Background: In thoracic surgery, postoperative pneumonia (POP) is the leading cause of postoperative morbidity and mortality. The clinical diagnosis of POP is difficult and conventional microbiological diagnostic tests perform poorly. The contribution of molecular diagnostic tests (multiplex PCR, mPCR) should be evaluated to optimize the diagnostic and therapeutic management of POP.

Objectives: The main objective is to describe the microbiological relationship between the existence of pre- (if available) and intra-operative bronchial and pulmonary bacterial colonization and the occurrence of POP. The secondary objectives are to analyze the contribution of the mPCR for the diagnosis of POP and to validate the predictive factors of POP described in the literature Material and methods: A monocentric prospective non-interventional research with minimal risks and constraints. The study population is represented by all the consecutive adult patients hospitalized for lung surgical resection (except surgical resection indicated for infectious disease) during one year. The preoperative respiratory samples within the 3 preceding months (date and type, pathogen and threshold) are recorded, if available. Intra-operative bronchial aspirate is performed for direct examination and culture (pathogen and threshold) and mPCR (PCR1). A mPCR is optionally performed on the surgical specimen (PCR2). In case of postoperative clinical suspicion of POP, invasive or non invasive samples of respiratory tract secretions are obtained for direct examination and culture (pathogen, threshold) and mPCR (PCR3). A clinical pulmonary infection score (CPIS) is calculated by integrating the results of conventional tests (CPIS1) and mPCR (CPIS2).

The pre / intra operative and postoperative microbiological relationship will be described qualitatively and quantitatively and analyzed using correlation tests. Concordances and discrepancies between conventional tests and mPCR will be studied to analyze the contribution of molecular tests in this context.

Background: Post operative pneumonia (POP) is a common and severe complication associated with a high morbidity and mortality, regardless of the type of surgery. In thoracic surgery, the global incidence of POP is estimated at 25%. Pre-operative bronchial and pulmonary bacterial colonization appear as a major risk factor for the occurrence of POP, according to the literature.

The diagnosis of POP is challenging, because the usual diagnostic criteria are poorly relevant and clinical diagnostic scores are not validated, which may explain the scarcity of published data in the literature.

Patients' fragility and comorbid conditions mainly due to smoking limit the opportunity to perform invasive microbiological diagnostic tests, and those latter perform poorly. Altogether, the pathogen(s) involved are identified in 14% to 50%.

The evaluation of the diagnostic contribution of molecular diagnostic tests (mPCR) is important in this context. Molecular diagnostic tests offer better performance (sensitivity and sensibility) than conventional test, and the results are not affected by previous exposure to antibiotics. These tests could be useful to analyze the microbiological relationship between pre or per-operative bronchial and pulmonary colonization and postoperative infection, to optimize the diagnostic and the management of POP after thoracic surgery.

Materials and methods: In this study, The investigators aim to describe the microbiological relationship between the existence of pre- (if available) and intra-operative bronchial and pulmonary bacterial colonization and the occurrence of POP. The investigators also intend to analyze the contribution of mPCR for the diagnosis of POP.

For more precision, primary and secondary outcomes descriptions are fully detailed in the corresponding section.

The investigators perform a monocentric prospective non-interventional research with minimal risks and constraints. The study population is represented by all the consecutive adult patients hospitalized for lung surgical resection (except surgical resection indicated for infectious disease) during one year at Tenon Hospital, a University teaching hospital in Paris, France. About 200 patients per year undergo a lung surgical resection. Depending on the POP incidence, the number of patients suspected of POP may vary between 25 and 50 during the study period. Study duration participation corresponds to the hospital length of stay.

PCR Film Array Pneumonia is the molecular test used in this study. It is designed to detect the most common and critical pathogens of pneumonia (bacteria, virus). The results are reported quantitatively. Patients' management is in line with recommendations and not modified by the research. The expected risks are those of the usual care.

The practical progress of the study is defined by:

• Preoperative. Patients' clinical characteristics, respiratory status and expected surgery data are collected. The preoperative respiratory samples (date and type, pathogen and threshold) are recorded if available within the 3 preceding months
• Per-operative. Intra-operative bronchial aspirate is performed for direct examination and culture (pathogen and threshold) and mPCR (PCR1). Surgery characteristics are collected. Per-operative antibioprophylaxis is in line with recommendations.
• Post operative. Identification of post operative complications (respiratory and extra respiratory) including POP. POP diagnosis is based on clinical and microbiological data (Clinical Pulmonary Infection Score). In case of clinical suspicion of POP, respiratory samples are performed for direct examination and culture (pathogen and threshold) and mPCR (PCR2). A CPIS score is calculated with conventional microbiological tests results (CPIS 1) and PCR results (CPIS2). Hospital health care and vital status at discharge are recorded.

Primary and secondary outcomes measures are fully detailed in the corresponding section.

Statistical analysis: The statistical analysis will be performed at the end of the study. The characteristics of the patients will be described and compared between two groups, i.e. patients with POP and patients without POP. Qualitative variables will be described by size and frequency, and quantitative variables by mean and standard deviation or median and inter-quartile range. Between-groups comparisons will be performed using a Chi2 or a Fisher exact test for qualitative variables, and a Student t test or Mann-Whitney U test for the quantitative variables.

For the primary outcome, the proportion of patients for whom the intra operative colonization strain and the postoperative infection strain are the same will be calculated (with their 95% confidence interval).

For the secondary outcomes, the overall concordance rate, and the qualitative and quantitative diagnostic discrepancies will be calculated with their 95% confidence interval between conventional and molecular tests. The proportions of patients with appropriate antibiotic therapy and with targeted antibiotic therapy will be estimated and compared between conventional and molecular tests.

The predictive factors for POP occurrence will be assessed by a logistic regression model.

Inclusion Criteria:

• Adult patients hospitalized for lung surgical resection
• Consenting to research

Exclusion Criteria:

• surgical resection indicated for infectious disease

All the consecutive adult patients hospitalized for lung surgical resection (except surgical resection indicated for infectious disease) for one year at Tenon hospital, a University teaching in Paris, Assistance publique - Hôpitaux de Paris, France.

About 200-250 patients undergo a thoracic surgery per year and are supported in the thoracic surgery department of Tenon hospital. If we consider a 25% incidence rate of post operative pneumonia, we expect 50-60 patients to be included during the study period.

individual participant data collected in this study will be used for publication.