Taylor et al showed that among the ‘reasons for discontinuing’ clozapine is the unfortunate outcome of death. ^{Reference Taylor, Douglas-Hall, Olofinjana, Whiskey and Thomas1} Out of the 21 deaths reported, five patients died from pneumonia (∼24%). Interestingly, ‘there was no evidence of neutropenia or agranulocytosis in any patients at the time of death’. ^{Reference Taylor, Douglas-Hall, Olofinjana, Whiskey and Thomas1}

The relationship between clozapine and infection is indeed complex. Links between clozapine agranulocytosis, and between agranulocytosis and the increased risk of infection are well established. Other possible indirect mechanisms of clozapine predisposition to infection, particularly aspiration pneumonia, include sialorrhoea ^{Reference Hinkes, Quesada, Currier and Gonzalez-Blanco2} and impairment of swallowing function with oesophageal dilatation and hypomotility. ^{Reference Maddalena, Fox, Hofmann and Hock3} However, less is known about whether clozapine has more direct pro-inflammatory effects. It has been argued that clozapine has a direct influence on different cytokines resembling an inflammatory reaction and that infection or inflammation could induce bioactivation of clozapine into its nitrenium ion. ^{Reference Haack, Bak, Beurskens, Maes, Stolk and Delespaul4} The latter can exert a toxic reaction that induces apoptosis and gives rise to elevated cytokine levels. ^{Reference Haack, Bak, Beurskens, Maes, Stolk and Delespaul4} However, these arguments are still awaiting robust research assessment.

Regardless of the cause of infection, a number of reports ^{Reference Haack, Bak, Beurskens, Maes, Stolk and Delespaul4–Reference De Leon and Diaz6} showed that infection leads to a rise of toxic levels in serum clozapine and its metabolites. This is likely to be mediated by cytokine suppression of cytochrome P450 1A2 (CYP1A2), the main hepatic microsomal system involved in clozapine metabolism; CYP1A2 is also involved in the metabolism of a number of antibiotics in common use for treating infections. This enhances further potentials for clozapine toxicity.

We wholeheartedly agree with Taylor et al ^{Reference Taylor, Douglas-Hall, Olofinjana, Whiskey and Thomas1} that any bronchial infection (or indeed other infections, including wound infection) should attract immediate attention. Clinicians should bear in mind that both the infection and the drug treatment of the infection (through drug–drug interactions at CYP1A2) can lead to very high and toxic levels of serum clozapine that may lead to more adversities. In such circumstances we recommend monitoring for signs of clozapine intoxication (e.g. speech dysfluency, myoclonus and increased sedation), obtaining clozapine levels, considering significant clozapine dose reduction and working closely with physical health physicians in deciding about the most appropriate antimicrobial therapy and required supportive measures.