Unusual complication of B.C.G. vaccination.

A man aged 26 had three Mantoux tests in France in January, March, and June 1966; they were negative and he was therefore advised to have a B.C.G. vaccination six months later. On 6 January 1967 he attended the casualty department of another hospital with an ampoule of B.C.G. vaccine (designed for percutaneous multiple puncture) from the Institut Pasteur. This is more than 10 times stronger than the English intradermal vaccine, and the instructions recommend that it should be given by scarification of three sites of the skin of the deltoid region: 2 to 4 drops on each site, separated by 1.5 cm, with a maximum total length of scratch of 20 cm. The casualty officer misinterpreted the French instructions and injected 0-4 ml intradermally into three separate areas over the right deltoid, thus giving a dose 40 times greater than that contained in the usual 0-1 ml of English vaccine. Over the next few weeks the patient's arm became inflamed, swollen, and painful, with pronounced axillary and supraclavicular lymphadenopathy. On 10 February he developed erythema nodosum, and four days later was admitted to the Royal Free Hospital. On examination he was pyrexial (38.5 C). There were three pustules, each 1 cm in diameter, in the right deltoid region surrounded by an area of erythema and induration, up to a total area of 12 by 12 cm. Right axillary and supraclavicular lymphadenopathy was present with glands up to 4 cm in diameter. There was erythema nodosum of both legs. No other abnormalities were found. Investigations.-Chest x-ray examination showed nothing abnormal. Mantoux reaction was 1:100.000 positive; a swab from the lesions yielded a growth of Staphylococcus epidermidis only; no acid-fast bacilli were seen and no mycobacteria were grown in culture. The E.S.R. was 35 mm in one hour (Westergren). Haemoglobin, white cell count, platelet count, blood urea, and urine were all normal. After a course of ampicillin some of the swelling and induration improved, possibly due to regression of secondary infection. He was started on isoniazid, 100 mg three times a day and during the next few weeks he gradually improved. In May 1967, however, he noticed a swelling in the right infraclavicular region, which was slightly tender. Apart from some lack of energy, he was otherwise well. On examination he was apyrexial; a soft fluctuant swelling 4 cm in diameter was found just below the right clavicle; a lymph node 1 cm in diameter was palpable in the right supraclavicular region; and there was induration of the right axilla without definite lymphadenopathy. Three small depressed scars were seen at the site of the original vaccination. Haemoglobin, white cell count, E.S.R., blood urea, and electrolytes were all nornal. Pus aspirated from the cold abscess showed pus cells and a few Gram-positive cocci, but there was no growth on culture. No acid-fast bacilli were seen, no mycobacteria were grown on culture, and guinea-pig inoculation was negative. He was started on full doses of strep-