To examine the relationship between radiographic features, serum beta-2-microglobulin (beta-2-M) levels, results of sputum-smear microscopy and outcome, we performed a retrospective study of 99 HIV-seropositive and 162 HIV-seronegative patients with pulmonary tuberculosis (TB) in Dar es Salaam, Tanzania. Radiographic features of primary TB were more common and features of postprimary TB less common in HIV-seropositive compared to seronegative patients (50% vs 31%, p < 0.002; and 40% vs 63%, p < 0.001), respectively). HIV infection had a strong independent effect on the beta-2-M levels. Among HIV-infected patients radiographic findings of primary TB were significantly more often associated with beta-2-M levels of > 4 mg/l than features of postprimary TB (71.1% vs 44.4%, p < 0.02). In patients with features of postprimary TB, acid-fast bacilli were more often detected in sputum smears than in patients with primary TB (65% vs 47%, p > 0.05, in HIV-seropositive patients; and 63% vs 31%, p < 0.001) in seronegative patients). The observed mortality was too low to identify radiographic predictors of survival. We conclude that HIV-infected patients with features of primary pulmonary TB are likely in an advanced stage of HIV infection and deserve close supervision during anti-tuberculous therapy.
PIP: A retrospective study enrolled 99 HIV-seropositive and 162 HIV-seronegative patients with pulmonary tuberculosis (TB) in Dar es Salaam, Tanzania, during July-October 1991 in order to examine the relationship between radiographic features, serum beta-2-microglobulin (beta-2-M) levels, results of sputum-smear microscopy and outcome. 179 (68.6%) were male. Radiographic features of primary TB were more common and features of postprimary TB were less common in HIV-seropositive compared to seronegative patients (50% vs. 31%, p 0.002; and 40% vs. 63%, p 0.001), respectively. HIV infection had a strong independent effect on the beta-2-M levels. Among HIV-infected patients, radiographic findings of primary TB were significantly more often associated with high beta-2-M levels of 4 mg/l than features of postprimary TB (71.1% vs. 44.4%, p 0.02). Serum was available for measurement of beta-2-M concentrations in 83 HIV-infected patients. Patients who presented with primary TB had high beta-2-M levels significantly more often than patients with postprimary TB (27/38 vs. 16/36, p 0.02). 6 of 9 patients with indeterminate TB had beta-2-M levels of 4 mg/l. Type of infiltration also showed a trend to be associated with beta-2-M levels, as 14 of 31 with patchy type, 19 of 29 with reticulonodular type, and 8 of 9 with homogenous consolidation had increased beta-2-M levels of 4 mg/l. In patients with features of postprimary TB, acid-fast bacilli were more often detected in sputum smears than in patients with primary TB (65% vs. 47%, p 0.05, in HIV seropositive patients; and 63% vs. 31%, p 0.001, in seronegative patients). The observed mortality was too low to identify radiographic predictors of survival. In conclusion, HIV-infected patients with features of primary pulmonary TB are likely in an advanced stage of HIV infection and deserve close supervision during anti-tuberculous therapy.