CHRYSLER DIAGNOSTIC APPLICATION CDA.rar
Camila Fonua
This review will focus on the use of flow cytometry in the routine clinicopathologic approach to the diagnosis of leukemias and lymphomas, emphasizing the relevant literature of the past 10 years. Some of the recent advances in flow cytometric monitoring of disease and treatment are shown in the last section. We will review the use of flow cytometry in the diagnosis of major disorders highlighting the prognostically important subgroups defined either morphologically or genetically. The discussion will focus not only on the use of flow cytometry in the differential diagnosis of a particular disorder, but also correlate immunophenotypic, molecular, and cytogenetic data in the delineation of biologically important subgroups. It is our intent that this review support a combined modality approach to the daily practice of hematology-oncology and hematopathology. A working knowledge of the basics of flow cytometry is assumed; thus, technical aspects of instrumentation, normal distribution of surface antigens, and methodologies are not included, but have recently been reviewed.2-13
Flow cytometric analysis of acute leukemia is interpretive, combining the patterns and intensity of antigen expression to reach a definitive diagnosis.1-13 Gating is critical to isolate the abnormal cells because the leukemic phenotype should be determined on as pure a population as possible. Most leukemias involve the analysis of bone marrow. Standard forward and side scatter gating is not optimal for separating bone marrow cells because of the overlap between monocytes, blasts, myelocytes, promyelocytes, and metamyelocytes. As bone marrow cells mature, they express increasing CD45. Thus, when CD45 is combined with side scatter, which separates lineages based on cytoplasmic complexity, the bone marrow sample is readily separated into its cellular constituents.14 Infiltration of marrow by immature cells or blasts is more easily recognized on a CD45 versus side-scatter plot than on traditional forward side-scatter gating (Fig 1).
Analysis of normal and leukemic bone marrow by CD45-side scatter analysis. (A) Normal marrow illustrating several normal populations. (B) Lymphoblasts as seen in ALL. (C) Treated CML illustrating transition to acute phase with increased myeloblasts and a reactive increase in erythroid precursors. (D) Low-grade lymphoproliferative disorder illustrated by CLL. These patterns are representative and are not specifically diagnostic in the absence of other data.
Acute myeloid leukemia (AML). AML is traditionally subclassified by morphology and cytochemistry according to the French-American-British (FAB) criteria as modified by the National Cancer Institute-Sponsored Workshop that incorporates immunophenotypic data.15 Although the major role of flow cytometry is to provide immunophenotypic data, cellular morphology can be examined by both forward-side scatter16,17 and CD45-side scatter analysis.14 We will summarize each of the major subtypes of AML below incorporating the morphologic, immunologic, and cytogenetic (MIC) approach.18,19
The ability of flow cytometry to identify myeloid versus lymphoid differentiation approaches 98%.20,21 However, the prognostic value of immunophenotypic data is controversial.21-26 Studies that failed to find prognostic value for immunophenotyping generally looked at the correlation of outcome with individual antigens and did not find clinically useful associations, although the utility of flow cytometry in defining myeloid differentiation was confirmed.21-24 Studies that found correlation with specific phenotypes were generally single institution results. Three of the four studies showing no correlation were in children,21-23 in whom there is some evidence that the t(8; 21) may not carry the same good prognosis as in adults.27 Additionally, differences in reagents, gating and staining techniques, and thresholds for positivity may account for discrepancy.
It is our impression that genetic phenotypes carry the most important prognostic information. Unfortunately, there are few entirely consistent relationships between morphology, immunophenotype, and specific genetic alterations.30,31 There exist trends that are discussed below. However, in general, for any given genetic phenotype, there are patients with more than one possible FAB subtype or immunophenotype. We have chosen to place genotypic information in the FAB section in which the particular genotype is most commonly observed. Thus, we will discuss the immunophenotype of AML in the context of morphology and correlate with genetic phenotype where appropriate, similar to the MIC approach to diagnosis.21,28,32,33 AML is summarized in Table 1 based on the most common phenotypes with inclusion of possible genetic associations.
M0. M0 blasts have low forward and side scatter and typically merge with the lymphoblast region on CD45-side scatter plots. By definition, the blasts are cytochemically negative but express at least one myeloid specific marker such as CD13, CD33, or CD11b.34 Detection of the cytoplasmic enzyme myeloperoxidase (MPO) by monoclonal antibody appears more sensitive than CD13 and CD33 combined.35 Blasts are generally negative for lymphoid markers, but may express CD7 or CD4.21,35 M0 blasts are almost always positive for HLA-DR and CD34.35,36 Several investigators have shown an association between CD7 as well as CD34 expression in AML and a worse prognosis.26,37-45 These antigens may relate to expression of drug resistance phenotypes discussed in the last section. M0 is associated with a high incidence of cytogenetic abnormalities, most of which are complex but frequently involve chromosomes 5 and 7.35,36,46
M1. The flow appearance of M1 is similar to M0 and probably not separable. There may be slightly more side-scatter reflecting the cytochemically positive granules, but this is not definitive in a single case. M1 blasts are usually CD13+, CD33+, and HLA-DR+, but may not express as much CD34 as M0. There may be partial CD15 expression and less commonly CD4.35
M2. The major difference between M2 and M1 is the presence of maturation and a reduced percentage of blasts. Typically, there is a spectrum of cells with varying degrees of light scatter. CD45-side scatter can show a continuum of cells from the myeloblast region to the maturing myeloid cell regions. CD34 is less prominent and CD15 is more prominent than in M1. Most cases are HLA-DR+. CD13 is sometimes expressed more brightly than CD33. CD45-side scatter may be useful in determining the percentage of blasts.
ATRA therapy is not entirely benign. A complication is the sometimes fatal retinoic acid syndrome. This syndrome has been correlated with the expression of CD13 in the pretreatment leukemia population.62-64
M4 and M5. These two categories are similar phenotypically although M4 is more often CD34+ than M5.35 M4 and M5 cells have more forward and side scatter than M0 and M1. By CD45-side scatter, the maturing cells merge into the monocytic region. Maturation into the myeloid region as well should occur with M4, but this is not entirely reproducible. Important phenotypic features are the presence of CD13, CD33, HLA-DR, CD14, and CD15.35 CD33 may be brighter than CD13. The combination of CD33 positivity with negative CD13 and CD34 is highly correlated with an M5 phenotype, but occurs in only a minority of patients.29 CD56 may be seen in some cases of M5.29 Subtle clues to monocytic differentiation may be weak CD7 or CD4 expression35 and in our experience nonspecific binding of κ and λ light chains and IgG but not IgD and to a lesser extent IgM. Some cases of M5b may be entirely in the monocyte region on CD45-side scatter. The presence of CD2 is correlated with an important subtype M4Eo that is associated with abnormalities of chromosome 16 and a better prognosis.65-69
M6. M6 is rare and not well characterized. HLA-DR, CD34, and possibly CD13 or CD33 are usually present. CD45-side scatter may show a prominent erythroid component. Antibodies to glycophorin may demonstrate erythroid differentiation.70,71
M7. Acute megakaryoblastic leukemia (M7) accounts for less than 1% of AML4 and is diagnosed when greater than 30% of the nonerythroid cells are megakaryoblasts. The megakaryoblastic nature of the blasts must be confirmed by ultrastructural demonstration of platelet peroxidase or by immunophenotyping. Micromegakaryocytes are not counted as blasts but raise the possibility of M7.15 Immunophenotyping is important because neither morphologic nor routine cytochemical features are pathognomonic and ultrastructural techniques are difficult.72 Megakaryoblasts are typically identified by the expression of CD61 (GpIIIa) and/or CD41 (GpIIb-IIIa).72-74 However, caution must be exercised as false-positive reactions may occur due to platelet adherence to leukemia blasts.72 In one study of more than 1,000 cases of AML, 38% were positive for CD41. Comparison to cytospin immunoflourescence in 37 cases showed that 85% of the apparent expression of CD41 was due to adherent platelets. Therefore, confirmation of flow cytometric results by cytospin immunofluorescence is probably necessary in cases of M7.72 An interesting approach to reduce binding of activated platelets is two-color flow cytometry for GPIIb/IIIa and CD34 in the presence of EDTA.75
Extramedullary leukemia. The increasing use of immunophenotyping in fine needle aspiration of solitary tissue masses makes it likely that extramedullary leukemia will be encountered. In a recent review of extramedullary leukemia, 46% of cases were initially misdiagnosed.76 Two-thirds of patients receiving chemotherapy for a solitary primary site of extramedullary leukemia never developed AML, whereas 97% not treated initially with chemotherapy progressed to AML, emphasizing the importance of systemic therapy.76 A second review also emphasized the importance of prompt chemotherapy over radiation or surgery.77 An association between chloromas and t(8; 21) may exist.78 Accurate initial diagnosis of extramedullary leukemia is important and is another potential use of flow cytometric immunophenotyping.
b37509886e