From January 2019 to August 2020, a total of 685 RP-EBUS were performed at our institution. We performed a retrospective observational study of 109 consecutive patients (15.9%) who underwent RP-EBUS in bronchus sign negative PPLs at Yeungnam University Hospital (a 930-bed, university-affiliated, tertiary referral hospital in Daegu, South Korea). All EBUS visualization lesions (n = 68) were biopsied with RP-EBUS-TBLB. Of the 41 EBUS non-visualization lesions, 30 were definitively diagnosed on clinical course, chest CT, bronchoalveolar lavage fluid, microbiological results, percutaneous transthoracic needle biopsy, EBUS-transbronchial needle aspiration, or surgical resection. Lung lesion that is neither benign nor malignant was defined as indefinite. From among all patients, 17 were without a definitive diagnosis (Fig. 1).

All patients underwent thin-section chest CT (0.75 mm slice thickness at intervals of 0.75 mm; SOMATOM Definition AS 64-slice CT system; Siemens Healthcare, Erlangen, Germany) less than 7 days ahead of EBUS-GS. Three experienced pulmonologists (Hong KS, Jang JG, Ahn JH) reviewed the chest CT images together before the procedure and planned a bronchial route to reach the target lesion without using a virtual bronchoscopic navigation (VBN) system.8 In all cases, we identified a target bronchus to approach the target lesion. The bronchus sign on CT was defined as the presence of a bronchus leading to the target lesion. Referring to previous literature, we defined as bronchus sign negative in case of type B or type C. In type A, the bronchus clearly approach inside of the target lesion. In type C, no bronchus detected in relation to the target lesion. Neither A nor C, it was categorized as type B.9 Three pulmonologists independently reviewed chest CT scan and determined the type of CT bronchus sign. In case of disagreement, it was decided through discussion. Of the 109 PPLs, 90 were agreed upon by three reviewers and 19 were disagreed upon and type of CT bronchus sign was decided through discussion. Using the concept of the bronchovascular bundle, an anatomical structure of the normal lung, the nearest bronchus was identified using the pathways of the pulmonary vessels travelling with the bronchus that entered the target PPL. We defined EBUS visualization as an image on EBUS with the probe within or adjacent to the lesion and non-visualization as no visible image of the lesion on EBUS. The distance from the lesion to the pleura was measured as the shortest distance on an axial plane CT scan, as in a previous study.10

All bronchoscopy procedures were performed by three pulmonologists, each with more than 5 years of experience in respiratory medicine. A 4 mm bronchoscope (BF P260F; Olympus, Tokyo, Japan) was used to reach the bronchus nearest the target lesion. Then an RP-EBUS (UM S20-17S; Olympus) inside a GS was inserted through the bronchoscope working channel. After the PPL was found, the RP was withdrawn, leaving the GS in place. Then a bronchial brush and biopsy forceps were introduced into the GS, and brushings and biopsy specimens were collected. If the PPL was not found, the examination was completed without a bronchial brush or biopsy. X-ray fluoroscopy was not used.

Fig. 2 shows representative cases of RP-EBUS-TBLB with GS in bronchus sign negative PPLs.

Continuous variables are expressed as means ± standard deviations and were compared with Student's t test or the Mann–Whitney U test. Categorical variables are described as frequencies (percentages) and were compared with the χ² test or Fisher's exact test. Fleiss' kappa was calculated to assess interobserver agreement of CT bronchus sign. We calculated the EBUS visualization yield by dividing the number of EBUS visualization cases by the total number of cases. We calculated the diagnostic yield by dividing the number of diagnostic successes by the total number of cases. To determine the factors for diagnostic yield, we divided the study population into two groups: a diagnostic success group (true-positive and true-negative results) and a diagnostic failure group (false-positive, and false-negative results). EBUS non-visualization and indefinite final diagnosis cases were also considered as false negative. Univariable and multivariable (using the factors with P < 0.1 in univariable analyses) logistic regression analyses were performed to identify the factors affecting the EBUS visualization and diagnostic yields. In all analyses, P < 0.05 under a two-tailed test was considered statistically significant. All statistical procedures were performed with SPSS (version 24.0; IBM, Chicago, IL, USA).

This study was conducted in accordance with the tenets of the Declaration of Helsinki, and its protocol was reviewed and approved by the institutional review board of Yeungnam University Hospital (YUH IRB 2020-09-025). The requirement for informed consent was waived because of the retrospective study design.

The baseline characteristics of the 109 patients according to EBUS visualization status are presented in Table 1. Their mean age was 66.8 ± 12.0 years, and 76 patients were men (69.7%). A total of 53 patients (48.6%) had underlying lung disease, including chronic obstructive pulmonary disease, asthma, idiopathic pulmonary fibrosis, combined pulmonary fibrosis and emphysema, old pulmonary tuberculosis, and pneumoconiosis. The most common location of the lung lesion was the left upper lobe (n = 31, 28.4%), and the mean distance from the pleura to the lung lesion was 12.9 ± 13.5 mm. The PPLs were significantly larger in the EBUS visualization group than in the EBUS non-visualization group (25.3 ± 11.5 mm vs. 19.3 ± 10.4 mm, P = 0.007). The lung lesions were classified as solid (n = 88, 80.7%), part-solid (n = 13, 11.9%), and ground-glass opacity (n = 8, 7.3%). Procedure times were significantly longer in the visualization group (18.2 ± 7.4 min vs. 10.2 ± 5.2 min, P < 0.001).

Table 2 shows the interobserver agreement of CT bronchus sign, and Fleiss' Kappa showed moderate agreement (0.559).

The initial pathologic results and final diagnosis are listed in Table 3. Of the 109 lung lesions, 53 (48.6%) were diagnosed as malignant, 39 (35.8%) as benign, and 17 (15.6%) as not definitive. The 41 lesions with biopsy failure due to EBUS non-visualization were ultimately diagnosed as malignant (n = 17), benign (n = 13), and not definitive (n = 11). Of the malignant lesions (n = 53), 29 (54.7%) were diagnosed by RP-EBUS-TBLB. Lesions with false-negatives (n = 7) and EBUS non-visualization (n = 17) were diagnosed as malignant by percutaneous transthoracic needle biopsy (n = 14), surgical resection (n = 6), EBUS transbronchial needle aspiration (n = 3), and bone biopsy (n = 1). Adenocarcinoma of the lung (n = 26, 49.1%) was the most common diagnosis among definitive malignant diagnoses. Among the benign lesions (n = 39), pneumonia (n = 16, 41.0%) was the most common diagnosis among definitive benign features. The diagnostic performance of RP-EBUS-TBLB in bronchus sign negative PPLs is summarized in Table 4. The sensitivity, specificity, positive predictive value, negative predictive value, and diagnostic accuracy of RP-EBUS-TBLB were 34.9% (29/83), 100% (26/26), 100% (29/29), 32.5% (26/80), and 50.5% (55/109), respectively. The diagnostic performance of RP-EBUS-TBLB was better in PPLs ≥ 20 mm.

In the patients with RP-EBUS performed in bronchus sign negative PPLs, we investigated the factors affecting EBUS visualization (Table 5). Univariable analyses revealed that larger size (≥ 20 mm; odds ratio [OR], 1.47; 95% CI, 1.08–2.02; P = 0.011), distance from the pleura > 10 mm (OR, 1.40; 95% CI, 1.05–1.87; P = 0.023), and a solid lesion (OR, 1.79; 95% CI, 1.02–3.15; P = 0.011) were significantly associated with EBUS visualization. In the multivariable analyses, larger size (≥ 20 mm; OR, 2.62; 95% CI, 1.16–5.93; P = 0.021) and distance from the pleura > 10 mm (OR, 2.37; 95% CI, 1.02–5.52; P = 0.045) were independent factors affecting EBUS visualization.

In patients with definitive final diagnoses, we investigated the factors affecting diagnostic success (Table 6). Univariable analyses revealed that larger size (≥ 20 mm; OR, 2.98; 95% CI, 1.37–6.49; P = 0.005), a solid lesion (OR, 4.21; 95% CI, 1.42–12.51; P = 0.007), and an EBUS image with the probe within the lesion (OR, 12.66; 95% CI, 4.37–36.66; P < 0.001) were significantly associated with diagnostic success. In the multivariable analyses, a solid lesion (OR, 14.58; 95% CI, 2.64–80.38; P = 0.002) and an EBUS image with the probe within the lesion (OR, 28.50; 95% CI, 6.26–129.85; P < 0.001) were independent factors affecting diagnostic success.

Among the total 109 patients, four cases of pneumothorax (3.7%) occurred, but all patients recovered after oxygen therapy and close observation. Hemoptysis occurred in one case (0.9%) and resolved spontaneously.