Clinical factors associated with the development of Staphylococcus aureus pneumonia in patients with airway colonization Authors: Reimi Mizushima, Yuta Kono, Haruka Motohashi, Mariko Kogami, Mayuko Ishiwari, Ryota Kikuchi, Kenichi Kobayashi, Shinji Abe Published: 2026 ABSTRACT BackgroundStaphylococcus aureus (SA) remains a major cause of infection-related mortality, and the majority of SA infections are thought to originate from strains that colonize the host prior to infection. However, clinical factors associated with the development of SA pneumonia in patients with airway SA colonization remain poorly defined. We aimed to identify clinical factors associated with the development of SA pneumonia in patients with airway SA colonization.MethodsWe conducted a retrospective observational study of 48 adult patients with airway SA colonization identified by sputum cultures at Tokyo Medical University Hospital between April 2015 and September 2022. Airway SA colonization was defined as detection of SA in two sputum cultures obtained at least one week apart, and cases of SA pneumonia developing within one year after colonization were evaluated.ResultsAmong 48 patients with SA colonization, 5 (10.4%) developed SA pneumonia. Patients who developed SA pneumonia had significantly higher stable-phase white blood cell (WBC), neutrophil, and eosinophil counts compared with those who did not. Receiver operating characteristic analysis using WBC counts to predict SA pneumonia yielded an area under the curve of 0.840 (95% confidence interval: 0.711–0.968). Although WBC counts showed discriminatory performance, their independent association with SA pneumonia could not be confirmed in this small cohort. Idiopathic pulmonary fibrosis (IPF) was significantly more frequent in patients who developed SA pneumonia and was associated with its development.ConclusionsThese findings suggest that progression from airway SA colonization to SA pneumonia is influenced by underlying lung disease, particularly IPF. WBC counts showed potential utility for risk stratification, although their independent predictive role remains uncertain. Highlight box Key findings • Patients with airway Staphylococcus aureus (SA) colonization and idiopathic pulmonary fibrosis (IPF) appeared to be at higher risk of subsequent SA pneumonia. • White blood cell (WBC) counts showed discriminatory performance for risk stratification. What is known and what is new? • Airway SA colonization is considered a potential source of subsequent infection, but predictors of progression to pneumonia are unclear. • This study suggests that structural lung disease, particularly IPF, may be an important determinant of SA pneumonia. What is the implication, and what should change now? • Patients with airway SA colonization and IPF may require closer monitoring. • Simple clinical markers such as WBC may help identify higher-risk patients. Introduction Staphylococcus aureus (SA) infection remains a major cause of morbidity and mortality worldwide (1). Most severe SA infections occur in hospitalized patients, with lower respiratory tract infections, bloodstream infections, and intra-abdominal infections accounting for the majority of cases. Previous studies have demonstrated that approximately 80% of severe SA infections originate from strains that have colonized the host prior to infection, establishing SA colonization as a major source of infection and an important risk factor for SA infection (2-4). SA colonization is commonly observed in healthy individuals, particularly in the nasal cavity, intestinal tract, skin, pharynx, and vagina. It is estimated that 20–30% of adults are colonized with SA in the nasal cavity or intestinal tract (5,6). A recent meta-analysis in low- and middle-income countries found that 26% of individuals are colonized with SA in the nasal cavity and/or throat, with methicillin-resistant Staphylococcus aureus (MRSA) accounting for 15% of isolates (7). Colonization at these sites has been shown to be associated with the development of SA infections (3,8,9). In contrast, the clinical significance of SA colonization in the respiratory tract has not been fully elucidated. SA is known to colonize not only the nasal cavity but also the pharynx (10). In patients with chronic lung diseases, such as bronchiectasis, SA colonization of the lower respiratory tract is occasionally observed (11). Current pneumonia guidelines, including those from the American Thoracic Society/Infectious Diseases Society of America, recommend consideration of empirical therapy against MRSA in patients with a prior history of MRSA detection in respiratory specimens (12,13), indicating that airway SA colonization is regarded as a clinically relevant risk factor for SA pneumonia. However, it remains unclear which patients with airway SA colonization are at increased risk of subsequently developing SA pneumonia. The aim of this study was therefore to identify clinical factors associated with the development of SA pneumonia in patients with airway SA colonization. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2026-1-0377/rc). Methods Study design and patients This retrospective observational study included adult patients (≥18 years) in whom airway SA colonization was identified through sputum cultures submitted to the Department of Respiratory Medicine at Tokyo Medical University Hospital between April 1, 2015, and September 30, 2022. Airway SA colonization was defined as the detection of SA in two sputum cultures obtained at least one week apart, as previously described (1). Patients in whom sufficient clinical data, including stable-phase blood test results, chest computed tomography (CT), and adequate follow-up (at least one year or until the development of SA pneumonia), were not available were excluded. Sputum cultures were performed as part of routine clinical care, primarily in patients with respiratory symptoms or for longitudinal microbiological monitoring in chronic lung diseases. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Ethics Committee of Tokyo Medical University (approval No. T2024-0105). Given the retrospective design, informed consent was obtained using an opt-out approach. Definition of SA pneumonia SA pneumonia was defined as pneumonia developing within one year after first confirmation of SA colonization and meeting all of the following criteria: new or worsening respiratory symptoms (e.g., cough, sputum production, or dyspnea), radiographic evidence of pneumonia, elevated inflammatory markers, and detection of SA in sputum cultures. When multiple microorganisms were isolated, the organism with the highest colony count was considered the causative pathogen. All diagnoses were retrospectively reviewed and confirmed by experienced respiratory physicians (M.R. and Y.K.) according to the predefined study criteria. Data collection and clinical variables Clinical data, including age, sex, smoking history, underlying respiratory diseases, chest CT findings, and laboratory data, were extracted from medical records. Blood test results obtained during clinically stable periods, defined as the absence of active infection, at the time point closest to the first detection of SA were used for analysis. Chest CT findings were evaluated using scans performed closest to the time of first SA detection. The association between the development of SA pneumonia and clinical variables was evaluated. Radiological assessment The presence of bronchiectasis and emphysema was evaluated using chest CT. Bronchiectasis was defined by any of the following findings: a bronchial diameter exceeding that of the accompanying pulmonary artery, lack of bronchial tapering toward the periphery, or visualization of bronchi within 10 mm of the pleural surface. Traction bronchiectasis associated with interstitial lung disease was excluded. Emphysema was defined as the presence of low-attenuation areas on chest CT. Statistical analysis Continuous variables are presented as medians with interquartile ranges. Categorical variables were compared using Fisher’s exact test. Continuous variables were analyzed using the Mann-Whitney U test. The predictive performance of white blood cell (WBC) counts for SA pneumonia was assessed using receiver operating characteristic (ROC) curve analysis, and the area under the curve (AUC) with 95% confidence interval (CI) was calculated. Univariable and multivariable analyses were performed using Firth’s penalized logistic regression as part of an exploratory analysis, given the very small number of events. Statistical analyses were performed using EZR, and a P value <0.05 was considered statistically significant. No formal sample size calculation was performed due to the retrospective design of the study. Results During the study period, 419 patients had sputum cultures positive for SA, of whom 60 met the definition of airway SA colonization. After excluding 12 patients with insufficient clinical data, 48 patients were included in the final analysis. All patients were followed for up to one year after the first confirmation of airway SA colonization. Among these patients, 5 developed SA pneumonia within one year after the first confirmation of SA colonization. In patients who did not develop SA pneumonia, 3 patients developed non-SA pneumonia, whereas 40 did not develop pneumonia. The causative pathogens of non-SA pneumonia were Streptococcus pneumoniae in one case, Klebsiella pneumoniae in one case, and were unidentified in one case. Patient characteristics and laboratory findings are summarized in Table 1. There were no significant differences in age, sex, or smoking history between patients who developed SA pneumonia and those who did not. In peripheral blood analyses, compared with patients who did not develop SA pneumonia, patients who developed SA pneumonia had significantly higher stable-phase peripheral blood WBC counts [8,300 (7,200–8,700) vs. 6,300 (5,075–7,300) cells/mm3, P=0.01], peripheral blood neutrophil counts [5,677 (4,686–6,177) vs. 3,965 (3,058–5,011) cells/mm3, P=0.03], and peripheral blood eosinophil counts [349 (226–398) vs. 130 (64–213) cells/mm3, P=0.02]. Table 1Differences in patient characteristics between patients who developed SA pneumonia and those who did notVariablePatients who developed SA pneumonia (n=5)Patients who did not develop SA pneumonia (n=43)P valueAge, years76 (68–79)68 (62–74.5)0.15Female2 [40]24 [56]0.64Smoking status Never smoker2 [40]26 [60]0.63Blood test White blood cell count (cells/mm3)8,300 (7,200–8,700)6,300 (5,050–7,300)0.01* Neutrophil proportion (%)67.2 (66.0–68.4)62.9 (57.3–72.0)0.51 Neutrophil count (cells/mm3)5,677 (4,686–6,177)3,965 (3,058–5,011)0.03* Eosinophil proportion (%)4.2 (2.6–5.6)2.5 (1.3–3.4)0.12 Eosinophil count (cells/mm3)349 (226–398)130 (64–213)0.02* Lymphocyte proportion (%)22.1 (20.5–23.2)26.1 (18.0–31.0)0.61 Lymphocyte count (cells/mm3)1,922 (1,702–2,020)1,380 (1,041–1,820)0.08 Albumin (g/dL)3.4 (3.3–3.6)3.9 (3.2–4.2)0.50 BUN (mg/dL)20.1 (13.3–24.0)15.2 (13.0–17.3)0.23 Creatinine (mg/dL)0.78 (0.75–1.37)0.66 (0.57–0.87)0.12 CRP (mg/dL)1.35 (0.42–2.75)0.30 (0.18–1.50)0.10Underlying respiratory disease Bronchiectasis1 [20]29 [67]0.059 Emphysema3 [60]15 [35]0.34 Asthma2 [40]5 [12]0.14 Idiopathic interstitial pneumonia2 [40]5 [12]0.14 IPF2 [40]1 [2]0.02* Non-IPF0 [0]4 [9]>0.99Data are presented as n [%] or median (interquartile range). *, significant, P<0.05. BUN, blood urea nitrogen; CRP, C-reactive protein; IPF, idiopathic pulmonary fibrosis; SA, Staphylococcus aureus. Differences in patient characteristics between patients who developed SA pneumonia and those who did not Data are presented as n [%] or median (interquartile range). *, significant, P<0.05. BUN, blood urea nitrogen; CRP, C-reactive protein; IPF, idiopathic pulmonary fibrosis; SA, Staphylococcus aureus. ROC analysis using WBC counts to predict SA pneumonia yielded an AUC of 0.840 (95% CI: 0.711–0.968) (Figure 1). A cutoff value of 7,100 cells/mm3 provided a sensitivity of 72.1% and a specificity of 100%. ROC curve analysis of white blood cell counts for predicting SA pneumonia. ROC curve showing the predictive performance of stable-phase peripheral blood white blood cell counts for the development of SA pneumonia in patients with airway SA colonization. The area under the curve was 0.840 (95% confidence interval: 0.711–0.968). A cutoff value of 7,100 cells/mm3 yielded a sensitivity of 72.1% and a specificity of 100%. ROC, receiver operating characteristic; SA, Staphylococcus aureus. Bronchiectasis was common in patients with SA colonization in the overall cohort (63%), but its prevalence tended to be lower in patients who developed SA pneumonia than in those who did not (20% vs. 67%, P=0.059). In contrast, IPF was infrequent in the overall cohort (6%), but was significantly more prevalent in patients who developed SA pneumonia (40% vs. 2%, P=0.02). In univariable Firth’s penalized logistic regression analysis, IPF was significantly associated with the development of SA pneumonia [odds ratio (OR), 20.2; 95% CI: 2.13–276; P=0.01], whereas WBC counts were not significantly associated (OR per 100 cells/mm3 increase, 1.02; 95% CI: 0.99–1.05; P=0.13). In multivariable Firth’s penalized logistic regression analysis, with SA pneumonia as the dependent variable and IPF and WBC counts as independent variables, IPF remained significantly associated with SA pneumonia (OR, 18.0; 95% CI: 1.9–249; P=0.01), whereas WBC counts were not significantly associated (OR per 100 cells/mm3 increase, 1.02; 95% CI: 0.99–1.05; P=0.14). Discussion In this retrospective observational study, stable-phase WBC counts were higher in patients who developed SA pneumonia, and WBC counts demonstrated discriminatory performance in ROC analysis, suggesting potential clinical utility as a marker for risk stratification of SA pneumonia. Moreover, the presence of IPF was associated with the subsequent development of SA pneumonia in patients with airway SA colonization. Previous studies have established that SA infections often arise from colonizing strains (2,3,8,9); however, factors that stratify the risk of pneumonia in airway-colonized populations remain poorly defined. Although current pneumonia guidelines consider prior MRSA detection in respiratory specimens when selecting empirical therapy (12,13), they do not provide guidance on risk stratification for the development of SA pneumonia in patients with airway SA colonization. Our findings suggest that simple, routinely available clinical parameters may help identify patients with airway SA colonization who are at increased risk of subsequent SA pneumonia. The ROC analysis of WBC counts demonstrated discriminatory performance for the prediction of SA pneumonia. However, WBC counts were not significantly associated with SA pneumonia in univariable analysis. Similarly, multivariable analysis did not demonstrate an independent association between WBC counts and SA pneumonia after adjustment for IPF, and these results should be interpreted with caution given the very small number of outcome events. This apparent discrepancy reflects fundamental differences between discrimination (ROC analysis) and estimation of independent associations (regression analysis), as well as the limited statistical power inherent to small sample sizes. While WBC counts may have potential utility for risk stratification, their independent predictive role remains uncertain. It is possible that WBC counts reflect non-specific host factors, including underlying conditions such as IPF, rather than serving as a direct indicator of disease risk. With regard to underlying respiratory diseases, bronchiectasis was frequently observed in patients with airway SA colonization but was not associated with the development of SA pneumonia. Previous reports have estimated that the prevalence of SA colonization is approximately 10% in patients with bronchiectasis (11,14,15). In contrast, the prevalence of bronchiectasis in patients with airway SA colonization has not been well characterized. In the present study, bronchiectasis was identified in 63% of patients with airway SA colonization, suggesting that bronchiectasis may be strongly associated with SA colonization of the airway. The role of airway SA colonization in disease severity in bronchiectasis remains unclear. One previous study failed to demonstrate an association between SA isolation and frequent exacerbations, reduced lung function, or hospital admissions in patients with bronchiectasis (16), whereas another study reported that airway SA colonization in patients with bronchiectasis was associated with reduced lung function, more extensive radiological involvement, and an increased frequency of mild-to-moderate exacerbations (17). However, the latter study did not find a clear association between SA colonization and severe exacerbations or pneumonia requiring hospitalization. These findings suggest that airway SA colonization does not necessarily contribute directly to the development of pneumonia in patients with bronchiectasis. Consistent with these observations, bronchiectasis was not associated with SA pneumonia in the present study. The heterogeneous nature of bronchiectasis, encompassing a wide range of disease severity, may have limited the ability to detect its impact on pneumonia risk. In addition, routine clinical management in patients with bronchiectasis—including airway clearance, regular microbiological surveillance, and early initiation of antibacterial therapy—may have contributed to preventing progression from airway colonization to pneumonia. In contrast, IPF was infrequent in the overall cohort but was significantly more common in patients who developed SA pneumonia. Previous studies have reported that SA is commonly isolated from respiratory specimens in hospitalized patients with IPF (18-20). Experimental models of pulmonary fibrosis have demonstrated impaired airway clearance of SA due to reduced neutrophil bactericidal activity and diminished macrophage phagocytic capacity (21). These findings suggest that advanced fibrotic lung disease may predispose patients to SA pneumonia by impairing local immune defense mechanisms and disrupting normal lung architecture, thereby facilitating progression from airway colonization to infection. Consistent with these observations, our results suggest that IPF may be a potentially important determinant of progression from airway colonization to SA pneumonia. This study has several limitations. First, this was a retrospective, single-center study conducted in a Japanese cohort with a small sample size, and the limited number of events restricted robust multivariable analyses. The very small number of outcome events increases the risk of type I error and may lead to overestimation of effect sizes, including the discriminatory performance observed in ROC analysis. In addition, residual confounding, including age, disease severity, and comorbidities, cannot be excluded, and caution is warranted when extrapolating these findings to other populations, such as Western cohorts and intensive care unit populations. Second, sputum cultures were obtained based on clinical indications rather than systematic screening, which may have introduced selection bias. Third, due to the retrospective design and lack of longitudinal data, we could not evaluate temporal changes in bacterial burden or clinical parameters preceding pneumonia onset. Fourth, the absence of a non–SA-colonized control group limits the ability to distinguish general risk factors for pneumonia from those specific to progression from SA colonization. Finally, the distinction between colonization and infection is inherently indistinct, and misclassification bias cannot be excluded. Conclusions Our results suggest that airway SA colonization does not uniformly confer a high risk of pneumonia. Rather, progression to SA pneumonia may be influenced by both underlying structural lung disease, such as IPF, and systemic inflammatory activity. Structural lung disease appeared to have a stronger association in this small cohort. Further large-scale prospective studies are needed to validate these findings and develop risk prediction tools. Supplementary The article’s supplementary files as Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2026-1-0377/prf