6-Thioguanine — efficacy and safety in chronic active Crohn's disease


Dr K. R. Herrlinger, Department of Gastroenterology, Hepatology and Endocrinology, Robert-Bosch-Hospital, Auerbachstrasse 110, D-70376 Stuttgart, Germany.
E-mail: klaus.herrlinger@rbk.de


Background : Azathioprine and mercaptopurine are commonly used in chronic active Crohn's disease. They share the disadvantage of a delayed onset of action and potentially serious side-effects, and are metabolized to thioguanine nucleotides which are thought to be the active metabolites. The direct use of 6-thioguanine may offer a more rapid and safer alternative. We conducted an open prospective study to investigate the efficacy and safety of 6-thioguanine in chronic active Crohn's disease.

Methods : Thirty-seven patients with chronic active Crohn's disease and a Crohn's disease activity index of > 150 were enrolled in this study. Inclusion criteria were steroid dependence (n = 19), steroid refractoriness (n = 9) and/or intolerance (n = 16) or refractoriness (n = 6) to azathioprine. Patients were treated with 40 mg/day of 6-thioguanine for 24 weeks; a dose escalation to 80 mg was allowed at week 12. Remission was defined as a Crohn's disease activity index of < 150 associated with a decrease of > 70 points; response was defined as a decrease of > 70 points in the Crohn's disease activity index.

Results : In the intention-to-treat analysis, 13 of 37 patients achieved remission (35%). Twelve of these 13 patients achieved remission after 4 weeks. Fifty-seven per cent of patients (21/37) achieved a response. The mean Crohn's disease activity index decreased from 284 ± 74 to 153 ± 101. 6-Thioguanine was more effective in azathioprine-intolerant than in azathioprine-refractory patients. Twelve of 16 patients intolerant to azathioprine tolerated 6-thioguanine. Adverse events included phototoxicity, pancreatitis, headache, nausea, alopecia, arthralgia, minor infections and reversible elevation of transaminases. Six patients required discontinuation of medication, two because of leucopenia.

Conclusions : In this patient group with chronic active Crohn's disease, 6-thioguanine appeared to be effective with acceptable short-term toxicity, but long-term controlled trials are clearly needed to further define its role.


More than 50% of patients with steroid-treated Crohn's disease suffer from steroid refractoriness or steroid dependence.1 For these patients, immunosuppressive therapy is warranted for induction and maintenance of remission and for steroid reduction. The thiopurine analogue azathioprine and its metabolite mercaptopurine represent the gold standard for immunosuppression in Crohn's disease.2–4 For unknown reasons, the onset of action is delayed for up to 6 months, representing a major problem in severe steroid-refractory disease, and, despite adequate treatment dosing and duration, at least one-third of patients fail to respond to therapy.

In addition, recent data have suggested that up to 28% of patients have to stop medication because of intolerable side-effects.3 Although the exact mechanism of toxicity is unknown, side-effects have been differentiated into ‘allergic-type’ and ‘dose-related’ toxicity. Mercaptopurine exerts less allergic-type reactions than azathioprine, but only the minority of idiosyncratic reactions can be averted by switching to mercaptopurine. In particular, drug-induced allergic pancreatitis is caused by an allergic reaction to an undefined moiety or metabolite of both drugs. Dose-dependent toxicity seems to be partly dependent on the enzyme thiopurine methyltransferase (TPMT). Low TPMT activity results in high 6-thioguanine levels with the risk of haematotoxic side-effects,5 whereas hepatotoxicity may be correlated with high levels of 6-methyl-mercaptopurine, the product of the TPMT-catalysed methylation of mercaptopurine.6

By using 6-thioguanine, which is also metabolized by TPMT, but serves as a direct precursor of 6-thioguanine nucleotides, it may be possible to avoid certain toxic metabolites. 6-Thioguanine has been widely used in paediatric leukaemia, and has been shown to produce very high 6-thioguanine nucleotide levels with little toxicity.7, 8 Data on its therapeutic effects in Crohn's disease are scarce, with only one promising pilot study in selected mercaptopurine-resistant patients.9

Therefore, we conducted a 6-month open-label study to evaluate the efficacy and tolerability of 6-thioguanine in chronic active Crohn's disease.

Patients and methods


Thirty-seven patients (14 males, 23 females; median age, 35.7 years; range, 22–61 years) with chronic active Crohn's disease and a Crohn's disease activity index (CDAI) of > 150 were enrolled in this open-label, prospective study (Table 1). The median duration of disease was 10.1 years (range, 0.5–28 years). Inclusion criteria included steroid dependence (n = 19), steroid refractoriness (n = 9) and/or intolerance (n = 16) or refractoriness (n = 6) to azathioprine. Reasons for azathioprine intolerance were pancreatitis (three patients), hepatotoxicity (two patients), nausea (four patients), abdominal pain (four patients), headache (one patient) and non-specified allergic reaction (two patients). Twenty-seven patients were on systemic steroids, five patients used budesonide alone and five patients used systemic and topical steroids in combination. Twenty-two patients had been treated with azathioprine prior to the study and 23 patients had received mesalazine.

Table 1.  Characteristics of the patients included in the study
  1. CDAI, Crohn's disease activity index. Results are expressed as median (range).

Male : female14 : 23
Age (years)35.7 (22–61)
Disease duration10.1 (0.5–28)
Site/extent of disease
 Small bowel 2
CDAI (baseline)283 (170–448)

Inclusion and exclusion criteria

All patients had active Crohn's disease proven by endoscopy and histological findings, with active disease documented by a CDAI of more than 150. Furthermore, patients had to fulfil one of the following criteria of chronic activity: (i) steroid dependence (at least one relapse following steroid tapering during the previous 6 months); or (ii) steroid refractoriness (active disease with a CDAI of > 150 despite treatment with steroids at a dose of > 10 mg prednisolone equivalent or 9 mg budesonide for at least 8 weeks and a stable dose during the 2 weeks before baseline); and/or (iii) intolerance to azathioprine; or (iv) azathioprine refractoriness (active disease with a CDAI of > 150 despite treatment with azathioprine at a dose of > 1.5 mg/kg body weight for at least 12 weeks before baseline).

Reasons for exclusion were short bowel syndrome, prior colostomy or ileostomy, symptomatic stenosis, bacterial or viral infection, pregnancy, serious second illness, active cancer or history of cancer, compromised hepatic function, TPMT deficiency and treatment with other immunomodulatory or experimental agents within 3 months before baseline.

Study design

Patients received 40 mg of 6-thioguanine daily for 24 weeks. At week 12, a dose escalation to 80 mg was allowed for non-responders. Treatment with azathioprine had to be stopped prior to study entry. Co-medication, including steroids, had to be kept stable during the final 2 weeks before the trial. Steroids had to be kept stable for the first 4 weeks and were tapered subsequently according to the clinical response. Aminosalicylates were allowed as stable co-medication throughout the trial.

The study medication was provided by Dr Falk-Pharma GmbH, Freiburg, Germany. The study was approved by the local ethical committees of all centres and written informed consent was given by every patient prior to study entry. The study was conducted according to the Declaration of Helsinki and in accordance with the guidelines of Good Clinical Practice.

Laboratory controls

Laboratory controls for blood counts, liver and pancreatic enzymes, C-reactive protein, erythrocyte sedimentation rate and urine analysis were performed at every visit.

TPMT activity

To minimize the risk of haematological toxicity of 6-thioguanine, TPMT phenotyping was performed by measuring the erythrocyte TPMT enzyme activity in every patient prior to study entry according to previously published protocols.10, 11 TPMT levels of > 23 U indicated normal enzyme activity. Four patients had intermediate TPMT activity (< 23 U and > 10 U). Patients were excluded from the study when the TPMT activity was < 2 U.

Statistical analysis

It was calculated (‘nQuery Advisor’, Statistical Solutions, USA) that a one-group chi-squared test with a 5% one-sided significance level would have 80% power to detect a difference between the proportions 0.33 (null hypothesis) and 0.54 (alternative hypothesis) if the sample size was 33. All patients who fulfilled the inclusion criteria and received at least one single dose of study medication were included in the intention-to-treat analysis (n = 37). If not indicated specifically, variables are described as medians with ranges throughout the text. The statistical analysis for CDAI was performed using Student's t-test for paired data. As the steroid dose was not normally distributed, the Wilcoxon test was used for the analysis of steroid reduction. A P value of < 0.05 was considered to be statistically significant.


Thirty-seven patients received treatment and were included in the intention-to-treat analysis. Nine patients had to be withdrawn from the study prematurely due to intolerable adverse events (six patients), lack of efficacy (two patients) and violation of inclusion criteria (one patient). Two patients whose steroid dose had to be increased in the final 4 weeks of the study period were included as failures, but excluded from the calculation of the final CDAI.

Clinical remission and response

In the intention-to-treat group, the mean CDAI decreased from 284 ± 74 to 153 ± 101 (ΔCDAI = 131, P < 0.001) (Figure 1). Thirteen patients (35%) achieved remission. Remarkably, 12 of the 13 patients achieving remission did so as early as 4 weeks; the median time to achieve remission was 29 days (14–126 days). Fifty-seven per cent of patients (21/37) achieved a clinical response, defined as a decrease in the CDAI of > 70 points. The mean Inflammatory Bowel Disease Questionnaire (IBDQ) score increased from 128 ± 25 points at baseline to 157 ± 42 points at the end of the study.

Figure 1.

Time course of Crohn's disease activity index (CDAI) at study visits (mean ± S.E.M.).

In a sub-group analysis, the efficacy was best for steroid-refractory patients, with five of nine achieving remission and seven showing a response. Only six of 18 steroid-dependent patients achieved remission (response in eight of 18). Fifteen patients were azathioprine-naive before study entry, seven of whom achieved remission (response in nine); this was in contrast with five of the azathioprine-intolerant patients and none of the azathioprine-refractory patients. In 12 patients without a clinical response to 40 mg thioguanine, the daily dose was increased to 80 mg, but only one achieved remission on dose escalation.

Steroid reduction

Twenty-seven patients were on systemic steroids, and 20 (74%) were able to reduce the dose. The median daily prednisolone equivalent dose was decreased from 13.75 mg to 5.0 mg (P < 0.03); the median individual decrease in prednisolone equivalent was 67%. Eight patients were completely weaned from systemic steroids. Five of 10 (50%) patients receiving budesonide at baseline were completely withdrawn from this topical steroid.


Two serious adverse drug reactions with a possible causal relationship to the study drug occurred: one moderate photoallergic reaction following a sun studio visit and a rotavirus infection. Both patients recovered and completed the study. Six patients had to be withdrawn from the study due to intolerable adverse drug reactions. The reasons for withdrawal due to intolerable side-effects were mild pancreatitis (one patient), headache (two patients), pneumonia (one patient) and leucopenia (two patients). One patient experienced erythema nodosum, which recovered after drug withdrawal and exacerbated with re-exposure. Remarkably, 12 of 16 patients previously intolerant to azathioprine tolerated 6-thioguanine. The clinically relevant adverse events are listed in Table 2.

Table 2.  Clinically relevant adverse events
Adverse eventn
  • SAE, serious adverse event.

  • Necessitating study termination.

Headache17 (2*)
Vomiting 2
Pruritus 3
Arthralgia 3
Alopecia 3
Dysaesthesia 1
Eczema 4
Exanthema 2
Photoallergic reaction (1 SAE) 2
Erythema nodosum 1
Rotavirus infection (SAE) 1
 Respiratory tract11 (1*)
 Urinary tract 4
 Genital candidiasis 2
 Herpes labialis 1
Elevation of transaminases 2
Leucopenia 3 (2*)
Anaemia 1
Thrombopenia 1
Pancreatitis 1*

The mean basic laboratory parameters remained essentially unchanged between baseline and week 24, with a drop in leucocytes and thrombocytes within the normal range (Table 3). However, myelosuppression occurred in three patients. In all patients, medication was discontinued. One patient developed leucopenia and thrombopenia and recovered 27 days after withdrawal. One patient experienced leucopenia and anaemia; this patient recovered from leucopenia within 2 weeks, but anaemia was still present when the patient was lost to follow-up. Pancytopenia developed in a third patient at the final visit. This patient recovered completely within 8 weeks after withdrawal. The last two patients were at higher risk for toxic events due to intermediate TPMT activity. All three patients were additionally treated with aminosalicylates.

Table 3.  Laboratory parameters at baseline and final visit
 Baseline (n = 37)Final visit (n = 27)
  1. Results are expressed as median (range).

Leucocytes (103/µL)10.30 (5.20–22.70)7.45 (3.00–15.30)
Erythrocytes (106/µL)4.72 (3.60–5.54)4.32 (2.67–5.17)
Haemoglobin (g/dL)13.40 (8.60–15.48)13.35 (9.19–15.30)
Thrombocytes (103/µL)345.0 (179.0–1005.0)286.0 (132.0–516.0)
C-reactive protein (mg/dL)1.20 (0.30–8.60)0.79 (0.20–3.80)

Clinically relevant elevation of transaminases occurred in two patients. Both patients recovered without a reduction in the thioguanine dose.


The thiopurines azathioprine and mercaptopurine are very effective in the treatment of chronic active Crohn's disease, but share the disadvantage of a delayed onset of action and several side-effects. The shorter pathway of 6-thioguanine towards the production of thioguanine nucleotides may avoid several potentially toxic metabolites, raising the possibility of better tolerability. Based on these theoretical considerations and encouraged by promising results from a small open-label study on 6-thioguanine in Crohn's disease,9 we tested for a beneficial role of 6-thioguanine in the treatment of Crohn's disease. The aim of this study was to investigate the efficacy and tolerability of 6-thioguanine in chronic active Crohn's disease.

Overall, 35% of patients achieved remission at the end of the study, and a clinical response was seen in 57%. These results are almost identical to those of a previously published small series on 6-thioguanine in Crohn's disease9 (remission rate, 40%; clinical response, 60%), as well as those of a recent controlled trial of azathioprine-induced remission after 16 weeks (remission in 36% of patients).12 In a meta-analysis of the treatment of Crohn's disease with standard thiopurines, remission was obtained in 54%.4 However, only four of the eight studies included used validated criteria for outcome measures, such as the CDAI. According to a 3-year retrospective analysis, the remission rate for azathioprine in Crohn's disease was 45%.3 Thus, 6-thioguanine is unlikely to be superior to the standard thiopurines. It should be noted that the proportion of patients achieving remission or response following 6-thioguanine is similar to that after methotrexate13 or infliximab.14

One reason for the low remission rate achieved with 6-thioguanine is probably the negative selection bias. It is important to note that only 41% (15/37) of patients were azathioprine naive prior to study entry, and, in the trial of Dubinsky et al.,9 all patients had received previous treatment with mercaptopurine. In the present study, the remission rate of azathioprine-naive patients reached 47%, whereas none of the azathioprine-refractory patients achieved a clinical response with 6-thioguanine. Therefore, from our results, 6-thioguanine cannot be recommended for patients refractory to other thiopurines. In addition, in the present study, only one patient refractory to 6-thioguanine responded to dose escalation.

It was hoped that 6-thioguanine would have a more rapid onset of action in responders and, in this regard, the results are very encouraging. All but one patient achieving remission did so within the first 4 weeks after starting treatment; the median time to remission was 29 days. This is much more rapid than reported on average for standard thiopurines, with a mean time to remission of 3.1 months.4

Although thiopurines are generally well tolerated, toxicity may be a major problem. According to recent data, up to 28% of patients with inflammatory bowel disease stop azathioprine treatment due to intolerable side-effects.3 Interestingly, in the present study, the patients experiencing side-effects were not the same as those who were azathioprine intolerant prior to study entry. The majority (12/16) of these patients tolerated 6-thioguanine, which may therefore represent an alternative for this sub-group of patients.

In our study, non-specific side-effects, such as nausea, vomiting or headache, were frequent, although study withdrawal occurred in only three patients. Dermatological side-effects occurred in eight patients, with one moderate photoallergic reaction, although this has also been described for conventional thiopurines. One patient developed pancreatitis and was withdrawn from the trial.

Elevated liver enzymes during treatment with azathioprine or mercaptopurine have been reported in 2–3% of patients.3, 15, 16 Most of these recover on dose reduction or drug cessation. During treatment with 6-thioguanine, two patients experienced mild elevation of liver enzymes, which normalized without dose reduction. Thiopurine-induced liver toxicity seems to be correlated at least in part to high levels of 6-methyl-mercaptopurine,6 which is the TPMT-dependent methylated product of mercaptopurine, but is not a metabolite of 6-thioguanine.

Bone marrow toxicity occurs in about 2–5% of patients treated with azathioprine or mercaptopurine.3, 15, 17 Severe haematological side-effects have been associated with a low TPMT activity, resulting in toxic levels of 6-thioguanine nucleotides.5 In the present study, three patients experienced myelosuppression. Two had intermediate TPMT activity and all patients were additionally treated with aminosalicylates; these are known to reversibly inhibit enzyme activity,18, 19 and cause leucopenia in combination with thiopurines.20

Few data are available on the long-term effects of 6-thioguanine, and virtually none in patients with Crohn's disease. Single reports exist on the development of non-cirrhotic portal hypertension following maintenance 6-thioguanine therapy of chronic myeloid and acute lymphoblastic leukaemia.21–23 Only one of our patients, in whom 6-thioguanine had been continued after the trial for another 6 months, developed elevated liver enzymes as well as thrombocytopenia. Nevertheless, despite this overall moderate toxicity, we suggest that the use of 6-thioguanine should be further evaluated only in controlled trials.

In conclusion, 6-thioguanine seems to induce rapid remission in some patients with Crohn's disease. In azathioprine-intolerant but not azathioprine-refractory patients, 6-thioguanine may offer a new alternative with acceptable short-term toxicity.


This study was supported by Dr Falk-Pharma GmbH, Freiburg, the Robert-Bosch-Foundation, Stuttgart and, in part, by grant 01 GG 9846 from the Ministry of Education and Science, Germany.