Enhanced bioavailability of azathioprine compared to 6-mercaptopurine therapy in inflammatory bowel disease: correlation with treatment efficacy


Dr C. Cuffari The Johns Hopkins Hospital, Division of Paediatric Gastroenterology, 600 N. Wolfe St, Baltimore, MD 21287, USA. E-mail: ccuffari@jhmi.edu



Azathioprine and 6-mercaptopurine have proven efficacy in the treatment of Crohn’s disease. Immunosuppression is mediated by their intracellular metabolism into active 6-thioguanine metabolites, and clinical responsiveness to therapy in patients with inflammatory bowel disease has been correlated with the measure of erythrocyte 6-thioguanine levels.

Aims and methods:

To perform a dosing equivalency analysis and comparison of clinical efficacy in 82 patients with inflammatory bowel disease on long-term (> 2 months) therapy with either branded azathioprine (Imuran) (n=26), generic azathioprine (n=38), or 6-mercaptopurine (n=18), based on the measurement of erythrocyte 6-thioguanine metabolite levels.


Disease remission was achieved in 51% (42 out of 82) of patients treated with either azathioprine or 6-mercaptopurine therapy, and correlated well with high erythrocyte 6-thioguanine levels (> 250 pmoles/8 × 108 RBCs). Patients treated with either branded azathioprine or 6-mercaptopurine achieved significantly higher erythrocyte 6-thioguanine levels than patients treated with generic azathioprine, thereby suggesting that branded azathioprine has improved oral bioavailability compared to generic azathioprine. These data are consistent with the putative immunosuppressive role of 6-thioguanine metabolites in the treatment of inflammatory bowel disease, and provides a basis for developing a therapeutic index of clinical efficacy based on the measurement of erythrocyte 6-thioguanine metabolite levels.


Our results suggest that differences in bioavailability may have clinical relevance when considering the need to optimize erythrocyte 6-thioguanine metabolite levels in patients deemed unresponsive to treatment on conventional drug dosages.


6-Mercaptopurine and its parent compound azathioprine have been widely used as immunosuppressant agents in the treatment of inflammatory bowel disease.1, 2 The immunosuppressive properties of 6-mercaptopurine and azathioprine are most likely mediated through their interference with protein synthesis and nucleic acid metabolism in the sequence that follows antigen stimulation, as well as by their cytotoxic effects on lymphoid cells.3, 4 Since 6-mercaptopurine and azathioprine are, by themselves, inactive, they must be transformed intra-cellularly into 6-thioguanine ribonucleotides, which function as purine antagonists. These anti-metabolites are then incorporated into DNA, inducing cytotoxicity and immunosuppression.5

Azathioprine is 55% of 6-mercaptopurine by molecular weight; once absorbed into the plasma, 88% is converted to 6-mercaptopurine. These pharmacological differences contribute to the conversion factor of 2.08 when converting 6-mercaptopurine to an equivalent dose of azathioprine, assuming 100% oral bioavailability.6

Although these medications share common metabolic pathways, and are often considered interchangeable, important differences in drug bioavailability exist that may influence drug efficacy and clinical responsiveness to therapy.7 6-Mercaptopurine undergoes rapid and extensive catabolic oxidation to 6-thiouric acid in the intestinal mucosa and liver by the enzyme xanthine oxidase. As proof, the bioavailability of 6-mercaptopurine ranges from 5 to 37%.7 In comparison, the intestinal absorption of azathioprine ranges from 50 to 72%.8 Once absorbed into the circulation, azathioprine is rapidly converted to 6-mercaptopurine and S-methyl-4-nitro-5-thioimidazole on exposure to sulfhydryl-containing compounds in the plasma and tissues. Twelve per cent of azathioprine is excreted in the form of S-methyl-4-nitro-5-thioimidazole, and thus represents first detoxification product known in its metabolism. This reaction has been shown to occur non-enzymatically.8 Beyond this metabolic step, in vivo studies in both animals and man have shown that azathioprine metabolism is identical to that of 6-mercaptopurine. It is not known whether the metabolism of generic azathioprine is the same as the non-generic Imuran.

The anabolic transformation of 6-mercaptopurine into its active metabolites occurs intra-cellularly along the competing routes catalysed by thiopurine methyl-transferase and hypoxanthine phosphoribosyl transferase, giving rise to 6-methyl-mercaptopurine, 6-methyl-thioinosine 5′-monophosphate and 6-thioguanine nucleotides, respectively ( Figure 1). The 6-thioguanine nucleotides are thought to be lymphocytoxic, and beneficial in the treatment of patients with leukaemia. In these patients, low erythrocyte 6-thioguanine levels have been associated with a low 6-mercaptopurine dose and with an increased risk of a disease relapse.9[10]–11

Figure ure 1.

. Azathioprine metabolism: azathioprine is non-enzymatically converted to 6-mercaptopurine (6-MP) in the plasma. The metabolism of 6-mercaptopurine occurs along competing routes catalysed by thiopurine methyl transferase (TPMT), xanthine oxidase (XO), and hypoxanthine phosphoribosyltransferase (HPRT). The incorporation of 6-thioguanine (6-TG) nucleotides into RNA and DNA induces immunosuppression.

In Crohn’s disease, the HPLC measurement of erythrocyte 6-thioguanine metabolite levels has now become a useful clinical tool for documenting patients’ compliance to therapy.12 Ongoing studies have developed the notion of a therapeutic window of efficacy and toxicity based on the measure of erythrocyte 6-thioguanine metabolite levels. Preliminary studies in patients with inflammatory bowel disease have shown that erythrocyte 6-thioguanine levels correlate well with responsiveness to therapy, and can be used clinically to optimize drug therapy.12[13]–14

Since azathioprine and 6-mercaptopurine share a common active metabolite, the purpose of this study was to perform an equivalency analysis comparing 6-mercaptopurine, azathioprine and the brand Imuran, based on the measure of erythrocyte 6-thioguanine metabolite levels. Previous studies have shown equivalency in drug bioavailability between generic and non-generic (Imuran) azathioprine, based on the measurement of plasma 6-mercaptopurine. These studies were limited by the short plasma half-life of 6-mercaptopurine, ranging anywhere from 1 to 2 h because of the rapid tissue absorption of plasma 6-mercaptopurine. Since the half-life of erythrocyte 6-thioguanine levels range between 1 and 2 weeks, its measure will reflect an accurate assessment of steady-state conditions.15

The specific aim of this pilot study was to compare differences in clinical efficacy and drug bioavailability between generic and non-generic forms of azathioprine, compared to 6-mercaptopurine, based on the measurement of erythrocyte 6-thioguanine levels.


A total of 82 patients (42 females; 40 males) with either Crohn’s disease (n=59) or ulcerative colitis (n=23) on long-term anti-metabolite therapy (> 2 month) followed at the Meyerhoff centre at The Johns Hopkins Hospital were studied. The diagnosis of Crohn’s disease and ulcerative colitis was established by standard clinical and histological findings.16 At the time of our study, a history and physical examination was performed by the attending gastroenterologist. In patients with Crohn’s disease, disease activity was measured by the Harvey–Bradshaw index (HBI). Remission was defined as an HBI of less than 5, in patients weaned off of corticosteroid therapy or on a low alternate day dose (20 mg every other day). An HBI of < 5 is equivalent to a Crohn’s disease activity index below 150, considered to be quiescent disease.17 A similar score was adapted from Lichtiger and co-workers when describing ulcerative colitis.18 In contrast to other studies, prednisone or budesonide therapy were not tapered on a fixed schedule.

Erythrocyte 6-mercaptopurine metabolite levels

Heparinized blood samples (10 cc) were obtained at random times after the daily administration of 6-mercaptopurine. After washing with phosphate buffered saline, red blood cells were centrifuged (1000 g, 5 min) and the number of red blood cells was determined from an aliquot. The packed red blood cells were frozen at –20 °C until assayed. A 50 μL aliquot of the lytic red blood cells mixture was diluted with 0.5 mL of water and mixed with 0.5 mL of 1.5 M H2SO4 and 0.3 mL of 10 m M (DL)-Dithiothreitol (DTT). This mixture was hydrolysed at 100 °C for 3 h to allow the release of free bases of 6-thioguanine and 6-methyl-mercaptopurine from their respective nucleoside and nucleotide moieties. The hydrolysate was then alkalinized with 0.5 mL of 3.4 M NaOH, and the thiobases were extracted by a 10-min mixing with 4 mL of CH2CL2 with 0.03% phenylmercuric chloride. A 3-mL aliquot of the organic layer was back-extracted by mixing with 0.2 mL of 0.1 M HCL for 5 min. A 150 μL aliquot of the aqueous phase was collected and a 100 μL fraction was analysed using the HPLC Gold System (Beckman). It consisted of a program solvent model (Model 126), an autosampler (Model 506), a programmable detector model (Model 166) and an analogue interface model. The analytes were resolved on a NOVA-PAK C184 μm (150 × 46 mm) column (Waters Associates), using a mobile phase which consisted of 50 m M H3PO4, 0.5 m M DTT (6-thioguanine at 340 nm; 6-methyl-mercaptopurine at 314 nm). Concentrations of the various thiobases were expressed as picomol/8 × 108 red blood cells (±S.E.M.).

Generic vs. non-generic

At the completion of the study, each patient was contacted by telephone by our nurse practitioner. At the time, she was unaware of their respective 6-thioguanine metabolite levels. Each patient was asked to specify their formulation of azathioprine therapy, thereby allowing a comparison of treatment efficacy and dosing equivalency analysis based on the measurement of erythrocyte 6-thioguanine metabolite levels. The content of azathioprine within the generic and brand Imuran tablets was analysed and found to be similar (difference of < 7% was not statistically significant).

Statistical analysis

Statistical analysis comparing erythrocyte 6-thioguanine metabolite levels, and drug dosing was made using Student’s t-test. A quartile analysis was used to assess clinical responsiveness to anti-metabolite therapy, based upon a presumed therapeutic erythrocyte 6-thioguanine level of 250 picomol/8 × 108 red blood cells.



A total of 38 patients were taking generic azathioprine, 26 taking the brand Imuran and 18 taking 6-mercaptopurine. All patients had their erythrocyte 6-thioguanine and 6-methyl-mercaptopurine metabolite levels measured by HPLC under ultraviolet detection. Each patient was contacted by telephone by our nurse practitioner so that the exact drug formulation could be determined prior to the measurement of metabolite levels.

Remission was achieved in 42 patients (51%) on either 6-mercaptopurine or azathioprine therapy. Patients who were deemed to be in disease remission, as defined by low indices scores whilst either taking or not taking low alternate day corticosteroid therapy, had median erythrocyte 6-thioguanine levels higher than those patients not responding to therapy ( Table 1). Interestingly, patients in clinical remission on low dose corticosteroid therapy had lower erythrocyte 6-thioguanine metabolite levels than patients in remission who were no longer taking steroids. These results would suggest the potential for further anti-metabolite dosage optimization to further reduce the need for low dose corticosteroid therapy. In all patients, a quartile analysis showed that erythrocyte 6-thioguanine levels > 250 picomol/8 × 108 red blood cells was associated with disease remission. Serial erythrocyte 6-thioguanine metabolite levels were performed in 25 patients in disease remission on a stable dose of anti-metabolite therapy. Their intra-patient variability in erythrocyte 6-thioguanine levels was less than 10%.

Table 1.  . Erythrocyte 6-thioguanine (6-TG) and 6-methyl-mercaptopurine (6-MMP) metabolite levels in patients with inflammatory bowel disease Thumbnail image of


An equivalency comparison between generic and non-generic forms of azathioprine therapy and 6-mercaptopurine is presented in Table 2. Patients taking either Imuran or 6-mercaptopurine therapy had median erythrocyte 6-thioguanine levels higher than the azathioprine treated patients despite no significant difference in drug dose ( Table 2). Higher erythrocyte 6-thioguanine levels also correlated well with a favourable clinical response to therapy ( Table 1). Although there was no statistically significant difference in median drug dosages in each treatment group, patients undergoing 6-mercaptopurine therapy were taking a lower median dosage.

Table 2.  . Equivalency analysis based on the measurement of erythrocyte 6-thioguanine (6-TG) metabolite levels Thumbnail image of


Overall, 51% of our patients with inflammatory bowel disease on long-term anti-metabolite therapy successfully achieved disease remission. This rate of remission in corticosteroid-dependant inflammatory bowel disease is lower than that previously reported as a consequence to the preferential referral of patients unresponsive to anti-metabolite therapy for erythrocyte 6-thioguanine metabolite measurement. The HPLC measurement of erythrocyte 6-mercaptopurine metabolites provided a useful tool in determining patient compliance to therapy.12 Furthermore, patients with erythrocyte 6-thioguanine levels > 250 picomol/8 × 108 red blood cells successfully achieved disease remission. In comparison, a recent study of 93 paediatric patients with inflammatory bowel disease showed that disease remission correlated well with erythrocyte 6-thioguanine levels > 230 picomol/8 × 108 red blood cells.13 These studies would support the putative immunosuppressive role of 6-thioguanine metabolites in patients with inflammatory bowel disease, and raise the notion of a therapeutic index of patient responsiveness to therapy, based on the measurement of erythrocyte 6-mercaptopurine metabolite levels.

Our pilot study would also suggest a temporal relationship between clinical responsiveness to anti-metabolite therapy and achieving therapeutic erythrocyte 6-thioguanine metabolite levels. Long-term therapy was defined as greater than 2 months, based on pharmacokinetic studies in patients with leukaemia on 6-mercaptopurine therapy. A recent study by Sandborn and co-workers have proposed that steady-state erythrocyte 6-thioguanine levels are achieved after 2 weeks of azathioprine therapy.19 In that study, patients were started from the onset on high dose (2 mg · day/kg) azathioprine therapy. Since most clinicians titrate anti-metabolite therapy biweekly to avoid toxicity, a 2-month waiting period would be optimal, prior to drug metabolite testing.

Our study is the first of its kind to perform an equivalency analysis between 6-mercaptopurine, Imuran and its generic formulation utilizing erythrocyte 6-thioguanine metabolite levels in patients with inflammatory bowel disease. In general, patients receiving either 6-mercaptopurine or Imuran therapy achieved significantly greater erythrocyte 6-thioguanine levels than the generic azathioprine treated patients. Interestingly, both the azathioprine and Imuran treated patients showed no differences in median drug doses. Although pharmacokinetic studies were not performed, this study would suggest that the non-generic azathioprine formulation has improved bioavailability. Previous studies on drug equivalency were based entirely on the measure of plasma 6-mercaptopurine production after the intestinal absorption of equal doses of either generic or non-generic forms of azathioprine. These studies are limited for several reason: first, the plasma half life of 6-mercaptopurine is measured in minutes and will not necessarily reflect the extent of tissue absorption; second, we have shown that 6-thioguanine is the active metabolite for azathioprine; third, efficacy studies have shown that erythrocyte 6-thioguanine levels can be used to monitor treatment efficacy, and can be potentially used to tailor drug therapy;20 and last, we have performed our equivalency analysis on patients with intestinal disease. In contrast, previous equivalency studies were performed either on animals or healthy adult volunteers.21 Unpublished data gathered in our laboratory did not show a significant difference in the amount of native azathioprine in either the generic or non-generic azathioprine tablets. Future crossover studies are needed to compare the oral bioavailability of these two formulations.

Similar differences in bioavailability between generic and non-generic drug formulations have been identified for the drugs phenytoin and digoxin. The non-generic formulations were found to achieve higher blood levels than their respective generic drugs. Since these drugs have well-established therapeutic windows of clinical efficacy and toxicity based on measurable blood levels, the implications in clinical practice are apparent.22, 23

In comparison, patients on 6-mercaptopurine therapy were able to achieve similar erythrocyte 6-thioguanine levels to the Imuran treated patients, with a relatively lower dose. However, when one makes an equivalency comparison based on a conversion factor of 2.08 when comparing 6-mercaptopurine to azathioprine, patients on Imuran therapy require 44% less drug in order to achieve an equivalent erythrocyte 6-thioguanine level. This may be attributed to relative differences in drug bioavailability. 6-Mercaptopurine is well recognized to have poor oral bioavailability compared to azathioprine, which could be dependant on the influences of food.24 Moreover, previous studies have also shown that the absorption of 6-mercaptopurine actually decreases with increasing doses of drug therapy.25 These differences may have clinical relevance when considering the need to tailor drug therapy in order to optimize erythrocyte 6-thioguanine metabolite levels.20 Unpublished data from our laboratory would support the role for optimizing anti-metabolite dosages to achieve an improved clinical response in patients with low erythrocyte 6-thioguanine metabolite levels (< 250 picomol/8 × 108 red blood cells).

Future controlled crossover studies are required to adequately study these putative differences in drug bioavailability between generic and non-generic forms of azathioprine, and 6-mercaptopurine therapy. These studies will also mandate that pharmacokinetic differences in drug metabolism (thiopurine methyl-transferase activity) be accurately measured for each patient recruited into the study.


This study was supported in part by the Pediatric Crohn’s and Colitis Association.