Comparison of Endobronchial Ultrasound, Positron Emission Tomography, and CT for Lymph Node Staging of Lung Cancer: Operability

Patients with proven or suspected lung cancer who were judged medically fit with stage I, II, or clinical minimal IIIa disease (defined as single-station N2 lymph node positive on EBUS-TBNA) were considered operable. These patients underwent lobectomy with complete thoracic lymphadenectomy. Those patients with extrathoracic spread disease, extensive N2 disease (bulky disease or multiple N2 positive nodes), or N3 disease proven by needle biopsy were considered inoperable.
For CT, PET, and EBUS-TBNA, mediastinal lymph node staging was individually assessed either as N0, N1, N2, or N3. The results of each modality were compared to the final surgical/ pathologic diagnosis confirmed by thoracotomy with complete mediastinal lymph node dissection. In case N3 or extensive N2 disease was confirmed by EBUS-TBNA and surgical resection not performed, the results were compared to clinical course consistent with malignant disease. The sensitivity, specificity, positive predictive value, negative predictive value, and diagnostic accuracy rate for prediction of lymph node staging was calculated using the standard definitions. The x2 test was used for comparison of the three modalities for the correct prediction of lymph node status. From December 2003 to March 2005, a total of 280 patients with suspected or pathologically established lung cancer were evaluated by the multidisciplinary group. Of these 280 patients, 178 patients were not further evaluated for presence of distant metastasis, tumor resectability, or medical operability. The remaining 102 patients with proven (n = 96) or suspected (n = 6) lung cancer fulfilled the criteria and underwent CT, PET, and EBUS-TBNA for mediastinal staging prior to surgery. The characteristics, final diagnosis, and the location of the target lesions in these 102 patients are shown in Table 1. In the six suspected lung cancer cases with negative bronchoscopy findings prior to EBUS-TBNA, final diagnosis of lung cancer was obtained from tissue sampling of lymph nodes. In all, EBUS-TBNA was successfully performed in 147 mediastinal and 53 hilar lymph nodes. Thirty-seven lymph nodes were malignant and 163 were benign based on EBUS-TBNA (Table 2). As a result, 24 patients were found to have N3 or extensive N2 disease. Of these 24 patients, 4 patients found to have minimal N2 disease had surgical resection of lung cancer with complete thoracic lymphadenectomy.

Table 1—Characteristics, Final Diagnosis, and Location of Target Lesions in Patients Enrolled in the Study

Variables Data
Patients, No. 102
Male/female gender, No. 81/21
Median age (range), yr 67.8 (44-85)
Final diagnosis, No.
Adenocarcinoma 72
Squamous cell carcinoma 24
Small cell carcinoma 3
Large cell neuroendocrine carcinoma 3
Location of lymph nodes targeted by 200
Highest mediastinal (#1) 1
Right upper paratracheal (#2R) 27
Left upper paratracheal (#2L) 7
Right lower paratracheal (#4R) 42
Left lower paratracheal (#4L) 13
Subcarinal (#7) 57
Right hilar (#10R) 19
Left hilar (#10L) 4
Right interlobar (#11R) 18
Left interlobar (#11L) 12
Lymph node size by CT (range), mm
Long axis 12.7 (5-25)
Short axis 8.7 (5-22)

Table 2—Lymph Node Location and Number of Abnormalities on CT, PET, and EBUS-TBNA

Lymph Node Location of Abnormalities, No CT (n = 92) PET (n = 89) EBUS-TBNA (n = 37)
Supraclavicular 0 4(3) 0
Highest mediastinal (#1) 0 0 0
Right upper paratracheal (#2R) 20(11) 15 (9) 7 (0)
Left upper paratracheal (#2L) 3 (2) 3(1) 1(0)
Right lower paratracheal (#4R) 20 (13) 17 (10) 8 (0)
Left lower paratracheal (#4L) 5(2) 5 (3) 4 (0)
Subaortic (#5) 0 1 (0) 0
Subcarinal (#7) 22 (15) 22 (11) 12 (0)
Right hilar (#10R) 6 (4) 9 (8) 1 (0)
Left hilar (#10L) 1(1) 2 (2) 0
Right interlobar (#11R) 12 (7) 8 (6) 3 (0)
Left interlobar (#11L) 3(1) 3 (3) 1 (0)

Leave a Reply

Your email address will not be published. Required fields are marked *