Visual Telerehabilitation in Children, Adolescents and Young Adults With Hemianopsia Consecutive to a Brain Tumour

Central nervous system tumors are the second most common malignancies in childhood. A brain tumor and its treatment can affect the visual system at different levels, from the optic nerves (through compression or infiltration), to sub-cortical structures such as the superior colliculus (SC) and lateral geniculate nuclei (LGN) to optic tracts, optic radiations and visual cortices. Children with brain tumors can present visual impairments like decreased visual acuity and contrast sensitivity, loss of color vision, and visual field loss such as hemianopsias. Individuals with hemianopsia present difficulties in detecting stimuli in the defective visual field and show defective scanning and exploration. Moreover, they show a rotation and compression of the auditory space leading to imprecise localization of sound across both hemispaces. Hemianopsia patients naturally develop oculo-3D-MOTor strategies to compensate for visual field loss, but visual rehabilitation procedure must still be developed to optimize/improve visual perception in the blind field. Several studies demonstrated that individuals with hemianopsia could improve visual perception in the damaged hemifield after a stimulation procedure where auditory and visual stimuli were temporally and spatially correlated. Such repetitive audiovisual stimulation programs induce a functional and anatomical reorganization of the visual connectivity in sub-cortical and cortical structures over time mediated by perceptual learning and synaptic plasticity.

The investigators have developed an audiovisual stimulation procedure using immersive virtual reality (IVR) in head-mounted display (HMD) as a delivery platform. IVR is an emerging and very promising approach for visual and auditory rehabilitation. There are currently limited practical results whether this technology is suitable for low-vision patients to use at home and if it can be deployed on a large scale. A few case report/series studies suggested a potential effectiveness of IVR on visual perception in teenagers and young adults but more information as to the potential of use and effectiveness of this technology is necessary.

Real-time measurement of task performance and neurophysiological signals while immersed in virtual reality can bring many benefits, from monitoring experience-related factors to the development of bio/neuromarkers. In Multiple Object Tracking in 3D (3D-MOT), visual tracking and detection is directly influenced by the speed of the spheres and is tailored to the participants' performance using an adapted design to efficiently reach perceptual threshold, therefore constantly challenging the patient's perceptual abilities. Progress made by the participants can be evaluated following real-time measures such as speed of the spheres, percentage of positive hits and reaction time. These outcomes naturally increase over sessions as a learning effect (generally around 5-6 sessions in our settings) but then quickly stabilize. Further progress of performance is related to improvement in vision as a strong correlation between performance at 3D-MOT and functional vision outcomes is observed in our pilot studies.

Eye-tracking measures eye position and movement using pupil position to infer fixation, gaze durations, saccadic velocities and saccadic amplitudes which altogether reveal strategies that lead to successful detection and tracking. Eye-tracking has been extensively used during the 3D-MOT task to describe the strategies of attention allocation during visual tracking. Target/head/eye-tracking measures (positions recorded every 14 ms - 90Hz) will be used to investigate the visual strategies (head movement, saccades, fixation, smooth pursuit) developed by the participants to track the targets in 3D-MOT.

Visual field recovery and restoration from hemianopsia has been postulated to occur from both improved function of perilesional tissue and recruitment of additional cortical structures to assume the function of the permanently damaged centers. However, longitudinal studies detailing functional and structural changes in the visual system before and after treatment are lacking. EEG recordings in individuals with hemianopsia revealed some residual activity in V1. Here, brain activity in V1 will be measured before and after treatment and in comparison between control and treated group using visually evoked potential (VEP). Retinal anatomy changes will be investigated between control and treated group by measuring the ganglion cell complex (GCC) integrity and the retinal nerve fiber layer (RNFL) thickness, indicative of retinal ganglion cell atrophy using spectral domain optic coherence tomography (SD-OCT), a non-invasive imaging technique. Brain scans (Diffusion MRI, Retinotopy) will be performed using a 3T MRI at the Toronto site (Advanced MRI, Toronto Western Hospital, University Health Network) to visualize the integrity (optic radiation) and activity (primary visual cortex) of the visual system between control and treated group.

Hypothesis: Home-based audiovisual 3D-MOT IVR stimulation program will improve visual perception (primary outcome Esterman binocular field test, +3 points perceived) in the blind field of individuals with hemianopsia and increase contrast sensitivity, fixation stability, reading speed and quality of life (secondary outcomes).

Following the ORBIT model, a phase IIa/b, single blind (to assessor), prospective, randomized, controlled, multi-centre, cross-over study involving 5 academic centres in Canada (St Justine - Montreal, CHUL - Quebec, Alberta Children's Hospital - Calgary, BC Children's Hospital - Vancouver and Hospital for Sick Children - Toronto) will be performed. Participants will be randomized into 2 groups: Group 1 receives no treatment for the first 8 weeks and then start the 3D-MOT IVR program. Group 2 starts with 3D-MOT IVR and then switches to an observation period for 8 weeks. Using this cross-over design, outcome measures at week 8 (W8 - Period 1) will assess the effectiveness of 3D-MOT stimulation between the control non-treated (G 1) and treated (G 2) groups (independent measures). Primary outcomes assessment at week 16 (W16 - Period 2) will measure the effectiveness of 3D-MOT IVR within Group 1 (internal control, repeated measures) and the sustainability of the treatment in Group 2 (repeated measures). Visual assessments will be performed at baseline, 8 and 16 weeks with follow-ups at 20 and 40 weeks (W21, W42).

Population:

Male and female individuals aged 10 - 40 years old diagnosed with hemianopsia consecutive to a brain tumour with no prior visual rehabilitation interventions.

Intervention:

3D-MOT in immersive virtual reality:

• 1 session every 2 days (± 1 day) for 8 weeks (28 sessions total)
• 1 session = 3 blocks of 15 audiovisual stimulation tasks (2 min break between each block).
• 1 task = 20 seconds audiovisual IVR stimulation (3D-MOT + correlated sound). Rest time is 30 sec. to 2 min. between blocks. 1 session lasts 19 min. The duration of subject's participation is 8 months, including follow-up. Treatment period: 8 weeks Follow-up period: 6 months The expected frequency and duration of study visits (anticipated time commitment) for study participants Screening/inclusion visit: 2-3 hours (visit 1) Period change visit: 1.5 hours (visit 2) Final visit: 2 hours (visit 3) Follow-up visit 1 month: 1.5 hours (visit 4) Follow-up visit 6 months: 1.5 hours (visit 5)

Primary outcome, the Esterman binocular field testing will be performed at the ophthalmology clinic of the participating centers at baseline, week 8, week 16 and follow-up at week 21 and week 42.

There are no risks for participants enrolled in the study. The use of IVR may cause moderate dizziness, nausea or disorientation for continuous stimulation above 10 minutes. If nausea, dizziness or disorientation is experienced during IVR stimulation, stopping the VR stimulation immediately restores normal condition.

Participants will be assessed for IVR sensitivity using the Virtual reality Induced Symptoms and Effects (VRISE) questionnaire score at inclusion (exclusion criteria: three (3) consecutive VRISE score <25). At-home continuous VR stimulation will be 5 minutes of continuous stimulation, below the critical 10 minutes threshold inducing effects and symptoms.

Bias minimization:

• Selection bias will be addressed by randomly assigning participants to treatment or reference group using a permuted block randomization approach (block size = 2).
• Attrition bias will be addressed by including all participants who were randomized into the study (intention-to-treat analysis).
• Performance and detection biases will be partially controlled for as primary outcome assessors will be blind of the assigned group (single blinding).
• Reporting bias will be minimized by reporting all statistically and clinically significant and non-significant results in publications.

Procedures for monitoring subject compliance. Compliance will be monitored in real-time after each block of tasks. Data are sent to dedicated and secured laboratory computer via Wi-Fi in a .csv file containing the date, time, duration and performance of the 3D-MOT IVR stimulation performed. If no files received for 72 hours, the research team will contact the participants by phone and/or email to inquire about the absence of data received. Action will be taken accordingly. Experience has shown that the main reason for the absence of data for 72 hours is a lost Wi-Fi connection. As soon as Wi-Fi connection is restored, all non-sent data will be sent automatically with timestamp.

Safety parameters:

AEs will be the safety endpoints. Known AEs induced by IVR stimulation are nausea, dizziness and disorientation. The severity of these AEs will be scored using the validated VRISE questionnaire40. Three (3) consecutive VRISE scores <25 is considered as an AE.

VRISE questionnaire will be recorded electronically (.csv file) after each home-based session (every 2 days) and sent via Wi-Fi to a dedicated and secured laboratory computer in real-time.

AE reporting will begin at the time of signing of the informed consent (screening) and will continue until discharge from the study. AEs will be elicited by:

• spontaneous report by participants, partner/caregiver and/or healthcare staff,
• by 3 consecutive VRISE scores < 25,
• by observation (by the investigators and/or healthcare staff). AEs due to IVR stimulation resolve as soon as the stimulation stops. Investigators will follow-up the participant at 24 hours by phone or email assessing the severity of the AEs using the VRISE questionnaire.

Statistical methods:

Analysis will be performed following an Intention-To-Treat (ITT) approach. Results will be reported using descriptive statistics (including frequency distributions, a measure of central tendency and a measure of dispersion) of the primary outcome, average VRISE scores and secondary outcomes. Data will be analyzed using the statistical software JASP. Statistical comparison between the 2 groups and strata for the primary outcome will be made using Bayesian and frequentist two-way ANOVA with repeated measures.