Indolent Systemic Mast Cell Disease in Adults: Immunophenotypic Characterization of Bone Marrow Mast Cells and Its Diagnostic Implications

A total of 12 BM samples were obtained from 10 patients suffering from indolent SMCD (six men and four women), with a mean age of 45 ± 14 years, ranging from 26 to 67 years. Cases of mastocytosis associated with hematologic disorders, aggressive mastocytosis, or mast cell leukemia were excluded from this study. The diagnosis of indolent SMCD was made according to the criteria proposed in previous reports.1,2,14-16 Table 1 shows the clinical and biological characteristics of each patient enrolled in the present study. Immunophenotypical studies were performed at diagnosis in eight cases (patients 1 to 7, and 9) and 1 and 8 years after the diagnosis in the other two cases (patients 10 and 8, respectively). At the moment of entering this study, five patients (2, 7, 8, 9, and 10) were under anti-H1 (chlorpheniramine maleate or hydroxyzine) and anti-H2 (ranitidine) therapy. Of these, three cases were receiving treatment with sodium cromoglycate (patients 7, 8, and 9) and one (patient 8) also received aspirin; no case was treated with interferon before entering this study. A total of 19 healthy volunteers undergoing either orthopedic surgery or BM harvest were sudied as normal controls.

In all cases, samples were collected in K₃-EDTA anticoagulant and immediately diluted 1/1 (vol/vol) in phosphate-buffered saline (PBS). After collection, the sample was passed several times through a 25-mm gauge needle to disaggregate the BM particles and cell concentration was adjusted to 7.5 × 10⁹ nucleated cells/L.

BM samples were analyzed by direct immunofluorescence using either triple or double-staining combinations of MoAb directly conjugated with fluorescein isothiocyanate (FITC), phycoerythrin (PE), and either the PE-cyanin 5 (PE-Cy5) fluorochrome tandem or PerCP. The following MoAb conjugates were used: (1) FITC-conjugated: CD2, CD10, CD13, CD14, CD15, CD16, CD19, CD20, CD22, CD25, CD33, CD34, CD44, CD71, (purchased from Becton Dickinson, San Jose, CA), CD38, CD65, (Caltag Laboratories, San Francisco, CA), CD11a, CD11b, CD11c, CD35, CD54, CD66 (CLB, Amsterdam, The Netherlands), CD29 (Coulter Corp, Miami, FL) CD21, CD30, CD42b, CD61 (Dakopatts, Copenhagen, Denmark), CD41a (Immunotech, Marseille, France), CD18, CD51 (Menarini Diagnostics, Barcelona, Spain), CD117 (Nichirei Corporation, Tokyo, Japan), CD9, CD43, CD49d, CD49e (Serotec, Oxford, UK) anti-IgE (The Binding Site, Birmingham, England), and BB4 (CD138) (Immunoquality Products, Gröningen, The Netherlands); (2) PE-conjugated: CD23 (purchased from Becton Dickinson), and CD117 (Nichirei Corporation); (3) PE/Cy5-conjugated: CD38, and HLA-DR (Caltag Laboratories); (4) Per-CP conjugated CD45 (Becton-Dickinson).

Briefly, 200 μL of the sample containing aproximately 1.5 × 10⁶ nucleated cells were incubated for 10 minutes at room temperature with the above mentioned MoAbs. After this, erythrocytes were lysed by incubating cells for 10 minutes (room temperature) with 2 mL/tube of FACS lysing solution (Becton Dickinson) diluted 1/10 vol/vol in distilled water. Isotype-matched mouse nonspecific immunoglobulins and a tube stained for CD117-PE were used as negative and positive controls to assess nonspecific binding and BMMC autofluorescence, respectively.

Data acquisition was performed on a FACScan flow cytometer (Becton Dickinson) using the LYSYS II software program (Becton Dickinson). Initially, a minimum of 10,000 events/tube from the total BM cells was acquired. In addition, a second acquisition step, through a SSC/CD117 live gate, was performed to increase the sensitivity of the method for the analysis of MC present at a low frequency.8For data analysis, the Paint-A-Gate PRO software (Becton Dickinson) was used. The quantitation of positivity for each of the markers tested was performed using QuickCal beads (Flow Cytometry Standards Corporation, San Juan, Puerto Rico) and results were expressed as the mean number of molecules equivalent of soluble fluorochrome (MESF) obtained specifically for the MC. The threshold for positivity was 5,000 MESF for both FITC and PE after subtracting the autofluorescence levels obtained for BMMC.

Mean values and their standard deviations were calculated for all variables in each group of samples. The Mann-Whitney U and the χ² tests were used to assess the statistical significance of the immunophenotypic differences observed between BMMC from SMCD patients and those of normal controls for continuous and dichotomic variables, respectively.

Despite their low frequency, MCs were clearly identified in all BM samples analyzed based on their strong CD117 expression (Fig 1). The mean number of BMMC in indolent SMCD patients was significantly higher (P > .001) than in normal controls (0.27% ± 0.19% and 0.021 ± 0.0025 of the nucleated BM cells analyzed, respectively). BMMC from indolent SMCD displayed a relatively homogenous FSC and SSC distribution, the values of both light scatter parameters being higher than those found for normal BMMC (Fig 2A and B). Moreover, the autofluorescence level was also higher in BMMC from indolent SMCD than it was in normal individuals (Fig 2C and D).

The immunophenotypical characteristics of BMMC from indolent SMCD are shown in Fig 2 and Table 2. As may be seen, three patterns of antigen expression were detected in BMMC from indolent SMCD patients: (1) markers that were constantly positive (CD2, CD9, CD11c, CD13, CD25, CD29, CD33, CD35, CD44, CD45, CD54, CD117, and FcεRI); (2) antigens that were always negative (CD10, CD11a, CD14, CD15, CD16, CD19, CD20, CD21, CD23, CD30, CD34, CD38, CD65, CD66, CD71, HLA-DR, and CD138); and (3) markers that were positive in a variable proportion of cases: CD11b (25%), CD18 (40%), CD22 (78%), CD41a (71%), CD42b (25%), CD43 (60%), CD49d (50%), CD49e (25%), and CD61 (28%). In addition, among those antigens that were positive, three groups could be identified according to the intensity of antigen expression detected: (1) strong positive markers (> 12,000 and > 50,000 MESF for FITC and PE, respectively): CD9, CD11c, CD25, CD33, CD44, CD49d, CD54, CD117 and FcεRI; (2) uniform dim positive antigens (from 5,000 to 12,000 MESF of FITC): CD2, CD11b, CD18, CD41a, CD49e, and CD51; and (3) antigens showing a variable intensity of expression (from 5,000 to 51,000 MESF of FITC): CD13, CD22, CD29, CD42b, CD43, and CD61.

After comparing the immunophenotype of BMMC from indolent SMCD patients and healthy subjects, significant differences were observed regarding both the incidence of positivity and the fluorescence intensity (Table2). Accordingly, in all indolent SMCD patients, BMMC expressed the CD2, CD25, and CD35 antigens, which were never found in BMMC from normal controls (P = .0001, .0001, and .006, respectively). In contrast, the CD71 molecule was constantly present in normal controls, but never detected in indolent SMCD cases (P = .004). Other molecules such as CD41a and CD42b were present in a variable proportion of SMCD cases, but constantly absent in normal controls (P = .6 and .06, respectively). Additional differences between BMMC from indolent SMCD cases and healthy individuals were observed upon analyzing the levels of mean fluorescence intensity estimated for several antigens: CD33 expression was significantly higher in SMCD patients (P = .02), while CD29 and CD117 expression was greater in healthy subjects (P = .03 and .008, respectively).

The present report represents a first attempt at the extensive characterization of the immunophenotype of BMMC from indolent SMCD adult patients using a large panel of MoAb. Our major goal was to explore the use of immunophenotyping for the differential diagnosis between SMCD and normal/reactive MC. Although, as expected, the mean number of BMMC in indolent SMCD was significantly higher than in normal controls,8 the overall number of BMMC in indolent SMCD was low. This is concordant with histologic studies that show a low number of MC^17,18 with focal distribution.2,17,19-23 

Interestingly, BMMC from SMCD displayed light scatter characteristics that were higher than those found for their normal counterpart, as well as a greater level of autofluorescence. Light scatter properties of cells analyzed at flow cytometry to a large extent reflect morphologically-related features such as cell size and internal complexity.24 As a matter of fact, a slightly larger size has been reported for BMMC in SMCD25 and this could explain the higher FSC values found in the present study. Interestingly, both piecemeal and anaphylactic degranulation have been described in vivo at the ultrastructural level in MC from different tissues,26-30 including BMMC31 from SMCD patients. Accordingly, MC undergoing degranulation display several changes regarding granule size, shape, and contents,29,30,32 and these morphologic changes, together with an increased content in endoplasmic membranes, Golgi apparatus and endoplasmic reticulum, may explain the increased SSC (internal complexity) of BMMC from SMCD found in the present study. Additionally, this would also help to increase the autofluorescence levels of SMCD BMMC. Isolation of mast cells from a SMCD using a fluorescence activated cell sorter according to previously described methods8 confirmed that these MC displayed an abnormal morphology (data not shown).

Increased autofluorescence of SMCD BMMC may limit the measurement of the fluorescence levels obtained for specific antigens detected through the use of immunofluorescence techniques. Therefore, the assessment of the immunophenotype of BMMC from SMCD patients should take into account the autofluorescence levels of these cells, and the fluorescence intensity for a specific marker should be calculated after subtracting the baseline autofluorescence of the cells under study. To allow the comparison of the results of the present report with those of other studies in which the same MoAb conjugates are used, results on the reactivity obtained for each of the markers analyzed were expressed as MESF after subtracting the MESF values corresponding to the mean autofluorescence level obtained for those specific mast cells.

From the immunophenotypic point of view, the most striking findings were the expression of CD2, CD25, and CD35 molecules by BMMC from all SMCD patients, markers that were constantly absent in normal BMMC.3,5,9,33,34 Coexpression of both CD2 and CD25 by BMMC could be considered as characteristic of indolent SMCD, but not normal and reactive BMMC,9 and it could be of great help for the differential diagnosis between SMCD and reactive mastocytosis. Nevertheless, reactivity for CD35 cannot be pathognomonic of SMCD, because although negative in MC from several tissues5,7,34-36 including normal BM,9 it has been observed to be weakly positive in BMMC from some patients (27%) with B-cell malignancies.9 On the other hand, the reactivity for CD71 (transferrin receptor) was shown to be constantly negative in indolent SMCD cases, while positive in BMMC from normal subjects,9 as well as in the MC from a case of mast cell leukemia.12 However, because CD71 expression in normal BMMC was always weak (from 5 to 6.5 × 10³ MESF), this would limit its use in the differential diagnosis of SMCD. It should be noted that this antigen is absent in human MC from different tissues.3,5,35 

The CD29 and the myeloid-associated markers CD33 and CD117, although present in BMMC from all indolent SMCD patients and normal controls studied, displayed a significantly different intensity of expression in both groups of individuals. In this sense, while reactivity for CD33 was higher in SMCD patients, CD29 and CD117 were expressed at a greater intensity in BMMC from control subjects.

In summary, our results show that multiparametric flow cytometry using direct immunofluorescence on erythrocyte-lysed whole BM is of great help for the diagnosis of adult indolent SMCD and its differential diagnosis with reactive mastocytosis. Simultaneous assessment of FSC/SSC and reactivity for the CD117, CD2, CD25, CD29, and CD33, in the presence of BM involvement, forms the basis for the immunophenotypic diagnosis of SMCD in adults.

We thank Professor Frank K. Austen for reading the manuscript.