In a retrospective study of patients with inflammatory bowel disease, van Dieren et al recently reported the absence of a correlation between genotypes for both inosine triphosphate pyrophosphatase (ITPA) and thiopurine methyltransferase (TPMT), with any side effects to azathioprine (AZA) (Gut 2005;54:1664). This contrasts with two other studies. A rigorous prospective study, published recently, has demonstrated a significant association between ITPA genotype and early dropout from AZA therapy.¹ Our original publication implicated ITPA in a number of adverse effects, which were independent of myelosuppression.² Another letter has reported non-association of ITPA with myelosuppression³ but thiopurine induced myelosuppression has been well documented over the past 25 years as associated with TPMT, not ITPA, status.⁴

However, we draw attention to a peculiar feature of the TPMT results of van Dieren et al that one patient—who suffered severe myelosuppression—was reported as TPMT*3B/*3B genotype. We previously published a meta-analysis of the incidence of the TPMT*3B (G460A) mutation,⁵ discovering that it is rare, and this has been confirmed by a recent large study, making the chance of homozygosity negligible.⁶ Indeed, our evidence suggested that even the few …