No abstract

Sir, Prospective cohort studies have demonstrated an increased risk of active tuberculosis in patients treated with tumor necrosis factor antagonists (anti-TNF agents) [1]. Current guidelines support chemoprophylaxis with isoniazid (INH) for 6-9 months prior to the initiation of anti-TNF agents in patients with evidence of latent tuberculosis [2]. We present two rheumatoid arthritis patients who had positive tuberculin skin tests (TST). Despite receiving an adequate course of INH for latent tuberculosis infection, both patients developed active tuberculosis after initiation of infliximab. Patient 1 is a 58-yr-old male Chinese immigrant with a 3 yr history of seropositive rheumatoid arthritis who was admitted to the hospital with three weeks of fevers, chills, productive cough, night sweats and altered mental status. His medical history was otherwise notable for chronic inactive hepatitis B and chronic obstructive pulmonary disease. Three yrs prior, he had a positive TST. Chest computed tomography (CT) revealed emphysematous changes and multiple calcified granulomas consistent with prior tuberculosis. He was treated with a 9 month course of INH, 300 mg/day. Despite treatment with methotrexate and sulfasalazine, his rheumatoid arthritis remained active, requiring frequent intra-articular steroid injections and oral corticosteroids. Fourteen months prior to admission he started infliximab infusions (6 mg/kg) in addition to methotrexate and sulfasalzine, and was successfully tapered off corticosteroids. On admission, the patient was ill appearing and thin. He had a temperature of 102.6° with normal respiratory rate and oxygen saturation. The breath sounds were decreased and there were diffuse expiratory wheezes in all the lung fields. There was no lymphadenopathy. Initial investigations included normal complete blood counts, electrolytes, liver and renal function tests, as well as negative tests for HIV, hepatitis C and rapid plasma reagent. Sputum stains for acid-fast bacilli, pneumocystis carinii and legionella species were negative. A chest radiograph and chest CT scan demonstrated no new infiltrates or pathological lymphadenopathy. A CT scan of the abdomen and pelvis revealed enlarged mesenteric and retro-peritoneal nodes. Bone marrow biopsy and right inguinal lymph node biopsy did not show evidence of malignancy or granulomatous disease.

The contribution of interleukins produced by most inflammatory cells to chronic arthritis is not well understood. Therefore, we investigated the influence of several human recombinant interleukins (IL‐1β, IL‐2 and IL‐6) on joint swelling, on the inflammatory process, and on serological parameters in a novel animal model of arthritis, the human/murine SCID arthritis. In this model an arthritis is induced by implanting human synovial tissue from patients with rheumatoid arthritis (RA) into the knee joint of mice with SCID. These mice tolerate the xenogeneic implant and develop a mixed human/murine pannus tissue. The interleukins were injected daily for 7 or 14 days after implantation. IL‐1β led to a significant increase in joint swelling. It intensified the inflammatory process accompanied by enhanced migration of murine inflammatory cells into the knee joint. The production of human IL‐6 in the transplanted tissue was stimulated through the application of IL‐1β, and the serum level of human IL‐6 was thus significantly higher than in controls. We could not observe a significant influence of IL‐1β on the production of human IgG or IgM by the implant. The application of human IL‐2 had a weak effect similar to that of IL‐1β, but without statistical significance. Although IL‐6 is a good marker for inflammation in RA, the application of recombined human IL‐6 had no influence on the inflammatory process in this model.

No abstract

No abstract