medications such as mood stabilizers or antidepressants, a history of treatment with clozapine, a history of unresponsiveness to haloperidol or olanzapine, and intramuscular antipsychotic treatment within the last year.
Previous medication was diminished during 1 week and then stopped. Olanzapine, 7.5 mg once daily, or haloperidol, 2.5 mg once daily, was assigned according to a randomized block design. These doses were chosen for their expected D2 receptor occupancy of 65% (23, 25–27) in patients with previous exposure to antipsychotic medication. This rate of D2 receptor occupancy is thought to be a threshold above which a satisfactory antipsychotic response can be expected (7–9).
Duration of the double-blind phase was 6 weeks. Only the benzodiazepine oxazepam was allowed as adjunctive medication. Treating physicians made the decision regarding the use of oxazepam on the basis of clinical considerations. The use of oxazepam was indicated for anxiety, agitation, and insomnia.
Subjective experience, extrapyramidal side effects, and psychopathology were assessed at baseline and after 6 weeks. The subjective experience of the patients was measured with the Subjective Well-Being Under Neuroleptics Scale (17, 28). This 38-item, self-rating scale assesses through 6-point Likert-type responses the subjective experience over the preceding 7 days. Naber (17) found a five-factor solution of the scale: emotional regulation, self-control, mental functioning, social integration, and physical functioning. (Sample questions: "I am full of energy and life." "I feel very comfortable in my body.") In a study of patients with schizophrenia (N=216), the Subjective Well-Being Under Neuroleptics Scale showed good practicability, reliability, and sensitivity (17). We recently replicated these results: 97% of patients with recent-onset schizophrenia (N=105) were capable of reproducing their subjective experience in a consistent way before as well as after medication stabilization. The internal consistency was high (Cronbach’s alpha=0.95). The test-retest reliability appeared to be high if medication was not changed, and the Subjective Well-Being Under Neuroleptics Scale appeared to be sensitive for changes in medication or dose (28). Severity of patients’ psychopathology was assessed with the Positive and Negative Syndrome Scale, the Montgomery-Åsberg Depression Rating Scale, and the Clinical Global Impression (CGI) scale. Extrapyramidal symptoms and akathisia were assessed with the Simpson-Angus Rating Scale and the Barnes Rating Scale for Drug-Induced Akathisia, respectively.
SPECT imaging to determine the levels of D2 occupancy was performed with an SME 810 brain-dedicated SPECT multidetector system. The transaxial resolution of this camera is 7.6 mm full width at half maximum of a line source in air, and the axial resolution is 13.5 mm full width at half maximum. All subjects received oral potassium iodide to block thyroid uptake of free radioactive iodide. [123I]IBZM (specific activity of >185 MBq/nmol; radiochemical purity of >95%) was injected intravenously at an approximate dose of 110 MBq. 123I labeling of IBZM was performed by Amersham Cygne (Technical University Eindhoven, the Netherlands). SPECT image acquisition was always performed 2 hours postinjection (29). SPECT imaging was performed after 6 weeks of continuous treatment with the study medication. All scans were carried out 12 to 13 hours after the administration of the last dose of the study medication. Specifications and imaging procedures have been described elsewhere (30).
Semiquantification of [123I]IBZM binding was performed by placing a template of fixed regions of interest over the striatum and occipital cortex (30). Specific striatal binding was defined as striatal binding divided by occipital binding (30). All analyses were performed blind to the clinical data.
A group of nine healthy comparison subjects meeting the following criteria—no current or past psychiatric, neurological, or medical illness; no use of medication; age 17–28 years—was included to obtain baseline dopamine receptor densities. Their mean [123I]IBZM binding ratio was 1.92 (SD=0.08), which was used to calculate the percentage of occupancy by medication as follows: (ratio striatum/occipital region)–1/(1.92–1)×(–100)+100 .
By computing the mean absolute difference between the positively and the comparative negatively formulated items of the Subjective Well-Being Under Neuroleptics Scale, we estimated the reliability of the patient’s assessment with the Subjective Well-Being Under Neuroleptics Scale before further analyses were done. If the mean absolute difference was 2 or higher, the patient’s assessment was considered to be not consistent, and these patients were excluded from analysis relating to subjective experience (17, 28).
Differences between groups were tested with a t test, and differences within patients between baseline and endpoint were tested with a paired t test. All tests were two-tailed.
Twenty-four patients with schizophrenia according to DSM-IV criteria were included. Subjects’ characteristics, including characteristics of the illness and Subjective Well-Being Under Neuroleptics Scale total score, psychopathology ratings, and percentage of D2 receptor occupancy at endpoint, are shown in t1. Characteristics of the participants (sex, diagnosis, number of episodes, duration of psychosis, and prior treatment) treated with olanzapine did not significantly differ at baseline from the characteristics of participants treated with haloperidol.
Three patients treated with olanzapine and one patient treated with haloperidol dropped out. Reasons for dropping out were patient decision (N=2, both olanzapine) and increase of psychosis (N=2, one patient treated with haloperidol, one patient treated with olanzapine). There were no significant differences between patients who dropped out and study completers with respect to diagnosis, number of psychotic episodes, duration of psychotic symptoms, psychopathology, or subjective experience ratings.
Mean absolute difference between the positively and the comparative negatively formulated items of the Subjective Well-Being Under Neuroleptics Scale was 0.58 (SD=0.34). Only one patient (haloperidol group) had a mean absolute difference of 2 or higher (2.9). This assessment was considered to be not consistent and was excluded from analysis relating to subjective experience.
Subjective Experience and D2 Receptor Occupancy
Mean baseline and change at endpoint scores for measures of the Subjective Well-Being Under Neuroleptics Scale did not differ significantly between participants treated with haloperidol or olanzapine except for a significant difference in the change at endpoint score for the self-control subscale in favor of haloperidol. Total score and four of the five subscale scores of the Subjective Well-Being Under Neuroleptics Scale improved significantly from baseline to endpoint in the patients treated with haloperidol (t2).
As seen in t3 and F1, patients with D2 receptor occupancy levels between 60% and 70% (N=6) had a significantly higher total score on the Subjective Well-Being Under Neuroleptics Scale than did subjects with D2 receptor occupancy levels higher than 70% (N=5). In addition, the total score on the Subjective Well-Being Under Neuroleptics Scale for subjects with D2 receptor occupancy levels between 60% and 70% was significantly higher when compared with that of subjects with D2 receptor occupancy levels either lower than 60% or higher than 70% (N=13; mean=151.2, SD=24.7) (t=3.3, df=17, p=0.004). The mean scores on all the subscales of the Subjective Well-Being Under Neuroleptics Scale were significantly higher in subjects with D2 receptor occupancy levels between 60% to 70% than in those with D2 receptor occupancy levels higher than 70% (t3).
Subjects treated with haloperidol accounted for most of the difference in scores on the Subjective Well-Being Under Neuroleptics Scale between participants with D2 receptor occupancy levels of 60%–70% (mean=211.3, SD=11.1, N=3) and those with D2 receptor occupancy levels of 71% or higher (mean=140.6, SD=10.8, N=5) (t=8.9, df=6, p<0.001).
The mean D2 receptor occupancy level in patients treated with olanzapine, 7.5 mg/day, was 51.0% (SD=11.1%, range=36%–67%). This was significantly lower than the mean D2 receptor occupancy level of 65.5% (SD=8.7%, range=45%–75%) seen in patients treated with haloperidol, 2.5 mg/day (t=3.3, df=18, p=0.004) (F2).
D2 receptor occupancy level was not related to any baseline variable (age, weight, nicotine use, coffee use, or severity of illness).
Psychopathology and Extrapyramidal Side Effects
Significant improvement in CGI scores was seen in the patients treated with olanzapine (mean change at endpoint=–1.3 [SD=1.0]; t=3.42, df=7, p<0.05) and in the patients treated with haloperidol (mean change at endpoint=–0.8 [SD=0.9]; t=?, df=10, p<0.05). For the patients treated with olanzapine and those treated with haloperidol, nonsignificant improvement was seen for total scores on the Positive and Negative Syndrome Scale (mean change at endpoint=–7.2 [SD=31.9] and –11.4 [SD=19.5], respectively) and the Montgomery-Åsberg Depression Rating Scale (mean change at endpoint=–2.8 [SD=12.1] and –1.2 [SD=3.6]). No significant differences in mean baseline (data not shown) and change at endpoint scores were found between participants treated with haloperidol and olanzapine.
We found no relationship between level of D2 receptor occupancy and response defined as difference in Clinical Global Impression, Positive and Negative Syndrome Scale, and Montgomery-Åsberg Depression Rating Scale scores between baseline and endpoint.
No relationship was found at endpoint between D2 receptor occupancy level and Positive and Negative Syndrome Scale total or subscale scores or Montgomery-Åsberg Depression Rating Scale scores.
At baseline, participants treated with haloperidol experienced more akathisia (mean=0.6, SD=0.8) than participants treated with olanzapine (mean=0.0, SD=0.0) (t=2.35, df=18, p<0.05) as measured with the Barnes Akathisia Scale. Change at endpoint did not differ significantly between groups. Within the haloperidol group, mean score on the Simpson-Angus Scale improved significantly (mean change at endpoint=–3.9, SD=5.0) (t=2.47, df=9, p<0.05).
At endpoint, akathisia and parkinsonism were generally absent. No relationship between the D2 receptor occupancy percentage and Barnes Akathisia Scale scores or Simpson-Angus Scale scores was found at endpoint.
The results of this study offer evidence that D2 receptor occupancy between 60% and 70% is optimal for subjective experience of patients with recent-onset schizophrenia. Exceeding a level of 70% D2 receptor occupancy may substantially increase the risk of inducing worse subjective experience and noncompliance and therefore psychotic relapse.
The very low incidence and severity of extrapyramidal side effects we found is in line with the D2 receptor occupancy threshold of 78% for development of extrapyramidal side effects. We were not able to replicate the relationship between D2 receptor occupancy and response as found by Kapur et al. (10). None of our patients had D2 receptor occupancy levels that exceeded 75%, whereas five of the 10 patients with CGI scores that were much or very much improved in the study of Kapur et al. had D2 receptor occupancy levels that exceeded 75%. This leaves clinicians with an almost insurmountable problem: the level of D2 receptor occupancy needed for response in some patients exceeds the level above which the risk of negative influence on subjective experience is probably high. However, since the duration of treatment in the study by Kapur et al. (10) was only 2 weeks, a D2 receptor occupancy level lower than 75% for a longer period of time may result in response in most patients. Moreover, most of our patients had been treated with antipsychotic medication before study entry, thereby increasing the risk of a higher percentage of patients with treatment resistance than in a neuroleptic-naive population. This may also explain why we could not replicate the relationship between D2 receptor occupancy and response.
We found a lower mean D2 receptor occupancy (65.5%) in patients treated with haloperidol, 2.5 mg/day, than did Kapur et al. (10), whose nine first-episode patients also received haloperidol, 2.5 mg/day (mean D2 receptor occupancy=75%). This difference could reflect the need for treatment with very low doses for neuroleptic-naive patients compared with patients who have been treated with antipsychotic medication before. As far as we know, it is not clear when the need for higher doses emerges during treatment with antipsychotic medication.
The substantial interindividual variation in D2 receptor occupancy at fixed amounts of low-dose haloperidol and olanzapine in this homogeneous group of subjects was in concordance with the wide interindividual variation at a given dose of haloperidol found in other studies (7, 10).
Mean D2 receptor occupancy in patients treated with olanzapine, 7.5 mg/day, was significantly lower than mean D2 occupancy reached during treatment with haloperidol, 2.5 mg/day. The mean D2 occupancy found with olanzapine, 7.5 mg/day, is in line with the saturation hyperbola as earlier described (23). In a study on the D2 receptor occupancy of olanzapine, more interindividual variation seemed to occur in patients receiving 5 mg/day of olanzapine than in patients receiving doses of 10 mg/day or higher (23). In our study, more than half of the patients treated with olanzapine, 7.5 mg/day, reached D2 receptor occupancy of 45% or less; however, one-third of these patients showed an occupancy of 60% or more.
Olanzapine, 7.5 mg/day, did not show superior subjective response compared with haloperidol, 2.5 mg/day, in patients with recent-onset schizophrenia. In fact, subjective experience improved significantly only during treatment with haloperidol. This is an unexpected finding given the advantages in subjective experience found during treatment with olanzapine in former studies. Inordinately high levels of D2 receptor occupancy in the typical comparison group have confounded these studies. However, our study may be biased by the relatively low D2 receptor occupancy associated with the treatment of olanzapine, 7.5 mg/day. We therefore cannot rule out the possibility that olanzapine in a dose that gives rise to D2 receptor occupancy beyond 70% may still result in superior subjective well-being. The decline in subjective well-being with an occupancy of striatal D2 receptors above 70% may be an issue only relevant for haloperidol. However, future studies are needed to test this hypothesis.
Fast dissociation from the D2 receptor could make an antipsychotic medication more accommodating of physiological dopamine transmission, permitting antipsychotic effect without major side effects (25). This proposed mechanism could also account for superior subjective experience associated with atypical antipsychotic drugs. Since we found no evidence for superior subjective experience associated with low-dose olanzapine compared with low-dose haloperidol, we suppose that besides faster dissociation, the level of D2 occupancy is also important for the influence of antipsychotic drugs on subjective experience. Moreover, subjective experience is probably influenced not only by effects on modulation of the D2 neurotransmission but effects on receptors other than D2 may also account for worse subjective experience. Probably an ideal antipsychotic medication binds selectively to the D2 receptor and dissociates relatively fast from it, without causing enduring occupancy beyond about 70%.
Our study has several limitations. First, mesolimbic D2 receptors might be more relevant to subjective experience than striatal D2 receptors. Since there are no indications for differences in striatal and extrastriatal D2 occupancy in patients treated with haloperidol, we think that striatal occupancy is a reasonable surrogate of mesolimbic D2 occupancy, at least in the case of haloperidol (31, 32). Second, most of the patients included used antipsychotic medication before study entry, probably altering their D2 binding capacity. We think that the study duration of 6 weeks makes confounding effects on subjective experience measures caused by former medication less probable. Third, one-half of the patients used oxazepam as adjunctive medication. However, occupancy in a PET study of lorazepam has been shown to have no influence on striatal [11C]raclopride binding (33). Fourth, the 7.5 mg/day dose of olanzapine was too low for most (not all) patients with recent-onset schizophrenia previously treated with antipsychotic medication. Comparison of haloperidol, 2.5 mg/day, with olanzapine, 10 mg/day, might give ideal D2 occupancy matching. Future comparison of subjective experience of typical and atypical antipsychotic medication in doses that lead to the same range of D2 receptor occupancy is needed. Fifth, only one out of 24 subjects was female. Finally, in the present study plasma levels of study medication were not measured. However, compliance with study medication was guaranteed, since nursing staff supervised intake of medication.
In summary, we found evidence that a level of D2 receptor occupancy between 60% and 70% is optimal for subjective experience of patients with recent-onset schizophrenia. We found substantial interindividual variation in D2 receptor occupancy at fixed amounts of low-dose haloperidol and olanzapine. Olanzapine, 7.5 mg/day, showed no superior subjective response over haloperidol, 2.5 mg/day. From the perspective of subjective experience of patients, haloperidol needs to be individually titrated in the very low dose range (probably 1.5–3 mg/day) to reach optimal occupancy, and olanzapine needs to be dosed higher than 7.5 mg/day for most patients with recent-onset schizophrenia previously treated with antipsychotic medication.
Received Sept. 24, 2001; revision received June 5, 2002; accepted Aug. 1, 2002. From the Academic Medical Center, University of Amsterdam Department of Psychiatry. Address reprint requests to Dr. de Haan, Academic Medical Center, University of Amsterdam Department of Psychiatry, Postbox 22700, 1100 DE Amsterdam, the Netherlands; firstname.lastname@example.org (e-mail). Supported by grant 28-1241-2 from the Dutch Health Research and Development Council.
Relationship Between Subjective Well-Being and D2 Receptor Occupancy Level in Patients With Recent-Onset Schizophrenia After 6 Weeks of Double-Blind Treatment With Either Olanzapine (7.5 mg/day) or Haloperidol (2.5 mg/day)
aHigher scores reflect better subjective experience (total score range=38–228).
D2 Receptor Occupancy Levels in Patients With Recent-Onset Schizophrenia After 6 Weeks of Double-Blind Treatment With Either Olanzapine (7.5 mg/day) or Haloperidol (2.5 mg/day)