Expression of RUNX1 and RUNX1T1 Correlates with Cytogenetic Subgroups in Pediatric AML: A Report From the Children's Oncology Group

Abstract

AbstractAbstract 1425 Background:The t(8;21) translocation results in the fusion oncogene RUNX1-RUNX1T1 (also called AML1-ETO) and confers a favorable prognosis in both pediatric and adult AML. Little is known about the expression of RUNX1 and its translocation partner RUNX1T1 in myeloid leukemia. In this study we examined the mRNA expression of both genes in children with newly diagnosed AML and correlated patient clinical features with gene expression. Methods:We isolated mRNA from diagnostic specimens of 206 patients enrolled on COG AAML03P1. RUNX1 and RUNX1T1 mRNA expression was measured using qRT-PCR. GUSB gene served as control for mRNA quality and to standardize expression. Expression was correlated with patient and disease characteristics and clinical outcomes. Results:Relative RUNX1 expression ranged from 0.09 to 11.32 with a median of 1.45. The patient population was divided into quartiles (n=51, 52, 51, 52) with quartile 1 (Q1) having the lowest expression and quartile 4 (Q4) the highest expression. There was no significant difference in sex, age, race, ethnicity, hematologic parameters at diagnosis, or CNS status. Cytogenetic groups t(8;21), inv(16), abnormal 11q23, and monosomy 7 differed significantly in RUNX1 expression. The percentage of patients with t(8;21) decreased for increasing quartiles of RUNX1 expression with 28% of Q1 having t(8;21) compared to 12% of Q2, 8% of Q3 and 0% of Q4 (p=<0.001). Conversely, the percentage of patients with inv(16) increased for increasing quartiles of RUNX1 expression with only 2% of Q1 having inv(16) compared to 12% of Q2, 27% of Q3 and 20% of Q4 (p=0.006). Percentage of patients with abnormal 11q23 decreased for increasing quartiles of RUNX1 expression (Q1=36%, Q2=22%, Q3=15% and Q4=7%, p=0.003). All 4 patients with monosomy 7 had RUNX1 expression levels in Q4 (p=0.005). RUNX1 expression did not differ by FLT3-ITD, CEBPa mutant and NPM mutant. On recent COG studies, low risk includes core binding factor (CBF) patients and CEBPa or NPM mutant, high risk -5/5q-, −7 or FLT3-ITD high AR, and standard risk being all others. There was no difference in distribution of low or standard risk patients. The high risk group had higher RUNX1 expression attributable to the high expression in monosomy 7 (p=0.027). There was no difference across quartiles for rates of CR, 5 yr EFS or OS.Relative RUNX1T1 expression had a wide range from 0 to 139,786 with 69% of patients having no detectable expression. Due to the number of patients with no expression and the logarithmic distribution of expression, we divided the cohort into no expression (“exp=0”, n=143), expression levels >0 to 1 (“exp>0–1”, n=23) and expression >1 (“exp>1”, n=40). Among these 3 groups, there was no difference in sex, age, race, ethnicity, or CNS status. WBC at diagnosis was higher (median 38,000/μL) for exp=0 compared to exp>0–1 (median 14,100/μL) and exp>1 (median 15,800/μL) (p=0.045). RUNX1T1 expression differed significantly in patients with t(8;21) and monosomy 7. Patients with t(8;21) made up a larger proportion of the high RUNX1T1 expression group including 35% (13/37) having exp>1 compared to 0% of patients with exp>0–1 and 7% (10/134) with exp=0 (p=<0.001). Inv(16) patients showed trend toward the lower RUNX1T1 expression group but not significantly (p=0.11). Monosomy 7 patients clustered in the higher expression group accounting for 8% (3/37) of patients with exp>1 compared to 0% with exp>0–1 and 1% (1/134) with exp=0 (p=0.017). Expression did not differ in the molecular groups or risk groupings. Rates for CR, OS and EFS did not differ significantly by RUNX1T1 expression. However, there was a trend toward higher relapse risk among those with higher RUNX1T1 expression in standard risk patients (exp=0, RR=34±14%; exp>0–1, RR=28±33%; exp >1, RR=68%±33; p=0.138). Conclusion:Both RUNX1 and RUNX1T1 expression correlate with cytogenetic subgrouping, particularly in CBF and monosomy 7. Patients with inv(16) and monosomy 7 make up a larger percentage of the high RUNX1 expression patients but t(8;21) patients tend to be in the low expression groups. For RUNX1T1, t(8;21) and monosomy 7 patients had higher percentages in the high expression group, while inv(16) patients showed a trend toward the lower expression group. RUNX1 or RUNX1T1 expression levels alone did not predict OS or EFS, but further study is warranted to understand the role of high RUNX1 and RUNX1T1 expression in the rare high risk pediatric patients with monosomy 7. Disclosures:No relevant conflicts of interest to declare.