Marked Bone Marrow Eosinophilia Post‐Azacitidine + Ivosidenib Treatment for Acute Myeloid Leukemia

Various histomorphologic changes can arise in the bone marrow after induction therapy for acute myeloid leukemia (AML). Common changes include regenerative hematopoiesis, suppressed hematopoiesis, dysplasia, fibrosis, and necrosis [1]. Certain cell populations may also be increased as part of a regenerative and/or reactive response to therapy, such as macrophages and lymphoid cells [1], the latter of which tend to be less sensitive to chemotherapy. Eosinophilia is not a typical bone marrow finding associated with induction therapy. In general, bone marrow eosinophilia warrants a broad differential diagnosis, including drug reactions, allergic reactions, infections, and neoplastic processes. We describe an unusual case of marked bone marrow eosinophilia in a 70-year-old female after treatment for AML. At the time of diagnosis, the patient presented with spontaneous bruising and pancytopenia and had 38% circulating blasts. An initial bone marrow biopsy showed 47% myeloblasts, dyserythropoiesis, dysmegakaryopoiesis, and a normal XX karyotype. Next-generation sequencing identified pathogenic variants in IDH1, DNMT3A, and STAG2, consistent with AML, myelodysplasia-related (AML-MR). The patient underwent induction therapy with azacitidine and the IDH1 inhibitor ivosidenib. She tolerated the treatment well, reporting only mild headaches after the first cycle of therapy. A day +35 bone marrow exam showed neutrophil dysplasia, but no evidence of residual AML. Interestingly, the bone marrow was notably hypercellular (Figure 1A). The aspirate predominantly contained eosinophils and eosinophil precursors with cytoplasmic vacuolization, large specific granules, and occasional “Harlequin” eosinophils that also contained basophilic primary granules (Figure 1B–D). There was no peripheral blood eosinophilia. A follow-up bone marrow examination performed after the third cycle of therapy (day +105) showed ~7% eosinophils, decreased granulopoiesis, dysplastic megakaryocytes, and 1% blasts, consistent with morphologic remission (Figure 2). This case is unique in the presentation of marked eosinophilia after a single cycle of azacitidine + ivosidenib. In this context, eosinophilia was considered indicative of a reactive or regenerative process in response to therapy. This interpretation was further supported by the transient nature of the eosinophilia. The patient had no recent travel history and there were no skin changes, respiratory distress, gastrointestinal symptoms, or constitutional symptoms that can be seen with various infections, neoplasms, allergic reactions, drug hypersensitivities and other conditions that may produce eosinophilia. The absence of peripheral blood eosinophilia would also make a chronic eosinophilic leukemia unlikely. Although some myeloid neoplasms may present with increased marrow eosinophils without peripheral blood eosinophilia, including AML with CBFB::MYH11 fusion and various myeloid/lymphoid neoplasms with eosinophilia and tyrosine kinase gene fusions, the marrow morphology and initial molecular investigations were not compatible with such entities. The patient's only other active medications were related to control of hypertension and gastroesophageal reflux disease, which had been prescribed several years prior to initiating treatment for AML. IDH1 inhibition–associated differentiation effects may plausibly contribute to the observed phenomenon, although a direct mechanistic link remains speculative. Documented side-effects of combined therapy with azacitidine and ivosidenib for IDH1-mutated AML include nausea, vomiting, arthralgia, prolonged QT, dyspnea, insomnia, differentiation syndrome, hematoma, neutropenia, thrombocytopenia, leukocytosis, hypertension and headache [2, 3]. However, significant changes in bone marrow histomorphology associated with these medications have not been well described. To our knowledge, marked bone marrow eosinophilia related to azacitidine + ivosidenib in the treatment of AML has not been reported. Our case, therefore, adds to the literature of therapy-related bone marrow changes, which are important to document and recognize to avoid over investigation or unnecessary alarm for what is otherwise an interesting but benign finding. All authors contributed to the paper's conception and design. Christopher Liwski and David Conrad collected clinical and histological data and drafted the manuscript. All authors read and approved the final manuscript. All authors acknowledge the work of the clinical team members caring for the patient and the laboratory technologists who prepared the histology and morphology slides used in this case. The authors have nothing to report. Informed consent has been obtained from the patient. The authors declare no conflicts of interest. The data that support the findings of this study are available from the corresponding author upon reasonable request.