Acute myeloid leukemia (AML) is one of those cancers so genetically complex, with its many chromosomal abnormalities and mutations, that it has been hard for doctors to find clinically meaningful information to sort out this diversity. So when three gene sequencing studies from Penn Medicine show patterns that may help detect problems earlier and potentially better treat the blood cancer, it’s worth sharing.
The findings were presented this week at the annual meeting of the American Society of Hematology, known to most as simply ASH, where thousands of researchers and physicians gathered from around the world to hear the latest findings on blood disorders and cancers. In three separate studies, physicians from the Abramson Cancer Center and researchers from Penn’s Center for Personalized Diagnostics looked at three groups of patients: younger ones, those with secondary AML and those with a skin condition known as leukemia cutis, using next generation sequencing to find answers in their genetic makeup.
“This modern, molecular tool is helping us better understand the heterogeneity of AML and show us the different patterns of behavior in the disease,” said Martin P. Carroll, MD, a medical oncologist and leukemia biologist in the Abramson Cancer Center. “What can these mutations actually tell us about how certain patients fare is the big question we are after now.”
First is a study from Carroll and Caroline E. Sloan, a medical student from the Perelman School of Medicine, that looks at new ways to improve prognostication in patients with AML. In 2012, a study in the New England Journal of Medicine led by researchers at Memorial Sloan Kettering used data from patients enrolled in a randomized clinical trial to create a new genetic profile using chromosomal abnormalities and mutations in nine genes. That is, they found genetic patterns among patients they believed could be used to better characterize survival for patients. Carroll and Selina M. Luger, MD, professor of Medicine and Director of the Leukemia Program, were among co-authors on that NEJM study.
Carroll and his colleagues put this profile to the test for the first time in a clinical setting with newly-diagnosed patients seen at the Hospital of the University of Pennsylvania since 2002. In the retrospective study, they found that six out of the nine mutations (FLT3-ITD, DNMT3A, NPM1, CEBPA, IDH1 and IDH2) improved characterization of survival in younger patients (those younger than 60), but not older. To effectively predict outcome, they modified the use of the markers compared to the NEJM study. The various combination of mutations accurately predicted whether their outcomes were favorable, intermediate, or unfavorable, the authors reported.
Identifying such mutations on the front end could help better predict survival outcomes in younger patients, who typically have more aggressive cancer at diagnosis, and thus better inform treatment decisions.
The second study, led by Luger, investigated a mutation known as ASXL1 for secondary AML, which refers to the development of AML following the history of a previous disease, such as a myelodysplastic syndrome (MDS) or a chronic myeloproliferative disorder (MPN). To study the impact of the mutation, Jingmei Hsu, MD, an instructor in the Hematology/Oncology division, looked at 254 AML patients treated at the Hospital of the University of Pennsylvania to see who was carrying the mutation, and then analyzed the molecular and clinical characteristics of that subgroup.
The retrospective analysis revealed that two-thirds of patients who had the mutation had previously had MDS, MPN, and or prior radiation/chemotherapy, another cause of secondary AML. This strongly suggests ASXL1 may be a molecular biomarker for secondary AML.
“These were striking results that help us better understand the mutation and point to a possible, much-needed biomarker in these patients,” said Luger. “More studies are necessary to further investigate ASXL1 and its independent prognostic significance in AML.”
Last, researchers from the division of Hematology/Oncology and department of Dermatology, including Marlise R. Luskin, MD, Sasha Perl, MD, and Misha A. Rosenbach, MD, presented results that identified an association between the NPM1 mutation and leukemia cutis (LC), a skin condition that occurs in about 10 to 30 percent of AML patients in which leukemic cells invade the skin before they appear in blood or bone marrow. There are currently no genetic predictors of LC.
For the study, the researchers identified 279 adult patients, and found three common mutations: NPM1, DNMT3A and FLT3-ITD. Skin involvement was present in 26 patients, and 14 had the NPM1 mutation.
This is a new association between NPM1 and AML, the researchers reported, and could serve as a potential molecular marker in patients who may be at risk for LC, particularly those with monocytic AML. There was no association found among the group for the other two mutations.
“This story is constantly evolving as we continue to learn more about the molecular characterization of AML,” Carroll said. “But findings like these bring us one step closer to an even more individualized approach to treating AML patients and other cancers. Increasingly, we understand that AML is actually a group of diseases with different behaviors and response to therapy. The diseases look similar under a microscope, but molecular fingerprinting helps us to tell them apart, allowing us to better predict what will happen in individual patients.”