Eosinophil Chimerism In The Differential Diagnosis Between DEK-CAN+ Acute Myeloid Leukemia Relapse and Chronic Graft Versus Host Disease

Eosinophilia is a rare recurrent finding in Acute Myeloid Leukemia (AML). The classical cases being: Core Binding Factor AMLs with chromosome 16 abnormalities, and AMLs with rearrangements of PDGFRA, PDGFRB and FGFR1. According to the 2008 WHO Classification, acute myeloid leukemia with t(6;9)(p23;q24) is a rare entity (0,7-1,8% of all AMLs) characterized by multilineage dysplasia, basophilia and unfavorable prognosis. The translocation leads to the development of the hybrid protein DEK-CAN. This cytogenetic alteration usually presents as a single abnormality, but can also be associated with complex karyotype. There are no descriptions of eosinophilia associated with DEK-CAN+ AML.

We report the case of a 16-year old male patient that presented with generalized cutaneous pruritus and fever associated with peripheral blood eosinophilia (absolute count: 7.830/mm3) on August of 2010. We ruled out the most common causes of secondary eosinophilia, and then submitted the patient to a bone marrow aspiration that showed erythroid hyperplasia (58,8%), with dyserythropoiesis, eosinophilia (16,4%), dysgranulopoiesis, and 9,6% of blast cells, compatible with the diagnosis of AML-M6 according to the FAB classification. Cytogenetic study showed the presence of t(6;9) (p23;q34), FISH was negative for inv (16) and molecular studies showed FLT3 and NPM1 to be wild type. PDGFRb rearrangement was negative.

The patient was treated with 2 cycles of cytarabin based chemotherapy, achieving hematological remission, but with sustained peripheral eosinophilia (absolute count: 920/mm3). Since the presence of t(6;9) is associated with dismal prognosis, the patient received a peripheral blood HLA matched related allogeneic stem cell transplantation (ASCT) in first remission.

Four months after the ASCT, he presented with symptoms of sinusopathy and worsening of the peripheral blood eosinophilia (absolute count: 10.655/mm3). Bone marrow aspirate revealed blast count < 5% with 50,8% of eosinophils, and decreasing donor chimerism. Cytogenetic study revealed persistence of t(6;9) (p23;q24) associated to other abnormalities (including monosomy 5), revealing a complex karyotype. He was then submitted to a therapeutic trial with imatinib without response. Afterwards he received one cycle of cytarabine plus azacytidine with donor lymphocyte infusion during post chemotherapy aplasia. The patient recovered counts with 100% donor chimerism and shortly developed signs of chronic graft versus host disease (cGVHD) affecting eyes, skin and liver. Bone marrow evaluation revealed complete morphological and cytogenetic remission with normal eosinophil counts and 100% donor chimerism. He was then treated with prednisone and tacrolimus with good control of GVHD.

Eleven months later, during tacolimus taper, the patients developed eosinophilia again. Bone marrow evaluation revealed morphologic and cytogenetic complete remission with 100% donor chimerism. Since eosinophilia had been a marker of AML activity in this patient, it became crucial to define whether this represented an impending relapse. We then performed eosinophil enrichment of peripheral blood eosinophils with Easysep (CD66b) beads, reaching a final sample with 80% of eosinophils (CD45high/CD16negative/CD13positive), and performed chimerism analysis in this sample. The results were compatible with 100% eosinophil donor chimerism, what ruled out the possibility of recipient neoplastic eosinophils. Tacrolimus tapering was stopped and two weeks later the patient developed worsening of cGVHD symptoms with sustained eosinophilia. Upon increasing the dose of tacrolimus, the eosinophilia resolved and we interpreted it to be secondary to cGVHD activity.

Afterwards we designed a PCR reaction specific for DEK-CAN rearrangement and tested retrospectively all bone marrow samples RNA we had stored. While the test was positive in all previous episodes of eosinophilia, the testing of the last sample was negative, in agreement with the eosinophil chimerism finding. The patient is now in complete molecular remission 24 months after DLI with good cGVHD control.

In conclusion, we have described a case of AML with DEK-CAN rearrangement secondary to t(6;9)(p23;q24) with eosinophilia. Additionally we have demonstrated how the study of eosinophil chimerism can be essential in the differential diagnosis between “neoplastic” and reactional eosinophilia.