Prediction of Clozapine-Induced Occurrences of Hematological Abnormalities Using Initial Hemodynamic Changes

Jeong, Seong Hoon; Kim, Yong Sik

Abstract

STUDY OBJECT: Clozapine has been known to induce variety of side effects. Amongst of all, the hematological abnormalities, especially agranulocytosis, has prohibited the wide-spread use of this drug. There have been a lot of efforts to predict the hematological abnormalities based upon the demographical and clinical characteristics. In addition, although only a few, prediction by the initial hemodynamic change was also attempted. This study was carried out as an attempt to find a reliable predictor of clozapine-induced hematological abnormalities based upon the baseline hematological status and the initial hemodynamic change by clozapine. METHOD: Regardless of the diagnosis, the complete blood count data of every patient who had received clozapine in Seoul National University Hospital from 1996 to the present were analyzed. The risk group was defined as the patient whose absolute neutrophil count(ANC) had dropped below 1500mm3 equal or more than 3 times during the one year after clozapine trial. The occurrences of future clozapine-induced hematological abnormalities were predicted by the baseline ANC and the initial ANC change. RESULT: The baseline ANC immediately before clozapine trial, the degree of ANC declining during the first week, and that of ANC rising during the 2nd and 3rd week all showed significant difference between the risk group and the safe group. The likelihood ratio of risk was 7.75(95% confidence interval: 2.77-21.3) implying significant risk of hematological abnormalities when the baseline ANC was below 2000mm3, and the likelihood ratio of risk was 0.372(0.159-0.798) when the baseline ANC was above 4000mm3. In likewise manner, the interval likelihood ratios of risk associated with the ANC rising during the first 3 weeks were calculated. These two predictor variables contributed mutually independent information in predicting future hematological abnormalities.
CONCLUSION
The baseline ANC and the initial hemodynamic change after clozapine trial could help predicting future clozapine-induced hematological abnormalities. If the more reliable predictors can be found, prescreening high risk patients using these predictors and the close hematological monitoring of these patients may not only decrease the risk associated with clozapine usage, but also widen the indication of clozapine by relieving much of the burden currently imposed upon the doctors.