The disease duration in our patients was between 2 and 26 years (mean 8.1 years), and all of them showed a rapid progression before starting MTX. In this particular group of patients, the severity of the disease was independent of the duration. This does not necessarily apply to an average population of RA patients.While our study does not demonstrate a superiority of MTX over parenteral gold, it does show that clinically effective treatment with MTX may slow down joint destruction in RA patients whose disease was previously not sufficiently responsive to parenteral gold treatment. A double-blind prospective study comparing intramuscular doses of MTX (15 mg/week) and gold sodium thiomalate (50 mg/week) in early RA did not show significant differences between the drugs after I year (4). Giant cell arteritis in Mediterranean countries: comment on the article by Salvarani et al To the Editor:We read with interest the article by Salvarani et al concerning giant cell (temporal) arteritis (GCA) in Italy (1). We recently investigated the incidence of biopsy-proven GCA from 1981 to 1990, in an area of northwestern Spain (2). The average annual incidence rate of biopsy-proven GCA was 6/100,000 population 50 or older, but if we had considered all patients diagnosed with GCA, i.e., those with a positive biopsy result plus those diagnosed by clinical criteria alone (3), the annual incidence would be 6.6/100,000 age 50 or older, which is very similar to the rate in Italy reported by Salvarani and colleagues. The similarity in these findings may be related to genetic factors.We raise two issues for consideration by Salvarani et al: First, they had previously published an article in which the annual incidence of GCA was reported to be 8.8/100,000 (4). and in their later article (I), the reported incidence was lower (6.9/100,000). In contrast, we observed a progressive increase of the annual incidence, which is similar to findings in other studies (5).The second issue is that, in the study by Salvarani and colleagues, only 46.5% of the patients had biopsyproven disease. Ninety percent of our patients had a positive biopsy result, which is slightly lower than the percentage found in a study in Olmsted County, Minnesota ( 5 ) . We believe these differences may be explained by the fact that Salvarani et al obtained artery specimens that were of insufficient size; we usually obtain biopsy specimens at least 3 cm in length.It has been proposed that, in patients with a negative biopsy result, determination of factor VIII-von Willebrand factor antigen might be of some help in supporting the diagnosis of GCA (6). although others have considered this to be of limited value (7). However, despite the segmented nature of the lesions in the arterial wall of patients with GCA, we are in favor of pathologic confirmation, since these patients, in the majority of cases, will require treatment with corticosteroids for long periods of time.Finally, similar to the findings reported by others (8,9), we have recently observed that treatment with lo...