Faculty Bibliography

STUDY OBJECTIVES: Patients with HIV-1 infection or AIDS may present with abnormal chest radiograph (CXR) findings in the absence of symptoms specific to the lung. The objective was to determine the spectrum of disease and the diagnostic modalities employed in these patients. METHODS: From 1996 to 1998, we identified patients with HIV-1 infection presenting to the Bellevue Hospital Chest Service with abnormal CXR findings, and absence of specific pulmonary symptoms. Charts were reviewed for presence of constitutional symptoms, CD4 lymphocyte count, use of Pneumocystis carinii pneumonia (PCP) prophylaxis, eventual diagnosis, and all diagnostic modalities employed. CXR findings were classified according to their predominant abnormalities: nodules, infiltrates, cavity, mass, adenopathy, or effusion. RESULTS: Forty-four patients were eligible for inclusion. Eight-six percent of patients had a CD4 lymphocyte count < 200 cells/microL, and 57% were receiving PCP prophylaxis. Nodular disease was the most common radiographic abnormality (57%), followed by adenopathy (17%). A definitive diagnosis was obtained in 86% of the patients. The most common diagnosis was tuberculosis (26%), followed by nontuberculous mycobacteria (NTM; 23%) and Kaposi sarcoma (12%). No patients had PCP or bacterial pneumonia. Sixty-two percent of patients required an invasive modality to establish a diagnosis. Only 18% of patients with tuberculosis (2 of 11 patients) received diagnoses by sputum analysis. CONCLUSIONS: Patients with HIV-1 infection, abnormal CXR findings, and lack of pulmonary symptoms have a high incidence of infectious disorders, especially pulmonary tuberculosis and infection due to NTM. The high prevalence of treatable and potentially communicable disorders warrants an aggressive diagnostic approach in these patients

The objective of this study was to evaluate the mechanisms of colchicine action in pulmonary fibrosis. The study included 10 patients with pulmonary fibrosis (idiopathic pulmonary fibrosis 5, asbestosis 4, and scleroderma 1) who had been admitted to Bellevue Hospital Center, a tertiary care public hospital in New York City. We administered colchicine 0.6 mg orally for 12 weeks to patients with pulmonary fibrosis. Symptoms, high resolution CT scans, pulmonary function tests, and bronchoalveolar lavage parameters were compared prior to and after treatment. Results showed declines in dyspnea index, selective improvement in several CT scans, but no statistically significant change in BAL cells, cytokines, fibronectin, or hydroxyproline. However, there was a decline in hydroxyproline in the BAL fluid in 8/10 patients. We concluded that colchicine has a mild antifibrotic effect which may be in inhibiting collagen formation since there was no effect on the inflammation that accompanies fibrosis

We conducted a prospective multi-institutional clinical study involving community hospitals and academic medical centers to more carefully define the value of computerized tomography (CT) of the chest with transbronchial needle aspiration (TBNA) in the staging of bronchogenic carcinoma (CA), and to assess the predictors of a positive aspirate. Of 360 individuals determined to have bronchogenic carcinoma, 50 of 81 (62%) with small cell carcinoma (SCC) and 135 of 279 (48%) with non-small cell carcinoma (NSCC) had positive aspirates (p = 0.034). TBNA precluded additional thoracic surgery in a total of 104 of 360 (29%) patients and was exclusively diagnostic of carcinoma in 65 of 360 (18%) cases. Right-sided tumors were more likely to have a positive mediastinal TBNA (p = 0.002 to 0. 01) as were histologic (67 of 118 [57%]) rather than cytology aspirates (228 of 532 [41%]) (p < 0.001). Sensitivity was > 57% in lymph nodes (LN) >/= 10 mm, and among LN of equivalent size, right paratracheal and subcarinal sites were most likely to establish malignancy. Preoperative CT is a valuable adjunct in the staging of CA by TBNA. Increasing LN size, right-sided tumors, right paratracheal and subcarinal locations, use of a histology needle, and the presence of SCC are the best predictors of a positive aspirate

STUDY OBJECTIVE: To determine if screening by specialists could reduce unnecessary test ordering and reduce costs related to diagnostic workup in patients undergoing evaluation for tuberculosis. DESIGN: Prospective evaluation of expert opinion in consecutive patients suspected of having tuberculosis. SETTING: A large municipal hospital. PATIENTS: Patients for whom sputum acid-fast smears were ordered. INTERVENTION: For patients from whom sputum acid-fast bacilli smears and cultures were requested, the chest radiograph and a brief clinical history were presented separately to two pulmonologists with considerable experience in tuberculosis. Each expert reviewed each case independently (and was blinded to the opinion of the other) and indicated if he thought sputum smear examination and culture was, in fact, necessary. Final clinical diagnosis and microbiological information were correlated with the experts' opinion. MEASUREMENTS AND MAIN RESULTS: Ninety-seven patients had sputum smears ordered and had chest radiographs available for review. The two experts believed that sputum examination (smear and culture) was indicated in only 51.5% and 52.6% of cases, respectively. Interobserver agreement was 84.4%. Ultimately, six cases of active tuberculosis were diagnosed. Each expert detected all proven cases of tuberculosis, although one case occurred in a patient with a poor quality radiograph about which the experts offered no opinion. CONCLUSIONS: Screening by experienced clinicians may be effective in reducing unnecessary test ordering and reducing costs related to diagnostic workup in patients evaluated for tuberculosis

Transbronchial needle aspiration (TBNA) of intrathoracic lymph nodes has been shown to be useful in the diagnosis and staging of bronchogenic carcinoma. With the exception of sarcoidosis, the usefulness of TBNA has not been widely investigated in other clinical settings. We investigated the utility of TBNA with a 19-gauge histology needle in HIV-infected patients with mediastinal and hilar adenopathy at Bellevue Hospital Center. We performed 44 procedures in 41 patients. Adequate lymph node sampling was obtained in 35 of 44 (80%), and diagnostic material was obtained in 23 of 44 (52%) procedures. TBNA was the exclusive means of diagnosis in 13 of 41 (32%) patients. Of the 44 procedures, 23 (52%) were performed in patients with mycobacterial disease, with TBNA providing the diagnosis in 20 of 23 (87%). In these patients, positive TBNA specimens included smears of aspirated materials for acid-fast bacilli in 11, mycobacterial culture in 14, and histology in 15. In other diseases, TBNA diagnosed sarcoidosis with noncaseating granulomata in 2 of 4 patients and non-small cell lung cancer in 1 of 2 patients. TBNA was not helpful in other diseases including Pneumocystis carinii pneumonia, infection with Cryptococcus or Nocardia, bacterial pneumonia, viral pneumonia, and Kaposi's sarcoma. No pulmonary diagnosis was established in five patients. No complications of TBNA occurred. We conclude that TBNA through the flexible bronchoscope is safe and effective in the diagnosis of intrathoracic adenopathy in HIV-infected patients, and is particularly efficacious in the diagnosis of mycobacterial disease. Furthermore, TBNA may provide the only diagnostic specimen in almost one-third of HIV-infected patients, thereby sparing these patients more invasive procedures such as mediastinoscopy

Isoniazid (INH) is one of the most important first line drugs in the treatment of tuberculosis. We utilized high performance liquid chromatography with a hydrazone extraction technique to measure INH in bronchoalveolar lavage (BAL) fluid specimens from six patients with active pulmonary tuberculosis. We found BAL fluid INH levels to be similar to 2-h peak serum levels. The concentration of INH in BAL fluid from lobes with infiltrate was similar to the concentration of INH in BAL fluid from lobes without infiltrate (0.062 microgram/ml and 0.073 microgram/ml, respectively). After adjusting for protein concentration in the BAL fluid, INH levels in lobes with infiltrate were threefold lower than in lobes without infiltrate. The correlation between the concentration of INH in serum and BAL fluid approached significance after correcting for protein (lobes with infiltrate, r2 = 0.60 (p = 0.07); lobes without infiltrate, r2 = 0.50 (p = 0.12). INH penetrates into bronchoalveolar fluid, and concentrations of INH in the BAL fluid suggest that assessment of the INH serum concentration is adequate to evaluate bioavailability of the drug in patients with pulmonary tuberculosis

Pulmonary alveolar proteinosis (PAP) is a rare disease characterized by an excessive accumulation of surfactant lipids and proteins in the alveolar space. In mice with a homozygous deletion of granulocyte macrophage-colony stimulating factor (GM-CSF), their phenotype mimics PAP. To evaluate whether the knockout mouse model mimics human disease, we evaluated GM-CSF expression in alveolar macrophages from a patient with PAP. We performed multiple whole lung lavages on a patient with PAP, and cultured BAL cells in the presence or absence of LPS. In contrast to the GM-CSF knockout mouse, human BAL cells from a patient with PAP expressed mRNA for GM-CSF following LPS stimulation. However, similar to the knockout mouse, GM-CSF protein release from BAL cells was undetectable with or without LPS. BAL cells from normal human controls released GM-CSF in abundance after LPS stimulation. In BAL cells from the patient with PAP, neutralization of interleukin-10 (IL-10) by anti-IL-10 antibody, resulted in enhanced GM-CSF production. Thus, alveolar macrophages from a PAP lung have deficient GM-CSF production analogous to the GM-CSF knockout mice; in contrast, human cells from a PAP lung have an intact GM-CSF gene. This case report illustrates an important difference between the knockout mouse model of PAP and the human disease

OBJECTIVE AND METHODS: While pulmonary aspergilloma has been well described in immunocompetent hosts, to date and to our knowledge, there has not been a description of pulmonary aspergilloma in the HIV-infected individual. A retrospective review of cases seen by the Bellevue Hospital Chest Service from January 1992 through June 1995 identified 25 patients with aspergilloma. To investigate the impact of HIV status on pulmonary aspergilloma, we compared clinical presentation, progression of disease, treatment, and outcome in the HIV-infected patient vs the HIV-negative patient with aspergilloma. RESULTS: Of the 25 patients identified, 10 were HIV-infected and 15 were HIV-negative. Predisposing diseases included tuberculosis (18/25, 72%), sarcoidosis (4/25, 16%), and Pneumocystis carinii pneumonia (3/25, 12%). All 25 patients had evidence of aspergilloma on chest CT. In addition, 17 of 25 patients had evidence of Aspergillus species in fungal culture, pathologic specimens, or immunoprecipitins. Hemoptysis was present in 15 of 25 (60%) (11/15 [73%] of the HIV-negative group vs 4/10 [40%] of the HIV-infected group). Severe hemoptysis (> 150 mL/d) occurred in 5 of 15 (33%) of the HIV-negative group vs 1 of 10 (10%) of the HIV-infected group. Disease progression occurred more frequently among the HIV-infected group (4/8, 50% vs 1/13, 8% in HIV-negative individuals). All patients with disease progression had lymphocyte subset CD4+ < 100 cells per microliter. Four of eight (50%) of the HIV-infected group vs 1 of 13 (8%) of the HIV-negative group died. SUMMARY AND CONCLUSIONS: We conclude the following: (1) although tuberculosis and sarcoidosis are the most prevalent predisposing diseases, P carinii pneumonia in the HIV-infected individual is a risk factor for pulmonary aspergilloma; (2) HIV-infected individuals with CD4+ < 100 cells per microliter are more likely to have disease progression despite treatment; and (3) HIV-negative patients are more likely to have hemoptysis requiring intervention