Towards a minimally invasive staging strategy in NSCLC: analysis of PET positive mediastinal lesions by EUS-FNA

Abstract

Study objective: To asses the value of endoscopic ultrasound guided fine needle aspiration (EUS-FNA) in the nodal staging of patients with (suspected) non-small cell lung cancer (NSCLC) and a ${}^{18}\text{FDG}$ positron emission tomography (PET) scan suspect for N2/N3 mediastinal lymph node (MLN) metastases. Background: Due to the imperfect specificity of positron emission tomography, PET positive MLN should be biopsied in order to confirm or rule out metastasis. Currently, invasive surgical diagnostic techniques such as mediastinoscopy/-tomy are standard procedures to obtain MLN tissue. The minimally invasive technique of EUS-FNA has a high diagnostic accuracy (90–94%) for the analysis of MLN in patients with enlarged MLN on computed tomography of the chest (CT). Design and patients: Thirty-six patients with proven (n=26) or suspected (n=10) non-small cell lung cancer and a PET scan suspect for N2/N3 lymph node metastases underwent EUS-FNA. When EUS-FNA did not confirm metastasis and the PET lesion was within reach of mediastinoscopy, a mediastinoscopy was performed. EUS-FNA negative patients with PET lesions beyond the reach of mediastinoscopy or those with a negative mediastinoscopy were referred for surgical resection of the tumour and MLN sampling or dissection. Results: EUS-FNA confirmed N2/N3 disease in 25 of the 36 patients (69%) and was highly suspicious in one. In the remaining 10 patients, one PET positive and one PET negative N2 metastasis was detected at thoracotomy. The PPV, NPV, sensitivity, specificity and accuracy of EUS-FNA in analysing PET positive MLN were 100%, 80%, 93%, 100% and 94%, respectively. No complications of EUS-FNA were recorded. Conclusions and significance: EUS-FNA yields minimally invasive confirmation of MLN metastases in 69% of the patients with potential mediastinal involvement at FDG PET. The combination of PET and EUS-FNA might qualify as a minimally invasive staging strategy for NSCLC.

Introduction

Accurate mediastinal staging is of major importance for the prognosis and assignment of patients to optimal treatment for non-small cell lung cancer (NSCLC). A recent retrospective study showed that up to 36% of thoracotomies performed for treatment of (presumed) non-small cell lung carcinoma are futile, due to locally advanced, irresectable or benign lesions, despite preceding staging procedures [1]. This observation stresses the need to improve preoperative staging in NSCLC in order to identify those patients who will benefit from surgical resection. Recent developments in positron emission tomography (PET) and endoscopic ultrasound guided biopsies (EUS-FNA) may provide new staging opportunities for NSCLC.

Mediastinal staging by ${}^{18}\text{F}$-fluorodeoxyglucose (${}^{18}\text{FDG}$) PET is superior in both sensitivity (79–91%) and specificity (86–91%) compared to computed tomography (CT) scan of the chest (60–75%) and (61–75%), respectively [2], [3], [4]. Moreover, the addition of a PET scan to conventional staging may prevent unnecessary surgery in one out of five patients, resulting in a 51% relative reduction of futile thoracotomies as compared to conventional work up alone according to prevailing guidelines [5]. With respect to mediastinal staging, the main value of PET is its high negative predictive value [4]. However, the false positive rate ranges from 9 to 39% [2], [3], [6], [7] so that PET positive lesions should be biopsied to preclude that patients are denied a potentially curative resection. To this end, invasive surgical procedures such as mediastinoscopy, mediastinotomy, VATS or even exploratory thoracotomies are performed.

Transesophageal endoscopic ultrasound guided fine needle aspiration has demonstrated high sensitivity (83–92%) and specificity (100% ) in analysing mediastinal lymph nodes (MLN) in NSCLC [8], [9], [10], [11], [12]. It should be noted that these results are largely obtained in selected patients with enlarged MLN on CT. EUS-FNA is a safe and minimally invasive technique which can be performed in an outpatient setting. EUS-FNA is complementary to mediastinoscopy in its diagnostic range. EUS-FNA provides easy access to the following MLN stations: the retrotracheal station (3), the lower paratracheal station on the left (4L), the aortopulmonary window (5), the subcarinal (7), the lower paraesophageal (8), as well as those stations in the ligamentum pulmonale (9), whereas mediastinoscopy provides good access to both the upper (2R, 2L), lower paratracheal areas (4R, 4L) and the anterior subcarinal area (7). The diagnostic reach of EUS-FNA and mediastinoscopy are largely complementary (Fig. 1). In a comparison study the combination of CT and PET was as accurate as EUS-FNA in the mediastinal staging of suspected lung cancer [12].

In this study, we assessed the feasibility and yield of EUS-FNA to demonstrate MLN metastases in patients with (suspected) NSCLC and possible N2/N3 involvement at FDG PET. If so, the combination of PET and EUS-FNA might qualify as a minimal invasive staging strategy, thereby reducing the need for invasive diagnostic surgical procedures. A staging strategy was used starting with PET, while PET is a non-invasive staging tool, has a high negative predictive value and can be helpful to detect distant metastasis [3].