Bronchial colonization profile of patients undergoing lung resection and its impact on infectious complications in the postoperative period. Is it necessary to evaluate the collection method and risk of contamination?
Introduction: The pattern of bronchial colonization in patients requiring pulmonary resections is little described in the literature. The primary objective of this study is to evaluate the profile of bronchial colonization in patients undergoing lung resection in a general hospital through the collection of the resected specimen only. The secondary objective is to evaluate prognostic factors of infectious complications after lung resection, including the colonization of the lower airways.
Methods: This was a prospective study that included all consecutive patients undergoing lung resection for noninfectious disease and without signs of acute respiratory infections. Intraoperative bronchial or lung parenchyma culture of the resected specimen was collected by the surgeon still under completely sterile conditions. A patient was considered colonized if the quantitative endobronchial culture was positive at 48h with a predominant microorganism exceeding a cutoff value of 104 colony-forming units.
Results: Negative cultures were found in 81.4% of patients. Colonization with predominant bacteria (at least 104 cfu/mL) was identified in 18.6%. The rate of infectious complications was 24.42%. The incidence of postoperative pneumonia (POP) was 20.9%. POP developed at a median of 4 days after surgery. Three patients developed empyema, and of the three, 2 had both POP and empyema. None of the studied factors was associated with postoperative infectious complications
Conclusion: We conclude there is colonization of the lower airways in patients undergoing lung resection in our institution. This finding, as well as the other analyzed factors, did not result in increased POP risk in this sample. The intraoperative collection method employed in this study should be further evaluated in larger studies.
The lungs of healthy humans have traditionally been considered to be sterile when examined by culture-based techniques (1). In patients with lung diseases, such as chronic obstructive pulmonary disease and cystic fibrosis, colonization of the lower airways (LAWs) has been well documented and associated to these diseases [2-5].
Regarding the subgroup of patients with lung diseases requiring thoracic surgeries, very few studies have described the profile of colonization of LAWs. And to the best of our knowledge, there has been no description of this subject in a Latin American country (3,6-8). However, the colonization of LAW has been associated with the risk of postoperative pneumonia (POP). Schussler et al., in a study with 507 patients and Bede et al., showed that there is significant association between bacteria in the lower airways and risk of POP [3,9], being an independent risk factor in the multivariate analysis.
A study was carried out including all consecutive patients undergoing pulmonary resections for noninfectious diseases to investigate the profile of LAW colonization. The secondary objective is to evaluate prognostic factors of infectious complications after lung resection, including LAW colonization.
Materials and Methods
This prospective study consecutively included all patients submitted to lung resection for noninfectious diseases, without signs of acute respiratory infections. Patients presenting at the date of admission for planned surgery with clinical and radiological signs of pulmonary infection (fever greater than 37.7°C, purulent sputum) were excluded from this study in cases of urgency surgery. Except for the urgencies, patients were treated with antibiotics and submitted to surgery at least 7 days after antibiotic discontinuation. Patients with a diagnosis of pulmonary tuberculosis were also excluded. The study was approved by the ethics committee of our institution under protocol number 209146141.
All data on patient characteristics, results of microbiological studies, treatment procedures and outcome (POP) were prospectively collected through our database. We studied the following risk factors for POP: age, gender, surgical risk classification of the American Society of Anesthesiology (ASA), cancer patients and the presence of bacteria in the lower airways.
All patients were intubated with a double-lumen endobronchial tube to undergo single-lung ventilation. Bronchial or pulmonary parenchyma culture of the resected specimen was collected by the surgeon still under completely sterile conditions. A patient was considered colonized if a 48-hour culture was positive with the presence of a microorganism. A policy of early extubation was systematically employed. Decisions concerning intensive care unit (ICU) admission after resection were established based on type of resection, predicted postoperative lung function and associated comorbidities. Postoperative analgesia was achieved through one of the following methods: use of epidural catheter intermittently or intermittent intravenous analgesia with morphine and dipyrone. A regular physical therapy program was started on the day of the surgery. Oral feeding was started on Postoperative Day 1.
Patients received antibiotic prophylaxis with second-generation cephalosporin (cefazolin, 2.0 g at anesthesia induction and postoperatively, 3 g/24 h for 24 h) except in cases of known or suspected allergy, or if a different type of prophylaxis was indicated [4,10]. Lung resections were performed according to standard techniques. The type of resection was recorded. The diagnoses were established by anatomopathological analysis of the resected lung specimens.
Patients were assessed twice a day. Chest roentgenograms were performed postoperatively once a day during the chest drainage period. POP definition was (1) abnormal radiographic findings (new or changing radiographic infiltrates that persisted after physical therapy or bronchoaspiration) and (2) fever greater than 37.8 oC or dyspnea, and (3) one of the following criteria: a new rise in C-reactive protein level, WBC count in the previous 24 h (with WBC count > 12×109/L)  or an increase and modification of the expectorate.
All postoperative pulmonary complications were secondarily reviewed by a clinician and a thoracic surgeon. Infections occurring within one month of surgery or during in-hospital surgical stay were recorded. Wound infection was defined as a reddened, painful, and indurated wound, not necessarily associated with bacterial isolation. Empyema was defined as the presence of purulent fluid in the pleural drainage or as the isolation of pathogens from the pleural cavity.
Results are expressed as percentages and means ± standard deviation (SD). The correlation between the studied variables and postoperative infection was performed by logistic regression analysis. Data processing and analysis were performed using the statistical software MedCalc, release 184.108.40.206. A p value less than 0.05 was considered statistically significant. The risk factors found to be predictive of POP at univariate analysis were entered into a multivariate regression analysis, to identify independent variables.
Between June 15, 2012, and December 15, 2013, 86 patients undergoing lung resections in our department were included in the study. Three patients were excluded from the study due to preexisting infections at the time of surgery (infectious interstitial lung disease) and six due to a diagnosis of pulmonary tuberculosis.
Antibiotic prophylaxis with drugs other than cefazolin was employed in three patients due to known allergy to penicillin/cephalosporins or because of coexistent cardiac valve disease. Demographic data on the remaining 86 patients, type of diseases and surgical procedures are shown in Table 1.
Cancer patients comprised 53.4% of the sample. Among them, primary lung tumors (non-small cell lung cancer) were treated through major lung resection (lobectomy or pneumonectomy). Pulmonary metastases were treated through segmentectomy. Also regarding cancer patients, some patients were submitted to segmentectomy for diagnosis of interstitial infiltrate, which showed to be lymphangitic carcinomatosis. Among the benign diseases, lobectomy and pneumonectomy were used in all patients with bronchiectasis. Segmentectomy was used for interstitial lung diseases and benign nodules that included patients with scar tissue and inflammatory nodules, in addition to hamartomas. Regarding the length of preoperative hospital length of stay, 52% of patients were admitted 1 (one) day prior to surgery (Figure 1).
Negative cultures were obtained in 70 of 86 (81.4%) patients. Colonization with predominant bacteria was identified in 16 of 86 cases (18.6%). Colonization by pathogenic bacteria occurred in 10.4% of all patients (9/86). Staphylococcus, Enterobacter, Acinetobacter baumannii and Escherichia coli species represented the most frequently involved microorganisms (Table 2). Gram-negative bacteria were obtained from cultures in nine cases. None of the patients had polymicrobial colonization.
The complication rate was 24.4% (21/86). The incidence of POP was 20.9% (18 /86). POP developed at a median of 4 (1-9) days after surgery. Empyema developed in 3 patients, and of the three, 2 had both POP and empyema. Patients with POP frequently required mechanical ventilation (40.2%), and in-hospital mortality was 22%. Of the 16 colonized patients, 4 (25%) had POP, whereas among non-colonized patients, POP occurred in 14 of 70 cases (20%) (Figure 2).
For all variables analyzed separately, the p value was greater than 0.05 and the 95% confidence interval (95%CI) included the number 1, indicating that no variable significantly influenced POP incidence, according to the available sample. Still, some alternatives were used aiming to identify some underlying statistical relevance, such as stratifying age ranges, transforming age into a categorical variable or considering the minimum value of p as 0.1, so that the variable would be subsequently analyzed in a multivariate model. However, no statistically significant impact was identified for any variable (Tables 3). Multivariate analysis was not performed due to the negative results of the univariate analysis.
The colonization of LAW in patients undergoing thoracic surgery is scarcely mentioned in the literature. It is known that patients with some lung diseases, such as cystic fibrosis and chronic obstructive pulmonary disease (COPD) have atypical pulmonary flora [11-18], but patients submitted to lung resection usually include a much wider range of diseases. The study of this profile in different institutions is important to better understand the predominant bacterial flora and whether there is or not impact on the risk of infectious complications and whether antibiotic prophylaxis needs to be reviewed. Some studies have shown that LAW colonization increases the risk of POP (2,8,9). Other studies even suggest that, as it influences the risk of POP, LAW colonization would require a change in antibiotic prophylaxis in thoracic surgery (3,7). However, it is necessary to evaluate the collection method and risk of contamination of the collected samples.
Our study described a heterogeneous flora, with many non-pathogenic microorganisms, but with an incidence (18.6%) that is comparable to that found in other studies. Yamada et al., in their study with 626 patients and Schussler et al., who assessed 478 patients in 2008, found 12.8% and 14.7% of LAW colonization, respectively (3,8). Belda et al. and Ionas et al. reported 83% and 41% of LAW colonization, respectively [9,6]. However, to the best of our knowledge, our study is the first that used the culture collection method performed by the surgeon, while still under completely sterile conditions in all patients. Previously, Ionas et al. used this technique, but in combination with protected specimen brush (PSB) through bronchoscopy in 41 patients (6). On the other hand, all studies used bronchoalveolar lavage (BAL) or PSB as collection method (2,3,6,8,9). Schussler et al. reported that they initially attempted to collect cultures from the resected specimens in the first 30 patients, but as culture results were negative, they gave up on the method, although it appears to be a more reliable result (2). It is also noteworthy that our study involved patients with different lung diseases and this fact may have influenced the incidence of bacteria in LAWs, unlike previous studies that were carried out in patients with the same disease, most with early-stage lung cancer.
According to the literature, BAL is influenced by factors such as: the collected volume, when less than 100 mL, can increase contamination by mucus and airway cells; smokers and patients with COPD may have decreased volume of the recovered fluid. This method has sensitivity and specificity values ranging between 42-93% and 45-100%, respectively . In addition to bronchoalveolar lavage (BAL), the PSB method is a procedure with greater specificity, due to lower chance of sample contamination caused by the bronchoscope passage through regions such as the oral mucosa or contact with tracheal and bronchial secretions, compared to unprotected BAL . On the other hand, the risk of sample contamination exists and operational costs are not feasible in most Latin American institutions. We understand that intraoperative collection eliminates the risk of contamination from other airway areas and the sterile conditions of the environment and operating team also warrants that the chance of contamination during material handling is also minimal.
We found no association between bronchial colonization and POP, perhaps because there were few patients with pathogenic bacteria (10.4%). Yamada et al. also found no association between LAW colonization and POP in their study (8). Belda et al. described 35.8% of patients colonized with pathogenic bacteria. Schusller et al., in 2006, first reported an incidence of 22.8% of LAW colonization by pathogenic bacteria (9,2). These studies showed an association between LAW colonization and POP. However, once again, differently from our study, they collected culture samples through PSB and BAL, increasing the chance of contamination with the upper airways and thus, possibly increasing the number of patients with positive culture and pathogens. Consequently, our results might represent the actual bacterial flora of LAWs more accurately. Another interesting factor is that Schussler et al. found a correlation between the colonizing bacteria and the causative agent of POP in only 5 of 50 patients and this finding was not statistically significant. Ionas et al. and Yamada et al. also found the same result regarding this correlation between colonizing bacteria and bacterial agent identified in patients who developed POP (3,6,8).
The incidence of POP was relatively high (20.9%), but compatible with literature data. Radu et al. described seventy-six cases (24.4%) of pulmonary resections that were complicated by postoperative pulmonary infections (7). Belda et al. described POP in 31% of the patients (9). Regarding mortality in patients with POP, the literature shows mixed results, with a mortality rate of up to 40% (2, 21). Possibly, the higher mortality is associated with the profile of operated patients. Belda et al. reported 13% of deaths in patients submitted to pulmonary resections only for early-stage primary lung cancer (9). Our study showed a mortality rate of 22% in individuals with POP. However, our sample included patients with metastatic cancer and severe inflammatory diseases.
Our study has some limitations. Ours is a small sample and, therefore, we believe that other studied factors did not influence the risk of infection. Moreover, we did not isolate microorganisms during the postoperative period in patients who developed POP, to be compared with LAW cultures collected during surgery. The analysis of antibiotic prophylaxis was not performed, because LAW colonization did not appear as a risk factor for POP. In the last decade, culture-independent DNA-based techniques have demonstrated that much more complex microbial communities reside in the lower airways, where bacterial culture has failed to reliably demonstrate resident bacteria. (22). Unfortunately, these techniques are not yet available in our institution.
We conclude that lower-airway colonization is found in patients undergoing lung resection in our institution. This finding, as well as the other analyzed factors, did not result in increased POP risk in this sample. The intraoperative collection method employed in this study should be further evaluated in larger studies to define the risk of POP associated with LAW colonization.
Table 1 – Preoperative clinical status and surgical procedures performed
Table 2 – Colonization profile of lower airways
Table 3 – Univariate analysis to identify possible risk factors for postoperative pneumonia.
Figure 1. Number of preoperative in-hospital length of stay.
Figure 2. Incidence of postoperative pneumonia (POP).