This single center study was conducted according to the principles expressed in the Declaration of Helsinki. Institutional review board approval was obtained. The requirement for informed consent was waived for this retrospective study.

Patients were identified by means of a full-text database query of all CT-scans conducted in our tertiary care university medical center over an 11-year period using the terms „CT-guided,” “lung” and “biopsy” in the Radiological Information System (Nexus.medRIS, Version 8.42, Nexus, Villingen-Schwenningen, Germany). Inclusion criteria were technically successful CT-PCB of lung lesions. Exclusion criteria were: preexisting pneumothorax and fluid specimen aspiration (in case of lung abscess).

All patients referred to the Department of Radiology were inpatients, and senior radiologists experienced in CT-guided percutaneous interventions performed or supervised the procedures. All interventions were performed under local anesthesia. Piritramide was additionally administered intravenously at the responsible interventionalist’s discretion and determination of dosage directly after positioning the patient on the CT table and connecting the patient to a surveillance monitor measuring the heart rate and oxygen saturation rate. By administration of piritramide level 1 to level 2 of sedation and analgesia according to the American Society of Anesthesiologists definition was induced [15]. Interventions were conducted either using sequential CT guidance or CT fluoroscopy, using one of the following CT scanners: Siemens Somatom Plus 4, Siemens Somatom Sensation 16 and Siemens Somatom Definition AS (Siemens Healthcare, Erlangen, Germany). All biopsies were core biopsies performed with semi-automatic notch sample devices and coaxial technique was the preferred biopsy method at our institution.

To rule out pneumothorax 1.) CT slices focused on the level of the biopsy site at the end of the procedure and 2.) chest x-ray about 3 h after the procedure were obtained. In case of a clinically relevant pneumothorax (depending on size and clinical symptoms), a chest tube was inserted. Patients were under observation for at least one night.

The analyzed data were collected by reviewing the medical records, procedural CT images and post-procedural chest x-ray images. The following patient-related data were noted: age, gender, weight and height with body mass index, history of smoking, previous thoracic surgery or tuberculosis, major comorbidities concerning the lungs and airways (asthma, chronic obstructive pulmonary disease and emphysema) and sleep apnea.

Concerning the target lesion the following parameters were recorded: location, size, pleural contact (yes/pleural tag/no), distance to the parietal pleura (measured along the needle path from the parietal pleura to the needle insertion point of the lesion), proximity to the diaphragm (defined by concomitant visibility of the diaphragm on a transversal CT slice at the level of the target lesion), cavitation (if present, wall thickness), presence of emphysema along the needle trajectory and histopathological diagnosis.

The following procedure-related data were noted: patient positioning (prone vs. supine vs. lateral), intravenous opioid administration (yes/no; dosage), caliber of the outmost biopsy needle (summarized to larger or equal to 18G and smaller or equal to 19G) and the number of pleural passages per procedure (including crossing of pleural fissures).

Complications were documented according to the standards of practice guidelines of the Cardiovascular and Interventional Radiological Society of Europe (CIRSE) [16]. In case of pneumothorax the largest distance of retraction of pulmonary surface was measured. Pneumothorax was classified into (1) mild asymptomatic and (2) symptomatic requiring chest tube placement (duration of chest tube therapy was documented). Nausea and vomiting as potential adverse effect of piritramide were documented.

Continuous variables are presented as mean ( ± standard deviation, SD) and categorical variables as absolute and relative frequencies. For single factor analysis of/to test for differences between continuous variables the unpaired Student t test was used, and for single factor analysis of/to test for differences between categorical data the Pearson’s chi-square test was applied. To assess risk factors for developing a pneumothorax, univariate logistic regression models were calculated in a first step. Afterwards, 8 variables with clinical relevance, which might affect the risk for pneumothorax (proximity of the target lesion to the diaphragm, emphysema along the needle trajectory, distance of the target lesion to the pleural surface) or with statistical significance in the univariate logistic regression model were selected and added to a multivariable model (full model). Since the number of events per variable is quite low ( < 10) in this model, a reduced multivariable logistic regression model was calculated, including only significant variables after using a forward selection model of the 8 pre-selected variables. Due to the smaller number of variables in the model, these have a higher power to show a significant effect. Both the full and the reduced model were calculated including 167 patients who had complete data for all considered variables. Four patients were not included in the multivariable analysis because the biopsy was performed in lateral position. For all logistic regression models, odds ratios (OR) and corresponding 95% confidence intervals (95%-CI) are reported as effect estimates. A p value < 0.05 was considered as statistically significant. All analyses were performed using IBM SPSS Statistics 25 (IBM, Armonk, New York, USA).