Predictive factors for tuberculous peripheral pulmonary lesions during radial endobronchial ultrasound.
Respir Med
; 234: 107805, 2024 Sep 13.
Article
em En
| MEDLINE
| ID: mdl-39265839
ABSTRACT
BACKGROUND:
Tuberculosis frequently poses diagnostic challenge when it presents as a peripheral pulmonary lesion (TB-PPL). The growing use of radial endobronchial ultrasound (rEBUS) for PPL biopsy highlights the need to identify predictive factors for TB-PPL, which is crucial for procedure safety.METHODS:
A six-year retrospective review at our institution on adult patients with TB and malignant-PPL diagnosed from rEBUS procedure from October 1, 2016, to December 31, 2022. Clinical, radiological, procedural, histological and microbiological data were extracted and analysed.RESULTS:
387 PPLs were included in our cohort, 32 % were TB-PPL and 68 % were malignant-PPL. The median age was 63 (IQR 55-70) years, with the TB-PPL group significantly younger. The median size of the target lesion was 2.90 (IQR 2.26-4.00) cm. The overall rEBUS diagnostic yield was 85.3 %, with a 1.3 % pneumothorax risk. Multivariate analysis identified independent predictors for TB-PPL, including age <60 years (adj OR 2.635), target lesion size <2 cm (adj OR 2.385), upper lobe location (adj OR 2.020), presence of a cavity on pre-procedural CT (adj OR 4.186), and presence of rEBUS bronchogram (adj OR 2.722). These variables achieved an area under the curve of 0.729 (95 % CI 0.673-0.795) with a diagnostic accuracy of 75.49 % (95 % CI 70.68-79.88).CONCLUSIONS:
Despite non-specific radiological findings in TB-PPL, our study identifies younger age, target lesion size less than 2 cm, upper lobe location, the presence of cavitation, and rEBUS bronchogram were independent clinical predictors for TB-PPL. This prediction model potentially helps mitigate the risk of accidental TB exposure during bronchoscopic procedures. A future prospective cohort study to validate these findings is essential to allow proper triaging of patient planning for rEBUS procedure.
Texto completo:
1
Base de dados:
MEDLINE
Idioma:
En
Ano de publicação:
2024
Tipo de documento:
Article