Clinical Challenges: Timely Testing for Genetic Mutations in AML

Acute myeloid leukemia (AML) is a genetically heterogeneous malignancy characterized by recurrent gene mutations. Genomic heterogeneity, patients’ individual variability, and recurrent mutations are major obstacles to optimum treatment.

“Genetic knowledge has been increasingly incorporated into the diagnosis, prognosis, and treatment of AML over the past decade. It’s still relatively new, although it’s become the standard of care at larger centers at diagnosis and at remission,” said John Mascarenhas, MD, of the Icahn School of Medicine at Mount Sinai in New York City. The presence of mutations such as IDH1 can help determine a patient’s risk of relapse and whether the risk is high enough to warrant bone marrow transplant in addition to chemotherapy, he told MedPage Today.

“But while mutations can assist in treatment decisions, the significance of a mutation is not black and white or cut and dry,” he added.

The varying genetic landscapes of patients with AML pose considerable challenges to clinicians even before treatment — not least the length of time needed to get the results of treatment-defining genetic tests before optimal therapy can begin.

“A sick patient walks in or is choppered in or brought into the clinic on a stretcher. They may need to start AML therapy urgently, but if they have one of the mutations that impact therapy you may not know that for 5 to 21 days,” said Mark Levis, MD, PhD, of Sidney Kimmel Cancer Center at Johns Hopkins Medicine in Baltimore. “The tests are time- and labor-intensive. “

And while a polymerase chain reaction (PCR) assay for one specific mutation can take just a few days, it takes 2 weeks to turn around a next-generation sequencing (NGS) panel on multiple variations, he told MedPage Today.

“The problem is,” Levis added, “standard intensive chemotherapy will almost certainly do harm if the patient has, say, a TP53 mutation.”

These mutations are associated with complex cytogenetic abnormalities, advanced age, chemoresistance, less response to bone marrow transplant, and poor outcomes, Levis explained. “So the dilemma is, do we wait to rule that out even if the patient is in desperate need of immediate treatment?”

One way around this is to test for complex karyotype, which most TP53 carriers have and which takes only 2 or 3 days, he said, noting that he is working “desperately” with the scientific labs, the health bureaucracy, and MBA consultants to speed things up and get these tests moved to the front of the line. “So if they do have a complex karyotype, they may well have a TP53 mutation, and I need to do everything I can to hold off treatment until that’s confirmed,” Levis said.

Then there’s the additional layer of getting timely authorization from insurers for testing, particularly for outpatients, who stay at home while awaiting test results. “We don’t want outpatients to get stuck with big bills for genetic tests the insurance company won’t authorize, so we’re working with the insurance bureaucracy on that,” he said.

Take the context of elderly AML patients who are not candidates for standard intensive chemotherapy, Levis continued. Knowing early on that they have mutated FLT3 can lead to the use of a FLT3 inhibitor instead of non-targeted treatment. “But you have to start the inhibitor by day 8 or you’ve missed your opportunity. You can get PCR results on one mutation back in a few days, but the panel takes 2 weeks, so we don’t order that,” he explained.

He said that in the absence of formal clinical trials, oncologists are experimenting “on the fly” with adding off-label FLT3 inhibition to treatment with azacitidine and venetoclax (Venclexta).

It’s an unorthodox approach, but doctors are trying it throughout the country. “I have 80-year-old AML patients in remission and out playing golf who would otherwise be dead. Regardless of how chaotic it seems, clearly we’re moving forward,” Levis said.

Aaron Gerds, MD, MS, of the Cleveland Clinic Taussig Cancer Institute, shares Levis’s frustration with the wrangling over insurance coverage, which can delay the critical diagnostic and prognostic tool of genomic testing.

“The challenge is getting universal coverage for genomic profiling. It’s something we still struggle with,” he told MedPage Today. “It’s critical for risk stratification. We need to understand the patient’s prognosis as soon as possible to know if transplantation will be needed or not.”

If new AML patients have mutated FLT3, it is crucial to know that in a timely fashion in order to get insurance authorization to complement chemotherapy with midostaurin (Rydapt) — a multitargeted oral kinase inhibitor — before day 8, according to recommendations from the RATIFY trial.

“But some won’t get authorization until well beyond day 8, day 10, or day 20 even, and the therapeutic window is lost,” Gerds said. “So in this era of targeted therapy we may have the treatment, but we’re not getting approval soon enough to follow the trial protocols.”

Timely tests are also important in the post-treatment phase, Mascarenhas said, noting that post-PCR, NGS is providing superior measurement of minimal residual disease and facilitating better treatment decisions.

Studies are underway to determine which mutational profiles will render leukemic cells more susceptible to targeted, gene-specific treatments, he added. “Mutations are heterogeneous and must be seen in the context of other mutations. We need to figure out how to better induce remission in those at genetic risk and achieve the longest complete remission and the best salvage therapy for relapsers.”

As trials move forward with gene-targeting done earlier in treatment and identification of new targeted treatments for different mutations and ways to better target existing agents, expediting genomic information at diagnosis will become even more important, Gerds said. “I can foresee a time not too far in the future when induction therapy for a patient with an IDH mutation may be an IDH inhibitor.”

The big jigsaw puzzle of AML genetic mutations may eventually be put together by artificial intelligence, Gerds continued. Machine learning may help clarify relationships between recurrent mutations in different populations, determine how to integrate them with patient variables, and help clinicians produce profile-based compilations of therapies that will work best for individuals.

“But that’s a few years off,” he said. “Machine learning is not there yet. Right now we need to build much better databases and better algorithms for better output. At the moment, the problem is ‘garbage in, garbage out.'”

Last Updated September 22, 2021