Table 1—

Results of studies that assess the effectiveness and safety of endobronchial ultrasound(EBUS)-guided transbronchial needle aspiration (TBNA) in bronchopulmonary cancer

First author [ref]	Type of study	Patients/procedures included	Reference/comparison test	Diagnostic performance				P	Adequate sample	Safety
First author [ref]	Type of study	Patients/procedures included	Reference/comparison test	S	Sp	PPV	NPV	P	Adequate sample	Safety
Lee 18^#	Prospective	91 patients with strongly suspected or histologically confirmed potential NSCLC with lymph nodes accessible to EBUS–TBNA (5–20 mm on CT). Exclusion of M1 disease, inoperable T4 disease, N3 disease with spread to supra-clavicular lymph node, lymph nodes >2 cm or extranodal invasion.	Cytology results in positive cases Surgery (open thoracotomy or video-assisted thoracic surgery) in patients without mediastinal metastasis. No mediastinoscopy performed	93.8%	100%	100%	96.9%	42%	92%
Wallace 19^¶	Prospective	138 consecutive patients with known or suspected lung cancer on the basis of CT abnormality but no proven extrathoracic metastases Blinded comparison	Pathological confirmation for positive results Surgical sampling by mediastinoscopy or thoracoscopy, open surgical exploration or 6–12 months of follow-up	PET				28%	100%	No complications
				24%		90%
				CT
				67%		53%
				TBNA
				36%	100%	100%	78%
				EUS–FNA
				69%	100%	100%	88%
				EBUS–TBNA
				69%	100%	100%	88%
				EUS–FNA + EBUS–TBNA
				93%	100%	100%	97%
Bauwens 20^#	Prospective collection and retrospective assessment	106 patients with proven lung cancer (staging) or suspected lung cancer on FDG–PET (33 combined PET/CT)	Cytology/histology results in positive cases Surgical staging (30 out of 46), diagnostic procedures or follow-up in negative cases Blind retrospective comparison	EBUS–TBNA (n = 90 patients)				58%	94%	Side-effects, notably cough, seldom encountered
				93%	100%	100%	91%
				FDG–PET (n = 41 patients)
				67%			100%
Vincent 21^#	Retrospective	152 procedures carried out in 152 patients with primary lung masses or/and mediastinal adenopathy	Cytology results in positive cases with or without surgical confirmation Mediastinoscopy, lung resection, CT, FDG–PET or metabolic activity	99.1%	100%	100%	97%	74.3%	95.4%	No complications
Herth 17^#	Prospective	97 patients highly suspicious for NSCLC with CT scans showing lymph nodes <1 cm and negative PET in mediastinum	Surgical results in all patients	89%	100%	100%	98.9%	9.2%
Monsó 22^#	Prospective	67 patients with pulmonary neoplasms and mediastinal and/or lobular lymph nodes >5 mm on ultrasound	Cytology	Diagnostic yield 92.5%				55.2%	87.8%	No complications
Nakajima 14^#	Retrospective	43 patients with suspected metastasis of lung cancer and lymph nodes ≥5 mm on CT	Cytology or histology results in positive cases Thoracotomy, thoracoscopy or clinical follow-up (≥6 months)	92%	100%	100%	90%	53.5%	95.3%	No complications
Herth 15^#	Prospective	502 patients with mediastinal or hilar adenopathies (>1 cm; unknown origin or staging of lung cancer, especially exclusion of N3 nodes)	Cytology in positive cases Surgery or clinical follow-up in negative cases	94%	100%		11%	98.2%	93.5%	No complications
Herth 16^#	Prospective	100 patients with suspected tumour of pulmonary origin evidenced by CT (T1–T4), without enlargement of lymph nodes (measuring <1 cm) or diagnosis of suspected or known NSCLC	Mediastinoscopy (15%) or thoracotomy (85%) with resection of mediastinal nodes in all patients	92.3%	100%		96.3%	17%	100%	No complications
Yasufuku 13^#	Prospective	102 patients with suspected or anatomopathologically-confirmed lung cancer who were considered candidates for curative thoracic surgery (patients with stage I, II, or clinical minimal IIIa disease). Extensive N2/N3 disease excluded	Thoracotomy performed or diagnosis made from dissection of lymph nodes in potentially operable patients, and follow-up conducted in other patients (N3 or extensive N2 disease).	EBUS–TBNA				100%	100%	No major complications
				92.3%	100%	100%	97.4%
				CT
				76.9%	55.3%	37%	87.5%
				FDG–PET
				80%	70.1%	46.5%	91.5%
Yasufuku 12^#	Prospective	108 patients with suspected NSCLC or proven lung cancer and enlarged mediastinal lymph nodes (>1 cm) or mediastinal lesions suspected of malignancy detected by CT (N2 or N3)	Positive cytologies regarded as final diagnosis Follow-up of the disease and confirmation by thoracic lymphadenopathy in negative cases	94.6%	100%	100%	89.5%	62.9%	97.2%	No complications
Rintoul 23^#	Prospective	18 patients with suspected or known lung cancer and detection of nodular enlargement or presence of paratracheal or parabronchial masses on CT	Positive cytologies regarded as final diagnosis Benign results confirmed by surgery (n = 4) or clinical follow-up (n = 1)	85%	100%	100%	71.4%	72.2%	100%	No complications
Vilmann 24^#	Prospective	33 patients for staging of lung cancer (n = 20) or diagnosis of mediastinal lesions (n = 13)	Thoracotomy (n = 9), clinical follow-up (n = 19)	EBUS–TBNA				93.9%	95%	No complications
				85%	100%		72%
				EUS–FNA
				80%	100%		66%
Yasufuku 9^#	Prospective	70 patients with hilar and/or mediastinal lymphadenopathies (>1 cm) and with suspected or confirmed lung cancer.	Positive cytologies regarded as final diagnosis In benign results confirmation by thoracotomy or surgery (n = 21) and by clinical follow-up (n = 3)	95.7%	100%			68.5%	96%	No complications

S: sensitivity; Sp: specificity; PPV: positive predictive value; NPV: negative predictive value; P: prevalence of lymph node malignancy; NSCLC: nonsmall cell lung cancer; CT: computerised tomography; M: metastasis; T: tumour; N: lymph node; PET: positron emission tomography; EUS: transoesophageal ultrasonography; FNA: fine-needle aspiration; FDG: fluorodeoxyglucose. ^#: patient-based analysis; ^¶: lesion-based analysis.