Association Between Acute Leukoencephalopathy and Long-Term Neurobehavioral and Brain Imaging Outcomes in Survivors of Childhood Acute Lymphoblastic Leukemia Treated with Chemotherapy Only

Introduction: Leukoencephalopathy is observed in a subset of children undergoing chemotherapy for acute lymphoblastic leukemia (ALL), though the impact of these white matter abnormalities on long-term behavior and brain integrity are unknown. This study examines associations between acute (on-therapy) leukoencephalopathy and neurobehavioral ratings and white matter integrity in long-term survivors of ALL treated with chemotherapy only.

Methods: 408 patients with newly diagnosed ALL were treated on St. Jude Total XV protocol which omitted cranial irradiation in all patients. Of the 369 patients who had prospective MRI scan of brain during active therapy, 294 were eligible for long-term follow-up and 189 (64%) participated in neurobehavioral assessment and brain imaging when ≥5 years post-diagnosis. Brain MRI’s during therapy and at follow-up were systematically coded by a Board Certified Neuroradiologist (blinded to the neurobehavioral outcomes) using the Common Terminology Criteria for Adverse Events (CTCAE) 4.03. At follow-up, survivors’ parents completed the Behavior Rating Inventory of Executive Function (BRIEF) to assess survivors’ neurobehavioral problems. Diffusion tensor imaging (DTI) was conducted to assess white matter integrity. Fractional anisotropy (FA), axial diffusivity (AD) and radial diffusivity (RD) were extracted from DTI voxels within the frontostriatal tract, given its association with executive function. Generalized linear models were used to examine associations among leukoencephalopathy, long-term neurobehavioral and DTI outcomes, adjusting for current age.

Results: Acute leukoencephalopathy was identified in 49 survivors (28.3%), 78% of whom continued to demonstrate leukoencephalopathy at follow-up. Compared to population norms, survivors had more severe problems with working memory (mean[SD] Z-score of 0.60 [1.27]), organization (0.31[1.05]), initiation (0.25[1.10]) and planning (0.33[1.19]), all p’s<0.001. Survivors who developed acute leukoencephalopathy displayed more neurobehavioral problems at follow-up than those who did not, adjusting for age at diagnosis and parents’ education (Table 1). Acute leukoencephalopathy was associated with reduced white matter integrity at follow-up: lower FA (p=0.03), higher AD (p=0.03) and higher RD (p=0.002). Lower FA at follow-up was associated with more neurobehavioral problems on initiation (Est -19.3, p=0.03), planning (Est -41.3, p=0.007), working memory (Est -40.6, p=0.002) and organization (Est -23.8, p=0.02). Leukoencephalopathy at follow-up was also associated with concurrent abnormalities in white matter integrity and more neurobehavioral problems on planning and organization.

Conclusions: Even without cranial radiation, approximately a quarter of ALL patients developed leukoencephalopathy during active therapy, and are at risk for long-term neurobehavioral problems and reduced white matter integrity in frontal brain regions. Survivors who develop early leukoencephalopathy may benefit from preventative cognitive and/or behavioral interventions.

Table 1

Survivor characteristics:	Acute Leukoencephalopathy N (%) / mean [SD]	No Acute Leukoencephalopathy N (%) / mean [SD]	P
Male (%)	27 (55)	62 (50)	0.55
Whites (%)	35 (71)	89 (72)	0.75
High risk (%)	23 (47)	48 (39)	0.32
Current age (years)	15.5 [4.8]	14.0 [4.6]	0.06
Age at diagnosis (years)	7.6 [5.0]	6.4 [4.0]	0.38
Time since diagnosis (years)	7.9 [2.0]	7.7 [1.7]	0.76
Total IV high-dose methotrexate (g/m2)	15.0 [4.4]	15.6 [7.4]	0.69
Total no. of intrathecal injections^	15.1 [4.0]	14.1 [4.0]	0.07
Total oral dexamethasone (mg/m2)	1066.3 [343.0]	1108.1 [286.9]	0.43
BRIEF domains:	Neurobehavioral problems Mean [SD]*
Initiation	0.46 [1.1]	0.17 [1.1]	0.04
Organization of materials	0.65 [1.0]	0.18 [1.0]	0.004
Planning	0.57 [1.1]	0.24 [1.2]	0.04
Working memory	0.74 [1.4]	0.54 [1.2]	0.16
Emotional control	0.03 [1.0]	0.08 [1.1]	0.68
Inhibition	0.06 [1.1]	0.08 [1.2]	0.74
Shift	0.16 [1.3]	-0.02 [1.1]	0.22
Monitor	0.08 [1.1]	0.02 [1.1]	0.33

*Age- and gender- adjusted scores with population mean=0 and SD=1. A higher score is indicative of more severe neurobehavioral problems.

^Inthrathecal combination of methotrexate, hydrocortisone and cytarabine