Limits of phenotypic markers for the diagnosis of megakaryoblastic leukemia

Diagnosis of megakaryoblastic and early erythroid leukemia requires the use of differentiation markers that in most cases permit their precise diagnosis. In some cases, their use can be misleading. Here we report and discuss some examples. A platelet peroxidase (PPO) activity is detected in most cases of early erythroid leukemias as well as in the CFU-E-like cells of normal marrow, thus providing evidence that PPO activity must be studied along with other (immunologic or ultrastructural) markers to permit a reliable diagnosis of megakaryoblastic leukemia. Ferritin molecules an erythroid marker, could be detected as a cluster at ultrastructural level in leukemic platelets and in micromegakaryocytes of one patient. However, in blasts of the erythroid lineage, ferritin molecules are also either dispersed in the cytoplasm or localized in theta granules. Immunologic markers have also their own limit. Indeed, in one patient, GB IIb and IIIa were detected on erythroid blasts, resulting in a phenotype very similar to HEL cells. Carbonic anhydrase (CA) I, an early erythroid marker, was detected in the platelets of four leukemic patients and was present along with an increased expression of CA II. This study emphasizes the fact that precise diagnosis of leukemia cannot be performed with a single marker of differentiation, but requires the simultaneous use of several lineage restricted markers.