Significance of Aspergillus spp. isolation from lower respiratory tract samples for the diagnosis and prognosis of invasive pulmonary aspergillosis in chronic obstructive pulmonary disease.

ABSTRACT

BACKGROUND: Chronic obstructive pulmonary diseases (COPD) is an emerging population at risk for invasive infection of Aspergillus. Isolation of Aspergillus from lower respiratory tract (LRT) samples is important for the diagnosis of invasive pulmonary aspergillosis (IPA). The purpose of this study was to investigate the value of Aspergillus isolation from LRT samples for the diagnosis and prognosis of IPA in COPD population. METHODS: Clinical record with Aspergillus spp. isolation in COPD and immunocompromised patients was reviewed in a retrospective study. Patients were categorized and compared according to their severity of illness (admitted to general ward or ICU) and immunological function (COPD or immunocompromised). RESULTS: Multivariate statistical analysis showed that, combined with Aspergillus spp. isolation, APACHE II scores > 18, high cumulative doses of corticosteroids (> 350 mg prednisone or equivalent dose) and more than four kinds of broad-spectrum antibiotics received in hospital may be predictors of IPA in COPD (OR = 9.076, P = 0.001; OR = 4.073, P = 0.026; OR = 4.448, P = 0.021, respectively). The incidence of IPA, overall mortality, mortality of patients with IPA and mortality of patients with Aspergillus spp. colonization were higher in COPD patients in ICU than in general ward, but were similar between COPD and immunocompromised patients. CONCLUSIONS: Aspergillus spp. isolation from LRT in COPD may be of similar importance as in immunocompromised patients, and may indicate an increased diagnosis possibility of IPA and worse prognosis when these patients received corticosteroids, antibiotics, and need to admit to ICU. Aspergillus spp. isolation from LRT samples combined with certain risk factors may be useful in differentiating colonization from IPA and evaluating the prognosis of IPA in COPD patients.