Researchers from Singapore General Hospital (SGH), A*STAR’s Genome Institute of Singapore (GIS), and Duke-NUS Medical School have used artificial intelligence (AI) to speed up the identification of vital biomarkers that can identify patients with chronic myeloid leukaemia (CML) at diagnosis who will not respond to standard therapy.
These patients may be eligible for a life-saving bone marrow transplant in the early stages of the illness with this favourable prognosis.
A genetic mutation that causes a tyrosine kinase enzyme to turn on permanently causes CML, a specific type of blood cancer. In the bone marrow, a blood stem cell experiences a mutation that transforms it into an aggressive leukaemic cell that eventually takes over the creation of healthy blood.
Tyrosine kinase inhibitors (TKI), which turn off the tyrosine kinase that the genetic mutation switched on as a result, are the standard treatment for CML. But not everyone reacts the same way to these medications. Some individuals respond very well to the point that their life expectancy would be regarded as typical, at the other end of the range.
Besides, some individuals do not respond at all, and their sickness develops into a severe condition known as a blast crisis that is resistant to all sorts of conventional therapy.
Finding out if a patient is resistant to TKI therapy earlier could make the difference between survival or early death because the only cure for blast crisis is a bone marrow transplant, which would be most successful when carried out during the early stages of the disease.
“Our work indicates that it will be possible to detect patients destined to undergo blast crisis when they first see their haematologist,” said the study’s senior author and associate professor, Ong Sin Tiong of Duke-NUS’ Cancer & Stem Cell Biology (CSCB) Programme.
He added this may save lives since bone marrow transplants for these patients are most effective during the early stages of CML.
Researchers made an “atlas” of cells by taking samples of bone marrow from six healthy people and 23 people with CML before they were treated. The map let them see the different types of cells in each sample and how many of each type there were. Researchers did RNA sequencing on a single cell and used machine-learning methods to figure out which genes and molecular processes were on and off in each cell.
The work found eight statistically important things about the bone marrow cells before treatment. These things were linked to either sensitivity to treatment with a tyrosine kinase inhibitor or strong resistance to it.
Patients were more likely to react well to treatment if their bone marrow samples showed a stronger tendency toward premature red blood cells and a certain type of “natural killer cell” that kills tumours. As the number of these cells in the bone marrow changed, so did the way the patient responded to treatment.
The study could lead to drug targets that could help people with chronic myeloid leukaemia avoid or delay treatment resistance and blast crisis.
Associate Professor Charles Chuah from Duke-NUS’s CSCB Programme, who is also a Senior Consultant at the Department of Haematology at SGH and National Cancer Centre Singapore (NCCS), cited that the results of treating chronic myeloid leukaemia have gotten much better over the years and that patients now have many options. Knowing which treatment works best for each patient will improve these results even more, and they are excited about the chance of doing so.
The team hopes to use the results to make a test that can be used regularly in hospitals to predict how well a treatment will work.
