Published September 2003 | Version v1
Miscellaneous

Clinical value of virtual bronchoscopy in early diagnosis of central malignant lung neoplasm in high risk patients

Description

Virtual bronchoscopy (CT-VB) is a recently developed 3D visualization technique that employs thin-section spiral computed tomographic data of the thorax for non-invasive evaluation of the tracheo-bronchial tree. The purpose of this study was to compare the efficacy of multislice-CT virtual bronchoscopy in detecting early endobronchial manifestation of central lung cancer with that of flexible fiberoptic bronchoscopy. For this purpose, a group of 50 high risk patients (cigarette pack-years, long standing history of COLD) underwent both FOB and CT-VB investigation. Multislice CT was performed using the following parameters: collimation, 1 mm; pitch factor, 0.937-1.375; reconstruction interval, 0.8 mm and postprocessing was based on volume rendering technique. Images of eight standard tracheo-bronchial sections were obtained from each patient both during FOB and CT-VB and were independently interpreted respectively by a pneumologist and a bronchiseptica. Airway evaluation was based on the analysis of changes in the mucosal architecture, the shape of the ostia and of the carinae and on the presence of secretion deposits as early signs of tumor onset. The sensitivity of virtual bronchoscopy in discriminating diseased from healthy patients was 94 % and its specificity 65 %. Its positive predictive value 56 % and its negative predictive value 96 %. According to the ROC analysis, except for the assessment of the secretion deposits, the other three criteria evaluated separately during virtual bronchoscopic assessment of the airways contributed equally in discriminating diseased from disease-free patients (areas under curve ranging between 0,797 and 0,845). The best AUC was associated with the evaluation of the sum of all three parameters together (0,858). Nevertheless, FOB and CT-VB results in the evaluation of the shape of the ostia resulted to agree moderately while just a fair agreement was observed with respect to the evaluation of the mucosal architecture and of the shape of the carinae (Kappa values ranging between 0,310 and 0,520). As to the evaluation of the secretion deposits, for example, the two diagnostic tests showed very poor agreement (negative Kappa value) concluding, this preliminary evaluation indicates that virtual bronchoscopy combined with the evaluation of MS-CT axial scans may be a promising and non-invasive diagnostic method for evaluation of patients at high risk of developing central lung neoplasm and can add important information about an intraluminal tumor and its relation to surrounding structures. Further studies should be undertaken to evaluate the potentials of this promising method as a tool to evaluate endoluminal growth in tumorous lung lesions. Even though it showed just fair agreement with flexible fiberoptic bronchoscopy, the evaluation of changes in the form of the carinae, in the shape of the ostia and the accurate analysis of modifications in the mucosal architecture as early signs of tumor growth, showed to be three valid separate diagnostic criteria for airway assessment by means of virtual bronchoscopy. From a technical point of view, multislice-CT allowed the optimization of examination parameters such as acquisition time and primary collimation resulting not only in a reduction of respiratory and pulsation artefacts but also in a dramatic improvement of the z-axis and hence a much better resolution of the 3-dimensional reconstructions. As a consequence, MS-CT virtual bronchoscopy provides enhanced detail fidelity and represents a valid diagnostic tool for the evaluation of the whole tracheo-bronchial tree up to its third generation bronchi. (author)

Availability note (English)

Available from Univ. Bibliothek Innsbruck, Innrain 50, 6010 Innsbruck (AT)

Additional details

Publishing Information

Imprint Pagination
59 p.

INIS

Country of Publication
Austria
Country of Input or Organization
Austria
INIS RN
36083694
Subject category
S62: RADIOLOGY AND NUCLEAR MEDICINE;
Resource subtype / Literary indicator
Thesis, Non-conventional Literature
Descriptors DEI
BRONCHI; CARCINOMAS; COMPUTERIZED TOMOGRAPHY; DIAGNOSIS; IMAGES; LUNGS; PATIENTS; THREE-DIMENSIONAL CALCULATIONS
Descriptors DEC
BODY; DIAGNOSTIC TECHNIQUES; DISEASES; NEOPLASMS; ORGANS; RESPIRATORY SYSTEM; TOMOGRAPHY

Optional Information

Notes
Reference number: DG37666