We compared the efficacy and complications of video-assisted mediastinoscopy (VAM) and video-assisted mediastinal lymphadenectomy (VAMLA) for mediastinal staging of lung cancer.
Between March 2006 and July 2008, a total of 157 patients with non-small-cell lung cancer (NSCLC) underwent VAM (n = 113, 72%) or VAMLA (n = 44, 28%). We studied them retrospectively. Data for the operating time, node stations sampled/dissected, number of biopsies, and the patients who were pN0 by mediastinoscopy and underwent thoracotomy were collected. The false-negative rate was calculated. Demographics and operative complications were analyzed.
The overall complication rate was 5.7% (n = 9). The most common complication was hoarseness (n = 8). Complications were seen significantly more often after VAMLA than after VAM (11.3% vs. 2.6%, P = 0.04). There were no deaths. The mean number of removed lymph nodes (8.43 ± 1.08) and the station numbers (4.81 ± 0.44) per patient were higher with VAMLA than with VAM (7.65 ± 1.68, P = 0.008 and 4.38 ± 0.80, P = 0.001, respectively). The mean operating time was 44.8 ± 6.6 min for VAM and 82.0 ± 7.8 min for VAMLA. Patients diagnosed as pN2 numbered 9 in the VAMLA group and 27 in the VAM group. The patients diagnosed as pN0 with mediastinoscopy then underwent thoracotomy (VAM 77, VAMLA 32). When they were investigated for the presence of mediastinal lymph nodes, there were three (3.8%) false-negative results in the VAM group and five (15.6%) in the VAMLA group. Sensitivity, accuracy, and negative predictive values for VAM and VAMLA were 0.90/0.97/0.96 and 0.64/0.87/0.84, respectively.
VAMLA was found to be superior to VAM with regard to the number of stations and lymph nodes. Complications after VAMLA were common. The sensitivity and NPV of VAM for mediastinal staging are significantly higher than those of VAMLA.