Author's reply

Abstract

Sirs, We welcome the comments by Kitiyakara et al. regarding the clinical utility of thiopurine methyltransferase (TPMT) genotype screening.1 Their argument illustrates that TPMT genotyping, like any screening test, is only cost-effective if fully understood by the clinician and performed properly. The literature (and our model) highlights that most leucopoenic episodes occur, for undetermined reasons, in patients with normal TPMT genotype. We therefore emphasize that TPMT genotyping must only be employed as an adjunct to regular full blood count monitoring. Although the majority of leucopoenic events occur in patients with normal genotype, patients who are homozygous for mutant TPMT alleles (0.3% of the population) have such a high risk of profound and life-threatening leucopoenia2 that their detection prior to commencing therapy represents a cost-effective use of the assay. These patients should not be given thiopurine therapy. The screening policy for the remaining patients commencing azathioprine therapy depends on careful monitoring of the full blood count. We agree that the optimal haematological monitoring strategy has not been defined, but are unaware of any data suggesting early leucopoenia is related to inadequate monitoring. We have concerns that patients with homozygous deficiency may develop early, severe leucopoenia despite frequent full blood count monitoring in the first few months of treatment and therefore advocate genotyping. It has been suggested that the detection of heterozygous deficiency (11% of the population) of TPMT activity would allow tailoring of the starting dose of azathioprine with slow and cautious dose escalation so as to avoid leucopoenia.3 This seems a reasonable but, as yet, unproven approach. Currently, there is no way of predicting the majority of leucopoenic episodes, which develop in patients with normal TPMT activity. The regular and careful monitoring of the full blood count remains the key to avoidance of leucopoenia. Audits that suggest poor compliance with this policy are certainly worrying. It is vital that clinicians should be fully conversant with the limited role of TPMT genotyping and not lead into a false sense of security in a patient with a normal genotype. The use of a clinical nurse specialist in this monitoring role is recommended.