Skip to main content
. Author manuscript; available in PMC: 2013 Dec 8.
Published in final edited form as: Ann Surg. 2012 Oct;256(4):10.1097/SLA.0b013e31826b4dcc. doi: 10.1097/SLA.0b013e31826b4dcc

TABLE 2.

Univariate and Multivariate Analysis of Clinicopathologic Variables Associated With DFS in 341 Patients Who Underwent Hepatectomy for Metachronous CLM

Variable Univariate Analysis
Multivariate Analysis*
N = 341 (%) Median DFS, mo P HR 95% CI P
Sex Male 60 19 0.340
Female 40 20
Age, y >65 35 19 0.268
≤65 65 20
Primary tumor Rectal 29 20 0.820
Colon 71 18
Lymph nodes for primary Positive 54 16 0.023 NS
Negative 46 25
Number CLM Multiple 61 15 0.065
Solitary 39 22
Adjuvant therapy for CRC None 28 29 < 0.0001 1.52 1.23–1.89 < 0.0001
5-FU 49 20
FOLFOX 23 12
Size, cm > 3 47 20 0.434
≤3 53 18
CEA, ng/mL > 5 51 17 0.924
≤5 49 16
Preoperative chemotherapy for CLM Yes 55 15 0.028 NS
No 44 25
Portal vein embolization Yes 7 11 0.174
No 93 20
Transfusions Yes 10 16 0.602
No 90 20
Positive surgical margins at CLM resection Yes 12 13 0.012 NS
No 88 20
Pathologic response to preoperative chemotherapy for CLM (n = 189) Major 48 16 0.089
Minor 52 12
Postoperative chemotherapy for CLM Yes 64 18 0.113
No 36 23
Complications Yes 27 19 0.687
No 73 20
*

Cox regression multivariate analysis included all variables with P < 0.05 in univariate analysis.

NS indicates not significant.