Objective: Despite technical difficulties due to mediastinal fibrosis,
repeat mediastinoscopy can be a valuable tool in the restaging of lun
g cancer. It provides essential pathological information on mediastina
l invasion when selecting patients for surgical resection after induct
ion chemotherapy in stage ma disease. The aim of our study was to eval
uate the feasibility, sensitivity and accuracy of repeat mediastinosco
py. Methods: From 1994 to 1997 we performed a repeat mediastinoscopy i
n 15 patients (13 men, two women) with bronchogenic carcinoma. Their a
ge ranged from 49 to 75 years. (mean 64.7). Seven patients had inducti
on chemotherapy for a nonsmall cell bronchogenic carcinoma with positi
ve N2 nodes on mediastinoscopy. Four patients had a second primary con
tralateral lung cancer, one had a locoregional recurrence of bronchoge
nic carcinoma. The other three had a first mediastinoscopy for other r
easons than lung cancer, repeat mediastinoscopy being performed for st
aging of malignant disease. Results: In all 15 patients it was possibl
e to perform a complete repeat mediastinoscopy. In one patient repeat
mediastinoscopy turned out to be false negative, so, in our series, se
nsitivity was 87.5%, specificity 100% and accuracy 93.7%. Conclusion:
Previous mediastinoscopy is no contraindication for a repeat one. Repe
at mediastinoscopy offers valuable pathological information in restagi
ng of lung cancer. (C) 1998 Elsevier Science B.V. All rights reserved.