← Cancer trial database
Trial result

PAX5-rearranged (PAX5r) B-ALL shows MRD-driven risk: 5-year EFS drops to 44% in high-risk kids

Cancer type
Acute Lymphoblastic Leukemia (ALL)
Published in
New in Oncology - April 30, 2026

Across consecutive AIEOP-BFM pediatric ALL trials/registries (ALL 2000/2009/2017; 2001-2024), this retrospective series captured 165 children (<18y) with PAX5r BCP-ALL spanning 65 fusion partners (most commonly PAX5::JAK2 n=22, PAX5::NOL4L n=18, PAX5::AUTS2 n=13, PAX5::ETV6 n=12). The phenotype skewed clinically high-risk (median age 3.4y; WBC>100×10^9/L in 22%; NCI high-risk in 53.3%; IKZF1plus in 15.1%) and biologically “slow to clear” disease—prednisone poor response 13%, PCR-MRD+ at end of induction 74.7% (28.6% with MRD≥5×10^-4) and still PCR-MRD+ at end of consolidation in 31.5%. In the AIEOP-BFM ALL 2009/2017 subset (n=116; median follow-up 3.6y), 5-year EFS/OS were 66.8±5.0%/95.3±2.1%; outcomes were sharply MRD-linked with 5-year EFS of 100% in protocol standard-risk vs 59.8±9.3% in medium-risk and 43.8±12.7% in high-risk—well below typical contemporary pediatric B-ALL benchmarks where OS is ~95% and EFS commonly ~85-90% in broadly treated cohorts (so the OS here looks “normal,” but the relapse/event signal concentrates in MRD+ non-standard-risk patients).

Take home

If PAX5r is validated prospectively as an “MRD-sensitive” high-risk genomics bucket, it argues less for blanket intensification and more for earlier MRD-guided escalation (and perhaps targeted add-ons in kinase-fusion subsets like PAX5::JAK2), because the standard-risk/MRD-negative group seems to do fine while the MR/HR groups clearly do not.

Source: Read the original publication

More in acute lymphoblastic leukemia

See all entries in acute lymphoblastic leukemia →