Immunoreactivity of MIC2 (CD99) and terminal deoxynucleotidyl transferase in bone marrow clot and core specimens of acute myeloid leukemias and myelodysplastic syndromes

Abstract

Context: MIC2 ("thymus leukemia") antigen has been shown to be expressed by T cells and monocytes, as well as B cells and granulocyte-lineage cells. It is most intensely expressed by the most immature thymus T-lineage cells and is more intensely expressed by CD34-positive/CD33-positive myeloid cells (compared to more mature myeloid cells) and the earliest CD34-positive/CD10-positive B-cell precursor cells (compared to cells of later B-cell precursor stages). CD99 (MIC2) is characteristically expressed in precursor B- and T-cell lymphoblastic lymphomas/leukemias, as well as in Ewing sarcoma/primitive neuroectodermal tumors (ES/PNET). It has also been shown to be expressed in a few terminal deoxynucleotidyl transferase (TdT)-positive myeloid processes, but has been uniformly negative in TdT-negative myeloid processes. A more recent study showed that 43% of acute myeloid leukemias (AMLs) and 55% of chloromas express CD99, concluding that CD99 is commonly expressed in AML and rarely seen in myeloproliferative disorders, myelodysplastic syndromes, or normal bone marrow. Although this study speculated that MIC2 expression was probably not limited to TdT-positive AML, there was no comparison with TdT reactivity in this study.

Objective: Since AML and high-grade myelodysplastic syndrome may occasionally be difficult to distinguish morphologically from acute lymphoblastic leukemia and ES/ PNET, we undertook a study to analyze MIC2 expression in conjunction with TdT reactivity in distinguishing AML or high-grade myelodysplastic syndrome from acute lymphoblastic leukemia and ES/PNET.

Design: We studied bone marrow core and clot paraffin specimens from AML (classified according to criteria of the World Health Organization; n = 49), myelodysplastic syndromes (n = 4), precursor B-cell acute lymphoblastic leukemia (n = 4), ES/PNET (n = 1), and normal bone marrow (n = 3) with MIC2 (CD99) and TdT immunohistochemistry.

Results: Overall, CD99 was expressed in 24 (49%) of 49 AML cases, including all (11/11) TdT-positive cases. CD99 was expressed in all subtypes of AML except M5. Myelodysplastic syndromes and normal bone marrow specimens were uniformly CD99 negative. Expression of TdT was limited to a subset of AML-M0, -M1, -M2, and -M4, and AML with multilineage dysplasia.

Conclusions: In contrast to a previous study, CD99 expression was not restricted to TdT-positive hematologic proliferations. In particular, the CD99-positive M3 and M7 AMLs were TdT negative. An M5 AML may likely be excluded based on a uniform TdT-negative/CD99-negative immunophenotype. In addition, in our experience, CD99 should be routinely evaluated on bone marrow clots, owing to decreased reactivity or loss of reactivity in rapid decalcifying (RDO) solution-decalcified specimens.