In recent years, endothelin receptor antagonism has emerged as an important therapeutic strategy for pulmonary arterial hypertension (PAH). Bosentan, an orally active endothelin receptor antagonist, has proved to be effective in primary PAH or PAH related to connective tissue diseases.^1–3

Of 18 patients with PAH related to systemic sclerosis (SSc) treated with bosentan at our division of rheumatology during the past 2 years, two were also affected by rheumatoid-like arthritis before starting bosentan. We here report the clinical course of these two cases.

The first patient, a 55 year old woman, with SSc since 1988, developed a non-erosive polyarthritis affecting the hands, knees, and ankles, with positive rheumatoid factor (RF; 585 U/l) in March 2002. The second patient, a 57 year old woman, with SSc since 2001, showed a symmetric non-erosive arthritis of hands and wrists, with high RF (414 U/l) in January 2004. Both patients were treated with methotrexate (10 mg weekly) and low dose methylprednisolone (8 mg daily), with favourable effects. In August and October 2004, respectively, both patients observed the onset of effort dyspnoea. High resolution computed tomography of the lung did not show signs of pulmonary fibrosis, while a Doppler echocardiogram estimated a pulmonary artery mean pressure of 45 and 40 mm Hg, respectively.

Bosentan treatment for PAH was given orally at a dose of 62.5 mg twice daily for 1 month and then 125 mg twice daily, without modifying methotrexate and methylprednisolone dosages. One month later, the two patients had an evident flare of arthritis. An increase in erythrocyte sedimentation rate (ESR) and C reactive protein (CRP) as well as RF levels was seen. Owing to the persistence of the arthritis flare, after a further month of treatment, we decided to withdraw bosentan in both patients. Four weeks later, a remission of the articular inflammation and a significant reduction of ESR, CPR, and RF was seen (table 1).

Bosentan, an endothelin receptor antagonist, is a useful and generally well tolerated agent for the treatment of PAH. Of 18 patients with SSc given this treatment at our division of rheumatology, four dropped out owing to side effects: one for a significant increase in liver enzymes and one for severe generalised tremors during the first month of treatment; the other two are the two cases described in this report. In these patients, both affected with rheumatoid-like arthritis associated with SSc and PAH, we observed a marked flare of articular inflammation during treatment and a rapid, subjective, and objective improvement of arthritis after withdrawing bosentan.

A possible hypothesis explaining the arthritis flare is that bosentan might inhibit cortisol secretion. Good evidence exists that normal human adrenal cortex cells are significantly stimulated by both endothelin A and B receptors.⁴

We may also suppose that the flare is related to pharmacological interactions. We know that methotrexate in serum is approximately 50% protein bound and may be displaced from albumin by various compounds, including sulphonamides. Bosentan, a benzene sulphonamide, might interact with the protein binding of methotrexate, increasing its renal excretion, and reducing its hepatic metabolisation to the active polyglutamated forms.⁵ For these reasons, the flare of arthritis in our two patients might be related to the reduced bioavailability of methotrexate. Clinical worsening may directly explain the increase of either inflammatory indices or RF, but it is also known that methotrexate treatment itself may reduce RF levels.⁶

Our observations suggest that special attention should be paid to patients affected with arthritis and treated with bosentan. Moreover, in these cases, the concomitant administration of methotrexate and bosentan should be avoided.