NPM1, FLT3, PML::RARA, the core-binding-factor fusions: they decide the treatment, the risk group, and whether a patient is eligible for a targeted drug. In most hospitals that panel takes three to ten days. In most of the world it never arrives.
While it runs, the patient waits, or treatment starts blind. And to get the sample that feeds it, a needle goes into the hip bone. For a child, under sedation. Then again at day fifteen to check response. Then again at the end of induction. Then again for residual disease.
Axiisium starts from a different fact. These mutations are not invisible. NPM1 mislocalises a nuclear protein and the nucleus folds into a cup. PML::RARA fills the cytoplasm with granules and Auer rods. The core-binding-factor fusions leave their own marks. Hematopathologists have suspected genotype from morphology for forty years. Axiisium is being built to do it systematically, on every cell of every case, and to state its own uncertainty gene by gene.
Predict molecular abnormalities from a routine blood smear, abstain when the visual evidence is insufficient, and progressively replace sequencing wherever prospective clinical trials prove it safe.
Three words, in order.
The model. Read the lesion off the cell, on Day 1, from the sample a blood draw already produces.
The discipline that makes the model safe to use, because a system that guesses on the cases it cannot see is worse than no system.
The last word for a reason: it is earned in a prospective trial, one gene at a time, or it is not earned at all.
As a child, our founder sat in a hospital ward beside a boy his own age with acute myeloid leukemia. It was Myanmar; the diagnosis was slow, and care was scarce. The boy never made it home.
That memory is the reason Axiisium exists. AML moves fast, and too often the answer cannot keep up. We are building the tool we wish that boy had had: one that reads what the cell already shows, in time to matter, and that can one day reach the kinds of places where the genetics never arrive at all.
Axiisium is early and in active development, a research-use tool today. We hold every claim to the same standard we hold the disease to, and we publish our results on real data as they stand, not before. Every number, with its conditions attached, is on the Evidence page.
Orange is the leukemia awareness color.