Review Article: Monitoring of Immunomodulators in Inflammatory Bowel Disease

F. N. Aberra; G. R. Lichtenstein


Aliment Pharmacol Ther. 2005;21(4):307-319. 

In This Article

Mercaptopurine and Azathioprine

Mercaptopurine and the nitroimidazole derivative of MP, AZA, are thiopurine analogues. AZA is non-enzymatically metabolized to MP. MP is metabolized to 6-thioinosine 5'-monophosphate (TIMP) by the enzyme hypoxanthine phosphoribyl transferase (HPRT). TIMP is eventually metabolized to 6-tioguanine (thioguanine) nucleotides (6-TGN) (see Figure 1). 6-TGN has been shown to bind to Rac 1 instead of guanine triphosphate and with co-stimulation with CD28 leads to inhibition of Rac 1 and apoptosis of T-lymphocytes.[1] MP is also metabolized to 6-methylmercaptopurine (6-MMP) by the enzyme thiopurine methyltransferase (TPMT) and 6-thiouric acid by the enzyme xanthine oxidase (XO). Both 6-thiouric acid and 6-MMP are inactive metabolites of MP. The three enzymes metabolizing MP are in constant competition for substrate and the concentration of the metabolites of MP are based on the concentrations of these enzymes. About 84% of MP is quickly metabolized by XO found in high concentrations in enterocytes and hepatocytes, leaving only 16% left to be catabolized by TPMT and HPRT.[2]

Metabolism of azathioprine (AZA) and mercaptopurine (MP). 6-TU, 6-thiouric acid; 6-MMP, 6-methyl mercaptopurine; 6-MMPR, 6-methyl mercaptopurine ribonucleotides; 6-TIMP, 6-thioinosine 5'-monophosphate; 6-TXMP, 6-thioxanthosine 5'-monophosphate; 6-TGN, 6-tioguanine nucleotides; XO, xanthine oxidase; TPMT, thiopurine methyltransferase; HPRT, hypoxanthine phosphoribosyltransferase; IMPDH, inosine monophosphate dehydrogenase; GMPS, guanosine monophosphate synthetase.

Genetic Variation of Metabolism. Mutations of the gene encoding TPMT leading to varying functional activity of the enzyme have been identified in the population. There are at least 12 mutant alleles responsible for TPMT deficiency and several silent and intronic mutations have been described for the TPMT gene located on chromosome 6.[3,4] The most common of these variants is the normal TPMT*1 or wild-type allele. TPMT*3A is the most common mutant allele and seen predominantly in Caucasians. TPMT*3C is the most prevalent variant in African-American and Asian populations.[5–10] Homozygous mutations of the TPMT gene produce enzyme with minimal activity and accounts for 0.3% of the general population.[11,12] Heterozygous mutations yield enzyme with moderate activity and accounts for 11% of the general population.[11] A recent study in patients with CD has shown the frequency of homozygous and heterozygous TPMT mutations parallels the frequency in the general population.[12] TPMT level is measured from lysed erythrocytes and the TPMT level corresponds to that in liver, kidney and lymphocytes.[13] TPMT activity appears to be inversely correlated with drug response. TPMT genotyping is also available but limited to the allelic variants that are known. More detail is given utilizing TPMT activity for tailoring treatment with AZA or MP in the following sections.

Indication and Dosing. There have been several clinical trials supporting the use of AZA and/or MP for the treatment of the following indications: active CD, maintenance of remission of CD, fistulizing CD, prevention of post-operative recurrence of CD, maintenance of remission of UC, and as a steroid-sparing agent. In a meta-analysis of randomized-controlled trials from 1966 to 1994 of AZA and MP in active and quiescent CD the odds ratio (OR) for a clinical response to therapy of active CD was 3.09 (95% CI: 2.45–3.91), 1.45 (95% CI: 1.12–1.87) without the one clinical trial with MP.[14] For quiescent CD the OR for maintaining remission if on AZA was 2.27 (95% CI: 1.76–2.93). Both longer duration and higher dose of AZA/MP was associated with improved response and a steroid-sparing effect was seen in subjects with active and quiescent disease. Several studies have also shown a steroid dose reduction with the addition of MP and AZA to therapy.[15–22] MP and AZA have also been found to be useful in the treatment of fistulizing CD with several studies showing a higher response of complete closure or decreased drainage of fistulas.[17,21,23] There is also emerging data supporting MP and AZA use for CD post-operative prevention of recurrence of in those with previous history of surgeries and perforating disease.[24–28] In UC, AZA and MP appear to be most useful in maintaining remission and as a steroid-sparing agent.[29]

The mercaptopurine and AZA have a significantly delayed onset of action with several studies showing clinical benefit after 2–3 months of treatment.[17,21] The best approach if rapid clinical response is needed is to start another form of therapy such as corticosteroids or infliximab until the period of clinical onset is achieved. It is not yet clear what the ideal length of therapy for maintenance of remission should be. A study by Bouhnik et al . showed a low relapse rate of <20% in patients in remission who used AZA for 4 years or more, 2 years after completing therapy. The relapse rate was similar to those that continued AZA therapy.[30] In a recent study published in abstract form by Lemann et al ., 19% relapsed at 18 months and 60% relapsed at 54 month after AZA was stopped for maintenance of remission of CD.[31]

For CD and UC the most effective doses appear to be 2.5 mg/kg of AZA and 1.5 mg/kg of MP, although there has not yet been a head-to-head comparison at various dose levels, or a comparative trial evaluating efficacy of MP vs. AZA in patients with IBD. There are two methods for starting therapy, dose escalating to the weight-based dose vs. starting immediately at the weight-calculated dose. Dose escalating came into practice due to the fear of dose-related toxicities such as leucopoenia, thrombocytopenia and hepatitis. Whether leucopoenia or other dose-related side-effects are less severe with a gradual escalation algorithm compared with starting at the weight-calculated dose has yet to be determined. Neither method prevents toxicity. In a study by Connell et al ., the range of time from starting AZA to onset of leucopoenia was 2–11 years with a median of 9 months.[32] Whereas in a study by Present et al . most of the cases of leucopoenia occurred within 1 month of starting MP.[33] The dose escalating method chosen is arbitrary, but a common practice is to start at 50 mg everyday and increase the dose by 25 mg every 1–2 weeks and monitor for leucopoenia and other potential adverse events.


Use of Metabolites: 6-TGN and 6-MMP. The metabolites 6-TGN and 6-MMP have been used to determine likelihood of therapeutic response. See Table 1 for metabolite levels and the associated response and/or toxicity. Non-responders to MP and AZA can be categorized into three groups based on levels of the metabolites 6-TGN and 6-MMP. (i) Low levels of both metabolites are likely a consequence of under dosing or non-compliance. (ii) Low 6-TGN and high 6-MMP may signify increased likelihood for hepatotoxicity and also suggest metabolic resistance to MP/AZA, shunting away from 6-TGN production.[34] Early recognition of this pattern may lead to early replacement of AZA with another medication. (iii) A normal 6-TGN level likely represents refractoriness to AZA or MP.

The initial studies of metabolite levels and associated clinical response were studied in paediatric IBD patients predominately with CD with various disease severities, type of disease and disease distribution. The application of measuring metabolite levels in UC studies has not been as well evaluated. Discriminant levels of 235–250 pmol/8 × 108 have been considered predictive of response. Yet, several studies have shown subjects that were in remission receiving MP or AZA with 6-TGN levels lower than 235–250.[35–39] Table 2 provides the sensitivity, specificity and positive predictive value of using discriminant levels (>230, 235 and 250 pmol/8 × 108) to predict clinical remission based on data from published studies. The positive predictive value for remission of 6-TGN level >230–250 are poor to fair at best ranging from 24 to 83%. Overall, therapeutic drug monitoring with measuring 6-TGN levels in patients treated with AZA or MP may be most useful and can be considered in the following selected settings: patients suspected of non-compliance and possibly patients who are failing to respond to standard doses of drug. The latter has yet to be proven in a prospective controlled fashion. There is now an ongoing trial comparing the benefit of dosing based upon 6-TG levels vs. weight-based dosing.

TPMT. The TPMT enzyme activity and genotype are tests that are available on a commercial basis for testing. An increasing number of TPMT mutations leading to reduced enzyme are being identified. Normal levels of enzyme activity appear to be associated with more response than high levels of TPMT enzyme activity but both groups have a widely variable clinical response.[40,41] Low enzyme levels can lead to development of severe myelosuppression in the setting of MP/AZA therapy.[42,43]

The relationship of leucopoenia and various TPMT genotypes in the setting of AZA or MP treatment has been the focus of several research studies. In a study by Colombel et al . of 75 patients whom developed leucopoenia while treated with AZA or MP, 10% of patients were homozygous for TPMT mutant alleles and 17% were heterozygous for TPMT mutant alleles.[44] The median time to leucopoenia was significantly shorter in those with homozygous TMPT mutations, 1 month (range: 0–1.5) compared with 4 months (range: 1–18) in those with one mutant allele and 3 months (range: 0.5–87) in those with normal TPMT.[44] It is not yet clear whether patients heterozygous for a TPMT mutation are more likely to develop leucopoenia compared to those with normal TPMT. In a study by Seddik et al . published in abstract form 75 subjects with CD that were starting AZA/MP treatment were assessed for TPMT mutations. Six (8%) of the subjects were heterozygous for a TPMT mutation and none was homozygous of a TPMT mutation. Seven (9%) of the subjects developed leucopoenia and only one subject was heterozygous for a TPMT mutation.[45] In a case–control study by Curvers et al . published as an abstract subjects with TPMT variant alleles were compared to subjects with normal alleles for the risk of developing leucopoenia and the OR was 1.5 (95% CI: 0.7–3.2) for developing leucopoenia.[46] In this study and others, the majority of myelosuppression cases occur with normal levels of TPMT enzyme activity.

The prevalence (one of 300) of low TPMT enzyme is high enough and the potential complications of myelosuppression severe enough for us to recommend obtaining a TPMT enzyme activity level prior to starting therapy. A few small studies have suggested that TPMT-deficient patients can safely use low doses of MP/AZA, but until more studies verify safety other immunomodulators should be considered.[42,47]

In patients with intermediate TPMT enzyme activity, the risk of myelosuppression is not clearly increased when compared to those with normal TPMT enzyme activity. Until more studies are performed formal evidence-based recommendations to vary treatment (e.g. maximal dose used or rapidity of achieving maximal dose) based upon whether an individual has the wild type (full TPMT enzyme activity) or intermediate TPMT enzyme activity cannot be formally established. It is not yet clear whether TPMT genotype or phenotype can also be used to predict other adverse effects of AZA such as nausea, vomiting, hepatitis and pancreatitis because of conflicting data.[48]

Mean Corpuscular Volume. Several studies have suggested that mean corpuscular volume (MCV) correlates with 6-TGN levels and perhaps may be used as surrogate marker for monitoring therapeutic dosing.[36,49,50] A study by Thomas et al ., showed that a change of MCV (mean change in MCV was 7.5 ± 6.3 fL) from baseline during treatment with AZA or MP was directly correlated with 6-TGN levels ( r = 0.33, P < 0.001) and inversely correlated with leucocyte counts ( r = 0.26, P = 0.001).[50] A study by Belaiche et al . also showed a correlation of MCV with 6-TGN concentrations ( r = 0.38, P = 0.048).[36] Further studies are needed to determine the MCV level equivalent to a 6-TGN level.

Several side-effects have been associated with MP/AZA use and include allergic reactions, pancreatitis, myelosuppression, nausea, infections, hepatoxicitiy and malignancy. Bone marrow suppression is related to levels of 6-TGN and may occur at any time during the duration of treatment.[32] Therefore, monitoring of complete blood counts at regular intervals throughout the duration of therapy is suggested. It has been described that leucopoenia can occur in the setting of 6-TGN levels <230 pmol/8 × 108, but has also been associated with non-response to MP/AZA.[39] Hepatoxicity is a rare complication and the pathophysiological mechanism of hepatic injury is unknown. Possibilities include drug-induced hepatitis, cholestasis, nodular regenerative hyperplasia and peliosis.[33,51] 6-MMP levels >5700 pmol/8 × 108 is associated with hepatotoxicity, but patients may have high levels with normal liver chemistries and at low 6-MMP levels hepatotoxicity may occur.[39] Routine measurement of 6-MMP for hepatotoxicity is not currently recommended, although the routine measurement of serum liver chemistries is recommended. A baseline measurement of serum liver chemistries should be obtained, then periodic measurements at least once every year thereafter. In patients developing nausea may be switched from AZA to MP or vice versa.[52] Patients developing biochemical hepatitis or pancreatitis should have AZA or MP treatment discontinued. There have been several cases of lymphoma reported in AZA/MP users and may be partially related to infection with Epstein–Barr virus, but it is still unclear if there is a definitive increased risk.[53–55]

Several categories of drugs have been shown to possibly interact with MP and AZA metabolism and include medications that have 5-aminosalicylates (5-ASA) as the active moiety [sulfasalazine, mesalazine (mesalamine), olsalazine and balsalazide], allopurinol, aspirin and furosemide. In a study by Lowry et al . RBC 6-TGN levels were slightly higher but not statistically significant in subjects taking mesalazine, sulfasalazine or olsalazine concurrently compared with subjects not on these medications, 182 vs. 153 pmol/8 × 108 respectively ( P = 0.10). The pathway of the drug interaction between MP/AZA and 5-ASA appears to be by inhibition of the TPMT enzyme which may lead to a higher risk for leucopoenia.[56–59] Genetic variants of an intestinal epithelial enzyme arylamine- N -acetyl transferase, involved in 5-ASA metabolism, also appear to increase the risk for adverse events of MP therapy.[60] Subjects on these medications should be more carefully monitored for myelosuppression.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.