Clinical, immunological and bacteriological characteristics of H7N9 patients nosocomially co-infected by Acinetobacter Baumannii: a case control study

Abstract

Background: Bacterial co-infection of patients suffering from influenza pneumonia is a key element that increases morbidity and mortality. The occurrence of Acinetobacter baumannii co-infection in patients with avian influenza A (H7N9) virus infection has been described as one of the most prevalent bacterial co-infections. However, the clinical and laboratory features of this entity of H7N9 and A. baumannii co-infection have not been systematically investigated.

Methods: We collected clinical and laboratory data from laboratory-confirmed H7N9 cases co-infected by A. baumannii. H7N9 patients without bacterial co-infection and patients with A. baumannii-related pneumonia in the same hospital during the same period were recruited as controls. The antibiotic resistance features and the corresponding genome determinants of A. baumannii and the immune responses of the patients were tested through the respiratory and peripheral blood specimens.

Results: Invasive mechanical ventilation was the most significant risk factor for the nosocomial A. baumannii co-infection in H7N9 patients. The co-infection resulted in severe clinical manifestation which was associated with the dysregulation of immune responses including deranged T-cell counts, antigen-specific T-cell responses and plasma cytokines. The emergence of genome variations of extensively drug-resistant A. baumannii associated with acquired polymyxin resistance contributed to the fatal outcome of a co-infected patient.

Conclusions: The co-infection of H7N9 patients by extensively drug-resistant A. baumannii with H7N9 infection is an important issue which deserves attention. The dysfunctions of immune responses were associated with the co-infection and were correlated with the disease severity. These data provide useful reference for the diagnosis and treatment of H7N9 infection.

Keywords: Acinetobacter baumannii; Avian influenza A(H7N9) virus; Extensively drug-resistant bacteria; Immune responses; Nosocomial infection; Pneumonia.

Fig. 1

The occurrence of nosocomial A. baumannii infection in H7N9 patients during the timeline of the 2013–2014 H7N9 epidemic wave in the hospital. a. The weekly reported cases infected by H7N9 and A. baumannii, respectively, from December 2013 to April 2014 in Shenzhen Third People’s Hospital, Guangdong China. The first week to the fourth week of each month were denoted as W1 to W4. b. The linear correlation of weekly reported numbers of H7N9 patients and A. baumannii patients. c. The timeline of the laboratory test and clinical treatment of the H7N9 and A. baumannii co-infected patients. The corresponding dates for timeline were shown below as the horizontal axis. Flu: The persistent period for positive result of H7N9 RNA detected by RT-PCR; Baumannii: The period for A. baumannii positive; Admission: the date for the admission of the patient; Antivirus: the period for the anti-influenza drug therapy; Antibiotics: the period for the antibiotic use; Glucocorticoid: the period for glucocorticiod pulse therapy; Invasive mechanical ventilation: the period for the manipulation of invasive mechanical ventilation

Fig. 2

The risk factors for H7N9-A. baumannii co-infection. a. The comparison of the manipulation of non-invasive ventilation (NV) and invasive mechanical ventilation (MV) between H7N9-A. baumannii co-infected cases (H7N9 + AB) and H7N9 control patients without bacterial co-infection (H7N9). b. The time from the disease onset to the use of antibiotics (Onset to use) and the total time for the antibiotic use (Use to stop) is compared between H7N9 + AB cases and H7N9 controls. c. The time from the disease onset to the use of glucocorticoid (Onset to use) and the total time for the glucocorticoid use (Use to stop) is compared between H7N9 + AB cases and H7N9 controls. d. The total dosages for glucocorticoid and gamma globulin compared between between H7N9 + AB cases and H7N9 controls

Fig. 3

The disease severity of H7N9 patients co-infected by A. baumannii. a. The oxygenation index (OI, PaO₂/FiO₂) on admission compared between H7N9-A. baumannii co-infected cases (H7N9 + AB), H7N9 control patients without bacterial co-infection (H7N9), and A. baumannii infected control patients with pneumonia (AB group). b. The lowest PaO₂/FiO₂ of H7N9 + AB cases, H7N9 controls and A. baumannii controls during the disease process. c. The days for the PaO₂/FiO₂ abnormity of the three patient groups. d. The longitudinal variation of chest radiographs scores (SRC) during the disease process of H7N9 + AB patients with fatal outcome (Left), survived H7N9 + AB patients (Middle) and H7N9 infection without bacterial co-infection (Right). The SRC = 15 on day 20 was denoted as gray dash lines

Fig. 4

Dysfunctional immunity in H7N9 patients co-infected by A. baumannii. The plasma cytokines IL-6 (a) and IL-8 (b) in the H7N9-A. baumannii co-infected patients (H7N9 + AB) are higher than the H7N9 control patients (H7N9 group) during week 2 (Day 8–14), week 3 (Day 15–21) and week 4 (Day 22–28) after the admission. In contrast, IL-6 and IL-8 in the H7N9 control patients were decreasing gradually from week 1 (Day 1–7) to week 4 (Day 22–28) after hospitalization. Through the analysis of T-lymphocyte subtypes, the ratio of T-cells in total white blood cells on admission (c) was compared between H7N9 + AB cases and H7N9 controls and A. baumannii pneumonia control patients (AB group). The ratio of CD4⁺ T-cells/CD8⁺ T-cells (d) was also calculated among cases and controls. The time for CD4⁺ T-cell count restoring was longer for H7N9 + AB patients (e). The CD3⁺, CD4⁺ and CD8⁺ T-cell counts were compared between the co-infected cases and controls (f). The influenza virus-specific T-cell responses (g) were detected through ELISPOT by using the freshly-isolated PBMCs from the patients. H7N9-specific peptide pools and conserved peptide pools were used as the stimulators