Bone marrow findings in rheumatoid arthritis patients with peripheral cytopenias

Objective The aim was to study bone marrow (BM) cellularity, morphological characteristics, and types of cellular infiltrates in rheumatoid arthritis (RA) patients associated with peripheral cytopenias. Background RA is a systemic, autoimmune disease leading to joint destruction. BM is an important compartment in RA, where there are autoreactive lymphocytes, abnormal production of cytokines, and abnormal BM microenvironment, which may affect the integrity of the BM and accordingly may require specific modalities in treatment. Patients and methods In the current study, we examined 60 RA patients with cytopenias (defined according to WHO as hemoglobin level below 12 g/dl (woman), 13 g/dl (man), platelet count below 150×109/l, absolute neutrophil count below 1.8×109/l, and lymphocytes below 1.0×109/l. The diagnosis of RA was made according to the American College of Rheumatology (2010) criteria. All patients were subjected to thorough history taking, physical examination, imaging studies including abdominal ultrasound, and chest radiography and laboratory investigations including routine investigations, blood film, antinuclear antibody, antidouble stranded DNA antibody, rheumatoid factor, anticyclic citrullinated peptide (anti-CCP) antibody, and BM aspiration and biopsy. Flow cytometry immunophenotyping was performed to detect lymphocytic infiltration in the BM and its subtypes. Results The patients were predominantly women (91.7%); 40 (66.6%) patients had normocellular BM, 16 (26.6%) patients had hypercellular and four (6.6%) patients had hypocellular BM. Dysmyelopoiesis was present in five (8.3%) patients; dyserythropoietic and megaloblastoid changes were present in 26 (43.3%) patients whereas dysmegakaryopoiesis was present in six (10%) patients. BM eosinophils mean±SD was 2.95±1.51%. BM plasma cell mean±SD was 4.0±2.35. BM lymphocytes mean±SD was 15.65±6.75% and was predominantly T-lymphocytes (mean±SD: 61.63±8.69%). BM-activated (CD25+) T-lymphocytes mean±SD was 23.8±11.5. Higher BM lymphocytes, T-cells, and plasma cell percentages were statistically associated with increased frequencies of positive anti-CCP, neutropenia, thrombocytopenia, and arthritis and were positively correlated to inflammatory markers (erythrocyte sedimentation rate, C-reactive protein, and ferritin).The percentage of activated (CD25+) BM T-lymphocytes was higher in RA patients with either positive C-reactive protein, positive anti-CCP, elevated ferritin, or pancytopenia. There was no significant relation between lymphocytic infiltration (either B-lymphocyte or T-lymphocyte) and type of BM cellularity. Conclusion Lymphocytic infiltration, especially T-lymphocytes is closely related to the activity markers in RA and to the type and degree of cytopenia in RA patients.


Introduction
Rheumatoid arthritis (RA) is a systemic autoimmune disease characterized by chronic inflammation of the joint synovial tissue and subsequent destruction of the associated bone, cartilage, and soft tissues. T-cellmediated immune response is considered as a critical contributor in RA initiation and progression. It has been hypothesized that autoreactive T-cells escaping negative selection can recognize arthrogenic antigens and lead to autoimmunity and tissue destruction [1].

The pathogenic roles of circulating and tissue-localized
This is an open access journal, and articles are distributed under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 License, which allows others to remix, tweak, and build upon the work non-commercially, as long as appropriate credit is given and the new creations are licensed under the identical terms.
B-cells in RA can occur through several mechanisms that include autoantibody production, T-cell activation, and cytokine synthesis. Besides, the potential contribution of B-cells to T-cell differentiation and function, there is now growing evidence indicating that B-cells in RA can directly contribute to the local synthesis of a wide spectrum of cytokines acting on other pathologically relevant cell types [2]. The localization of inflammation to the synovium is primarily responsible for the dysregulated cellular and molecular mechanisms that ultimately lead to the typical features of RA, including joint pain, swelling, and structural changes. It is becoming increasingly apparent that a 'synoviocentric' model of RA could be limiting. Indeed, other anatomic compartments appear to be involved at all stages of the disease. The most important example comes from the recognition that clinical arthritis and subclinical synovitis are anticipated by a prearticular immunologic phase possibly developing in lymphoid tissues as well as in the lungs [3,4]. Full-blown joint disease itself spreads well beyond synovial tissue inflammation. Established arthritis can be characterized by the involvement of at least two other compartments that are in direct contact with the joint space. These include the draining lymph nodes [5] and, of special relevance, the subchondral bone marrow (BM) [6]. The aim of this work was to study the morphological characteristics of the BM and the types of cellular infiltrates in the BM in patients with RA associated with peripheral cytopenias.

Patients and methods
Sixty patients were selected from the outpatient and inpatient clinics of Rheumatology and Hematology Units, Menoufia University Hospital, in the period from January 2017 to January 2018. Informed consents were obtained from all patients in accordance with the local ethics committee. The diagnosis of RA was made according to the American College of Rheumatology (2010) [7]. Cytopenias were defined according to WHO as hemoglobin level below12 g/dl (woman) and 13 g/dl (man), platelet count below 150×10 9 /l, absolute neutrophil count below 1.8×109/l, and lymphocytes below 1.0×10 9 /l [8]. All patients were subjected to medical history taking and complete physical examination. Investigations were done for all patients including routine investigations (kidney function tests, liver function tests, fasting and postprandial blood glucose and complete blood count), blood film, iron profile, anti-HCV antibody, HBsAg, direct antiglobulin test, serology [rheumatoid factor (RF), antinuclear antibody (ANA), anti-dsDNA, antiphospholipid, and anticyclic citrullinated peptide (anti-CCP) antibodies], and BM aspiration and biopsy. Abdominal ultrasound was performed to detect organomegaly (splenomegaly and hepatomegaly). Chest radiography was done to detect any suspected hilar lymphadenopathy or pulmonary fibrosis. Flow cytometric analysis of BM lymphocytes was studied. Patients aged less than 18 years old, patients known to have other causes of cytopenias (like thyroid disorders, chronic kidney disease, iron deficiency anemia, myelodysplastic syndrome), patients with HCV Ab, or HBsAg-positive patients, patients of hematological malignancies, or other collagen vascular disease were excluded from this study. RA patients without cytopenias were also excluded. A volume of 8 ml of blood was drawn through a venous access under strict aseptic conditions; 2 ml was delivered into an EDTA tube, mixed well, and were used to perform CBC (Sysmex XN-1000 hematology analyzer; Sysmex Corporation, Kobe, Japan), blood smears (to detect morphological features and exclude spurious cytopenias), and direct antiglobulin test. The remaining 6 ml was delivered into two plain tubes, left to clot then centrifuged, and sera were separated. Sera were used to perform the other investigations including liver functions, renal functions, blood sugar, iron profile, HCV Ab, and HBsAg (Cobas 6000 analyze series; Roche Diagnostics International Ltd, Rotkreuz, Switzerland). ANA test was done by immunofluorescence by incubating dilutions of patient sera with a monolayer of fixed, permeabilized cultured HEp-2 cells. Antibodies adherent to the cell monolayer were visualized with an antihuman immunoglobulin fluorescent reagent. The presence or absence of nuclear staining and the pattern of nuclear staining by fluorescence microscopy was identified. ELISA method was used to detect and measure anti-dsDNA antibody quantified in units/ml. RF was detected by its ability to attach to the Fc portion of IgG coating latex particles. A positive result is reported if agglutination occurs when equal volumes are mixed. Anti-CCP was performed on Cobas 6000 analyzer series (Roche Diagnostics International Ltd). BM aspiration was done from the posterior superior iliac crest under local anesthesia using lignocaine after taking the patient's consent using Salah and Klima aspiration needles to study the M/E ratio, morphology and adequacy of hematopoietic cells, the percentage of blasts, lymphocytes, histiocytes and plasma cells and BM iron. BM biopsy was done by Islam BM biopsy needles under local anesthesia to asses BM cellularity, infiltrations, bone trabeculae, abnormal localization of immature precursors, and BM fibrosis using reticulin stain. BM aspirate flow cytometry immunophenotyping was performed on BD FACS Calibur (BD Immunocytometry Systems, San Jose, California, USA). Quantification of B was done using BD Pharmingen conjugated mouse monoclonal antihuman CD19, isotype IgG1, κ, catalog number 555413, while quantification of T was done using BD Pharmingen FITC-conjugated mouse monoclonal antihuman CD3, isotype IgG2a, κ, catalog number 555339. To study lymphocyte activation status BD Pharmingen, PerCP mouse IgG1, κ, antihuman CD25 immunoglobulin, catalog number 560503 was used. One hundred microliters of BM aspirates was added to a tube; then 5 μl of each of the monoclonal antibodies was added. An isotypic control and an auto control tube were involved. The tubes were well and gently mixed and incubated for 15 min in the dark at room temperature (18°-25°c). Cells were subjected to red blood cell lysis by 2 ml of lysing solution for 3 min. Then the tubes were washed three times in PBS and finally the cells were suspended in 250 μl of PBS for final flow cytometric analysis. Gating was done on lymphocyte population and the percentage of each subpopulation was measured.

Statistical analysis
Data were collected, tabulated, and statistically analyzed using an IBM personal computer with the statistical package for social sciences (SPSS) version 22 (SPSS Inc., Chicago, Illinois, USA). Quantitative data were presented in the form of mean, SD, and range. Qualitative data were presented in the form of numbers and percentages and analyzed by applying χ 2 and Fisher's exact tests. Independent sample t and analysis of variance tests were used for comparison between means of normally distributed quantitative variables. Mann-Whitney and Kruskal-Wallis tests were used for not normally distributed quantitative variables. Spearman's correlation was used to assess the strength and direction of correlation among nonparametric variables. A P value of less than 0.05 was considered statistically significant.

Results
Descriptive demographic and clinical data of RA patients are detailed in Table 1. The mean age of RA patients was 40.01±9.05 years with a range of 24-55 years. They showed female predominance with women constituting 91.7%. Arthralgia and morning stiffness were the most frequent clinical symptoms (85 and 83.3%, respectively), while arthritis and splenomegaly were the most frequent clinical signs (66.7 and 50%, respectively) ( Table 1). All patients were RF positive while anti-CCP was positive only in 34 (56.7%) patients. Erythrocyte sedimentation rate (ESR) was elevated in all patients with nine patients having an ESR higher than 100 mm/first hour while the C-reactive protein (CRP) was positive in 49 (81.75%) patients with five patients having CRP levels more than 100 mg/l. Similarly, serum ferritin was elevated in 48 patients with 13 patients having elevated serum ferritin of more than 1000 ng/ ml. ANA was positive in three patients. All patients were negative for anti-dsDNA and antiphospholipid antibodies (Table 2). Hemogram and BM findings are detailed in Table 3. Hemoglobin mean±SD was 8.56 ±1.12 g/dl while the mean corpuscular volume mean±SD was 86.90±7.59 fl. Platelet count mean±SD was 198.41 ±105.60 ×10 3 /μl while the total leukocytic count mean±SD was 6.09±2.53×10 3 /μl. Peripheral blood lymphocyte absolute count mean±SD was 2.03 ±0.75×10 3 /μl, while the neutrophil count mean±SD was 3.52±1.86×10 3 /μl. Reticulocytic percentage mean ±SD was 1.25±0.64%. All patients had anemia with isolated anemia seen in 38 (63.3%) patients, while 20 (33.3%) patients had thrombocytopenia and 18 (30%) patients had neutropenia. No patients had isolated thrombocytopenia or isolated neutropenia. Bicytopenia (anemia and thrombocytopenia) was present in four (6.7%) patients and bicytopenia (anemia and neutropenia) was present in two (3.3%) patients. Pancytopenia was present in 16 (26.7%) patients. All patients were anemic (Table 3), (Fig. 1). BM study has shown that 40 (66.7%) patients had normocellular BM, 16 (26.6%) patients had hypercellular, and four (6.7%) patients had hypocellular BM. M/E ratio was normal in 38 (63.3%) patients, decreased in 16 (26.6%) patients, and high in six (10%) patients. Dysplastic changes were seen in RA patients but dysplasia never affected 10% or more of lineage cells. Dysmyelopoiesis was present in five  (Table 3). Higher BM lymphocytes, T-cells, and plasma cell percentages were statistically associated with increased frequencies of positive anti-CCP, neutropenia, thrombocytopenia, and arthritis and were positively correlated to inflammatory markers (ESR, CRP, and ferritin); however, B-cells had exactly the opposite associations. The percentage of BM eosinophils was higher in anti-CCP-negative RA patients and was negatively correlated to ESR. Neither of BM lymphocytes, T-cells, B-cells, plasma cells, or eosinophil percentages had associations with splenomegaly or BM cellularity (Table 4). The percentage of activated (CD25 + ) BM T-lymphocytes was higher in RA patients with positive CRP, positive anti-CCP, elevated ferritin, and pancytopenia (Table 5). Pancytopenia patients had statistically higher incidence of arthritis, positive anti-CCP, and inflammatory markers (hyperferritinemia, positive CRP) than monocytopenia patients. On the other hand, no significant difference was present between monocytopenia, bicytopenia, and pancytopenia patients as regards BM cellularity (P=0.256), BM dysplasia (P=0.605), or splenomegaly (P=0.491) ( Table 6) Discussion Alongside its hematopoietic function and its role in the early selection of lymphocytes, the BM is an immuneregulatory organ involved in migration, selective retainment, and function of innate and adaptive All patients had negative anti-dsDNA and antiphospholipid antibodies. ANA, antinuclear antibody; CCP, cyclic citrullinated peptide; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate.

Figure 1
Type of cytopenias in the studied patients with rheumatoid arthritis.
immune cells [9]. A variety of morphological, immunophenotypic, and functional abnormalities in BM cells in RA have been detected as a consequence of augmented local production of inflammatory cytokines and cell-cell interactions. Both myeloid and lymphoid lineage cells appear affected at some stage of the disease [10]. Previous studies by Rudnicka et al. [11], Kuca-Warnawin et al. [12] and Kurowska and Kuca-Warnawin [13] showed the presence of cellular infiltrates that replace yellow adipose tissue in the BM of RA patients, described on magnetic resonance scans as BM edema. The cellular infiltrates found in RA BM consist of immunological cells that may form aggregates resembling germinal centers in secondary lymphoid organs. Flow cytometric analysis showed an increased number of mononuclear cells and accumulation of activated T-cells and B-cells in the BM of RA patients. Our results showed more that the BM T lymphocyte population was higher than the Blymphocyte population (61.63±8.69 vs. 35.66±8.57%, respectively). We found lymphoid infiltrate in most patients included in this study and this agreed with Tomita et al. [14] who found that the absolute number of mononuclear cells including lymphocytes was markedly increased in BM aspirates in RA patients; and agreed with Kuca-Warnawin et al. [12] and Engels et al. [15] who found expanded lymphoid aggregates in BM biopsy of RA patients. The lymphoid infiltrate was mainly T-lymphocytes and this may support the significant role of T-cells in RA pathogenesis; this agrees with Tomita et al. [14] and Doita et al. [16] who found abnormal accumulation of Tlymphocytes in the BM of RA patients. We found significant relation between BM lymphocytes, T-cells, and plasma cell percentages with increased frequencies of positive anti-CCP, neutropenia, thrombocytopenia, and higher levels of inflammatory markers (ESR, CRP, and ferritin). This may indicate a significant relation of lymphoid infiltrate in the BM with RA activity and may confirm the new consideration of the role of BM in RA as it is considered now as an important compartment in RA as mentioned by Kurowska and   Kuca-Warnawin [13]. The same significant relation was present between lymphoid infiltrate and arthritis. This again may indicate a significant role of the BM as one of the important compartments in RA and its relation to synovitis. This agrees with McQueen and Ostendorf [17] and Bugatti et al. [18] who suggest that inflammation in BM arises independently of the pathologic processes operating 'inside the joint,' and represents an early immunopathological event in RA. Moreover, Bugatti et al. [18] mentioned that data from experimental animal studies showed the enlargement of canals in the cortical bone that connects BM to the synovium in the preclinical phase of RA. This phenomenon may facilitate migration of precursor cells of synoviocytes and osteoclasts from BM directly to the synovial membrane to excite an inflammatory response and destruction processes.
Activated T-lymphocytes were found in the BM of our patients. This agrees with Kuca-Warnawin et al. [12] The percentage of activated (CD25 + ) BM T-lymphocytes was higher in RA patients with each of positive CRP, positive anti-CCP, elevated ferritin, and pancytopenia. This may support the significant role of T-cells in RA pathogenesis. Our results showed that 30 patients had splenomegaly (50%) and this is in agreement with Nishiya et al. [19], who showed that 52% (26/50) had splenomegaly. Nishiya et al. [19] also noticed increased splenomegaly in patients with leukopenia, but there was no significant difference as regards hemoglobin concentrations or platelet cell counts. This disagrees with our results that showed no significant difference between the presence of splenomegaly and type of cytopenia. In our study, splenomegaly was found in 44.7% of patients with anemia and 50% in patients with bicytopenia and 62.5% of patients with pancytopenia. Although it was insignificant, splenomegaly was present more in RA patients with pancytopenia. Spleen may be enlarged to remove immune complexes and the resultant pancytopenia may be due to hypersplenism. The present study showed that 40 (66.7%) patients had arthritis. Arthritis was present in 50% of patients who presented with anemia, 83.3% of patients with bicytopenia (anemia with thrombocytopenia or leukopenia), and in 100% of patients with showed that anti-CCP positivity is more with the increasing number of cytopenia as 42.1% of patients with anemia, 83.3% of patients with bicytopenia and 93.8% of patients with pancytopenia had positive anti-CCP. Anti-CCP usually predicts severe progressive disease, so the number and degree of cytopenia may predict the progressive disease. This also may add to our understanding of the etiopathogenesis of the different types of cytopenia in RA, which is directly related to the disease activity and may be related to BM suppression by inhibitory cytokines rather than hypersplenism or peripheral destruction.All the patients included in this study had anemia, and we found the increase of lymphocytic infiltration with the increased number of cytopenias to be more evident in patients with pancytopenia. Some of these patients had normocellular, hypercellular, or hypocellular marrow. Some had dysplastic features but the exact relation between lymphocytic infiltration and pancytopenia is not clear. Moreover, there was significant relation between lymphocytic infiltration and thrombocytopenia and neutropenia. T-lymphocytic infiltrate was more prominent than B-lymphocytic infiltrate, which confirms the more important role of T-lymphocyte in the pathogenesis of RA.

Conclusion and recommendations
Lymphocytic infiltration especially T-lymphocytes is closely related to the activity markers and to the type and degree of cytopenia in RA patients. We recommend BM aspiration and biopsy in patients with RA with unexplained cytopenia or if cytopenia persists after adequate treatment. We also recommend further studies on BM microenviroment in RA patients to unravel its role in disease pathogenesis and progression.

Financial support and sponsorship
Nil.

Conflicts of interest
There are no conflicts of interest.