Flow Cytometric Analysis of CD5+ B Cells A Frame of Reference for Minimal Residual Disease Analysis in Chronic Lymphocytic Leukemia

Transcription

1 Hematopathology / CD5+ -CELL NLYSIS Y FLOW CYTOMETRY Flow Cytometric nalysis of CD5+ Cells Frame of Reference for Minimal Residual Disease nalysis in Chronic Lymphocytic Leukemia Ritu Gupta, MD, 1 Paresh Jain, MD, 1 S.V.S. Deo, MS, 2 and tul Sharma, MD, DM 3 Key Words: CD5+ cells; Flow cytometry; Chronic lymphocytic leukemia; Minimal residual disease DOI: /T5EM9QU9CM8F57 bstract Recent reports suggest that CD5+ cells constitute up to 47% of the total cells in normal peripheral blood (P), a finding that would restrict the sensitivity of the CD5/CD19 flow cytometric assay for minimal residual disease (MRD) analysis in chronic lymphocytic leukemia (CLL). We studied 40 normal samples (P, 20; bone marrow [M], 20) using CD5 fluorescein isothiocyanate (FITC)/CD19- phycoerythrin (PE) immunostaining to evaluate the reference range of CD5+ cells. The mean percentage of CD5+ cells per total number of cells was 12.2% (range, 3.6%-23.9%) in P and 11.7% (range, 4.4%- 19.5%) in M. On serial dilution, this assay could detect 1 CLL cell in 1,000 leukocytes (sensitivity, 0.1%). distinct bright CD5+ -cell subpopulation, consistent with a CLL-like-phenotype, was observed in 3 samples. Our results suggest that the CD5- FITC/CD19-PE assay has a clinically useful sensitivity for MRD analysis in CLL. The usefulness of this assay as a screening tool to identify the earliest stage of indolent CLL needs further study. Chronic lymphocytic leukemia (CLL) is an indolent lymphoproliferative disorder characterized by clonal expansion of CD5+ cells expressing weak surface immunoglobulins. The conventional therapy protocols in CLL are not aimed at cure, and most patients in complete remission have substantial levels of disease detectable by modern techniques. 1-3 With the advent of better therapeutic modalities such as purine analogues, monoclonal antibodies, and hematopoietic stem cell transplantation, much more profound responses are achieved in patients with CLL. The depth of remission in these patients correlates with overall and progression-free survival. 4 Thus, response to these novel therapeutic approaches needs to be defined by more sensitive techniques such as flow cytometry and polymerase chain reaction. lthough highly sensitive, polymerase chain reaction techniques are labor intensive, expensive, and not ideal for routine assessment of response to therapy. 4 Flow cytometric analysis, on the other hand, is a relatively rapid, easy, and more readily available technique for minimal residual disease (MRD) analysis in leukemia. Detection of MRD by flow cytometric analysis is based mainly on the presence of an aberrant phenotype of the neoplastic cells. Overexpression of CD5 in CLL cells, therefore, often has been exploited to detect residual disease by flow cytometric analysis. 2,5 CD5+ cells constitute a distinct -cell subset. 6 t birth, most lymphocytes in cord blood coexpress CD5, and, by young adulthood, CD5+ cells constitute only about one fifth of normal peripheral blood (P) cells. 7,8 Thus, most studies have used a cutoff of 25% of CD5+ cells per total number of cells in P to define positivity for MRD in CLL. 2,5 However, in 2 recent studies, CD5+ cells constituted up to 47% of total cells in normal P, a finding that would tend to restrict the sensitivity of the CD5/CD19 flow cytometric assay 368 m J Clin Pathol 2004;121: DOI: /T5EM9QU9CM8F57

2 Hematopathology / ORIGINL RTICLE for MRD analysis in CLL. 9,10 The present study was conducted with the aim of setting up a reference range for CD5+ cells in our laboratory and to evaluate the sensitivity of this technique for detecting leukemic cells. the results by using the Wilcoxon rank sum test: (1) 40 years or older and (2) younger than 40 years. Materials and Methods P samples from 20 healthy adults (age, years) were obtained in EDT. one marrow (M) aspirate samples were obtained in EDT from 20 patients (age, years) in whom staging M aspiration was done for solid tumors. Inclusion criteria were as follows: (1) normal total and differential leukocyte counts in the samples, (2) normal hemoglobin level, (3) normal platelet count, and (4) bone marrow biopsy free of malignancy. Samples were prepared by using a standard whole blood lysis technique. 11 We immunostained 1 million cells with 10 µl each of pretitrated fluorescein isothiocyanate (FITC)- labeled CD5 (L1a; Immunotech, Marseilles, France) and phycoerythrin (PE)-labeled CD19 (J4.119; Immunotech) monoclonal antibodies. Monoclonal antibodies of the same IgG subclass were used as negative controls in all experiments. Samples were acquired on a flow cytometer (Epics Elite, Coulter, Hialeah, FL) and analyzed with Expo 32 software (pplied Cytometry Systems, Sheffield, England). Compensation was set using CD4/CD8 double-labeled lymphocytes. cquisition was performed in 2 stages. cquisition first was done on an ungated population, and 10,000 cells were acquired. second acquisition was performed with a live gate on a lymphoid population identified by low forward and low side scatter characteristics and was continued until all cells in the tube were acquired. nalysis was performed using sequential gating strategy Image 1. n initial region (R1) was set around cells with low forward and side scatter characteristics; a second region (R2) was set on cells with high CD19 expression and low side scatter; and a third region (R3) was set to exclude cells that were binding equivalent amounts of CD19 and CD5; such cells fall in a diagonal line that indicates nonspecific binding. Further analysis was performed on events that fell within all 3 of these regions and comprised total CD19+ cells. To determine the level of sensitivity of the CD5- FITC/CD19-PE assay for MRD analysis in CLL, a serial dilution experiment was performed by adding increasingly low numbers of clonal cells obtained from a patient with CLL to a normal P sample. The dilution factor ranged from 1: to 1:10 5. cquisition was continued until 10,000 CD19+ events were acquired. To study whether age could influence the reference range of CD5+ cells per total number of cells, we also analyzed the data separately for 2 age groups and compared Results CD5+ cells were identified in all P and M samples. CD5 expression in cells was dim compared with that seen in T cells. In 3 cases, an additional bright CD5+ subpopulation of cells with CD5 intensity comparable to that of T cells was noted. These 3 cases were excluded from analysis while calculating the reference range for CD5+ cells and were analyzed separately. The CD5+ cells represented a mean of 12.2% (range, 3.6%-23.9%) and 11.7% (range, 4.4%- 19.5%) of the total cells in P and M samples, respectively Table 1. The cutoff value for defining MRD positivity, calculated as the mean percentage of CD5+ cells per total number of cells plus 2 SD, was 25.0% in P and 21.1% in M. Of the total cells, CD5+ cells constituted a mean of 1,023 FS C 0 R1 0 1,023 SS R3 1,023 SS 0 R2 Image 1 The gating strategy used for identification of lymphocytes., n acquisition region was drawn on cells with low forward scatter (FS) and low side scatter (SS) characteristics (R1), and all such cells were acquired. -D, t analysis, cells with low SS and bright CD19 expression were selected (R2), and then cells binding equivalent amounts of both antibodies were excluded and remaining cells (R3) analyzed. FITC, fluorescein isothiocyanate; PE, phycoerythrin. D m J Clin Pathol 2004;121: DOI: /T5EM9QU9CM8F57 369

3 Gupta et al / CD5+ -CELL NLYSIS Y FLOW CYTOMETRY Table 1 Flow Cytometric nalysis of Cells in Normal Peripheral lood and one Marrow Samples * CD19+ Cells/Total CD5+CD19+ Cells/Total CD5+CD19+ Cells/ Sample Lymphocytes (%) Lymphocytes (%) CD19+ Cells (%) Peripheral blood (n = 19) 7.3 ± 3.3 ( ) 1.0 ± 0.9 ( ) 12.2 ± 6.4 ( ) one marrow (n = 18) 14.9 ± 7.2 (4.8-32) 1.8 ± 1.3 ( ) 11.7 ± 4.7 ( ) * Data are given as mean ± SD (range). One peripheral blood sample and 2 bone marrow samples exhibited bright CD5+ cells and were excluded from analysis. 10.3% (range, 3.6%-23.7%) in P and 12.5% (range, 4.4%- 19.5%) in M in the group younger than 40 years and 14.4% (range, 6.3%-23.9%) in P and 11.2% (range, 5.7%-18.2%) in M in the group 40 years or older Table 2. When the age groups were compared, the CD5+ cell percentage did not differ significantly in the P (P =.21) or M (P =.44) samples. CD5+ -cell subpopulation with bright CD5 was identified in 3 (8%) of 40 samples (P, 1; M, 2) Image 2. ll 3 samples were from the group aged 40 years or older. The percentage of CD5+ cells per the total number of cells in these samples ranged from 13.3% to 32.4% Table 3. The bright CD5+ -cell subpopulation constituted 1.0% to 2.0% of all cells. This subpopulation was not characterized further for clonality or other CLL-associated markers. The results of the serial dilution experiment are shown in Image 3. y using 25.0% CD5+ cells per total cells (mean + 2 SD) as the cutoff value in P, the CD5/CD19 assay could identify 1 neoplastic CLL cell among 1,000 normal leukocytes (sensitivity, 10 3 ; 0.1%). Discussion In our study, CD5+ cells constituted 3.6% to 23.9% and 4.4% to 19.5% of the total cells in normal P and M samples, respectively. These results compare favorably with those reported by Cabezudo et al 2 Table 4. Vuillier et al 12 analyzed 30 normal P samples and reported CD5 coexpression in up to 20% of the total number of cells. However, in 2 recent studies, CD5+ cells constituted a considerably higher proportion of total cells. 9,10 Sanchez et al 10 found CD5+ cells to constitute up to 37.1% (in P) and 49.2% (in M) of total cells. Similarly, Maloum et al 9 noted CD5 coexpression in up to 47% of total cells in normal P samples. Table 2 Comparison of CD5+ Cells in Two ge Groups * CD5+CD19+ Cells/CD19+ Cells (%) ge Group Peripheral lood one Marrow <40 y 10.3 ± 5.7 (n = 10) 12.5 ± 6 (n = 7) 40 y 14.4 ± 6.8 (n = 9) 11.2 ± 3.8 (n = 11) * Data are given as mean ± SD. Image 2 Double labeling with CD5-FITC/CD19-PE in lymphocytes., typical CD5-FITC/CD19-PE dot plot from a normal bone marrow sample showing a distinct dim CD5+ -cell population (arrow)., Dot plot from another bone marrow sample showing a distinct clone of lymphocytes (arrowhead) expressing bright CD5, which can be distinguished easily from normal CD5+ cells (arrow). FITC, fluorescein isothiocyanate; PE, phycoerythrin. review of these studies, as summarized in Table 4, suggests fluorochrome selection as an important variable that might influence CD5+ -cell estimation and the cutoff above which the MRD might be considered present. The Table 3 Flow Cytometric nalysis in Normal Samples With right CD5+ Cells Lymphocytes/ CD5+ Cells/ll ll CD5+ Cells/ right CD5+ Cells/ Sample ge (y) ll Lymphocytes (%) Lymphocytes (%) Lymphocytes (%) Lymphocytes (%) Peripheral blood one marrow one marrow m J Clin Pathol 2004;121: DOI: /T5EM9QU9CM8F57

4 Hematopathology / ORIGINL RTICLE sensitivity of a particular marker (or marker combination) for MRD analysis by flow cytometric analysis depends on its ability to discriminate normal cells from neoplastic cells and is inversely proportional to the percentage of normal cells expressing it. ecause a CD5-FITC/CD19-PE combination consistently identifies CD5 coexpression in about 20% to 25% of normal cells compared with CD5-PE or CD5- CyChrome (Pharmingen, San Diego, C) monoclonal antibodies that identify it in up to half of normal cells, the former is more sensitive for MRD analysis in CLL. Whether CD5-FITC underdetects CD5 expression or CD5-PE and CD5-CyChrome overdetect is unknown. correlation of flow cytometric data with CD5 messenger RN assay in cells might be helpful in this regard because a high correlation between CD5 surface and messenger RN expression has been demonstrated in murine lymphocytes. 13 Previous studies have demonstrated the usefulness of a CD5/CD19 assay for MRD analysis in CLL. 2,5,14 In the present study, we determined the in vitro sensitivity of this assay. On serial dilution of CLL cells in normal cells, the dual-color assay had a sensitivity of This is about 2 logs more sensitive than morphologic examination because the distinction of CLL cells from normal lymphocytes is difficult when neoplastic cells are present in low numbers. CD5/CD19 dual staining is more sensitive than clonal excess for MRD analysis in CLL because CLL lymphocytes express low amounts of surface immunoglobulins. 14 However, this dual staining might not always be able to discriminate residual neoplastic cells from -cell progenitors that often are increased in the bone marrow after stem cell transplantation. 15 clonality assay for light chain restriction, quantitative analysis of CD5 expression, or an assay with higher sensitivity and specificity is required to make this distinction. 2,4 4-color assay (CD19/CD5/CD20/CD79b) with improved prediction of outcome and a sensitivity of 10 5 has been described recently. 4 However, this assay entails the use of dual laser technology that might not be readily available at many centers. Thus, for routine clinical practice, dual-color MRD analysis may be performed as a useful adjunct to clinical and hematologic criteria for assessing the response to therapy and subsequent monitoring in patients with CLL. Normal CD5+ cells have dim CD5 expression compared with CLL cells that stain brighter for CD5. 16 In our study, 3 of 40 normal samples had a bright CD5+ -cell subpopulation. The percentage of CD5+ cells per total number of cells was less than our cutoff for MRD in 2 samples and more than the cutoff in 1 sample. So, it was not the percentage of CD5+ cells but the bright CD5 expression that distinguished these three samples from other normal samples. It is possible that bright CD5+ cells represent a subclinical clone of CLL cells. Unfortunately, we C E Image 3 Dilution experiment in which sequentially lower numbers of chronic lymphocytic leukemia (CLL) cells were added to normal peripheral blood samples to evaluate the sensitivity of a 2-color flow cytometric assay to detect CLL cells., % CLL cells., 1% CLL cells. C, 0.1% CLL cells. D, 0.01% CLL cells. E, 0.001% CLL cells. F, % normal cells. -F, CD5+ cells representing 45.4%, 27.4%, 23.8%, 21.8%, and 21% of total cells, respectively. y using 25% CD5+ cells/total cells (mean + 2SD) as the cutoff value in peripheral blood samples, the CD5/CD19 assay could identify 1 neoplastic CLL cell among 1,000 normal leukocytes (sensitivity, 10 3 ; 0.1%). FITC, fluorescein isothiocyanate; PE, phycoerythrin. D F m J Clin Pathol 2004;121: DOI: /T5EM9QU9CM8F57 371

5 Gupta et al / CD5+ -CELL NLYSIS Y FLOW CYTOMETRY Table 4 Comparative nalysis of Normal CD5+ Cells in Different Studies CD5+CD19+ Cells/CD19+ Cells (%) * Fluorochrome uthor and Year ge Range (y) Peripheral lood one Marrow Used (CD5/CD19) Vuillier et al, <20 (n = 30) ND FITC/PE Cabezudo et al, (n = 11) 1-15 (n = 10) FITC/PE Sanchez et al, (n = 10) (n = 10) PE/PE-Cy5 Maloum et al, Not reported 6-47 (n = 16) ND CyChrome/PC Present study, (n = 19) (n = 18) FITC/PE PC, allophycocyanin; FITC, fluorescein isothiocyanate; ND, not done; PE, phycoerythrin; PE-Cy5, phycoerythrin CyChrome 5. * Data are given as range. could not evaluate the cells for clonality or other CLLspecific markers because this was a chance observation made at the time of analysis. Rawstron et al 17,18 recently reported a monoclonal -cell population with a CLL immunophenotype in 3.5% of healthy adults and in 13.5% of healthy blood relatives of patients with CLL and argued that the presence of an abnormal clone represented the earliest stage of indolent CLL. longitudinal study in such individuals might further our understanding of the pathogenesis of CLL. 17 In view of the simplicity and wider availability of the CD5/CD19 assay, it might be worthwhile to evaluate its usefulness in screening healthy people for subclinical CLL. Our results suggest that the CD5/CD19 assay has clinically useful sensitivity for assessing response to treatment and monitoring MRD in patients with CLL. Fluorochrome selection is an important variable that might influence the sensitivity of the MRD assay. The usefulness of the CD5/CD19 assay as a screening tool to identify the earliest stage of indolent CLL in healthy people requires further evaluation. From the 1 Laboratory Oncology Unit and the Departments of 2 Surgical and 3 Medical Oncology, Institute Rotary Cancer Hospital, ll India Institute of Medical Sciences, New Delhi. ddress reprint requests to Dr Jain: 4/42, zad partments, Sri urobindo Marg, New Delhi - 116, India. References 1. Cheson D, ennett JM, Grever M, et al. National Cancer Institute sponsored Working Group guidelines for chronic lymphocytic leukemia: revised guidelines for diagnosis and treatment. lood. 1996;87: Cabezudo E, Matutes E, Ramrattan M, et al. nalysis of residual disease in chronic lymphocytic leukemia by flow cytometry. Leukemia. 1997;11: Pfitzner T, Engert, Wittor H, et al. real-time PCR assay for the quantification of residual malignant cells in cell chronic lymphatic leukemia. Leukemia. 2000;14: Rawstron C, Kennedy, Evans P, et al. Quantitation of minimal disease levels in chronic lymphocytic leukemia using a sensitive flow cytometric assay improves the prediction of outcome and can be used to optimize therapy. lood. 2001;98: Vuillier F, Claisse JF, Vandenvelde C, et al. Evaluation of residual disease in -cell chronic lymphocytic leukemia patients in clinical and bone marrow remission using CD5- CD19 markers and PCR study of gene rearrangements. Leuk Lymphoma. 1992;7: oumsell L, ernard, Lepage V, et al. Some chronic lymphocytic leukemia cells bearing surface immunoglobulins share determinants with T cells. Eur J Immunol. 1978;8: Kipps TJ, Vaughan JH. Genetic influence on the levels of circulating CD5 lymphocytes. J Immunol. 1987;139: Kipps TJ. The CD5 cell. dv Immunol. 1989;17: Maloum K, Sutton L, audet S, et al. Novel flow-cytometric analysis based on CD5+ subpopulations for the evaluation of minimal residual disease in chronic lymphocytic leukaemia. r J Haematol. 2002;119: Sanchez ML, lmeida J, Vidriales, et al. Incidence of phenotypic aberrations in a series of 467 patients with chronic lymphoproliferative disorders: basis for the design of specific four-color staining to be used for minimal residual disease investigation. Leukemia. 2002;16: Matutes E, Morillo R, Catovsky D. Immunophenotyping. In: Lewis SM, ain J, ates I, eds. Dacie and Lewis Practical Haematology. 9th ed. London, England: Churchill Livingstone; 2001: Vuillier F, Scott-lgara D, Dighiero G. Extensive analysis of lymphocyte subsets in normal subjects by three-color immunofluorescence. Nouv Rev Fr Hematol. 1991;33: Tung JW, Kunnavatana SS, Herzenberg L, et al. The regulation of CD5 expression in murine T cells. MC Mol iol. 2001;2: Lenormand, izet M, Fruchart C, et al. Residual disease in -cell chronic lymphocytic leukemia patients and prognostic value. Leukemia. 1994;8: ntin JH, ult K, Rappeport JM, et al. lymphocyte reconstitution after human bone marrow transplantation: Leu-1 antigen defines a distinct population of -lymphocytes. J Clin Invest. 1987;80: Lavabre-ertrand T, Janossy G, Exbrayat C, et al. Leukemiaassociated changes identified by quantitative flow cytometry, II: CD5 over-expression and monitoring in -CLL. Leukemia. 1994;8: Rawstron C, Green MJ, Kuzmicki, et al. Monoclonal lymphocytes with the characteristics of indolent chronic lymphocytic leukemia are present in 3.5% of adults with normal blood counts. lood. 2002;: Rawstron C, Yuille MR, Fuller J, et al. Inherited predisposition to CLL is detectable as subclinical monoclonal -lymphocyte expansion. lood. 2002;: m J Clin Pathol 2004;121: DOI: /T5EM9QU9CM8F57