多変量分散分析を用いた抗精神病薬の治療効果，副作用の予測に関する臨床薬理学的研究 Ｂｒｏｍｐｅｒｉｄｏｌについて

Abstract

In this study, we attempted to identify significant baseline factors for the clinical efficacy and side effects of bromperidol (BPD) using multivariate analysis of variance (MANOVA).Subjects for this study consisted of 61 psychiatric patients, almost all of whom were schizophrenic. BPD was administered to the patients for 6 weeks by a non-fixed open method (dose and duration not fixed).Psychiatric symptoms and side effects were assessed using the Brief Psychiatric Rating Scale (BPRS) and UKU Side Effects Rating Scale at the time of initiation and after 6 weeks of the therapy. The background baseline characteristics of the patients were also evaluated: sex, age, height, weight, hereditary disposition of psychiatric disorders, diagnosis, duration of illness, number of recurrent episodes, acute exacerbation, responses to previous antipsychotics, daily dosage of BPD, concomitant antiparkinson drugs, and the plasma levels of BPD and prolactin.For the statistical analysis, to identify these baseline factors closely associated with a good prognosis and/or side effects, a multivariate analysis of variance was applied, using the Statistical Analysis System (SAS).A higher improvement rate of BPRS total score was obtained in the group with good response to previous antipsychotics, low daily dosage, high plasma levels of BPD, and low initial score of “anergia”, and in those without an acute exacerbation (the chronic course).The chronic course was associated with a significantly higher improvement rate of “hostile-suspiciousness”. Good response to previous antipsychotics was related to “anergia”, “thought-disturbance”, “activation” and “hostile-suspiciousness”. Low daily dosage was found to contribute to the improved “anxiety-depression”, “thought-disturbance”, “activation”. Higher plasma levels of BPD were related to “anxiety-depression”. A higher “anergia” subscale score was especially associated with the lower improvement rate of each subscale except “anxiety-depression” of BPRS. Regarding side effects, there were no identifiable factors contributable to side effects.These methods of multivariate analysis seem to be quite useful in the field of clinical psychopharmacology for the analysis of the great volume of data consisting of qualitative and quantitative factors, and also to identify those factors contributing to the effcacy and side effects.