Abstract:Single institution studies of hepatic resection have implied that some subsets of patients with metastases from colon and rectum adenocarcinoma can be cured by surgery. However, it is unknown how such patients can be selected and what, in fact, the ultimate cure rate is. The objective of this multi-institutional study was to define in a prospective manner how many patients predicted to have resectable liver metastases actually could undergo curative resection and what the disease-free and overall survival were. A prospec tive 15-institution treatment plan included all patients who were predicted to have resectable metastases from colon and rectum cancer primaries. Results were compared among patients who underwent successful resection, patients who underwent resection with pathologically defined inadequate margins, and patients who underwent exploration but were found at surgery not to have resectable liver metastases. Fourteen institutions throughout North America and one in Milan who were members of the Gastrointestinal Tumor Study Group at the inception of this protocol in July of 1984 participated in the study. The patient population consisted of all those with metachronous or synchronous metastases from colon or rectum adenocarcinoma seen during a consecutive 31 year period with protocol registry at the time liver metastases were predicted to be isolated and resectable. Intervention consisted of the application of standard preoperative screening tests, abdominaI exploration, and resection of all patients who had liver-only surgically resectable disease. The main outcome measures were: Accuracy of preoperative staging, acute and chronic operative morbidity, and disease-free, as well as overall, survival among the three treatment groups. The minimum follow up on all surviving patients after curative resection was 2.2 years, with a median follow up of 3.2 years. Median survival times for patients receiving curative and noncurative resections and for those receiving no resection were estimated to be 35.7, 21.2, and 16.5 months, respectively. No statistically significant difference in the survival distribution of the noncuratively resected
Offprint requests to: G. Steele JrReceived for publication on Nov. 15, 1994; accepted on Dec. 7, 1994 or the non-resected patients was observed. The survival distribution of curatively resected patients remains statistically superior to the distribution of noncuratively resected patients and those receiving no resection (P = 0.01). These results confirmed that, although liver resection for hepatic metastases from colorectal car cinoma is safe to do and presently constitutes the only standard curative treatment for potentially resectable disease, it ultimately cures few patients.