Report on the second Latin American consensus conference for flow cytometric immunophenotyping of hematological malignancies
On May 3 and 4, 2005, the Second Latin American Consensus Conference for the Immunophenotyping of Hematological Malignancies took place in Queretaro, México, with representatives from 10 countries of the region and two external consultants. This document summarizes the major conclusions for which consensus were achieved. Major differences regarding the recommendations from the first conference, which took place 9 years ago, concern the medical indications and the antibody panels for immunophenotyping. The aim of disseminating these guidelines to the international community is based on the potential interest for other countries with similar socioeconomic conditions. © 2005 Wiley-Liss, Inc.
Several protocols and guidelines provided by experts from Europe, the USA, and Canada in the field of immunophenotyping in hematological disorders are available; however, they might not prove to be suitable for Latin American countries in which economic conditions are precarious. Of the many factors having an impact on the final costs of a diagnostic test in the region, a very serious concern in health provision, its appropriate use in terms of medical indications and reagent panels are doubtlessly the major elements. The First Latin American Consensus Conference for Flow Cytometric Immunophenotyping of Leukemia was organized with the primary goal of defining the exact area where leukemia Immunophenotyping was undoubtfully useful clinically and the most appropriate methodology for obtaining such clinically relevant information, at the lowest possible cost, without having a negative impact on quality. The recommendations of such conference aimed to provide only the necessary but sufficient immunophenotyping information that could be used routinely in making clinical decisions, but cautiously enough to avoid decreasing the informative value of the assays (1).
The recommendations of the First Conference proved useful to ensure the quality of the tests' results, to allow their comparability in multicenter national or regional studies; however, it was unanimously agreed by professionals in the region that such recommendations needed to be updated. Therefore, the Second Latin American Consensus Conference for Immunophenotyping of Hematological Malignancies was organized to revise the already existing recommendations and adequate the new ones to current information and needs, but with exactly the same background philosophy and aims of the previous one. The major changes, concern medical indications, classification goals and panels of combinations of antibody reagents are reviewed later in this paper (1).
Seventy professionals, including both laboratorians and clinicians, from 10 Latin American countries participated in the conference. An open invitation was electronically mailed to most centers in the region known to be performing flow cytometric immunophenotyping of leukemia/lymphoma, and also displayed in the web pages of several such institutions. The represented countries are listed in alphabetical order: Argentina, Bolivia, Brazil, Colombia, Costa Rica, Ecuador, Mexico, Panama, Peru, and Uruguay. Again, to act as moderators of the discussions, Dr. Alberto Orfao from Spain, and Dr. Ricardo E. Duque, from the United States, were invited.
The items to be discussed were distributed to pre-registered attendants several weeks in advance, and to non pre-registered participants at the opening of the conference. Discussions comprised medical indications, samples, antibody panels, fluorochromes, staining procedures, data acquisition strategies, interpretation of results, and final report. Consensus on these items was defined as agreement by the majority of attendants. Items regarding biosafety, quality control procedures, sample suitability and rejection criteria, were not further discussed, since it was agreed that the recommendations from the First Conference on these matters are still valid being in line with widely acknowledged recommendations (2, 3).
Four different aspects of clinical utility were established for six different disease categories. Consensus guidelines are summarized in Table 1.
Table 1. Medical Indicactions of Flow Cytometric Immunophenotyping
|Prognosis and Extension||Yes||Yes||Yes||No||No||Yes|
The main difference between these recommendations and those proposed in the first conference is the now recognized clinical utility of immunophenotyping for the diagnosis, classification, and monitoring of patients with myelodysplastic syndromes (MDS). Monitoring comprises changes in the progression of the disease, evaluation of response to treatment, and detection of minimal residual disease.
It was agreed that erythrocyte lysis and fixation procedures of whole samples should be used in most instances, while density gradient separation might be limited only for cases where sample contamination with necrotic, stromal, or fat cells could interfere with the test interpretation. Consensus was reached that it is not necessary to stain excessive numbers of cells with each antibody admixture, hence the volumes of the sample, antibodies, and lysing and fixing solutions might be reduced as far as: (a) the relative proportion of sample and reagents is kept constant, and (b) the efficiency of such reduction is experimentally validated in each laboratory prior to use it for the routine workload.
Panels of Combinations of Monoclonal Antibodies
It was agreed that laboratories performing acute leukemia phenotyping should be capable to recognize: (a) T-lineage acute lymphoblastic leukemias, (b) B-cell precursor lineage acute lymphoblastic leukemias, (c) Acute myeloblastic leukemias, and (d) Biphenotypic acute leukemias.
In cases of T-cell acute lymphoblastic leukemia (T-ALL), consensus existed on that definition of T-cell lineage and cell immaturity were the necessary and sufficient goals; combined staining for CD2, CD7, and cytoplasmic CD3 was recommended as the most appropriate combination of markers to define T-cell lineage, which would require coexpression on the leukemic cells of cytoplasmic CD3 and at least one of the other two antigens. The maturity stage should be defined by investigating the presence of CD34, TdT, and the intensity of expression CD45, this later marker being also useful for gating purposes (4). Further sub classification of T-ALL was considered to be of no clinical relevance, at present.
Regarding B-cell precursor ALL, besides definition of the blast cell lineage and maturation stage, the relevance of the EGIL sub classification of B-cell precursor ALL into the Pro-B (B-I), Common (B-II), Pre-B (B-III), and mature B-cell (B-IV) ALL was recognized because of its prognostic and therapeutic implications (5); because of this the EGIL criteria were adopted. Accordingly, the use of four distinct categories of antibodies, three of them mandatory, was recommended, as summarized in Table 2.
Table 2. Categories of Antibodies for Immunophenotyping B-cell Precursor Acute Lymphoblastic Leukemia
|CD79α (cytoplasmic)||TdT|| ||CD38|
| ||CD34||Surface Ig|| |
| ||CD45||Cytoplasmic μ chains|| |
In acute myeloblastic leukemia (AML), definition of blast-cell lineage and maturation stage, together with the identification of cases with a strong likeliness of bearing the PML/RARα fusion (t15;17), were considered to be the goals of immunophenotyping at diagnosis because of their therapeutic implications (6, 7). Additional markers could be used for the definition of other myeloid cell lineages within AML, but these were considered optional. Accordingly, as for B-cell precursor ALL the use of three different categories of essential antibodies, together with a fourth optional group, was recommended, as described in Table 3.
Table 3. Categories of Antibodies for Immunophenotyping Acute Myeloblastic Leukemia
|CD33||CD45|| || |
|CD117a|| || || |
In those cases showing promyelocytic differentiation (CD34− or dim/HLA-DR− or dim/CD13+heterogenous/CD33++/ CD15− or dim) the presence of the PML/RARα fusion should be confirmed either by fluorescent in situ hybridization (FISH), by reverse transcription and polymerase chain reaction (RT-PCR), or conventional cytogenetics; this recommendation should be included in the report (vide infra).
Cases fulfilling the criteria of commitment into two or more of the above-mentioned cell lineages are to be defined as biphenotypic acute leukemia. Highly specific markers for each lineage are CD3 for T cells, CD19 and CD79α for B cells, and myeloperoxidase for myeloid cells (8).
Chronic lymphoproliferative disorders.
For immunophenotyping chronic lymphoproliferative disorders (CLPD) a decision tree was agreed upon. This algorithm might be operational, as described in a step-by-step procedure, or merely definer, according to each institution's environment and policies. For T cell CLPD, definitive demonstration of clonality might rely either upon the phenotypic analysis of the TCR repertoire, with a rather large collection of monoclonal antibodies in case of TCR αβ+ expanded/aberrant cells, or on the identification of a clonal rearrangement of the T-cell receptor variable genes through molecular biology techniques. Both approaches might be well beyond the scope of the routine clinical laboratory, but national or regional reference centers might perform these tests for smaller or less technically sophisticated sites. The feasibility to ship DNA samples to distant places, without significant deterioration of its analytical quality, makes molecular analysis of gene rearrangements the more suitable approach. The consensual decision algorithm is shown in Table 4 (9–14).
Table 4. Algorithm for Immunophenotyping Chronic Lymphoproliferative Disorders
|CD4||If a T cell lineage CLPD is suspected:||Confirmation of clonality:|
|CD8|| CD7|| Vβ repertoire|
|CD3a|| αβ TCR|| TCR rearrangements|
|CD19|| γδ TCR|
|CD56a||If a B-CLL/SLL is suspected:|| |
|Kappa light chains|| CD5|
|Lambda light chains|| CD22|
|If B cell FL/MCL/BL are suspected:|
|If hairy cell leukemia/marginal zone lymphoma are suspected:|
|If an NK cell lineage CLPD is suspected:|
Consensual recommendation was the addition of CD45, surface IgM as well as kappa and lambda light chains, to the previously recommended panel consisting of CD19, CD38, and CD56 (15). The main goal of these additions is the detection of clonally restricted IgM+ plasma cells in cases of Waldenström's macroglobulinemia within this group of patients; in addition, CD45 has been considered as useful in helping to distinguish normal and clonal plasma cells in multiple myeloma and monoclonal gammopathy of undetermined significance (MGUS).
General comments on the panels of monoclonal antibodies.
During the meeting it was emphasized that for the detection of phenotypic aberrations for appropriate follow-up of minimal residual disease, in most cases suffering from different hematopoietic malignancies, additional markers to those mentioned earlier, might be required. In addition, it was noted that each combination of monoclonal antibodies provides a higher amount of information that the sum of each antibody once stained separately and changes in the fluorochrome conjugated to a specific monoclonal antibody in a combination of reagents may vary the quality of the information obtained.
The aims to encourage immunophenotyping of MDS are: (a) to discriminate a normal bone marrow from a myelodysplastic syndrome; (b) to discriminate a myelodysplastic syndrome from an acute leukemia; (c) to enumerate the proportion of blast cells and their lineage, and (d) to classify the type of myelodysplastic syndrome. The first consensual agreement concerning this item was that at present only bone marrow aspirates are suitable samples for the immunophenotypic study of MDS. Consensus was reached on that the neutrophil and monocytic lineages are the two main lineages where immunophenotyping might provide robust information for the discrimination among normal hematopoiesis, myelodysplasia, and leukemia (16–18). Accordingly, an antibody panel for assessing maturation within these cell populations was agreed upon, as summarized in Table 5.
Table 5. Antibody Combinations for the Analysis of the Differentiation/Maturation of Neutrophil and Monocytic lineage cells
|°Tube 1||CD13, CD11b, CD45, CD34|
|°Tube 2||CD15, CD16, CD45, CD34|
|°Tube 3||HLA-DR, CD 117, CD45, CD34|
|°Tube 4||CD56, CD36, CD45, CD34|
|°Tube 5||CD64, CD14, CD45, CD34|
The aforementioned admixtures allow for a clear definition of normal maturation routes and hence, they facilitate the identification of both quantitative as well as qualitative abnormalities in bone marrow samples from MDS patients. In addition, abnormally granulated neutrophils, commonly observed in MDS, may demonstrate aberrant light scatter patterns on the flow cytometer (19).
It was agreed that for the analysis of red cell differentiation and maturation, the finding of a high proportion of CD45-negative nucleated cells in the S phase of the cell cycle, along with low or absent peripheral blood reticulocytes, are highly consonant with frustrated erythropoiesis.
Additionally, it was recommended that the analysis of MDS samples should be performed by specifically trained personnel because of its relatively high complexity, frequently requiring flow cytometry instruments capable of measuring 4 or more colors simultaneously.
Selection of Fluorochromes
Inasmuch as the instruments' color detection capabilities are very different among laboratories in the region, it was not considered practical to recommend specific fluorochrome-labeled antibodies to compose the different panels. As a general recommendation, however, it was consensual opinion that antibodies directed to antigens whose normal expression is rather dim (e.g. CD7, CD10, CD11b, CD13, CD23, CD34, CD64, CD117, or TdT) should be tagged with the most brilliant fluorochormes: phycoerythrin and, if instrumentation makes it possible, allophycocyanin.
Data Acquisition Strategies
It was agreed that gating immature cells in a CD45/side scatter histogram might prove useful in the vast majority of cases of acute leukemia and MDS. Therefore, whenever permitted by the instruments' capabilities, an antibody directed against CD45 should be included in each tube. Concerning the number of events to be collected, consensual opinion was that the previously recommended “rules” on the minimum number of interesting events should be followed. These rules indicate that at least 100 events of interest should be collected so as to rely on the reproducibility, coefficient of variation, and count of a given population of cells (20). This must be taken in consideration particularly when dealing with the quantification of minimal residual disease, where a minimum of 100 residual malignant cells should be collected. This goal, in many instances, requires acquisition of several hundred thousands to millions of total events.
Data Interpretation and Report of Results
In contrast to the brief report that was recommended in the previous consensus, a rather comprehensive and interpretative report, following the suggestions made by Del Vecchio et al. (21), was agreed upon. The report should include the following sections
Identifying code, gender, and age, as well as name of the referring laboratory and/or physician.
Suspected diagnosis as established by the referring clinician and potentially useful clinico-biological information associated with the patient.
Type and quality of the sample.
A brief description of the type and analytical quality of the sample and annotations regarding the presence of hemolysis, visual clots, or deterioration, should be made in this section.
Should include a list of the antibody combinations, including fluorochrome, used in the particular reported case.
A succinct, although comprehensive, description of the neoplastic cells should include light scatter properties as well as the presence or absence of each of the tested antigens. Information regarding the intensity, modality, and coexpression of several antigens should be provided as well. A description of the normal cell populations is to be included within this section. In cases of MDS, this section should contain a description of the qualitative and quantitative abnormalities observed in each of the different hematopoietic lineages; in cases of minimal residual disease, the estimated number of malignant cells should be stated. For these descriptions, terms of common medical usage should be preferred, while the use of flow cytometry jargon should be avoided (e.g. “cell size” instead of “forward scatter” or “FALS”; and “cellular complexity/granularity” instead of “side scatter”).
The concise immunophenotypic diagnosis is to be given, when applicable.
Exclusively indications regarding further studies will contribute to the final diagnosis or classification of the reported patient. This might concern other types of analysis (e.g. ploidy, chromosomal, genetic or serologic studies recommended), or new immunophenotypic analysis of additional samples.
The inclusion of one or more representative cytograms of the core findings was considered optional.
The original agenda was covered in this two day conference and consensus was reached on all proposed issues. Substantial changes were made to the recommendations arisen from the First Latin American Consensus Conference, mainly in the indications for immunophenotyping and the antibody panels. The main goal of the conference was to update the previous recommendations, after revision of the many publications concerning the informative value of immunological markers, as well as the growing expertise of scientists in the region. We feel comfortable that the current recommendations will prove cost-effective by increasing the clinical utility while keeping the cost of immunophenotyping as low as possible, but we are aware that the additions or amendments will become necessary in the future because, as has been nicely stated by Paietta, “although immunophenotyping by flow cytometry has become a routine approach to the diagnosis of hematological malignancies, it is a discipline that is still in development (22)”. Changes, perhaps some of them even dramatic, are yet to be expected in this field, and further efforts to keep providing updated useful information for patients and clinicians will remain the responsibility of laboratory professionals.
LIST OF PARTICIPANTS
Sylvia Maureen Abblitt Luengas, Ma. de Lourdes Baranda Cándido, Ana Maria Barquin Orozco, Carlos Bejar Lozano, Carlos Roberto Bueno Enriquez, Ofelia Camacho Del Monte, Yolanda Campos Garcia, Teresa de Jesús Cuenca Roldan, Judith De Bernal, José Luis Delgado Lamas, Richard Dyer Velarde-Alvarez, Lucia Esparza Silva, Moisés Espino Durán, Gabriela Fernandez Castillo, Nicté Guadalupe Figueroa Vega, Susana Fiorentino, Sonia Fuentes Ruiz, Silvia Galvan Meléndez, Hugo Giordano Cortazzo, Saúl F. Hernandez Dominguez, Sonia Hernandez Rodriguez, Gregorio Ignacio Ibarra, Arturo Jhann A., Esther Layseca Espinosa, Rubén D. Lobato Tolama, Mónica Patricia Londoño, Hilda Luna Zaízar, Luis Maldonado Noriega, Oscar Martinez Nieto, Nereida Méndez Ramírez, Enrique Cirio Meredijo Trujillo, Laura Arcelia Montiel Cervantes, Luis Mora Bermudez, Yolanda Muñoz De Saavedra, Virginia Nieto Pérez, Yaneth Rocío Orduz, Jessica Paola Ovalle Rodriguez, Yuritzi Pacheco Chavez, Laura Pardo, Rogelio A. Paredes Aguilera, Elías Pérez Becerra, María Perfecto Rios, Carmen Preciado Sanchez, Laura J. Rabelo Carrasco, Martín Gilberto Ramirez, Martha Ana Reyes Lopez, Ramón Rivas Llamas, Leticia Rivera Pérez, Ana María Rodriguez, Rosa Maria Rodriguez Delgado, Ana María Rodriguez Garcia, Silvia Margarita Rodriguez Mejorada, Alain R. Rodriguez Orozco, Aurora Elizabeth Romero Coronel, Lina Teresa Romero Guzman, Yadira Romero Lopez, Víctor Hugo Rosales Garcia, Carlos Saavedra Andrade, Rosario Salazar Riojas, Lucía Sanchez, Luis Alberto Santoscoy Tovar, Raúl Solis Martínez, Nydia Strachman Bacal, Héctor Fernando Tovar Tovar, Dunia Romy Valdivia Ferrufino, Bertha Valverde Rojas, Adriana villegas, Diana Marcela Zamora Rangel, Cecilia Zepeda Saavedra.