Genomic Classification of T-Cell Acute Lymphoblastic Leukemia: Current Paradigms and Future Biomarkers

Over the past 20 years, long-term cure rates in T-lymphoblastic leukemia/lymphoma (T-ALL) have improved from 60% to 65% to >80%, largely due to the intensification of chemotherapy regimens. In B-lymphoblastic leukemia/lymphoma (B-ALL), outcomes have been improved through risk stratification, integrating clinical and demographic features, response to therapy, and disease biology to identify patients more or less likely to respond to conventional treatment, allowing for the use of regimens adapted to relative risk. Unfortunately, in T-ALL, few features have been identified that can independently predict the risk of poor outcomes beyond response to therapy. Accordingly, most cooperative groups do not use disease biology for T-ALL risk stratification. Recently, several T-ALL genomic initiatives have improved understanding of disease biology, enhanced the ability to classify T-ALL into different biological subtypes, and identified genomic alterations that predict outcome independent of response to treatment. Among these biomarkers is a newly defined high-risk "ETP [early T-cell precursor]-like" transcriptional subtype, which reclassifies immature T-ALL based on transcriptomic rather than immunophenotypic features; a treatment-refractory "bone-marrow-progenitor-like" leukemia fraction enriched in relapsed/refractory T-ALL; and intron-mediated noncanonical NOTCH1 dysregulation. The integration of these new biomarkers into clinical treatment holds promise to inform rapid upfront risk stratification, support new targeted therapeutic approaches, and guide future validation and preclinical studies for high-risk T-ALL.