Adaptive neuroplastic responses in early and late hemispherectomized monkeys

Abstract

Behavioural recovery in children who undergo medically required hemispherectomy showcase the remarkable ability of the cerebral cortex to adapt and reorganize following insult early in life. Case study data suggest that lesions sustained early in childhood lead to better recovery compared to those that occur later in life. In these children, it is possible that neural reorganization had begun prior to surgery but was masked by the dysfunctional hemisphere. The degree of neural reorganization has been difficult to study systematically in human infants. Here we present a 20-year culmination of data on our nonhuman primate model (Chlorocebus sabeus) of early-life hemispherectomy in which behavioral recovery is interpreted in light of plastic processes that lead to the anatomical reorganization of the early-damaged brain. The model presented here suggests that significant functional recovery occurs after the removal of one hemisphere in monkeys with no preexisting neurological dysfunctions. Human and primate studies suggest a critical role for subcortical and brainstem structures as well as corticospinal tracts in the neuroanatomical reorganization which result in the remarkable behavioral recovery following hemispherectomy. The non-human primate model presented here offers a unique opportunity for studying the behavioral and functional neuroanatomical reorganization that underlies developmental plasticity.

Figure 1

The complete removal of the left hemisphere was performed during infancy. Here, we show the results of the near complete removal of the left hemisphere four years after the initial surgery in the axial plane (A and C), lateral view (B), and orbital view (D).

Figure 2

Perimetry: infant-lesioned monkeys (a) were able to detect visual stimuli at 45° in the “blind” hemifield, whereas no visual response could be elicited in the adult-lesioned subjects (b) in the contralateral hemifield. Normal-sighted monkeys had a 75% visual perimetry at 90° in both hemifields (c) (adapted from [13]).

Figure 3

Normal gait was significantly affected by the removal of one hemisphere. The ipsilateral side of the body acts as a control for normal motor movements and did not show any paresis, also referred to as the nonparetic side of the subject. Panels (a)–(d) are indicative of a typical gait sequence. As the forelimb moves forward, the contralateral appendages drag on the ground (arrow panel a). The hand dragging continues through the entire forward motion (arrow panels b and c). The hind limb is fully removed from the ground and does not drag (arrow d). At each time point and lesion group there is a significant effect of the hemispherectomy on upper limb movement defined as arm drags/total forward arm movements (e). A number of abnormal leg movements (drags and limps) as a percentage of total leg movements was significantly elevated over the expected rate of zero in the adult-lesioned group only (f). Years 1, 2, and 3 refer to the age of the infant-lesioned subjects at testing which also corresponds to the years after lesion. For the adult-lesioned subjects, testing occurred 3 years after initial surgery when the subjects were 7 years old. *P < 0.001 adapted from [14].

Figure 4

At a young age, the subjects made a significant number of unsuccessful attempts to grab the horizontal bar with their hands and would transverse the bar upside down (a and e). The subjects would typically overreach the bar with their arms and glide their arms across the bar until their hands were able to latch on (white arrows in a and b). The lower limb also had difficulty latching onto the bar during the first two postoperative years after the surgery (black arrow in panels a and g). At two years of age, the monkeys began not to attempt to use the contralateral upper limb to transverse the bar (black arrows in panel c). Typically the subject would successfully use the affected hind limb (black arrow in panels d and h) and would not attempt to use the front limb (white arrow in panels d and f) to transverse the horizontal bar. By 2 years after surgery the subjects were able to walk upright across the bar. Dashed line (e and g) represents the expected value for a normal monkey. % successful was determined as successful latches (hand or foot)/total attempts to latch onto the horizontal bar. The ipsilateral side of the body acts as an internal control for normal motor movements and did not show any paretic movements. Years 1, 2, and 3 refer to the age of the infant-lesioned subjects at testing which also corresponds to the years after lesion. For the adult-lesioned subjects, year 4 refers to the number of years after initial surgery. Adapted from [14].

Figure 5

Thermal sensitivity: thermal sensitivity was tested in the infant-lesioned subjects at 3 years of age. The ipsilateral side of the body acts as a control for normal response to thermal stimuli and, as stated earlier, did not show any paretic movements. Withdrawal times were significantly longer for the contralateral (paretic side) than for the ipsilateral (nonparetic) limb (a and b). The contralateral lower limb tended to have higher response rates than upper limb to thermal sensitivity (c and d). Withdrawal response rates did not differ between upper and lower ipsilateral limbs. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 6

Suggested anatomical pathway for visual field recovery following early hemispherectomy (based on a series of anatomical studies performed on the same monkeys). The deafferented left temporal hemiretina (T) still contains ganglion cells [57] that send their axons to the remaining left dLGN as a dead end [12]. Retinofugal projections to the left superior colliculus (SC) are still robust and maintain a normal metabolic rate as measured by cytochrome oxydase immunochemistry [59]. The information reaching the left SC is transferred to the right SC via the intertectal commissure (IC), the right Pulvinar (P), and the occipital cortex of the remaining right hemisphere. The deafferented right nasal hemiretina (N) sends crossed projection through the optic chiasm (OC) to the appropriate layers of the remnants of the left dLGN and to the left SC [15, 58, 59, 62]. The left Pulvinar is severely atrophied and receives no retinal projections. Projections from the afferented portions of the retinae (left nasal and right temporal) reach their subcortical targets in a normal fashion en route to the occipital cortex of the right hemisphere.

Figure 7

Design-based stereology neuronal counts in the dorsal column nuclei. There were no neuronal population differences between ipsi- and contralateral subdivisions of the dorsal column nuclei.