1 - 8: Mechanisms of Functional Recovery after Stroke

Great strides have been made in clinical stroke research over the last decade. The therapeutic time window of intravenous recombinant tissue-type plasminogen activator has been extended to 4.5 h, and the new Solitaire flow restoration device achieves better recanalization in patients with large vessel intracranial occlusion [1,2]. However, the majority of patients with ischemic stroke still do not benefit from these advancements because of the narrow therapeutic indications. In patients with intracerebral hemorrhage, treatment with aggressive blood pressure control and hemostatic agents using activated factor VII has failed to translate into improvement in functional outcome [3,4]. Meanwhile, stroke still is the leading cause of disability worldwide [5].

The functional status of stroke patients spontaneously improves over 6 months after onset. More specifically, rapid recovery is achieved during the first month [6]. From the patient's perspective, rehabilitation is a process of regaining and relearning lost functions. Therefore, functional improvement, augmented by active rehabilitation, overlaps with motor learning in terms of underlying mechanisms [7]. Motor learning is associated with structural changes, such as axonal or dendritic growth along with new synapse formation and functional modulation including long-term potentiation or long-term depression, which may enhance or suppress synaptic activities. Together with the mechanisms above, cortical reorganization develops in the damaged brain, which plays an important role in recovery from acute stroke. Herein, we will elaborate on the functional recovery mechanisms after stroke.

A myriad of evidence from animal experiments suggests that neurogenesis does occur after stroke. Neuroblasts usually originate from their source location in the brain, such as the subgranular zone in the dentate gyrus of the hippocampus and the subventricular zone. In rodent stroke models, neuroblasts divert from the rostral migratory system and move to the ischemic penumbra. These migrated neuroblasts may replace injured neurons or glial cells, and help with remodeling and reorganization processes [8]. This has long been considered a unique process in animals; however, recent evidence shows that neuronal migration occurs in adult human brains as well. Brain biopsy and autopsy studies in humans have shown that neurogenesis occurs after stroke [9]. However, it still remains to be elucidated whether the neurogenesis directly translates into clinical functional benefit in the human brain.

Neuronal death after vascular occlusion is a major underlying pathophysiology of ischemic brain injury. Newly formed blood vessels might help with augmenting nutrient supply and repair processes [10]. Simply, proangiogenic balance is associated with mild neurologic deficit and antiangiogenesis status predicts a worse long-term functional outcome in humans [11]. However, it is still elusive whether angiogenesis is a since qua non for neurologic recovery. Proangiogenic growth factors promote survival of the neuronal, glial and endothelial cells in the peri-infarct tissues, and transient neovascularization in the ischemic brain helps with the clearance of damaged tissues. Moreover, it may create a vascular niche for neuroblast migration [10]. Therefore, angiogenesis has multiple beneficial roles in the ischemic brain tissue rather than simple blood flow augmentation. Decreased angiogenesis is frequently seen in elderly and those with hypertension or diabetes mellitus, which is associated with poor functional recovery after stroke [10]. Taken together, angiogenesis may be necessary, but not sufficient for neurologic recovery. More studies are needed to verify its clinical utility in humans.

Axonal sprouting and regeneration also play a significant role in neurologic recovery. The major stimuli for this process are thought to be peripheral deafferentation. Axonal sprouting is mainly driven by the balance between a growth-promoting status and reduction of growth-inhibitory environment. Axonal sprouting may alter cortical sensory or motor maps, and robust evidence exists to show that new connections are formed in peri-infarct cortex areas [12]. Nogo-A protein is closely related with this process. It limits plasticity via inhibiting neurite outgrowth. Anti-Nogo-A antibody enhances functional recovery and promotes reorganization of the corticospinal tract with axonal plasticity [13]. Therefore, it is currently a hot topic for modulating regeneration.

In intracerebral hemorrhage, extravasated blood forms a clot and generates thrombin which is a potent source for post-hemorrhage inflammation. However, recent animal research shows that thrombin might be important in the functional recovery process by stimulating neuroblasts, enhancing neurogenesis, promoting secretion of nerve growth factors, and affecting neurite outgrowth [8]. Thrombin also enhances angiogenesis and synaptic remodeling, and has a strong effect on brain plasticity. By contrast, Hirudin, a specific inhibitor of thrombin, decreases neurogenesis in a rat intracerebral hemorrhage model, suggesting the importance of thrombin in neurogenesis. Moreover, statin has a pleiotropic effect, and has strong beneficial effects on angiogenesis, neurogenesis and synaptogenesis in animal models. However, this should be re-evaluated in prospective clinical trials.

Advanced functional imaging helps us understand the underlying mechanisms of functional recovery from a neurologic deficit. The suggested mechanisms of cortical functional reorganization are peri-infarct reorganization, recruitment of ipsilesional or contralesional cortex, changes in interhemispheric interactions, or bihemispheric connectivity [14]. Active rehabilitation treatment might improve the neurologic deficit mediated by one of the above mechanisms.

Several functional imaging studies using SPECT or PET have demonstrated that functionally connected but structurally distant brain regions acted suboptimally after primary brain injury, which is called diaschisis [15]. After the acute phase, spontaneous neurologic recovery happens with the reversal of this type of functional impairment. Therefore, reversal of diaschisis is one of the mechanisms of spontaneous functional improvement. The most common form is crossed cerebellar diaschisis which occurs in the contralateral cerebellum after hemispheric stroke, mediated by the descending glutamatergic crossed corticopontocerebellar pathway. In middle cerebral artery infarction, the degree of crossed cerebellar diaschisis is well correlated with the neurologic deficit early after stroke [16]. Moreover, functional inhibition may occur ipsilaterally to the subcortical lesion (thalamocortical diaschisis), which is regarded as an underlying mechanism of subcortical aphasia or neglect [17].

Perilesional Cortex

Experimental studies in nonhuman primates showed that the representative hand areas in the motor cortex started to shrink after lesioning, and the cortical areas representing elbow or shoulder expanded [18]. Even in humans, ipsilateral perilesional cortical activation including premotor or supplementary motor area is a common finding after primary motor cortex injury. The descending fibers from the premotor area are less dense and less excitatory, and project to the proximal part of the arm [5]. Therefore, there is a possibility that chronic ipsilateral premotor area activation sometimes competitively inhibits distal hand motor recovery. Studies from well-recovered stroke patients suggest that ipsilateral perilesional cortical activation is associated with functional recovery, at least in the acute period. Inhibition of those recruited areas using transcranial magnetic stimulation resulted in reappearance of previous neurologic deficit. Even in cases of aphasia, the major component of recovery is associated with perilesional tissue activation, which underscores the importance of the integrity of perilesional brain issues [9,19].

Contralesional Cortex

In the recovery phase, the corresponding area in the contralateral cortex frequently shows coactivation. However, it is still debatable whether contralateral cortical activation is beneficial. In patients with aphasia, the contralateral nondominant hemisphere helps with neurologic recovery [20]. Studies from aphasic patients showed that cerebral blood flow was increased in the right inferior frontal lobe along with recovery. Other studies showed bihemispheric temporal and frontal engagement in auditory verbal processing during the recovery process. Meanwhile, a new balance in the cortical activation is needed in the chronic stage. Therefore, a decrease in the activation in the contralateral cortex is observed in patients with better functional recovery. Continuous coactivation of the mirror cortex represents maladaptive cortical mapping, which is related with nonoptimal functional recovery. The underlying mechanisms of change in contralateral cortical activation share similar physiologic changes such as unmasking of latent synapse, facilitation of alternating network, synaptic remodeling, and axonal sprouting [21].

Uncrossed Fibers from the Contralesional Hemisphere. A growing number of evidence supports that the contralesional (ipsilateral) motor cortex was activated after stroke [22]. Although the exact mechanism of coactivation of the contralesional motor cortex is still elusive, the disinhibition hypothesis is the most widely accepted [23]. With the development of hemispheric stroke, interhemispheric transcallosal inhibition is decreased from the affected side, which is translated into more activation of the contralesional motor cortex. The potential descending motor pathway from the contralesional hemisphere to the ipsilateral arm is via uncrossed ipsilateral descending corticospinal fibers, or noncorticospinal fibers, which is the corticoreticular projection, fibers passing through the red nucleus and pontine and olivary nucleus [24].

Generally, the neurologic outcome of the patients who recovered with ipsilateral (contralesional) motor cortex activation is worse than of those who recovered with perilesional reorganization [25]. Moreover, those patients experience mirror movements with recovery, which is attributed to the ipsilateral motor pathway [26]. The severity of mirror movements showed a reverse correlation with hand motor function. Therefore, abnormal involuntary mirror movement, or proximal-distal interjoint coupling may have a detrimental effect on functional recovery. Even with these conflicting results, the ipsilateral descending pyramidal tract helps trunk muscle recovery, and is an important factor in motor recovery in children.

In patients who recovered from unilateral cerebellar infarction, it seems that the cerebellocortical loop on the opposite side might be important [27]. When recovering from thalamic infarction, a somatosensory gaiting process plays a significant role in sensory improvement [28].

With the help of a sound understanding of the underlying mechanisms of the neurologic recovery and neural plasticity, pharmacological and nonpharmacological approaches to augment neurologic recovery were attempted.

Amphetamine is a monoamine agonist which increases norepinephrine, dopamine, and serotonin levels in the brain. Animal experimental studies using rats and cats showed that administration of amphetamine concomitantly with motor practice accelerated recovery from cortical injuries. Although amphetamine is a potent psychomotor stimulator, this effect is thought to be independent of its psychostimulatory effect, which is mediated by dopamine. Several human randomized clinical trials were performed to identify the beneficial effect of amphetamine on neurologic recovery. Although several anecdotal reports support that it may help a ‘speedy recovery' in small numbers of patients, it is still inconclusive whether amphetamines are beneficial for the quality of stroke recovery [29].

Antidepressants may promote neuroplastic changes mediated by surges of the amount of synaptic monoamines. Based on this, a pivotal randomized controlled clinical trial was performed and the results were recently published [30]. Patients treated with fluoxetine and physiotherapy showed better distal motor power improvement and less dependency at 3 months, compared with those with physiotherapy alone. Although the precise underlying mechanisms are unknown, fluoxetine seems to be effective via modulating brain plasticity. With the positive results, it is still unclear whether other selective serotonin reuptake inhibitors have a similar effect on neurologic recovery, or whether the routine use of fluoxetine is justifiable in patients without post-stroke depression. More studies are needed.

A randomized single-blind crossover trial was done before using levodopa administration in the chronic stage of stroke patients. Although the treated dose was low (100 mg per day), the treatment group showed better motor performance at 5 weeks after treatment, and better cortical excitability measured by repetitive transcranial magnetic stimulation [31]. This study was based on a small number of patients; therefore, it needs to be verified in a larger study.

Repetitive transcranial magnetic stimulation or transcranial direct current stimulation are noninvasive cortical stimulation methods to modulate cortical excitability in humans [32]. These noninvasive cortical stimulation techniques administered alone or in combination with various methods of neurorehabilitation were reported to be safe in the short term. However, more studies are needed to verify their long-term effect on motor recovery.

After severe motor stroke, patients may preferentially use the nonaffected limbs. This pattern of movement activates the contralesional hemisphere which may inhibit the damaged hemisphere via interhemispheric transcallosal inhibition. Constraint-induced movement therapy consists of forced use of the paretic arm aiming to decrease transcallosal inhibition in the affected hemisphere. Reduced unwanted inhibition improves the latent pathway and helps motor recovery via unmasking of the latent pathway. With constraint-induced movement therapy, expansion of ipsilesional motor maps with concomitant decreases in contralesional motor cortex activation was observed, strongly correlating with motor gains [33].

Here, we briefly reviewed the basic neurologic recovery mechanisms after stroke. Modern functional imaging helped with the understanding of basic mechanisms underlying functional improvement; however, more studies are needed to better understand the optimal mechanism in individual patients.

This study was supported in part by a grant from the Korea Healthcare Technology R&D Project, Ministry of Health & Welfare, Republic of Korea (A102065). The authors thank Dr. H. Alex Choi for constructive comments on this article.