The DNMT3A R882 Mutation’s Impaired DNA-Damage Sensing: A Key to Overcoming Chemoresistance in AML

A study presented at the 2015 European Hematology Association Congress suggests an important role for DNMT3A R882 mutations in chemoresistance in acute myeloid leukemia (AML).

“Unfortunately, apart from some exceptional sub-types, the prognosis [for AML] hasn’t really changed for the last 30 years,” said lead study author Olga Guryanova, MD, PhD, from the Memorial Sloan Kettering Cancer Center in New York City. “Although the majority of patients respond to standard chemotherapy initially, it is believed that a small set of chemoresistant-cells will persist in the body, later giving rise to relapse.”

While low-level residual AML is seen in many patients even at clinical remission, the mechanisms underlying persistence of AML cells after chemotherapy are not fully understood. In a multivariate analysis of 398 patients with AML from the ECOG E1900 study, Dr. Guryanova and colleagues sought to determine which mutations, if any, were predictive of minimal residual disease (MRD) followed by relapse, and how such mutations promote resistance to chemotherapy.

Patients were younger than 60 years old and had received induction chemotherapy with cytarabine plus high-dose or standard-dose daunorubicin for the treatment of AML. MRD was assessed by flow cytometry at day 28 in patients with complete pathologic remission. In the trial, patients were followed for a median of 47.4 months.

“There was only one particular mutation that robustly correlated with the presence of MRD: a very specific mutation on the gene DNMT3A,” Dr. Guryanova reported. Presence of the DNMT3A R882 mutation was associated with a non-significant increased risk of MRD at the time of complete remission (p=0.07). By contrast, non-R882 mutations in DNMT3A did not predict for MRD followed by relapse – nor did any other recurrent AML genetic mutations.

After identifying the DNMT3A R882 mutation as a predictor of MRD followed by relapse, Dr. Guryanova and researchers then developed a novel mouse model to study its role in the pathogenesis and chemoresistance of AML.

By introducing DNMT3A-mutant hematopoietic stem cells into mice and treating them with daunorubicin, the researchers observed a survival advantage for DNMT3A-mutant cells compared with wild-type controls, with preserved potential for proliferation after in vitro and in vivo exposure to daunorubicin. Treating cells with chemotherapy induces DNA damage. In the case of DNMT3A R882 mutant cells, Dr. Guryanova explained, “The damage persists because the cell doesn’t know that those [DNA breaks] are there and is unable to repair them.” Because the mutant cells fail to survey their DNA for the signs of damage, the damaged cells survive and accumulate additional mutations after chemotherapy, but are not eliminated by apoptosis.

Introducing DNMT3A did not cause leukemia in mice; however, when combined with other mutations that are commonly present in AML (FLT3-ITD and NPM1), the disease was much more aggressive than with these mutations alone.

“Maybe this particular mutation doesn’t cause leukemia, but it makes really good stem cells that can outcompete all other stem cells in the body,” Dr. Guryanova said. This was particularly true under stress conditions, including when mice were exposed to chemotherapy.

“This defect in DNA damage repair means that maybe we can find novel ways to target this specific vulnerability,” Dr. Guryanova concluded. In the E1900 trial, she noted, patients with the DNMT3A R882 mutation who were resistant to standard-dose daunorubucin benefitted from the increased dose, while patients without this mutation saw increased toxicity but no additional benefit with the higher dose. “Now, we have a mechanistic basis for understanding why increased doses of chemotherapy in these patients would be beneficial.”

Dr. Guryanova also mentioned other therapeutic implications for this finding, including targeting other components of the DNA damage machinery or combining chemotherapy with BCL-2 inhibitors to promote apoptosis of damaged cells.


Guryanova O, Shank K, Garrett-Bakelman F, et al. DNMT3A R882 mutation promotes chemoresistance and therapeutic relapse through impaired DNA-damage sensing. Abstract #S473. Presented at the 2015 European Hematology Association, Vienna, Austria, June 12, 2015.