Prognostic factors of survival in repeat liver resection for recurrent colorectal metastases: review of sixty-two cases treated at a single institution.

Purpose: Liver metastases develop in 50 percent of patients with colorectal carcinoma. Recurrent liver disease is usual. Repeat liver resection remains the only curative treatment. The aim of this study was to review our data on repeat hepatectomy and to analyze potential prognostic factors of survival.

Method: Patients who underwent repeat liver resection for metastases of colorectal carcinoma between January 1992 and August 2007 were identified from a prospective database and their medical records were analyzed.

Results: Of 62 patients who underwent a second hepatectomy, 15 underwent a third hepatectomy, and two underwent a fourth hepatectomy. There was no perioperative mortality. Morbidity was less than 20 percent for the first and second hepatectomies. Overall 5-year survival rate after first hepatectomy was 40 percent. Univariate analysis identified three risk factors confirmed by log-rank test and multivariate Cox regression analysis: serum carcinoembryonic antigen concentrations >5 ng/ml at first hepatectomy (HR = 2.265; CI = 1.140-4.497; P = 0.020), anatomic resection (HR = 2.124; CI = 1.069-4.218; P = 0.031), and tumors > or =3 cm at the second resection (HR = 2.039; CI = 1.013-4.103; P = 0.046).

Conclusion: Our study shows that repeat hepatectomy for liver metastases of colorectal carcinoma may be performed with low mortality and morbidity. Preoperative concentration of carcinoembryonic antigen at first hepatectomy, tumor size, and type of anatomic resection are independent prognostic factors.