Abstract
Novel imaging modalities are currently available for the noninvasive evaluation of the tracheobronchial tree. This study was undertaken to compare the diagnostic potentials of conventional CT scanning, super high-resolution CT (SHR-CT) scanning, and virtual bronchoscopy (VB) directly with fiberoptic bronchoscopy (FB) for the detection of tracheobronchial neoplasms.
Prospective observer study, in which 44 consecutive patients with thoracic malignancies were evaluated using several diagnostic imaging modalities. Images of the thorax were interpreted by individuals blind to the results of FB for the detection of endoluminal, obstructive, or mucosal lesions.
Image acquisition and simulation of the tracheobronchial anatomy were created successfully in all patients. Thirty-two patients who underwent both SHR-CT scanning and VB had correlative FBs within 1 month. In all nine patients who had a normal anatomy, SHR-CT scanning and VB accurately correlated with the FB findings. However, CT scanning demonstrated two false-positive obstructive lesions in one patient. Twenty-three patients had a total of 35 abnormal FB findings. The sensitivities of SHR-CT scanning and VB for the detection of endoluminal, obstructive, and mucosal lesions were 90%, 100%, and 16%, respectively. The overall sensitivities and specificities of SHR-CT scanning and VB were 83% and 100%, respectively. In contrast, CT scanning had sensitivities of 50%, 72%, and 0% for the detection of endoluminal, obstructive, and mucosal lesions with an overall sensitivity and specificity of 59%, and 85%, respectively. There was no case in which conventional CT scanning was better at detecting lesions than either SHR-CT scanning or VB.
SHR-CT scanning and VB are accurate, noninvasive methods for identifying obstructions and endoluminal lesions within the respiratory tract. Thus, these novel imaging techniques are valuable as complementary modalities to FB, providing information that is useful for the detection and management of airway malignancies.
