Skip Navigation

Complex Case Study: Using Genomic Testing to Tailor Treatment for Acute Myeloid Leukemia

  • Advances in Hematologic Malignancies Issue 4 Summer 2016
  • Advances in Hematologic Malignancies Issue 4, Summer 2016

    David Steensma, MD

    David Steensma, MDDavid Steensma, MD

    A 69-year-old man with a long-standing history of mild normocytic anemia and thrombocytopenia presented to his hematologist in Rhode Island for a routine follow-up visit and was noted to have new severe pancytopenia. Bone marrow aspiration and biopsy, which had previously been unrevealing on several occasions, was now consistent with acute leukemia, but the lineage of the leukemia was ambiguous.

    Although the patient's blasts expressed CD34 and TdT, subsets of blasts also expressed CD117, CD79b, CD11b, HLA-DR, CD7, CD22, CD25, CD38, CD13, and CD33 — a peculiar mixture of myeloid, lymphoid, monocytic and other lineage-associated markers. While the blasts did not express myeloperoxidase, the long prodrome of cytopenias and the karyotype of isolated trisomy 13 made a myeloid lineage leukemia more likely, so the patient was treated as if he had acute myeloid leukemia (AML).

    Although he tolerated induction chemotherapy with daunorubicin and cytarabine (the standard "3&7" regimen) well, a day 14 bone marrow exam showed more blasts than had been present prior to chemotherapy. Re-induction with a high-dose cytarabine-containing regimen also had no effect on the proportion of leukemic blasts, and the patient developed febrile neutropenia and radiographic evidence of pneumonia. Given the poor outlook, enrollment in hospice was discussed.

    CCGD BioInformatics and Profile diagram

    "I was really discouraged by this patient's initial course," said David Steensma, MD, the leukemia specialist at Dana-Farber/Brigham and Women's Cancer Center who cares primarily for this patient. "The chemotherapy didn't seem to touch his leukemia at all, which was tremendously disappointing to the patient and his family, as well as our medical team. Thankfully, by the time we were at the point where we were trying to decide whether to treat him with a hypomethylating agent or send him home for supportive care alone, molecular profiling results returned from a marrow aspirate that I had obtained from the patient when he first arrived. This report showed mutations in 6 genes, all associated with AML and related myeloid neoplasms: ASXL1, SETBP1, RUNX1, IDH2, NPM1, and SRSF2; several of these genes were identified in 2015 by R. Coleman Lindsley, MD, as highly specific for secondary AML, using samples from a trial of amonafide led by Richard Stone, MD. Not only did this genotyping result provide confirmation of the fact that it was appropriate to think of him as having AML and also confirm the conventional karyotype result, more importantly, the result suggested a therapeutic possibility: enrollment in a trial of a small molecule targeted at isocitrate dehydrogenase 2 (IDH2), which is a leukemia driver mutation detectable in about 10% of patients with AML."

    The copy number analysis showed gain of chromosome 13 (dark gray) and only one copy of the X chromosome since the patient was male (red). Seven different somatic mutations in 6 genes were also detected.

    The patient was found to be eligible for enrollment in Dana-Farber protocol 13-371, a Phase I study of the IDH2 inhibitor AG-221 — a targeted agent that was the subject of "The Transformation," a general-interest essay by Jerome Groopman, MD, in the September 15, 2014, issue of The New Yorker. During treatment with AG-221, the patient experienced prompt improvement in his blood counts, clearance of peripheral blood blasts, and (over the course of several months) a gradual reduction in marrow blasts from 70-90% to <10%, without any serious adverse effects. His performance status improved to the point where he was able to get back on his beloved John Deere tractor to mow his lawn and take a summer vacation with his family on Cape Cod. Eventually, though his blood counts never fully normalized and a subset of the other non-IDH2 mutations were still detectable in his marrow by the Rapid Heme Panel — a test developed at Dana-Farber and Brigham and Women's Hospital — he was able to be referred for reduced intensity conditioning allogeneic stem cell transplantation. Eighteen months later, he remains in complete remission and has an excellent performance status.

    "The clinical response of this very nice man to AG-221 is exciting on several levels. First, the response along with others supported the validity of the IDH2 drug target in AML and the activity of this newer agent. It is also a good example of how we don't always need to see a complete response from a drug in order for patients with leukemia to benefit," says Dr. Stone, leader of Dana-Farber's Adult Leukemia Program, who served as the Principal Investigator of the AG-221 trial at Dana-Farber, and has also been directly involved in the care of this patient. "This patient's marrow showed several leukemic clones and subclones at diagnosis — the sort of biological complexity we often see in older patients with AML. With treatment, he did not have normalization of his peripheral counts, probably since the IDH2 inhibitor only suppressed the IDH2-mutant clone and subclones so that the other pre-leukemic clonal cells were still present. Still, his blasts dramatically reduced, and his condition improved enough for him to get out of the hospital and do some of the things he enjoys, and even to move forward with transplant."

    Daniel DeAngelo, MD, PhD, director of clinical and translational research in the Adult Leukemia Program, adds, "Relatively few Phase I agents lead to the sort of striking response that we observed with this patient. But this is the direction that the hematological malignancies field, and medical oncology generally, is moving: towards use of narrowly targeted agents directed at driver mutations in molecularly well-defined subsets of disease. In our Leukemia Program, we try to have trials of new drugs or new combination approaches available for patients with all major classes of disease. In the case of AML, we have studies for people with newly diagnosed AML, relapsed or refractory AML, intensive regimens for younger patients and non-intensive regimens for older or frailer patients. Increasingly, our trials are also directed at specific genotypes of AML."