Lineage ambiguity, infidelity, and promiscuity in immunophenotypically complex acute leukemias: genetic and morphologic correlates.

The laboratory characterization of acute leukemia has evolved as new technologies have emerged and been applied in the routine evaluation of these hematologic malignancies. With the use of morphology, cytochemistry, flow cytometry, cytogenetics, and molecular genetics, it is possible to characterize acute leukemias in an extremely sophisticated and (ideally) therapeutically rational manner. Such routine approaches have unraveled, in part, the extreme and sometimes bewildering heterogeneity of the acute leukemias, with the identification of increasingly numerous subtypes. Much of this complexity is defined by the underlying genetic lesions, the determination of which is assuming an even greater role in the definition of specific entities. For example, more than 200 different recurrent cytogenetic abnormalities have been described in one broad category of acute leukemia, the acute myeloid leukemias (AMLs).1 Of note, these karyotypically detectable abnormalities occur in just more than half of all AMLs, with the remaining approximately 45% of these cases having normal cytogenetics. Molecular genetic insights into the latter category have accelerated in the past 5 years, with up to 80% of cases with a normal karyotype already revealing fascinating submicroscopic molecular aberrations.2 This will yield additional tiers of complexity, accompanied, hopefully, by clarity of the underlying pathogenic mechanisms and, ultimately, improved therapy. Nevertheless, the major initial decision to be made when assessing an acute leukemia is whether it is lymphoid or myeloid because this usually straightforward assignment of lineage is central to guiding remission induction chemotherapy and subsequent therapy. Whereas morphologic studies and cytochemistry were originally the primary tools used in such decisions, they have been largely if not completely superceded by immunophenotypic analysis, most often performed by flow cytometry. This allows the vast majority (>95%) of acute leukemias to be definitively classified as AML or acute lymphoblastic leukemia (ALL), with the latter quite easily further assigned to Bor T-cell lineage.3 However, the immense diagnostic power of flow cytometry may not always be absolute, and the data it provides are not always, at first pass, absolutely clear. Thus, the following become evident: (1) morphology retains its value, not only in the initial current “gold-standard” detection and enumeration of blasts, but also with an astute microscopist being able to recognize morphologic features that herald an immunophenotype or genotype; (2) immunophenotypic analysis by multiparametric flow cytometry has sometimes revealed greater complexity than what may have been desired or required for initial diagnostic purposes. Indeed, at least 20% each of AMLs and ALLs coexpress lymphoid or myeloid antigens, respectively,3 in what may be termed aberrant or cross-lineage expression. Such scenarios should not be (but unfortunately sometimes are) confused with the fewer than 5% of acute leukemia cases that have more profoundly complex immunophenotypes and that have been referred to by a variety of not always synonymous terms, including acute leukemia of ambiguous lineage, acute mixed-lineage leukemia, hybrid acute leukemia, biphenotypic acute leukemia, and acute bilineal leukemia. Ultimately, of course, it is the genetic abnormality that largely defines the morphologic, immunophenotypic, clinical, and prognostic phenomena and, hence, the need to appreciate that acute leukemias, as with cancers in general, are genetic diseases. In fact, some morphologic findings and immunophenotypic profiles, with or without so-called cross-lineage antigen