Polymorphic Variant in GATA3 gene Is a Hallmark of PAR1-Deleted BCP-ALL and Associates with Poor Prognosis Among Pediatric Patients Treated with the BFM Backbone Protocols

Introduction

Germline variant at rs3824662 in GATA3 gene was associated with early response to treatment as well as relapse risk in childhood BCP-ALL treated with COG protocols. This effect resulted from the strong link between the GATA3 polymorphism and the presence of somatic defects in leukemic cells including IKZF1 deletions, CRLF2 rearrangements and JAK2 gene mutations, which promote aggressive course of the disease.

Aims

The study aimed to evaluate an association between GATA3 gene polymorphism and clinical and biological features of pediatric BCP-ALL treated with the BFM backbone protocols.

Methods

Between November 2010 and June 2016, 957 consecutive children with newly diagnosed BCP-ALL were enrolled into the study. 822 patients treated according to ALL-IC BFM2009 protocol (n=594) and ALL-IC BFM2002 protocol (n=228) in 15 centers of the Polish Pediatric Leukemia/Lymphoma Study Group were enrolled into the study (median age 4.5 yrs, median follow-up time 2.5 yrs). Patients with BCR-ABL1 and MLL gene rearrangements were excluded from the analysis. The rs3824662 polymorphism of the GATA3 gene was genotyped using TaqMan probes. GATA3 mRNA expression level in leukemic cell was evaluated in BCP-ALL cases using qPCR with FAM-MGB probes (n=136). Targeted copy number screening of selected 23 loci was performed using the P335-B2 SALSA MLPA kit (MRC-Holland, Netherlands) in n=807 available DNA leukemia samples. MRD at day 15 and 33 was measured by flow cytometry with EuroFlow 8-color antibody panels.

Results

In the study group genotypes distribution withinrs3824662 of GATA3 was as follows: AA: n=44 (5.4%); CA: n=266 (32.4%); CC: n=512 (62.3%). Median MRD15 and MRD33 were 0.31% and 0.001% respectively. IKZF1 deletion were found in 18%, PAX5 in 20%, PAR1 in 12%, CDKN2A in 24%, CDKN2B 20%, BTG1 in 6.3%, ETV6 in 21%, EBF1 in 4% and RB1 in 6% of cases.

Leukemic cells harbouring AA variant withinrs3824662 showed GATA3 mRNA expression 1.6 and 2.2 times higher compared to cells with CA and CC variants respectively, with the difference close to significant (ANOVA p=0.06).

The presence of AA variant was not related to any gene deletion apart from microdeletions of the pseudoautosomal region PAR1 (Xp22 and Yp11) which occurred more frequently in AA carriers as compared to CA and CC carriers (11/41 vs. 34/242 vs. 52/468, p=0.013).

We did not find any association of clinical features such as initial WBC, steroid response, sex and age at diagnosis among BCP-ALL patients with GATA3 genotype. However, AA carriers had a higher risk for MRD>10% at day 15 OR (95%CI)=3.93 (1.37-11.23) as well as MRD>0.01% at day 33, OR(95%CI)=2.95 (1.12-7.73) as compared CC carriers.

Cox model of survival analysis was done for variables with univariate significance of p<0.15 (risk group, sex). The results of Cox regression showedthe AA genotype remained associated with an increased risk of death regardless of risk group (assigned based on ALL-IC BFM09 protocol), HR(95%C)=2.95(1.12-7.73), P=0.028).

Conclusions

We showed that carriers of AA genotype in GATA3 are prone to develop PAR1 -deleted BCP-ALL. Moreover our study confirmed an association of GATA3 AA rs3824662 homozygosity with poor early response to treatment as well as risk of death among pediatric BCP-ALL patients treated with the BFM backbone protocols.