Histopathologic and microbiologic aspects of ventilator-associated pneumonia.

BACKGROUND

The relationship between microbiology and histology in patients with ventilator-associated pneumonia has been sparsely described.

METHODS

Twenty-five patients who died in the intensive care unit after their lungs had been mechanically ventilated for 72 h were studied. Twenty of the 25 died with clinical suspicion of pulmonary infection. A total of 375 immediate postmortem pulmonary biopsies were obtained after death and processed for quantitative microbiology and histology. Four evolutionary stages of pneumonia were defined: early, intermediate, advanced, and resolution.

RESULTS

At least one specimen with histologic evidence of pneumonia was found in all but two patients (92%). Histologic pneumonia was a widespread and frequent process (46%) of biopsies examined) involving predominantly the lower lobes (55% of all biopsies with pneumonia) and showing different histopathologic stages of progression coexisting in the same lung lobes. Lung cultures were frequently polymicrobial (149 of 375, 40% of the pulmonary biopsy cultures, and 20 of 25, 80% of the cases) and not always yielding the same pathogen (19 microorganisms) when comparing one lung to the other. Histopathology and microbiologic biopsy cultures showed a weak relationship (28% and 49% of species had counts > or = 10(3) cfu/g in samples without pneumonia from patients with and without prior antibiotic treatment, respectively). Histopathologic evolutionary stages were not associated with any differences in quantitative culture results of pulmonary biopsies, independently of prior administration of antibiotics. Higher bacterial concentrations of biopsy cultures were associated with the absence of prior antibiotic treatment.

CONCLUSIONS

Ventilator-associated pneumonia is a frequent diffuse and polymicrobial process showing different coexisting degrees of evolution and involving preferentially the lower lobes. Microbiology and histology can be dissociated even in the absence of prior antibiotic treatment. Lung histology appears more reliable than bacteriology as a diagnostic reference test.