Skip to main content
. Author manuscript; available in PMC: 2012 Mar 1.
Published in final edited form as: Nat Genet. 2011 Feb 6;43(3):237–241. doi: 10.1038/ng.763

Table 3.

Multivariate Analysis for Risk of ALL Relapse

Patient Characteristics P Valuea Hazard Ratiob (95% CI)
MRD Positive <×10−6 3.73 (2.98, 4.68)
Leukocyte count at diagnosis (≥50,000/ul) 6.00×10−6 1.93 (1.45, 2.57)
DNA Index (≥1.16) 2.58×10−4 0.59 (0.45, 0.79)
Native American ancestry 0.017 1.84 (1.12, 3.04)
Age at diagnosis (≥10 years) 0.010 1.39 (1.08, 1.78)
ETV6-RUNX1 c 0.012 0.67 (0.49, 0.91)
T-cell lineagec 0.146 0.67 (0.40, 1.15)
BCR-ABL c 0.454 1.48 (0.53, 4.12)
TCF3-PBX1 c 0.793 0.93 (0.53, 1.63)
MLL rearrangementsc 0.935 0.96 (0.41, 2.25)

Abbreviations: MRD: minimal residual disease; CI: confidence interval

a

Associations with risk of relapse (any relapse) were assessed using the Fine and Gray’s regression model.

b

Hazard ratio: the relative difference (increase or decrease) in risk of ALL relapse when the patient is positive for the clinical feature of interest (e.g., 1.84-fold increase in relapse risk for every 100% increase in NA ancestry).

c

Terms refer to cancer characteristics of the ALL cells that may have prognostic significance and are used to subdivide cases. All prognostic features are dichotomized for presence vs absence of the patient characteristic except for NA ancestry, which is treated as a continuous variable. Supplementary Table S2 includes identical multivariate analyses with dichotomized variables used for all variables, including ancestry.