Discussion and Conclusions
Within the past two decades there have been rapid changes in the management of IBD. The traditional tiered 'step-up' approach in treatment starting with steroids has now been replaced by the use of time-critical structured approach, the 'accelerated step-up' approach and for some patients with a severe phenotype, 'top-down treatment' starting with anti-TNF agents.
However, each of the drugs in the treatment ladder is associated with risks and potential significant adverse side events. Azathioprine remains an integral part of IBD management and yet intolerance is common. The evidence for alternatives remains weaker. Some groups advocate escalation straight to anti-TNF agents, others using a personalised pharmacometabonomics approach using allopurinol, others thioguanine and there is limited evidence for methotrexate However, many of these approaches lack large-scale studies and long-term safety profiles and patients and clinicians are often wary to escalate to these treatments.
Investigators have previously reported thiopurine intolerance of 25–40% Our experience suggests this percentage is 28% with the most common reasons being nausea/vomiting, deranged LFTs and headaches. The UC cohort had a much higher incidence of deranged LFTs than the CD cohort and this has not been previously noted. Previous authors have reported that smoking is a risk factor for azathioprine intolerance, and although these data were not collected and we cannot comment on this finding, this study indicates that female sex and age 50–70 years are additional further risk factors.
Azathioprine-intolerant patients with CD were not at greater risks of developing fistulas/strictures or require surgery, and similarly, azathioprine-intolerant patients with UC were not at increased risks of requiring colectomy. This similar surgery risk was unexpected, but not entirely unsurprising as patients with azathioprine intolerance were often escalated to anti-TNF agents and there is evidence for early azathioprine use only as a disease-modifying drug. However, azathioprine intolerance in patients with CD was associated with significantly poorer symptom control with fewer patients in remission and more with moderate–severe active disease. This effect was less marked in patients in UC and probably not statistically significant owing to the small sample size.
The study has some limitations. It reflects the management of patients in a single UK hospital that uses an 'accelerated step-up' strategy. Furthermore, only one in seven of the patients with CD was receiving infliximab and this lower usage reflects the limited anti-TNF licence in the UK. The generalisation of our findings outside this setting is not clear. Nevertheless, the strengths of this study derive from robust long-term follow-up of a cohort of patients from a database, allowing in-depth analysis of outcomes and causes of intolerance. Further research into the causes and risk factors for azathioprine intolerance is required and the results of the PRED4 study will help to determine underlying causes.
In conclusion, azathioprine intolerance acts as a surrogate marker for patients with CD who may in future have poorer symptom control. The most common reasons for intolerance are nausea/vomiting, deranged LFTs and headaches. Female sex and an older age are associated with intolerance and initiating the drug in this cohort requires caution. Azathioprine intolerance marks out a cohort of patients in whom increased vigilance of symptom control is warranted and early escalation of treatment required.