To the editor:

With great interest we have read the paper by Schmiegelow and colleagues describing the increased risk of developing a second malignant neoplasm, especially acute myeloid leukemia and myelodysplastic syndromes, due to maintenance therapy with 6-mercaptopurine (6-MP) and methotrexate after childhood acute lymphoblastic leukemia.1  An impaired thiopurine S-methyltransferase (TPMT) activity, high 6-thioguaninenucleotides (6-TGN) and elevated 6-methylmercaptopurine-ribonucleosides (6-MMPR) levels were found to be risk factors. An increasing number of relatively young patients with different autoimmune diseases (eg, inflammatory bowel disease) or after organ transplantations are treated with a life-long thiopurine regime. The exact role of thiopurines in developing hematologic malignancies is unknown, but it has been postulated that the inactivation of the DNA mismatch repair system may lead to an increased rate of spontaneous mutations.2 

Here we propose an additional theory from a more pharmacodynamic point of view. An intrinsic complication of thiopurine therapy is that this cytotoxic and apoptosis-inducing agent can target almost all human cells, particularly myeloid precursor cells, and may induce genetic alterations. Children with acute lymphocytic leukemia treated with chemotherapy, including 6-MP, developed a high number of mutations at the hypoxanthine-guanine phosphoribosyltransferase (HGPRT)–reporter gene.3  In addition, azathioprine therapy in insulin-dependent diabetes mellitus patients resulted in an increase of HGPRT T cell–mutant frequencies that was statistically correlated with duration of therapy.4  All clinically used thiopurines have to be metabolized by the enzyme HGPRT to become pharmacologically active. Impaired HGPRT activity due to mutations may therefore lead to a relative failure of therapy. More importantly, mutated cells are thereby thiopurine refractory and thus positively selected and amplified, as all other cells will become apoptotic due to therapy. This process of selection of mutated cells due to thiopurine administration may give rise to an increased risk of developing hematologic malignancies such as leukemia or myelodysplastic syndromes. The risk factors diminished TPMT activity, high 6-TGN, and 6-MMPR levels that were observed by Schmiegelow and colleagues fit in with this theory. All these pharmacodynamic aspects indicate at an increased rate of elimination of nonmutated cells due to high levels of the pharmacologically active thiopurine metabolites, which eventually leads to an accelerated selection of HGPRT-mutated cells. These patients are probably at risk to develop a hematologic malignancy earlier than patients with a more common and more prevalent type of thiopurine metabolism, that is, those patients using usual, average TPMT activity. Studies are warranted to further explore the field of genotoxicity and mutagenicity of thiopurine therapy, as more and more patients are being treated with a lifelong regime.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: D. P. van Asseldonk, MD, Department of Gastroenterology and Hepatology, VU University Medical Center, P O Box 7057, 1007 MB, Amsterdam, The Netherlands; e-mail: [email protected].

1
Schmiegelow
 
K
Al-Modhwahi
 
I
Andersen
 
MK
, et al. 
Methotrexate/6-mercaptopurine maintenance therapy influences the risk of a second malignant neoplasm after childhood acute lymphoblastic leukemia: results from the NOPHO ALL-92 study.
Blood
2009
, vol. 
113
 (pg. 
6077
-
6084
)
2
Karran
 
P
Attard
 
N
Thiopurines in current medical practice: molecular mechanisms and contributions to therapy-related cancer.
Nat Rev Cancer
2008
, vol. 
8
 (pg. 
24
-
36
)
3
Rice
 
SC
Vacek
 
P
Homans
 
AH
, et al. 
Genotoxicity of therapeutic intervention in children with acute lymphocytic leukemie.
Cancer Res
2004
, vol. 
164
 (pg. 
4464
-
4471
)
4
Falta
 
MT
Atkinson
 
MA
Allegretta
 
M
, et al. 
Azathioprine associated T-cell mutations in insulin dependent diabetes mellitus.
Scand J Immunol
2000
, vol. 
51
 (pg. 
626
-
633
)
Sign in via your Institution