Tumor size as a predictive factor.

Is tumor size a predictive factor in carcinoma of the pancreas? In order to answer this question, the data of the Japan Pancreatic Cancer Registry from 1981 to 1987 were retrospectively analyzed. Of the 7687 cases that were registered, 5826 cases (75.7%) were laparotomized. Out of the 5826, 2311 (39.7%) underwent resection. Postoperative deaths within 1 mo among resected cases numbered 104, and the operative mortality rate was 4.5%. The postoperative survival rates were calculated by Kaplan-Meier's method. They clearly showed that the rates for small carcinomas (greater than 2 cm [T1]) were significantly higher than those for the tumors larger than 2 cm (T2, T3, T4) according to the Greenwood formula. Although the survival-rate calculation excluded operative deaths and deaths of causes other than pancreatic cancer, and included islet cell carcinoma, cystadenocarcinoma, and acinar cell carcinoma, it seems that tumor size may be a predictive factor for postoperative prognosis. However, our previous study on long-term survivors (more than 5 yr) after resection on the exocrine pancreatic cancer) revealed that there was no relation between the size of tumor and the postoperative prognosis. In those long-term survivors, there was no definite or severe cancerous invasion of either the retroperitoneal tissue or the portal venous system. It is concluded that a tumor 2 cm or smaller has a significantly better prognosis after resection than does a tumor larger than 2 cm, and it seems, in general, that tumor size may be a predictive factor. However, if the tumor does not invade the retroperitoneal tissue or the portal venous system, tumor size per se is not always a predictive factor.