Noboru Sakumoto, MD, Department of Neuropsychiatry, Faculty of Medicine, University of the Ryukyus, Okinawa 903–0215, Japan. Email: firstname.lastname@example.org
Abstract Previous reports have shown that both A1 allele carriers of TaqI A and Del allele non-carriers of −141C Ins/Del for dopamine D2 receptor (DRD2) gene polymorphisms have a better antipsychotic drug response. The present study aimed to examine the validity of a combination of these two DRD2 polymorphisms as predictors for response to DRD2 antagonists. The subjects consisted of 49 acutely exacerbated inpatients with schizophrenia treated with bromperidol (30 cases, 6–18 mg/day) or nemonapride (19 cases, 18 mg/day) for 3 weeks. Brief Psychiatric Rating Scale and Udvalg for Kliniske Undersøgelser side-effects rating scale were used for clinical assessments. DRD2 genotypes were determined using a polymerase chain reaction method. In the overall 49 subjects, combined DRD2 polymorphisms weakly predicted the response to DRD2 antagonists (Fisher exact test, P = 0.049), that is, good response in A1(+) or Del(–) subjects and poor response in A1(–) plus Del(+) subjects. In the former subjects, non-responders with A1(+) or Del(–) showed higher scores of psychic, extrapyramidal and total side-effects. At therapeutic doses (6–8 mg/day haloperidol equivalent dose) in 30 subjects, the predictability of response was greatly increased (Fisher exact test, P < 0.0045) with higher positive and negative predictive values (78.3% and 85.7%, respectively). These findings suggest that combined DRD2 polymorphisms can be used as a pretreatment marker for response to DRD2 antagonists at therapeutic doses, and that A1(+) or Del(–) subjects are highly sensitive to DRD2 antagonists, expressed as either treatment responders or non-responders vulnerable to extrapyramidal symptoms.
Most typical antipsychotic drugs have a strong antagonistic property for dopamine D2 receptor (DRD2). Accordingly, central nervous dopaminergic hyperactivity has been long regarded as an underlying important pathogenesis of schizophrenia.1 Studies using positron emission tomography (PET) have demonstrated that a threshold DRD2 occupancy (65–75%) is required to obtain an antipsychotic effect,2 suggesting that DRD2 occupancy during neuroleptic treatment is a determinant for neuroleptic response.
The representative DRD2 antagonist, haloperidol, has its therapeutic concentration range at 5.6 to 16.9 ng/mL.3 It has been suggested that plasma haloperidol concentrations accurately reflect DRD2 occupancy.4 Therefore, dose setting based on plasma drug monitoring may be to some extent useful to obtain optimal antipsychotic effects during haloperidol treatment from pharmacokinetic aspects.3,5
Besides, it is also possible that antipsychotic response differs interindividually because of variations in sensitivity at the site of drug action from the pharmacodynamic point of view. In particular, genetic factors that can affect the density and function of DRD2 may be responsible for the interindividual variability in neuroleptic response.
The human DRD2 gene is located at 11q22-q23. It contains a TaqI restriction fragment length polymorphism (TaqI A polymorphism), creating A1 and A2 alleles.6 It has been reported that A1 allele carriers shows lowered DRD2 density in the postmortem brain7,8 and decreased binding potential for DRD2 in an in vivo PET.9 In addition, the A1 allele has been suggested to diminish dopaminergic activity in the central nervous system, for example, prolonged P300 latency and reduced brain glucose metabolism.7 Therefore, TaqI A polymorphism may functionally affect the central nervous dopaminergic system.
Meanwhile, Arinami et al.10 found a single base pair cytosine insertion/deletion polymorphism at position −141 (−141C Ins/Del polymorphism) in the 5′ flanking region of the DRD2 gene, yielding the Ins and Del alleles. An in vitro study10 showed that the Del allele of the −141C Ins/Del DRD2 polymorphism was directly related to DRD2 expression. In vivo studies using PET11 showed that decreased striatal DRD2 density was observed in subjects who had no Del allele. Accordingly, the −141C Ins/Del polymorphism may be directly responsible for the regulation of DRD2 expression and DRD2 function.
These findings have prompted several studies focusing on associations between DRD2 polymorphisms and antipsychotic drug response. It has been shown that the A1 allele of TaqI A polymorphism is associated with greater improvement in positive symptoms of schizophrenia during treatment with DRD2 antagonists.12,13 Meanwhile, the Del non-carriers of −141C Ins/Del polymorphism have shown greater reduction in anxiety-depression symptoms during DRD2 antagonist treatments14 and better response to sedative and anxiolytic antipsychotics such as chlorpromazine15 and clozapine.16 Furthermore, the combined genotyping of these DRD2 polymorphisms revealed treatment-resistance to DRD2 antagonists in A1(–)/Del(+) subjects.17 However, there has been no basic data regarding the accuracy for the prediction of antipsychotic drug response by using these genetic markers.
The DRD2 antagonist bromperidol is a close structural analog of haloperidol, while nemonapride is a substituted benzamide antipsychotic drug developed in Japan. Both drugs have potent antagonistic effects for dopamine D2-like receptors in contrast to very weak affinity for other receptors such as dopamine D1, 5-HT2, norepinephrine, and acetylcholine.18,19 The authors previous studies20,21 showed that these drugs considerably increased prolactin concentrations in patients with schizophrenia, providing in vivo evidence for the strong DRD2 blockade property. Therefore, these two drugs have similar pharmacological properties despite their chemical structural differences. The two drugs selectivity and potency for DRD2 antagonism make bromperidol and nemonapride suitable probe drugs to investigate the effects of the above-mentioned two DRD2 gene polymorphisms on efficiency of DRD2 antagonism.
Therefore, the present study aimed to examine the validity of a combination of these two DRD2 polymorphisms as predictive markers for therapeutic response to antidopaminergic agents. The dose effects of the drugs were also taken into consideration when analyzing the predictability of antipsychotic response by the DRD2 polymorphisms.
Subjects and drug treatments
In total, 61 subjects were initially enrolled in this study. In six cases, blood samples were used up after repeated confirmative determination of TaqI A genotypes. Another three subjects dropped out of the 3-week treatment in each DRD2 antagonist. So, subjects in the present study finally consisted of 49 acutely exacerbated inpatients with schizophrenia (26 males, 23 females), who fulfilled the 4th edition, Diagnostic and Statistical Manual criteria22 for schizophrenia. None had received any medication for at least 1 month before this study. The study protocol was approved by the Ethics Committee of Hirosaki University Hospital, University of the Ryukyus, and written informed consent to participate in the study was obtained from the patients and their families.
The subjects consisted of 30 cases treated with bromperidol (6 mg/day in 11 cases, 12 mg/day in 11 cases and 18 mg/day in 8 cases) and 19 cases treated with 18 mg/day of nemonapride for 3 weeks. No other drugs were given except biperiden 3–6 mg/day for extrapyramidal side-effects (n = 28), flunitrazepam 2–8 mg/day for insomnia (n = 38) and sennoside 12–48 mg/day for constipation (n = 9). The drug compliance was confirmed by nursing staff. The mean ± standard deviation (SD) of age (years), bodyweight (kg) and duration of illness (months) were 36.1 ± 11.8, 59.7 ± 11.8 and 106.8 ± 96.7, respectively.
Haloperidol equivalent doses of bromperidol23 and nemonapride24 were 1 and 2.25, respectively. Accordingly, 30 cases treated with 6 mg/day of bromperidol and 18 mg/day of nemonapride (8 mg/day as a haloperidol equivalent dose) were regarded as subjects at therapeutic dosages, while the remaining 19 subjects treated with 12 and 18 mg/day of bromperidol were supratherapeutic cases.
Clinical symptoms before and 3 weeks after the treatment were assessed by Brief Psychiatric Rating Scale (BPRS).25 This scale has five scale steps (0–4) for each item according to the severity of symptoms. Symptom reduction after the treatment was expressed by percentage improvement, that is, amelioration scores after 3 weeks/pretreatment score ×100. A responder was defined as a patient who showed more than 50% reduction in total BPRS scores after 3-week treatment.
Adverse effects were also scored before and 3 weeks after the treatment using the Udvalg for Kliniske Undersøgelser (UKU) side-effects rating scale26 to assess their effects on therapeutic response. A pretreatment score was subtracted from one obtained during treatment and was recorded as a drug-associated adverse effects.
The clinical assessments were performed by well-trained psychiatrists who were unaware of drug regimens and genotypes. Therefore, the clinical status of patients was rated irrespective of the determination of the genotypes. Interrater reliability was found to be high (r = 0.895 in BPRS scores; r = 0.945 in UKU side-effects rating scale scores) for two psychiatrists who interviewed 20 inpatients with schizophrenia in a training session before the initiation of the study.
A total of 10 mL of blood was obtained from each subject, and DNA was isolated from peripheral leukocytes by a guanidinium isothiocyanate method. The TaqI A genotypes (A1 and A2 alleles) and −141C Ins/Del genotypes (Ins and Del alleles) were determined by the polymerase chain reaction methods as previously described by Grandy et al.27 and Arinami et al.,10 respectively. The determination of the genotypes was completed by researchers who were blind to the results of clinical assessments.
Measurements of plasma drug concentrations
Blood samplings for measurements of the drugs were performed just before the morning dose on the same day as the clinical symptoms were rated. Plasma concentrations of bromperidol were measured in duplicate using the high-performance liquid chromatography method of Hikida et al.28 Plasma concentrations of nemonapride and an active metabolite, desmethylnemonapride, which has the one-third potency of nemonapride in the inhibition of [3H] nemonapride binding,18 were measured in duplicate using the high-performance liquid chromatography method of Nagasaki et al.29 Plasma concentrations of nemonapride plus desmethylnemonapride were regarded as the active moiety of nemonapride.
Subjects were initially classified into four subgroups according to the presence or absence of the A1 allele of TaqI A polymorphism (A1 carriers or A1 non-carriers) and Del allele of the −141C Ins/Del polymorphism (Del carriers or Del non-carriers), that is, (1) A1(+)/Del(–), (2) A1(+)/Del(+), (3) A1(–)/Del(–), and (4) A1(–)/Del(+). Thereafter, according to the different effects of genotypes of TaqI A (A1 carriers vs A1 non-carriers) or −141C Ins/Del (Del carriers vs Del non-carriers) DRD2 gene polymorphisms on DRD2 density as reported in previous studies,7,9,11 the former three groups were integrated into one group (subjects with lower DRD2 density), that is, A1 carriers or Del non-carriers (A1(+) or Del(–) group), and the remaining cases were A1 non-carriers with Del allele (A1(–) plus Del(+) group).
Categorical data (gender, type of drugs and concomitant drugs) and the association between DRD2 genotype combinations (A1(+) or Del(–) group vs A1(–) plus Del(+) group) and treatment outcome (response/non-response), were analyzed using χ2 test or Fisher's exact test. Comparisons of age, bodyweight, duration of illness, pretreatment BPRS scores, total and subgrouped UKU side-effects rating scale scores after 3 weeks were performed by anova followed by Tukey test as a post-hoc analysis. Spearman rank test was used to assess the relationship between haloperidol equivalent doses or plasma concentrations of drugs and percentage improvement in BPRS symptoms and UKU scores after 3 weeks.
A two-tailed P-value of 0.05 or less was regarded as statistically significant. SPSS 11.0.1 J for Windows (SPSS Japan Inc., Tokyo, Japan) was used for these statistical analyses.
Genotype distribution of the subjects was as follows, A1/A1 (n = 5), A1/A2 (n = 25) and A2/A2 (n = 19) for TaqI A polymorphism, and Ins/Ins (n = 35) and Ins/Del (n = 14) for −141C Ins/Del polymorphism. There was no linkage disequilibrium between TaqI A and −141C Ins/Del DRD2 gene polymorphisms (delta value = 0.034, P = 0.098, n = 49). The overall subjects were divided into the two groups such as A1(+) or Del(–) group (n = 40) and A1(–) plus Del(+) group (n = 9) based on the differences in DRD2 density as demonstrated in previous reports.7,9,11
No significant differences were found in age, gender distribution, bodyweight, duration of illness, type of neuroleptic drugs, concomitant drugs and baseline BPRS scores between the two groups. Neither the haloperidol equivalent doses nor plasma concentrations of bromperidol or nemonapride correlated with the percentage improvement in BPRS symptoms or UKU side-effects rating scale scores after 3 weeks.
In the overall 49 subjects, a weak correlation was found between DRD2 genotype combinations and treatment outcome (P = 0.049; Fig. 1). The sensitivity and efficiency for prediction of therapeutic response were 90.6% and 71.4%, respectively, although the specificity was relatively low (35.3%). Positive predictive value (PPV) and negative predictive value (NPV) were 72.5% and 66.7%, respectively (Fig. 1). In 30 subjects at therapeutic doses (6 and 8 mg/day of haloperidol equivalent doses), correlation between DRD2 genotype combinations and treatment response became much stronger (P = 0.0045; Fig. 1). All of the parameters for prediction of response were elevated, that is, sensitivity 94.7%, specificity 54.5%, efficiency 80.0%, PPV 78.3% and NPV 85.7% (Fig. 1). In contrast, in 19 subjects at supratherapeutic doses (12 and 18 mg/day of haloperidol equivalent doses), no correlation was found between DRD2 genotype combinations and treatment outcome (Fig. 1).
Regarding side-effects, pseudo-positive subjects (non-responders with A1(+) or Del(–)) had significantly higher scores in total (P = 0.014), psychic (P = 0.025) and extrapyramidal (P = 0.034) side-effects scores than responders with A1(+) or Del(–) (Table 1). No differences were found in type and doses of drugs or plasma drug concentrations between these two groups.
Table 1. Udvalg for Kliniske Undersøgelser side-effects rating scale scores after 3-week treatment with dopamine receptor D2 antagonists in relation to therapeutic response and dopamine receptor D2 genotypes
anova followed by Tukey test as a post-hoc analysis.
Significantly different from responders with A1(+) or Del(–) ( *P = 0.014;**P = 0.025;***P = 0.034).
Although the second-generation antipsychotic drugs are currently used as the first-line treatment of schizophrenia, traditional antipsychotics still have a role in the treatment of schizophrenia because of their established therapeutic effects30 and lower costs in the treatment of schizophrenia.
Plasma drug monitoring is clinically used for optimal dose-settings of classical antipsychotics.3,5 However, these pharmacokinetic markers are only partially useful in clinical situations, and time lag for trial and error in dose setting cannot be avoided. Pretreatment prediction of therapeutic response to conventional antipsychotic drugs, for example, selective dopamine antagonists, seems important in selecting the second-line options for some treatment-resistant cases.
Relatively smaller doses of haloperidol (2–5 mg/day) are expected to induce high DRD2 occupancy (60–80%),31 which seems enough to obtain over threshold DRD2 occupancy for antipsychotic effect (65–75%).2 However, in reality, there exist a number of poor responders to haloperidol even at the above-mentioned therapeutic dose ranges. Some pharmacodynamic insensitiveness at drug action sites might be involved in the mechanisms of poor neuroleptic response.
Both TaqI A and −141C Ins/Del DRD2 gene polymorphisms are associated with modification of DRD2 density and function.7–9,11 Therefore, these two DRD2 polymorphisms can alter therapeutic response to antidopaminergic agents. In fact, previous studies have revealed that TaqI A polymorphism affects antipsychotic effects of DRD2 antagonists12,13 while −141C Ins/Del polymorphism does their anxiolytic and antidepressive effects.14 Therefore, the next necessary step appears to examine the validity of a combination of these two DRD2 gene polymorphisms for prediction of total response to DRD2 antagonists.
The present results clearly demonstrated good response to DRD2 antagonists in A1(+) or Del(–) group and poor response in A1(–) plus Del(+) group regarding the overall efficacy (Fig. 1). High sensitivity and efficiency justified usefulness of determining combined DRD2 polymorphisms for prediction of therapeutic response to DRD2 antagonists. Especially in 30 subjects at therapeutic doses (6 and 8 mg/day of haloperidol equivalent doses), the predictability of response was prominently improved (Fig. 1). Furthermore, a weak correlation was found between DRD2 genotype combinations and treatment outcome in each DRD2 antagonist (data not shown). These results suggest possible clinical implication of combined DRD2 polymorphisms as routine predictive markers in subjects who are going to receive therapeutic doses of DRD2 antagonists.
In contrast, pseudo-positive (non-responders with A1(+) or Del(–)) and pseudo-negative (responders with A1(–) plus Del(+)) cases confounded the prediction of response at supratherapeutic doses (12 and 18 mg/day of haloperidol equivalent doses). In the former cases, greater side-effects including extrapyramidal symptoms (as discussed later) might have been at least partly attributable to treatment failure at higher doses of DRD2 antagonists, as previously suggested by Van Putten et al.32 Meanwhile, higher doses may also have masked poor responsiveness to DRD2 antagonists in the latter cases.
Scores in total, psychic and extrapyramidal side-effects were the highest in non-responders with A1(+) or Del(–) (Table 1). Although good therapeutic response was not achieved in this group, instead, the susceptibility to extrapyramidal symptoms may be another reflection of efficient DRD2 blockade. Therefore, A1(+) or Del(–) subjects are highly sensitive to DRD2 antagonists, expressed as either treatment responders or non-responders vulnerable to extrapyramidal symptoms.
The present data should be carefully interpreted because of the small number of subjects in each subgroup. Therefore, it is noted that clinical usefulness of DRD2 polymorphisms for prediction of therapeutic response to therapeutic doses of antidopaminergic agents should be confirmed by further research with a larger number of subjects.
The present study suggests that the combined DRD2 polymorphisms can be used as a pretreatment marker for response to DRD2 antagonists at therapeutic dosages, and that A1(+) or Del(–) subjects are highly sensitive to DRD2 antagonists, expressed as either treatment responders or non-responders vulnerable to extrapyramidal symptoms.
This study was supported by Grants in Aid from the Japanese Ministry of Education, Culture, Sports, Science and Technology (12670919 and 16790696).