Accurate staging of the mediastinum in non–small cell lung cancer is vital to guide management, as confirmation of mediastinal disease (N2/3) generally precludes curative surgery.1
Minimal-invasive staging with endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) has become routine, due to reduced morbidity and mortality,2 lower costs,3 and equivalent sensitivity when compared with mediastinoscopy.4 However, it remains unclear whether further sampling by mediastinoscopy is required following a negative EBUS-TBNA.
We report on a retrospective cohort of consecutive patients who underwent curative-intent resection (lobectomy or pneumonectomy) of non–small cell lung cancer with intraoperative mediastinal lymph node sampling, following mediastinal staging with EBUS-TBNA. Our intent was to determine the prevalence of N2 disease in this cohort to establish the role of mediastinoscopy following negative EBUS-TBNA.
This study was approved by the Melbourne Health Human Research Ethics Committee. In accordance with institutional policy, the requirement for informed consent was waived for this retrospective observational study. Individual patient information was deidentified before analysis. Patients had a computed tomography (CT) scan of the chest and fluorodeoxyglucose positron emission tomography (FDG-PET) performed before EBUS-TBNA and these results were available for planning of the bronchoscopy.
Mediastinal lymph node assessment was undertaken by EBUS-TBNA with a targeted approach. N3 nodes were examined initially and only nodes avid on PET or >6 mm were sampled. If no positive N3 nodes were identified, suspicious N2, then N1 nodes were sampled, as previously described.5 Each aspirate was subjected to rapid-onsite cytologic evaluation, as previously described.6 The procedure was terminated if rapid-onsite cytologic evaluation demonstrated malignancy.
No patients received neoadjuvant therapy before surgery. Mediastinal lymph node sampling was performed intraoperatively at the time of thoracoscopic lobectomy or pneumonectomy.
We identified 57 patients who had 1 or more matched mediastinal lymph node stations sampled by EBUS-TBNA and during intraoperative sampling between December 2009 and June 2014 across 3 hospitals in Melbourne, Australia. The mean age of patients was 65 years (range, 46 to 82 y) and 72% of patients were male. Thirty-two patients (56%) were “low risk” with no evidence of mediastinal disease on staging CT/PET (cN0/1 disease), whereas the remaining 25 patients (44%) were “moderate-high risk” (cN2/3 disease). EBUS-TBNA sampled 103 stations and intraoperative sampling sampled 118 stations. All patients had sufficient material on biopsy for diagnostic evaluation.
Eighty-two stations were sampled by both methods and these are detailed in Table 1. Of the 82 stations sampled by both methods, EBUS-TBNA and surgical sampling returned concordant negative results in 78 cases. These included 6/6 station 2 nodes, 28/30 station 4 nodes, and 44/46 station 7 nodes. Malignancy was identified in 3 of 57 patients (5.3%) and in 4 of 82 stations where EBUS had demonstrated benign findings, and hence the prevalence of mediastinal disease following negative EBUS was 4.9%. EBUS-TBNA false-negative results are outlined in detail in Table 2. The mean size of the malignant deposits within lymph nodes was 5.5 mm. This resulted in a per-node negative predictive value for EBUS-TBNA of 78/82=0.95. There was no N3 disease detected.
Of the 25 patients with moderate-high risk based on staging CT/PET, 4/35 lymph nodes were found to be malignant following surgical resection, resulting in a disease prevalence of 11% and a per-node negative predictive value (NPV) of 31/35=0.89. The remaining 32 patients had low-risk disease and no malignant nodes (out of 47 sampled) were found at surgical resection, resulting in a prevalence of 0% and a per-node NVP of 47/47=1.0. Stations sampled within these groups are shown in Table 3.
Four patients with cN2/3 disease and negative EBUS-TBNA underwent mediastinoscopy to assess for disease in stations sampled by EBUS-TBNA. In all 4 cases nonmalignant histology was confirmed and patients proceeded to curative resection.
The prevalence of mediastinal nodal disease following negative EBUS-TBNA is very low (4.9%). Decision making following negative EBUS-TBNA is influenced by understanding the NPV of the test. In turn, NPV is highly contingent on disease prevalence, which we have shown is low even in moderate-high risk patients (based on noninvasive imaging) following negative EBUS-TBNA.
We demonstrate an extremely high per-node negative predictive value of 0.95 in mediastinal nodes sampled by both methods. This figure increases to 1.0 in low-risk (cN0/1) patients, whereas the NVP in moderate-high risk (cN2/3) patients is 0.89. Previous studies have presented diagnostic accuracy of EBUS-TBNA on a per-patient basis; however, these include lymph nodes inaccessible to EBUS-TBNA and are therefore less useful in guiding decision making in light of a negative result at a specific lymph node site. Our figure is consistent with recent studies demonstrating high NPV of EBUS-TBNA,7–9 and supports the conclusion of Clementsen and colleagues that a negative EBUS-TBNA of suspicious mediastinal lymph nodes identified on PET/CT does not require further sampling of those stations before proceeding to surgical resection with curative intent.
The randomized ASTER trial10 also demonstrated the prevalence of mediastinal disease was low (9%) following negative EBUS-TBNA, consistent with our findings. False-negative results were seen only in patients with microscopic metastases. Furthermore, the sensitivity of mediastinoscopy following negative EBUS-TBNA was only 60%,10 resulting in a high number needed to treat to preoperatively detect mediastinal disease following a negative EBUS-TBNA. If applied to our moderate-high risk cohort, this approach would result in the performance of 25 mediastinoscopies for the detection of just 2 patients with microscopic mediastinal deposits.
There are several limitations to this study. This is a retrospective cohort study with limitations inherent to this type of study design. The number of stations sampled by both approaches (1.4 per patient) may have been increased by a systematic, rather than targeted approach to sampling. While such an approach may improve the per-patient NPV, it is unlikely to affect the per-node outcomes and therefore we believe this does not detract from our results.
We are unable to calculate sensitivity and positive predictive value, as positive EBUS-TBNA results were generally not confirmed surgically. However, these figures have previously been reported in meta-analysis.11 Finally, NPV may be influenced by prevalence. While there was a low prevalence of mediastinal disease in this study, our results are consistent with other recent studies on this subject.
The prevalence of mediastinal metastasis in lymph nodes sampled by EBUS-TBNA where sampling demonstrates benign findings is extremely low, at 4.9% overall, 0% in patients staged cN0/1 by integrated CT/PET, and 11% in patients staged cN2/3. Per-node NVP of EBUS-TBNA is 0.95. Given the low prevalence of disease and the reported low sensitivity of mediastinoscopy following negative EBUS-TBNA,10 we suggest that negative EBUS-TBNA of mediastinal lymph nodes, regardless of PET/CT findings, does not require confirmatory sampling of those stations by mediastinoscopy before proceeding to resection with curative intent.
1. Ramnath N, Dilling TJ, Harris LJ, et al.. Treatment of stage III non-small cell lung cancer: diagnosis and management of lung cancer, 3rd ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest. 2013;143(suppl):e314S–e3140.
2. Silvestri GA, Gonzalez AV, Jantz MA, et al.. Methods for staging
non-small cell lung cancer: diagnosis and management of lung cancer, 3rd ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest. 2013;143(suppl):e211S–e2150.
3. Steinfort DP, Liew D, Conron M, et al.. Cost-benefit of minimally invasive staging
of non-small cell lung cancer: a decision tree sensitivity analysis. J Thorac Oncol. 2010;5:1564–1570.
4. Yasufuku K, Pierre A, Darling G, et al.. A prospective controlled trial of endobronchial ultrasound
-guided transbronchial needle aspiration compared with mediastinoscopy for mediastinal lymph node staging
of lung cancer. J Thorac Cardiovasc Surg. 2011;142:1393.e1–1400.e1.
5. van der Heijden EH, Casal RF, Trisolini R, et al.. Guideline for the acquisition and preparation of conventional and endobronchial ultrasound
-guided transbronchial needle aspiration specimens for the diagnosis and molecular testing of patients with known or suspected lung cancer. Respiration. 2014;88:500–517.
6. Steinfort DP, Russell PA, Tsui A, et al.. Interobserver agreement in determining non-small cell lung cancer subtype in specimens acquired by EBUS-TBNA
. Eur Respir J. 2012;40:699–705.
7. Clementsen PF, Skov BG, Vilmann P, et al.. Endobronchial ultrasound
-guided biopsy performed under optimal conditions in patients with known or suspected lung cancer may render mediastinoscopy unnecessary. J Bronchology Interv Pulmonol. 2014;21:21–25.
8. Sanz-Santos J, Andreo F, Castellà E, et al.. Representativeness of nodal sampling with endobronchial ultrasonography in non-small-cell lung cancer staging
. Ultrasound Med Biol. 2012;38:62–68.
9. Liberman M, Sampalis J, Duranceau A, et al.. Endosonographic mediastinal lymph node staging
of lung cancer. Chest. 2014;146:389–397.
10. Annema JT, van Meerbeeck JP, Rintoul RC, et al.. Mediastinoscopy vs. endosonography for mediastinal nodal staging
of lung cancer: a randomized trial. JAMA. 2010;304:2245–2252.
11. Gu P, Zhao YZ, Jiang LY, et al.. Endobronchial ultrasound
-guided transbronchial needle aspiration for staging
of lung cancer: a systematic review and meta-analysis. Eur J Cancer. 2009;45:1389–1396.