Abstract
Background
Pneumothorax is consistently reported as the most prevalent complication of computed tomography-guided transthoracic needle biopsy. While most cases of pneumothorax resolve with conservative management strategies, some cases are linked to a more protracted clinical trajectory necessitating active intervention. However, few reports have described the clinical and procedural determinants of prolonged hospitalization after post-biopsy pneumothorax.
Methods
This retrospective study analyzed 1,457 CT-guided TTNB procedures performed at a tertiary referral center between January 2021 and December 2025. The incidence of post-biopsy pneumothorax was 20.5%. Among 82 patients with prolonged hospitalization due to pneumothorax, their management strategies were categorized into two groups: (1) high-concentration oxygen therapy (n = 54) and (2) invasive interventions, including percutaneous catheter drainage (PCD) or chest tube insertion (n = 28). Logistic regression models with backward selection were used to identify predictive factors for prolonged hospitalization.
Results
Among the 82 patients with prolonged hospitalization due to pneumothorax, the median age was 75 years and 75.6% were male. In multivariable analysis, a lower diffusing capacity for carbon monoxide (DLCO) was independently associated with prolonged hospitalization (adjusted odds ratio [aOR], 0.980; P = 0.004). In the high-concentration oxygen group, lower DLCO (aOR, 0.974; P = 0.007) and a trans-fissural needle route (aOR, 3.968; P = 0.028) were independently associated with prolonged hospitalization. Conversely, in the PCD/chest tube group, reduced forced expiratory volume in 1 s was independently associated with prolonged hospitalization (aOR, 0.961; P = 0.002).
Conclusions
In patients with post-biopsy pneumothorax, impaired pulmonary function may be linked to a prolonged clinical course. Incorporating these pre-procedural physiological markers may facilitate improved risk stratification and optimized hospital resource allocation.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1186/s12890-026-04534-2.
Keywords: Computed tomography, Biopsy, Pneumothorax, Hospitalization, Respiratory function tests
Introduction
Computed tomography-guided transthoracic needle biopsy (CT-guided TTNB) remains a widely used minimally invasive procedure for tissue diagnosis of focal pulmonary lesions, particularly in patients with suspected lung cancer [1–4]. Given the escalating identification of pulmonary nodules driven by the expansion of lung cancer screening programs and the heightened necessity for histological validation and molecular profiling, CT-guided TTNB persists in its significant contribution to modern pulmonary medicine [1–4]. Despite its diagnostic benefits, this procedure is associated with several complications. Notably, pneumothorax is consistently the most common and clinically significant adverse effect [4–6]. Although the reported incidence rates of post-biopsy pneumothorax vary across clinical settings and study populations, evidence from large-scale cohorts and meta-analyses indicates that pneumothorax occurs in a substantial proportion of cases; most cases of post-biopsy pneumothorax resolve under conservative management, while others may lead to a more complicated clinical course and prolonged hospitalization [4–12].
Over the past two decades, numerous studies have examined the factors associated with post-biopsy pneumothorax. Patient-related predictors reported in the literature include advanced age, history of tobacco use, and previous respiratory conditions such as emphysema, interstitial lung abnormalities (ILA), and usual interstitial pneumonia (UIP) [6, 12–17]. Additionally, lesion- and procedure-related factors, including greater lesion depth, lack of pleural contact, fissure traversal, and respiratory motion of the target lesion, are associated with an increased risk of pleural complications [2, 7–10, 15, 17, 18]. Based on these observations, several predictive models and nomograms have been proposed to estimate the likelihood of pneumothorax before biopsy and to aid in procedural planning and risk stratification [15, 19–22]. These findings indicate that both patient vulnerability and technical characteristics of the biopsy procedure are critical factors in the development of pneumothorax [15, 19, 20]. However, most previous studies have primarily focused on the occurrence of pneumothorax itself or on immediate procedure-related outcomes such as chest drainage or chest tube insertion [3–12, 15, 16, 19, 20]. While previous studies have extensively evaluated the risk factors for development of post-biopsy pneumothorax, clinical determinants of the subsequent recovery course and the necessity for prolonged hospitalization remain poorly characterized. In practice, post-biopsy pneumothorax is not a uniform entity; some patients recover after brief observation, whereas others require prolonged hospitalization because of persistent oxygen demand or the need for pleural drainage [7–10]. This distinction is clinically important, because a more prolonged course can increase patient discomfort, impede recovery, and place greater demands on inpatient resources. In addition, although pre-procedural pulmonary function test (PFT) abnormalities have been implicated in biopsy-related complications, the specific contributions of physiological parameters such as forced expiratory volume in 1 s (FEV1) and diffusing capacity for carbon monoxide (DLCO) to the severity and recovery trajectory of post-biopsy pneumothorax remain insufficiently defined [23, 24].
This study aimed to identify clinical and procedural predictors of prolonged hospitalization in patients who developed pneumothorax after CT-guided TTNB. In addition, we sought to determine whether the predictors of a prolonged clinical course differed according to the intensity of management, specifically high-concentration oxygen therapy versus pleural drainage or chest tube insertion.
Methods
Study design
This single-center retrospective study was conducted at Gil Medical Center, a tertiary referral hospital in Incheon, Republic of Korea. We reviewed a series of consecutive CT-guided TTNB procedures performed on focal lung lesions between January 2021 and December 2025. Of the 1,481 patients who underwent CT-guided TTNB during the study period, 24 were excluded because of a history of prior CT-guided TTNB (n = 8) or the presence of a chest drain catheter before biopsy (n = 16). Among the remaining 1,457 patients, 299 post-biopsy pneumothorax cases developed. The study population was stratified based on whether prolonged hospitalization occurred because of post-biopsy pneumothorax. Patients who did not experience prolonged hospitalization served as the reference group (n = 217), whereas those requiring extended stays formed the case group (n = 82). Additional subgroup analyses were performed within the case group, focusing on patients who received high-concentration oxygen therapy (n = 54) and those who required percutaneous catheter drainage (PCD) or chest tube insertion (n = 28) (Fig. 1). PCD and chest tube insertion were combined into a single group as they both represent active, invasive mechanical drainage strategies.
Fig. 1.
Flow diagram of the study population. CT, computed tomography; PCD, percutaneous catheter drainage
Procedures and pneumothorax management
All patients underwent chest CT before TTNB. Pre-procedural CT images were reviewed to determine optimal patient positioning and guidance modalities based on the location of the target lesion. Specific guidance modalities were selected at the discretion of the operator. Typically, a single pleural puncture is performed. Tissue samples were obtained using a coaxial core-needle biopsy technique with a 16-gauge (G) guide sheath (Coaxial Guide Needle; TSK Laboratory, Tochigi, Japan) and an 18-G semi-automatic biopsy needle (StarCut standard type; TSK Laboratory). Four to six core samples were collected during each procedure. All biopsies were conducted in a dedicated angiography suite under local anesthesia by one of three experienced interventional radiologists [2]. Upon completion, tract sealing was performed by injecting 2–5 mL of normal saline during extraction of the guide sheath. The patient was immediately placed at the puncture site in the dependent (down) position. Finally, immediate post-procedural cone-beam CT was performed to assess potential complications.
Post-biopsy pneumothorax was identified on immediate post-procedural cone-beam CT, and follow-up chest radiography was performed approximately two hours later. Management strategies for post-biopsy pneumothorax were guided by the patient’s clinical presentation, oxygen requirements, and radiographic changes over time. Conservative management was applied to clinically stable patients with limited pneumothorax, and there was no evidence of progression on serial assessments. High-concentration oxygen therapy was administered to patients with symptomatic pneumothorax or an ongoing oxygen requirement when there was no immediate need for pleural drainage [25]. Oxygen was titrated to a target peripheral oxygen saturation of 94%–98% in patients without a risk of hypercapnic respiratory failure and 88%–92% in those at risk [25]. PCD or chest tube insertion was performed in patients with progressive, large, or symptomatic pneumothorax, including those with increasing pneumothorax size on follow-up imaging or inadequate response to conservative management [14]. Prolonged hospitalization was defined as a hospital stay > 24 h post-procedure due to the clinical management of biopsy-related pneumothorax [26]. This cutoff was selected because patients without complications or with minor, stable pneumothorax are routinely discharged within 24 h at our institution; thus, an extended stay directly indicates the need for active clinical intervention or prolonged surveillance.
Data collection
Clinical, radiological, and procedural variables were retrospectively extracted from electronic medical records and procedural imaging archives. Baseline clinical characteristics included patient age, sex, smoking history, and body mass index (BMI). Pre-biopsy pulmonary function test results, including forced vital capacity (FVC), FEV1, and DLCO were also collected. Underlying lung disease variables included emphysema and interstitial lung disease. Lesion-related variables included lesion location, size, morphological characteristics (solid, sub-solid, or ground-glass opacity), and pleural contact. Anatomical measurements included pre-biopsy skin-to-lesion, skin-to-pleura, and pleura-to-lesion distances. The procedural variables recorded were patient positioning, use of a non-coaxial approach, trans-fissural trajectory, number of biopsy specimens obtained, and total procedure duration. Pathological findings were classified as primary lung cancer, metastatic disease, benign disease, or non-diagnostic/inadequate specimens. Non-diagnostic or inadequate specimens were defined as those exhibiting non-specific benign findings, insufficient material for a specific diagnosis, or malignancy that was subsequently confirmed via additional biopsies [27].
Ethics approval
The Institutional Review Board (IRB) of Gil Medical Center, Gachon University College of Medicine, reviewed and approved the study protocol (IRB number: GCIRB2026-143). All methods and procedures were performed in accordance with the relevant guidelines and regulations, including the Declaration of Helsinki. The IRB officially exempted this investigation from the need to acquire written informed consent from the participants, referencing two principal considerations: the retrospective design of the research and the use of solely de-identified clinical data for analysis.
Statistical analysis
Continuous variables are presented as median values with interquartile ranges (IQR) and categorical variables are presented as counts and percentages. Baseline characteristics were compared between the groups using the Mann-Whitney U-test for continuous variables and either the Chi-squared test or Fisher’s exact test for categorical variables, depending on the data distribution. To identify predictive factors associated with prolonged hospitalization, we conducted multivariable analyses using logistic regression models with backward selection. Variables with clinical relevance or P < 0.10 in the univariable analysis were entered into the backward selection model. To determine the optimal cut-off value of DLCO and FEV1 as predictive factors for prolonged hospitalization related to post-biopsy pneumothorax, a Receiver Operating Characteristic (ROC) curve analysis was performed. The Area Under the Curve (AUC) was calculated to evaluate the overall diagnostic performance of the parameters. The optimal threshold was identified using Youden indices. All statistical tests were two-tailed, and P-values < 0.05 were considered statistically significant. All analyses were performed using IBM SPSS Statistics for Windows (v.27.0; IBM Corp., Armonk, NY).
Results
Baseline characteristics
Participant baseline characteristics are presented in Table 1. Among the 299 patients with post-biopsy pneumothorax, the median age was 72 years and 70.9% were male. The median FVC, FEV1, and DLCO were 3.36 L, 2.27 L, and 80%, respectively. The median target lesion size was 31 mm and the most common morphology was solid (90.0%). Additionally, patients with prolonged hospitalization were significantly older than those without prolonged hospitalization (75 vs. 70 years, P = 0.024). Pulmonary function was also significantly impaired in the prolonged hospitalization cohort, with a lower FEV1 (2.17 L vs. 2.34 L, P = 0.042) and a lower DLCO (71% vs. 83%, P = 0.005). The incidence of emphysema was significantly higher in patients with prolonged hospitalization (36.6% vs. 24.4%, P = 0.043).
Table 1.
Baseline characteristics
| Pneumothorax | Prolonged hospitalization due to Pneumothorax | P-value | ||
|---|---|---|---|---|
| Total | No | Yes | ||
| (N = 299) | (n = 217) | (n = 82) | ||
| Sex, male | 212 (70.9) | 150 (69.1) | 62 (75.6) | 0.319 |
| Age, years | 72 (65–78) | 70 (64–78) | 75 (69–78) | 0.024 |
| Ever-smoker | 208 (69.6) | 149 (68.7) | 59 (72.0) | 0.673 |
| BMI, kg/m2 | 23.5 (21.5–25.6) | 23.7 (21.6–25.8) | 23.4 (20.6–25.3) | 0.195 |
| Pulmonary function | ||||
| FVC, % | 89 (78–101) | 89 (81–101) | 86 (75–101) | 0.229 |
| FVC, L predicted | 3.36 (2.71–3.97) | 3.35 (2.70–3.96) | 3.41 (2.73–4.01) | 0.908 |
| FEV1, % | 89 (74–101) | 91 (78–101) | 83 (68–101) | 0.054 |
| FEV1, L predicted | 2.27 (1.90–2.80) | 2.34 (1.95–2.83) | 2.17 (1.64–2.51) | 0.042 |
| DLCO, % | 80 (64–99) | 83 (68–100) | 71 (52–94) | 0.005 |
| Underlying lung disease* | ||||
| Emphysema | 83 (27.8) | 53 (24.4) | 30 (36.6) | 0.043 |
| Interstitial lung disease | 37 (12.4) | 22 (10.1) | 15 (18.3) | 0.075 |
| Lesion location | ||||
| LUL | 59 (19.7) | 42 (19.3) | 17 (20.7) | 0.871 |
| LLL | 54 (18.1) | 41 (18.9) | 13 (15.9) | 0.615 |
| RUL | 84 (28.1) | 61 (28.1) | 23 (28.0) | > 0.999 |
| RML | 27 (9.0) | 19 (8.8) | 8 (9.8) | 0.822 |
| RLL | 75 (25.1) | 54 (24.9) | 21 (25.6) | 0.882 |
| Lesion diameter (mm) | 31 (21–45) | 30 (21–45) | 33 (23–43) | 0.614 |
| Lesion characteristics | ||||
| Solid | 269 (90.0) | 192 (88.5) | 77 (93.9) | 0.199 |
| Sub-solid | 16 (5.4) | 13 (6.0) | 3 (3.7) | 0.570 |
| Ground glass opacity | 14 (4.7) | 12 (5.5) | 2 (2.4) | 0.364 |
Values are presented as medians (interquartile range) or counts (percentages)
The values presented in bold indicate statistical significance (P < 0.05)
BMI body mass index; FVC forced vital capacity, FEV1 forced expiratory volume in 1 s, DLCO diffusing capacity for carbon monoxide
*Cases are duplicated
Procedural details and pathological results
Details of clinical procedures and diagnostic results are presented in Table 2. The median distance from the pleura was 18 mm, and lesion contact with the pleura and the trans-fissural route were noted in 16.4% and 9.0% of cases, respectively. Contact between the lesion and pleura was significantly less frequent in the prolonged hospitalization group than in the reference group (8.5% vs. 19.4%, P = 0.023). Other procedure-related factors showed no significant differences between the two groups. CT-guided TTNB revealed malignancy in 70.6% and non-diagnostic specimens in 13.4% of cases (Table 2). The frequency of malignant and non-diagnostic specimens did not differ significantly between the two groups.
Table 2.
Procedure and biopsy results
| Pneumothorax | Prolonged hospitalization due to Pneumothorax | P-value | ||
|---|---|---|---|---|
| Total | No | Yes | ||
| (N = 299) | (n = 217) | (n = 82) | ||
| Lesion contact with pleura | 49 (16.4) | 42 (19.4) | 7 (8.5) | 0.023 |
| Pre-biopsy skin to lesion distance (mm) | 54 (42–69) | 53 (40–69) | 55 (43–71) | 0.445 |
| Pre-biopsy skin to pleura distance (mm) | 34 (29–41) | 35 (29–42) | 33 (27–40) | 0.141 |
| Pre-biopsy pleura to lesion distance (mm) | 18 (8–34) | 17 (6–32) | 19 (10–36) | 0.096 |
| Patient position | ||||
| Supine position | 116 (38.8) | 83 (38.2) | 33 (40.2) | 0.791 |
| Prone position | 177 (59.2) | 131 (60.4) | 46 (56.1) | 0.512 |
| Lateral decubitus | 6 (2.0) | 3 (1.4) | 3 (3.7) | 0.351 |
| Noncoaxial approach | 88 (29.4) | 65 (30.0) | 23 (28.0) | 0.778 |
| Trans-fissural route | 27 (9.0) | 16 (7.4) | 11 (13.4) | 0.116 |
| Number of core needle biopsy specimens | 4 (4–6) | 4 (4–6) | 4 (4–6) | 0.177 |
| Total procedure time, min | 10 (8–15) | 10 (8–15) | 10 (8–15) | 0.766 |
| Length of stay, days | 3 (2–5) | 2 (2–3) | 5 (4–8) | < 0.001 |
| Biopsy results | ||||
| Malignancy | ||||
| Primary lung cancer | 207 (69.2) | 151 (69.6) | 56 (68.3) | 0.888 |
| Metastasis | 4 (1.3) | 2 (0.9) | 2 (2.4) | 0.303 |
| Benign disease | 48 (16.1) | 35 (16.1) | 13 (15.9) | > 0.999 |
| Non-diagnostic/Inadequate | 40 (13.4) | 29 (13.4) | 11 (13.4) | > 0.999 |
Values are presented as medians (interquartile range) or counts (percentages)
The values presented in bold indicate statistical significance (P < 0.05)
Clinical characteristics by intervention type
When the high-concentration oxygen therapy group was compared with the reference cohort without prolonged hospitalization (Supplementary Table 1), most baseline variables were similar. One statistically significant difference identified was diminished DLCO in the high-concentration oxygen therapy group (72% vs. 83%, P = 0.047). Age showed borderline significance (75 vs. 70 years, P = 0.050), whereas FEV1, lesion contact with the pleura, the trans-fissural route, and other lesion- or procedure-related variables were not significantly different.
By contrast, the PCD and chest tube groups displayed more pronounced differences than the control group (Supplementary Table 2). These patients had a lower BMI (22.2 kg/m² vs. 23.7 kg/m², P = 0.005), lower FVC (79% vs. 89%, P = 0.018), lower FEV1 (70% vs. 91%, P < 0.001), and lower DLCO (62% vs. 83%, P = 0.018). Emphysema was more common in the PCD and chest tube groups (50.0% vs. 24.4%, P = 0.007), whereas lesion contact with the pleura was less frequent (3.6% vs. 19.4%, P = 0.036). A direct comparison between patients receiving high-concentration oxygen therapy and those requiring pleural drainage or chest tube insertion is shown in Supplementary Table 3. Compared to the high-concentration oxygen therapy group, patients requiring invasive intervention showed lower BMI, FVC, and FEV1, whereas most other clinical and procedural variables were broadly similar. In the adjusted analysis, lower BMI and FVC were independently associated with the need for invasive intervention (Supplementary Table 4).
Predictive factors for overall prolonged hospitalization
In univariable analysis, lower FEV1 (odds ratio [OR], 0.986; 95% confidence interval [CI], 0.973–0.999; P = 0.029), lower DLCO (OR, 0.980; 95% CI, 0.967–0.994; P = 0.005), the presence of emphysema (OR, 1.785; 95% CI, 1.034–3.081; P = 0.037), and lesion contact with the pleura (OR, 0.389; 95% CI, 0.167–0.905; P = 0.028) were significantly associated with prolonged hospitalization (Table 3). In the multivariable analysis, lower DLCO remained the only significant independent predictor of prolonged hospitalization (adjusted OR [aOR], 0.980; 95% CI, 0.966–0.994; P = 0.004) (Table 3). ROC curve analysis for DLCO demonstrated limited discriminative ability with an AUC of 0.627. Based on Youden’s index, the optimal cut-off value for DLCO was identified as 73.5%. At this threshold, the model yielded a sensitivity of 56.36% and a specificity of 67.08% (Youden’s index = 0.2344) (Fig. 2a). Lesion contact with the pleura showed a protective trend, but was not statistically significant after adjustment (aOR, 0.386; 95% CI, 0.140–1.062; P = 0.065).
Table 3.
Predictive factors related to prolonged hospitalization due to pneumothorax (N = 299)
| Variables | Univariable analysis | Multivariable analysis | ||
|---|---|---|---|---|
| OR (95% CI) | P-value | aOR (95% CI) | P-value | |
| Sex, male | 1.385 (0.775–2.474) | 0.272 | ||
| Age, years | 1.025 (0.997–1.053) | 0.083 | ||
| BMI, kg/m2 | 0.948 (0.876–1.026) | 0.183 | ||
| Pulmonary function | ||||
| FVC, % | 0.990 (0.973–1.006) | 0.218 | ||
| FEV1, % | 0.986 (0.973–0.999) | 0.029 | ||
| DLCO, % | 0.980 (0.967–0.994) | 0.005 | 0.980 (0.966–0.994) | 0.004 |
| Emphysema | 1.785 (1.034–3.081) | 0.037 | ||
| Lesion contact with pleura | 0.389 (0.167–0.905) | 0.028 | 0.386 (0.140–1.062) | 0.065 |
| Pre-biopsy skin to lesion distance (mm) | 1.004 (0.992–1.016) | 0.488 | ||
| Trans-fissural route | 1.946 (0.862–4.392) | 0.109 | ||
The values presented in bold indicate statistical significance (P < 0.05)
OR odds ratio, aOR adjusted odds ratio, CI confidence interval, BMI body mass index, FVC forced vital capacity, FEV1 forced expiratory volume in 1 s, DLCO diffusing capacity for carbon monoxide
Fig. 2.
ROC curves of DLCO and FEV1 for predicting prolonged hospitalization and drainage requirements related to post-biopsy pneumothorax. a DLCO for overall prolonged hospitalization (AUC = 0.627). b DLCO for prolonged hospitalization in the subgroup requiring high-concentration oxygen therapy (AUC = 0.606). c FEV1 for the need for PCD or chest tube insertion (AUC = 0.752). AUC, area under the curve; DLCO, diffusing capacity for carbon monoxide; FEV1, forced expiratory volume in 1 s; PCD, percutaneous catheter drainage; ROC, receiver operating characteristic
Predictors specific to high oxygen therapy or drainage requirement
In the high-concentration oxygen therapy group (n = 54), lower DLCO was significantly associated with prolonged hospitalization in both univariable analysis (OR, 0.983; 95% CI, 0.967–0.999; P = 0.042) and multivariable analysis (adjusted OR, 0.974; 95% CI, 0.955–0.993; P = 0.007) (Table 4). ROC curve analysis for DLCO demonstrated limited discriminative ability with an AUC of 0.606. Based on Youden’s index, the optimal cut-off value for DLCO was identified as 73.5%. At this threshold, the model yielded a sensitivity of 55.56% and a specificity of 67.08% (Youden’s index = 0.2264) (Fig. 2b). In addition, the trans-fissural route was an independent predictor in multivariable analysis (aOR, 3.968; 95% CI, 1.157–13.607; P = 0.028). In the PCD or chest tube group (n = 28), several variables were significant in univariable analysis, including lower BMI (OR, 0.830; 95% CI, 0.730–0.945; P = 0.005), lower FVC (OR, 0.966; 95% CI, 0.944–0.989; P = 0.004), lower FEV1 (OR, 0.961; 95% CI, 0.942–0.980; P < 0.001), lower DLCO (OR, 0.973; 95% CI, 0.952–0.995; P = 0.017), and emphysema (OR, 3.094; 95% CI, 1.386–6.907; P = 0.006) (Table 5). However, in the multivariable analysis, a lower FEV1 remained the only independent predictor of drainage requirement (aOR, 0.961; 95% CI, 0.937–0.985; P = 0.002). ROC curve analysis for FEV1 demonstrated acceptable diagnostic accuracy with an AUC of 0.752. Based on Youden’s index, the optimal cut-off value of FEV1 was determined to be 73.5%. At this threshold, the parameter yielded a sensitivity of 65.38% and a specificity of 82.26% (Youden’s Index = 0.4764) (Fig. 2c).
Table 4.
Predictive factors related to prolonged hospitalization due to pneumothorax: high-concentration oxygen therapy (N = 54/271)
| Variables | Univariable analysis | Multivariable analysis | ||
|---|---|---|---|---|
| OR (95% CI) | P-value | aOR (95% CI) | P-value | |
| Sex, male | 1.161 (0.599–2.250) | 0.658 | ||
| Age, years | 1.023 (0.991–1.055) | 0.163 | ||
| BMI, kg/m2 | 1.012 (0.922–1.112) | 0.796 | ||
| Pulmonary function | ||||
| FVC, % | 1.008 (0.986–1.030) | 0.498 | 1.023 (0.997–1.050) | 0.084 |
| FEV1, % | 1.003 (0.986–1.020) | 0.748 | ||
| DLCO, % | 0.983 (0.967–0.999) | 0.042 | 0.974 (0.955–0.993) | 0.007 |
| Emphysema | 1.303 (0.673–2.524) | 0.433 | ||
| Lesion contact with pleura | 0.521 (0.209–1.298) | 0.161 | ||
| Pre-biopsy skin to lesion distance (mm) | 1.003 (0.989–1.017) | 0.677 | ||
| Trans-fissural route | 1.871 (0.728–4.805) | 0.193 | 3.968 (1.157–13.607) | 0.028 |
The values presented in bold indicate statistical significance (P < 0.05)
OR odds ratio, aOR adjusted odds ratio, CI confidence interval, BMI body mass index, FVC forced vital capacity, FEV1 forced expiratory volume in 1 s, DLCO diffusing capacity for carbon monoxide
Table 5.
Predictive factors related to prolonged hospitalization due to pneumothorax: PCD or chest tube insertion (N = 28/245)
| Variables | Univariable analysis | Multivariable analysis | ||
|---|---|---|---|---|
| OR (95% CI) | P-value | aOR (95% CI) | P-value | |
| Sex, male | 2.055 (0.749–5.636) | 0.162 | ||
| Age, years | 1.028 (0.984–1.074) | 0.217 | ||
| BMI, kg/m2 | 0.830 (0.730–0.945) | 0.005 | 0.863 (0.730–1.020) | 0.084 |
| Pulmonary function | ||||
| FVC, % | 0.966 (0.944–0.989) | 0.004 | ||
| FEV1, % | 0.961 (0.942–0.980) | < 0.001 | 0.961 (0.937–0.985) | 0.002 |
| DLCO, % | 0.973 (0.952–0.995) | 0.017 | ||
| Emphysema | 3.094 (1.386–6.907) | 0.006 | ||
| Lesion contact with pleura | 0.154 (0.020–1.168) | 0.070 | 0.149 (0.016–1.409) | 0.097 |
| Pre-biopsy skin to lesion distance (mm) | 1.007 (0.988–1.025) | 0.480 | ||
| Trans-fissural route | 2.094 (0.647–6.777) | 0.218 | ||
The values presented in bold indicate statistical significance (P < 0.05)
PCD percutaneous catheter drainage, OR odds ratio, aOR adjusted odds ratio, CI confidence interval, BMI body mass index, FVC forced vital capacity, FEV1 forced expiratory volume in 1 s, DLCO diffusing capacity for carbon monoxide
Discussion
The present study was designed to identify clinical and procedural factors associated with prolonged hospitalization among patients who developed pneumothorax after CT-guided TTNB. Three findings merit emphasis. First, a lower DLCO was an independent predictor of prolonged hospitalization in the study population and remained independently associated with the subgroup requiring high-concentration oxygen therapy. Second, lower FEV1 was an independent predictor of PCD or chest tube insertion. Third, the trans-fissural needle route independently increased the risk of a high oxygen requirement, whereas lesion contact with the pleura was less frequent in the drainage group and demonstrated a protective direction in regression analysis, although it did not remain an independent predictor after adjustment.
Differential roles of DLCO and FEV1 in patients requiring CT-guided TTNB are clinically plausible. DLCO reflects the gas-exchange reserve and functional integrity of the alveolar-capillary membrane; thus, a lower DLCO likely identifies patients with diminished alveolar-capillary reserve, who develop hypoxemia more readily in response to modest pleural air accumulation and transient lung volume loss. This may explain why DLCO tracked both overall prolonged hospitalization and the need for high-concentration oxygen therapy. However, it is important to note that the overall diagnostic accuracy of DLCO showed limited discriminative ability (AUC = 0.627), indicating that it should not be utilized as a strong standalone predictor. Instead, a reduced DLCO is more appropriately viewed as an adverse physiological marker that contributes to comprehensive risk stratification when evaluated alongside other clinical and technical parameters. By contrast, FEV1 more directly reflects expiratory flow limitation and ventilatory mechanics, which may become especially relevant when the pneumothorax progresses to a level that requires drainage. Patients with abnormal pulmonary function may tolerate pleural air poorly because they begin with a reduced physiological reserve, and resolution may be delayed by abnormal parenchymal mechanics. In emphysema, reduced elastic recoil and bullous destruction may impair spontaneous sealing of the visceral pleural defect, whereas in fibrotic lung abnormalities such as ILA or UIP, reduced parenchymal compliance may limit lung re-expansion, contributing to a more persistent clinical course. This interpretation is supported by previous work showing that abnormal pulmonary function, including reduced diffusion capacity, is associated with major TTNB-related complications [23, 24], as well as by recent studies linking fibrotic lung abnormalities, including ILA and UIP, with more severe post-biopsy complications [13, 14].
The effect of the trans-fissural route in our study is also pathophysiologically plausible. Crossing a fissure requires the violation of opposing visceral pleural surfaces and may create a longer or more unstable air-leak pathway than a purely intraparenchymal route. Such additional pleural injury may not always mandate immediate drainage, but it can plausibly increase the oxygen requirement by promoting persistent or progressive pleural air accumulation. This interpretation is concordant with prior literature showing that a trans-fissural trajectory is associated with prolonged chest tube requirement and an increased risk of pneumothorax or chest drainage after CT-guided lung biopsy [2, 9, 10, 15]. Conversely, contact of the lesion with the pleura likely lowers the severity of the pneumothorax because the needle traverses little or no normally aerated lung parenchyma. With reduced trans-pulmonary distance, the likelihood of enduring alveolar-pleural air leakage is diminished, rendering pleural-based lesions less prone to evolve into clinically significant pneumothorax. This is consistent with meta-analyses demonstrating that a lack of pleural contact increases the risk of pneumothorax, as well as recent predictive models identifying pleural attachment, fissure traversal, and tract-related factors as key determinants of post-biopsy pneumothorax [12, 15, 20–22]. In this study, lesion contact with the pleura was significantly less common in patients requiring drainage and showed a protective direction in regression analysis, supporting the same anatomical principle.
Most previous studies concerning TTNB-related pneumothorax have focused on predicting its occurrence or the need for chest drainage rather than the post-event clinical trajectory [5–12, 15, 19, 20, 23]. While these prior studies are valuable, they do not fully address the clinically relevant question of which patients will follow a more burdensome clinical trajectory after the development of a pneumothorax. The originality of the present study lies in shifting the focus from occurrence to post-event severity and in further stratifying severity according to the intervention level, namely high-concentration oxygen therapy versus drainage requirement. This interpretation was further supported by supplementary analyses showing that, among patients with prolonged hospitalization, those requiring pleural drainage or chest tube insertion had a lower BMI and more impaired pulmonary function than those managed with high-concentration oxygen therapy alone. From a practical standpoint, our findings suggest that pre-biopsy risk stratification should go beyond lesion size and anatomical location. Comprehensive pulmonary function assessments, particularly DLCO and FEV1, may facilitate the identification of patients who are predisposed to require prolonged observation, escalated oxygen therapy, or pleural drainage in the event of pneumothorax. Procedural planning should likewise minimize trans-fissural trajectories whenever an alternative route is available, and patients with low DLCO, low FEV1, emphysema, or fibrotic lung abnormalities may warrant closer post-biopsy monitoring even when the initial pneumothorax appears modest.
This study had certain limitations. First, its retrospective, single-center design introduces the possibility of selection bias and limited generalizability. Second, the subgroup requiring PCD or chest tube insertion was relatively small (n = 28), which may have reduced the statistical power of the multivariable analysis and may partly explain why some variables with clinically plausible effects, such as lesion contact with the pleura, did not exhibit statistical significance after adjustment. This modest sample size reflects the real-world safety profile of our tertiary center, where the majority of post-biopsy pneumothoraces were successfully stabilized via conservative means. Despite this limitation, the observed statistical significance of FEV1 suggests that it may serve as a clinically relevant marker associated with the need for invasive intervention. Third, management decisions for oxygen therapy or drainage are inevitably influenced by real-world clinical judgment and institutional practice patterns. A baseline presence of pre-existing respiratory impairment could have biased clinicians toward early, proactive invasive interventions to prevent sudden deterioration in patients with limited reserve, which may have partially influenced the strength of the observed associations. Fourth, regarding procedural characteristics, our institutional protocol mandates obtaining 4 to 6 core samples to ensure adequate tissue acquisition for advanced molecular profiling and predictive biomarker testing in suspected lung malignancies. Although a lower number of biopsies (e.g., 2–3 cores) might theoretically reduce pleural trauma, our analysis demonstrated that within the 4–6 core range, the exact number of specimens did not significantly affect the clinical course or prolong hospitalization. However, because a lower biopsy frequency cohort was not available for comparison due to our standardized protocol, further comparative studies are warranted to evaluate the minimum necessary core number that balances diagnostic yield with complication severity. Finally, because this study included only patients who had already developed post-biopsy pneumothorax, the findings should be interpreted as predictors of the clinical course after pneumothorax rather than of pneumothorax occurrence. The present study provides evidence that integrating pre-procedural physiological markers with anatomical and procedural features can refine risk stratification after TTNB. In particular, DLCO, FEV1, and needle pathway characteristics appear to be key determinants of whether post-biopsy pneumothorax remains self-limiting or progresses to a prolonged and intervention-requiring course. Incorporating these physiological and anatomical variables into pre-biopsy evaluations may ultimately support better patient management, more targeted post-procedural surveillance, and more efficient use of hospital resources [7, 9, 23].
Conclusion
This study identified clinical and procedural factors that predict prolonged hospitalization due to pneumothorax after CT-guided TTNB. Reduced DLCO was found to be a significant independent predictor of both extended length of stay and the need for high-concentration oxygen therapy. Furthermore, a lower FEV1 was closely associated with a higher likelihood of requiring invasive interventions such as PCD or chest tube insertion. The use of the trans-fissural route has also emerged as a critical procedural risk factor for prolonged hospitalization due to pneumothorax and the need for high-concentration oxygen therapy. Consequently, these findings suggest that comprehensive pre-biopsy pulmonary function assessment and careful needle-path planning may help identify patients at higher risk of clinically significant prolonged hospitalization due to post-biopsy pneumothorax.
Supplementary Information
Acknowledgements
Not applicable.
Abbreviations
- aOR
Adjusted odds ratio
- AUC
Area under the curve
- BMI
Body mass index
- CI
Confidence interval
- CT
Computed tomography
- CT-guided TTNB
Computed tomography-guided transthoracic needle biopsy
- DLCO
Diffusing capacity for carbon monoxide
- FEV1
Forced expiratory volume in 1 s
- FVC
Forced vital capacity
- ILA
Interstitial lung abnormality
- IQR
Interquartile range
- IRB
Institutional review board
- PCD
Percutaneous catheter drainage
- PFT
Pulmonary function test
- ROC
Receiver operating characteristic
- UIP
Usual interstitial pneumonia
Authors’ contributions
UWK and BS contributed to the conceptualization of the study. UWK, MH, SP, JHH, and BS were involved in data curation and investigation. UWK and BS performed the formal analysis. BS contributed to the methodology and provided supervision. UWK and BS wrote the original draft of the manuscript. All authors contributed to the review and editing of the manuscript. All authors read and approved the final manuscript.
Funding
Not applicable.
Data availability
The datasets used and analyzed during the current study are available from the corresponding author upon reasonable request.
Declarations
Ethics approval and consent to participate
The Institutional Review Board (IRB) of Gil Medical Center, Gachon University College of Medicine, reviewed and approved the study protocol (IRB number: GCIRB2026-143). All methods and procedures were performed in accordance with the relevant guidelines and regulations, including the Declaration of Helsinki. The IRB officially exempted this investigation from the need to acquire written informed consent from the participants, referencing two principal considerations: the retrospective design of the research and the use of solely de-identified clinical data for analysis. We would like to clarify that our study is a retrospective observational study that analyzed historical clinical data.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Ahrar K, How I, Do It. CT-guided Percutaneous Transthoracic Lung Biopsy. Radiology. 2025;317(1):e243593. [DOI] [PubMed] [Google Scholar]
- 2.Yoon J, Park S, Kim JH, Hwang JH, Park C, Park S. Percutaneous core needle biopsy of 566 peripheral lung lesions: analysis of factors associated with biopsy failure and postprocedural pneumothorax. Clin Radiol. 2025;85:106911. [DOI] [PubMed] [Google Scholar]
- 3.Constantinescu A, Stoicescu ER, Iacob R, Chira CA, Cocolea DM, Nicola AC, et al. CT-Guided Transthoracic Core-Needle Biopsy of Pulmonary Nodules: Current Practices, Efficacy, and Safety Considerations. J Clin Med. 2024;13:23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Lentz RJ, Frederick-Dyer K, Planz VB, Koyama T, Aboudara MC, Avasarala SK, et al. Navigational Bronchoscopy or Transthoracic Needle Biopsy for Lung Nodules. N Engl J Med. 2025;392(21):2100–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Yoon SH, Park CM, Lee KH, Lim KY, Suh YJ, Im DJ, et al. Analysis of Complications of Percutaneous Transthoracic Needle Biopsy Using CT-Guidance Modalities In a Multicenter Cohort of 10568 Biopsies. Korean J Radiol. 2019;20(2):323–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Vachani A, Zhou M, Ghosh S, Zhang S, Szapary P, Gaurav D, et al. Complications After Transthoracic Needle Biopsy of Pulmonary Nodules: A Population-Level Retrospective Cohort Analysis. J Am Coll Radiol. 2022;19(10):1121–9. [DOI] [PubMed] [Google Scholar]
- 7.Lim WH, Park CM, Yoon SH, Lim HJ, Hwang EJ, Lee JH, et al. Time-dependent analysis of incidence, risk factors and clinical significance of pneumothorax after percutaneous lung biopsy. Eur Radiol. 2018;28(3):1328–37. [DOI] [PubMed] [Google Scholar]
- 8.Hiraki T, Mimura H, Gobara H, Shibamoto K, Inoue D, Matsui Y, et al. Incidence of and risk factors for pneumothorax and chest tube placement after CT fluoroscopy-guided percutaneous lung biopsy: retrospective analysis of the procedures conducted over a 9-year period. AJR Am J Roentgenol. 2010;194(3):809–14. [DOI] [PubMed] [Google Scholar]
- 9.Moreland A, Novogrodsky E, Brody L, Durack J, Erinjeri J, Getrajdman G, et al. Pneumothorax with prolonged chest tube requirement after CT-guided percutaneous lung biopsy: incidence and risk factors. Eur Radiol. 2016;26(10):3483–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Ruud EA, Stavem K, Geitung JT, Borthne A, Soyseth V, Ashraf H. Predictors of pneumothorax and chest drainage after percutaneous CT-guided lung biopsy: A prospective study. Eur Radiol. 2021;31(6):4243–52. [DOI] [PubMed] [Google Scholar]
- 11.Heerink WJ, de Bock GH, de Jonge GJ, Groen HJ, Vliegenthart R, Oudkerk M. Complication rates of CT-guided transthoracic lung biopsy: meta-analysis. Eur Radiol. 2017;27(1):138–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Huo YR, Chan MV, Habib AR, Lui I, Ridley L. Pneumothorax rates in CT-Guided lung biopsies: a comprehensive systematic review and meta-analysis of risk factors. Br J Radiol. 2020;93(1108):20190866. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Balbi M, Righi L, Culasso NC, Capelli S, Caroli A, Impala GA, et al. CT-guided core needle biopsy of focal pulmonary lesions with coexisting interstitial lung abnormalities: a case-control study. Eur Radiol. 2026;36(6):4366–77. [DOI] [PMC free article] [PubMed]
- 14.Park J, Lee JH, Hong W, Hwang EJ, Yoon SH, Goo JM, et al. Usual Interstitial Pneumonia: Associations With Complications After Percutaneous Transthoracic Needle Lung Biopsy. AJR Am J Roentgenol. 2024;222(2):e2329938. [DOI] [PubMed] [Google Scholar]
- 15.Weon J, Robson S, Chan R, Ussher S. Development of a risk prediction model of pneumothorax in percutaneous computed tomography guided transthoracic needle lung biopsy. J Med Imaging Radiat Oncol. 2021;65(6):686–93. [DOI] [PubMed] [Google Scholar]
- 16.Chan MV, Afraz Z, Huo YR, Kandel S, Rogalla P. Manual aspiration of a pneumothorax after CT-guided lung biopsy: outcomes and risk factors. Br J Radiol. 2023;96(1148):20220366. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Elshafee AS, Karch A, Ringe KI, Shin HO, Raatschen HJ, Soliman NY, et al. Complications of CT-guided lung biopsy with a non-coaxial semi-automated 18 gauge biopsy system: Frequency, severity and risk factors. PLoS ONE. 2019;14(3):e0213990. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kim D, Han JY, Baek JW, Lee HY, Cho HJ, Heo YJ, et al. Effect of the respiratory motion of pulmonary nodules on CT-guided percutaneous transthoracic needle biopsy. Acta Radiol. 2023;64(7):2245–52. [DOI] [PubMed] [Google Scholar]
- 19.Zhao Y, Wang X, Wang Y, Zhu Z. Logistic regression analysis and a risk prediction model of pneumothorax after CT-guided needle biopsy. J Thorac Dis. 2017;9(11):4750–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Yang L, Liang T, Du Y, Guo C, Shang J, Pokharel S, et al. Nomogram model to predict pneumothorax after computed tomography-guided coaxial core needle lung biopsy. Eur J Radiol. 2021;140:109749. [DOI] [PubMed] [Google Scholar]
- 21.Wang S, Tu J, Chen W. Development and Validation of a Prediction Pneumothorax Model in CT-Guided Transthoracic Needle Biopsy for Solitary Pulmonary Nodule. Biomed Res Int. 2019;2019:7857310. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Qin Y, Zhao Z, Zhang L, Shao X, Wang Q. Clinical and Procedural Predictive Model for Pneumothorax Risk After CT-Guided Lung Biopsy. Ther Clin Risk Manag. 2026;22:579698. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Rothman A, Lim S, Hasegawa D, Steiger D, Patel R, Lee YI. Abnormal Pulmonary Function Testing as an Independent Risk Factor for Procedural Complications During Transthoracic Needle Biopsies. J Bronchol Interv Pulmonol. 2022;29(3):213–9. [DOI] [PubMed] [Google Scholar]
- 24.Ruud EA, Heck S, Stavem K, Soyseth V, Geitung JT, Ashraf H. Low diffusion capacity of the lung predicts pneumothorax and chest drainage after CT-guided lung biopsy. BMC Res Notes. 2022;15(1):353. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.O’Driscoll BR, Howard LS, Earis J, Mak V. BTS guideline for oxygen use in adults in healthcare and emergency settings. Thorax. 2017;72(Suppl 1):ii1–90. [DOI] [PubMed] [Google Scholar]
- 26.Bae K, Ha JY, Jeon KN. Pneumothorax after CT-guided transthoracic lung biopsy: A comparison between immediate and delayed occurrence. PLoS ONE. 2020;15(8):e0238107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Lim WH, Lee JH, Park H, Park CM, Yoon SH. Effect of smoking on the diagnostic results and complication rates of percutaneous transthoracic needle biopsy. Eur Radiol. 2024;34(10):6514–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets used and analyzed during the current study are available from the corresponding author upon reasonable request.