Interaction between Vestibular Compensation Mechanisms and Vestibular Rehabilitation Therapy: 10 Recommendations for Optimal Functional Recovery

Abstract

This review questions the relationships between the plastic events responsible for the recovery of vestibular function after a unilateral vestibular loss (vestibular compensation), which has been well described in animal models in the last decades, and the vestibular rehabilitation (VR) therapy elaborated on a more empirical basis for vestibular loss patients. The main objective is not to propose a catalog of results but to provide clinicians with an understandable view on when and how to perform VR therapy, and why VR may benefit from basic knowledge and may influence the recovery process. With this perspective, 10 major recommendations are proposed as ways to identify an optimal functional recovery. Among them are the crucial role of active and early VR therapy, coincidental with a post-lesion sensitive period for neuronal network remodeling, the instructive role that VR therapy may play in this functional reorganization, the need for progression in the VR therapy protocol, which is based mainly on adaptation processes, the necessity to take into account the sensorimotor, cognitive, and emotional profile of the patient to propose individual or "à la carte" VR therapies, and the importance of motivational and ecologic contexts. More than 10 general principles are very likely, but these principles seem crucial for the fast recovery of vestibular loss patients to ensure good quality of life.

Keywords: adaptation; anxiety and stress; critical period; ecologic contexts; habituation; motivation; vestibular compensation mechanisms; vestibular rehabilitation therapy.

Figure 1

(A–D) The crucial role of early and active retraining in animal models of vestibular loss. Data from animal models showing the effects of restraining the post-lesion sensorimotor activity. (A) The effects of sensorimotor restriction (SMR) applied just after a unilateral vestibular neurectomy (UVN) in the monkey (baboon: Papio papio) for a short (4 days) or longer (20 days) time duration on the behavioral state of recovery (states I–IV for maximal to minimal posture-locomotor deficits, respectively: ordinates). The behavioral recovery was frozen as long as the SMR was applied, and the time to full recovery was strongly delayed compared with unrestrained animals, the more so the longer the SMR [modified from Lacour et al. (47)]. (B,C) The effects of 1 week of SMR applied at different time windows after UVN on the recovery of posture (B) and of posturo-locomotor performance (C) in the cat model. The SMR was applied very early after UVN (SMR 1: day 3 to day 9) or later during the compensatory stage (SMR 2: day 14 to day 20). Postural asymmetry (B) as well as dynamic equilibrium (C) were strongly delayed in the cats submitted to SMR1 and SMR2 compared with unrestrained animals. More drastic effects were observed for the dynamic equilibrium function with a final level of recovery 1 year post-lesion limited to 40 and 50% of the preoperative maximal performance for the SMR 1 and SMR 2, respectively [modified from Xerri and Lacour (48)]. (D) The role of dynamic visual cues on the neuronal response of vestibular nuclei cells to optokinetic stimulation in the UVN cat. In the intact animal, the neuronal response (in impulses per second) was limited to the low frequency range of optokinetic stimuli (0–0.5 Hz). In the UVN cats submitted to passive visual optokinetic stimulation, there was no change compared with the controls, while the UVN cats dynamically receiving the optokinetic stimulation showed strongly increased neuronal responses, the more so the higher the frequency. In cats moving freely in their optokinetic environment, the visual cues substituted for vestibular input at high frequencies, up to 1 Hz, a result never seen in intact cats, which indicates that the post-lesion experience may determine new neuronal properties that alter the recovery process [modified from Zennou-Azogui et al. (49)].

Figure 2

Post-lesion vestibular plasticity and vestibular rehabilitation therapy: inter-relations during the critical period. This figure illustrates the plastic events occurring in the vestibular nuclei (VN) after unilateral vestibular deafferentation in animal models. The first event was the up-regulation of immediate early genes (IEGs) in the very early hours and days, with Fos immunoreactivity peaking 2 h post-lesion (open histograms) and Zif-268 immunoreactive neurons peaking 1–3 days post-lesion (hatched histograms). Both IEG expressions declined progressively within 3 days to 1 week [modified from Gustave Dit Duflo et al. (53) and Lacour and Tighilet (24)]. Many plastic events showed up-regulation peaking at 3 days. Neurotrophin immunoreactivity occurred 1 day post-lesion, with peak expression of the brain-derived nerve factor (BDNF: yellow histograms) at 3 days in the VN and related structures, and a return toward basal expression within the first post-lesion week. Similar spatio-temporal patterns were found for the nerve growth factor, the neurotropin 3, and their respective TrKA/TrKC receptors [modified from Lacour and Tighilet (24)]. Using bromodeoxyuridine (BrdU) as a newborn cell marker, intense cell proliferation was found in the deafferented VN after total and sudden unilateral loss of vestibular function, with a peak of cell proliferation at 3 days (dark blue histograms). Later on, cell proliferation was followed by cell differentiation (GABAergic neurons, microglial cells, and astrocytes). At 3 months, 70% of the newborn cells survived (light blue histograms) [modified from Tighilet et al. (54) and Lacour and Tighilet (24)]. Immunolabeling of markers of inflammatory responses such as tumor necrosis factor alpha (TNF alpha: black histograms), and of markers of neuroprotection such as manganese superoxide dismutase (MnSOD: gray histograms), showed up-regulations detectable as early as 4 h post-lesion, peaking at 8 h to 1 day and regaining normal values at 3 days to 1 week for TNF alpha. A more delayed expression for MnSOD was observed at 1 day, peaking at 3 days and returning to control values at 15 days [modified from Liberge et al. (55)]. Plasticity of the hypothalamo-pituitary-adrenal axis (HPA or stress axis) was characterized by an increased immunostaining for corticotrophin-releasing factor (CRF) and arginine vasopressine in the paraventricular nucleus of the hypothalamus. CRF up-regulation was observed as early as 1 day and persisted during the whole compensatory stage, until 1 month post-lesion (green histograms). An opposite pattern with down-regulation of the number of CRF-immunoreactive neurons was seen in the VN (not illustrated). The long-lasting activation of the HPA axis reflects chronic stress that was no longer present when the animals were totally compensated at 3 months [modified from Tighilet et al. (56)]. Increased histidine decarboxylase (HDC) mRNA expression in the TM nuclei was observed acutely with a peak at 7 days (350% of the basal level) and a return toward control values at 3 months (red histograms). HDC up-regulation during the whole compensatory stage (the enzyme synthesizing histamine) points to the role of the histaminergic system in behavioral recovery [modified from Tighilet et al. (57) and Lacour and Tighilet (24)]. The up-regulation of the GABAergic and cholinergic systems (not illustrated) found in the acute (at 1 week) and chronic (at 3 months) stages of vestibular compensation supports the idea that they play a significant role in the maintenance of compensation in totally compensated animals. All of the plastic events occurring during the first post-lesion month are indicative of a critical period (arrow), which should benefit vestibular rehabilitation and during which vestibular plasticity could be shaped by vestibular rehabilitation therapies.