Fungal empyema thoracis: an emerging clinical entity.

STUDY OBJECTIVES To analyze the clinical spectra, pathogenesis, treatment, outcome, and prognostic factors of fungal empyema thoracis. DESIGN The medical records of patients with positive fungal cultures from pleural effusions were retrospectively analyzed. SETTING A university-based tertiary care hospital in Taipei, Taiwan. PATIENTS AND METHODS From January 1990 through December 1997, patients diagnosed with fungal empyema were included in this study. The criteria for diagnosis of fungal empyema thoracis were as follows: (1) isolation of a fungal species from the pleural effusion; (2) significant signs of infection, such as fever (body temperature > 38.3 degrees C) and leukocytosis (white blood cell > 10,000/microL); and (3) isolation of the same mold species from pleural effusion on more than one occasion, or from pleural effusion and other specimens such as blood, sputum, or surgical wounds that showed evidence of tissue invasion. RESULTS Sixty-seven patients with fungal empyema thoracis were included. Their mean age was 54 years (range, 2 weeks to 93 years), and 64% (43 patients) were men. Fifty-seven patients (85%) had various underlying diseases, and 18 (27%) had more than one immunocompromising condition. A total of 73 fungal isolates were recovered from pleural effusion; the most commonly encountered were Candida species (47 isolates, 64%), Torulopsis glabrata (13 isolates, 18%), and Aspergillus species (9 isolates, 12%). Candida albicans (28 isolates) was the most common Candida species, followed by Candida tropicalis (13 isolates). Six patients (9%) had two fungal strains isolated, and 16 (24%) had concomitant bacterial empyema thoracis. Eighteen patients (27%) had concurrent fungemia. Most (56 patients, 84%) cases of fungal empyema thoracis were nosocomial, and many case (43 patients, 64%) were acquired in ICUs. Abdominal disease (20 patients, 30%), especially previous abdominal surgery and GI perforation (12% and 10%, respectively), was the most common cause of fungal empyema thoracis, followed by bronchopulmonary infection (15 patients, 22%) and chest surgery (12 patients, 18%). Forty-nine patients (73%) received systemic antifungal therapy, and 38 (57%) underwent closed drainage therapy. Eleven patients (16%) underwent pleural irrigation with normal saline solution, povidone-iodine solution, or antifungal agents. Six patients (9%) finally received decortication. All patients receiving surgery or pleural irrigation with antifungal agents survived. Despite the aforementioned management, the crude mortality was high (73%). Multivariate analysis showed a significantly increased risk of death in immunocompromised patients (relative risk, 1.58; p < 0.005) and those with respiratory failure (relative risk, 2.31; p < 0.001). Systemic antifungal therapy was associated with a significantly lower risk of death (relative risk, 0.69; p < 0.05). CONCLUSION These data imply an increasing incidence of fungal empyema thoracis in recent years and the necessity for aggressive treatment of patients with this disease.