Background: We examined the potential prognostic value of free malignant ce
lls in the peritoneal cavity of patients with colon adenocarcinoma. Methods
: In 1991 and 1992, 88 patients underwent peritoneal washing with cytologic
analysis immediately before elective colon resection for adenocarcinoma wi
th no evidence of peritoneal metastases. Peritoneal washing fluid was colle
cted before abdominal exploration, centrifuged immediately, and stained wit
h the May-Grunwald-Giemsa and Harris Shorr methods. Malignancy was defined
as recommended by Papanicolaou. Results: Free malignant cells in peritoneal
fluid (positive peritoneal washing) were found in 25 of 88 (28%) patients
(Dukes A, 0 of 11; Dukes B, 10 of 31; Dukes C, 11 of 37; Dukes D, 4 of 9).
The positivity rate was 24 of 75 (32%) among patients with rumours involvin
g the serous layer and 1 of 13 (8%) among the others (P = 0.0989). With a f
ollow-up of 66.7 to 78.7 months the 5-year survival rates were 48% (95% con
fidence interval, 28.4-67.6) and 68% (95% CI, 56.2-79.8) among patients wit
h positive and negative peritoneal washing, respectively (P = 0.09). Positi
ve peritoneal washing was not significantly associated with survival in mul
tivariate analysis. Conclusion: The presence of free malignant cells in the
peritoneal cavity of patients with colon cancer provided no further progno
stic information relative to the Dukes classification in this study. Nevert
heless, further study is needed, particularly in a larger series of Dukes B
patients, for whom a new prognostic factor would be useful for deciding ad
juvant therapy.