Dr. Bhawan received consultant fees (more than $10,000) from Galderma for contract work for a research project. Dr. Kantarci holds a patent for “Delivery of H2 Antagonists” (patent WO2006/071659A1).
Multicentric reticulohistiocytosis: A systemic osteoclastic disease?†
Article first published online: 29 FEB 2008
Copyright © 2008 by the American College of Rheumatology
Arthritis Care & Research
Volume 59, Issue 3, pages 444–448, 15 March 2008
How to Cite
Codriansky, K. A., Rünger, T. M., Bhawan, J., Kantarci, A. and Kissin, E. Y. (2008), Multicentric reticulohistiocytosis: A systemic osteoclastic disease?. Arthritis & Rheumatism, 59: 444–448. doi: 10.1002/art.23320
Presented at the New England Dermatological Society meeting, Boston, Massachusetts, November 2006.
- Issue published online: 29 FEB 2008
- Article first published online: 29 FEB 2008
- Manuscript Accepted: 6 SEP 2007
- Manuscript Received: 22 MAY 2007
Multicentric reticulohistiocytosis (MRH) is a rare systemic disorder of unknown etiology, characterized by erosive polyarthritis and papulonodular lesions on the skin, mucous membranes, and internal organs. MRH is the most destructive chronic inflammatory arthritis, progressing to arthritis mutilans in 45% of cases (1), compared with 5% for rheumatoid arthritis and psoriatic arthritis (2, 3). Cutaneous lesions are pink-brown, asymptomatic, firm papules and nodules, most commonly located over the dorsal aspects of proximal and distal interphalangeal joints of the hands, on elbows, and on the forehead. Such papules located along the proximal nailfold have been described as “coral beads” and are very characteristic of the disease (1). Histologically, cutaneous nodules and synovium demonstrate infiltrates of multinucleated giant cells with eosinophilic, periodic acid–Schiff–positive granular material in the cytoplasm, giving the appearance of “ground glass” (4). MRH nodules are reported to stain with CD68 and CD45, but not with S100 or CD1a (5), and are therefore thought to be of monocyte/macrophage lineage. These cells also have been shown to stain positively with the osteoclast marker tartrate-resistant acid phosphatase (TRAP) (5). MRH is difficult to control with common immunosuppressive agents (5–7). Recently, there have been several reports of intravenous bisphosphonates showing efficacy in MRH arthritis and nodulosis (5–7), further implicating osteoclasts in the pathogenesis of MRH.
We present an additional case of MRH with novel histologic evidence of osteoclast-like cells in nodular skin lesions, as well as response to treatment with zoledronic acid, an intravenous bisphosphonate.
A 40-year-old woman with an 8-year history of Sjögren's syndrome manifested by xerostomia and xerophthalmia with positive anti-Ro, anti-La, and antinuclear (1:1,280; speckled pattern) antibodies developed inflammatory polyarthritis involving the proximal interphalangeal (PIP), metacarpophalangeal (MCP), wrist, elbow, and knee joints. Knee aspiration revealed 9,000 white blood cells/μl, 54% mononuclear cells, 36% polynuclear cells, and 10% lymphocytes with no crystals seen on polarized light microscopy. After serologic tests yielded positive rheumatoid factor (75 IU/ml) and anti–cyclic citrullinated peptide antibody (29 units), the patient was initially diagnosed with rheumatoid arthritis. Treatment with a combination of prednisone, hydroxychloroquine, and methotrexate failed to control the arthritis. Eight months after the start of joint symptoms, the patient suddenly developed asymptomatic, skin-colored to slightly erythematous, firm, nontender papules and nodules over the proximal and distal interphalangeal joints, elbows, forehead, and ears (Figure 1A and 1B). Characteristic “coral beads” (Figure 1B) were noted over the proximal nailfolds. The distal interphalangeal joints (DIP), PIP joints, MCP joints, and knee joints were effused with decreased range of motion, warmth, and tenderness to palpation. The patient's conjunctivae were injected and her oral mucosa was dry with poor dentition. Examination of her heart, lungs, and abdomen was unremarkable. The timeline of clinical progression is illustrated in Figure 1D.
Laboratory studies revealed a white blood cell count of 12,000/∝l, normocytic anemia with a hematocrit of 34%, and a platelet count of 820,000/μl. Liver enzymes, electrolytes, creatinine, urinalysis, creatine phosphokinase, C3, C4, and thyroid function tests were all within normal limits. Radiologic examination of the hands revealed well-circumscribed marginal erosions at the DIP joints (Figure 1C).
Histologic and immunohistochemical methods and findings.
A skin biopsy sample of a cutaneous nodule from a dorsal finger showed a dermal nodular infiltrate composed predominantly of multinucleated histiocytes with “glassy” cytoplasm. These cells stained faintly and focally with CD68 (Figure 2A) (mouse monoclonal antibody, DakoCytomation, Carpinteria, CA) as a macrophage marker, whereas S100 (rabbit polyclonal, Ventana Medical Systems, Tucson, AZ) and CD1a (Leu-6; mouse monoclonal, Immunotech, Miami, FL), a calcium-binding protein and neurotrophic factor and a surface antigen of Langerhans' cells, respectively, were negative, confirming the diagnosis of MRH. These cells also stained strongly with the osteoclast markers TRAP (Sigma, St. Louis, MO) and cathepsin K (mouse monoclonal, Vision Biosystems, Norwell, MA), a cysteine protease with high matrix-degrading activity (Figures 2C and 2E). Comparative staining of a rheumatoid nodule showed only trace cathepsin K staining in the region of fibroblasts surrounding the nodule, and no significant TRAP staining (Figures 2D and 2F), while no staining for cathepsin K or TRAP was detected in normal control skin (data not shown).
Followup and treatment.
Computed tomography scans of the chest and abdomen, positron emission tomography scan, transvaginal ultrasound, mammogram, and gynecologic examinations did not reveal an associated malignancy. Treatment with intravenous infusions of 4 mg zoledronic acid resulted in improvement of arthritis and concurrent reduction in the size and number of cutaneous MRH nodules (Figure 3). However, this response was transient, with worsening of both arthritis and skin nodules 8 weeks after infusion. Repeat infusions resulted in less dramatic, transient improvement of joint and skin symptoms. However, over a period of 8 months with zoledronic acid infusions, hand radiographs did not show any progression of bony erosions.
MRH is a rare autoimmune disease most frequent in women in their fourth decade of life (60–75%) (4). It has been associated with a wide variety of malignancies in 15–28% of cases (1), but does not run a course parallel to the neoplasm (4). MRH is also associated with autoimmune diseases (15% of reported cases), including hypothyroidism, Sjögren's syndrome, diabetes mellitus, primary biliary cirrhosis, systemic sclerosis, idiopathic inflammatory myopathy, and systemic lupus erythematosus (1, 8). It is unclear whether MRH is also associated with rheumatoid arthritis (8), as it is often misdiagnosed as such due to similarities in arthritis presentation and association with cutaneous nodules. In the present case, the presence of both rheumatoid factor and the highly specific anti–cyclic citrullinated peptide antibody makes a strong argument for co-occurrence of the 2 diseases.
Internal organ involvement has been described, with 20% of cases having pulmonary lesions. Other described organ involvement includes thyroid, salivary glands, heart, kidney, liver, and gastrointestinal tract (4). MRH can present with arthritis only, skin nodules only, or both, and commonly involves the hands, shoulder, knees, and wrists, but can affect any joint (4). Similar to rheumatoid arthritis, MRH is an erosive arthritis, with radiographs commonly showing well-circumscribed periarticular “punched out” erosions and resorption of the juxtaarticular zone but without periarticular osteopenia typical in rheumatoid arthritis or heterotrophic new bone typical in the spondylarthritides and gouty arthritis (1). MRH is a highly destructive disease with up to 45% of cases progressing to arthritis mutilans before the disease remits, usually after 5–10 years (1).
The cause of the severe, erosive nature of MRH has not been well established. Prior reports indicate that cells of a monocyte/macrophage lineage infiltrate both skin lesions and synovium (1).
Similarly, in rheumatoid arthritis, mononuclear cells are found in areas of pannus erosion into bone marrow. In MRH, these cells exhibit properties of osteoclasts such as expression of TRAP and cathepsin K (9). Our case provides evidence of the osteoclastic nature of multinucleated cells infiltrating the skin nodules in MRH, as they strongly express the osteoclast, tissue lytic markers TRAP and cathepsin K.
Speculation that decreasing osteoclast formation and activity by blocking stimulatory signals, such as RANKL, might attenuate the progression of destructive arthritis (9), and findings that RANKL levels can be reduced by bisphosphonate therapy (10) have led to trials of bisphosphonates to prevent focal bone resorption in inflammatory arthritis. These studies demonstrated that bisphosphonates can decrease bone erosion in both animal models of inflammatory arthritis (11) and in humans with rheumatoid arthritis (12). In MRH, synovial fluid macrophages have been shown to differentiate into osteoclasts after stimulation with a combination of macrophage colony-stimulating factor (M-CSF) and RANKL (5), whereas such macrophages obtained after treatment with pamidronate were less likely to be induced to differentiate into osteoclasts by M-CSF/RANKL treatment and had lower bone-eroding activity after stimulation (5). We also confirmed reports of skin and joint disease improvement after treatment with an intravenous bisphosphonate (5–7). Although the palliation of disease observed in our patient is modest, we are encouraged by the lack of progressive joint erosion on hand radiographs. We speculate that the efficacy of zoledronic acid is mediated through its effect on osteoclast-like cells in the skin and synovium. In addition to effects on RANKL, an aminobisphosphonate may promote osteoclast apoptosis directly by inhibiting farnesyl pyrophosphate synthase in the mevalonate pathway and impairing isoprenylation of proteins (13).
Our observations are consistent with the report by Goto et al (6), who also demonstrated the TRAP osteoclast marker in skin and synovial tissue from a patient with MRH. Our article is the first to report that the skin-infiltrating cells also highly express cathepsin K, a strong cysteine proteinase that mediates bone resorption. The cathepsin K staining of the cutaneous MRH nodule was very different from the staining seen in a rheumatoid nodule (Figure 3), where palisading histiocytic cells surrounding the fibrinoid necrosis were cathepsin K and TRAP negative. Only surrounding fibroblasts, similar to those seen in cutaneous scars (14), were cathepsin K positive (but TRAP negative).
Cathepsin K has been shown to be present in low concentrations in normal synovium, but up-regulated in synovial fibroblasts of patients with rheumatoid arthritis, at sites of bone invasion and destruction (15). Furthermore, a transgenic mouse model has demonstrated that overexpression of the cathepsin K gene increases susceptibility to both synovitis and destruction of articular cartilage (16).
While TRAP and cathepsin K expression is highly suggestive of osteoclast differentiation, it is not completely specific (9). TRAP is also detected in macrophages that lack bone resorptive properties (9). While cathepsin K has long been considered to be expressed only in osteoclasts, it has more recently also been shown to be expressed in lung fibroblasts, synovial fibroblasts, and skin fibroblasts during wound healing, presumably to counteract the profibrotic processes through degradation of the extracellular matrix (14, 15).
However, the combined expression of both tissue lytic markers by the multinucleated cells found in the highly erosive MRH strongly suggests osteoclastic activity. Cathepsin K inhibition is currently under investigation for the treatment of osteoporosis and might be an alternative treatment approach for MRH.
Dr. Kissin had full access to all of the data in the study and takes responsibility for the integrity of the data.
Study design. Codriansky, Rünger, Kissin.
Acquisition of data. Codriansky, Rünger, Bhawan, Kantarci, Kissin.
Analysis and interpretation of data. Codriansky, Rünger, Bhawan, Kantarci, Kissin.
Manuscript preparation. Codriansky, Rünger, Bhawan, Kissin.
- 8Multiple cutaneous reticulohistiocytomas in a patient with rheumatoid arthritis. Dermatol Online J 2004; 10: 11., , , .