GM volume increases in some areas of the healthy hemisphere, mainly in the frontal lobe, in pediatric patients with heterogeneous tumors after surgery. The healthy hemisphere had some structural changes after operation; however, four indices of the SBM analysis did not pass correction, which might indicate that the GM volume measured by VBM is a composite result combining thickness, sulcus depth, gyrification, FD, and other indicators not provided by CAT12. This is the first study to demonstrate brain structural and functional changes in the healthy hemisphere after brain surgery in children within 6 months.
The mixed measurements increased sensitivity compared to analyses using each measurement method alone, which corroborated to a previous study (Voets et al., 2008). The main active brain region after surgery was the frontal lobe, which supported the fact that the frontal lobe was the main area of change. These functional and structural changes may be due to neuroplasticity mechanisms and can be explored and studied by means of brain imaging (Xu et al., 2017). Plasticity is the intrinsic adaptation of the central nervous system to changes in the internal and external environment. It occurs throughout the life cycle, during development, new learning, and adaptive mechanisms to support functional recovery after brain injury (Draganski et al., 2004; Maguire et al., 2000; Payne & Lomber, 2001). Therefore, after an injury, the brain activates its plasticity mechanisms to compensate to minimize the effects of the injury.
Our VBM and functional imaging analyses indicated that the compensation of the healthy hemisphere was mainly in the frontal lobe, especially the dorsolateral superior frontal gyrus. Liu et al. (Liu et al., 2020) found that unilateral frontal glioma invasion of the frontal lobe in adults caused structural and functional reorganization of the contralateral structure of the posterior cognitive control network, especially the contralateral superior parietal lobe. While results in insular glioma adult patients supports the plasticity mechanism of homologous recombination which means the damage of the lesion side brain region was compensated by the corresponding brain region of the healthy side (Almairac et al., 2018), ours results differ, as do the compensatory brain regions. There may be differences in the mechanisms of plasticity between adults and children that have influenced the results. With regards to GM density/volume during brain development (Sowell et al., 2001), the local decrease in GM density is mainly distributed in the dorsal frontal and parietal lobes during childhood and adolescence. Additionally, between adolescence and adulthood, a sharp increase in local GM density in the frontal lobe is observed (Sowell et al., 2001). Further, the loss of GM in the parietal lobe decreases relative to the early period. We found that GM density decreases in the dorsal frontal and parietal lobes as a developmental trend toward maturity from childhood to adolescence. In adult development, frontal lobe density increases and parietal lobe density decreases, which implies that there are structural changes in different brain regions at different ages. Conversely, increased GM density in the dorsal frontal or parietal lobes in children and adolescents after a certain treatment may represent reduced maturity in that region, or reverse development. We found increased GM density/volume in the frontal lobe. Therefore, the mechanism of the plasticity recombination of the healthy hemisphere after focal injury in children may be to reduce the developmental stage of some brain regions of the frontal lobe, to promote functional compensation and functional recombination, and reduce the impact of surgical injury.
For brain damage in adults, an increase in GM density/volume might be observed in the parietal lobe (Liu et al., 2020). GM volume in the healthy parietal lobe increased significantly when the tumor invaded the right frontal lobe compared with the control group. Although this phenomenon is is only found in one hemisphere, it is in line with our findings. This developmental stage difference may be one of the reasons for the difference in postoperative compensation between adults and children. Some studies support our hypothesis, and events that occur after injury are similar to those observed during normal development (Cramer & Chopp, 2000; Dromerick et al., 2015; Kollen et al., 2009; Pollock et al., 2007; Shehadah et al., 2014; Zeiler et al., 2016). However, Xu et al. (Xu et al., 2017) found compensatory sites in the right cuneus, left thalamus, and right globus pallidus in glioma adult patients, and Huang et al. (Huang et al., 2021) found increased contralateral insula GM volume in glioma adult patients with IDH mutations. These results cannot be well explained by our theory since the influencing factors are complex; hence, further research is needed in this regard.
Previous studies on brain structure changes in survivors of pediatric brain tumor focused on posterior fossa tumors, and few investigated only surgical injuries (Ailion et al., 2017); however, most of them investigated the combined effects of surgery, chemotherapy, and radiotherapy. Horská et al. (Horská et al., 2010) investigated the volume changes of the vermis of the cerebellum in tumor patients, and found that the volume decreased after surgery, radiation, or chemotherapy. Leung et al. (Leung et al., 2004) conducted a study on a similar cohort, and found that the integrity of several white matter areas in the whole brain decreased after treatment. Jayakar et al. (Jayakar et al., 2015) observed a decrease in bilateral hippocampal, putamen, and whole brain volume in pediatric patients after undergoing radiation or chemotherapy. Patients in these studies experienced extensive brain damage, such as that following radiotherapy or chemotherapy. Further, the GM and WM were damaged, which differs from the results of structural compensation in our study. We hypothesize that focal injury can stimulate structural compensation in the brain. However, if the damage pattern is too extensive, it is difficult to compensate within the structure.
Extensive atrophy of the WM was also found in a study of TBI (Dennis et al., 2017). Interhemispheric transfer time, measured as an event-related potential, was used to divide patients into a TBI-slow group and a TBI-normal group. Some areas of the TBI-slow group, such as the superior frontal gyrus, had increased GM volume compared with the TBI-normal group. The WM damage in the TBI-slow group was more severe than that in the TBI-normal group, and the more severely damaged group showed an increase in GM volume. All our surgical patients suffered WM damage, and GM volume increased in the healthy hemisphere, which was similar to the aforementioned results. We assume that focal WM injury may increase GM volume in some areas of the brain to reduce the effects of injury. However, if the damage scope is too wide, this mechanism may not exist.
For the cognitive assessment, PsyMoSp measures how well a subject perceives, attends, responds to visual-perceptual information, and performs motor speed and fine motor coordination; and MS measures the ability to perform movements to produce and satisfy an intention toward a manual action and goal. Further, MS can influence PsyMoSp. The difference in calculation of these two indicators lies in the score of SDC CR, which we compared with no statistical difference using paired T-test (Table 3). The decline in motor speed may cause a decline in psychomotor speed, and other cognitive functions required by psychomotor speed have not been impaired or have been restored. However, there were no corresponding brain imaging results in the healthy hemisphere to explain this phenomenon, which may be caused by surgical injury in the affected hemisphere.
There is a period of vulnerability after brain damage (Madhavan et al., 2019). From the perspective of the whole brain, the decline of PsyMoSp could indicate that the brain is in a vulnerable stage (Amieva et al., 2019). As for the reasoning score, the battery uses a non-verbal reasoning test to assess reasoning ability, which is a higher cognitive function mainly involving the frontal lobe (Dosenbach et al., 2010; Paz-Alonso et al., 2014; Soderqvist et al., 2012). Our results showed structural and functional changes in the healthy frontal lobe, which is consistent with the trend of reasoning score changes after surgery. There is sufficient evidence that substantial spontaneous recovery occurs weeks or months after the sudden onset of brain injury, including cognitive ability (Fasotti, 2017). Therefore, we hypothesize that the recovery of some higher cognitive functions can be seen within 6 months after brain surgery in children, and the related brain regions are mainly in the frontal lobe.
The current study had some limitations. The unwillingness of the guardians of the children to cooperate with follow-up made it challenging to acquire more extended follow-up data. Additionally, the participants’ poor self-control made it harder to minimize head movements in the lengthy MRI scan, resulting in incomplete data. Further studies could redo the rs-fMRI scan, retest to understand the changes after 6 months, and expand the sample size to reduce the effects of noise.