The Impact of Leukoencephalopathy on the White Matter Tracts of Long-Term Survivors of Childhood Acute Lymphoblastic Leukemia Treated with Chemotherapy Only
Noah D. Sabin1, Yin Ting Cheung2, Wilburn E. Reddick1, Deepa Bhojwani3, Wei Liu4, John O. Glass1, Tara M. Brinkman2, Scott N. Hwang1, Deokumar Srivastava4, Ching-Hon Pui5, Leslie L. Robison2, Melissa M. Hudson2, and Kevin R. Krull2

1Diagnostic Imaging, St. Jude Children's Research Hospital, Memphis, TN, United States, 2Epidemiology & Cancer Control, St. Jude Children's Research Hospital, Memphis, TN, United States, 3Children's Center for Cancer and Blood Diseases, Children's Hospital Los Angeles, Los Angeles, CA, United States, 4Biostatistics, St. Jude Children's Research Hospital, Memphis, TN, United States, 5Oncology, St. Jude Children's Research Hospital, Memphis, TN, United States


Survivors of acute lymphoblastic leukemia (ALL) can develop neurocognitive deficits and leukoencephalopathy. On-therapy and follow-up MRI examinations of the brain for 173 ALL survivors were reviewed for leukoencephalopathy. At follow-up, the survivors also underwent neurocognitive testing and brain diffusion tensor imaging (DTI). DTI parameters were associated with leukoencephalopathy in multiple regions of the brain. Although there were no associations between neurocognitive performance and leukoencephalopathy, increased mean diffusivity (MD) in certain fiber tracts was associated with neurocognitive impairment. DTI, in particular MD, may better detect loss of white matter integrity associated with neurocognitive deficits in ALL survivors than leukoencephalopathy.


Survivors of childhood acute lymphoblastic leukemia (ALL) are at risk for neurocognitive deficits and leukoencephalopathy.1-3 Longitudinal assessment of leukoencephalopathy, its association with brain diffusion tensor imaging (DTI) in major white matter tracts and its association with neurocognitive outcomes, have not been extensively examined in survivors of childhood ALL treated with chemotherapy only. We hypothesized that survivors with leukoencephalopathy would demonstrate poorer DTI indices of white matter microstructural integrity compared to those without leukoencephalopathy, and that these DTI indices would be associated with neurocognitive impairment.


One-hundred-seventy-three survivors of ALL (median [range] age 5.3 [1.0-18.2] years at diagnosis) treated on a single institutional chemotherapy-only protocol underwent brain magnetic resonance imaging (MRI) during active therapy, and repeat imaging including whole brain DTI and neurocognitive testing, at a median of 7.5 [5.1-12.5] years since diagnosis. All MRI examinations were reviewed by a board certified neuroradiologist, blinded to neurocognitive test performance and DTI results. Leukoencephalopathy was graded according to criteria of the Common Terminology Criteria for Adverse Events (version 4.0). Neurocognitive testing evaluated executive function, intelligence, processing speed, attention, memory and fine motor dexterity. Survivors who performed > 1.5 SD below age-adjusted normative data on two neurocognitive measures or >2 SD below on one measure were considered to have global neurocognitive impairment. Demographic and treatment information abstracted from patient medical records were analyzed using Chi-square and Mann-Whitney U tests for associations with the presence of leukoencephalopathy. Neurocognitive outcomes were compared to norms among all participants, and between those who presented with and without leukoencephalopathy both during active therapy and follow-up. Fractional anisotropy (FA) and mean diffusivity (MD) were compared between survivors with and without persistent leukoencephalopathy using general linear modeling, adjusted for current age, and corrected for false discovery rate within the global tracts. Only tracts that were significantly associated with leukoencephalopathy after correction for false discovery rate were tested for associations with treatment characteristics and global neurocognitive impairment.


Leukoencephalopathy was identified in 52 of 173 subjects (30.0%) during therapy and persisted at long-term follow-up in 41 of 52 (78.8%). No patient developed new leukoencephalopathy. DTI parameters were associated with the presence of persistent leukoencephalopathy in multiple regions of the brain including the corona radiata (FA p=0.001, MD p<0.001), superior longitudinal fasciculi (FA p=0.02, MD p<0.001), and superior fronto-occipital fasciculi (FA p=0.006, MD p<0.001). The total number of intrathecal administrations of chemotherapy was associated with increased MD values in several fiber tracts and with decreased FA in the anterior left corona radiata. Survivors performed worse than population norms on measures of executive function, memory and motor and visual-motor processing speed (almost all p<0.001). There were no significant associations between neurocognitive performance measures and the presence of leukoencephalopathy. However, increased MD in the genu of the corpus callosum (p=0.04), corona radiata (p=0.02), and superior fronto-occipital fasciculi (p=0.02) was associated with global neurocognitive impairment.


To our knowledge, this is the largest study of leukoencephalopathy and tractography-based examination of white matter integrity in long-term survivors of childhood ALL treated with chemotherapy only. The associations between DTI parameters in multiple supratentorial white matter tracts and the presence of leukoencephalopathy in our survivors confirms lower anisotropy, increased diffusion and, presumably, poorer structural white matter integrity in survivors with leukoencephalopathy. Despite its associations with decreased FA and increased MD, leukoencephalopathy was not significantly associated with neurocognitive test performance while MD was associated with neurocognitive function. The fact that only MD, and not FA, was associated with neurocognitive deficits suggests that the amount of diffusion, rather than the degree of anisotropy in white matter has a greater effect on neurocognitive abilities. The larger number of associations between MD in multiple fiber tracts and the total number of intrathecal administrations of chemotherapy, compared with FA, also suggests that chemotherapy has a greater effect on the amount of diffusion in those tracts than the degree to which it is anisotropic.


DTI may serve as a more sensitive imaging technique for detecting loss of white matter integrity that is associated with neurocognitive deficits in ALL survivors than the presence of leukoencephalopathy on standard MRI sequences. MD, in particular, could be a potential measure for identifying patients at risk for neurocognitive impairment.


No acknowledgement found.


1 Pui CH, Campana D, Pei D, et al. Treating childhood acute lymphoblastic leukemia without cranial irradiation. N. Engl. J. Med. 2009;360(26):2730-2741.

2 Bhojwani D, Sabin ND, Pei D, et al. Methotrexate-induced neurotoxicity and leukoencephalopathy in childhood acute lymphoblastic leukemia. J. Clin. Oncol. 2014;32(9):949-959.

3 Reddick WE, Glass JO, Helton KJ, et al. Prevalence of leukoencephalopathy in children treated for acute lymphoblastic leukemia with high-dose methotrexate. AJNR Am. J. Neuroradiol. 2005;26(5):1263-1269.

Proc. Intl. Soc. Mag. Reson. Med. 26 (2018)