An attempt to predict the response to gold therapy in rheumatoid arthritis

• Published in: Arthritis and rheumatism 1984-11, Vol.27 (11), p.1210-1217
• Main Authors: O'Duffy, J. D., O'Fallon, W. M., Hunder, G. G., Mcduffie, F. C., Moore, S. B.
• Format: Article
• Language: English
• Subjects: Arthritis, Rheumatoid - drug therapy; Biological and medical sciences; Bones, joints and connective tissue. Antiinflammatory agents; Female; Gold - therapeutic use; Gold - toxicity; HLA Antigens - analysis; Humans; Immunoglobulins; Male; Medical sciences; Middle Aged; Pharmacology. Drug treatments; Time Factors
• ISSN: 0004-3591; 1529-0131
• DOI: 10.1002/art.1780271102

In 54 rheumatoid arthritis patients undergoing conventional chrysotherapy, we prospectively sought predictors of response, using strict clinical and laboratory criteria of improvement. Forty‐five patients who completed 6–12 months of therapy were classified into 3 outcome categories: group 0, not significantly improved (18 patients); group 1, improved (18 patients); and group 2, markedly improved (9 patients). Sixty‐two entry variables were tested in univariate and multivariate analysis for predictor function. No continuous variable was predictive. The discrete variables HLA‐A3 positivity and HLA‐DR4 negativity were the best predictors of response to gold. In a multivariate analysis using these 2 univariates (A3 and DR4) plus hemoglobin, we developed a discriminant function that correctly predicted outcome in 21 of 23 patients in groups 0 and 2. We also observed that of 15 DR blank patients (10 of whom were DR4 blank), none entered remission.