Skip to main content
NCBI home page
Human Brain Mapping logoLink to Human Brain Mapping
. 2018 Apr 20;39(8):3375–3387. doi: 10.1002/hbm.24082

Advanced MR diffusion imaging and chemotherapy‐related changes in cerebral white matter microstructure of survivors of childhood bone and soft tissue sarcoma?

Charlotte Sleurs 1,2,3,✉, Jurgen Lemiere 1, Daan Christiaens 4, Thibo Billiet 5, Ronald Peeters 2, Stefan Sunaert 2,6, Anne Uyttebroeck 1,3, Sabine Deprez 2,6
PMCID: PMC6866599  PMID: 29675944

Abstract

With the increase of survival rates of pediatric cancer patients, the number of children facing potential cognitive sequelae has grown. Previous adult studies suggest that white matter (WM) microstructural changes may contribute to cognitive impairment. This study aims to investigate WM microstructure in childhood bone and soft tissue sarcoma. Differences in (micro‐)structure can be investigated using diffusion MRI (dMRI). The typically used diffusion tensor model (DTI) assumes Gaussian diffusion, and lacks information about fiber populations. In this study, we compare WM structure of childhood bone and soft tissue sarcoma survivors (n = 34) and matched controls (n = 34), combining typical and advanced voxel‐based models (DTI and NODDI model, respectively), as well as recently developed fixel‐based models (for estimations of intra‐voxel differences, apparent fiber density [AFD] and fiber cross‐section [FC]). Parameters with significant findings were compared between treatments, and correlated with subscales of the WAIS‐IV intelligence test, age at diagnosis, age at assessment and time since diagnosis. We encountered extensive regions showing lower fractional anisotropy, overlapping with both significant NODDI parameters and fixel‐based parameters. In contrast to these diffuse differences, the fixel‐based measure of AFD was reduced in the cingulum and corpus callosum only. Furthermore, AFD of the corpus callosum was significantly predicted by chemotherapy treatment and correlated positively with time since diagnosis, visual puzzles and similarities task scores. This study suggests altered WM structure of childhood bone and soft tissue sarcoma survivors. We conclude global chemotherapy‐related changes, with particular vulnerability of centrally located WM bundles. Finally, such differences could potentially recover after treatment.

Keywords: childhood solid tumors, diffusion‐weighted imaging, non‐CNS‐directed chemotherapy, white matter

Abbreviations

AFD

apparent fiber density

DMRI

diffusion weighted magnetic resonance imaging

DTI

diffusion tensor model

FA

fractional anisotropy

FBA

fixel‐based analysis

FC

fiber cross‐section

FOD

fiber orientation distribution

NODDI

neurite orientation dispersion and density index

NDI

neurite density index

ODI

orientation dispersion index

VISO

isotropic volume fraction

WM

white matter

1. INTRODUCTION

As survival rates increase for oncological patients, potential late adversive effects of the treatment become more and more important to address (Krull et al., 2008; Kunin‐Batson, Kadan‐Lottick, & Neglia, 2014). In children specifically, brain development could be challenged during cancer, due to the disease and/or treatment. Hence, it becomes essential to trace potential neurotoxic mechanisms. For pediatric oncology, it is well‐known that cranial irradiation leads to cognitive deficits in brain tumor patients (Chang et al., 2009; McDuff et al., 2013) and leukemia patients (Buizer Buizer, De Sonneville, Van Den Heuvel‐Eibrink, & Veerman, 2005; Giralt et al., 1992; Krull et al., 2013; Sleurs et al., 2016, 2016b). Consequently, treatment for leukemia patients was replaced by CNS‐directed chemotherapy.

Still, evidence also exists for cognitive decline due to CNS‐directed chemotherapy, even without radiotherapy (Khong et al., 2006; Krull et al., 2013; and Moleskiski, 2000). Most difficulties were reported for executive functioning (Buizer et al., 2005; Caron et al., 2009; Kadan‐Lottick et al., 2010). In line with these cognitive studies, recent neuroimaging studies have also evidenced white matter (WM; Edelmann et al., 2014; Elalfy et al., 2014; Kesler, Gugel, Huston‐Warren, & Watson, 2016; Schuitema et al., 2012; Schuitema et al., 2013) and grey matter changes due to CNS‐directed chemotherapy (Koppelmans, Breteler, Boogerd, Seynaeve, & Schagen, 2013; Krull et al., 2016; Tamnes et al., 2015). In contrast to the large number of reports about late neurocognitive sequelae in leukemia (van der Plas et al., 2015), only limited research has been conducted on childhood solid non‐CNS tumors (Sleurs et al., 2016a). Recently, a few studies have reported cognitive outcomes of childhood solid non‐CNS tumor survivors. Mohrmann and colleagues have shown that one‐third of solid non‐CNS tumor patients experience cognitive complaints later in life (Mohrmann et al., 2015). For rhabdomyosarcoma, a recent survivor study also evidenced higher risk for depression and task efficiency problems (Schapiro et al., 2015). Patients using psychopharmaca for anxiety or depression appeared to be most vulnerable. In addition, a recent study performed by Edelman and colleagues, demonstrated lower reading skills, attention, memory and processing speed in survivors of osteosarcoma (Edelmann et al., 2016).

Not only the number of behavioral studies remains limited for these patients, neuroimaging studies in childhood non‐CNS solid tumor survivors have not been performed yet. Still, adult studies have reported alterations of the WM microstructure in solid tumor patients (including breast and testicular cancer; Ahles et al., 2014; Amidi et al., 2016; Deprez et al., 2011; Deprez, Billiet, Sunaert, & Leemans, 2013; McDonald et al., 2010). Nowadays, MR diffusion‐weighted neuroimaging is used to characterize the WM microstructure, and assess possible changes after treatments. These studies typically implement the diffusion tensor imaging (DTI) model, modeling the diffusion of water molecules in each voxel as a 3D Gaussian distribution (Madden, Bennett, & Song, 2009).

In the cancer patient groups previously described, DTI‐derived parameters such as fractional anisotropy (FA) or mean diffusivity often showed differences in multiple regions, scattered throughout the brain (Abraham et al., 2008; Koppelmans et al., 2014; Partridge et al., 2010; De Ruiter et al., 2012; Schuitema et al., 2013). However, the cellular microstructure in biological tissue can lead to hindered, non‐Gaussian diffusion, for which the DTI model is inadequate. Hence, advanced models should include information of more specific microstructural compartments of the tissue. One of these compartment models is Neurite Orientation Dispersion and Density Imaging (NODDI; Zhang et al., 2012). This is a three‐compartment model of WM microstructure: intra‐axonal space, extra‐axonal space, and the cerebrospinal fluid. According to the model, water movement in the intra‐axonal space depends on the orientation dispersion of tracts. Anisotropic and isotropic Gaussian distributions of movement are assumed for the extra‐axonal space and cerebrovascular fluid, respectively. Using these biophysical assumptions, the NODDI model derives parameters for Neurite Density (NDI), orientation dispersion index (ODI) and isotropic volume fraction (VISO) to characterize WM microstructure.

Still, these parameters are voxel‐based, while WM voxels are typically occupied by multiple fiber populations, resulting in crossing fibers (Jeurissen et al., 2013). In other words, changes measured by voxel‐based parameters (i.e., based on the DTI or NODDI model), are not fiber‐specific. To overcome this limitation, Raffelt and colleagues have recently developed the so‐called “Fixel‐based” analysis (Raffelt et al., 2016). Instead of calculating voxel‐based parameters, the concept “fixel” was introduced to refer to individual fiber populations within a voxel, segmented from the fiber orientation distribution (FOD) within a voxel. Hence, fiber‐specific WM microstructure can be modeled in each direction, and provide more detailed information about specific fiber bundles. Specifically, fixel‐based parameters can model changes at both microstructural level (e.g., apparent fiber density, AFD) and macroscopic level (e.g., fiber cross‐section, FC).

In this study, we compare WM microstructure between childhood bone and soft tissue sarcoma survivors and healthy matched controls, by using both fixel‐based and voxel‐based analyses. In addition, we investigate the link between IQ subscales and diffusion metrics. This way, we aim to investigate the WM microstructure as underlying neural substrate for cognitive functioning in more detail. We hypothesize that if WM development is affected by chemotherapy during childhood, differences are expected in the diffusion‐derived parameters. Specifically, if WM differences are different at macroscopic versus microscopic level, the fixel‐based analysis (FBA) could demonstrate such distinctions and show fiber‐specific changes.

2. METHODS

2.1. Participants

We acquired dMRI in 34 survivors of pediatric solid non‐CNS tumors and 34 healthy age‐ and gender‐matched controls. The patient group consisted of survivors of bone and soft tissue sarcomas (n = treated according to the chemotherapy protocols EURAMOS1 (n = 13), RMS2005 (n = 3), NRSTS2005 (n = 1), Euro‐Ewing99 (n = 7), MMT95 (n = 5), EORTC2001 (n = 3), IVAD (n = 1), which included intravenous chemotherapy only (1 survivor was excluded due to cranial irradiation). Adult survivors were included at least 2 years after treatment at the University Hospitals Leuven. Patients were excluded in case of mental retardation, or psychiatric complaints in daily life (hyperactivity, depression, anxiety, etc.) requiring psychopharmacological treatment. The age at diagnosis ranged between 4.00 and 17.75 years (mean = 12.97 years, SD = 3.33). Healthy age‐matched control participants were recruited using local electronic advertisements. All participants were aged between 16.14 and 35.28 years (mean = 22.51, SD = 3.97) at the moment of acquisition. Time since diagnosis ranged between 2.08 and 19.92 years (mean = 9.92, SD = 4.72). The study was approved by the Ethical Committee of University Hospitals Leuven and conducted according to the Declaration of Helsinki.

2.2. Data acquisition

Images were acquired on a 3T Philips Achieva MRI scanner with a 32‐channel phased‐array head coil. The echo‐planar, multi‐shell diffusion imaging scheme consisted of b values 700, 1000, and 2800 s/mm2, applied along 25, 40, and 75 uniformly distributed gradient directions respectively, in addition to 10 non‐diffusion‐weighted images (b = 0). This multi‐shell scheme including a high b value, allows us to apply the advanced diffusion models. For all diffusion series, the following acquisition parameters were constant: TR/TE = 7800/90 ms, 50 slices, voxel resolution of 2.5 × 2.5 × 2.5 m. The total acquisition time for the diffusion‐weighted scans was ∼25 min.

In addition to the neuroimaging data, the WAIS‐IV intelligence test was acquired in all subjects on the same day. The main subscales were investigated as neurocognitive outcome of interest, including Verbal Comprehension Index subscales (i.e., Similarities, Vocabulary, Information), Perceptual Reasoning Index (i.e., Block Design, Matrix Reasoning, Visual Puzzles), Working Memory (i.e., Digit Span and Arithmetic) and Processing Speed (i.e., Symbol Search and Coding). Descriptive statistics of the main scales for this population and included data, see Supporting Information Table S1).

2.3. Data processing

To investigate the WM of survivors in detail, both voxel‐based and fixel‐based analyses are used. For a schematic overview of these analyses, see Figure 1.

Figure 1.

Figure 1

Open in a new tab

Schematic overview of voxel‐based and fixel‐based parameter maps that were calculated for all participants. NDI = neurite density index, ODI = orientation dispersion index, VISO = isotropic volume fraction, FA = fractional anisotropy. For fixel‐based analyses, FODs = fiber orientation distributions are used to calculate fixel‐based parameters including: FC = fiber cross‐section, AFD = apparent fiber density [Color figure can be viewed at http://wileyonlinelibrary.com]

2.4. Voxel‐based analyses

DWI preprocessing including motion‐ and eddy current induced distortion correction was performed using ExploreDTI (Leemans, Jeurissen, Sijbers, & Jones, 2009), DWI bias field correction (Tustison et al., 2010) and global intensity normalization (Raffelt et al., 2012) was performed using MRtrix3 (Tournier, Calamante, & Connelly, 2012). The corrected DWI images were then fitted with the (a) diffusion tensor model (DTI) to generate FA maps using MRtrix3 and (b) the NODDI model using the NODDI Matlab toolbox (Zhang et al., 2012), resulting in three additional parameter maps: NDI, VISO, ODI.

The corrected DWI images were then used to compute FODs, using robust constrained spherical deconvolution (Tournier et al., 2013) with a group average response function (based on the highest shell data [b = 2,800]). To be able to make interindividual comparisons at group‐level, individual maps should be registered to the same space. Hence, individual FODs were registered diffeomorphically to a population‐based FOD‐atlas (Raffelt et al., 2011), which was created based on 20 individual FOD‐maps (10 patients and 10 controls; mean age of subgroup: 20.83 years old [range 16.14–25.69]; see Figure 2). This selection was required to limit the duration of the atlas calculation. Transformation fields of these registrations were applied to individual FA‐, NDI‐, ODI‐, and VISO‐maps.

Figure 2.

Figure 2

Open in a new tab

Population‐based atlas of FODs. To be able to make interindividual comparisons, individual FODs were calculated, using robust constrained spherical deconvolution (Tournier, Calamante, & Connelly, 2013). Individual FOD maps were diffeomorphically registered to the population‐based FOD‐atlas [Color figure can be viewed at http://wileyonlinelibrary.com]

2.5. Fixel‐based analyses

Fixel‐based parameters were calculated as described by Raffelt and colleagues, using MRtrix3 (Raffelt et al., 2015). FODs were used to segment the fiber populations (fixels) in each voxel. More specifically, the first fixel‐based parameter, AFD, is calculated as the integral of each FOD lobe (Raffelt et al., 2012). This measure is correlated with the relative intra‐axonal volume of fibers along the main direction of that lobe. The second fixel‐based parameter, fiber bundle cross‐section (FC), is a macroscopic measure representing changes in the number of voxels that a specific tract bundle occupies. FC is computed based on the non‐linear FOD registration step, as the cross‐sectional change in the plane perpendicular to the fixel orientation. Smoothing and cluster enhancement was applied in a fixel‐based manner, instead of voxel‐based (Raffelt et al., 2015).

2.6. Statistics

WM microstructure was compared between the solid non‐CNS tumor survivor group and healthy controls by using multiple t tests (p < .05). Threshold‐free cluster enhancement was applied for voxel‐based parameters (i.e., FA‐, ODI‐, NDI‐, and VISO‐ maps; Smith, & Nichols, 2009) and connectivity‐based fixel enhancement for measures of FD and FC in all WM fixels (i.e., AFD and FC). 12 t tests tests were performed (i.e., FA, ODI, NDI, VISO, AFD, and FC [6] × patients > controls and vice versa [2]). p values were calculated for each fixel using non‐parametric permutation testing (5,000 permutations), and family‐wise error corrected for multiple comparisons. Finally, stringent bonferroni‐correction (p < .05/12) was applied to investigate most significant differences more in detail. In Figure 3, we depict the expected value changes in case of specific WM changes.

Figure 3.

Figure 3

Open in a new tab

Schematic presentation of potential WM changes after therapy and expected related parameter changes. In case of chemotherapy‐induced WM microstructural damage, parameters derived from diffusion models, are expected to change. In voxel‐based analyses, typically lower FA (DTI model), lower NDI and higher VISO (NODDI model) are expected. In fixel‐based analysis, FC and AFD are expected to change, according to macroscopic and microscopic changes, respectively. NDI = neurite density index, ODI = orientation dispersion index, VISO = isotropic volume fraction, FA = fractional anisotropy, FC = fiber cross‐section, AFD = apparent fiber density [Color figure can be viewed at http://wileyonlinelibrary.com]

In addition, to investigate the link between clinical data and diffusion parameters, Pearson correlation analyses were performed. More specifically, the region showing most significant differences in diffusion parameters was selected for further analyses with clinical measures within the patient group. The most significant parameter in this region was correlated with age at diagnosis, time since treatment, age at assessment, task scores on the subscales of the WAIS‐IV assessment (Matrix Reasoning [MR], Visual Puzzles [VP], Similarities [S], Vocabulary [V], Information [I], Digit Span [DS], Arithmetic [A], Symbol Search [SS], Coding [C]), and was compared between chemotherapy protocols (SPSS toolbox, v.23).

3. RESULTS

3.1. Voxel‐based analyses

Voxel‐based analyses based on the NODDI model, demonstrated significant group differences for the NDI and VISO. More specifically, elevated VISO was encountered distributed throughout the central WM in the patient group, even after Bonferroni‐corrections. These regions surround a small region of higher NDI in the corticospinal tract. Additionally, FA is lower in similar WM regions (see Figure 4).

Figure 4.

Figure 4

Open in a new tab

Voxelwise group comparisons. Regions showing significant different WM paramaters in bone and soft tissue sarcoma survivors when compared to healthy matched controls using voxel‐wise permutation testing with threshold‐free cluster enhancement (FWE‐corrected p < .05). Upper panel shows original results. Lower panel shows results after stringent additional Bonferroni corrections (adapted for 12 tests). Significant higher NDI and VISO in patients is depicted in orange and purple, respectively. Lower FA is shown in green [Color figure can be viewed at http://wileyonlinelibrary.com]

3.2. Fixel‐based analyses

Given that the typical single‐tensor (i.e., DTI) measures can be affected by the amount of fiber populations in the voxels, FBA was performed. Similar to the widespread differences of FA and VISO, FC is also lower throughout the brain in the patient group. As shown in Figure 5, this measure overlaps with lower FA, higher VISO and higher NDI values. By contrast, AFD is lower in specific tracts, including the corpus callosum, the cingulum and a small part of the corticopontine tract. In conclusion, the scattered differences of FA and VISO overlap with both fixel parameters FC and fiber density, but differences in AFD seem limited to single tracts (see Figure 6). AFD in the corpus appeared most significantly different, which remained after Bonferroni corrections. Finally, the fixel‐based and voxel‐based parameters indicated additional significant regions, which were not detected by the other modality (i.e., voxel‐based and fixel‐based, respectively).

Figure 5.

Figure 5

Open in a new tab

Results of fixel‐based group comparisons of AFD and FC. Regions with lower AFD in the corpus callosum and cingulum of patients compared to controls (p < .05) are depicted in orange. Lower FC is presented in blue. In the lower panel, results of the group comparisons are shown after stringent Bonferroni corrections (adapted for 12 tests) [Color figure can be viewed at http://wileyonlinelibrary.com]

Figure 6.

Figure 6

Open in a new tab

Results of fixel‐based group comparison of AFD between survivors and control participants. (a) Lower AFD in survivors is depicted (colored by direction: left‐right, anterior‐posterior and superior–inferior are represented by red, green, blue, respectively; p < .05). (b) A mask was created in the corpus callosum, for the significant region of AFD only, in order to calculate mean AFD values in this region for all subjects. These mean values were correlated with timing variables (time since diagnosis, age at diagnosis and age at assessment) [Color figure can be viewed at http://wileyonlinelibrary.com]

3.3. WM structure, intellectual outcome, therapy dose, and timing

The region showing significant differences (p < .05) in the most diffusion parameters, was the posterior part of the corpus callosum body (Figures 5 and 6). AFD was most significant in this region (most significant fixels in this region showed p < .01). Hence, this parameter was calculated for each individual within this region. The mean AFD was specifically calculated within the central part of the corpus callosum (coronal view), which showed significantly lower AFD. Probabilistic fiber tracking in this region was performed, based on the FOD template, to use as mask (see Figure 6). This specific AFD measure was compared between chemotherapy protocols (using a univariate ANOVA) and used to correlate with timing variables: age at diagnosis, time since treatment and age at assessment, as well as with the subscales of the WAIS‐IV. First, AFD was significantly predicted by the treatment protocol (F(7,1) = 3.080, p = .008). More specifically, AFD appeared lowest for patients treated according to the EURAMOS1 and RMS05 treatment (see Figure 7). However, for specific chemotherapy doses within these protocols, no clear associations were encountered (see Supporting Information Figure S1).

Figure 7.

Figure 7

Open in a new tab

AFD, treatment protocol and cognitive outcomes. (a) Bar plots with average AFD values for the control group in blue (n = 33). The red bars represent equivalent values for patients, treated according to the following chemotherapy protocols: RMS05 (n = 3), EURAMOS1 (n = 13), MMT95 (n = 5), EuroEwing99 (n = 7), EORTC2001 (n = 3), NRSTS05 (n = 1), IVAD (n = 1). (b) Scatter plots indicating the significant linear associations between IQ subscales and AFD, for both controls and patients [Color figure can be viewed at http://wileyonlinelibrary.com]

Second, correlations were calculated within the patient group between mean AFD in the CC, timing variables and IQ subtask scores. Mean AFD did not correlate significantly with age at diagnosis, age at assessment (r = −.068, p = .708; r = .030, p = .809, respectively; see Figure 8), nor with most subscales of the WAIS (see Supporting Information Table S2). In contrast, time since diagnosis and task scores on visual puzzles and similarities were positively correlated with mean AFD (r = .366, p = .036; r = .262, p = .032; r = .288 p = .018, respectively; see Figure 7).

Figure 8.

Figure 8

Open in a new tab

Scatter plots depicting the relationship between timing variables and AFD in the corpus callosum. Timing variables including time since diagnosis, age at diagnosis and age at assessment are depicted against AFD that was measured in the significant region within the corpus callosum [Color figure can be viewed at http://wileyonlinelibrary.com]

4. DISCUSSION

This study showed the first findings of potential WM alterations in childhood solid non‐CNS tumor survivors, that is bone and soft tissue sarcomas, on average 10 years after chemotherapy‐treatment, when compared to healthy controls. This was the first childhood cancer survivor study investigating long‐term microstructural changes using advanced imaging methods. By expanding the standard voxel‐based analyses of FA (i.e., based on the DTI model) to NODDI and FBA, distinctions could be made between different origins of WM changes. This study showed widespread differences between patients and controls, with lower values of FA (DTI), increased VISO (NODDI) and reduced FC in patients, while lower AFD appeared only in specific WM tracts in patients (FBA). However, after stringent Bonferroni corrections, only widespread VISO and AFD in the corpus callosum remained significant. The fiber density measure also appeared to be related to time since diagnosis and verbal and visual subtask scores. Using these advanced methods, we demonstrated microstructural changes even in cases who did not show WM lesions.

4.1. Widespread differences in voxel‐based measures

First, voxel‐based comparison of FA‐maps showed lower FA in multiple brain regions in survivors when compared to matched controls. This finding is in line with previous findings in cancer survivors (Abraham et al., 2008; Deprez et al., 2011; Simó, Rifà‐Ros, Rodriguez‐Fornells, & Bruna, 2013) describing reduced FA linked with impaired cognitive performance suggestive for decreased WM integrity. In earlier studies, regions showing lower FA values were also heterogenous, suggesting widespread WM microstructural changes. One hypothesis of chemotherapy‐induced neurotoxicity is accelerated aging of fiber populations due to chemotherapy (Schuitema et al., 2012). Healthy aging studies typically report decreases in FA values throughout lifespan after the age of 30 years (Kochunov et al., 2012). The differences in FA did not remain after Bonferroni correction. However, this correction could be a rather stringent method in this case, given that the NODDI analyses showed an overlap between lower FA and higher VISO. This latter measure clearly remained significant after corrections. A recent study of Billiet and colleagues yielded similar results related to aging throughout lifespan, with increases in VISO and NDI and decrease in FA (Billiet et al., 2015). Similar to our study, they did not encounter differences in ODI. Since ODI did not differ between our survivor group and controls, we assume that the diffuse pattern of higher VISO and lower FA is not due to increased neurite dispersion but due to the relative amount of fluid in the extra‐cellular space. Such elevated levels of extra‐cellular fluid could suggest widespread changes due to WM atrophy or cerebral edema. Periventricular changes related to fluid retention, could also explain altered NDI of the corticospinal tract. Neurotoxic mechanisms including ischemic hypoxia, encephalopathy, cardiac arrest, inflammatory processes, etc. (Abbott, 2004) could result in elevated interstitial fluid. Although the VISO parameter suggests increased interstitial fluid, voxel‐based parameters are affected by multiple tissue changes, including axonal loss, axonal density, axon diameter, myelination, gliosis etc., which cannot be easily disentangled (Concha, 2014). Furthermore, in crossing fiber regions, the degeneration of tracts can even lead to higher anisotropy, which makes the interpretation of voxel‐based measures difficult. Tensor‐derived metrics, such as FA are also more affected by periventricular partial volume effects given the remaining csf signal in lower b values. Given that only the higher shell b = 2800 data were used for CSD‐derived metrics, these effects should be smaller. Therefore, we assume the fixel‐based measure to be less dependent on the overall amount of WM, and partial volume fractions. By using fixel‐based analyses, we could provide essential additional information in this study.

4.2. Distinction between global and local differences in fixel‐based measures

In accordance with the diffuse differences in VISO and FA, the fixel‐based measure of the fiber‐cross‐section also appeared to be widely affected. The lower FC values support the hypothesis of decreased axon bundle thickness, which could be due to atrophy or interstitial fluid retention. In contrast, with regard to the microstructural measure of AFD, differences were encountered only in the single tracts of the corpus callosum, the cingulum and the corticopontine tract. Since AFD is different in specific fiber populations, one could hypothesize that these centrally‐located tracts are specifically vulnerable for microstructural changes in pediatric chemotherapy. The vulnerability of central tracts could be explained by their dense axonal packing and their high vascular supply (Garg et al., 2015). More specifically, after Bonferroni corrections, AFD appeared only significant in the corpus callosum. As described by Garg and colleagues (2015), the corpus callosum would be especially vulnerable to demyelination and inflammatory processes, which could reduce the intra‐axonal compartment volume fraction in these regions. This region was already suggested to be vulnerable to toxicity earlier due to chemotherapy (Deprez et al., 2013), but also due to alcohol abuse as well (Kapogiannis et al., 2012; Smith et al., 2017). Such specific sensitivity of the corpus callosum could explain the observed decrease in AFD in this region. Although metrics derived from higher shell data (b = 2800) are less affected by CSF volume effects, one should note that highly variable partial volumes in small structures might affect significance levels. By contrast, larger structures such as the corpus callosum would be less affected by such volume effects (Vos, Jones, Viergever, & Leemans, 2011).

4.3. Microstructural differences, intelligence scores, chemotherapy, and timing

Finally, the specific measure of fiber density in the corpus callosum was associated with chemotherapy protocol, time since diagnosis and intelligence subtask scores. Specifically, survivors treated according to the EURAMOS1 and RMS05 protocol, showed lower apparent fiber densitiy in the corpus callosum. Similarly, Stouten‐Kemperman and colleagues (Stouten‐Kemperman et al., 2015) showed that patients exposed to higher doses of chemotherapy are probably more vulnerable to neurotoxicity. However, the protocols included in our study prescribe different chemotherapy agents. The first protocol mainly includes methotrexate, cisplatin and doxorubicin (Bielack et al., 2015; Whelan et al., 2015), whereas the second protocol mainly prescribes ifosfamide, vincristine and actinomycin (Ferrari et al., 2017). As our survivor group was limited to 33 subjects, this study was not able to investigate the effects of specific agents included in these protocols. Multiple chemotherapy protocols were included in this study, which resulted in heterogeneity in the patient group. Still, all treatments included at least one high‐dose chemotherapy agent (methotrexate and/or ifosfamide). Furthermore, as higher chemotherapy doses are administered in case of larger, more aggressive tumors, or metastases, these factors could also indirectly be related to potential neurotoxicity. Nevertheless, additional figures did not show clear trends in AFD related to cumulative chemotherapy doses (see Supporting Information Figure S1). This suggests that neurotoxicity probably arises in case of certain combinations of agents, and depending on an individual's vulnerability, rather than due to only the dose of one specific agent. In addition, the significant group difference in AFD was measured between survivors and control participants, at group level. During recruitment of patients, 52% of the existing database of patients (inclusion criterium: treated according to the predefined chemotherapy protocols, >2 years out of treatment) participated. Given that times since diagnosis were similar between the participating and non‐participating patients, we assume that this dataset was representative for all patients. Nonetheless, we note that the ratio of osteosarcoma patients was higher in participating than in non‐participating patients. Given that these patients often receive high‐dose chemotherapy, we cannot exclude this higher ratio to have affected our results. Furthermore, although this study showed group‐based effects for bone and soft tissue sarcoma patients, larger samples are necessary to investigate vulnerability of specific subgroups and differentiate between specific chemotherapy agents. At this stage, it remains uncertain which individual factors play a role in therapy‐induced toxicity affecting the WM. Besides neurotoxicity, also differences in SES could have an important impact on brain development during childhood. In this regard, one should note that SES scores were significantly lower in patients. This might highly affect the group comparison results. Although it is complicated to investigate the independent effect of this measure on imaging values or any cognitive value, since these are highly correlated, additional analyses showed that significant group differences in diffusion metrics still remain after statistical correction for SES. With regard to cognitive measures, visual puzzles and similarities subtask scores were positively correlated with corpus callosal fiber density. Both these subtasks require matching skills and logical reasoning. Puzzles require visuospatial matching and manual coordination. Previous studies have clearly indicated the important role of the corpus callosum in bimanual coordination (Gooijers & Swinnen, 2014) . By contrast, the ability to find similarities between words depends more on conceptual matching (cognitive semantics). As our measure of AFD was derived from the posterior part of the corpus callosum, we mainly investigated the link between parietal and occipital lobe connectivity and intellectual outcome. This region could play an important role in conceptual and visual matching skills. However, we cannot exclude that the global macrostructural changes (as measured with VISO, FA, and FC) can affect multiple other cognitive domains as well.

Finally, we encountered a significant correlation between time since diagnosis and AFD of the corpus callosum, suggestive of a recovery pattern after treatment. In this respect, Armstrong and colleagues recently summarized the processes of axon regeneration and remyelination in case of brain injury (Armstrong et al., 2016a). They demonstrate the natural tendency of the WM to remyelate after inflammation or vascular damage (Armstrong et al., 2016b). Since chemotherapy could induce such neurotoxic processes (Ahles & Saykin, 2007; Sleurs et al., 2016a), leading to microstructural changes, recovery processes could arise in childhood cancer patients as well. Furthermore, recently a similar potential recovery was detected in FA values 3 years after treatment in breast cancer patients (Billiet et al., 2015). This would explain a positive trend in AFD throughout time. Although a positive trend with age at diagnosis was expected as well, this correlation was not significant. Still, other pediatric oncology (i.e., brain tumor and leukemia) studies also evidenced higher neurotoxic vulnerability of the younger patients at diagnosis (Buizer et al., 2005; Caron et al., 2009; Sleurs et al., 2016b; von der Weid et al., 2003). Patients in our study were on average 13 years old at diagnosis. Although the age at diagnosis ranged between 5 and 18 years, the majority of the patients were older than 10 years at diagnosis (91%). It can be assumed that underrepresentation of this younger patient group would explain the lack of correlation between age at diagnosis and AFD. If future studies would show similar lack of correlation between age at diagnosis and diffusion parameters in younger patients at diagnosis as well, recovery processes could provide a possible explanation.

4.4. Future directions

By using multiple analysis methods, we aimed to distinguish between several potential tissue changes. First, based on voxel‐based analyses, we showed elevated isotropic fractions and lower FA, presumably due to interstitial edema. Second, based on the fixel‐based analyses, we showed diffuse macroscopic WM morphological changes (as measured by FC), while microscopic changes (as measured AFD) were limited to densely packed WM tracts. In summary, these findings suggest that densely packed WM tracts are mainly vulnerable for microscopic changes, while macrocopic changes are diffuse. Both types of WM changes could lead to lowered ability to transfer information efficiently. In this respect, lower FA values were related to lower processing speed (Abraham et al., 2008) and attentional functioning (Billiet et al., 2015; Deprez et al., 2011) in breast cancer patients. To investigate which type of WM change leads to specific cognitive difficulties, and processing speed in particular, future studies will be necessary including behavioral measurements. Furthermore, this is the first pilot study showing potential long‐term WM microstructural changes in children treated with chemotherapy, using advanced diffusion imaging. Hence, this patient group covered a variety of subgroups of patients (i.e., bone as well as soft tissue sarcoma), treated according to different chemotherapy protocols. In this study, we did not report a relationship between chemotherapy dose and diffusion metrics. Since exploratory analyses did not show differences in AFD of the corpus callosum according to the cumulative doses of individual chemotherapy agents (i.e., methotrexate, cisplatin, carboplatin, doxorubicin, epirubicin, or ifosfamide), statistical analyses were not included in this study to investigate this in detail. Additional diffusion‐weighted MRI studies will be necessary to draw firm conclusions about potential predictors for neurotoxic vulnerability. Although this prospective study suggests potential WM changes due to treatment, the underlying physiological processes can not be identified using dMRI alone. Not only specific chemotherapeutic agents could be more toxic, also individual patient characteristics such as therapy metabolism, cardiotoxicity, and genetic predisposition should be further investigated. Finally, we suggest to implement longitudinal studies to investigate potential recovery processes more in detail, since this study only included a cross‐subject correlation between time since diagnosis and AFD.

5. CONCLUSION

In summary, our study provides the first evidence of changes in the WM structure of childhood survivors of bone and soft tissue sarcoma.We showed widespread differences in diffusion parameters, suggesting macroscopic changes. Furthermore, microscopic differences in centrally‐located tracts could suggest specific vulnerability to vascular or inflammatory mechanisms after disease or treatment, associated with chemotherapy protocol, visual and verbal matching task scores, and time since diagnosis. Advanced diffusion models, such as fixel‐based analyses bring additional insights in the mechanisms of disease‐ or therapy‐induced neurotoxicity and enabled us to differentiate between WM toxicity at the macroscopic versus microscopic level. Still, histological animal and post‐mortem studies are necessary to complement these findings.

CONFLICTS OF INTEREST

The authors do not have any conflict of interest to declare.

Supporting information

Additional Supporting Information may be found online in the supporting information tab for this article.

Supporting Information 1

Click here for additional data file. (143.6KB, docx)

ACKNOWLEDGMENTS

First, we thank Iris Elens for her help with the data collection for this study, and for her medical insights on this topic. Second, we are very grateful to Dr. Donald Tournier and Dr. David Raffelt for their excellent support when using their MRtrix toolbox, and the fixel‐based analyses specifically. Finally, we thank the Kinderkankerfonds Leuven for their funding to be able to perform this work.

Sleurs C, Lemiere J, Christiaens D, et al. Advanced MR diffusion imaging and chemotherapy‐related changes in cerebral white matter microstructure of survivors of childhood bone and soft tissue sarcoma? Hum Brain Mapp. 2018;39:3375–3387. 10.1002/hbm.24082

Funding information Kinderkankerfonds Leuven

REFERENCES

  1. Abbott, N. J. (2004). Evidence for bulk flow of brain interstitial fluid: Significance for physiology and pathology. Neurochemistry International, [DOI] [PubMed] [Google Scholar]
  2. Abraham, J. , Haut, M. W. , Moran, M. T. , Filburn, S. , Lemiuex, S. , & Kuwabara, H. (2008). Adjuvant chemotherapy for breast cancer: Effects on cerebral white matter seen in diffusion tensor imaging. Clinical Breast Cancer, 8(1), 88–91. http://www.ncbi.nlm.nih.gov/pubmed/18501063. [DOI] [PubMed] [Google Scholar]
  3. Ahles, T. , Li, Y. , McDonald, B. , Schwartz, G. , Kaufman, P. , Tsongalis, G. , … Saykin, A. (2014). Longitudinal assessment of cognitive changes associated with adjuvant treatment for breast cancer: The impact of APOE and smoking. Psycho‐oncology, 23(12), 1382–1390. http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=ovftp&NEWS=N&AN=01445499-201412000-00007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Ahles, T. , & Saykin, A. (2007). Candidate mechanisms for chemotherapy‐induced cognitive changes. Nature Reviews. Cancer, 7(3), 192–201. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Amidi, A. , Agerb??K, M. , Wu, L. M. , Pedersen, A. D. , Mehlsen, M. , Clausen, C. R. , Demontis, D. , … Zachariae, R. (2016). Changes in cognitive functions and cerebral grey matter and their associations with inflammatory markers, endocrine markers, and APOE genotypes in testicular cancer patients undergoing treatment. Brain Imaging and Behavior, 11(3), 769–783. [DOI] [PubMed] [Google Scholar]
  6. Armstrong, R. C. , Mierzwa, A. J. , Marion, C. M. , & Sullivan, G. M. (2016a). White matter involvement after TBI: Clues to axon and myelin repair capacity. Experimental Neurology, 275(Pt 3), 328–333. [DOI] [PubMed] [Google Scholar]
  7. Armstrong, R. C. , Mierzwa, A. J. , Sullivan, G. M. , & Sanchez, M. A. (2016b). Myelin and oligodendrocyte lineage cells in white matter pathology and plasticity after traumatic brain injury. Neuropharmacology, 110(Pt B), 654–659. [DOI] [PubMed] [Google Scholar]
  8. Bielack, S. S. , Werner, M. , Tunn, P. U. , Helmke, K. , Jürgens, H. , Calaminus, G. , … Dhooge, C. (2015). Methotrexate, doxorubicin, and cisplatin (MAP) plus maintenance pegylated interferon alfa‐2b versus MAP alone in patients with resectable high‐grade osteosarcoma and good histologic response to preoperative MAP: First results of the EURAMOS‐1 good respons. Journal of Clinical Oncology, 33(20), 2279–2287. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Billiet, T. , Vandenbulcke, M. , M??Dler, B. , Peeters, R. , Dhollander, T. , Zhang, H. , … Emsell, L. (2015). Age‐related microstructural differences quantified using myelin water imaging and advanced diffusion MRI. Neurobiol Aging, 36(6), 2107–2121. [DOI] [PubMed] [Google Scholar]
  10. Buizer, A. I. , De Sonneville, L. M. J. , Van Den Heuvel‐Eibrink, M. M. , & Veerman, A. J. P. (2005). Chemotherapy and attentional dysfunction in survivors of childhood acute lymphoblastic leukemia: Effect of treatment intensity. Pediatric Blood & Cancer, 45(3), 281–290. [DOI] [PubMed] [Google Scholar]
  11. Caron, J. E. , Krull, K. R. , Hockenberry, M. , Jain, N. , Kaemingk, K. , & Moore, I. M. (2009). Oxidative stress and executive function in children receiving chemotherapy for acute lymphoblastic leukemia. Pediatric Blood & Cancer, 53(4), 551–556. http://www.ncbi.nlm.nih.gov/pubmed/19499584. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Chang, E. L. , Wefel, J. S. , Hess, K. R. , Allen, P. K. , Lang, F. F. , Kornguth, D. G. , … Meyers, C. A. (2009). Neurocognition in patients with brain metastases treated with radiosurgery or radiosurgery plus whole‐brain irradiation: A randomised controlled trial. The Lancet. Oncology, 10(11), 1037–1044. [DOI] [PubMed] [Google Scholar]
  13. Concha, L. (2014). A macroscopic view of microstructure: Using diffusion‐weighted images to infer damage, repair, and plasticity of white matter. Neuroscience, 276, 14. [DOI] [PubMed] [Google Scholar]
  14. De Ruiter, M. B. , Reneman, L. , Boogerd, W. , Veltman, D. J. , Caan, M. , Douaud, G. , … Schagen, S. B. (2012). Late effects of high‐dose adjuvant chemotherapy on white and gray matter in breast cancer survivors: Converging results from multimodal magnetic resonance imaging. Human Brain Mapping, 33(12), 2971–2983. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Deprez, S. , Amant, F. , Yigit, R. , Porke, K. , Verhoeven, J. , Stock, J. , Van den, … Sunaert, S. (2011). Chemotherapy‐induced structural changes in cerebral white matter and its correlation with impaired cognitive functioning in breast cancer patients. Human Brain Mapping, 32(3), 480–493. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Deprez, S. , Billiet, T. , Sunaert, S. , & Leemans, A. (2013). Diffusion tensor MRI of chemotherapy‐induced cognitive impairment in non‐CNS cancer patients: A review. Brain Imaging and Behavior, 7(4), 409–435. [DOI] [PubMed] [Google Scholar]
  17. Edelmann, M. N. , Daryani, V. M. , Bishop, M. W. , Liu, W. , Brinkman, T. M. , Stewart, C. F. , … Krull, K. R. (2016). Neurocognitive and patient‐reported outcomes in adult survivors of childhood osteosarcoma. JAMA Oncology, 2(2), 201–208. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Edelmann, M. N. , Krull, K. R. , Liu, W. , Glass, J. O. , Ji, Q. , Ogg, R. J. , … Reddick, W. E. (2014). Diffusion tensor imaging and neurocognition in survivors of childhood acute lymphoblastic leukaemia. Brain, 137(11), 2973–2983. http://www.ncbi.nlm.nih.gov/pubmed/25125614. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Elalfy, M. , Ragab, I. , Azab, I. , Amin, S. , & Abdel‐Maguid, M. (2014). Neurocognitive outcome and white matter anisotropy in childhood acute lymphoblastic leukemia survivors treated with different protocols. Pediatric Hematology and Oncology, 31, 194–204. http://ovidsp.ovid.com/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=emed11&AN=2014136810%5Cnhttp://opurl.bib.umontreal.ca:9003/sfx_local?sid=OVID:embase&id=pmid:&id=doi:10.3109%2F08880018.2013.871763&issn=0888‐0018&isbn=&volume=31&issue=2&spage=194&pages=1. [DOI] [PubMed] [Google Scholar]
  20. Ferrari, A. , Trama, A. , De Paoli, A. , Bergeron, C. , Merks, J. H. M. , Jenney, M. , … Bisogno, G. (2017). Access to clinical trials for adolescents with soft tissue sarcomas: Enrollment in European pediatric Soft tissue sarcoma Study Group (EpSSG) protocols. Pediatric Blood & Cancer, 64(6), e26348. [DOI] [PubMed] [Google Scholar]
  21. Garg, N. , Reddel, S. W. , Miller, D. H. , Chataway, J. , Riminton, D. S. , Barnett, Y. , … Hardy, T. A. (2015). The corpus callosum in the diagnosis of multiple sclerosis and other CNS demyelinating and inflammatory diseases. Journal of Neurology, Neurosurgery, and Psychiatry, 1–9. http://jnnp.bmj.com/cgi/doi/10.1136/jnnp-2014-309649. [DOI] [PubMed] [Google Scholar]
  22. Giralt, J. , Ortega, J. J. , Olive, T. , Verges, R. , Forio, I. , & Salvador, L. (1992). Long‐term neuropsychologic sequelae of childhood leukemia: Comparison of two CNS prophylactic regimens. International Journal of Radiation Oncology, Biology, Physics, 24(1), 49–53. [DOI] [PubMed] [Google Scholar]
  23. Gooijers, J. , & Swinnen, S. P. (2014). Interactions between brain structure and behavior: The corpus callosum and bimanual coordination. Neuroscience and Biobehavioral Reviews, 43, 1. [DOI] [PubMed] [Google Scholar]
  24. Jeurissen, B. , Leemans, A. , Tournier, J. D. , Jones, D. K. , & Sijbers, J. (2013). Investigating the prevalence of complex fiber configurations in white matter tissue with diffusion magnetic resonance imaging. Human Brain Mapping, 34(11), 2747–2766. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Kadan‐Lottick, N. S. , Zeltzer, L. K. , Liu, Q. , Yasui, Y. , Ellenberg, L. , Gioia, G. , … Krull, K. R. (2010). Neurocognitive functioning in adult survivors of childhood non‐central nervous system cancers. Journal of the National Cancer Institute, 102(12), 881–893. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Kapogiannis, D. , Kisser, J. , Davatzikos, C. , Ferrucci, L. , Metter, J. , & Resnick, S. M. (2012). Alcohol consumption and premotor corpus callosum in older adults. European Neuropsychopharmacology: The Journal of the European College of Neuropsychopharmacology, 22(10), 704–710. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Kesler, S. R. , Gugel, M. , Huston‐Warren, E. , & Watson, C. (2016). Atypical structural connectome organization and cognitive impairment in young survivors of acute lymphoblastic leukemia. Brain Connectivity, 6(4), 273–282. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Khong, P.‐L. , Leung, L. H. T. , Fung, A. S. M. , Fong, D. Y. T. , Qiu, D. , Kwong, D. L. W. , … Chan, G. C. F. (2006). White matter anisotropy in post‐treatment childhood cancer survivors: Preliminary evidence of association with neurocognitive function. Journal of Clinical Oncology: Official Journal of the American Society of Clinical Oncology, 24(6), 884–890. [DOI] [PubMed] [Google Scholar]
  29. Kochunov, P. , Williamson, D. E. , Lancaster, J. , Fox, P. , Cornell, J. , Blangero, J. , & Glahn, D. C. (2012). Fractional anisotropy of water diffusion in cerebral white matter across the lifespan. Neurobiology of Aging, 33(1), 9–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Koppelmans, V. , Breteler, M. M. B. , Boogerd, W. , Seynaeve, C. , & Schagen, S. B. (2013). Late effects of adjuvant chemotherapy for adult onset non‐CNS cancer; cognitive impairment, brain structure and risk of dementia. Critical Reviews in Oncology/Hematology, 88(1), 87–101. [DOI] [PubMed] [Google Scholar]
  31. Koppelmans, V. , de Groot, M. , de Ruiter, M. B. , Boogerd, W. , Seynaeve, C. , Vernooij, M. W. , … Breteler, M. M. B. (2014). Global and focal white matter integrity in breast cancer survivors 20 years after adjuvant chemotherapy. Human Brain Mapping, 35(3), 889–899. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Krull, K. R. , Brinkman, T. M. , Li, C. , Armstrong, G. T. , Ness, K. K. , Kumar Srivastava, D. , … Hudson, M. M. (2013). Neurocognitive outcomes decades after treatment for childhood acute lymphoblastic leukemia: A report from the St jude lifetime cohort study. Journal of Clinical Oncology, 31(35), 4407–4415. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Krull, K. R. , Cheung, Y. T. , Liu, W. , Fellah, S. , Reddick, W. E. , Brinkman, T. M. , … Hudson, M. M. (2016). Chemotherapy pharmacodynamics and neuroimaging and neurocognitive outcomes in long‐term survivors of childhood acute lymphoblastic leukemia. Journal of Clinical Oncology, 34(22), 2644–2653. http://www.ncbi.nlm.nih.gov/pubmed/27269941. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Krull, K. R. , Okcu, M. F. , Potter, B. , Jain, N. , Dreyer, Z. , Kamdar, K. , & Brouwers, P. (2008). Screening for neurocognitive impairment in pediatric cancer long‐term survivors. Journal of Clinical Oncology: Official Journal of the American Society of Clinical Oncology, 26(25), 4138–4143. [DOI] [PubMed] [Google Scholar]
  35. Kunin‐Batson, A. , Kadan‐Lottick, N. , & Neglia, J. P. (2014). The contribution of neurocognitive functioning to quality of life after childhood acute lymphoblastic leukemia. Psycho‐oncology, 23(6), 692–699. [DOI] [PubMed] [Google Scholar]
  36. Leemans, A. , Jeurissen, B. , Sijbers, J. , & Jones, D. (2009). ExploreDTI: A graphical toolbox for processing, analyzing, and visualizing diffusion MR data. Proceedings of the 17th Scientific Meeting of the International Society for Magnetic Resonance in Medicine, 17, 3537 http://www.mendeley.com/research/exploredti-a-graphical-toolbox-for-processing-analyzing-and-visualizing-diffusion-mr-data/%5Cnhttp://www.exploredti.com/ref/ExploreDTI_ISMRM_2009.pdf. [Google Scholar]
  37. Madden, D. J. , Bennett, I. J. , & Song, A. W. (2009). Cerebral white matter integrity and cognitive aging: Contributions from diffusion tensor imaging. Neuropsychology Review, 19(4), 415–435. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. McDonald, B. C. , Conroy, S. K. , Ahles, T. A. , West, J. D. , & Saykin, A. J. (2010). Gray matter reduction associated with systemic chemotherapy for breast cancer: A prospective MRI study. Breast Cancer Research and Treatment, 123(3), 819–828. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. McDuff, S. G. R. , Taich, Z. J. , Lawson, J. D. , Sanghvi, P. , Wong, E. T. , Barker, F. G. , … Chen, C. C. (2013). Neurocognitive assessment following whole brain radiation therapy and radiosurgery for patients with cerebral metastases. Journal of Neurology, Neurosurgery & Psychiatry, 84(12), 1384–1391. http://www.ncbi.nlm.nih.gov/pubmed/23715918. [DOI] [PubMed] [Google Scholar]
  40. Mohrmann, C. , Henry, J. , Hauff, M. , & Hayashi, R. J. (2015). Neurocognitive outcomes and school performance in solid tumor cancer survivors lacking therapy to the central nervous system. Journal of Personalized Medicine, 5(2), 83–90. . [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Moleski, M. (2000). Neuropsychological, neuroanatomical, and neurophysiological consequences of CNS chemotherapy for acute lymphoblastic leukemia. Archives of Clinical Neuropsychology: The Official Journal of the National Academy of Neuropsychologists, 15(7), 603–630. [PubMed] [Google Scholar]
  42. Partridge, S. C. , Ziadloo, A. , Murthy, R. , White, S. W. , Peacock, S. , Eby, P. R. , … Lehman, C. D. (2010). Diffusion tensor MRI: Preliminary anisotropy measures and mapping of breast tumors. Journal of Magnetic Resonance Imaging: Jmri, 31(2), 339–347. [DOI] [PubMed] [Google Scholar]
  43. Raffelt, D. A. , Smith, R. E. , Ridgway, G. R. , Tournier, J. D. , Vaughan, D. N. , Rose, S. , … Connelly, A. (2015). Connectivity‐based fixel enhancement: Whole‐brain statistical analysis of diffusion MRI measures in the presence of crossing fibres. Neuroimage, 117, 40–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Raffelt, D. , Tournier, J. D. , Fripp, J. , Crozier, S. , Connelly, A. , & Salvado, O. (2011). Symmetric diffeomorphic registration of fibre orientation distributions. NeuroImage, 56(3), 1171–1180. [DOI] [PubMed] [Google Scholar]
  45. Raffelt, D. , Tournier, J. D. , Rose, S. , Ridgway, G. R. , Henderson, R. , Crozier, S. , … Connelly, A. (2012). Apparent Fibre Density: A novel measure for the analysis of diffusion‐weighted magnetic resonance images. NeuroImage, 59(4), 3976–3994. [DOI] [PubMed] [Google Scholar]
  46. Raffelt, D. A. , Tournier, J. D. , Smith, R. E. , Vaughan, D. N. , Jackson, G. , Ridgway, G. R. , & Connelly, A. (2016). Investigating white matter fibre density and morphology using fixel‐based analysis. NeuroImage, 144(Pt A), 58–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Schapiro, M. , Hayashi, R. , Liu, W. , Srivastava, D. , Jeffe, D. , Perkins, S. , … Leisenring, W. (2015). Neurocognitive, emotional, and quality of life outcomes in long‐term survivors of rhabdomyosarcoma: A report from the Childhood Cancer Survivor Study (CCSS). Journal of Clinical Oncology, 33(15), 10072 http://meeting.ascopubs.org/cgi/content/abstract/33/15_suppl/10072?sid=a1f9e3a2-0a0d-494c-896f-69f5cc91b53c%5Cnhttp://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=emed13&NEWS=N&AN=72015867. [Google Scholar]
  48. Schuitema, I. , Deprez, S. , Van Hecke, W. , Daams, M. , Uyttebroeck, A. , Sunaert, S. , … de Sonneville, L. M. J. (2013). Accelerated aging, decreased white matter integrity, and associated neuropsychological dysfunction 25 years after pediatric lymphoid malignancies. Journal of Clinical Oncology, 31(27), 3378–3388. [DOI] [PubMed] [Google Scholar]
  49. Schuitema, I. , Deprez, S. , Van Hecke, W. , Daams, M. , Uyttebroeck, A. , Sunaert, S. , … De Sonneville, L. (2012). Accelerated aging, decreased white matter integrity and associated cognitive dysfunction 25 years after childhood acute lymphoblastic leukemia. Pediatr Blood Cancer, 59, 965–1152. http://doi.wiley.com/10.1002/pbc.24295. 22987436 [Google Scholar]
  50. Simó, M. , Rifà‐Ros, X. , Rodriguez‐Fornells, A. , & Bruna, J. (2013). Chemobrain: A systematic review of structural and functional neuroimaging studies. Neuroscience and Biobehavioral Reviews, 37(8), 1311–1321. [DOI] [PubMed] [Google Scholar]
  51. Sleurs, C. , Deprez, S. , Emsell, L. , Lemiere, J. , & Uyttebroeck, A. (2016). Chemotherapy‐induced neurotoxicity in pediatric solid non‐CNS tumor patients: An update on current state of research and recommended future directions. Critical Reviews in Oncology/Hematology, 103, 37–48. 10.1016/j.critrevonc.2016.05.001. [DOI] [PubMed] [Google Scholar]
  52. Sleurs, C. , Lemiere, J. , Vercruysse, T. , Nolf, N. , Van Calster, B. , Deprez, S. , … Uyttebroeck, A. (2016b). Intellectual development of childhood ALL patients: A multicenter longitudinal study. Psychooncology, 26(4), 508–514. http://doi.wiley.com/10.1002/pon.4186. [DOI] [PubMed] [Google Scholar]
  53. Smith, K. W. , Gierski, F. , Andre, J. , Dowell, N. G. , Cercignani, M. , Naassila, M. , & Duka, T. (2017). Altered white matter integrity in whole brain and segments of corpus callosum, in young social drinkers with binge drinking pattern. Addiction Biology, 22(2), 490–501. [DOI] [PubMed] [Google Scholar]
  54. Smith, S. M. , & Nichols, T. E. (2009). Threshold‐free cluster enhancement: Addressing problems of smoothing, threshold dependence and localisation in cluster inference. NeuroImage, 44(1), 83–98. [DOI] [PubMed] [Google Scholar]
  55. Stouten‐Kemperman, M. M. , de Ruiter, M. B. , Koppelmans, V. , Boogerd, W. , Reneman, L. , & Schagen, S. B. (2015). Neurotoxicity in breast cancer survivors ≥10 years post‐treatment is dependent on treatment type. Brain Imaging Behavior, 9(2), 275–284. http://www.ncbi.nlm.nih.gov/pubmed/24858488. [DOI] [PubMed] [Google Scholar]
  56. Tamnes, C. K. , Zeller, B. , Amlien, I. K. , Kanellopoulos, A. , Andersson, S. , Due‐Tonnessen, P. , … Fjell, A. M. (2015). Cortical surface area and thickness in adult survivors of pediatric acute lymphoblastic leukemia. Pediatric Blood & Cancer, 62(6), 1027–1034. [DOI] [PubMed] [Google Scholar]
  57. Tournier, J. D. , Calamante, F. , & Connelly, A. (2012). MRtrix: Diffusion tractography in crossing fiber regions. International Journal of Imaging Systems and Technology, 22(1), 53–66. [Google Scholar]
  58. Tournier, J. D. , Calamante, F. , & Connelly, A. (2013). Determination of the appropriate b value and number of gradient directions for high‐angular‐resolution diffusion‐weighted imaging. NMRi N Biomedicine, 26(12), 1775–1786. [DOI] [PubMed] [Google Scholar]
  59. Tustison, N. J. , Avants, B. B. , Cook, P. A. , Zheng, Y. , Egan, A. , Yushkevich, P. A. , & Gee, J. C. (2010). N4ITK: Improved N3 bias correction. IEEE Transactions on Medical Imaging, 29(6), 1310–1320. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. van der Plas, E. , Nieman, B. J. , Butcher, D. T. , Hitzler, J. K. , Weksberg, R. , Ito, S. , & Schachar, R. (2015). Neurocognitive late effects of chemotherapy in survivors of acute lymphoblastic leukemia: Focus on methotrexate. Journal of Canadian Academy of Child and Adolescent Psychiatry, 24, 25–32. http://www.ncbi.nlm.nih.gov/pubmed/26336377%5Cnhttp://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=PMC4357331. [PMC free article] [PubMed] [Google Scholar]
  61. von der Weid, N. , Mosimann, I. , Hirt, A. , Wacker, P. , Nenadov Beck, M. , Imbach, P. , … Wagner, H. P. (2003). Intellectual outcome in children and adolescents with acute lymphoblastic leukaemia treated with chemotherapy alone: Age‐ and sex‐related differences. European Journal of Cancer, 39(3), 359–365. http://ac.els-cdn.com/S0959804902002605/1-s2.0-S0959804902002605-main.pdf?_tid=81ac3662-1e8c-11e2-9b75-00000aab0f01&acdnat=1351160139_cb1fdca7dc67a7f34ea796825672efd2. [DOI] [PubMed] [Google Scholar]
  62. Vos, S. B. , Jones, D. K. , Viergever, M. A. , & Leemans, A. (2011). Partial volume effect as a hidden covariate in DTI analyses. NeuroImage, 55(4), 1566–1576. [DOI] [PubMed] [Google Scholar]
  63. Whelan, J. S. , Bielack, S. S. , Marina, N. , Smeland, S. , Jovic, G. , Hook, J. M. , … Teske, C. (2015). EURAMOS‐1, an international randomised study for osteosarcoma: Results from pre‐randomisation treatment. Annals of Oncology, 26(2), 407–414. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Zhang, H. , Schneider, T. , Wheeler‐Kingshott, C. A. , & Alexander, D. C. (2012). NODDI: Practical in vivo neurite orientation dispersion and density imaging of the human brain. NeuroImage, 61(4), 1000–1016. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Additional Supporting Information may be found online in the supporting information tab for this article.

Supporting Information 1

Click here for additional data file. (143.6KB, docx)

Articles from Human Brain Mapping are provided here courtesy of Wiley

RESOURCES