Abstract

Background: Clozapine is known to cause agranulocytosis. Mandatory monitoring schemes are aimed at reducing the risk of agranulocytosis and of the consequences of agranulocytosis. All cases of agranulocytosis occurring in people prescribed clozapine are assumed to be caused by clozapine. Methods: In a previous study, we examined a cohort of patients listed on our hospital database as having had clozapine-induced agranulocytosis and applied specific criteria to identify those with confirmed clozapine-related, life-threatening agranulocytosis. In this study, we examine the cases not meeting these specific criteria. Results: In the original study, 9 of 23 cases met the criteria for clozapine-induced, life-threatening agranulocytosis. Of the 13 remaining cases for whom data were available, 5 were probably caused by clozapine but were not life-threatening. Three cases were the result of concomitant cancer chemotherapy. Three were anomalous results probably related to measurement error. For the remaining two cases, the cause was not identified. Conclusion: Not all cases of agranulocytosis occurring in people taking clozapine are caused by clozapine. The widely used threshold criterion-based diagnosis overestimates the risk of agranulocytosis. True clozapine-related agranulocytosis is best identified by pattern-based criteria: rapid fall in neutrophil counts over around 2 weeks to below 0.5 × 109/L for two consecutive days (unless clozapine is stopped very early or granulocyte colony stimulating factor is given) where other possible causes (benign ethnic neutropenia, cancer chemotherapy) can be ruled out.