Abstract

BACKGROUND--Open lung biopsy (OLB) is rarely necessary for investigation of HIV positive patients with acute respiratory episodes because of the high yield from fibreoptic bronchoscopy with bronchoalveolar lavage (BAL). METHODS--A retrospective review of OLB in HIV positive patients admitted to a specialist inpatient unit with acute respiratory symptoms was carried out in order to define clinical indications, diagnostic yield, impact on management, complications and outcome. RESULTS--OLB was performed in 23 patients; 21 had undergone one or more bronchoscopies with BAL (5 also had negative results from transbronchial biopsy). Indications for OLB were: Group A, 15 patients thought clinically to have pneumocystis pneumonia but not responding to treatment; Group B, 4 patients with focal chest radiographic abnormalities; Group C, 4 patients with diffuse radiographic abnormalities and miscellaneous conditions. Preoperative PaO2 (on air) ranged from 4.4 to 14.5 (mean = 9.5) kPa. The results of OLB were in Group A 5 patients had non specific interstitial pneumonitis (NIP), 1 also had Kaposi's sarcoma, 4 had pneumocystis pneumonia (1 also had bronchiolitis obliterans organising pneumonia [BOOP]), 3 had Kaposi's sarcoma and 1 had BOOP and emphysema, 1 had pulmonary infarction and no infection and 1 had normal lung tissue. In Group B diagnoses were NIP, B cell lymphoma, occult alveolar haemorrhage and Pseudomonas aeruginosa pneumonia with BOOP; In Group C 2 patients had NIP and 2 had pneumocystis pneumonia (1 also had cytomegalovirus pneumonitis). All patients survived surgery and none required mechanical ventilation. OLB results significantly affected management; in Group A inappropriate treatment was discontinued in 11 patients found not to have pneumocystis pneumonia, and alternative therapy was begun in the 4 with pneumocystis and in Groups B and C 6 patients began specific therapy; unnecessary therapy was avoided in one and antimicrobial treatment was modified in one. CONCLUSIONS--Open lung biopsy in HIV positive patients with focal and diffuse radiographic abnormalities has a high diagnostic yield and low morbidity. This investigation should be considered in those with acute respiratory episodes and negative results from bronchoscopic investigations or who have contra-indications to this procedure.