|Home | About | Journals | Submit | Contact Us | Français|
An increase in leucocyte apoptosis and impaired clearance of apoptotic cells has been observed in patients with systemic lupus erythematosus (SLE). Apoptotic cells are likely to be a key source of autoantigens in SLE as they express many of the nuclear autoantigens (in surface blebs and apoptotic bodies) that are relevant to this disease. The clearance of apoptotic cells is usually a rapid process, such that few cells are usually seen in the extracellular environment in vivo. We report a case in which multiple apoptotic bodies were observed in the bone marrow of a patient with SLE that was complicated by an immune‐mediated pancytopenia. We have subsequently examined the frequency of apoptotic cells, identified morphologically, and by caspase‐3 staining in bone‐marrow trephine samples taken from patients with SLE over a 10‐year period of follow‐up. A high proportion of bone marrows contained apoptotic debris. The novel demonstration of apoptotic bodies in vivo in patients with SLE is unusual and supports the notion that the marrow may be a target organ in the disease. Their abundance is also consistent with the hypothesis that normal clearance mechanisms are defective and/or overwhelmed in SLE.
Apoptosis, an active process of programmed cell death, does not release intracellular components into the extracellular space, minimizing the risk of a damaging inflammatory response. The clearance of apoptotic cells is a rapid process, so apoptotic cells are rarely identified in vivo and impaired clearance has been implicated in the pathogenesis of systemic lupus erythematosus (SLE).1 Apoptotic cells express many of the nuclear autoantigens that are relevant to SLE in surface blebs and apoptotic bodies.2 Delayed apoptotic cell clearance leads to prolonged exposure of autoantigens and predisposes to autoantibody production. We report a case in which multiple apoptotic bodies were observed in the bone marrow (BM) of a patient with SLE that was complicated by an immune‐mediated pancytopenia. We subsequently reviewed all SLE BM trephines obtained during 1993–2003, with particular reference to evidence of apoptosis in vivo, which has not previously been reported in the literature.
A 42‐year‐old female, who developed SLE in 1988, presented with arthralgia, malar rash, antinuclear antibodies (ANA), anti‐dsDNA and anti‐Ro antibodies. There was no major organ involvement and treatment consisted of low‐dose prednisolone and hydroxychloroquine. She relapsed in 2003, presenting with fevers, headaches and a worsening rash following prolonged sun exposure. She was pyrexial (40°C), with an erythematous desquamating rash on the face, chest and forearms, livedo reticularis on the legs and cervical lymphadenopathy. There was no focus of infection and the urinary sediment was inactive.
Investigations revealed pancytopenia, with a haemoglobin concentration of 9.4 g/dl, an MCV of 91 fl, a total leukocyte count of 0.9 × 109/l, a neutrophil count of 0.5 × 109/l and a platelet count of 47 × 109/l. The ESR was 55 mm/hr and the C‐reactive proten was 8 mg/l. She was hypocomplementemic, with a CH50 of 35% pooled normal plasma (reference 50–120), a C3 of 0.47 g/l (0.70–1.70) and a C4 of 0.05 g/l (0.16–0.54). The anti‐dsDNA titre was 66 U/ml (<30) and antibodies against nRNP and C1q were identified. Anticardiolipin antibody and lupus anticoagulant tests were negative. The Coombs test was positive, with IgG antibodies against erythrocytes. BM aspirate confirmed the presence of neutrophils with multiple precursors and normal maturation. The BM trephine was hypocellular, with signs of myeloid maturation and neutrophil production. CD34+ cells comprised 0.2% of the marrow cellularity. There was no evidence of lymphoma. There was increased reticulin, reticulin sclerosis and hyperlobulated megakaryocytes. Multiple apoptotic bodies were identified on H&E and Giemsa‐stained sections by their characteristic morphology (fig. 1A–C). Subsequent staining with an antibody against the activated form of caspase‐3 demonstrated that these cells had undergone primary apoptosis (fig. 1D1D).
Treatment was initiated with intravenous methylprednisolone, followed by oral prednisolone 1 mg/kg/day, in addition to intravenous tazocin, gentamicin and fluconazole. However, high fevers persisted and pancytopenia worsened, with a neutropenia of 0.3 × 109/l and thrombocytopenia of 30 × 109/l. She subsequently received cyclophosphamide 1 g intravenously and, 10 days later, the neutrophil and platelet counts were 0.2 × 109/l and 30 × 109/l, respectively. A total of four units of packed erythrocytes were required for anaemia. No organisms were cultured from blood, sputum or urine, and parvovirus B19 serology was negative. A direct platelet immunofluorescence test revealed IgG, IgM and IgA anti‐platelet antibodies, and neutrophil and lymphocyte indirect immunofluorescence tests were positive.
Due to persistent neutropenia and high fevers, recombinant human granulocyte colony‐stimulating factor (G‐CSF) (19.2 mU/m2) was given for 3 days. She became afebrile and the neutrophil count increased, reaching 11.2 × 109/l prior to discharge. She received four further pulses of cyclophosphamide, and remains well with normal platelet and neutrophil counts on maintenance therapy of low‐dose prednisolone and mycophenolate mofetil.
We subsequently examined BMs taken from ten SLE patients with cytopenias seen in our Unit in the preceding 10 years. Apoptotic cells and bodies were identified as above and clinical data obtained for each patient at the time of BM examination.
Ten patients fulfilling American College of Rheumatology criteria were studied (9 females, 1 male; 23–50 years).
Haematological disease developed 1–10 years (median, 4 years) after SLE was diagnosed. Nephritis was present in 5/10 and CNS disease in 3/10 patients.
Three patients had been treated with azathioprine prior to BM examination, but none had received cyclophosphamide.
The haematological diagnoses were: pancytopenia, 4/10; leucopenia, 3/10; anaemia, 2/10. Neutropenia was seen in 6/10 and lymphopenia 7/10 patients (Table 11).
The ESR was raised in all patients (range, 40–130 mm/hr). Haematinics were normal in all patients, with the exception of iron deficiency in two patients.
Anti‐nRNP was positive in 7/10, anti‐Ro in 5, anti‐C1q in 4 and anti‐dsDNA in 9/10 patients, and 4/10 patients had a positive Coombs test. C4 was low in 9/10 and C3 in 6/10 patients. In the majority of cases, the BM appeared reactive. Hypocellularity was observed in all cases. Megakaryocyte clusters were present in 5/10, plasmacytosis in 6/10 and increased reticulin in 7/10 patients.
Apoptotic bodies were present in 8/10 BM samples (fig. 11).). Four of these patients were receiving azathioprine. Disease activity at the time of bone‐marrow sampling was measured using the SLEDAI. Disease was regarded as active (SLEDAI >10) in 8/10 patients. Although apoptotic bodies were detected in the marrow of six of the eight patients with active disease, there was no direct correlation with SLEDAI. A preponderance of T cells was observed in 3/10 marrows, most marked in the index case, in which T cells constituted 30% of marrow cells. The main BM findings are summarised in Table 11.
The predominant BM findings of hypocellularity and plasmacytosis were consistent with previous studies of SLE BM.3,4 However, most striking was the novel finding of significant apoptotic bodies in the marrow. In the index case, ultraviolet light exposure was the likely precipitant. UVB light renders human keratinocytes apoptotic and complement facilitates the clearance of apoptotic cells through the binding of C1q to surface blebs.5 The complement system is of central importance in apoptotic cell clearance in SLE and we have shown that C1q‐deficient mice develop ANA and glomerulonephritis associated with apoptotic bodies in glomeruli6, and mice deficient in either C1q or C4 have impaired uptake of apoptotic thymocytes by peritoneal macrophages.7 Impaired uptake of apoptotic cells by monocyte‐derived macrophages from patients with SLE has been observed1,8 and defective phagocytosis of apoptotic cells by macrophages deficient in C1q reported,9 with the defect corrected by addition of C1q. These studies illustrate the importance of the classical pathway in apoptotic cell clearance.
The removal of apoptotic cells in vivo is normally so rapid that few are seen in tissues such as the thymus, where 95% of cells undergo apoptosis. Their presence in BM is rare and they are not typically seen in normal BM, although we have reported their occurrence in myelodysplasia.10 Few studies have reported the presence of apoptotic debris in the peripheral blood or tissues from SLE patients. Increased spontaneous apoptosis of SLE neutrophils in culture8 and peripheral blood lymphocytes in patients has been observed.11,12 A reduction in CD34+ cells in the SLE BM has been reported, correlating with an increased number of apoptotic CD34+ cells and CD34+/Fas+ cells in cultures of BM mononuclear cells.13 SLE patients had fewer G‐CSF units and impaired formation of colony‐forming cells in BM cultures. These observations, together with our demonstration of multiple apoptotic bodies, suggest that the BM is a target organ in SLE. This may be mediated by autoreactive lymphocytes and antibodies against pluripotent stem cells,14 a hypothesis supported by our observation of a high proportion of T cells in the BM. Herein, we have identified apoptotic bodies by their characteristic morphology and by caspase‐3 staining.
Apoptotic debris has been observed in the germinal centres of lymph nodes from patients with SLE when compared with benign follicular hyperplasia controls.15 Apoptotic material was associated with the surface of follicular dendritic cells and tingible body macrophages containing apoptotic cells were reduced in SLE. The latter observation was consistent with impaired apoptotic cell clearance in SLE, leading to the unexpected demonstration of apoptosis in vivo. No correlation between apoptotic debris in vivo and disease activity was found.15 In our series, SLE was active in 8/10 patients and apoptotic bodies in the marrow were more common in the presence of pancytopenia. Apoptotic cells were not detected in the BM of two patients with cytopenias. This is in concordance with data from lymph nodes in SLE15 and is likely to reflect different clearance rates between individuals.
In conclusion, we report a case of SLE complicated by immune‐mediated pancytopenia and an increase in apoptotic cells in the BM of a cohort of SLE patients. The demonstration of apoptotic bodies in vivo is unusual and supports the notion that the marrow may be a target organ in SLE. Autoantibodies may have a role in scavenging apoptotic cells and their abundance suggests that normal clearance mechanisms are defective or overwhelmed in such cases. This may provide a source of autoantigens, initiating an autoimmune response, complement activation, inflammation and tissue damage.
ANA - antinuclear antibodies
BM - bone marrow
ESR - Erythrocyte Sedimentation Rate
MCV - Mean Corpuscular Volume
SLE - systemic lupus erythematosus
SLEDAI - SLE disease activity index
Conflicts of interest: None declared.