To the Editors:In a recent study by Breen et al., a possible causal role of the key enzyme thiopurine methyltransferase (TPMT) in inducing nodular regenerative hyperplasia (NRH) during azathioprine (AZA) treatment in liver transplant patients was considered. 1 Both patients who developed this histological liver abnormality had one mutant TPMT allele, probably leading to an impaired ability to methylate AZA-generated metabolites and subsequently to an overproduction of the pharmacologically active end metabolites 6-thioguaninenucleotides (6-TGN). 2 We believe that NRH may not be caused by the TPMT activity itself but is attributable to the generated 6-TGN concentrations. Recently, the use of the thiopurine 6-thioguanine (6-TG) in inflammatory bowel disease (IBD) patients has been discarded due to its presumed role in inducing NRH. 3 A major pharmacokinetic difference between AZA and 6-TG is that during 6-TG administration (40 to 80 mg), much higher 6-TGN levels (Ͼ1,000 pmol/10 8 per red blood cell [RBC]) are achieved than during AZA treatment (normally between 100 and 500 pmol/10 8 RBC). 2 These relatively high 6-TGN levels may explain the alleged elevated risk of developing NRH during 6-TG therapy. Our liver histological data concerning the use of low-dose 6-TG for treating IBD patients intolerant or refractory to AZA therapy support this hypothesis (Table 1). At our hospital, a liver biopsy is routinely performed in patients who use 6-TG for at least one year. In addition, 6-TGN levels are regularly monitored. The fact that we have not encountered one case of NRH may well be explained by the relatively low 6-TGN levels (mean 442 pmol/10 8 RBC) achieved as a consequence of the prescribed 6-TG dosages (approximately 0.3 mg/kg). The induction of NRH during thiopurine treatment in general may be 6-TGN dependent. This hypothesis explains the relatively low but well-documented incidence of NRH during AZA treatment as high 6-TGN levels are merely found in those patients with impaired TPMT activity (approximately 10% of the population) and would also provide an answer for the high incidence of NRH during highdose 6-TG treatment. We therefore consider low-dose 6-TG (0.3 mg/kg) as a therapeutic alternative, but until more long-term toxicity data are available, 6-TG use should preferably be restricted to trials with a regular performance of liver biopsies.