Major Depressive Disorder (MDD) may be composed of some symptom clusters with distinct neurochemical disturbances, suggesting the importance of the factor analysis of depressive symptoms; however, the results of previous studies using the Montgomery-Asberg Depression Rating Scale (MADRS) have been inconsistent. In the present study, factor analysis of the MADRS was performed in 132 Japanese patients (range 23-74 years, mean 47.6 years) with MDD without any psychiatric comorbidity. The principal component analysis with Varimax rotation identified three factors, accounting for 61% of the total variance: The first factor, labeled dysphoria, included pessimistic thoughts, suicidal thoughts, and reported sadness; the second factor, labeled retardation, included lassitude, inability to feel, apparent sadness, and concentration difficulties; and the third factor, labeled vegetative symptoms, included reduced sleep, reduced appetite, and inner tension. The score of the vegetative factor showed a significant positive correlation with age and was significantly higher in females than in males. This study suggests that the symptoms of MDD, as assessed by the MADRS, cluster into three factors (dysphoria, retardation, and vegetative symptoms).

The present study suggests that there is a therapeutic threshold for the Css of FLV and probably also for the Css of FLA, and the Css of FLV+FLA above 180 ng/ml best predicts a good therapeutic response.

The effects of the cytochrome P450 (CYP) 2D6 genotype and cigarette smoking on the steady-state plasma concentrations (C(ss)) of fluvoxamine (FLV) and its demethylated metabolite fluvoxamino acid (FLA) were studied in 49 Japanese depressed patients receiving FLV 200 mg/d. The C(ss) of FLV and FLA were measured by HPLC, and the wild-type allele (*1) and two mutated alleles causing absent (*5) or decreased (*10) CYP 2D6 activity were identified by PCR methods. The patients were divided into three genotype groups by the number of mutated alleles: 12 cases with no (*1/*1), 27 cases with one (*1/*5 and *1/*10), and 10 cases with two (*5/*10 and *10/*10) mutated alleles. The means +/- SD of the C(ss) of FLV and FLA and the FLA/FLV ratio of all patients were 169.1 +/- 147.5 ng/mL, 83.9 +/- 52.7 ng/mL, and 0.71 +/- 0.50, respectively. The C(ss) of FLV and FLA were not significantly different among the three genotype groups. However, the FLA/FLV ratio was significantly lower in the patients with one (P < 0.05) and two (P < 0.01) mutated alleles than in those with no mutated allele. There was no significant difference between nonsmokers (n = 34) and smokers (n = 15) in these values. In the stepwise multiple regression, the C(ss) of FLA (P < 0.05) and FLA/FLV ratio (P < 0.001) showed significant negative correlations with the number of mutated alleles, and the FLA/FLV ratio was significantly (P < 0.05) lower in women than in men. The present study suggests that the CYP 2D6 genotype and cigarette smoking have no major impact on the C(ss) of FLV and FLA, though CYP 2D6 is involved in the demethylation of FLV.

Several classical antipsychotic drugs, i.e., chlorpromazine, haloperidol, perphenazine, thioridazine and zuclopenthixol; and some new neuroleptic drugs, i.e., risperidone and sertindole, are metabolized predominantly by cytochrome P450 (CYP) 2D6. Significant relationships have been reported between the steady state plasma concentrations (Css) of some classical neuroleptics and the CYP2D6 activity or genotype. Several of these drugs also potently inhibit the CYP2D6 activity. These facts explain several drug metabolic interactions of the classical drugs. Two studies failed to show that the CYP2D6 activity predicts the therapeutic effects of haloperidol or perphenazine. Some studies have suggested that the poor metabolizer phenotype is associated with the development of oversedation during treatment with the classical drugs, but other studies have been inconsistent or negative. The CYP2D6 phenotyping and genotyping appear to be useful in predicting the Css of some classical drugs, but their usefulness in predicting clinical effects must be further explored.