Table 2.

Review of neuroimaging studies in adult survivors of pediatric cancer

First author	Year of publication	Journal	Sample size (N)	Survivors (n)	Healthy controls (n)	Age at assessment (years)	Age at diagnosis in years (mean, SD, [range])	Time since diagnosis/treatme nt in years (mean, SD,	Type of cancer
1 Chen et al.	2016	Neuroimage Clinical	32	16	16	22.5 (5.2) [1734]	7.6 (5.1)	14.9 (7.3)	Cerebellar tumor
2 Tamnes et al.	2015	Pediatric Blood and Cancer	260	130	130	29.3 (7.3) [18.6–46.5]	6.2 (4.0) [0.3–16.0]	23.0 (7.7) [7.4–40.0]	ALL
3 Jayakar et al.	2015	Neuropsychology	94	35	59	24.10 (4.93) [17–36]	8.17 (4.43) [117]	15.38 (5.34) [524]	Brain tumor (posterior fossa, pituitary, frontal, temporal, other)
4 King et al.	2015b	PLoS ONE	54	27	27	22.7 (4.5) [18–32]	9.0 (5.14)	13.7 (5.37)	Brain tumor (posterior fossa, temporal, occipital, hypothalamic, third ventricle)
5 King et al.	2015a	Journal of the International Neuropsychological Society	34	17	17	23.2 (5.9) [1735]	7.65 (4.90) [117]	15.5 (7.6) [4.5–30]	Posterior fossa brain tumor
6 Smith et. al.	2014	Neuropsychology	37	18	19	24.19 (4.51) [19–40]	7.22 (4.57) [117]	17.13 (5.43) [525]	Brain tumor (n = 6 MB n = 1 pineoblastoma, n = 9 astrocytoma, n = 1 ganglioglioma, and n = 1 craniopharyngioma). Location of tumor: n = 13 in posterior fossa, n = 1 pareital lobe, n = 1 occipital
7 Edelmann et al.	2014	Brain	98	75	23	CT alone: 24.9 (3.6), CRT: 26.7 (3.4)	CT alone: 9.97 (3.99), CRT: 2.81 (1.73)	CT alone: 15.0 (1.7), CRT: 23.9 (3.1)	ALL
8 Schuite ma et al	2013	Journal of Clinical Oncology	142	93	49	CRT and CT: 31.2 (4.8), CT: 26.7 (5.1)	CRT and CT: 5.7 (3.7), CT: 5.3 (3.5)	CRT and CT: 25.4 (3.2), CT: 21.4 (2.9)	ALL, lymphoma
9 Zeller et al.	2013	Journal of Clinical Oncology	260	130	130	28.4 [18.646.5]	5.3 [0.3–15.9]	22.5 [7.4–40.0]	ALL
10 Edelmann et al.	2013	Pediatric Blood and Cancer	38	38	__	Without dex: 24.6 [20.432.4], with dex: 24.6 [19.731.2]	Without dex: 8.7 [3.8–16.9], with dex: 11.8 [5.818.6]	Without dex: 15.9 [14.8–17.9], with dex: 13.3 [12.015.1]	ALL
11 Monje et al.	2013	Pediatric Blood and Cancer	20	10	10	30.8 [22–40]	10.2 [1.83–17.0]	Not stated	ALL
12 Armstrong et al.	2013	Journal of the National Cancer Institute	85	85	__	36.7 (6.4)	6.6 (4.1)	30.1 (6.2)	ALL
13 Krull et al.	2012	Journal of Clinical Oncology	62	62	__	42.2 (4.77) [1855]	15.1 (3.30)	>= 15	Hodgkin lymphoma
14 Brinkman et al.	2012	Neuro-oncology	20	20	__	29 [21–36]	[2–17]	18 [12–25]	MB
15 Dellani et al.	2008	Journal of Magnetic Resonance Imaging	27	13	14	17–37	2–16	16–23	ALL
16 Porto et al.	2008	European radiology	41	20 (n=10 male, n=10 female)	21		Male: 22.0 (3.2) [18–28], Female: 23.2 (3.5) [18–27]	Male: 8.1 (3.8) [2.2–13.7], Female: 8.1 (3.8) [3.0–13.3]	Not stated
First author	Type of treatment		Neuroimaging modalities	Type of study	Physical and physiological outcome measures	Cognitive and behavioral outcome measures	Emotional and quality of life measures	Main findings
1 Chen et al.	Surgical resection (n = 16), CT (n = 7), CRT (n = 8)		Resting-state fMRI	Cross-sectional	__	__	__	Pediatric brain tumor survivors showed altered rsFC in frontal functional networks (executive control, salience, default mode) compared to controls. Even at rest, survivors’ hyperconnectivity in frontal functional networks may reflect recruitment of more brain regions due to continuous needs of a higher level of cognitive effort (“all hands on deck” approach).
2 Tamnes et al.	CT (n=130), CT and CRT (n=18), CT and stem-cell transplantation (n=3)		Structural MRI	Cross-sectional	__	Executive functioning	__	ALL survivors had smaller surface area in several cortical regions but reduced cortical thickness in only one region, compared to controls. Cortical surface area/thickness in these regions was not associated with disease or treatment variables. In survivors, reduced cortical surface area in prefrontal regions was associated with more self-reported problems in executive functioning (particularly difficulties in emotional control and self-monitoring).
3 Jayakar et al.	Surgical resection (n=35), CT (n=12), CRT (n=16)		Structural MRI	Cross-sectional	__	Verbal memory, IQ	__	Compared to controls, brain tumor survivors had lower hippocampal, putamen, and whole brain volumes as well as lower verbal memory scores (auditory attention list span and final list learning). Hippocampal volume in survivors was significantly correlated with auditory attention. In both groups, hippocampal and putamen volumes were significantly correlated with each other but not with total brain volume.
4 King et al.	CRT with CT (n=11), CRT without CT (n=3), no CRT (n=12), CT and no CRT (n=1)		Structural MRI, DTI	Cross-sectional	__	IQ	__	Brain tumor survivors treated with CRT had lower IQ and a lower level of white matter integrity relative to survivors without CRT and healthy controls. Non-CRT treated survivors had lower mean FA compared with healthy controls. IQ and cumulative neurological factors were related to white matter disruption in the CRT-treated group of survivors.
5 King et al.	Neurosurgery (n=17), CT (n=8), CRT (n=9)		Task-based fMRI (N-back paradigm)	Cross-sectional	__	Working memory, IQ	__	Survivors of childhood posterior fossa brain tumors had lower IQ scores and working memory performance compared with controls. Among survivors, increased prefrontal activation (left superior/middle frontal gyri) during a working memory task was associated with increased working memory demands and re duced working memory performance.
6 Smith et. al.	Surgical resection (n = 18), CT and CRT (n = 8), CT (n = 1), and CRT (n = 1)		DTI	Cross-sectional	__	Reading achievement, processing speed, skilled motor speed	__	Among brain tumor survivors and controls, white matter FA values of the parietotemporal-occipitotemoral (PT-OT) tract were associated with word reading, and FA values in the inferior fronto-occipital fasciculus (IFOF) were associated with reading in survivors only. Among survivors only, processing speed mediated the relation between white matter FA (in PTOT and IFOF) and word reading skill.
7 Edelmann et al.	CT alone (n=36), CT and CRT (n=39)		Structural MRI, DTI	Cross-sectional	__	IQ, academic performance, attention, memory, processing speed, executive function	__	CRT-treated ALL survivors performed lower than survivors treated with CT alone on only 3 of 20 neurocognitive measures. Compared to healthy controls, ALL survivors had lower neurocognitive performance, reduced gray and white matter, and higher FA in fibre tracts within the left hemisphere. Frontal and temporal lobe volumes correlated with vocabulary and academic ability; frontal, parietal, and temporal white matter volumes were associated with memory. Higher FA in the left longitudinal fasciculus and left uncinate fasciculus were associated with lower memory and learning performance. Increased FA in the left sagittal stratum was associated with better sustained attention
8 Schuite ma et al	CT (n=49), CT and CRT (n=44)		DTI	Cross-sectional	__	IQ, speed and accuracy of information processing, attention, working memory	__	Compared with controls, ALL and lymphoma survivors treated with CRT showed decreased FA in frontal, parietal, and temporal WM tracts. Decreased FA was asociated with poorer neuropsychological performance. Trends for lower WM FA were seen in the CT-treated survivors. Similarly, CRT-treated survivors performed significantly lower on all neuropsychological measures compared to controls whereas survivors treated with CT alone did not differ significantly from controls. Reduced WM integrity in CRT-treated survivors was associated with younger age at CRT and higher dosage.
9 Zeller et al.	CT (n=130), CT and CRT (n=18), CT and stem-cell transplantation (n=3)		Structural MRI	Cross-sectional	__	IQ, processing speed, executive function, verbal learning/memory	__	ALL survivors showed smaller volumes of cortical gray matter, cerebral white matter, amygdala, caudate, hippocampus, thalamus, and intracranial volume compared with controls – with the strongest effect found in the caudate. These neuroanatomic volumes were not effected by age at diagnosis nor treatment variables. Survivors also showed reduced processing speed, executive function, and verbal learning/memory in survivors compared with controls. Reduced neurocognitive performance correlated with smaller volumes of cortical gray matter, caudate, and thalamus and intracranial volume in survivors
10 Edelmann et al.	CT (n=20 with prednisone and no dexamethasone, n=18 with dexamethasone)		Structural MRI, task-based fMRI (memory recognition tasks: word & face)	Cross-sectional	__	IQ, academic performance, memory	__	Compared to survivors treated with only prednisone (without dexamethasone), ALL survivors treated with dexamethasone had lower performance on several memory-dependent measures in cluding story memory and word recognition. Decreased neural activation in the left retrosplenial brain region was associated with dexamethasone treatment; story memory was associated with altered activation in left inferior frontal-temporal brain regions.
11 Monje et al.	CT and CRT (n=10)		Structural MRI; task-based fMRI (memory encoding task)	Cross-sectional	__	Recognition memory	__	Compared to healthy controls, ALL survivors demonstrated lower recognition memory, greater hippocampal atrophy and altered memory-related hippocampal activation. Survivors showed increased neural activation in several brain regions during unsuccessful memory encoding, which may reflect ineffective compensatory neural recruitment.
12 Armstrong et al.	CT (n=85), 24 Gy CRT (n=36), 18 Gy CRT (n=49)		Task-based fMRI (memory task), DTI	Cross-sectional	__	Memory, IQ, cognitive status	__	ALL survivors who received 24 Gy (but not 18 Gy) CRT had reduced cognitive status and memory performance. The mean score for long-term narrative memory among survivors who received 24 Gy CRT was equivalent to that for individuals older than 69 years. Memory impairments were associated with smaller temporal lobe white matter volumes, thinner parietal and frontal cortices, increased radial diffusivity (inverse measure of WM integrity) in parietal and temporal lobes, and reduced hippocampal volume. Neural activation during memory retrieval in the left anterior hippocampus was correlated with design memory impairment in all survivors, and was driven by the the 24 Gy rather than the 18 Gy group. Neural activation during
13 Krull et al.	CT and thoracic radiation		Structural MRI	Cross-sectional	__	IQ, attention, memory, processing speed, executive function	__	Lymphoma survivors showed lower performance on sustained attention, short-term memory, long-term memory, working memory, naming speed, and cognitive fluency compared with age-adjusted norms. Leukoencephalopathy was present in 53% of survivors, who demonstrated lower cognitive fluency than those without leukoencephalopathy. Evidence of cerebrovascular injury was present in 37% of survivors.
14 Brinkman et al.	CRT (n=20), CT (n=15)		Structural MRI, DTI	Cross-sectional	__	IQ, academic skills, memory, attention, processing speed, motor function, executive function	__	IQ in medulloblastoma survivors was lower than population norms, and 75% of survivors showed executive function impairment. Lower performance on executive function tasks was correlated with reduced white matter intergrity in multiple brain regions. Radial diffusivity in the frontal lobes was correlated with shifting attention and cognitive flexibility, whereas volume and cortical thickness were not correlated with neurocognitive function. Neurocognitive impairment was common and involved many domains.
15 Dellani et al.	Total brain radiation (18–24 Gy) and CT		DTI	Cross-sectional	__	__	__	ALL survivors had significantly reduced white matter FA values in the temporal lobes, hippocampi, and thalami along with significant white matter volume loss. Although survivors did not show reductions in gray matter, they did show decreased total brain volume and intracranial volume compared to controls. Concerning structural white matter changes (as indexed by global and frontal WM mean diffusivity), adult survivors of childhood ALL show the same age dependence as controls - no age dependence of radiation damage was found.
16 Porto et al.	CT (n = 10), CT and CRT (n = 10)		Structural MRI, DTI	Cross-sectional	__	__	__	Compared to controls, ALL survivors showed reduced white matter FA, with the most severe effects apparent in those who had received both CRT and CT (compared to CT alone). Survivors treated with CRT had reduced WM volumes and gray matter concentration within the caudate nucleus and thalamus.

Abbreviations: ALL, acute lymphoblastic leukemia; PA, pilocytic astrocytoma; MB, medulloblastoma; PNET, primitive neuroectodermal tumor; WM, white matter; NAWM, normal appearing white mater; LGG, low-grade glioma; CT, chemotherapy; CRT; cranial radiation therapy; IV-MTX, intravenous methotrexate; CSI, craniospinal irradiation; BMT, bone marrow transplant; DTI, diffusion tensor imaging; fMRI, functional magnetic resonance imaging; SPECT, single-photon emission computed tomography; DWI, diffusion weighted imaging; ASL, arterial spin labelling; IQ, intelligence quotient; FA, fractional anisotropy,