THE LANCET 1298 Vol 350 • November 1, 1997 blind placebo-controlled study of the effect of inhibition of nitric oxide synthesis in bradykinin-induced asthma. Lancet 1996; 348: 374–77. 3 Figini M, Ricciardolo FLM, Javdan P, et al. Evidence that epithelium- derived relaxing factor released by bradykinin in the guinea-pig trachea is nitric oxide. Am J Respir Crit Care Med 1996; 153: 918–23. 4 Kharitonov SA, Yates D, Robbins RA, Logan-Sinclair R, Shinebourne EA. Barnes PJ. Increased nitric oxide in exhaled air of asthmatic patients. Lancet 1994; 343: 133–35. 5 Moncada S, Higgs A. The L-Arginine-nitric oxide pathway. N Engl J Med 1993; 329: 2002–12. Institute of Respiratory Disease, University of Catania, 95125 Catania, Italy (G U Di Maria); and Department of Clinical and Experimental Medicine, Unit of Pharmacology, University of Ferrara, Ferrara respectively. 5 Also, HIV-1-negative patients with TB showed during the course of treatment a much higher mortality rate in Bangui (8·2%) than in most reports (Haiti 1%, 3 Abidjan 0·4% 4 ). 1 Pablos-Méndez A, Sterling TR, Frieden TR. The relationship between delayed or incomplete treatment and all-cause mortality in patients with tuberculosis. JAMA 1996; 276: 1259–60. 2 Elliott AM, Halwiindi B, Hayes B, et al. The impact of human immunodeficiency virus on mortality of patients treated for tuberculosis in a cohort study in Zambia. Trans R Soc Trop Med Hyg 1995; 89: 78–82. 3 Chaisson RE, Clermont HC, Holt EA, et al. Six-month supervised intermittent tuberculosis therapy in Haitian patients with and without HIV infection. Am J Respir Crit Care Med 1996; 154: 1034–38. 4 Ackah AN, Coulibaly D, Digbeu H, et al. Response to treatment, mortality, and CD4 lymphocyte counts in HIV-infected persons with tuberculosis in Abidjan, Côte d’Ivoire. Lancet 1995; 345: 607–10. 5 World Health Organization Global Tuberculosis Programme. Global tuberculosis control, WHO report 1997, WHO/TB/97.225. Institut Pasteur de Bangui, BP923-Bangui, Cental African Republic (B Garin); and Hôpital de l’Amitié, Bangui High mortality rates among patients with tuberculosis in Bangui, Central African Republic B Garin, P Glaziou, E Kassa-Kelembho, S Yassibanda, P Mbelesso, J Morvan A National Tuberculosis Control Programme (NTP) was implemented in the Central African Republic (CAR) in 1995. We report here survival time to death of a prospective cohort of 224 randomly selected adult patients with cultures positive for tuberculosis (TB), recruited in Bangui between 1993 and 1994 among 1492 notified TB cases, of whom 1029 were culture positive. All bacteriological examinations were done at the Bangui Institut Pasteur. TB strains were tested for drug susceptibility. All patients were admitted to hospital at onset of treatment, and underwent HIV-1 testing and CD4-cell count, after giving informed consent. Treatment regimen was isoniazid, rifampicin, ethambutol, and pyrazinamide for 2 months, followed by isoniazid and ethambutol for another 6 months. Directly observed short- course treatment strategy (DOTS) has not been adopted in CAR. Sex ratio (M/F) was 1 and median age 30 years (range 17–70). Among the 224 patients, six presented with extrapulmonary TB (2·7%), 186 were smear positive (83%), 139 were HIV-1 seropositive (62%), 205 were new cases of TB (91·5%), and 23 had multidrug-resistant (MDR) TB strains (10·3%), defined as resistance to at least rifampicin and isoniazid, among whom 14 were new cases. Treatment outcomes in HIV-1-positive and negative patients at the end of 8 months were: 46·4% vs 67·1% treatment success, 39·1% vs 8·2% deaths, 3·6% vs 4·7% treatment failures, and 10·9% vs 20% defaulters. 24 months after start of treatment, the cumulative death rate was 20% in HIV-1-seronegative patients and 58% in HIV-1-seropositive patients (figure) whose median life expectancy was 15 months. A Cox regression model showed that both age and HIV-1 seropositivity were risk factors for death, whereas MDR TB was not associated with increased mortality (p=0·35). In addition, low CD4 cell count was found to be associated with a higher mortality in HIV-1 seropositive patients (p=0·007). Among the 119 patients who survived the recommended duration of treatment (8 months), those who completed treatment had a lower subsequent mortality rate (10·2% vs 47·6%, RR=0·58, 95% CI: 0·39–0·88). We found similar predictors of decreased survival as described elsewhere, namely age, HIV-1 serostatus, 1,2 treatment completion and CD4 cell count in HIV-1 seropositive patients, 3,4 but not MDR TB. 1 This is probably due to the small numbers and the high overall mortality rate, which is the highest in Africa. Non-DOTS African countries reported 9% mean mortality rate to WHO for the year 1994. 5 Congo and Zaire reported 5% and 13·1% mortality rates, 1·0 0·9 0·8 0·7 0·6 0·5 0·4 0·3 0·2 0·1 0 Cumultative proportion surviving 0 2 4 6 8 10 12 14 16 18 20 22 24 Months Log rank test, p=0·001 HIV-1 negative HIV-1 positive = Died = Censored Cumulative survival probabilities of patients with TB according to HIV-1 serostatus in Bangui Autoimmune hepatitis type 2 induced by HCV and persisting after viral clearance Sandro Vento, Francesca Cainelli, Carlo Renzini, Ercole Concia HCV infection may be associated with type 1 autoantibodies against liver-kidney microsomes (LKM1). 1 These autoantibodies (the hallmark of type 2 autoimmune hepatitis) 2 recognise cytochrome P450IID6. 3 It is unclear, however, whether there is a connection (accidental or causal) between HCV and autoimmune hepatitis type 2 and no cases of autoimmune liver disease have been so far described after HCV infection. We have been looking after a 29-year-old nurse (HLA A*0101, B*0801, DQB1*0201, DQA1*0501, DRB1*0301, DRB3*0101) who had an accidental needlestick injury while caring for an HCV-infected patient in December, 1989. At that time her aminotransferase and  globulin concentrations were normal and she had negative anti-HCV, anti-nuclear (ANA), anti-smooth muscle (ASMA), and anti-LKM1 antibodies, and HCV-RNA (RT-PCR). In March, 1990, anti-HCV antibodies and serum HCV-RNA had become positive, aspartate (AST 362 IU/L) and alanine (ALT 548