Shortening the duration of treatment for cervical tuberculous lymphadenitis.

The aim of the study was to determine the optimal duration of treatment for patients with tuberculous lymphadenitis. The Medline database was searched for relevant articles published between 1978-1997. Inclusion criteria were study populations of patients with predominantly cervical tuberculous lymphadenitis in whom the diagnosis had been confirmed bacteriologically and/or histologically, or was made probable by using clinical and laboratory markers. Treatment management had to include at least isoniazid, rifampicin and pyrazinamide and a follow-up of at least 12 months after the end of treatment. Patients with resistance to rifampicin and pyrazinamide and previous treatment for tuberculosis were excluded. The number of patients who relapsed after treatment was calculated. The study population in eight out of the 35 articles retrieved were suitable for analysis. Some concerned comparative studies. There were eight treatment schedules of 6 months' duration and three schedules of 9 months' duration. Treatment for 6 months resulted in a tuberculous lymphadenitis relapse rate of 13/422=3.3% (95% confidence interval: 1.7-5.5), with a mean follow-up of 31 months after completion of treatment. Treatment for 9 months resulted in a relapse rate of 3/112=2.7% (95% confidence interval: 0.6-7.8), with a mean follow-up of 20 months. Despite the limitations of the literature available, 6 months of therapy is probably sufficient for patients with tuberculous lymphadenitis.