Optimising outcome on thiopurines in inflammatory bowel disease by co-prescription of allopurinol

Transcription

1 Journal of Crohn's and Colitis (2012) 6, Available online at Optimising outcome on thiopurines in inflammatory bowel disease by co-prescription of allopurinol Melissa A. Smith a, Paul Blaker a, Anthony M. Marinaki b, Simon H. Anderson a, Peter M. Irving a, Jeremy D. Sanderson a, a Department of Gastroenterology, Guy's & St. Thomas' NHS Foundation Trust, London, UK b Purine Research Laboratory, Guy's & St. Thomas' NHS Foundation Trust, London, UK Received 8 December 2011; received in revised form 7 February 2012; accepted 7 February 2012 KEYWORDS: Azathioprine; 6-Mercaptopurine; Thioguanine nucleotides; Inflammatory bowel disease; Allopurinol Abstract Background and aims: Azathioprine and mercaptopurine remain first line immunomodulatory treatments for inflammatory bowel disease. Toxicity and non-response are significant issues. Co-prescription of allopurinol with reduced-dose (25 33%) azathioprine or mercaptopurine may overcome these problems. We present the outcome of co-prescription in a large singlecentre cohort. Method: Patients on thiopurine/allopurinol co-prescription were identified. Indication for and outcome on combination treatment were established. Blood parameters and metabolite results were compared on single agent and combination treatment. Toxicity associated with combination treatment was sought. Results: 110 patients on combination treatment were identified. Clinical remission was achieved in 60/79 (76%) of patients in whom the effect of thiopurine could be studied in isolation. 20/25 patients with hepatotoxicity tolerated combination treatment and normalised their liver function tests. 24/28 patients with atypical side effects tolerated co-therapy. 13/ 20 non-responders responded to combination treatment. In patients started on combination treatment as first line therapy, 15/23 achieved clinical remission. Thioguanine nucleotides Abbreviations: AZA, azathioprine; MP, mercaptopurine; MeMP, methylated thiopurine metabolites; TGN, thioguanine nucleotide; IBD, inflammatory bowel disease; UC, ulcerative colitis; CD, Crohn's disease; IBD-U, inflammatory bowel disease unclassified; OFG, oral facial granulomatosis; TPMT, thiopurine methyltransferase; XDH, xanthine oxidase/dehydrogenase; LFTs, liver function tests; ADRs, adverse drug reactions; TEN, toxic epidermal necrolysis; LFTs, liver function tests; PRL, Purine Research Laboratory; GSTFT, Guy's and St. Thomas' NHS Foundation Trust. Conference presentation: An interim analysis of 35 patients was presented at the BSG and DDW An abstract calculating the potential gain from allopurinol co-prescription, using data from this cohort, has been submitted to DDW Corresponding author at: Department of Gastroenterology, 1st Floor, College House, St Thomas' Hospital, Lambeth Palace Road, London, SE1 7EH, UK. addresses: melissa.smith@gstt.nhs.uk (M.A. Smith), paul.blaker@gstt.nhs.uk (P. Blaker), tony.marinaki@gstt.nhs.uk (A.M. Marinaki), simon.anderson@gstt.nhs.uk (S.H. Anderson), peter.irving@gstt.nhs.uk (P.M. Irving), jeremy.sanderson@kcl.ac.uk (J.D. Sanderson) /$ - see front matter 2012 European Crohn's and Colitis Organisation. Published by Elsevier B.V. All rights reserved. doi: /j.crohns

2 906 M.A. Smith et al. were significantly higher and methylated metabolites significantly lower on combination therapy. Mean cell volume was higher and total white cell and neutrophil counts lower on combination treatment. 13 adverse events occurred, including 6 specific to co-therapy (3 rash, 2 abnormal liver function tests, 1 dosing error). All were minor and self-limiting. Conclusion: This is the largest published experience of the use of allopurinol to optimise outcomes on thiopurine treatment. Combination therapy permitted successful treatment of a significant number of patients who would otherwise have been labelled as thiopurine failures. A few self-limiting side effects were encountered European Crohn's and Colitis Organisation. Published by Elsevier B.V. All rights reserved. 1. Introduction The purine analogues azathioprine (AZA) and mercaptopurine (MP) have become established as first line immunomodulatory therapy for inflammatory bowel disease (IBD), and up to 60% of patients diagnosed with IBD will take thiopurines at some point in their disease course. 1 When effective, thiopurines have advantages over biologic treatments, not just in terms of cost and convenience but potentially also safety and durability of response. They also retain an important role as concomitant immunomodulation alongside biologic treatment. However, thiopurines can cause significant toxicity and a proportion of patients will fail to respond. Strategies designed to overcome these issues would be highly advantageous, ensuring that the greatest possible number of patients benefit from this key treatment option. Allopurinol, an inhibitor of xanthine oxidase/dehydrogenase (XDH), is a designer drug which was originally developed alongside thiopurines in an attempt to optimise thiopurine metabolite profiles and improve treatment response. However, when the original studies in leukaemia failed to demonstrate additional benefit from allopurinol co-prescription, the strategy was abandoned and allopurinol instead found its niche in the treatment of gout. 2 Decades later in the 1990s, coprescription of azathioprine and allopurinol in the context of renal transplant immunosuppression improved graft survival 3 and optimised thiopurine metabolite profiles. 4 Interest in combination treatment with AZA and allopurinol has recently been revived in the IBD literature, following the characterisation of patients demonstrating a predominant methylator phenotype 5 mostly in association with a high thiopurine methyltransferase (TPMT) activity. Patients with this phenotype have a higher risk of both hepatotoxicity 6 and thiopurine non-response. 7,8 On allopurinol/aza combination treatment however, the unfavourable metabolite profile is reversed (i.e. higher TGN level and lower methylated metabolites), hepatotoxicity and other side effects 9 avoided and a successful response to treatment recaptured. Allopurinol affects the metabolism of thiopurines by blocking the activity of XDH which is a major contributor to thiopurine catabolism, (Fig. 1). The increase in thiopurine drug thus available for conversion to TGNs requires a dose reduction to approximately 25% of the usual target AZA dose to avoid dose-related toxicity, particularly myelotoxicity. 7 However, the mechanism by which it so dramatically reduces the production of methylated metabolites of thiopurines is the subject of much debate. Combination therapy has been used in our institution for a number of years consequent on an interest in thiopurine metabolism. We have previously published two early series of patients. 6,9 Here, we report our experience in a new large cohort of patients with IBD prescribed thiopurine/allopurinol co-treatment, describing indications for co-treatment, treatment outcome and toxicity encountered. 2. Methods Patients prescribed combination treatment with allopurinol and AZA or MP were identified from pharmacy records of co-prescription, records of a pharmacist-led dedicated thiopurine monitoring clinic, by reviewing case notes in the IBD clinic and from TGN monitoring requests received by the Purine Research Laboratory (PRL), GSTS Pathology, Guy's and St Thomas' NHS Foundation Trust (GSTFT), London, UK. Clinical data was collected retrospectively from the clinical notes and electronic patient records to establish patient and disease characteristics, indication for allopurinol cotreatment and clinical outcome. Where available, blood results were compared before and after the addition of allopurinol, particularly liver function tests, full blood counts and TGN and MeMP levels. TPMT activity, TGN and MeMP levels were measured by the Purine Research Laboratory at GSTFT. TPMT activity was measured as previously described. 10 Activity b10 pmol/h/ mghb was considered to reflect TPMT deficiency, (although Figure 1 The effect of allopurinol on azathioprine metabolism. AZA azathioprine, MP mercaptopurine, MeMP methylmercaptopurine,tgns thioguanine nucleotides, XDH xanthine oxidase/dehydrogenase,tpmt thiopurine methyltransferase.

3 Allopurinol/Thiopurine Co-prescription in IBD 907 no patients in this range were included). TPMT activity in the range pmol/h/mghb was considered intermediate (with a target AZA dose of mg/kg) and activity N25 pmol/h/mghb was considered to represent normal TPMT metabolism (target AZA dose mg/kg). Those patients with TPMT activity N35 pmol/h/mghb were considered to have high TPMT activity and were considered for primary therapy with combination AZA and allopurinol. TGN and MeMP measurements were taken as the hydrolysed base in 0.5 ml whole blood according to the PCA hydrolysis method reported by Dervieux and Boulieu. 11 The therapeutic range for TGN measurements in our laboratory was considered to be pmol/ RBC Protocol for combination treatment Combination treatment was considered in the following circumstances: Hepatotoxicity on thiopurine monotherapy new abnormality of liver function tests (LFTs) occurring on thiopurine therapy which the treating physician attributed to the thiopurine drug. All patients developing hepatotoxicity were investigated for other causes of abnormal LFTs. Other side effects on thiopurine monotherapy the occurrence of any adverse event co-incident with thiopurine therapy which resolved on ceasing the culprit agent, the event must have been severe enough to warrant cessation of thiopurine agent. Where the thiopurine was tolerated but side effects limited dose escalation and therefore clinical response, patients were included as sub-optimal responders. Sub-optimal response inability to achieve clinical remission on thiopurine monotherapy with thiopurine metabolite profile either suggesting hyper-methylation or under-dosing which could not be corrected due to intolerance of higher doses. This group also included patients with a historical label of non-response to thiopurine monotherapy and a TPMT activity N35 pmol/h/mghb. Predominant methylation only patients with a ratio of MeMP:TGNN11 but not currently experiencing a loss of response or toxicity. High pre-treatment TPMT activity patients with a TPMT greater than 35 pmol/h/mghb (the threshold we have previously demonstrated predicts non-response 12 ) were considered for combination treatment as first line thiopurine therapy. The initial dosing protocol for combination therapy used in this study is shown in Table 1. Allopurinol was given at 100 mg per day Outcome measures The primary outcome measure was successful circumvention of the problem encountered on thiopurine monotherapy, which in some cases was clinical remission, but in others drug tolerance, metabolite profile etc. Additional analysis was conducted to seek remission rates for the whole group, regardless of their indication for combination therapy. This was a retrospective analysis therefore Table 1 Dosing strategy for low dose thiopurine and allopurinol, adapted from. 9 TPMT genotype formal clinical disease activity measures (Harvey Bradshaw Index, Simple colitis activity index etc.) were not generally available. Clinical response was assessed at one year of treatment and determined from the physician's global assessment according to indication for treatment. Any recourse to steroids, biologics or surgery indicated treatment failure. A secondary clinical response analysis was also conducted to include those on therapy for greater than 4 months but not yet reaching one year of therapy. Adverse Drug Reactions (ADRs) were defined as any adverse event occurring on combination therapy severe enough to require the cessation of combination treatment Monitoring on combination treatment Blood counts and LFTs were monitored at weeks 2, 4, 8 and 12 following the switch to combination therapy and TGN/ MeMP levels checked at approximately 4 weeks after the switch to allow informed dose adjustment. Following this bloods were monitored every 3 months in the same manner as for standard azathioprine treatment Statistical analysis Means were compared by two-tailed t-test where the data sets passed normality testing and Wilcoxon rank where this criteria was not met, using Instat, Graphpad software. 3. Results TPMT phenotype pmol/h/ mghb AZA dose with 100 mg allopurinol MP dose with 100 mg allopurinol Wild-type mg/kg mg/kg Heterozygous mg/kg mg/kg Homozygous b10 Avoid Avoid 110 patients on combination thiopurine/allopurinol treatment were identified. Of these, one patient was lost to follow up, so 109 were available for analysis of outcome. The indications for co-prescription were: hepatotoxicity (n = 24), other side effects (n = 28), inadequate response with low TGN/high methylated metabolites (n = 27), adverse metabolite profile with good clinical response (n = 8) and primary treatment due to high TPMT, (n=23). The average duration of follow up on combination treatment was 16 months, (range months, short time-frames were due to treatment discontinuation as a result of ADRs). The male:female ratio was 51:59, age range years, disease duration 1 34 years. There were 69 patients with Crohn's disease (CD), 33 with ulcerative colitis (UC) and 4 with IBD-unclassified (IBD-U). Additionally we identified

4 908 M.A. Smith et al. one patient treated with thiopurines for eosinophilic colitis and 3 from our joint oral medicine/gastroenterology clinic with orofacial granulomatosis (OFG). Amongst those patients with UC, 3 had proctitis, 12 had left-sided colitis, and 18 subtotal/pancolitis. Amongst those with CD, 7 had ileal disease, 31 ileo-colonic and 30 colonic disease, 18 patients had peri-anal disease (including 3 with isolated peri-anal disease), 5 had upper GI disease, 3 had oral CD and one cutaneous CD (several patients had involvement of multiple parts of the GI tract). 26 patients were receiving concomitant oral mesalazine, 4 rectal mesalazine, 21 were on prednisolone and 4 on oral beclomethasone. 30 patients were concurrently receiving biologic therapy (22 infliximab, 8 adalimumab), 2 were on ciclosporin A and one was weaning from mycophenolate mofetil. 69 patients had been treated with single agent AZA and 8 had received single agent MP before allopurinol coprescription. 12 had unsuccessfully switched from AZA to MP for non-specific side effects. 20 patients had no recorded prior exposure to thiopurine monotherapy. In individuals with normal TPMT activity, single agent AZA doses ranged from 0.6 to 2.5 mg/kg, average 1.9 mg/kg, (those on low doses were unable to escalate as a result of drug toxicity). On combination therapy, AZA doses ranged from 0.16 to 1.0 mg/kg, average 0.54 mg/kg, MP doses ranged from 0.37 to 1.77 mg/kg, average 1.03 mg/kg Success in overcoming problem with monotherapy Overall, 64/78 (82%) patients receiving combination therapy successfully overcame the problem encountered on thiopurine monotherapy. This analysis excludes those patients that were started on combination therapy as a first line thiopurine treatment, those lost to follow up and those where response would be the outcome measure but this could not be accurately judged due to either the duration of treatment less than 4 months or use as concomitant immunomodulation with a biologic. If those starting combination treatment as first line were included, with remission at one year as their outcome measure, then 78/96 (81%) of the combination treatments were a success Overall clinical response rate When only those patients completing a full year (or longer) of successful therapy are counted as treatment successes, response rate is 39/63 (62%). When all 109 patients are analysed, 59/83 (71%) where the effect of thiopurine could be studied in isolation, had achieved clinical remission at one year (or the latest follow up if treated for less than one year). Of the 23 patients dependent on oral steroids at the time of initiating combination therapy, 14 were able to completely stop steroids and a further 2 were able to successfully reduce their steroid dose at one year. If only those who had failed on thiopurine monotherapy are analysed, (i.e. excluding those never given a trial of thiopurine monotherapy and those switched solely on the basis of metabolite results), response rate is 39/59 (66%) Thiopurine metabolite profiles TGN levels increased from a median of 213 to 397 pmol/ RBC on co-treatment, pb Meanwhile, MeMP levels decreased from a median of 3559 pmol/ RBC on single agent treatment to just 115 on combination therapy, p b Where paired metabolite levels were available for individual patients they are shown graphically, (Figs. 2 and 3) TPMT The median TPMT activity for the whole cohort was 39.5 pmol/ h/mghb. Excluding those that were started on combination treatment due to a high pre-treatment TPMT activity, the median reduced to 37 pmol/h/mghb Blood counts Total white blood cell count reduced slightly on transfer to combination treatment, from a mean of 6.9 to 6.4 (p=0.02), however, this was not, as might be anticipated, due to a reduction in lymphocyte count, which was unchanged, p = 0.27, but due to a reduction in neutrophil count (from 4.7 to 4.3, p=0.02). MCV measurements increased slightly on co-therapy, (rising from a mean of 92.3 to 92.6, pb0.0001). MCV/WBC ratio, which has been proposed as a surrogate marker of thiopurine response increased from 15.1 to 16.8, p = Figure 2 The effect of co-prescription of allopurinol on Thioguanine nucleotide levels in pmol/ RBC. Excluding the patient who did not reduce his azathioprine dose appropriately.

5 Allopurinol/Thiopurine Co-prescription in IBD 909 Figure 3 The effect of allopurinol co-prescription on Methylated metabolite levels pmol/ RBC Adverse drug reactions on combination therapy In total, thirteen adverse events were reported. Three patients had to discontinue allopurinol due to the development of a rash; in all cases this was mild and self-limiting. No cases of toxic epidermal necrolysis were encountered. Two patients developed abnormal liver function tests on combination treatment, which appeared most likely to have been due to the allopurinol. One of these patients was also found to have early PSC, but on reverting back to single agent thiopurine his LFTs improved. The other patient underwent colectomy for UC and has discontinued thiopurine therapy. One patient misunderstood the dosing instructions and continued to take full dose AZA alongside the allopurinol. This was detected early (whilst blood counts remained normal) by metabolite monitoring and the patient experienced no adverse effects. In addition, six patients experienced nausea and vomiting, alongside other minor side effects and discontinued therapy with no long term problems. 5 of the 6 patients had experienced similar side effects on thiopurine monotherapy. One further patient discontinued therapy as it was felt to be contributing to frequent exacerbations of COPD. 4. Results according to indication for combination treatment 4.1. Hepatotoxicity Of the 25 patients switched to combination treatment for hepatotoxicity, 20 patients were able to tolerate combination therapy with a normalisation of their LFTs. One discontinued therapy immediately due to nausea (which he had also experienced on single agent treatment) and one was lost to follow up. Three patients had an alternative/additional explanation for abnormal LFTs discovered during their work up, 1 genetic haemochromatosis and 2 early PSC. These three patients continued on combination treatment on the rationale that it minimised the risk of their treatment contributing to liver damage, however, one of the PSC patients was eventually tried back on single agent AZA with some improvement in his LFTs suggesting that the allopurinol may have been contributing to the liver function abnormality. 9/21 patients with LFTs available on single agent treatment had the switch to combination treatment made before transaminases reached double the upper limit of normal. However, in the hepatotoxic subgroup ALT was significantly reduced after introduction of combination treatment, from a mean of 151 IU/L to 30 IU/L, (mean difference 124 IU/L, 95%CI 61 to 187, p=0.0006). Likewise GGT was reduced from a mean of 163 IU/L to 49 IU/L, (mean difference 133 IU/L, 95%CI , p=0.005), and ALP from a mean of 134 IU/L to 76 IU/L, (mean difference 50 IU/L, 95%CI 8 93, p=0.02) Other adverse drug reactions 28 patients were transferred to co-therapy for ADR's other than hepatotoxicity. The side effects experienced by these individuals on monotherapy were GI disturbance, rash, alopecia, flu-like symptoms, tremor, headaches, myalgia, arthralgia, myelotoxicity, fatigue and frequent respiratory infections. Most patients experienced multiple side effects. 24/28 (86%) were able to tolerate co-therapy, 3 had to discontinue combination treatment due to side effects (in 2 patients a recurrence of the side effects they experienced on monotherapy [arthralgia, nausea and fatigue] and in 1 abnormal LFTs) and 1 progressed rapidly to colectomy for uncontrolled UC. In those where a judgement about the clinical outcome of combination treatment could be made (no concomitant biologic, adequate duration of therapy) 11/20 (55%) achieved a clinical remission at 1 year (or the latest available assessment if treatment duration for less than one year), the pure one year remission rate was 47%. Patients discontinuing treatment due to side effects were considered to be treatment failures Non-response 27 patients were switched to combination therapy for inadequate response to monotherapy. In 18 patients this was in conjunction with predominant methylation on TGN monitoring, 6 had a historical label of non-response to azathioprine in conjunction with a high TPMT activity. 3 patients were partial responders, tolerating thiopurine well at low doses but in whom dose-optimisation could not occur due to side effects. In those where a judgement about the clinical outcome of combination treatment could be made (no concomitant biologic, adequate duration of therapy), 10/17, (59%) were in clinical remission at one year on combination therapy (including those in remission but yet to complete one year of treatment, this figure rose to 13/20 [65%]). Those

6 910 M.A. Smith et al. discontinuing combination treatment due to side effects were considered to be treatment failures Primary therapy Of the 23 patients started on combination therapy for high TPMT activity, 5 could not be assessed for clinical response (1 concomitant biologic, 2 inadequate duration of treatment and 2 non-adherent by TGN), 14 of the remaining 18 patients (78%) achieved clinical remission on combination treatment. One had their treatment discontinued due to frequent exacerbations of COPD, but no other side effects were encountered Adverse metabolite profile 8 patients were changed to combination treatment despite a good clinical response to single agent treatment on the basis of an adverse metabolite profile. One was unable to tolerate allopurinol (rash) and was transferred back to single agent treatment without any loss of response or long term problems, the remaining 7 patients improved their metabolite profile and maintained clinical remission (one of these patients was also on a biologic throughout). 5. Discussion In this larger study, we have confirmed the findings of previous smaller studies, that the use of allopurinol alongside appropriately reduced doses of AZA and MP can allow those developing both hepatotoxicity and other side effects to benefit from thiopurine treatment. It is now time for a well-designed randomised controlled study. The emerging role for thiopurine/allopurinol co-prescription to circumvent atypical side effects is important, as these account for many patients unable to tolerate thiopurine treatment. We have demonstrated that many of these patients are able to tolerate thiopurine when administered in this way. This is important not just for patients starting thiopurine therapy for the first time, but also for those with a historical label of thiopurine intolerance. In this cohort, combination therapy successfully recaptured a clinical response in a significant number of those failing to respond to thiopurine monotherapy. Despite the advent of biologic therapies in the treatment of IBD, thiopurines remain an important independent treatment option and also have a key role as concomitant immunosuppression with biologics Hence, any strategy which can increase the number of patients able to tolerate and benefit from thiopurine therapy is a valuable addition to our current treatment options. Taking these two indications together it is possible to estimate the proportion of patients that could stand to benefit from this strategy. In a cohort of 207 prospectively recruited patients with IBD patients had to discontinue thiopurine monotherapy due to non-specific side effects, 8 due to hepatitis and 32 were non-responders with a TPMT activity greater than 35 pmol/h/mghb. This is a total of 100 patients with clear indications for combination treatment with allopurinol. Our results predict that 53 of these patients could have achieved remission on combination treatment. As this represents 26% of the whole original cohort this is an important development. A potential problem with this analysis is that the side effect rate reported in this study was relatively high. A more representative rate would perhaps be 10% 16 with 2.8% due to side effects unsuitable for combination therapy (pancreatitis and myelotoxicity). Using this rate of side effects (non-response by TPMT is not available in other studies) would still give a figure of 12% of all patients started on AZA, that would previously have failed AZA treatment, but would achieve 1 year remission on combination therapy. The original rationale for use of combination therapy in hepatotoxicity was to circumvent toxicity due to high methylated metabolite levels. However, hepatotoxicity on thiopurine treatment is multifactorial and combination treatment could also reduce liver toxicity by allowing reduced doses of the culprit drug, blocking the production of oxygen free radicals 7 and by raising hypoxanthine levels which could lead to improved cell salvage/repair. 17 Nonetheless, it is also interesting to note that abnormal liver function developed in 2/109 of our patients in association with allopurinol treatment, one of whom initially had the deterioration in LFTs attributed to PSC. The mechanisms underlying many of the ADRs encountered on thiopurines are poorly understood. Nausea is heterogeneous but is often dose-related, worsening on escalation to target thiopurine dose in many cases. Likewise, the myalgic flu-like syndrome experienced by a number of patients, appears dose-related and could be due to accumulation of thiopurine metabolites, particularly thioitp. 12,18 These issues can sometimes be circumvented by a switch to MP 19 or very slow dose escalation, akin to desensitisation in hypersensitivity. 20 However, it is clear from the current series that the significantly reduced parent drug concentration permitted by combination therapy is a very effective means of circumventing these adverse reactions. Blood monitoring remains as important on combination therapy as it is for monotherapy. Indeed, it is probably more important, as small dose adjustments whilst on allopurinol have a much larger impact. Our practice is to adopt the same protocol for blood count and LFT monitoring used for patients on monotherapy. Monitoring of TGN/MeMP levels is also important on combination therapy, as it aids careful monitoring and facilitates finely tuned dose adjustment. Indeed, checking TGN levels at 4 weeks avoided a potentially serious dosing issue in this cohort. We recommend checking TGNs 4 weeks after switching to combination therapy, 4 weeks after any dose adjustment, to investigate non-response or adverse events on therapy and every 6 12 months for patients stable on combination treatment. Safety concerns in the literature over the use of combination allopurinol/aza treatment relate to dose-related toxicity, 8 this should therefore be avoidable by correct dosing, supported by TGN measurement. Allopurinol, importantly, can itself be associated with a variety of side effects including gastro-intestinal disturbance, malaise, headache, vertigo, drowsiness, visual and taste disturbances, hypertension, alopecia, hepatotoxicity, paraesthesia and neuropathy, gynaecomastia and blood disorders. Perhaps the most significant adverse events however are rashes. Whilst most of these are mild and self-limiting, they can be more serious. Toxic epidermal necrolysis (TEN)

7 Allopurinol/Thiopurine Co-prescription in IBD 911 is of particular concern and may be more common in individuals of Asian extraction, particularly among the Han Chinese population. 21 Recent studies in European populations have shown an incidence of approximately 1: for TEN with approximately 17% of these cases being attributable to allopurinol, making it the commonest culprit drug, 22 reassuringly, doses less than 200 mg appear to be safer. 22 An association with HLA-B*5801 has been demonstrated, 23 and since TEN carries a significant mortality, 24 pre-treatment pharmacogenetic testing is being proposed by some groups. 23 In our cohort the adverse event profile mirrored this experience, and fortunately no serious toxicity was encountered. The use of combination AZA and allopurinol therapy as a first line treatment is new and potentially controversial. It was adopted, after prolonged discussion, in our department after a series of patients with very high TPMT activity developed the complications of hyper-methylation (hepatotoxicity and non-response) on single agent therapy, exactly as predicted. It was agreed that these patients were being exposed to unnecessary, predictable risks (which could be easily prevented with combination treatment) and incurring avoidable delay in achieving a successful clinical response. Furthermore, the use of allopurinol and azathioprine and the negative influence of very high TPMT are both now well described in the literature, meriting translation to clinical practice. However, it is becoming clear that TPMT activity alone does not provide a complete explanation for hypermethylation, with some patients showing predominant methylation despite a TPMT activity below the population median and, conversely, a few patients with very high TPMT activity showing a normal TGN/MeMP levels. At present therefore, first line combination therapy, whilst safe, cannot be recommended for general use on the grounds of a very high pre-treatment TPMT activity alone, but should wait until predominant methylation is demonstrated on TGN monitoring on treatment. Further studies will hopefully identify which additional factors, genetic or otherwise, might predict hypermethylation. A prospective trial, ideally comparing single agent thiopurine against combination therapy in all patients initiating thiopurines, will be required before combination therapy can be more widely recommended. It is not immediately clear why blocking XDH with allopurinol should have such a profound effect on methylated metabolite levels. Allopurinol is not thought to have a direct impact on TPMT activity, 8 although published data is lacking in this area. Early evidence suggests, however, that thioxanthine is a TPMT inhibitor and higher thioxanthine production in the presence of allopurinol could provide the explanation for this profound suppression of the production of MeMP. 25 It is also possible that the pharmacokinetics of TPMT are such that, at the lower drug doses permitted by co-administration of allopurinol, methylation of thiopurines and their metabolites is much less efficient. This is supported by the partial reversal of predominant methylation that can be achieved by splitting the dose of single agent thiopurine. 26,27 The shift in balance between TGN and methylated products may also be partly explained by decreased breakdown of the TGNs, for which XDH and aldehyde oxidase are responsible, (Fig. 1). It is interesting to note that in a few patients TGN levels dropped when they were started on combination therapy. This is likely to represent cautious initial dosing. Doses can be titrated based on TGNs in the same way as for monotherapy but, as for initial doses, any dose adjustment should be reduced to 25 33% of that which would normally be attempted on monotherapy. In conclusion, allopurinol and AZA combination treatment provides a safe and effective treatment option for the majority of patients either intolerant or unresponsive to thiopurines, increasing the number of patients that can achieve remission on these first line agents. As long as the dose adjustment for AZA is correctly applied, co-treatment appears to be safe and can be monitored in the same way as standard AZA treatment. TGN/MeMP measurements are however important to identify suitable candidates for cotreatment, to facilitate dose adjustments and to avoid doserelated toxicity. Statement of authorship JS and AM conceived the study, MS collected and analysed the data and drafted the manuscript. MS, PB, SA, PI and JS recruited patients, AM provided advice and consultation. All authors participated in the study design and read, critically revised and approved the final manuscript. Conflict of interest The authors have no conflict of interest to declare. Acknowledgement We acknowledge funding support from Guy's and St Thomas' Charity who had no other input into the study or its publication. References 1. Cosnes J, Nion-Larmurier I, Beaugerie L, Afchain P, Tiret E, Gendre JP. Impact of the increasing use of immunosuppressants in Crohn's disease on the need for intestinal surgery. Gut 2005;54: Elion GB. The purine path to chemotherapy. Science 1989;244: Chocair P, Duley J, Simmonds HA, Cameron JS, Ianhez L, Arap S, et al. Low-dose allopurinol plus azathioprine/cyclosporin/- prednisolone, a novel immunosuppressive regimen. Lancet 1993;342: Chrzanowska M, Krzymanski M. Determination of 6-thioguanine and 6-methylmercaptopurine metabolites in renal transplantation recipients and patients with glomerulonephritis treated with azathioprine. Ther Drug Monit 1999;21: Dubinsky MC, Yang H, Hassard PV, Seidman EG, Kam LY, Abreu MT, et al. 6-MP metabolite profiles provide a biochemical explanation for 6-MP resistance in patients with inflammatory bowel disease. Gastroenterology 2002;122: Ansari A, Elliott T, Baburajan B, Mayhead P, O'Donohue J, Chocair P, et al. Long term outcome of using allopurinol co-therapy as a strategy for overcoming thiopurine hepatotoxicity in treating inflammatory bowel disease. Aliment Pharmacol Ther 2008;28: Sparrow MP, Hande SA, Friedman S, Lim WC, Reddy SI, Cao D, et al. Allopurinol safely and effectively optimizes tioguanine metabolites in inflammatory bowel disease patients not

8 912 M.A. Smith et al. responding to azathioprine and mercaptopurine. Aliment Pharmacol Ther 2005;22: Sparrow MP, Hande SA, Friedman S, Cao D, Hanauer SB. Effect of allopurinol on clinical outcomes in inflammatory bowel disease nonresponders to azathioprine or 6-mercaptopurine. Clin Gastroenterol Hepatol 2007;5: Ansari A, Patel N, Sanderson J, O'Donohue J, Duley JA, Florin TH. Low-dose azathioprine or mercaptopurine in combination with allopurinol can bypass many adverse drug reactions in patients with inflammatory bowel disease. Aliment Pharmacol Ther 2010;31: Breen DP, Marinaki AM, Arenas M, Hayes PC. Pharmacogenetic association with adverse drug reactions to azathioprine immunosuppressive therapy following liver transplantation. Liver Transpl 2005;11: Dervieux T, Boulieu R. Simultaneous determination of 6-thioguanine and methyl 6-mercaptopurine nucleotides of azathioprine in red blood cells by HPLC. Clin Chem 1998;44: Ansari A, Arenas M, Greenfield SM, Morris D, Lindsay J, Gilshenan K, et al. Prospective evaluation of the pharmacogenetics of azathioprine in the treatment of inflammatory bowel disease. Aliment Pharmacol Ther 2008;28: Colombel JF, Sandborn WJ, Reinisch W, Mantzaris GJ, Kornbluth A, Rachmilewitz D, et al. Infliximab, azathioprine, or combination therapy for Crohn's disease. N Engl J Med 2010;362: Sokol H, Seksik P, Carrat F, Nion-Larmurier I, Vienne A, Beaugerie L, et al. Usefulness of co-treatment with immunomodulators in patients with inflammatory bowel disease treated with scheduled infliximab maintenance therapy. Gut 2010;59: Vermeire S, Noman M, Van AG, Baert F, D'Haens G, Rutgeerts PJ. The effectiviness of concomitant immunosuppressive therapy to suppress formation of antibodies to infliximab in Crohn's disease. Gut 2007;56(9): Prefontaine E, Macdonald JK, Sutherland LR. Azathioprine or 6-mercaptopurine for induction of remission in Crohn's disease. Cochrane Database Syst Rev 2010:CD Butler R, Morris AD, Belch JJ, Hill A, Struthers AD. Allopurinol normalizes endothelial dysfunction in type 2 diabetics with mild hypertension. Hypertension 2000;35: Marinaki AM, Duley JA, Arenas M, Ansari A, Sumi S, Lewis CM, et al. Mutation in the ITPA gene predicts intolerance to azathioprine. Nucleosides Nucleotides Nucleic Acids 2004;23: Hindorf U, Johansson M, Eriksson A, Kvifors E, Almer SH. Mercaptopurine treatment should be considered in azathioprine intolerant patients with inflammatory bowel disease. Aliment Pharmacol Ther 2009;29: Korelitz BI, Reddy B, Bratcher J. Desensitization of patients with allergic reactions to immunosuppressives in the treatment of inflammatory bowel disease. Expert Opin Drug Saf 2010;9: Lee HY, Ariyasinghe JT, Thirumoorthy T. Allopurinol hypersensitivity syndrome: a preventable severe cutaneous adverse reaction? Singapore Med J 2008;49: Halevy S, Ghislain PD, Mockenhaupt M, Fagot JP, Bouwes Bavinck JN, Sidoroff A, et al. Allopurinol is the most common cause of Stevens Johnson syndrome and toxic epidermal necrolysis in Europe and Israel. J Am Acad Dermatol 2008;58: Somkrua R, Eickman EE, Saokaew S, Lohitnavy M, Chaiyakunapruk N. Association of HLA-B*5801 allele and allopurinol-induced Stevens Johnson syndrome and toxic epidermal necrolysis: a systematic review and meta-analysis. BMC Med Genet 2011;12: Atzori L, Pinna AL, Mantovani L, Ferreli C, Pau M, Mulargia M, et al. Cutaneous adverse drug reactions to allopurinol: 10 year observational survey of the Dermatology Department Cagliari University (Italy). J Eur Acad Dermatol Venereol 2011, doi: /j x [Epub ahead of print]. 25. Blaker P, Arenas M, Fairbanks L, Irving P, Marinaki AM, Sanderson JD. A biochemical mechanism for the role of allopurinol in TPMT inhibition. Gastroenterology 2011;140(Suppl 1): Tu Bell BA, Brockway GN, Shuster JJ, Erdmann G, Sterikoff S, Bostrom B, et al. A comparison of red blood cell thiopurine metabolites in children with acute lymphoblastic leukemia who received oral mercaptopurine twice daily or once daily: a Pediatric Oncology Group study (now The Children's Oncology Group). Pediatr Blood Cancer 2004;43: Bradford K, Shih DQ. Optimizing 6-mercaptopurine and azathioprine therapy in the management of inflammatory bowel disease. World J Gastroenterol 2011;17: