Diffusion tensor imaging and neurocognition in survivors of childhood acute lymphoblastic leukaemia

Abstract

Survivors of childhood acute lymphoblastic leukaemia are at risk for neurocognitive impairment, though little information is available on its association with brain integrity, particularly for survivors treated without cranial radiation therapy. This study compares neurocognitive function and brain morphology in long-term adult survivors of childhood acute lymphoblastic leukaemia treated with chemotherapy alone (n = 36) to those treated with cranial radiation therapy (n = 39) and to healthy control subjects (n = 23). Mean (standard deviation) age at evaluation was 24.9 (3.6) years for the chemotherapy group and 26.7 (3.4) years for the cranial radiation therapy group, while time since diagnosis was 15.0 (1.7) and 23.9 (3.1) years, respectively. Brain grey and white matter volume and diffusion tensor imaging was compared between survivor groups and to 23 healthy controls with a mean (standard deviation) age of 23.1 (2.6) years. Survivors treated with chemotherapy alone had higher fractional anisotropy in fibre tracts within the left (P < 0.05), but not in the right, hemisphere when compared to controls. Survivors of acute lymphoblastic leukaemia, regardless of treatment, had a lower ratio of white matter to intracranial volume in frontal and temporal lobes (P < 0.05) compared with control subjects. Survivors of acute lymphoblastic leukaemia treated with chemotherapy alone performed worse in processing speed (P < 0.001), verbal selective reminding (P = 0.01), and academics (P < 0.05) compared to population norms and performed better than survivors treated with cranial radiation therapy on verbal selective reminding (P = 0.02), processing speed (P = 0.05) and memory span (P = 0.009). There were significant associations between neurocognitive performance and brain imaging, particularly for frontal and temporal white and grey matter volume. Survivors of acute lymphoblastic leukaemia treated with chemotherapy alone demonstrated significant long-term differences in neurocognitive function and altered neuroanatomical integrity. These results suggest substantial region-specific white matter alterations in survivors of acute lymphoblastic leukaemia possibly resulting in restricted radial diffusion due to the compaction of neuronal fibres.

Keywords: anisotropy; brain fraction; cancer; neuroimaging; neuropsychology.

Figure 1

Hippocampal volume and shape in survivors treated with CRT or chemotherapy alone compared to controls. (A) Mean and 95% confidence intervals for left and right hippocampal volumes by group. (B) Differences in hippocampal shape between controls and ALL survivors treated with CRT or chemotherapy alone. The T-value colour map displays positive and negative T-values representing the outward and inward shape differences.

Figure 2

Comparison of fibre tract fractional anisotropy between controls and survivors treated with CRT or chemotherapy alone. Mean and 95% confidence intervals for fibre tract fractional anisotropy. Statistically significant differences between a survivor group and controls are identified by an asterisk (*P < 0.05). Statistically significant differences between survivor groups were identified by a hash symbol (^#P < 0.05). Fractional anisotropy of fibre tracts primarily located in cortical (A) or subcortical (B) regions are plotted by hemisphere. Sup = superior; Fr-occip = fronto-occipital; Fasc = fasciculus; Long = longitudinal; Inf = inferior; Cereb = cerebellar; Ped = peduncle, Med = medial; Sag = sagittal; Post = posterior; Thal = thalamic; Rad = radiation; Mid = middle.