Base Excision Repair Gene OGG1 Affects the Relapse Risk of Acute Myeloid Leukemia

Background: Approximately 80% of acute myeloid leukemia (AML) patients can achieve complete remission, but around half of them relapse within five years. Recent studies have shown that AML relapse is associated with additional genetic mutation calls gclonal evolutionh in leukemic cell population (Ding, et al. Nature. 2012 & Parkin B, et al. Blood. 2013). These studies suggested that cytotoxic chemotherapy damaged cellular DNA and caused genetic mutation. In fact, anthracycline can produce 8-oxoguanine (8-OG) through induction of oxidative stress. 8-OG is most common DNA damage, which cause G:C to T:A transversion mutation. It is reported that transversion mutation more frequently observed in relapsed AML than primary AML. Base excision repair (BER) plays important role to correct base lesion including 8-OG and suppress genetic mutation. Therefore, we hypothesized BER gene polymorphisms may affect the risk of AML relapse, and focused five major functional polymorphisms: OGG1 S326C, MUTYH Q324H, APE1 D148E, XRCC1 R194W and XRCC1 R399Q. Material & Method: Ninety-four consecutive adults with AML who had achieved their first complete remission were recruited (male: 52, female: 42, age: 15-83 years, median age: 55.7 years). To remove the bias of the group, we also evaluated the risk in patients of non-M3 and under 65 years old (male: 22, female: 19, age: 15-64 years, median age: 49.0 years) (trimming group). These patients treated on Japan Adult Leukemia Study Group (JALSG) treatment protocols consisted of daunorubicin or idarubicin plus cytarabine (JALSG AML95, AML97, AML201, AML209). Genotyping was performed by PCR-RFLP method. The X2-test was used to compare the distribution of genotype and allele frequencies in patients. Leukemia-free survival (LFS) was calculated using the Kaplan-Meier method. Survival curves were compared using the log-rank test. In multivariate analysis, a stepwise selection procedure was performed using the proportional hazards Cox model for LFS. The variables were chosen with reference to previous studies; age, sex, white blood cell count and lactate dehydrogenase at diagnosis, number of induction courses, stem cell transplantation, MRC classification and history of tumor. This study was approved by the Institutional Review Board of Gunma University Hospital. Results: The OGG1 S326C CC genotype was observed significantly more often in the relapsed group (28.9% vs. 8.9%, OR = 4.10, 95% CI = 1.35-12.70, p = 0.01). In trimming group, the CC genotype was also observed more frequently in the relapsed group (50.0% vs. 6.9%, OR = 13.5, 95% CI = 2.17-84.0, p = 0.002). In addition, the OGG1 S326C CC genotype experienced a shorter median LFS than those with a non-CC genotype (CC vs. non-CC = 27.0 months vs. not reached, p = 0.02) (Figure 1). This genotype was also associated with poor LFS in trimming group (CC vs. non-CC = 11.0 months vs. not reached, p < 0.001) (Figure 1). Furthermore, multivariate analysis of LFS revealed OGG1 S326C CC genotype as an independent prognostic factor (HR = 4.32, 95% CI = 1.70-11.0, p = 0.002), like age, number of induction courses, and MRC classification (Table 1). Other polymorphisms had no significant effect on the risk of relapse. Conclusion: We previously reported that mutations by 8-OG were more efficiently suppressed in OGG1-S326 transduced cells than in OGG1-C326 transduced cells. Therefore, we hypothesized that low OGG1 activity promotes relapse of AML. To the best of our knowledge, this is the first report to show an association between BER gene polymorphisms and the relapse of AML. Our data suggest that OGG1 S326C can be a prognostic factor for AML relapse. Disclosures No relevant conflicts of interest to declare.