One hundred and ninety five consecutive, potentially curative resectio
ns for adenocarcinoma of the stomach were performed in one surgical de
partment between 1970 and 1989: 76 patients underwent gastrectomy with
splenectomy and 119 gastrectomy without splenectomy. Operative mortal
ity was 12% after gastrectomy with splenectomy, but only 2.5% after ga
strectomy without splenectomy (p<0.05). Postoperative complications we
re also significantly more common when splenectomy was combined with g
astrectomy (41% v 14%, p<0.01). Cumulative five year survival was 45%
after gastrectomy with splenectomy, compared with 71% after gastrectom
y alone (p<0.01). When the results of the two groups of patients were
compared, stage for pathological stage, no evidence was found that spl
enectomy improved survival. Application of Cox's proportional hazards
model, which makes allowance for other variables such as the T and N s
tages, showed that splenectomy had an adverse influence on patients' s
urvival. Splenectomy does not benefit the patient and its routine use
in the course of radical resections for carcinoma of the stomach shoul
d be abandoned.