Abstract
Children of Hispanic/Latino ethnicity have the highest risk of acute lymphoblastic leukemia (ALL) in the U.S., with a 30-40% and >2-fold increased incidence than in non-Latino Whites and non-Latino Blacks, respectively, yet the underlying basis of this increased risk remains largely uncharacterized. We hypothesized that differences in the presence or frequencies of germline genetic risk alleles may contribute to these disparities. Here, we leveraged the largest-available multi-ancestry genome-wide association study (GWAS) for childhood ALL to identify population-specific signals. Interestingly, rather than identifying a novel genomic region, we identified and characterized a new independent signal near the IKZF1 gene, which has previously been shown to have independent variants that increase the risk for ALL.
Two independent associations in the IKZF1 gene locus were identified in prior European-centric or multi-ancestry GWAS of ALL. Using genome-wide SNP array data from 1,878 childhood ALL cases and 8,441 controls of self-reported Latino ethnicity from California, imputed with data from the TOPMed consortium, we identified a third independently associated variant in this locus (P = 4.67 x 10-11, odds ratio [OR] = 1.44). The newly identified risk allele is rare in European and African populations (frequency ~0.2% and ~3.3%, respectively), but is found commonly in Latinos (17.6%) and East Asians (18.8%) in the Genome Aggregation Database. We did not identify this risk locus in a GWAS of childhood ALL in non-Latino whites (1,162 cases and 57,341 controls), in which only two independent IKZF1 risk loci were detected. Among Latinos, the new risk allele was significantly associated with both global and local Indigenous American ancestry (P < 8.8 x 10-7), and had a larger effect on childhood ALL risk in Latinos with ≥1 copy of the Indigenous American haplotype (OR = 1.55, P = 1.98 x 10-8) than in those with zero haplotype copies (OR = 1.18, P = 0.59).
Two of the three IKZF1 risk loci in Latinos, including the new third risk allele, are tightly linked (R2>0.96) to two nearby SNPs downstream of IKZF1 that are only 26 bp apart and that reside within the same regulatory region that demonstrates selective accessibility in developing B lymphoid progenitors and precursors, most notably pro- and pre-B cell populations, but which is largely inaccessible across other hematopoietic cell states. This regulatory region shows selective interaction with the promoter of IKZF1 using promoter capture Hi-C data, demonstrating direct regulatory interactions. To delineate the functionality of this element, we have been performing a number of assays, including exogenous reporter assays and genome/ base editing-mediated perturbation to examine activity in an appropriate cellular context.
In summary, our study reveals a genetic basis for the long-standing observed ethnic and racial disparities in childhood ALL risk through genetic fine-mapping coupled to functional follow up. Critically, these findings suggest opportunities not only to understand the basis for some disparities in ALL risk, but also suggest avenues that could be pursued for preventive strategies targeting childhood ALL.
Disclosures
Ma:Bristol Myers Squibb: Consultancy. Sankaran:Sana Biotechnology: Consultancy; Ensoma: Consultancy; Forma Therapeutics: Consultancy; Branch Biosciences: Consultancy; Novartis: Consultancy; Cellarity: Consultancy.
Author notes
Asterisk with author names denotes non-ASH members.