Skip to main content
. 2013 Feb 5;121(18):3547–3553. doi: 10.1182/blood-2012-09-454694

Table 3.

Pairwise Cox proportional hazards models

Patients
Pairwise multivariate analysis
Variable No. % HR 95% CI P value
Pairwise analysis 1
 Pretreatment plasma EBV(+) 54 20 1.8 1.1-3.1 .02
 Age > 45 y 61 22 1.9 1.2-3.2 .01
Pairwise analysis 2
 Pretreatment plasma EBV(+) 54 20 1.9 1.1-3.1 .02
 Male gender 154 56 1.7 1.0-2.9 .03
Pairwise analysis 3
 Pretreatment plasma EBV(+) 54 20 1.8 1.1-3.1 .02
 Stage IV disease 82 30 1.7 1.1-2.8 .03
Pairwise analysis 4
 Pretreatment plasma EBV(+) 54 20 2.0 1.2-3.3 .01
 Albumin < 4 g/dL 185 68 1.2 0.7-2.1 .45
Pairwise analysis 5
 Pretreatment plasma EBV(+) 54 20 1.9 1.1-3.2 .02
 Hemoglobin < 10.5 g/L 52 19 1.5 0.9-2.6 .15
Pairwise analysis 6
 Pretreatment plasma EBV(+) 54 20 2.1 1.2-3.6 .01
 Leukocytosis ≥ 15 000 WBCs/mm3 45 16 1.4 0.8-2.7 .27
Pairwise analysis 7
 Pretreatment plasma EBV(+) 54 20 2.0 1.2-3.4 .01
 Lymphocytes < 600/mm3 or < 8% WBCs 62 23 1.3 0.8-2.3 .27

Pairwise multivariate analyses were performed for the cohort of 274 patients with pretreatment plasma specimens and each IPS factor, using Cox proportional hazard regression models. Pretreatment plasma EBV(+) was defined as >60 viral copies/100 μL of plasma. P values are for the correlation between each variable and FFS.

CI, confidence interval; HR, hazard ratio; WBC, white blood cells.