A high α1-antitrypsin/interleukin-10 ratio predicts bacterial pneumonia in adults with community-acquired pneumonia: a prospective cohort study
Taiga Miyazaki,
Kiyoyasu Fukushima,
Kohji Hashiguchi,
Shotaro Ide,
Tsutomu Kobayashi,
Toyomitsu Sawai,
Kazuhiro Yatera,
Yoshihisa Kohno,
Yuichi Fukuda,
Yoji Futsuki,
Yuichi Matsubara,
Hironobu Koga,
Tomo Mihara,
Eisuke Sasaki,
Nobuyuki Ashizawa,
Tatsuro Hirayama,
Takahiro Takazono,
Kazuko Yamamoto,
Yoshifumi Imamura,
Norihito Kaku,
Kosuke Kosai,
Yoshitomo Morinaga,
Katsunori Yanagihara,
Hiroshi Mukae
Affiliations
Taiga Miyazaki
Division of Respirology, Rheumatology, Infectious Diseases, and Neurology, Department of Internal Medicine, Faculty of Medicine, University of Miyazaki
Kiyoyasu Fukushima
Japanese Red Cross Nagasaki Genbaku Isahaya Hospital
Kohji Hashiguchi
Japanese Red Cross Nagasaki Genbaku Hospital
Shotaro Ide
Department of Respiratory Medicine, Nagasaki University Hospital
Tsutomu Kobayashi
Sasebo Chuo Hospital
Toyomitsu Sawai
Nagasaki Harbor Medical Center
Kazuhiro Yatera
Department of Respiratory Medicine, University of Occupational and Environmental Health, Japan
Yoshihisa Kohno
Kouseikai Hospital
Yuichi Fukuda
Sasebo City General Hospital
Yoji Futsuki
Saiseikai Nagasaki Hospital
Yuichi Matsubara
Juko Memorial Nagasaki Hospital
Hironobu Koga
Aino Memorial Hospital
Tomo Mihara
Nagasaki Medical Center
Eisuke Sasaki
Ureshino Medical Center
Nobuyuki Ashizawa
Department of Respiratory Medicine, Nagasaki University Hospital
Tatsuro Hirayama
Department of Respiratory Medicine, Nagasaki University Hospital
Takahiro Takazono
Department of Respiratory Medicine, Nagasaki University Hospital
Kazuko Yamamoto
Department of Respiratory Medicine, Nagasaki University Hospital
Yoshifumi Imamura
Department of Respiratory Medicine, Nagasaki University Hospital
Norihito Kaku
Department of Laboratory Medicine, Nagasaki University Graduate School of Biomedical Sciences
Kosuke Kosai
Department of Laboratory Medicine, Nagasaki University Graduate School of Biomedical Sciences
Yoshitomo Morinaga
Department of Laboratory Medicine, Nagasaki University Graduate School of Biomedical Sciences
Katsunori Yanagihara
Department of Laboratory Medicine, Nagasaki University Graduate School of Biomedical Sciences
Hiroshi Mukae
Department of Respiratory Medicine, Nagasaki University Hospital
Abstract Background Current microbiological tests fail to identify the causative microorganism in more than half of all pneumonia cases. We explored biomarkers that could be used for differentiating between bacterial and viral pneumonia in patients with community-acquired pneumonia (CAP). Methods In this prospective cohort study conducted in Japan, data obtained from adult patients with bacterial pneumonia, including bacterial and viral coinfections (bacterial pneumonia [BP] group), and purely viral pneumonia (VP group) at diagnosis were analyzed using multivariate logistic regression analysis to identify predictors of bacterial pneumonia. Furthermore, a decision tree was developed using the predictors. Results A total of 210 patients were analyzed. The BP and VP groups comprised 108 and 18 patients, respectively. The other 84 patients had no identified causative microorganism. The two groups shared similar characteristics, including disease severity; however, a significant difference (p < 0.05) was observed between the two groups regarding sputum type; sputum volume score; neutrophil counts; and serum levels of interleukin (IL)-8, IL-10, and α1-antitrypsin (AAT). Sputum volume score (p < 0.001), IL-10 (p < 0.001), and AAT (p = 0.008) were ultimately identified as predictors of BP. The area under the curve for these three variables on the receiver operating characteristic (ROC) curve was 0.927 (95% confidence interval [CI]: 0.881–0.974). The ROC curve for sputum volume score and an AAT/IL-10 ratio showed a diagnostic cutoff of 1 + and 65, respectively. Logistic regression analysis using dichotomized variables at the cutoff values showed that the odds ratios for the diagnosis of BP were 10.4 (95% CI: 2.2–50.2) for sputum volume score (absence vs. presence) and 19.8 (95% CI: 4.7–83.2) for AAT/IL-10 ratio (< 65 vs. ≥ 65). Conclusions Considering that obtaining a definitive etiologic diagnosis with the current testing methods is difficult and time consuming, a decision tree with two predictors, namely sputum volume and the AAT/IL-10 ratio, can be useful in predicting BP among patients diagnosed with CAP and facilitating the appropriate use of antibiotics. Trial registration UMIN000034673 registered on November 29, 2018.
