We conducted a retrospective analysis of 211 patients who underwent bronchoscopy at Henan Provincial People’s Hospital between July 2018 and August 2025. Eligible patients met the following criteria: (1) chest computed tomography (CT) showing mediastinal or hilar lymph node enlargement with a short-axis diameter ≥ 1 cm within the puncture range of EBUS-TBNA, or positron emission tomography (PET)-CT demonstrating abnormally increased lymph node metabolism requiring further diagnostic clarification; (2) no contraindications to bronchoscopy. Exclusion criteria were as follows: (1) severe cardiovascular disease, severe psychiatric disorder, active massive hemoptysis, or anesthesia intolerance contraindicating bronchoscopy; (2) coagulation dysfunction (platelet count < 20 × 10⁹/L), considered a relative contraindication for bronchoscopy; (3) intraoperative ultrasound revealing lesions adjacent to abundant vasculature or major vessels, where biopsy was deemed high-risk for bleeding.

All patients fasted for at least 6 h prior to the procedure and received intravenous anesthesia. The anesthetic regimen included sufentanil (0.2 μg/kg), propofol (1.5–2.0 mg/kg), and rocuronium bromide (0.3 mg/kg) to ensure adequate analgesia, sedation, and muscle relaxation. Continuous electrocardiographic monitoring was performed throughout the procedure, and an artificial airway was established using a laryngeal mask, followed by mechanical ventilation through a T-joint extension tube. Initially, a conventional white-light bronchoscopy was performed to conduct a comprehensive examination of the airways. When endobronchial lesions were identified, bronchoalveolar lavage (BAL) and/or transbronchial lung biopsy (TBLB) were performed as clinically indicated. BAL was preferred when infectious or granulomatous disease was suspected, as it enabled microbiological evaluation and facilitated exclusion of infectious etiologies. TBLB was generally performed when mediastinal lymphadenopathy coexisted with parenchymal lung lesions, enabling concurrent sampling of lung parenchyma for a more comprehensive clinical assessment. The decision to perform BAL and/or TBLB was made by the bronchoscopist based on real-time bronchoscopic findings and clinical indication. Subsequently, an ultrasound bronchoscope (BF-UC260F-OL8 or BF-UC260FW; Olympus, Tokyo, Japan) was introduced. Under real-time ultrasound guidance, the lobar to segmental bronchi were examined sequentially, and mediastinal and hilar lymph nodes were evaluated according to the American Thoracic Society (ATS) lymph node map, in conjunction with preoperative imaging findings. The EBUS Doppler function was employed to detect perinodal blood flow, and lymph nodes with favorable safety profiles and accessibility were preferentially selected as puncture targets.

EBUS-TBNA was performed under general anesthesia using a 21G puncture needle (Olympus, Tokyo, Japan). Under real-time ultrasound guidance, the target lymph nodes were punctured, and after confirming that the needle tip had entered the lesion, the needle was moved back and forth while applying negative pressure to obtain tissue cores (Figure 1A). Each lymph node was punctured at least four times and no more than seven times. Rapid on-site cytological evaluation (ROSE) was performed, with smear results independently assessed by a pathologist. In parallel, tissue cores were fixed in formalin and submitted for histological examination.

During EBUS-TBFB or EBUS-TBCB procedures, 1.5-mm biopsy forceps or 1.1-mm cryoprobes were introduced into the target lymph node via the tract established by EBUS-TBNA, with tissue samples obtained under real-time ultrasound guidance. In our center, the choice between TBCB and TBFB during EBUS-guided procedures followed predefined procedural criteria. Rapid on-site evaluation (ROSE) was routinely performed during EBUS-TBNA. When ROSE indicated malignancy or granulomatous inflammation but the histological specimen was insufficient, particularly when additional tissue was required to enable molecular testing, an additional biopsy using either TBCB or TBFB was undertaken. TBCB was performed when the target lymph node or lesion was adjacent to the airway wall, appeared firm on EBUS, or required a larger, intact tissue core for histopathological or molecular analysis (8). TBFB was selected for lesions near major vessels, for softer lesions on ultrasound, or when cryoprobe deployment was deemed unsafe due to anatomical constraints (9). The final decision was made by the bronchoscopist based on real-time EBUS findings and procedural safety considerations. For each lymph node station, 1–2 tissues were obtained for TBCB, and 2–3 forceps biopsy samples were taken for TBFB, depending on lesion characteristics and procedural factors (10, 11). The number of biopsies was based on tissue adequacy, lesion accessibility, and real-time safety assessment. Previous studies have reported that in approximately 28% of cases, biopsy instruments could not be advanced into the lymph node, most commonly due to limited penetration of the airway wall or nodal capsule (4, 12). In such cases, high-frequency electrocautery or laser-assisted window opening was employed to enlarge the entry site, thereby facilitating smooth passage of instruments into the target node (Figures 1B–H). This approach enabled the acquisition of larger and more structurally intact specimens, suitable for molecular pathology and immunohistochemical analysis (13).

After biopsy completion, a conventional bronchoscope was reintroduced to inspect the puncture sites and surrounding tissues for active bleeding. If bleeding was observed, local hemostatic measures were applied, including topical thrombin, iced saline lavage, or diluted norepinephrine.

Benign and malignant tumors (including primary and metastatic) were diagnosed based on histopathological findings, which is considered the gold standard. Tissue specimens were obtained by bronchoscopic biopsy (e.g., EBUS-TBNA, EBUS-TBCB, or EBUS-TBFB, TBLB), surgical procedures, or other invasive techniques, and all cases were confirmed by the pathology department. The diagnostic criteria for lymph node tuberculosis included: (1) histopathological evidence, such as granulomatous inflammation with caseous necrosis or the presence of acid-fast bacilli; or (2) positive results from pathogen detection, including TB-DNA PCR, Xpert MTB/RIF, or metagenomic next-generation sequencing (mNGS). The diagnostic criteria for pulmonary sarcoidosis were: (1) histopathological evidence of non-caseating epithelioid granulomatous inflammation; (2) compatible clinical presentation and radiological findings; (3) exclusion of other granulomatous diseases such as tuberculosis or fungal infections.

Specimens obtained by the combined EBUS-TBNA and EBUS-TBCB or EBUS-TBFB approach were considered positive when malignant cells were detected and negative when no malignant cells were present. Diagnostic outcomes were further validated through subsequent clinical follow-up. Definitions were as follows: True positive (TP): cases in which malignancy was detected by the combined approach and subsequently confirmed by surgical or pathological evaluation, or by clinical follow-up. True negative (TN): cases in which a benign result from the combined approach was confirmed by surgical or pathological evaluation, or by clinical follow-up. False negative (FN): cases in which a benign result from the combined approach was later confirmed to be malignant by additional examination or clinical follow-up.

During and after EBUS-TBNA combined with EBUS-TBCB or EBUS-TBFB procedures, various adverse events or complications may arise from the procedure or patient-related factors. The primary complications are bleeding, infection, and pneumothorax. Complications related to EBUS-TBCB or EBUS-TBFB are similar to those of EBUS-TBNA, and these procedures are generally considered safe. The incidence and mortality rates are significantly lower than those of mediastinoscopy (14). The common complications and incidence rates reported in existing literature are as follows: cough approximately 14%, bleeding 10–14%, hemoptysis 2%, pneumothorax 1%, mediastinal emphysema 1%, and dyspnea 1% (15, 16). No cases of mediastinal infection were observed. Both mediastinal emphysema and pneumothorax typically resolved spontaneously without intervention. Salcedo Lobera et al. (17) reported that in their study of 50 patients, 2 cases developed hypoxemia and 1 case developed vocal cord hematoma. Symptoms improved after symptomatic treatment.

The grading of bleeding is based on the standards established by Ernst et al. and Yarmus et al. (18, 19), as follows: Level 0: No suctioning required; Level 1: Requires suctioning and bronchoscopic wedging for ≤ 2 min; Level 2: Requires bronchoscopic wedging for ≥ 3 min; Level 3: Requires local instillation of epinephrine or iced saline; Level 4: Requires hemodynamic support, including transfusion of blood products, selective main bronchial intubation, placement of bronchial occluders, hospital admission, or surgical intervention.

In clinical practice, most bleeding cases are mild (grades 0–1) and do not require special treatment. Bleeding usually resolves spontaneously with bronchoscopic suctioning or can be controlled by instillation of cold normal saline. If necessary, diluted norepinephrine solution can be applied topically for hemostasis.

Statistical analyses were conducted using IBM SPSS Statistics version 26.0 (IBM, Armonk, NY, United States). With the final pathological diagnosis or clinical follow-up as the reference standard, the sensitivity, specificity, positive predictive value, negative predictive value, and overall diagnostic accuracy of the combined EBUS-TBNA and EBUS-TBCB or EBUS-TBFB approach were calculated. Continuous variables with normal distribution are presented as mean ± standard deviation (mean ± SD) and compared using independent-sample t-tests. Non-normally distributed continuous variables are expressed as median and interquartile range (median [Q1, Q3]) and compared using the Wilcoxon rank-sum test. Categorical variables are presented as frequencies and percentages, and differences between groups were analyzed using chi-square tests. All statistical tests were two-sided, and a p-value < 0.05 was considered statistically significant.

A total of 211 patients were included in this study, comprising 133 men (63.0%) and 78 women (37.0%), with a mean age of 59.5 ± 9.1 years (range, 27–82 years). Ninety-six patients (45.5%) had a history of smoking. The main presenting symptoms were cough (n = 106), chest tightness (n = 49), fever (n = 24), chest pain (n = 18), hemoptysis (n = 17), and abdominal pain (n = 3). Comorbid conditions included hypertension (n = 73), diabetes (n = 27), coronary heart disease (n = 21), cerebral infarction (n = 16), chronic obstructive pulmonary disease (COPD) (n = 4), and asthma (n = 4).

A total of 569 lymph node biopsies were performed on 211 patients, with station 7 lymph nodes as the most commonly sampled site, followed by station 4R lymph nodes (Table 1).

A total of 211 patients underwent combined EBUS-TBNA with either EBUS-TBCB or EBUS-TBFB. All biopsy specimens were systematically evaluated for their suitability for molecular testing on the basis of tissue quantity and quality. Molecular analyses were performed according to clinical indications, specimen integrity, and patient-specific factors. For benign diseases, mycobacterial DNA testing and acid-fast staining were performed, particularly in cases with granulomatous inflammation. For malignant diseases, immunohistochemistry (IHC) was initially conducted based on the initial histopathological assessment to determine tumor type and origin. Subsequently, next-generation sequencing (NGS) and PD-L1 IHC were performed on lung cancer specimens when clinically indicated.

This study included 211 patients, all of whom received definitive diagnoses. Among them, 130 cases were malignant lymph node lesions, comprising 124 cases of lung cancer (119 primary and 5 metastatic) and 6 cases of lymphoma. The remaining 81 cases were benign lymph node lesions, including 53 cases of pulmonary sarcoidosis, 22 cases of lymphadenitis, and 6 cases of lymph node tuberculosis.

A total of 204 patients were successfully diagnosed using EBUS-TBNA combined with additional sampling techniques. Among them, 106 patients underwent EBUS-TBNA plus EBUS-TBCB, 65 underwent EBUS-TBA plus EBUS-TBFB, and 3 underwent both EBUS-TBCB and EBUS-TBFB in addition to EBUS-TBNA. In addition, 30 patients underwent EBUS-guided electrocautery- or laser-assisted window biopsy.

Among the 204 diagnosed patients, 121 had lung cancer, including 55 adenocarcinomas, 35 small cell carcinomas, and 24 squamous cell carcinomas. Additionally, there were 2 cases of cervical squamous cell carcinoma metastasizing to the lung, 1 case each of renal clear cell carcinoma, rectal cancer, and cardia gastric fundus adenocarcinoma metastasis to the lung, as well as 1 case of neuroendocrine tumor and 1 case of SMARCA4-deficient undifferentiated carcinoma. Other diagnoses included 50 cases of pulmonary sarcoidosis, 22 cases of lymphadenitis, 6 cases of lymph node tuberculosis, and 5 cases of lymphoma (Table 2).

Among 53 patients with pulmonary sarcoidosis, 50 were confirmed by histopathology to have non-caseating epithelioid granulomatous inflammation. Combined with clinical manifestations and imaging findings (e.g., bilateral hilar lymph node enlargement), the diagnosis was established after excluding alternative causes such as tuberculosis and fungal infection. The remaining 3 cases did not yield typical pathological specimens but were diagnosed based on clinical presentation, imaging features, and response to empirical glucocorticoid therapy during follow-up. Twenty two patients with lymphadenitis were diagnosed with benign nonspecific changes on histopathology, and the final diagnoses were confirmed through clinical manifestations, imaging, and response to antibiotic therapy. Six patients with lymph node tuberculosis were diagnosed by histopathology (necrotizing granulomatous inflammation or positive acid-fast staining) and/or microbiological testing (TB-DNA PCR, Xpert MTB/RIF, or mNGS). There were four false-negative cases that were ultimately confirmed as malignant by other methods: two squamous cell carcinomas identified by follow-up bronchoscopy biopsy, one adenocarcinoma confirmed by surgery, and one lymphoma diagnosed by ultrasound-guided cervical lymph node biopsy (Table 3).

To further evaluate the diagnostic performance of EBUS-guided sampling, we assessed the accuracy of ROSE on TBNA specimens. A total of 10 false-negative ROSE cases were identified. Among these, four cases showed negative ROSE smears and negative histological findings from specimens obtained using the combined EBUS-TBNA with EBUS-TBCB or EBUS-TBFB approach. No atypical nuclei were observed on the ROSE smears, but these cases were ultimately diagnosed as malignant based on additional diagnostic procedures. The remaining six cases yielded definitive histological diagnoses using the combined sampling approach, including two lymphomas, two small cell lung cancers, one squamous cell carcinoma, and one adenocarcinoma. For malignant cases, ROSE demonstrated a sensitivity of 92.31% (120/130), a specificity of 100% (81/81), and an overall diagnostic accuracy of 95.26% (201/211), which demonstrated excellent concordance with the final cytopathological or histopathological diagnoses (Table 4).

In this study, EBUS-TBCB and EBUS-TBFB were categorized as generalized lymph node biopsy techniques, which include EBUS-guided electrocautery- or laser-assisted window biopsy, EBUS-TBCB, EBUS-TBFB, or their combination. A total of 108 cases were performed with EBUS-TBNA combined with EBUS-TBCB, 68 with EBUS-TBNA combined with EBUS-TBFB, 32 with EBUS-guided electrocautery or laser-assisted window biopsy, and 3 with EBUS-TBNA combined with both EBUS-TBCB and EBUS-TBFB. Due to the small sample size and lack of statistical power in the electrocautery/laser-assisted window biopsy group and the combined EBUS-TBCB plus TBFB group, intergroup comparisons were not conducted. Nevertheless, the overall diagnostic yield in the electrocautery/laser-assisted window biopsy group was 93.75% (30/32), and all 3 cases in the combined EBUS-TBCB and TBFB group were successfully diagnosed. No significant difference in diagnostic yield was observed between the EBUS-TBNA plus EBUS-TBCB group and the EBUS-TBNA plus EBUS-TBFB group. However, for pulmonary sarcoidosis, the diagnostic yield of EBUS-TBNA combined with EBUS-TBCB (96.67%) was higher than that of EBUS-TBNA combined with EBUS-TBFB (87.50%) (Table 5).

In this study, no severe complications, including massive bleeding, severe hypoxemia, respiratory distress, or mediastinal infection, were observed with EBUS-TBNA combined with EBUS-TBCB or EBUS-TBFB. Minor bleeding events were effectively controlled under bronchoscopy. Specifically, three patients experienced grade 3 bleeding at the puncture site, which was successfully managed with iced saline irrigation and topical application of diluted norepinephrine, with no recurrence. Ten patients developed mild hemoptysis, presumed to result from residual airway blood, which resolved spontaneously without systemic hemostatic therapy. Two patients reported chest pain, considered a local symptom related to the biopsy site; symptoms resolved spontaneously without medication. Three patients developed transient low-grade fever (<38.5 °C), which normalized after supportive treatment, without recurrence. Four patients experienced transient postoperative cough, likely associated with general anesthesia and bronchoscope passage through the glottis, which subsided spontaneously. One patient developed mediastinal emphysema, possibly due to tissue adhesion and tearing during cryoprobe withdrawal, leading to a transmural airway wall defect and air leakage into the mediastinal connective tissue. The patient remained stable and recovered with observation and oxygen therapy, without respiratory distress or chest tightness. Overall, EBUS-TBNA combined with EBUS-TBCB or EBUS-TBFB demonstrated a favorable safety profile, with only mild, manageable adverse events observed (Table 6).