Observations on the use of allopurinol in combination with azathioprine or mercaptopurine

Sirs, The report by Sparrow et al. of a manoeuvre using allopurinol to increase tioguanine (thioguanine) nucleotide (TGN) levels in inflammatory bowel disease (IBD) patients unresponsive to thiopurine drugs is important. In 1993, we published in The Lancet the results of the first trial of thiopurine/allopurinol combination therapy, in renal transplant patients. The combination was well tolerated. We thus applaud the work of Sparrow et al., as there has been no follow-up of this drug combination until now, in large part because the literature says the combination is unsafe. We now draw attention to several important points. The synergistic effect of allopurinol is enigmatic, for the following two reasons. First, patients who are genetically thiopurine methyl-transferase (TPMT)-deficient respond well and accumulate high red cell 6-TGN levels on doses of about 5% of normal, indicating that as much as 95% of a normal thiopurine dose is inactivated by this methylation pathway. Secondly, renal transplant patients who have clinical gout are often started on a reduced thiopurine dose, and the rule of thumb is to reduce the normal dose to about a third. This experience corresponds with the dose reduction reported by Sparrow et al., but implies that about two-thirds of a normal thiopurine dose is inactivated by the oxidation pathway of xanthine oxidase. Thus, the inactivation pathways do not add up. Importantly, Sparrow et al. have additionally shown that methylated metabolites of mercaptopurine (6-mercaptopurine; MP) are greatly reduced by the presence of allopurinol. Therefore, we are led inexorably to the unexpected conclusion that allopurinol inhibits TPMT. But what is the mechanism for this? As quoted in Sparrow et al. s report, unpublished work by Prometheus Laboratories found that allopurinol does not inhibit TPMT, presumably assayed as haemolysate activity in vitro. The first published assays of TPMT by Weinshilboum et al. also included allopurinol in the assay mixture, ostensibly to inhibit xanthine oxidase. We devised a simplified version of the TPMT assay, omitting the allopurinol (and an ion-exchange step), as it is a common error to use allopurinol as an in vitro inhibitor, particularly in cell-free assays. Allopurinol is an inactive prodrug. Its primary active form in vivo is oxypurinol, which is formed from allopurinol by aldehyde oxidase. Aldehyde oxidase and xanthine oxidase are effectively absent in human red cells, the major site of activity being liver. In addition, oxypurinol is converted in vivo to form the purine nucleotide analogue oxypurinol riboside monophosphate and its nucleoside, oxypurinol riboside – these are thought to be the active metabolites of allopurinol that inhibit the enzyme orotic phosphoribosyl transferase and provide the basis of the allopurinol load test for diagnosis of the metabolic disease ornithine transcarbamylase deficiency. Significantly, oxypurinol riboside monophosphate is a 6-oxo analogue of 6-thioinosine monophosphate, a known substrate for TPMT in vivo. Thus, there is a high probability that allopurinol provides oxypurinol metabolites in vivo that inhibit TPMT. This needs to be resolved. Sparrow et al. concluded by expressing their concern that there may be a heightened risk of nodular regenerative hyperplasia (NRH) with thiopurine plus allopurinol co-therapy. We suggest that this risk may actually be reduced, for the following two reasons. First, it has been shown that increased risk of hepatoxicity with MP or azathioprine (AZA) is associated with higher Aliment Pharmacol Ther 2005; 22: 1161–1166.