Prognostic value of microscopic peritoneal dissemination: comparison between colon and gastric cancer

We evaluated the incidence and prognostic relevance of microscopic intraperitoneal tumor cell dissemination of colon cancer in comparison with dissemination of gastric cancer as a rational for additive intraperitoneal therapy.

Peritoneal washouts of 90 patients with colon and 111 patients with gastric cancer were investigated prospectively. Sixty patients with benign diseases and 8 patients with histologically proven gross visible peritoneal carcinomatosis served as controls. Intraoperatively, 100 ml of warm NaCl 0.9 percent were instilled and 20 ml were reaspirated. In all patients hematoxylin and eosin staining (conventional cytology) was performed. Additionally, in 36 patients with colon cancer and 47 patients with gastric cancer, immunostaining with the HEA-125 antibody (immunocytology) was prepared. The results of cytology were assessed for an association with TNM category and cancer grade, based on all patients, and with patient survival, among the R0 resected patients.

In conventional cytology 35.5 percent (32/90) of patients with colon cancer and 42.3 percent (47/111) of patients with gastric cancer had a positive cytology. In immunocytology 47.2 percent (17/36) of patients with colon cancer and 46.8 percent (22/47) of patients with gastric cancer were positive. In colon cancer, positive conventional cytology was associated with pT and M category (P = 0.044 and P = 0.0002), whereas immunocytology was only associated with M category (P = 0.007). No association was found between nodal status and immunocytology in colon cancer and with the grading. There was a statistically significant correlation between pT M category and conventional and immunocytology in gastric cancer (P < 0.0015/P = 0.007 and P < 0.001/P = 0.009, respectively). Positive immunocytology was additionally associated with pN category (P = 0.05). In a univariate analysis of R0 resected patients (no residual tumor), positive immunocytology was significantly related to an unfavorable prognosis in patients with gastric cancer only (n = 30). Mean survival time was significantly increased in patients with gastric cancer with negative cytology compared with positive cytology (1,205 (standard error of the mean, 91) vs. 771 (standard error of the mean, 147) days; P = 0.007) but not in patients with colon cancer (1,215 (standard error of the mean, 95) vs. 1,346 (standard error of the mean, 106) days; P = 0.55).

Because microscopic peritoneal dissemination influences survival time after R0 resections only in patients with gastric but not with colon cancer, our results may provide a basis for a decision on additive, prophylactic (intraperitoneal) therapy in gastric but not colon cancer.