Causative agents of drug eruptions are frequently unknown, and skin tests with candidate drugs would be useful before systemic challenge. It remains to be clarified how phostosentive lichenoid drug eruptions are induced, but allergy, including delayed type allergy, has been suggested. Two patients who had taken anti-tuberculous drugs developed a lichenoid drug eruption, primarily on sun-exposed skin. Patch and photopatch tests were performed with each of the ingested drugs (10% in petrolatum). Photopatch tests to isoniazid (INH) were positive. These were confirmed by oral challenge followed by irradiation with UVA. In conclusion, photopatch tests facilitated identification of the causative drug in two patients with photosensitive lichenoid eruptions to INH.

Leukocytoclastic vasculitis is characterized by a palpable purpura showing microscopically necrotizing vasculitis of the small dermal blood vessels. The cause has been considered to be a deposition of immune complexes under conditions of slight antigen-excess. 1 Infections are proposed to be a causative factor and leukocytoclastic vasculitis due to Mycobacterium tuberculosis has been reported previously 2-5 but is uncommon. We describe two cases of the condition associated with tuberculosis. The association with M. tuberculosis was identified by the polymerase chain reaction (PCR) and response to antituberculous therapy. Case report Case 1A 15-year-old Korean girl presented with a one-month history of palpable purpura of the lower extremities, along with oedema and painful knee joints on walking. Histological examination of a lesion showed leukocytoclastic vasculitis. Laboratory examination including platelet count was normal as were the serum levels of immunoglobulin and complement, except for an increase in IgA and a decrease in C3. Active pulmonary tuberculosis was detected on chest X-ray and by the PCR of sputum for M. tuberculosis. A biopsy specimen also prepared for PCR was unfortunately lost, but the PCR of peripheral blood mononuclear cells for M. tuberculosis was positive. Peripheral blood mononuclear cells had been prepared with a Ficoll-paque (Pharmacia Biotech). DNA was extracted with a QIAamp Blood Kit and Tissue Kit (Qiagen) and the IS6110 insertion sequence specific for M. tuberculosis (but which also recognizes M. bovis) was amplified by PCR. One to 10 pg of DNA extracted from Bacille-Calmette-Guérin (BCG) material was used as a positive control and distilled water as a negative control. The patient was treated with antituberculous regimens for 6 months without recurrence of the lesions at 18-month follow-up. Case 2A 13-year-old Korean girl had extensive palpable purpura of the upper and lower extremities and buttocks with leg oedema and associated abdominal pain for 2 days. The histopathology of a lesion showed leukocytoclastic vasculitis. Laboratory examination also showed blood neutrophilia, haematuria and occult blood in the stool. Oesophagogastroscopy revealed severe haemorrhagic gastritis without infection with Helicobacter pylori. There was no evidence of pulmonary tuberculosis on chest X-ray or repeated sputum examination, and skin testing with purified protein derivative (PPD) was negative. PCR of peripheral blood mononuclear cells for M. tuberculosis was positive. The patient was therefore treated with an antituberculous regimen for approximately 7 months without recurrence over the next 5 months. Repeated skin testing with PPD was positive and further PCR of the blood for M. tuberculosis negative at the end of treatment. DiscussionPCR has been useful for the identification of mycobacteria, Borrelia burgdorferi, Listeria monocytogenes, and Chlamydia. 6 Such amplification of the IS6110 insertion fragments specific for M. tuberculosis in skin biopsies [7][8][9] and peripheral blood mononuc...