Leukemia Subtype Once Deemed High Risk Gets a Genomic Reclassification

Early T-cell precursor acute lymphoblastic leukemia, a rare and biologically unusual form of T-cell leukemia, has spent more than a decade carrying a reputation it may not fully deserve. First identified in 2009 through gene expression profiling at St. Jude Children’s Research Hospital and the Italian pediatric oncology group AIEOP, the subtype was initially linked to dismal survival and induction failure. Yet a comprehensive new review published in Blood Research argues that the picture has changed dramatically: when treated with contemporary intensive chemotherapy, patients with this subtype fare far better than the original reports suggested, and the immunophenotypic label itself is being overtaken by a more precise genomic classification.

The subtype, commonly abbreviated ETP-ALL, was defined by a gene expression signature resembling that of early T-cell precursors in the thymus—immature thymocytes that have recently left the bone marrow and retain hematopoietic stem cell-like characteristics. In that landmark 2009 analysis of 139 consecutive patients enrolled in St. Jude Total Therapy Studies XIII, XIV, and XV between 1992 and 2006, with independent validation in 100 patients from the AIEOP ALL-2000 protocol, a cluster of 13 cases stood out. Those blasts co-expressed hematopoietic stem cell and myeloid markers while underexpressing genes driving T-lineage commitment, including CD1, CD3, CD4, CD8, RAG1, and NOTCH3, and overexpressing stem cell and myeloid programs such as CD44, CD34, KIT, GATA2, and CEBPA.

Translating that molecular profile into routine diagnostics, clinicians established a practical immunophenotypic definition using flow cytometry. ETP-ALL is characterized by positivity for CD7 and cytoplasmic CD3, absence of CD1a and CD8, weak or absent CD5 expression below 75 percent of normal T-cell levels, and co-expression of at least one myeloid or stem cell marker such as HLA-DR, CD13, CD33, CD34, or CD117. A related category, near-ETP, was later added for cases that otherwise fit the criteria but express higher CD5 levels. By these definitions, ETP-ALL accounts for 5.5 to 17.1 percent of pediatric T-ALL cases and 15.8 to 47.3 percent of adult cases, with near-ETP adding another 14 to 16 percent.

The initial prognosis was grim. In the St. Jude cohort, 10-year overall survival for ETP-ALL was just 19 percent compared with 84 percent for non-ETP T-ALL, and 10-year event-free survival was 22 percent versus 69 percent, both differences highly significant. The AIEOP cohort showed similarly stark gaps, with 5-year overall survival of 45 percent and event-free survival of 22 percent for ETP-ALL against 90 percent and 71 percent for conventional T-ALL. A Japanese study, TCCSG L99-15, reported a 4-year event-free survival of 40 percent in ETP-ALL compared with 70 percent in non-ETP disease, though overall survival did not differ significantly, likely because three of four relapsed patients were salvaged with allogeneic stem cell transplantation.

Contemporary prospective trials have substantially revised that outlook. In the UKALL 2003 trial, which included 222 patients with T-ALL, the 35 patients with ETP-ALL showed a non-significantly inferior 5-year event-free survival of 76.7 percent versus 84.6 percent and overall survival of 82.4 percent versus 90.9 percent compared with non-ETP cases. The authors concluded that ETP-ALL carried an intermediate-risk outcome that did not warrant experimental treatment or allogeneic transplantation in first remission. The largest prospective pediatric trial, COG AALL0434, enrolled 1,256 patients, including 145 with ETP-ALL, 209 with near-ETP, and 902 with non-ETP disease. Although ETP and near-ETP patients had higher rates of induction failure at 6.2 percent each compared with 1.2 percent for non-ETP, the three groups achieved comparable event-free and overall survival, with no significant differences between them.

That paradox—high induction failure but ultimately similar survival—points to the central role of minimal residual disease monitoring, which the review identifies as the most powerful prognostic tool for ETP-ALL. In the AALL0434 analysis, day-29 MRD levels between 0.01 and 0.1 percent did not predict inferior outcomes, and even day-29 MRD of 0.1 percent or higher was associated with worse outcomes only in near-ETP and non-ETP patients, not in ETP patients. Within the ETP group, only day-29 MRD of 10 percent or higher predicted inferior outcome, suggesting that ETP blasts clear more slowly during induction, likely reflecting relative glucocorticoid resistance and slower cytoreduction kinetics rather than ultimate treatment failure. End-of-consolidation MRD of 0.01 percent or higher, however, was highly predictive of treatment failure across the overall cohort, with 5-year event-free survival of 52.4 percent versus 80.9 percent for those below that threshold.

The most consequential shift, however, is genomic. In 2024, Pölönen and colleagues presented a classification derived from whole-genome, whole-exome, and whole-transcriptome sequencing of 1,309 patients treated in the COG AALL0434 trial, delineating 15 T-ALL subtypes with distinct drivers and oncogene expression patterns. The immunophenotypic ETP status was distributed across multiple genomic subtypes, confirming that immunophenotype alone is suboptimal for classification. Only about 70 percent of immunophenotypically defined ETP-ALL cases carry an ETP-like genomic signature, while roughly 28 percent of genomically defined ETP-like cases have a non-ETP immunophenotype. The ETP-like genomic group itself comprised nearly equal distributions of ETP, near-ETP, and non-ETP immunophenotypic cases.

Prognostic implications vary dramatically across these genomic subtypes. The ETP-like and SPI1-rearranged subtypes show the highest rates of MRD positivity at the end of induction and inferior event-free survival, whereas TAL1-, TLX1-, BCL11B-, and KMT2A-rearranged subtypes generally have favorable outcomes. An integrated risk framework combining genomic subtype with MRD status places ETP-like cases with KMT2A, SPI1, or LMO2 γδ-like features in a very high-risk category with 5-year event-free survival of 45 percent when MRD-positive and 58 percent when MRD-negative, while other ETP-like variants fall into low or very low-risk groups with event-free survival reaching 91 to 98 percent. Notably, genomic subtypes provided prognostic information independent of MRD: BCL11B-rearranged T-ALL and the ZFP36L2 ETP-like subtype had poor MRD response yet favorable outcomes, whereas the SPI1-rearranged subtype and patients with PI3K pathway mutations had favorable MRD responses yet poorer survival. These subtypes were derived from a single cohort without independent validation, and further studies are required to confirm their robustness.

Treatment recommendations reflect this evolving evidence base. Dexamethasone-based multi-agent induction following the Berlin-Frankfurt-Münster protocol, incorporating vincristine, dexamethasone, L-asparaginase, and an anthracycline, remains the standard approach, with preclinical data suggesting that relative glucocorticoid resistance in ETP-ALL is mediated through IL-7/JAK/STAT5 signaling that upregulates BCL-2. In the COG AALL0434 trial, nelarabine improved outcomes in non-ETP patients and showed a trend toward benefit in near-ETP, but patients with ETP did not show significant benefit, though the authors caution that ETP patients were more frequently removed from the study and underwent allogeneic transplantation. Escalating-dose Capizzi methotrexate without leucovorin rescue showed benefits across T-ALL regardless of ETP status. For patients achieving good MRD response, allogeneic hematopoietic stem cell transplantation in first complete remission is not required, but persistent MRD positivity after consolidation triggers a transplant recommendation, based in part on AALL1231 findings that very high-risk patients had 4-year event-free and overall survival both below 35 percent.

For relapsed or refractory disease, the review highlights several promising agents. Bortezomib, a proteasome inhibitor, achieved a second complete remission rate of approximately 68 percent in the COG AALL07P1 pilot study of relapsed or refractory T-ALL, though the successor AALL1231 trial found no frontline benefit. Daratumumab, an anti-CD38 monoclonal antibody, demonstrated antileukemic efficacy in 14 of 15 T-ALL patient-derived xenograft models including an ETP-ALL model, and the phase II DELPHINUS study reported an overall response rate of approximately 82.8 percent in pooled T-cell ALL at the end of cycle 2, with a median event-free survival of approximately 8.9 months in childhood T-ALL. Venetoclax and navitoclax exploit the dependence of ETP-ALL on BCL-2 signaling: a phase I navitoclax-based reinduction study in 47 patients reported an overall complete remission rate of 60 percent, with five of seven ETP-ALL patients achieving complete remission, while a separate trial of venetoclax with homoharringtonine reported complete remission or remission with incomplete hematological recovery in all 18 enrolled ETP-ALL patients. Dasatinib, by contrast, is not recommended as a salvage strategy because ETP-ALL samples demonstrated resistance in ex vivo models, likely due to their stem cell-like epigenetic state and reduced LCK dependency. The overarching goal remains allogeneic transplantation in MRD-negative second complete remission, supported by evidence that transplantation in first remission yields significantly better post-transplant survival than later attempts.

Subject of Research: Genomic classification and clinical management of early T-cell precursor acute lymphoblastic leukemia

Article Title: Understanding and clinical approach to early T-cell precursor acute lymphoblastic leukemia

Article References: Ju, H. Y. (2026). Understanding and clinical approach to early T-cell precursor acute lymphoblastic leukemia. Blood Research, 61(1), Article 64. https://doi.org/10.1007/s44313-026-00181-y

Image Credits: AI Generated

DOI: 10.1007/s44313-026-00181-y

Keywords: ETP-ALL, T-cell acute lymphoblastic leukemia, genomic subtypes, minimal residual disease, allogeneic HSCT, nelarabine, daratumumab, venetoclax, bortezomib, immunophenotype, pediatric oncology, relapsed leukemia