Predicting Relapse in Leukemia with Digital Droplet PCR

Is there a better way to identify measurable residual disease?

Despite aggressive and highly toxic chemotherapy, measurable (or minimal) residual disease, or MRD, is seen in a variety of cancers and has become a powerful tool for predicting leukemia relapse and survival in patients. These leukemia cells that stick around after treatment are the likely cause of most relapses within five years of diagnosis.

Current approach for predicting leukemia relapse

Pathologists in the U.S. currently identify a majority of MRD with flow cytometry — a method that allows for detection of 1 in 10,000 leukemia cells by identifying two or more proteins that are never expressed together on normal cells. Pathologists in Europe use other molecular methods, but the goal of these techniques is the same: to more accurately measure the disease that predicts relapse.

Flow cytometry helps risk stratify patients in a far more accurate way than is possible using a microscope, the historical way to detect disease. In fact, hundreds of papers describe different groups of patients for whom flow cytometry has consistently proven to be a predictor of relapse-free survival. But there's something it doesn't take into account.

"Cell surface markers can change over time, particularly after leukemia cells have seen chemotherapy or immunotherapy," says Children's Colorado pediatric oncologist and hematologist Amanda Winters, MD, PhD. "This makes it difficult to distinguish a malignant white blood cell population from a variety of normal ones."

New approach for predicting leukemia relapse

She says a more reliable way to identify MRD is to measure the abundance of leukemia-associated mutations, which are not present in normal cells.

To accomplish this, she's working in collaboration with Dan Pollyea, MD, MS, a leukemia physician at UCHealth, and in the lab of Craig Jordan, PhD, a leukemia researcher at University of Colorado, to create a special type of polymerase chain reaction (PCR) called digital droplet PCR, or ddPCR. It's highly sensitive and can detect one mutant gene copy in 10,000 to 100,000 copies. That's at least as sensitive as flow cytometry, says Dr. Winters, and it's often much easier to interpret.

The effect the ddPCR creates is similar to a typical PCR, except that, as the name implies, it emulsifies into individual droplets when it combines with an oil, capturing individual copies of DNA in each droplet. Different color fluorescent tags help distinguish the mutant versions of the gene from the normal, or wild type, allowing Dr. Winters to calculate abundance of mutant DNA.

Expanding research on prediction of leukemia relapse to kids

So far, she's been able to validate more than 60 mutation-specific ddPCR assays to identify MRD in a cohort of adult patients with acute myeloid leukemia. She says these assays can differentiate patients who will ultimately show signs of leukemia relapse symptoms from those who have lower relapse risk based on presence or absence of acute myeloid leukemia mutations. She also sees the same correlation between MRD status and outcomes in these adult patients after a bone marrow transplant.

On the pediatric side, although some patients with acute myeloid leukemia have overlapping mutations with adults, most have chromosomal translocations that adults don't have. An important next step in the research is for Dr. Winters to design ddPCR assays for these translocations so that she can more reliably monitor MRD in pediatric patients.