Skip to main content
Premium Trial:

Request an Annual Quote

Leukemia Trial to Assess DNA Methylation, Mutation Tests to Stratify Patients


NEW YORK – Researchers at the University of California, San Francisco will study the ability of tests for DNA methylation and certain gene variants to stratify patients with juvenile myelomonocytic leukemia (JMML) in a Phase I/II therapeutic interventional trial.

In the trial, which is sponsored by the Therapeutic Advances in Childhood Leukemia Consortium, researchers will evaluate the safety of Novartis' MEK inhibitor Mekinist (trametinib) with the chemotherapy azacitidine as a treatment for patients with newly diagnosed, lower-risk JMML, and Mekinist-azacitidine plus traditional chemotherapy in patients with newly diagnosed, high-risk JMML. Patients in the high-risk group will receive a stem cell transplant after two courses of therapy, while the lower-risk group will have the option to avoid a transplant if they respond after 12 months of treatment. They are aiming to enroll 64 patients over a period of five years.

Patients will be assigned to one cohort or the other based on results from blood tests for gene mutations and DNA methylation. Those with one clonal alteration and a low DNA methylation classification will be enrolled in the lower-risk group. Those with more than one mutation in certain genes or intermediate/high DNA methylation classification or both will be in the high-risk group.

JMML is the third most common type of leukemia in children, and it has a median onset age of 18 months. The prognosis for children diagnosed with JMML is poor, with around 50 percent of patients surviving two years after diagnosis.

Elliot Stieglitz, a UCSF pediatric oncologist and the lead investigator of the trial, said he became interested in studying JMML because the disease is characterized by mutations in the RAS pathway. According to Stieglitz, the trial his group is launching stands to be the first one in JMML that requires patients to test positive for a RAS pathway mutation to be eligible.

"It has taken years to get to the point where we can say every single patient needs to have a mutation, because we just didn't have the ability to test for it, and there were mutations we didn't know about," Stieglitz said. "But over time, our group as well as others have figured out every single patient has a mutation, and we can now test for it in a clinical format."

Stieglitz's group is using a next-generation sequencing panel at the UCSF Health Center for Clinical Genetics and Genomics to test for germline mutations in RAS pathway genes such as NRAS, KRAS, PTPN11, CBL, and NF1. The DNA methylation test for the trial is conducted by the National Cancer Institute Center for Cancer Research's Neuro-Oncology Branch.

In addition to testing for RAS pathway mutations, the trial will also be the first to guide patients' treatments based on results from a DNA methylation test or a test for gene mutations. Stieglitz noted that previous studies have shown that the number of mutations a patient has, regardless of the type, is an important predictor of outcomes. "We see very clear differences between patients who only have one mutation in their [cancer cells] versus patients that have more than one," Stieglitz said. "This is going to be the first trial that actually gives different amounts of therapy based on how many mutations a patient has."

Altered patterns of DNA methylation are also common in JMML patients with a more aggressive disease course. Stieglitz and collaborators a few years ago found that DNA methylation is frequently altered in the presence of secondary mutations in genes related to DNA methylation, RAS pathway genes, and others. Additionally, they reported that high methylation levels predicted worse outcomes in patients, while those with lower levels of DNA methylation fared better.

Subsequently, researchers from multiple groups who had separately published similar findings on DNA methylation in JMML came together, combined data on almost 300 patients, and established an international consensus criteria defining low, intermediate, and high methylation. "Now every trial that's run around the world [in JMML] can use the same exact criteria," Stieglitz said. "Even if we're not participating in the same clinical trial, we can compare apples to apples." 

The primary objective of the UCSF trial is to evaluate the safety of the respective therapeutic regimens in lower-risk and high-risk patients. The researchers will also track event-free survival in patients with lower-risk JMML who do not proceed to stem cell transplant, and they will measure the pre-transplant rate of molecular response to therapy in the high-risk patient group.

Stieglitz said his team is particularly interested in how many patients are in molecular remission after treatment with the Mekinist combination, "because we know that patients who enter transplant in molecular remission do better." Lastly, they will collect patients' samples longitudinally to evaluate genetic, biochemical, and functional responses to the treatments over time.

A stem cell transplant carries significant risks for the patient, including graft failure, organ damage, infections, new cancers, and graft-versus-host disease. Stieglitz and his colleagues hope to show that patients with low DNA methylation and only one mutation can safely avoid a stem cell transplant and survive long term.

The trial investigators will evaluate the lower-risk patients after 12 months of treatment. Patients who have progressed would proceed to a stem cell transplant, but those who have no evidence of disease after a year will stop treatment and undergo evaluation to gauge if they are cured of their leukemias. Stieglitz said his team is hoping that more lower-risk patients will be in the "cured" group. If these patients have progressed following therapy, they would then proceed to a stem cell transplant.

In contrast, patients in the high-risk group will receive a stem cell transplant after one or two cycles of Mekinist-azacitidine-chemotherapy, and researchers will evaluate whether that combination of therapies improves outcomes after transplant. Stieglitz specified that the type of transplant, whether it is autologous or allogeneic, and the source of cells, such as cord blood or bone marrow, are not mandated by the trial protocol.

Stieglitz believes an approach similar to the one taken in this JMML trial could be fruitful in other cancers. For example, he pointed out that breast cancer is actually many different diseases if you consider the driver mutations. If other cancers are distinguished according to genomic subtypes, he said that testing for DNA methylation and the number of gene mutations may also predict outcomes and play a larger role in clinical trials in the future.