Skip to main content
. 2015 Dec 1;5:17717. doi: 10.1038/srep17717

Table 1. Baseline demographics and clinical characteristics across different treatment groups.

Clinical characteristics N (%) Discovery Cohort Validation Cohort
Bevacizumab group Control group P-valuea
No. of patients 64 91 95
Age at diagnosis 52 (26–77) 53 (21–76) 55 (26–83) 0.93
Sex       0.45
 male 35 (56.5) 59 (60.2) 56 (63.6)  
 female 27 (43.5) 39 (39.8) 32 (36.4)  
Location of primary tumor       0.54
 colon 42 (67.7) 62 (68.1) 60 (63.2)  
 rectum 20 (32.3) 29 (31.9) 35 (36.8)  
ECOG performance status       0.94
 0–1 84 (95.5.) 92 (97.9) 84 (95.5)  
 ≥2 4 (4.5) 2 (2.1) 4 (3.5)  
Pathology       0.73
 moderately differentiated 40 (65.6) 55 (69.6) 60 (73.2)  
 poorly differentiated 21 (34.4) 24 (30.4) 22 (26.8)  
Matestasis       0.30
 single 31 (50.0) 46 (50.5) 56 (58.9)  
 multiple 31 (50.0) 45 (49.5) 39 (41.1)  
Curative-intent matestasis resection 6 (9.7) 17 (18.7) 19 (20.0) 0.86
Backbone chemotherapy       0.28
oxaliplatin-based 36 (56.2) 63 (70.0) 57 (61.3)  
irinotecan-based 28 (43.8) 27 (30.0) 36 (38.7)  
Prior adjuvant chemotherapy 17 (27.4) 30 (31.6) 23 (25.8) 0.42
Recurrent diseaseb 17 (27.4) 23 (25.3) 31 (32.6) 0.33
First-line duration of CT (median, mos) 5 (3.0–18.0) 4.5 (1.5- 25.0) 3.8 (2.5–24.0) 0.18
Total duration of CT (median, mos) 12 (3.0–22.0) 9.5 (2.0–25.0) 11.5 (2.5–26.0) 0.60
Second-line treatment 23 (39.0) 17 (23.3) 14 (15.9) 0.27
Anti-EGFR treatment following first PD 3 (4.8) 22 (24.7) 31 (32.6) 0.26
Maintenance treatmentc 20 (32.3) 10 (13.3) 14 (15.6) 0.83

aCompared patient characteristics between the bevacizumab group and the control group.

bPatients who had metastatic disease more than 6 months elapsed from adjuvant chemotherapy.

cMonotherapy of capecitabine or bevacizumab, or combined with both. Abbreviations: ECOG PS, eastern cooperative oncology group performance status; EGFR, epidermal growth factor receptor; CT, chemotherapy; mos, months. Statistical significance was set at 0.05 based on a two-sided test. P-values listed in bold were notable for possible association with clinical outcomes.