1197O - Selective mediastinal node irradiation in NSCLC in the IMRT/VMAT era: added value of EBUS-TBNA-mapping to PET-CT

Date 29 September 2014
Event ESMO 2014
Session NSCLC, locally advanced and metastatic
Topics Non-Small-Cell Lung Cancer, Locally Advanced
Imaging, Diagnosis and Staging
Surgery and/or Radiotherapy of Cancer
Presenter Dirk De Ruysscher
Citation Annals of Oncology (2014) 25 (suppl_4): iv417-iv425. 10.1093/annonc/mdu348
Authors S. Peeters1, C. Dooms2, J.F. Vansteenkiste2, H. Decaluwe3, P. De Leyn3, K. Nackaerts2, W. De Wever4, C. Deroose5, D. De Ruysscher1
  • 1Radiation Oncology, University Hospitals Leuven - Campus Gasthuisberg, 3000 - Leuven/BE
  • 2Department Of Pulmonology And Leuven Lung Cancer Group, University Hospitals Leuven - Campus Gasthuisberg, Leuven/BE
  • 3Department Of Thoracic Surgery, University Hospitals Leuven - Campus Gasthuisberg, Leuven/BE
  • 4Department Of Radiology, University Hospitals Leuven - Campus Gasthuisberg, Leuven/BE
  • 5Nuclear Medicine, University Hospitals Leuven - Campus Gasthuisberg, 3000 - Leuven/BE



FDG-PET-CT based selective lymph node (LN) irradiation is the standard when using 3D-conformal techniques (3D-CRT) for locally advanced NSCLC. With 3D-CRT, adjacent LN not included in the target volume still receive a substantial radiation dose. However, with current new techniques (IMRT/VMAT), radiation dose to non-involved LNs decreases, which raises the question whether selective nodal irradiation based on FDG-PET-CT is still safe. We therefore evaluated the impact of adding EBUS-TBNA to FDG-PET-CT in selective nodal irradiation.


Literature data that included sensitivity and specificity of EBUS-TBNA in FDG-PET-CT staged NSCLC were identified. As false negative (FN) LNs have the largest consequence to avoid geographical miss, FN rates for different constellations of CT, PET and EBUS-TBNA were calculated. This algorithm was tested on 325 mediastinal LN in 25 consecutive NSCLC-patients with N2/N3 disease based on PET-CT who underwent full mediastinal EBUS-TBNA mapping. Results are expressed as mean +/- SD and range.


A practical algorithm when to include LNs in the GTV was made based on data from 5 large meta-analyses. The expected cancer prevalence, taking into account the FN rate of EBUS of 20%, were calculated. The resulting algorithm shows that for PET-positive and EBUS-negative LNs, FN rates are 16% for enlarged LNs, and 14% for normal-sized LNs. Furthermore, adding EBUS to FDG-PET-CT decreases the FN-rate of FDG-PET: from 13% to 3% for enlarged LNs, and from 6% to 1% for normal-sized LNs. In our population, 325 mediastinal LNs were visually analyzed on PET-CT scan. Sixty-seven were enlarged (≥10mm), of which 63 were PET-positive. Twelve normal-sized LNs were PET-positive. All accessible LNs were reviewed with EBUS. Fifty suspicious LNs were biopsied with EBUS-TBNA (mean: 2/patient +/-0.96; 1-5): 28 were malignant, 22 normal. EBUS-TBNA detected 1 cancer-containing normal-sized LN without FDG-uptake, thus 1/25 geographical miss (4%).


When the incidental nodal irradiation is low, such as in IMRT or VMAT, EBUS-TBNA should be added to FDG-PET-CT for mediastinal staging. PET-positive and EBUS-negative LNs should be included in the GTV as FN-rates remain high. Adding EBUS to FDG-PET-CT decreases the risk of geographical miss.


All authors have declared no conflicts of interest.