We wish to take the opportunity to respond to Piel et al, who have pointed out their differing experience of hepatic veno-occlusive disease (VOD) in patients taking thiopurines during maintenance therapy for childhood acute lymphoblastic leukaemia (Stoneham et al, 2003).
They also found a higher incidence of VOD, but at a lower frequency than we reported. The overall incidence in their group was 5·5% compared with 12% in ours. The male sex bias that we saw was not evident in their cohort.
More concerning perhaps, is the frequency of long-term sequelae that the authors describe. This contrasts markedly with our group who appeared mostly to suffer a mild, reversible toxicity.
Possible explanations for a differing incidence and outcome might be related to the patient cohorts considered. We had limited our patient review to all patients that had received a minimum of 1 year of maintenance therapy. In addition, we had excluded all patients who had developed VOD or significant liver dysfunction related to the intensification blocks. Our decision to limit this patient group was based on the consideration that there may be additional confounding variables at play during intensification that could contribute to the hepatotoxicity of thiopurine-related VOD. We have had two patients who have developed VOD related to intensification blocks containing thioguanine. Both patients have persistent features of hepatotoxicity that have not resolved, including abnormal liver function tests, splenomegaly and thrombocytopenia.
Thus, in Dr Piel's cohort, six of nine patients would not have been eligible for our review. Our two index cases were diagnosed retrospectively after cessation of therapy. Both patients took longer to resolve their splenomegaly and thrombocytopenia than the remainder of the cohort; perhaps indicative of greater toxicity and subsequent liver damage. Following the recognition that thrombocytopenia may be a sentinel sign of VOD, all patients at our centre with persistent thrombocytopenia on maintenance therapy were actively investigated for VOD.
In other words, we used different inclusion criteria and encountered a higher incidence of milder reversible disease, probably because we developed a low threshold for recognizing the problem and thus were looking for it. We would not be surprised that VOD associated with multiagent intensification blocks or prolonged exposure to thioguanine in the face of thrombocytopenia (or both) may be more severe and likely to produce persistent or progressive liver damage.
The gender difference we saw may, of course, be due to chance. However, as the Marsden patients are not really comparable, we feel that there may still be an important question to answer here concerning male:female differences in thiopurine metabolism.