The Influence of Genetic Ancestry on Disease Biology in Pediatric T-Cell Acute Lymphoblastic Leukemia

Introduction:

Pediatric T-cell Acute Lymphoblastic Leukemia (T-ALL) is a biologically heterogenous disease that can be grouped into 16 subtypes based on integration of gene expression and somatic genomic alterations that have variable prevalence of activated targetable pathways (Pölönen P., et al ASH 2022). In both adult cancers and pediatric B-cell ALL, genetic ancestry has been associated with tumor biology and clinical outcomes; however, there is a paucity of data exploring the influence of genetic ancestry in T-ALL. Thus, we sought to determine the association of genetic ancestry with tumor biology in pediatric T-ALL.

Methods:

Whole genome sequencing (germline and tumor) and bulk RNA-sequencing was performed to characterize T-ALL subtype and pathway aberrations among 1312 newly diagnosed T-ALL patients treated on Children's Oncology Group trial AALL0434. Genetic ancestral composition was determined for each individual using the iAdmix program comparing allele frequencies between patients and reference genomes from the 1000 Genomes Project. Coefficients from the 5 reference populations (European (EUR), African (AFR), East Asian (EAS), South Asian (SAS), and American Admixture (AA)) were assumed to sum to 100%. Categorical genetic ancestral groups were classified from prior methods: "European" (EUR >90%), "African" (AFR >70%), "East Asian" (EAS >90%), "South Asian" (SAS >70%), and "Amerindian" (AA >10% and AA greater than AFR), with the rest defined as "Other." We evaluated associations of genetic ancestry with T-ALL molecular subtypes and pathways. Ancestry was first examined as a categorical variable using Fisher's exact test for its association with subtypes and pathways. Ancestry was also analyzed as a continuous variable using logistic regression for association with pathways and multinomial regression for association with subtype (TAL1-RB subtype as reference). The effect of ancestry was summarized as odds ratios (OR) for every 25% increase in a non-European ancestry, with a concurrent 25% decrease in European ancestry and remaining ancestries fixed.

Results:

Categorical ancestral composition in this cohort was 59% European, 11% African, 15% Amerindian, 3% East Asian, 2% South Asian, 10% Other. The prevalence of T-ALL subtypes differed by genetic ancestry (Figure 1). For example, HOXA9 and BCL11B, which are linked to ETP T-ALL, were significantly enriched in children of African genetic ancestry as compared to European ancestry (HOXA9: 11% AFR vs. 6% EUR; BCL11B: 3% AFR vs. 1.5% EUR). ETP phenotype was also more common in children of African descent as compared with European (11.9% vs. 7.5%, p =0.031).

In analyses of ancestry as a continuous variable (with European as the reference), four subtypes and four pathways exhibited differences in odds of expression (Table 1). For every 25% increase in African ancestry, the odds of BCL11B subtype was 41% higher (p=0.034), odds of HOXA9 subtype was 26% higher (p=0.020), odds of MLLT10_TRA was 68% higher (p<0.001), and odds of tumor microenvironment (TME) aberrant subtype was 33% higher (p=0.027). For every 25% increase in African ancestry, patients were also 1.13-fold more likely to harbor JAK/STAT mutations (p=0.049) and 1.18-fold for MAPK/RAS pathway alterations (p=0.011), with decreasing odds of NOTCH mutations (OR=0.88, p=0.014) or ribosomal gene alterations (OR=0.69, p=0.003). For increasing East Asian ancestry, patients were also less likely to have NOTCH pathway alteration (OR=0.85, p=0.027), whereas there was no difference for increasing South Asian ancestry.

Conclusion:

Disease biology varies substantially among children of different genetic ancestries with T-ALL. Given that subtypes may inform future risk stratification, and pathways may guide novel treatment strategies, examining subtype and pathway expression for children of all ancestral backgrounds is critical. Important biological indicators of risk may be missed when studying patients of a single genetic ancestral group or when focusing on averages for the entire cohort without taking into consideration variation in genetic ancestry. Analyses of survival by genetic ancestry within genomic groups are ongoing.

Gupta:Jazz Pharmaceuticals: Honoraria. Raetz:Pfizer: Research Funding; BMS: Other: Data and Safety Monitoring Board. Hunger:Amgen: Current equity holder in private company; Amgen: Honoraria; Jazz: Honoraria; Servier: Honoraria. Mullighan:Abbvie: Research Funding; Pfizer: Research Funding; Amgen: Honoraria; Illumina: Honoraria; Consulting: Honoraria; BEAM: Honoraria; FAZE: Honoraria. Yang:Takeda: Research Funding. Teachey:BEAM Therapeutics: Consultancy; Sobi: Consultancy.