85P - Performance of EBUS-TBNA in NSCLC mediastinal staging stratified according to ACCP radiographic groups on CT

Date 15 April 2016
Event European Lung Cancer Conference 2016 (ELCC) 2016
Session Poster lunch
Topics Staging procedures (clinical staging)
Thoracic malignancies
Basic Principles in the Management and Treatment (of cancer)
Presenter Timothy Edwards
Citation Journal of Thoracic Oncology (2016) 11 (supplement 4): S57-S166. S1556-0864(16)X0004-4
Authors T. Edwards, H. Al-Najjar, P. Crosbie, J. Martin, R. Booton, M. Evison
  • North West Lung Centre, Wythenshawe Hospital-South Manchester University Hospitals Trust, M23 9LT - Manchester/GB



The American College of Chest Physicians (ACCP) use radiographic groups based on Computed Tomography (CT) of the thorax to predict the probability of mediastinal nodal metastases and inform the need for pathological nodal staging. Group A is a peripheral tumour with a normal mediastinum, Group B a central tumour or N1 lymphadenopathy but normal mediastinum and Group C is discrete mediastinal lymphadenopathy. This study investigated the performance of EBUS-TBNA stratified by these radiographic groups.


Prospective data is collected for all EBUS-TBNA procedures at out tertiary UK lung cancer centre. The radiographic group is recorded at the time of the procedure based on the index staging CT. EBUS-TBNA outcome data and subsequent data from mediastinoscopy, intra-operative lymph node sampling and 6 months of clinical-radiological follow-up is then recorded to allow calculation of performance (using the presence or absence of N2/3 disease as the denominator). This study is a retrospective analysis of the prospectively maintained database for the period 01/01/2014–31/12/2014.


See the table. 85PT1

 Group AGroup BGroup C
Prevalence of N2/30%21%69%
Reasons for false negative procedure:   
 Inaccessible nodes (5, 6, 8, 9)50%33%
 False negative sampling50%50%
 Accessible nodes not sampled12%


EBUS-TBNA is highly capable of identifying mediastinal disease in NSCLC patients with enlarged lymph nodes. This performance however does appear to reduce significantly in the radiologically normal mediastinum. This is not due to a lack of lymph node sampling but a combination of disease in lymph node stations inaccessible to EBUS and missing disease in nodes that have been sampled.

Clinical trial identification

Legal entity responsible for the study

University Hospitals of South Manchester


Manchester Thoracic Oncology Centre


All authors have declared no conflicts of interest.