The effect of vestibulo-ocular reflex deficits and covert saccades on dynamic vision in opioid-induced vestibular dysfunction

Abstract

Patients with bilateral vestibular dysfunction cannot fully compensate passive head rotations with eye movements, and experience disturbing oscillopsia. To compensate for the deficient vestibulo-ocular reflex (VOR), they have to rely on re-fixation saccades. Some can trigger "covert" saccades while the head still moves; others only initiate saccades afterwards. Due to their shorter latency, it has been hypothesized that covert saccades are particularly beneficial to improve dynamic visual acuity, reducing oscillopsia. Here, we investigate the combined effect of covert saccades and the VOR on clear vision, using the Head Impulse Testing Device-Functional Test (HITD-FT), which quantifies reading ability during passive high-acceleration head movements. To reversibly decrease VOR function, fourteen healthy men (median age 26 years, range 21-31) were continuously administrated the opioid remifentanil intravenously (0.15 µg/kg/min). VOR gain was assessed with the video head-impulse test, functional performance (i.e. reading) with the HITD-FT. Before opioid application, VOR and dynamic reading were intact (head-impulse gain: 0.87±0.08, mean±SD; HITD-FT rate of correct answers: 90±9%). Remifentanil induced impairment in dynamic reading (HITD-FT 26±15%) in 12/14 subjects, with transient bilateral vestibular dysfunction (head-impulse gain 0.63±0.19). HITD-FT score correlated with head-impulse gain (R = 0.63, p = 0.03) and with gain difference (before/with remifentanil, R = -0.64, p = 0.02). One subject had a non-pathological head-impulse gain (0.82±0.03) and a high HITD-FT score (92%). One subject triggered covert saccades in 60% of the head movements and could read during passive head movements (HITD-FT 93%) despite a pathological head-impulse gain (0.59±0.03) whereas none of the 12 subjects without covert saccades reached such high performance. In summary, early catch-up saccades may improve dynamic visual function. HITD-FT is an appropriate method to assess the combined gaze stabilization effect of both VOR and covert saccades (overall dynamic vision), e.g., to document performance and progress during vestibular rehabilitation.

Figure 1. Eye and head movements during head-impulses: representative subject (A) and subject triggering covert saccades (B).

Before remifentanil administration (A and B, left), the vestibulo-ocular reflex (VOR) was intact. Eye movements (black) compensated for passive head rotations (grey). Figure A, right, shows the behavior of a typical subject during remifentanil administration. The VOR was impaired (head-impulse gain 0.56±0.11). The eyes were not sufficiently compensating for the head rotation, moving in the direction of the head movement. The subject initiated re-fixation saccades. These saccades occurred after the head movement when the visual stimulus was switched off (dashed lines indicate the mean time interval during which the visual stimulus was displayed). Reading ability during head motion was impaired (HITD-FT rate of correct answers 13%). One subject (B, right) could perform catch-up saccades during stimulus presentation in 60% of the head thrusts. Reading ability was intact in spite of a decreased VOR (head-impulse gain 0.59±0.03; HITD-FT rate of correct answers 93%). Eye and head velocity traces were aligned to stimulus start.

Figure 2. Correlation between vestibulo-ocular reflex (VOR) gain and HITD-FT scores during remifentanil administration.

HITD-FT rate of correct answers and head-impulse gains are shown; each dot represents mean results from one subject, values computed during left and right head rotations were pooled. Empty dots show data for head-impulse gain and HITD-FT scores before remifentanil administration. Black dots represent data during remifentanil administration. Gray dashed lines indicate the changes within each subject. The percentage of covert saccades performed during stimulus presentation is reported for each subject. The black vertical dashed line indicates the threshold for a pathological head-impulse gain (<0.7, in analogy to [11]). For 12/14 subjects, HITD-FT scores and head-impulse gains correlated: low gain, reflecting a deficient VOR, affected the ability of reading during head movements, leading to a low HITD-FT score (solid line, R = 0.63, p = 0.03). Two subjects showed high HITD-FT scores, which fell out of the one-tailed 99% confidence interval of the HITD-FT scores from all subjects during remifentanil administration. These subjects were also the only ones being detected as outliers (see Methods); box shows interquartile range, whiskers extend to the most extreme values not considered outliers (MATLAB standard boxplot). One of these subjects showed a high HITD-FT score with a normal head-impulse gain (right upper corner). The other subject (black box, example B in Figure 1) could trigger “covert” saccades during the stimulus presentation in 60% of the head rotations. The HITD-FT was intact despite a pathological head-impulse gain.