Journal of Clinical Pharmacy and Therapeutics (2010) 35, 93–97 doi:10.1111/j.1365-2710.2009.01041.

PHARMACOGENETICS

Thiopurine S-methyltransferase genotype and the use of

thiopurines in paediatric inflammatory bowel disease
Greek patients
M. Gazouli* PhD , I. Pachoula MD , I. Panayotou MD , G. Mantzaris MD PhD ,
V. P. Syriopoulou MD PhD , N. Goutas§ MD PhD , D. Vlachodimitropoulos§ MD PhD ,
N. P. Anagnou*– MD PhD and E. Roma-Giannikou MD PhD
*Department of Biology, School of Medicine, University of Athens, Athens, First Department of Pediatrics,
‘‘Aghia Sophia’’ Children’s Hospital, School of Medicine, University of Athens, Athens, Department of
Gastroenterology, Evagelismos Hospital, Athens, Greece, §Department of Pathology, School of Medicine,
University of Athens, Athens and –Laboratory of Cell and Gene Therapy, Center of Basic Research II,
Foundation for Biomedical Research of the Academy of Athens (IIBEAA), Athens, Greece

Conclusions: In this small cohort of subjects, no

SUMMARY
association was found between TPMT polymor-
Background and objective: Azathioprine (AZA) phisms and the occurrence of thiopurines-related
and 6-mercaptopurine (6MP) are used in the adverse events.
treatment of paediatric inflammatory bowel
disease (IBD). Genetic variations in thiopurine Keywords: 6MP, azathioprine, inflammatory
S-methyltranfarase (TPMT) gene have been bowel disease, TPMT genotype
correlated with enzyme activity and with the
occurrence of adverse events to AZA and 6MP.
The aim of the present study was to investigate INTRODUCTION
the frequency of the functional TPMT polymor-
Azathioprine (AZA) and 6-mercaptopurine (6MP)
phisms and their association with the occurrence
are widely used as immunosuppressive drugs in
of adverse events during azathioprine therapy in
the treatment of children with inflammatory bowel
a paediatric IBD cohort.
disease (IBD) [Crohn’s disease (CD) and ulcerative
Methods: Ninety-seven thiopurine-treated paedi-
colitis (UC)]. However, it is known that a high
atric IBD patients (41Æ24% boys and 58Æ76% girls)
percentage of patients develop adverse effects,
with a mean age 11Æ25 years (range 3–16), were
such as leucopoenia, bone marrow suppression,
assessed for TPMT polymorphisms and adverse
hepatotoxicity and pancreatitis (1–3). Thiopurines,
events.
through a series of enzymatic steps, are converted
Results: Of the 97 patients enrolled in the study,
into the active compound 6-thioguanine (6TGN).
18 (18Æ56%) were heterozygous mutated; two
Clinical response to AZA and 6MP treatment cor-
(2Æ06%) were homozygous for a mutated TPMT
relates with the levels of 6TGN. However, high
gene. Ten patients (10Æ31%) developed adverse
6TGN levels are also associated with increased
effects, and four of them (40%) had one of the
toxicity (4). The thiopurine S-methyltrasferase
variant alleles.
(TPMT) enzyme plays a role in the inactivation of
6TGN and determines the final levels of the 6TGN
and of the hepatotoxic metabolite, methylmercap-
Received 26 September 2008, Accepted 09 February 2009 topurine (4).
Correspondence: Dr E. Roma-Giannikou, First Department of
Thiopurine S-methyltranfarase enzyme activity
Pediatrics, ‘‘Aghia Sophia’’ Children’s Hospital, School of
Medicine, University of Athens, Thivon and Papadiamantopoulou, is largely influenced by polymorphisms in the
11527 Goudi, Athens, Greece. Tel.: + 30 210 7467892; fax: + 30 210 TPMT gene. Over 23 polymorphisms in the TPMT
7462231; e-mail: roma2el@med.uoa.gr gene have been associated with a decrease in

 2009 Blackwell Publishing Ltd 93

94 M. Gazouli et al.
enzyme activity (5). Three polymorphic alleles,
TPMT*2 (a single nucleotide transversion in exon 5,
G238C), TPMT*3A (two transition mutations, one
in exon 7, G460A and the other in exon 10, A719G)
and TPMT*3C (a single transition mutation in exon
10, A719G), account for 80–95% of intermediate or
low activity populations (5). Even if the overall
prevalence of TPMT deficiency is similar among
different ethnic groups, the frequency of variant
alleles differs between different populations. For
example, there is a higher representation of the
TPMT*3A allele in South Asians compared with
the high frequency of the TPMT*3C allele in East
and West Africans (6). The TPMT*1 (wild-type
allele, 96%) and TPMT*3 (G460A and ⁄ or A719G,
4%) alleles are the most common in Caucasians (7).
It has been reported that patients with inherited
TPMT deficiency treated with standard doses of
thiopurines present with higher levels of the thio-
guanine active metabolites and have an increased
risk for adverse events. Unless patients with two
defective alleles are treated with 10- to 15-fold
lower doses of this medication, potentially fatal,
haematopoietic toxicity, which requires immediate
discontinuation of treatment may follow (8–10).
Whether carriers of TPMT polymorphisms are
more likely to develop adverse effects, such as
leucopoenia, bone marrow suppression, hepato-
toxicity or pancreatitis compared with those with
the wild-type alleles, is still controversial and more
studies are needed to establish treatment recom-
mendations, based upon TPMT genotype (11).
Therefore, as no report on TPMT polymorphisms
in the Greek population, is available, the aim of the
present study was to investigate it and any rela-
tionship between adverse effects of treatment with
AZA or 6MP in paediatric IBD patients.
PATIENTS AND METHODS
Patients with IBD who presented between Febru-
ary 2007 and August 2008 at the First Department
of Pediatrics of the ‘Aghia Sophia’ Children’s
Hospital were consecutively enrolled. IBD was
diagnosed based on clinical, endoscopic, radiolog-
ical and histological criteria (12). The study inclu-
ded patients who had been taking AZA or 6MP for
at least 3 months or who had experienced adverse
effects during treatment with these drugs. Thiop-
urine dosage had to be 0Æ3–2 mg ⁄ kg. Bone marrow
 2009 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 35, 93–97
suppression was defined as leucopoenia (WBC
<3000 ⁄ mm3) and ⁄ or thrombocytopenia (platelets
<100 000 ⁄ mm3), hepatotoxicity by serum alanine
transaminase levels greater than twice of the upper
normal limit resolving after withdrawal of thiop-
urine drug, and pancreatitis by severe abdominal
pain and hyperamylasemia resolving after with-
drawal of thiopurine drug. The study was
approved by the Ethics Committees of the partici-
pant centres.
For genotype analysis, venous blood samples
(2 mL from each paediatric patient) were collected.
Genomic screening was accomplished by a poly-
merase chain reaction (PCR) and restriction frag-
ment-length polymorphisms assay as previously
described (13). DNA of the patients was screened
for TPMT*3A (both G460A and A719G mutation),
TPMT*3B (only G460A mutation), TPMT*3C (only
A719G mutation) and TPMT*2 (G238C mutation).
Differences in allele frequencies were compared
with the chi-square test (GRAPHPAD V. 3.00; GraphPad
Software, San Diego, CA, USA). The association
between adverse effects of TPMT polymorphisms
was tested using two-sided Fisher’s exact test, and
compared using the odds ratios and 95% confi-
dence intervals. Strong association (significance)
was assumed at P < 0Æ05.
RESULTS
The clinical data of the 97 patients enrolled in the
study are presented in Table 1. Among the 97
patients enrolled in the study, 18 (18Æ56%) were
heterozygous mutated; two (2Æ06%) were homo-
zygous for a mutated TPMT gene (Table 2). In 13 of
97 patients (13Æ40%), the thiopurine drug (AZA or
6MP) was stopped. In 10 patients (10Æ31%), this
was because of adverse events, whereas for the
other three patients (3Æ09%) it was insufficient
response. Adverse events consisted of leucopoenia
(n = 6, three of which were transient), pancreatitis
(n = 3) and bone marrow aplasia (n = 1). The three
cases of transient leucopoenia were resolved with
dose reduction; all other adverse effects resolved
with drug withdrawal. Four of the 10 patients
(40%) who developed adverse events had mutated
TPMT gene. None of the subjects who developed
pancreatitis and transient leucopoenia was carrier
for a mutated TPMT genotype. The three patients
who had an insufficient response to thiopurines
 TPMT genotype and thiopurine treatment 95

Table 1. Patient characteristics

Patients with Patients without All patients
adverse effect adverse effects (n = 97)
(n = 10) (n = 87)

Median age [years (range)] 12Æ5 [3–14] 13Æ6 [3–16] 11Æ25 [3–16]
Gender (male; female) 5; 5 35; 52 40; 57
IBD diagnosis (CD; UC; IC) 6; 3; 1 63; 14; 10 69; 17; 11
Thiopurine used (AZA; 6MP) 8; 2 82; 5 90; 7
Median thiopurine dose 1Æ13 [0Æ3–2] 1Æ41 [0Æ5–2] 1Æ38 [0Æ3–2]
[mg ⁄ kg ⁄ day (range)]
Median length of treatment 31Æ9 [2–84] 20Æ24 [0Æ5–60] 22Æ3 [0Æ5–84]
[months (range)]

CD, Crohn’s disease; UC, ulcerative colitis; IC, indeterminate colitis; IBD, inflammatory
bowel disease.

Table 2. Polymorphic alleles of TPMT gene and their treated with conventional doses of thiopurine
frequency among 97 paediatric inflammatory bowel drugs (9). Additionally, 10–30% of patients cannot
disease (IBD) patients (n = 194 alleles) tolerate thiopurine therapy because of other effects,
such as hepatotoxicity, pancreatitis, influenza-like
Homozygotes ⁄ symptoms, nausea and other side-effects (15, 16).
Allelic heterozygotes
Among the 97 patients examined in the present
variants n % (n)
study, two subjects were homozygous for a TPMT
TPMT*1 172 88Æ66 77 ⁄ 18 mutation (2Æ06%) and 18 (18Æ56%) were heterozy-
TPMT*2 6 3Æ19 0⁄6 gous. We found higher TPMT allele frequencies than
TPMT*3A 5 2Æ66 0⁄5 earlier studies (14, 17), this is probably because of the
TPMT*3B 9 4Æ78 2⁄5 limited number of patients tested and hence sam-
TPMT*3C 2 1Æ06 0⁄2 pling varability. TPMT allele frequencies are not
known in the general Greek population.
carried the wild type TPMT gene. All 16 patients There have been several studies on the relation-
with mutated TPMT gene who did not develop ship between reduced TPMT enzymatic activity
adverse events responded promptly to therapy; and the serious side-effects of thiopurines (14).
the median dose was 1 mg ⁄ kg ⁄ day (range Some have shown a significant increase in occur-
0Æ5–2 mg ⁄ kg ⁄ day) and lower than the one used in rence of leucopoenia and bone marrow toxicity in
the 81 responder patients with a wild type geno- carriers of a TPMT gene polymorphism, but other
type (1Æ52 mg ⁄ kg ⁄ day, range 1–2 mg ⁄ kg ⁄ day, side effects (i.e. hepatotoxicity, pancreatitis)
P = 0Æ0001). appeared not to be related to TPMT polymor-
phisms (18). In our series of patients, only those
who developed leucopoenia and bone marrow
DISCUSSION
aplasia carried a mutated TPMT gene. However,
Despite the clearly beneficial effects of AZA and overall no significant association between TPMT
6MP in the management of IBD, therapy is often polymorphisms and adverse effects of thiopurines
limited by common thiopurine-related adverse was found in the present study, in agreement with
effects. Individual differences have been observed previous studies (17, 19) suggesting the involve-
in tolerance of thiopurines, and attributed to dif- ment of factors other than a mutated TPMT gene.
ferent intracellular concentrations of 6TGN, a Indeed, recently, Fankoury et al. (20) reported that
cytotoxic metabolite of the thiopurines (14). It is apart from genotype, environmental factors are
well recognized that patients who inherit TPMT also important in influencing TPMT activity and
deficiency are at greater risk of severe and poten- for the observed intra-individual variability in
tially life-threatening hematopoietic toxicity when response seen in patients receiving thiopurines.

 2009 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 35, 93–97
96 M. Gazouli et al.
Those authors also suggested that measurement of
TMPT activity and ⁄ or determination of genotype
remain important before treatment to identify
patients totally deficient in the enzyme and to
protect them from adverse effects.
Interestingly, among the 81 responders to ther-
apy and with no side-effects, the 16 patients with
a mutated TPMT gene were treated with a lower
dose of thiopurine drug than patients who carried
the wild type gene. This finding supports the
suggestion of Stocco et al. (17) that carriers of
TPMT mutations respond to thiopurines at a
lower dose than those with a normal TPMT
genotype.
CONCLUSION
In this small cohort of subjects, no association was
found between TPMT polymorphisms and the
occurrence of thiopurines-related adverse events.
However, our findings are consistent with the
results of previous studies (14, 17), and suggest
that treatment for paediatric IBD patients with
wild-type TPMT can be started at a conventional
dose (2–2Æ5 mg ⁄ kg ⁄ day), whereas those who are
carriers of mutated TPMT may require a lower
dose (1–1Æ6 mg ⁄ kg ⁄ day). If, there is no clinical
response a different immunosuppressive drug
should be used.
REFERENCES
1. Fuentes D, Torrente F, Keady S et al. (2003) High-
dose azathioprine in children with inflammatory
bowel disease. Alimentary Pharmacology and Thera-
peutics, 177, 913–921.
2. Barabino A, Torrente F, Ventura A et al. (2002)
Azathioprine in paediatric inflammatory bowel
disease: an Italian multicentre survey. Alimentary
Pharmacology and Therapeutics, 16, 1125–1130.
3. Fraser AG, Orchard TR, Jewell DP (2002) The
efficacy of azathioprine for the treatment of
inflammatory bowel disease: a 30 year review. Gut,
50, 485–489.
4. Cuffari C, Hunt S, Bayless T (2001) Utilisation
of erythrocyte 6-thioguanine metabolite levels to
optimise azathioprine therapy in patients with
inflammatory bowel disease. Gut, 48, 642–646.
5. Schaeffeler E, Eichelbaum M, Reinisch W, Zanger
UM, Schwab M (2006) Three novel thiopurine
S-methyltransferase allelic variants (TPMT*20, *21,
 2009 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 35, 93–97
*22) – association with decreased enzyme function.
Human Mutation, 27, 976.
6. McLeod HL, Siva C (2002) The thiopurine S-meth-
yltransferase gene locus-implications for clinical
pharmacogenomics. Pharmacogenomics, 3, 89–98.
7. Gearry RB, Barclay ML, Burt MJ et al. (2003) Thiop-
urine S-methyltransferase (TPMT) genotype does not
predict adverse drug reactions to thiopurine drugs in
patients with inflammatory bowel disease. Alimen-
tary Pharmacology and Therapeutics, 18, 395–400.
8. Evans WE, Horner M, Chu YQ, Kalwinsky D, Roberts
WM (1991) Altered mercaptopurine metabolism, toxic
effects, and dosage requirement in a thiopurine
methyltransferase-deficient child with acute lym-
phocytic leukaemia. Journal of Pediatrics, 119, 985–989.
9. Evans WE, Hon YY, Bomgaars L et al. (2001) Pre-
ponderance of thiopurine S-methyltransferase defi-
ciency and heterozygosity among patients intolerant
to mercaptopurine or azathioprine. Journal of Clinical
Oncology, 19, 2293–2301.
10. Evans WE (2004) Pharmacogenetics of thiopurine
S-methyltransferase and thiopurine therapy. Thera-
peutic Drug Monitoring, 26, 186–191.
11. Aberra FN, Lichtenstein GR (2005) Review article:
monitoring of immunomodulators in inflammatory
bowel disease. Alimentary Pharmacology and Thera-
peutics, 21, 307–319.
12. Lennard-Jones JE (1989) Classification of inflamma-
tory bowel disease. Scandinavian Journal of Gastroen-
terology Supplement, 170, 2–6.
13. Hiratsuka M, Inoue T, Omori F, Agatsuma Y, Miz-
ugaki M (2000) Genetic analysis of thiopurine
methyltransferase polymorphism in a Japanese
population. Mutation Research, 448, 91–95.
14. Al Hadithy AF, de Boer NK, Derijks LJ, Escher JC,
Mulder CJ, Brouwers JR (2005) Thiopurines in
inflammatory bowel disease: pharmacogenetics,
theurapeutic drug monitoring and clinical recom-
mendations. Digestive Liver Disease, 37, 282–297.
15. Sanderson J, Ansari A, Marinaki T, Duley J (2004)
Thiopurine methyltransferase: should it be mea-
sured before commencing thiopurine drug therapy?
Annals of Clinical Biochemistry, 41, 294–302.
16. Kirschenr BS (1998) Safety of azathioprine and
6-mercaptopurine in paediatric patients with inflam-
matory bowel disease. Gastroenterology, 115, 813–821.
17. Stocco G, Martelossi S, Barabino A et al. (2005) TPMT
genotype and the use of thiopurine in paediatric
inflammatory bowel disease. Digestive Liver Disease,
37, 940–945.
18. Schwab M, Schäffeler E, Marx C et al. (2002) Aza-
thioprine therapy and adverse drug reactions in
patients with inflammatory bowel disease: impact of
 TPMT genotype and thiopurine treatment 97

thiopurine S-methyltransferase polymorphism. 20. Fakhoury M, Gallien-Andreu J, Mahr A et al. (2007)

Pharmacogenetics, 12, 429–436.
19. De Ridder L, Van Dieren JM, Van Deventer HJ et al.
(2006) Pharmacogenetics of thiopurine therapy in
paediatric IBD patients. Alimentary Pharmacology and
Therapeutics, 23, 1137–1141.
Should TPMT genotype and activity be used to
monito 6-mercaptopurine treatment in children with
acute lymphoblastic leukaemia? Journal of Clinical
Pharmacy Therapeutics, 32, 633–639.

 2009 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 35, 93–97