Therapy-related myeloid leukemias (t-MLs) occur in as many as 5% to 10% of patients who are otherwise cured of a primary neoplasm with aggressive multimodal regimens. Many patients with t-MLs or therapy-related myelodysplastic syndrome (t-MDS) have previously received chemotherapeutic agents that alkylate DNA, such as bis-chloroethyl-nitrosourea (BCNU) and cyclophosphamide (CP).1 We and other investigators have shown CP to be mutagenic in preclinical studies both in vitro and in vivo.2
Temozolomide (TMZ) is a more recently developed alkylating agent that has been effective in the treatment of adult high-grade glioma and refractory leukemia.3,4 TMZ is now being incorporated into initial therapy in more than 40 studies for a range of cancers, including glioblastoma and melanoma.5 TMZ's cytotoxicity depends on the methylation of guanine bases at the O6 position, resulting in O6-methylguanine and G:C→A:T transitions.6 Despite initial hopes that TMZ would be less leukemogenic than traditional alkylating agents, 2 groups have recently reported secondary myeloid malignancies after TMZ treatment in clinical studies.7,8 As TMZ moves into the front line of our chemotherapeutic armamentarium, further investigation of its in vivo mutagenic potential is warranted.
We used a transgenic mutation indicator mouse strain (small blue mouse) to compare the in vivo mutagenic potential of TMZ on bone marrow (BM) cells with that of CP. In this mouse model, the mutational target is the nontranscriptionally active lacZ portion of the plasmid pUR288. The mutation frequency was determined with a plasmid rescue procedure applied to genomic DNA derived from BM and with a subsequent selection for lac-Z negative clones, according to published protocols.9 The type of mutation was further determined by PCR amplification and restriction digestion.
Animals were treated with TMZ (175 mg/kg/d intraperitoneally for 5 days), CP (200 mg/kg intraperitoneally either once or weekly for 6 weeks), or phosphate-buffered saline (PBS), and BM was harvested 10 days after the last treatment (Figure 1). TMZ and CP doses were chosen by treating C57BL/6 mice in groups of 10 to 20 until the development of neutropenia without mortality. Determination of the mutation frequency revealed that the 1-day CP treatment increased the mutational load in BM 2-fold over the control, whereas the TMZ regimen resulted in a 22-fold increase over control. BM cells in animals treated 6 times with CP did not show an increase in the mutation frequency over animals treated with only a single dose of CP. As we expected from TMZ's mechanism of action, over 90% of all mutations in response to TMZ treatment were point mutations.10 Fewer than 30% of the mutations in BM cells from animals treated with CP were point mutations, with the remaining mutations being either translocations or deletions. These data emphasize TMZ's mutagenic potential for BM cells in vivo in the mouse model system and may indicate that TMZ's mutagenic potential is the underlying cause of the recently reported t-MLs in TMZ-treated patients. We suggest close long-term hematological monitoring of patients receiving TMZ in clinical trials. Further investigation is warranted to determine whether the increased mutation rate seen with TMZ exposure results in a comparable elevated risk of therapy-induced leukemia.
Generation of the data was supported by The Mouse Core of the Division of Experimental Hematology, the Somatic Mutation Frequency Core at the Cincinnati Children's Hospital Medical Center (CCHMC), and funds from the Translational Research Initiative at CCHMC. We thank Jan Vijg and Martijn Dollé for their support in establishing the mutation assay at CCHMC. TMZ was obtained from the National Cancer Institute (NCI) Developmental Therapeutics Program.