Start Your Search
Cristina Martín Cabeza
OA01 - Advanced Diagnostic Approaches for Intrathoracic Lymph Nodes and Peripheral Lung Tumors (ID 117)
- Event: WCLC 2019
- Type: Oral Session
- Track: Interventional Diagnostics/Pulmonology
- Presentations: 1
- Now Available
- Moderators:Marioara Simon, Tim Murgu
- Coordinates: 9/08/2019, 10:30 - 12:00, Melbourne (1991)
OA01.03 - Probability Model for Malignancy in Hilar and Mediastinal Lymph Nodes in Lung Cancer Based on PET-CT and EBUS (Now Available) (ID 133)
10:30 - 12:00 | Author(s): Cristina Martín Cabeza
The mediastinal lymph nodes (LN) staging is routinely performed by PET-CT and EBUS- TBNA. Nevertheless, there are no studies that explore the diagnostic capacity of both techniques together. This study aims is to find an algorithm based on combined PET-CT and EBUS image variables together with clinical criteria that provides the most accurate probability of malignancy for each LN explored.Method
Retrospective study of mediastinal staging of non-small cell lung cancer, based on PET-CT and EBUS-TBNA. The LN were identified by level (N1, N2 and N3) and by anatomical region (AR) (subcarinal, not subcarinal, and hilar). Standardized Uptake Value (SUV) was determined for each sampled LN (maximum, medium and peak) as well as for pulmonary mass, liver, and blood pool. The ultrasound features collected were: diameter in the short axis (DSA), morphology, border, ecogeneicity and presence of the vascular hilium. For the construction of the predictive algorithm a mixed model of logistic regression of Firth was used.Result
116 consecutive patients were included and a total of 358 LN were evaluated. The set of variables that presented the best discrimination were: age, DSA, SUVmax and AR. The model determines the probability for malignancy for each LN, using the following formula = (-9.26) constant + (-0.21) Age + (4.29) SUVmax + (0.52) DSA + AR. The discrimination power of the model measured by the Area Under the Roc curve was = 0.95.
The model including age, DSA, SUVmax and AR provide the probability of malignancy for each LN with the highest accuracy. All other variables can be discarded when combining PET-CT and EBUS image features.
Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.
Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.