Background

The proto-oncogene ERG is located on chromosome 21q22 and overexpression of ERG was shown to have an adverse impact on outcome in cytogenetically normal AML (CN-AML). Acquired gain of chromosome 21 (+21) is a recurrent cytogenetic abnormality in AML, however, the pathogenetic impact remains elusive. Data indicate that +21 could be the underlying mechanism for altered expression of genes located in the respective regions, such as ERG, which might contribute to the pathogenesis in myeloid malignancies with +21.

Aims

1) To investigate a possible relationship between ERG copy numbers and variations in ERG mRNA expression levels in patients (pts) with myeloid malignancies. 2) Analysis of ERG expression in CN-AML to reveal a possible association between ERG overexpression and other relevant molecular markers and to ascertain the prognostic impact of ERGin context of these markers.

Methods

ERG mRNA expression and ERG copy numbers were analyzed using a hydrolysis probe based real-time PCR assay. 1) To address ERG expression in relation to +21 the following cohorts were analyzed: 62 AML pts with a complex karyotype including +21 (CK+21); 44 AML pts with a non-complex aberrant karyotype including +21 (AK+21); 19 pts with various myeloid malignancies (10 de novo AML, 3 MDS, 1 s-AML, 2 t-AML, 1 MDS/AML, 1 t-MDS, 1 MDS/MPN) all of them showing ERG amplification by interphase FISH and array CGH (ERG-Amp). In addition, 32 CK-AML and 330 CN-AML pts without +21 or ERGamplification were analyzed. Results were expressed as mean values±SEM. Expression levels/DNA copy numbers were compared by t-test. 2)ERG expression was assessed in a cohort of 330 patients (<65 years) with de novo CN-AML. Female/male ratio was 169/161. To distinguish low from high ERG expressers the median %ERG/ABL1 level was used. BAALC expression levels were analyzed accordingly. Expression levels were correlated with clinical outcome and with the presence of mutations (mut) in ASXL1 (n=330), CEBPA (n=330), DNMT3A (n=261), FLT3-TKD (n=330), IDH1R132 (n=328), IDH2R140 (n=328), IDH2R172 (n=328), NPM1 (n=330), NRAS (n=330), RUNX1 (n=329), TET2 (n=166), WT1 (n=329) and FLT3-ITD (n=330), and with MLL-PTD (n=330) and BAALCexpression (n=328).

Results

1) Analysis of pts with +21 or ERG-Amp revealed significantly higher expression levels of ERG in pts with +21 (AK+21 and CK+21 combined, 331 ± 28) and ERG-Amp (606 ± 127) as compared to pts with CN-AML (229 ± 10; p=0.001, p=0.008, respectively) or CK-AML (177 ± 36; p=0.001, p=0.004, respectively). Mean ERG expression was even higher in pts with ERG-Amp (606 ± 127) as compared to pts with +21 (331 ± 28, p=0.047). Quantification of ERG copy numbers on DNA level showed good correlation to ERG mRNA expression, with abundantly higher ERG DNA amount in +21 (3.25 ±0.14) and moreover in ERG-Amp (9.21 ±1.05) as compared to CN-AML and CK-AML (combined: 2.20 ± 0.05; p<0.001 and p<0.001, respectively). 2) In CN-AML, %ERG/ABL1 levels ranged from 0.078 to 1,016.027 (median: 188.904). High ERG expression was associated with lower age (mean: 48.7 vs. 52.4 years, p=0.002), higher white blood cell (WBC) count (mean: 64.8 vs. 44.4 x 109/L, p=0.019), FLT3-ITDmut/wt ratio≥0.5 (51/165, 30.9% vs. 24/165, 14.5%, p=0.001), IDH2R172mut (6/164, 3.7% vs. 0/164, 0.0%, p=0.030) and high BAALC expression (104/164, 63.4% vs. 59/164, 36.0%, p<0.001), as compared to low ERG expression. In contrast, NPM1mut (91/165, 55.2% vs. 118/165, 71.5%, p=0.003), IDH1R132mut (10/165, 6.1% vs. 29/163, 17.8%, p=0.001) and TET2mut (10/76, 13.2% vs. 22/90, 24.4%, p=0.077) were less frequent in high ERG expressers. Survival analysis revealed inferior overall survival (OS at 3 years: 52.2% vs. 68.7%, p=0.021) and event free survival (EFS at 3 years: 34.6% vs. 43.1%, p=0.052) for high ERG expressers as compared to low ERG expressers. In a multivariate analysis adjusted for age, WBC, BAALC expression, FLT3-ITDmut/wt ratio≥0.5, MLL-PTD and WT1mut, high ERG expression revealed a trend towards an adverse impact on OS (p=0.069), while no impact on EFS was observed.

Conclusions

1) Gain of chromosome 21 and especially amplification of chromosomal band 21q22 is a mechanism for ERG overexpression. This might indicate ERG as an important factor contributing to the pathogenesis and progression of myeloid malignancies with gain of chromosome 21. 2) In CN-AML, ERG overexpression is associated to several molecular markers and has a negative impact on OS and EFS.

Disclosures:

Weber:MLL Munich Leukemia Laboratory: Employment. Haferlach:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Noel:MLL Munich Leukemia Laboratory: Employment. Alpermann:MLL Munich Leukemia Laboratory: Employment. Kern:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Haferlach:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Schnittger:MLL Munich Leukemia Laboratory: Employment, Equity Ownership.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution