Current issues of ACP Journal Club are published in Annals of Internal Medicine


Minocycline was effective for mild-to-moderate rheumatoid arthritis

ACP J Club. 1995 May-June;122:69. doi:10.7326/ACPJC-1995-122-3-069

Source Citation

Tilley BC, Alarcón GS, Heyse SP, et al. for the MIRA Trial Group. Minocycline in rheumatoid arthritis. A 48-week, double-blind, placebo-controlled trial. Ann Intern Med. 1995 Jan 15;122:81-9.



To study the safety and efficacy of minocycline in adults with active, mild-to-moderate rheumatoid arthritis (RA).


Randomized, placebo-controlled 48-week trial.


6 clinical centers.


219 patients (mean age, 54 y; 78% women; 66% white) with active, mild-to-moderate RA according to the 1987 American College of Rheumatology criteria. Patients had to have ≥ 9 tender joints and ≥ 6 swollen joints, had to be older than 18 years of age, and could not be taking disease-modifying anti-rheumatic drugs (DMARDs). Patients receiving nonsteroidal anti-inflammatory drugs (NSAIDs) or low-dose prednisone continued their medication. Exclusion criteria were previous use of ≥ 1 DMARDs; hypersensitivity to, or need for, tetracyclines; possible pregnancy; antacids containing aluminum, calcium, or magnesium; anticoagulants; or other serious disease. Follow-up was 94%.


109 patients were assigned to minocycline, 100 mg twice daily on an empty stomach for 48 weeks; 110 patients were assigned to placebo. Doses could be adjusted if side effects occurred.

Main Outcome Measures

≥ 50% improvement in joint swelling and tenderness and side effects.

Main Results

At 48 weeks, 79% of patients in the minocycline group and 78% of patients in the placebo group were receiving their assigned medication. In the minocycline group compared with the placebo group, 15% more patients had ≥ 50% improvement in joint swelling (54% vs. 39%; {95% CI for the 15% absolute difference, 2% to 28%; relative risk improvement [RRI], 38%; CI, 4% to 86%; number needed to treat [NNT], 7; CI, 4 to 55}*; P = 0.023), more patients had improvements in joint tenderness (56% vs. 41%; {CI for the 15% difference, 2% to 28%; RRI, 37%; CI 4% to 82%; NNT, 7; CI, 4 to 55}*; P = 0.021), and more patients had improvements in laboratory values (sedimentation rate, hematocrit, platelet count, and IgM rheumatoid factor; P for all comparisons < 0.001). The groups did not differ for grip strength, physician or patient overall assessments, morning stiffness, Modified Health Assessment Questionnaire scores, and frequency of side effects. No serious toxicity occurred.


Minocycline was safe and effective in patients with active, mild-to-moderate rheumatoid arthritis.

Sources of funding: National Institutes of Health and, in part, Lederle Laboratories (medication and placebo).

For article reprint: Dr. S.P. Heyse, National Institutes of Health, Natcher Building, Room 5AS-53, 45 Center Drive, MSC 6500, Bethesda, MD 20892-6500. FAX 301-402-2406.

*Numbers calculated from data in article.


Minocycline and other tetracyclines possess antirheumatic properties (1), and several small studies assessing the effect of minocycline in RA have been encouraging (2-4). In the well-executed study of Tilley and colleagues, small differences were found in major outcome variables, including joint swelling and joint tenderness. Patients assigned to receive minocycline were only 1.38 times more likely to have fewer active joints after 48 weeks of treatment than those assigned to placebo. Even more important from the patient's perspective, no difference existed between the treatment groups for morning stiffness, grip strength, and activities of daily living, and no patients went into remission. The authors attributed the lack of important side effects with minocycline to the design of the study, which allowed for dose modifications. No mention was made, however, of the study group's mean minocycline dose. This may have resulted in a lower daily dose than reported (200 mg/d), which may partly explain the less than remarkable effectiveness of minocycline. Unlike methotrexate, which usually begins to relieve the symptoms of RA within 3 weeks, minocycline acted very slowly, showing its first clinical effects only after 3 months. Lacking substantial short-term benefits, the future application of minocycline may be as a DMARD. Until its properties are confirmed, however, minocycline will probably be of limited use in the treatment of RA.

Roland Staud, MD
University of Florida Gainesville, Florida