ABSTRACT
<p><b>OBJECTIVE</b>To study the clinical significance of virulence genes exo U and exo S of type III secretion system (TTSS) of Pseudomonas aeruginosa (PA).</p><p><b>METHODS</b>One hundred and eighty-nine clinical isolates of PA were collected from five hospitals. The incidence of virulence genes exo U and exo S in PA were determined with PCR. Minimum inhibitory concentration of anti-bacterial drug for PA was determined with microdilution method. The clinical features and outcomes of 60 hospitalized patients colonized or infected with exo U+/exo S- positive or exo U-/exo S+ positive PA isolated from sputum were analyzed retrospectively. Data were processed with chi-square test.</p><p><b>RESULTS</b>Among the 189 PA isolates, 85.2% (161/189) harbored TTSS genes, including exo U-/exo S+ type (120 isolates), exo U+/exo S- type (31 isolates), exo U-/exo S- type (7 isolates), and exo U+/exo S+ type (3 isolates). 72.0% (72/100) isolates from sputum and 81.5% (44/54) isolates from blood belonged to exo U-/exo S+ genotype. Compared with those of TTSS-negative isolates, the antimicrobial resistance of TTSS-positive isolates to cefoperazone/sulbactam, ceftazidime, amikacin, and cefepime were lower (with χ² value respectively 10.1, 16.1, 9.3, 33.8, P values all below 0.01). The antimicrobial resistance to all examined drug between exo U-/exo S+ type and exo U+/exo S- type isolates was close (with χ² values from 0.08 to 2.04, P values all above 0.05). Patients detected with exo U+/exo S- positive PA isolated from sputum were significantly associated with PA infection, and they usually had history of tracheal intubation, ICU hospitalization, and combined use of drugs for anti-infection treatment. Patients detected with exo U-/exo S+ positive PA isolated from sputum were significantly associated with PA colonization, which had basic lung disease and better outcome than the former infection type.</p><p><b>CONCLUSIONS</b>The TTSS exists in most clinical isolates of PA. Detection of exo U or exo S of PA isolated from sputum is helpful for the analysis of clinical features and outcome of patients.</p>