Table 2.

Higher dexamethasone clearance and anti–asparaginase antibodies were associated with cumulative incidence of CNS relapse

Factors*	Hazard ratio	95% CI	P†
Dexamethasone CL/F (L/h per m2) per 50-unit change‡	1.93	1.1-3.27	.014
Anti–ASP antibody positive vs negative	6.58	1.36-31.9	.019
Older than 10 years vs 1-10 years	0.37	0.09-1.5	.16
Initial leukocyte count ≥ 100 cells/mm3 vs < 100 cells/mm3	1.56	0.28-8.67	.61
Black vs white§	1.81	0.51-6.35	.36
Other vs white§	1.56	0.14-17.6	.72
T-lineage vs B-lineage‖	3.55	0.65-19.3	.14
t(1;19)[TCF3-PBX1] vs absent¶	6.40	0.93-43.9	.06
Others vs CNS1#	4.62	1.55-13.8	.0061
SR/HR vs LR	9.33	0.74-117	.084

ASP indicates asparaginase.

^*Prognostic features included in this model are known to be associated with treatment outcome in St Jude Total XV cohort by univariate analysis.⁹

^†P value from multivariate analysis using Fine and Gray regression model.¹⁸

^‡Dexamethasone clearance is continuous variable, and the hazard ratio is calculated for every 50-unit change in clearance.

^§Genetically determined race as described.¹⁰

^‖Acute lymphoblastic leukemia immunophenotype.

^¶t(1;19)[TCF3-PBX1] included in analysis because it was associated with CNS relapse.⁹

^#CNS1 (no detectable blast cells in a sample of cerebrospinal fluid) vs other CNS status: CNS2 (< 5 leukocytes per cubic millimeter with blast cells in a sample with < 10 erythrocytes per cubic millimeter) and CNS3 (≥ 5 leukocytes per cubic millimeter with blast cells in a sample with < 10 erythrocytes per cubic millimeter or the presence of a cranial nerve palsy), and traumatic lumbar puncture with blast cells (≥ 10 erythrocytes per cubic millimeter with blast cells.