European Journal of Cardiovascular Medicine

The diagnosis and management of tracheobronchial airway diseases, including benign and malignant stenosis, endobronchial tumors, and tracheobronchomalacia, rely on accurate anatomical and pathological assessment. For decades, flexible fiberoptic bronchoscopy (FOB) has been the undisputed gold standard for evaluating the airway lumen [1]. Its ability to provide direct real-time visualization, assess mucosal characteristics, and, most critically, obtain tissue samples for histopathological diagnosis makes it an essential tool in respiratory medicine [2]. However, FOB is an invasive procedure associated with patient discomfort, the need for sedation, and potential complications such as laryngospasm, hemorrhage, pneumothorax, and adverse reactions to anesthesia, albeit with a low incidence [3]. Furthermore, its utility can be limited by an inability to pass the scope beyond high-grade stenoses, leaving the distal airway and parenchyma unexamined.

The advent and rapid evolution of multidetector computed tomography (MDCT) have revolutionized thoracic imaging. By acquiring volumetric, high-resolution, isotropic datasets of the chest, MDCT allows for sophisticated two-dimensional (2D) multiplanar reconstructions and three-dimensional (3D) post-processing [4]. One of the most significant applications of this technology is virtual bronchoscopy (VB), a 3D rendering technique that generates endoluminal "fly-through" images of the tracheobronchial tree, simulating the view obtained during conventional FOB [5]. As a non-invasive technique, VB avoids the risks associated with sedation and instrumentation, can be performed quickly, and provides a comprehensive roadmap of the entire airway, including segments distal to impassable obstructions and the relationship of the airway to adjacent mediastinal structures [6].

Numerous studies over the past two decades have compared VB with FOB, consistently demonstrating its high accuracy in detecting central airway stenosis and endoluminal masses [7, 8]. Some meta-analyses have reported pooled sensitivities and specificities exceeding 90% for the detection of malignant airway stenosis [9]. VB has also proven valuable in guiding interventional procedures, such as stent placement and tumor debulking, by providing precise measurements of lesion length, diameter, and distance from key anatomical landmarks like the carina [10].

Despite this growing body of evidence, a research gap persists. Many earlier studies were retrospective, utilized older-generation CT scanners with lower resolution, or focused on a single pathology, such as lung cancer staging. There is a need for prospective studies using state-of-the-art MDCT technology to evaluate the performance of VB across a mixed spectrum of benign and malignant airway diseases, directly comparing its quantitative and qualitative findings with FOB in the same patient cohort. Such an evaluation is critical to delineating the optimal role of VB in the modern diagnostic algorithm for patients with suspected airway disease.

Therefore, the aim of this study was to prospectively compare the diagnostic accuracy and quantitative assessment capabilities of 256-slice MDCT-VB with the gold standard FOB in patients referred for suspected central airway pathology. We hypothesized that VB would demonstrate high concordance with FOB for detecting and measuring significant structural lesions, proving its value as a non-invasive first-line or complementary diagnostic tool.

Study Design and Population
This was a prospective, single-center, comparative diagnostic accuracy study conducted at the Department of Pulmonology and the Department of Radiology of a tertiary university hospital between January 2023 and June 2024.

We enrolled consecutive adult patients (≥18 years) who were clinically referred for bronchoscopy due to symptoms (e.g., progressive dyspnea, stridor, persistent cough, hemoptysis) or chest radiography findings suggestive of a central airway (trachea to segmental bronchi) abnormality.

Inclusion Criteria:

1. Age ≥ 18 years.
2. Clinical and/or radiological suspicion of a central airway lesion.
3. Ability to provide informed consent.
4. Scheduled for elective fiberoptic bronchoscopy.

Exclusion Criteria:

1. Contraindication to MDCT (e.g., known severe allergy to iodinated contrast media).
2. Severe renal insufficiency (estimated glomerular filtration rate < 30 mL/min/1.73m²).
3. Pregnancy or lactation.
4. Hemodynamic instability or acute respiratory failure requiring emergency intervention.
5. Patient refusal.

MDCT and Virtual Bronchoscopy Protocol
All patients underwent MDCT on a 256-slice scanner (Revolution CT, GE Healthcare). Patients were instructed to hold their breath at full inspiration during the scan. The scan was performed from the thoracic inlet to the lung bases.

• Scan Parameters: Collimation: 256 × 0.625 mm; tube voltage: 100–120 kVp (with automated dose modulation); tube current: auto mA (Noise Index: 25); rotation time: 0.5 s; pitch: 0.98.
• Contrast Administration: For patients with suspected vascular involvement or tumors, 80 mL of non-ionic iodinated contrast medium (Iohexol, 350 mgI/mL) was injected intravenously at a rate of 3.5 mL/s, with scanning initiated using a bolus-tracking technique.
• Image Reconstruction: Axial images were reconstructed with a slice thickness of 0.625 mm and an interval of 0.5 mm using both standard soft-tissue and high-resolution lung algorithms.

VB images were generated from the source axial dataset on a dedicated post-processing workstation (Advantage Workstation 4.7, GE Healthcare) using commercially available software (Navigator). The process involved airway segmentation, centerline extraction, and generation of 3D endoluminal fly-through views. Two board-certified thoracic radiologists with over 10 years of experience, who were blinded to the patients' clinical history and FOB results, independently reviewed the VB and associated multiplanar reformatted images. Any discrepancies were resolved by consensus. They recorded the presence, location (by airway segment), morphology (stenosis, endoluminal mass, external compression, mucosal irregularity), and, for stenoses, the percentage of luminal reduction and longitudinal length.

Fiberoptic Bronchoscopy (FOB) Protocol
Within one week of the MDCT scan, all patients underwent FOB performed by one of two senior pulmonologists with over 15 years of experience, who were blinded to the detailed VB findings (but not the initial CT report indicating a need for bronchoscopy). Procedures were performed under conscious sedation using midazolam and fentanyl, with topical lidocaine anesthesia to the upper airway. A standard flexible video bronchoscope (Olympus BF-1T190) was used.

The entire tracheobronchial tree was systematically inspected down to the subsegmental level where possible. Findings were documented, including the location and nature of any lesions. For stenoses, the degree of obstruction was estimated visually as a percentage of the normal lumen, and the length was measured using markings on biopsy forceps or probes. Biopsies and/or brushings were taken from suspicious lesions for histopathological or cytological analysis. The findings of FOB, combined with histology where available, were considered the reference standard for this study.

Statistical Analysis
Data were analyzed using SPSS Statistics for Windows, Version 28.0 (IBM Corp., Armonk, NY). Patient demographics were summarized using descriptive statistics (mean ± standard deviation [SD] for continuous variables, and frequencies and percentages for categorical variables).

The diagnostic performance of VB was evaluated on a per-lesion basis, with FOB as the reference standard. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) with 95% confidence intervals (CIs) were calculated for the detection of airway stenosis, endobronchial masses, and external compression.

For the quantitative assessment of airway stenoses, the correlation between measurements of stenosis degree (%) and length (mm) obtained by VB and FOB was assessed using Pearson’s correlation coefficient (r). A paired samples t-test was used to compare the mean measurements between the two methods. A p-value of < 0.05 was considered statistically significant. Inter-observer agreement for VB interpretation between the two radiologists was assessed using Cohen’s kappa coefficient (κ).

Patient Characteristics
A total of 125 patients were initially screened for eligibility. Thirteen patients were excluded (7 refused consent, 4 had contraindications to contrast, 2 were medically unstable). The final study cohort comprised 112 patients, including 68 males (60.7%) and 44 females (39.3%), with a mean age of 58.4 ± 11.2 years (range: 22–81 years). The primary indications for bronchoscopy were stridor/dyspnea (42.0%), suspected lung malignancy (33.0%), hemoptysis (16.1%), and persistent atelectasis (8.9%). The demographic and clinical characteristics of the study population are summarized in Table 1.

Table 1. Demographic and Clinical Characteristics of the Study Population (N=112)

Diagnostic Performance of Virtual Bronchoscopy
FOB, the reference standard, identified a total of 128 significant lesions in 98 of the 112 patients. The remaining 14 patients had normal airway examinations. The lesions included 63 airway stenoses, 41 endobronchial masses, 15 cases of significant external compression, and 9 cases of diffuse mucosal irregularity/inflammation.

The diagnostic accuracy of VB for detecting these lesions is presented in Table 2. VB demonstrated excellent performance for detecting stenosis and endobronchial masses. It correctly identified 60 of 63 stenoses (false negatives = 3, all mild, <30% narrowing) and 37 of 41 endobronchial masses (false negatives = 4, all small <5mm or flat lesions). There were 2 false-positive findings for stenosis (mucus plugs mimicking stenosis) and 1 for an endobronchial mass. For external compression, VB identified all 15 cases and correctly identified the cause (e.g., lymph nodes, aortic aneurysm) by reviewing the accompanying axial images. VB's performance was notably lower for detecting subtle or diffuse mucosal changes, identifying only 5 of 9 such cases. The inter-observer agreement between the two radiologists for VB interpretation was excellent (κ = 0.91).

Table 2. Diagnostic Performance of MDCT-VB for Detecting Airway Lesions (Per-Lesion Analysis)

Quantitative Assessment of Airway Stenosis
Of the 63 stenoses identified by FOB, 60 were quantitatively assessed by both methods. Three were too mild (<20%) to be precisely measured by FOB. For the 60 comparable lesions, the measurements of stenosis degree and length showed strong concordance between VB and FOB. The mean degree of stenosis measured by VB was 61.5% ± 18.2%, compared to 60.1% ± 19.5% by FOB. The mean length of stenosis was 15.8 ± 7.1 mm by VB and 13.7 ± 6.8 mm by FOB.

As shown in Table 3, Pearson’s correlation analysis revealed a very strong positive correlation for both degree of stenosis (r=0.92, p<0.001) and length (r=0.89, p<0.001). A paired t-test showed no statistically significant difference in the measurement of stenosis degree (mean difference 1.4%, p=0.21), but the length as measured by VB was slightly longer on average, though this difference did not reach statistical significance (mean difference 2.1 mm, p=0.08).

Table 3. Correlation and Comparison of Quantitative Stenosis Measurements (n=60)

Additional Findings
A key advantage of VB was demonstrated in 11 patients with high-grade stenosis (>90% luminal narrowing). In all 11 cases, the bronchoscope could not be advanced past the obstruction. VB, however, provided clear visualization of the airway distal to the stenosis, identifying synchronous lesions in two patients (one with a distal tumor nodule, one with bronchiectasis) that were missed by the initial FOB. Complications related to FOB were minor and occurred in 5 patients (4.5%), including transient hypoxemia (n=3) and minor self-limiting bleeding after biopsy (n=2). No adverse events were associated with the MDCT or VB procedure.

This prospective study confirms the high diagnostic utility of 256-slice MDCT-VB as a non-invasive tool for the evaluation of central airway diseases. Our primary finding is that VB demonstrates excellent sensitivity and specificity for the detection of clinically significant lesions, such as airway stenosis and endobronchial masses, with performance metrics comparable to the invasive gold standard, FOB. This reinforces the growing consensus that VB can play a pivotal role in the diagnostic pathway of patients with suspected tracheobronchial pathology [7, 11].

Our reported sensitivity of 95.2% for airway stenosis and 90.2% for endobronchial masses aligns with, and in some cases exceeds, the results from previous studies and meta-analyses. For example, a meta-analysis by a large consortium reported a pooled sensitivity of 93% for detecting tracheobronchial stenosis [9]. Our slightly higher performance may be attributed to the use of a latest-generation 256-slice CT scanner, which provides superior spatial and temporal resolution, enabling more accurate airway segmentation and visualization of smaller lesions compared to older 4- or 16-slice systems used in earlier research [12]. The false-negative cases for VB in our cohort were exclusively small (<5 mm) or very flat, superficial lesions, a known limitation of the technology, as VB cannot assess mucosal color, vascularity, or subtle surface changes that are readily apparent on direct bronchoscopic visualization [13]. This underscores that while VB is excellent for structural abnormalities, FOB remains superior for detecting early mucosal disease.

A particularly significant finding of our study is the strong correlation in the quantitative assessment of stenoses. The concordance between VB and FOB for measuring both the percentage of luminal narrowing (r=0.92) and the length of the stenotic segment (r=0.89) is crucial for clinical practice. Accurate pre-procedural measurement is paramount for planning interventions like tracheal resection and anastomosis, laser therapy, or the selection of an appropriately sized airway stent [10, 14]. While FOB measurements can be subjective and operator-dependent, VB provides objective, reproducible data. The slight, non-significant overestimation of stenosis length by VB in our study has been noted previously and may relate to perilesional inflammation or the partial volume averaging effect at the lesion's margins on CT [8]. Regardless, this information provides an invaluable roadmap for the interventional pulmonologist or thoracic surgeon.

Perhaps the most compelling clinical advantage of VB highlighted by our results is its ability to "see beyond the obstruction." In 11 cases of high-grade stenosis, where FOB could not traverse the lesion, VB provided a complete anatomical survey of the distal airways. This capability is not merely academic; it directly impacts patient management by allowing for the detection of synchronous lesions, assessing the extent of post-obstructive pneumonia or bronchiectasis, and planning the full scope of an intervention [6]. This unique advantage can prevent therapeutic surprises and may obviate the need for a second, post-intervention bronchoscopy to evaluate the distal tree.

The clinical implications of our findings suggest a paradigm shift in the diagnostic algorithm. Rather than proceeding directly to invasive FOB in all cases, MDCT with VB can be employed as a primary diagnostic test after an abnormal chest radiograph. If VB is normal, a significant structural lesion is highly unlikely (NPV >95% in our study), and an invasive procedure might be avoided, especially in frail patients. If VB identifies a clear-cut pathology like a long-segment benign stenosis or extrinsic compression, it may provide sufficient information for surgical planning without a pre-operative diagnostic FOB. In cases of suspected malignancy, where tissue is required, VB serves as a complementary guide, directing the bronchoscopist to the precise location for biopsy and providing crucial information about extraluminal extent and nodal involvement from the source CT data [15].

This study has several strengths, including its prospective design, the inclusion of a heterogeneous patient population reflecting real-world clinical practice, the use of state-of-the-art MDCT technology, and the blinding of interpreters for both VB and FOB. However, we acknowledge certain limitations. First, it is a single-center study, which may limit the generalizability of our findings. Second, the "gold standard" of FOB itself has limitations, particularly in the subjective estimation of stenosis degree and the inability to pass severe obstructions. Third, the radiation dose associated with MDCT, although minimized with modern modulation techniques, remains a consideration, particularly in younger patients or those requiring serial scans. Finally, VB is unable to provide a histopathological diagnosis, which remains the exclusive and indispensable domain of FOB.

In conclusion, this prospective comparative study demonstrates that MDCT virtual bronchoscopy is a robust, accurate, and non-invasive imaging modality for the diagnosis and characterization of central airway diseases. It shows excellent concordance with fiberoptic bronchoscopy for detecting and quantifying airway stenosis and endobronchial masses. Its unique ability to evaluate the airway distal to high-grade obstructions and to provide a comprehensive 3D map for pre-procedural planning solidifies its role as a powerful complementary tool in modern respiratory medicine. While it cannot replace fiberoptic bronchoscopy for tissue acquisition or the assessment of subtle mucosal disease, integrating virtual bronchoscopy into the diagnostic algorithm can optimize patient care by reducing diagnostic uncertainty, guiding interventions, and potentially limiting the use of invasive procedures to cases where they are therapeutically or histologically essential.

1. Prakash UBS, Offord KP, Stubbs SE. Bronchoscopy in North America: the ACCP survey. Chest. 1991;100(6):1668-75. DOI: 10.1378/chest.100.6.1668.
2. Du Rand IA, Blaikley J, Booton R, et al. British Thoracic Society guideline for diagnostic flexible bronchoscopy in adults: accredited by NICE. Thorax. 2013;68 Suppl 1:i1-i44. DOI: 10.1136/thoraxjnl-2013-203618.
3. Facciolongo N, Patelli M, Gasparini S, et al. Incidence of complications in bronchoscopy. Multicentre prospective study of 20,986 bronchoscopies. Monaldi Arch Chest Dis. 2009;71(1):8-14. PMID: 19402484.
4. Boiselle PM, Ernst A. Recent advances in central airway imaging. Chest. 2002;121(5):1651-60. DOI: 10.1378/chest.121.5.1651.
5. Finkelstein SE, Summers RM, Nguyen DM, Stewart JH, Tretler C, Schrump DS. Virtual bronchoscopy for evaluation of malignant tumors of the thorax. J Thorac Cardiovasc Surg. 2002;123(5):967-72. DOI: 10.1067/mtc.2002.121404.
6. Hoppe H, Walder B, Sonnenschein M, Vock P, Dinkel HP. Multidetector CT virtual bronchoscopy to grade tracheobronchial stenoses. AJR Am J Roentgenol. 2002;178(5):1195-200. DOI: 10.2214/ajr.178.5.1781195.
7. Luo L, He J, Zhang J, Li Y, Ai S. Comparison of virtual bronchoscopy and fiberoptic bronchoscopy for diagnosis of tracheobronchial stenosis. Chin Med J (Engl). 2009;122(14):1639-43. PMID: 19715694.
8. Ferretti GR, Kocier M, Calvat S, et al. Clinical policy-making and routine practice for the use of virtual bronchoscopy. Br J Radiol. 2009;82(973):31-9. DOI: 10.1259/bjr/53549247.
9. Xiong M, Wu G, Ai H, et al. A meta-analysis of the accuracy of virtual bronchoscopy for the detection of tracheobronchial stenosis and malignancy. J Thorac Dis. 2015;7(6):1038-48. DOI: 10.3978/j.issn.2072-1439.2015.04.53.
10. Geltner C, Hohlrieder M, Pinter M, et al. Value of virtual bronchoscopy for planning of interventional bronchoscopy in the treatment of malignant central airway stenosis. J Bronchology Interv Pulmonol. 2011;18(3):218-24. DOI: 10.1097/LBR.0b013e31822765be.
11. Boiselle PM, Feller-Kopman D, Ashiku S, et al. Tracheobronchial stents: choosing the right one. J Bronchology Interv Pulmonol. 2009;16(2):106-16. DOI: 10.1097/LBR.0b013e3181a3d931.
12. Lawler LP, Fishman EK. Multi-detector row CT of the aorta and major arteries. Radiology. 2001;221(3):583-93. DOI: 10.1148/radiol.2213010313.
13. Lacasse Y, Martel S, Hébert A, et al. Accuracy of virtual bronchoscopy to detect endobronchial lesions. Ann Thorac Surg. 2004;77(5):1754-60. DOI: 10.1016/j.athoracsur.2003.10.021.
14. Murgu SD, Colt HG. Description of a classification system for benign central airway stenosis. J Bronchology Interv Pulmonol. 2013;20(1):36-44. DOI: 10.1097/LBR.0b013e31827e7135.
15. Choi YW, Jeon SC, Park CK, et al. The clinical role of virtual bronchoscopy in patients with central airway lesions. Chest. 2004;126(6):1798-804. DOI: 10.1378/chest.126.6.1798