Despite seemingly radical surgery many patients operated on for bronch
ial carcinoma will die from their disease. Some patients might benefit
from postoperative treatment and a prognostic factor that could ident
ify those with an increased risk for tumor relapse would be of great c
linical importance. One possible such factor is the occurrence of mali
gnant cells in pleural lavage performed at operation. To test this hyp
othesis 224 consecutive patients who had been operated on due to verif
ied or strongly suspected bronchial carcinoma, preoperatively staged a
s stage I or II, were investigated. After opening the thorax and befor
e manipulation or palpation of the lungs, 300 ml of physiological sali
ne solution was installed into the pleura. After excluding patients wh
o were not radically operated, there remained 138 patients with histol
ogically confirmed lung cancer (carcinoids excluded) and 12.3% showed
tumour cells in the washings. Two of 18 patients with metastatic lung
disease (11%) and one of ten patients with carcinoid tumor also showed
malignant cells in the lavage. The patients with lung cancer have bee
n followed for 3 years or until death. After three years 60.2% of thos
e without malignant cells in the pleural lavage were still alive, whil
e this figure was 41.2% in the other group. The difference was not sta
tistically significant. Other factors, such as spread to local lymph n
odes, size of tumor, etc. were related to the occurrence of malignant
cells in the pleura, and these factors were also better prognostic one
s. We conclude that the clinical use of pleural lavage cytology is lim
ited. (C) 1998 Elsevier Science Ireland Ltd. All rights reserved.