Double blind, randomised controlled trials have shown that minocycline is an effective disease modifying antirheumatic drug (DMARD) in rheumatoid arthritis (RA), compared with placebo^1–,⁴ or hydroxychloroquine.⁵ Minocycline was first used on the premise that RA may be caused by an infection but, subsequently, it was also shown to possess other properties such as matrix metalloproteinase inhibition and immunomodulation. Despite reported proof of its efficacy, most rheumatologists do not favour the use of minocycline in RA, possibly owing to availability of other “standard” DMARDs.

We performed a retrospective review of the case notes of 28 patients with RA who were prescribed minocycline. Treatment with minocycline in these patients began before the widespread availability of biological agents. Our aim was to assess the efficacy and safety of this drug in our hands compared with published trials. Our patients included 24 women and four men, aged between 43 and 80 years (mean 60). Their disease duration ranged from 2 to 48 years (mean 18). Rheumatoid factor status was known in 26 patients, of whom 21 were seropositive. Minocycline was used only after at least two to eight DMARDs (mean five drugs) had failed. None of these patients were receiving concomitant treatment with other DMARDs at the time of starting minocycline. We used minocycline in a dose of 100 mg twice daily.

As this was a retrospective review of case notes, improvement in disease activity could only be assessed from the information in clinic letters. Clinical improvement was assessed by factors such as improvement in joint pain and swelling, duration of early morning stiffness, function, physician’s global assessments, and general wellbeing of the patient, while improvement in laboratory measures was assessed by change in erythrocyte sedimentation rate (ESR) and haemoglobin.

In the opinion of the rheumatologist the drug was considered effective in 10 (36%) patients, of whom seven were still taking it at the time of performing this study. Three of these 10 patients had to stop taking minocycline because of side effects. Benefit was noted after a mean duration of 4 months (range 2–6) and was sustained for a mean duration of 14 months (range 8–24). Stopping treatment owing to a lack of efficacy occurred in only 7/28 (25%) patients and they had taken the drug for a mean duration of 6 months (range 3–11). No differences in disease duration, number of DMARDs tried before starting minocycline, or rheumatoid factor status were found between responders and non-responders (also including patients who stopped minocycline owing to toxicity, but had received the drug for at least 4 months).

There was documented improvement in clinical measures in all patients who responded. Laboratory data were available for 24 patients, of whom 18 had taken the drug for at least 4 months (eight responders, 10 non-responders). Among the eight responders, ESR values improved by more than 40 mm/1st h in four patients (reduced to 13, 25, 31, and 31 mm/1st h), while haemoglobin improved by more than 20 g/l in two patients. We did not note any deterioration of ESR or haemoglobin values in any of the other responders. However, the ESR and haemoglobin values either remained the same or deteriorated in all non-responders save for one patient.

Thirteen (46%) patients, including the three patients in whom the drug was considered effective, stopped taking the drug because of side effects. There were no serious or long term adverse effects. The side effects that were directly attributable to minocycline included dizziness (four patients), nausea (three patients), dizziness and nausea, allergic rash, and reversible grey pigmentation (one patient each). Three patients stopped the drug owing to problems not directly related to minocycline (atrial fibrillation, allergic skin rash to trimethoprim, and non-specific chest pain). The reason for stopping minocycline was not clear from the notes for one patient.

As far as we know, no one has reported their experience with the use of minocycline in patients with RA outside a research setting. If the fact that minocycline was only tried in our patients after they had failed to respond to other DMARDs is taken into account, it can be considered as a moderately efficacious drug. Studies in future should examine the role of minocycline in early RA either on its own or as part of combination DMARD treatment.