[Surgical resection of colorectal liver metastases: Gold standard for solitary and radically resectable lesions].

From 1960 to 1993, a total of 1.766 patients with liver metastases from colorectal carcinoma was recorded. Five-hundred-and-eight patients (28.8%) underwent hepatic resection which was performed with curative intent in 473 patients (26.8%). 30-day mortality in this group was 4.5%, being 2.6% (4 out of 155) since 1990. Significant morbidity was observed in 16% of patients with a decrease to 7% for the last 4 years. A 99.5 percent follow-up until January 1, 1996, was achieved. Excluding operative mortality there are 376 patients with "potentially curative" initial liver resection, and 65 corresponding patients with minimal macroscopic (n = 19) or microscopic (n = 46) residual disease. The latter group demonstrated a poor prognosis with median and maximum survival times of 14.8 and 56 months, respectively. Among the 376 patients having potentially curative resection the actuarial five, ten, and twenty year survival was 39 +/- 3, 26 +/- 5 and 21 +/- 13 percent, respectively. Tumor-free survival was 34 +/- 3 percent at 5 years. In the univariate analysis, the following factors were associated with decreased crude survival: Presence and extent of mesenteric lymph node involvement (p = 0.0001), poor grading of the primary tumor (p = 0.008), synchronous diagnosis of metastases (p = 0.004), satellite metastases (p < 0.0001), an increasing metastasis diameter (p < 0.0001), preoperative CEA elevation (p = 0.0002), a resection margin of less than 1 cm (p = 0.018), extrahepatic disease (p = 0.02), non-anatomical procedures (p = 0.008), and an operative blood loss exceeding 2.000 ml (p = 0.02). With respect to disease-free survival, extrahepatic disease (p = 0.09) failed to achieve statistical significance, while patients with colon cancer and with delayed resection of synchronous metastases did significantly better than those with rectal cancer (p = 0.02) and with a simultaneous procedure (p = 0.04), respectively. Multiplicity and bilobar involvement did not affect prognosis. Similarly, no significant predictive value of an increasing number of metastases (1-3 vs > or = 4) on either overall (p = 0.35) or disease free survival (p = 0.55) was found after a radical excision of all detectable disease. Using Cox's multivariate regression analysis, presence of satellite metastases, anatomical vs non-anatomical approach, primary tumor grade and diameter of the largest metastasis all independently affected both crude and tumor-free survival (p < 0.05). With respect to survival, this was complemented by the margin of clearance (0.05 < p < 0.1), while for disease-free survival primary tumor site and time of metastasis diagnosis had some additional influence. Twenty-six patients with R0-reresection of the liver, and 32 patients with radical excision of extrahepatic recurrent disease had a subsequent 5-year survival of 57 +/- 15 percent and 32 +/- 12 percent, respectively. This confirms the effectiveness of a close follow-up policy.