Abstract
Background: In treatment-refractory schizophrenia, pre-dose plasma clozapine concentrations are, on average, lower in smokers than in nonsmokers. This may relate to the incidence of adverse drug reactions (ADR).
Methods: Generalized estimating equation analyses were used to investigate clozapine TDM data from 1993 to 2017 for ADR with ≥30 mentions on request forms.
Results: There were 1154 mentions of an ADR (1058 patients), 36,470 controls ("baseline" samples, 14,572 patients). In the ADR group, median plasma clozapine concentration (mg/L) was significantly lower in smokers (0.41, N=586) than nonsmokers (0.56, N=368) (P<0.001). The constipation subgroup, for example, had higher median plasma clozapine concentrations [0.56 (0.22-1.27) mg/L] than the controls [0.45 (0.18-0.93) mg/L] (P=0.025). Higher plasma clozapine concentrations were also associated with hypersalivation, hypotension, and neurological ADR (seizures and myoclonus), but no associations were found in relation to cardiac, low white cell count, metabolic, and nausea/vomiting subgroups. In the neurological ADR subgroup, plasma clozapine concentrations were higher in smokers (P<0.001) and in nonsmokers (P=0.019) compared with controls. In the constipation subgroup, plasma clozapine was only significantly higher in nonsmokers, while in the hypotension and sedation subgroups, the opposite applied. Plasma norclozapine showed a similar pattern of results to that of clozapine. The plasma clozapine:norclozapine ratio showed no consistent pattern between the ADR subgroups.
Implications: ADR such as constipation and seizures may be more likely to occur at lower plasma clozapine concentrations in smokers. The underlying mechanism(s) require further investigation.