Phenformin (Dibotin) in polyarthritis.

Since 1956 biguanides have received attention as oral hypoglycaemic agents (Krall and Bradley, 1959). Ungar, Madison, and Carter (1960), using phenformin (Dibotin, phenethyl biguanide), showed their effect on peripheral utilization of glucose, Butterfield, Fry, and Holling (1958) having previously demonstrated increased uptake of insulin in hepatic and peripheral tissues. Since then the drug has been used with sulphonylureas for oral control of diabetes and with insulin in brittle diabetics and when they are partly insulin resistant. Fearnley and Chakrabarti (1964) showed that phenformin increased fibrinolytic activity in occlu-sive heart disease and Fearnley, Chakrabarti, and Hocking (1965) used this drug in rheumatoid arthritis in hopes of removing fibrin from the inflammatory lesions. In a follow-up of twenty cases treated with phenformin 100 mg. and ethyloestrenol 8-16 mg. for 5 to 14 months, (Fearnley, Chakrabarti, and Evans, 1966), they found some clinical improvement in twelve. In seventeen patients there was a 20 to 63 per cent reduction of plasma fibrinogen and in twelve the erythrocyte sedimentation rate (ESR) decreased by over 30 per cent. Material and Methods In view of these results it was decided to carry out a double-blind cross-over trial; 26 patients with rheumatoid arthritis were selected and allocated randomly to start treatment either with 100 mg. phenformin tablets or with placebo tablets of identical appearance. All were patients with typical, active, sero-positive, rheumatoid arthritis and throughout the trial no change was made in treatment except in the number of analgesic tablets. The number taken in each period was recorded but there was little variation and this did not influence the trial. The patients' conditions were assessed before treatment and after 6 weeks on each medication, the assessment including the patients' estimate of pain and stiffness (graded 0-3), strength of grip, ring-size measurement of the worst affected finger, ESR, haemoglobin, and weight. In four patients, who also had diabetes, blood sugar estimations were made three hours after the same light breakfast. Results Twenty patients completed the trial with little or no toxic symptoms, while six discontinued treatment. Two of the six were withdrawn because pre-existing mental instability made records unreliable and one with Felty's syndrome for splenectomy; the remaining three refused further phenformin treatment because of dyspepsia (1), diarrhoea (1), and general malaise (1). In two patients who completed the course, the dose of phenformin was reduced from 100 to 50 mg. per day, because of indigestion and diarrhoea, and …