A Genome-Wide Association Study of Susceptibility to Acute Lymphoblastic Leukemia in Adolescents and Young Adults

Acute lymphoblastic leukemia (ALL) in adolescents and young adults (AYA) is characterized by distinct presenting features and inferior prognosis compared to pediatric ALL. Age, as a continuous variable, is negatively correlated with prognosis, in spite of risk-adapted combination chemotherapy. To better understand the etiology of ALL in this age group, we performed the first genome-wide association study (GWAS) to comprehensively examine germline single nucleotide polymorphisms (SNPs) for their association with susceptibility to B-ALL in AYAs.

In the discovery GWAS, we compared genotype frequency at 635,297 SNPs between 308 AYA ALL cases (age 16-39 years, treated on the Children’s Oncology Group [COG], the Alliance-Cancer and Leukemia Group B, Eastern Cooperative Oncology Group, MD Anderson Cancer Center, and St. Jude Children’s Research Hospital trials) vs. 6,661 non-ALL controls. The association between genotypes at each SNP and ALL susceptibility was tested by using a logistic regression model after adjusting for genetic ancestries to control for population stratification. SNPs that reached P≤ 5×10^-8 in the discovery GWAS were tested in an independent cohort of 82 AYA ALL cases from the COG protocols and 5,755 non-ALL controls.

We identified a single genome-wide significant susceptibility locus on 10p14 signified by two SNPs within the GATA3 gene: rs3824662, P=2.8x10^-10, odds ratio (OR)=1.77; rs3781093, P=3.2x10^-9, OR=1.73, both of which were validated in the replication cohort (P=1.9x10^-8 and P=4.3x10^-5, respectively). We also examined the association signals in AYAs for susceptibility loci previously identified in pediatric ALL: ARID5B, IKZF1 and PIP4K2A variants were nominally significant in AYAs in the discovery GWAS and/or in the replication analysis, whereas CEBPE or CDKN2A/CDKN2B were not significant. These results imply both similarities and differences in genetic predisposition to ALL between children and AYAs.

At the GATA3 locus, rs3824662 risk variant was over-represented in Philadelphia chromosome (Ph)-like ALL in AYAs (P=0.02), confirming our previous report of Ph-like ALL susceptibility variants in GATA3 (Nat Genet 45:1494). Importantly, even after excluding Ph-like cases, rs3824662 remained associated with the risk of developing ALL in AYAs, suggesting that the influence of the GATA3risk variant on ALL susceptibility in AYAs extends beyond the predisposition to Ph-like subtype.

We next examined the relationship between GATA3 risk allele frequency and age at diagnosis in a cohort of unselected childhood and adolescent ALL cases enrolled in the COG P9900 protocols (N=1,827). Dividing patients into four consecutives age groups (<5, 5-10, 10-15 and >15 years), we observed a clear progressive increase in the risk allele frequency at rs3824662 (P=6.29×10^-11) with increasing allelic odds ratio (i.e. relative risk of ALL conferred by each copy of the risk allele). This correlation between genotype and age was evident regardless of genetic ancestry, although the risk variant was more common among individuals with higher Native American ancestry. In contrast, the frequency of ALL susceptibility variant in ARID5B decreased progressively with increasing age at diagnosis (P=0.006), whereas PIP4K2A, CDKN2A/CDKN2B, IKZF1 and CEBPE variants were not related to age. Finally, we compared rs3824662 risk variant frequency by age in the COG P9900 protocols after stratifying the ALL cases into TCF3-PBX1, ETV6-RUNX1, hyperdiploid, MLL-rearranged and B-other. There was a trend that the risk allele was more frequent in cases older than 16 years compared to those below 16 in the five subtypes examined.

In conclusion, we have identified inherited GATA3 genetic variants that strongly influence ALL susceptibility in adolescent and young adults, indicating potential age-related differences in ALL biology.