Regional Cerebral Blood Flow in Children From 3 to 5 Months of Age

Discussion

Understanding the relationship between brain and behavior in early childhood requires a probe of functional brain development. Because CBF reflects regional metabolism, ASL offers a noninvasive approach to studying brain function. In this report, we demonstrate the feasibility of the use of ASL in a group of healthy infants. Even within the narrow age range studied, ASL identifies significantly greater CBF in occipital and parietal lobes and the primary sensorimotor cortex compared with more anterior brain regions. In addition, we report significant age-related changes in CBF, which suggests that CBF change over time may be an important measure of functional brain development.

Histologic studies demonstrate the dynamic nature of early brain development, with rapid changes in neuronal³ and synaptic² attenuation occurring first in primary sensory (primary visual cortex, Heschl gyrus) and later in higher-order association regions (prefrontal). Recent MR imaging studies by use of high-resolution sequences show a similar pattern in early adolescence¹⁴ and as early as the first 2 months of life.¹⁵ More recent techniques, such as DTI, characterize structural white matter development through childhood.¹⁶ These histologic and imaging studies provide powerful insights into the complex nature of early structural brain development.

Structure, however, does not equal function. Although our understanding of the rich tapestry of structural brain development has increased, the relation to brain function and behavior remains unclear. A study of functional brain development by use of PET to assess regional cerebral glucose metabolism^5,17 found that the primary sensory and motor cortices, along with deeper cerebral areas, had the highest metabolic rates during the first months of life. Next, the parietal and temporal cortex developed increased glucose utilization. The frontal cortex demonstrated increased glucose metabolism by the end of the first year. These studies were ethically possible because of the sole inclusion of children requiring PET for clinical reasons that involve the substantial radiation exposure. PET is not feasible for the study of normal brain development. Other functional imaging modalities require radiation (SPECT) or subject cooperation (stimulus-evoked fMRI). Recently, functional connectivity MR imaging has been used to identify network connectivity in children by assessment of the coherence of fluctuating levels of blood oxygenation at rest.¹⁸

Tight coupling of brain metabolism and blood flow¹⁹ suggests that CBF can be used as a marker of regional metabolic activity. With the use of magnetically labeled arterial water as a tracer, ASL quantifies CBF noninvasively.²⁰ ASL has been used for study of brain changes in disease states including vascular disease, dementia, and epilepsy.²⁰ Noguchi et al²¹ demonstrated that increased CBF measured by ASL may be related to increased microvascular attenuation in subjects with brain tumors. It is possible that in healthy children, increased CBF may be related to increased regional synaptogenesis, which, through neurovascular coupling, increases regional microvascular attenuation. Because ASL is noninvasive and requires no external tasks or exposure to radiation, it is well suited for use in pediatric samples. Wang et al¹⁰ were the first to use ASL in children, establishing feasibility at 1.5T. More recently, ASL has shown that preterm children have higher perfusion than term children.²² In healthy individuals from 4 years to adulthood, there is a decrease in CBF—most prominent in gray matter—by midadolescence.⁸ Taki and Kawashima²³ reported a nonlinear relationship between age and CBF from 5–18 years. They noted that these findings were similar to the reported gray matter volume increase in prepuberty followed by a postpuberty decrease.²⁴ Wang et al¹⁰ compared 8 children at 7 and 13 months of age. Regional CBF increased significantly at 13 months in the dorsolateral frontal lobes and decreased in the sensorimotor areas, consistent with the PET studies of Chugani.^5,17

We report the first large study of regional CBF by using ASL in young children. Lack of MR imaging contrast between gray and white matter at this age prohibited detailed regional segmentation, however by dividing the brain into broad anatomic areas¹³ we were able to demonstrate significantly greater CBF in the sensorimotor cortex and parietal/occipital lobes compared with dorsolateral prefrontal, orbitofrontal, and subgenual regions. Children were 3–5 months of age and therefore it is difficult to compare our findings with those of Limperopoulos et al¹³ However, our findings are consistent with the prior PET work of Chugani,⁵ who found increased cerebral metabolism at this age in the parietal and occipital regions and reduced frontal lobe metabolism. Our findings are also consistent with histologic studies demonstrating primary sensorimotor regions maturing earlier than higher association regions.^2,3

Our cohort included 61 subjects 3–5 months of age, allowing us to evaluate regional change in CBF over time. We found that CBF increased significantly with age in every region except the premotor and sensorimotor regions. Two of the regions in which there was the greatest age effect (orbital frontal and subgenual) had lower overall CBF. We speculate that this changing blood flow may reflect increasing metabolism that precedes a corresponding maturation of brain structure as occurs in the more anterior brain regions toward the end of the first year of life. Further studies are required to determine whether regional changes in CBF predict structural development, which could potentially serve as a marker of abnormal development before it becomes clinically evident.

Our findings are also consistent with normal clinical development. Clinical skills developing at 3–5 months include visual-motor tasks such as reaching, rolling over, and increasing control of movement.²⁵ Facial recognition and social interaction are highly visually dependent and are actively developing at this stage of childhood.²⁶ These skills involve the sensorimotor and occipital/parietal regions, which had the highest CBF in our subjects. The lowest flow was found in the anterior brain regions, which subserve later developing skills such as working memory, impulse inhibition, and self-regulation.^27,28

We did not find a sex difference in CBF at this age; however only 39% of our study cohort were girls. This is a limitation that may have been partially mitigated by our relatively large sample size, but the role of sex in early CBF requires further evaluation. Likewise, our cohort reflected the local demographics with a proportionally large Hispanic population. Although this may be a limitation, we are aware of no evidence that ethnicity affects regional CBF. Our unsuccessful scans were largely caused by excessive movement; thus, children who had difficulty sleeping in our MR imaging environment were not included in the data analysis. This is a difficult limitation to avoid. By use of an experienced scanning team and providing families with multiple options for aiding sleep, our success rate of 74% probably reflects a range of normally developing children. Though our sample size was substantial, individual variability in brain development at this age probably warrants a larger sample size to ensure power to detect all differences of clinical import. In addition, data obtained from the cerebellum was markedly inconsistent, and therefore this region was not included in our study. Finally, this study is limited by the broadly defined regions studied, the inclusion of gray and white matter in the measurements, and the use of only T1 sequences. With additional data from a T2 sequence, a more physiologically meaningful segmentation may have been possible. These limitations must be considered in the interpretation of our results; however, the consistency of the findings across the study group strengthens the possibility that these differences may not have contributed importantly to the findings. However, it is impossible to know how these factors may bias our results. The validity of our findings is strengthened by their consistency with those found in a previous study of regional cerebral glucose utilization by use of PET scanning.^5,17 This implies that ASL may provide a noninvasive means of determining regional brain functioning with the same accuracy as PET scanning without radiation exposure.

Our findings highlight the potential for use of ASL to assess functional brain development in early childhood. Future longitudinal studies including motor and cognitive measures may offer insights into early brain/behavioral relationships that have relevance for understanding normal and abnormal child development. Although these studies are inherently difficult, given the challenges of scanning children at this age without sedation, ASL offers the possibility of establishing early markers of abnormal development that may allow earlier, more effective treatment for children at risk for developmental disorders.