Childhood Leukemia 2026: 90% Survival Isn't Good Enough

Acute lymphoblastic leukemia (ALL) is the most common childhood cancer — and one of medicine's great success stories. In the 1960s, it was nearly universally fatal. Today, 5-year survival rates exceed 90% in high-income countries with access to modern chemotherapy protocols. But "90% survival" still means thousands of children die each year — and in low-income countries, where treatment is expensive and often unavailable, survival rates fall to 40–60%.

The two remaining challenges: drug resistance in relapsed cases (which accounts for most pediatric leukemia deaths in wealthy countries) and treatment accessibility in low-income settings. Distributed computing is attacking both through gene expression analysis.

What Gene Expression Analysis Tells Us About Resistance

When childhood leukemia relapses after treatment, it's because a sub-clone of leukemia cells survived the chemotherapy — carrying genetic or epigenetic changes that made them resistant. Identifying the gene expression signatures of resistant clones before treatment — so doctors can choose different protocols upfront — is the goal of the Solvexoria childhood leukemia problem.

The network is building a gene expression atlas: a map of which genes are active or inactive in leukemia cells from 400,000 computed genomic alignment tasks. By comparing expression profiles of sensitive vs. resistant cases, the model identifies gene expression patterns that predict resistance with high accuracy.