Linear acceleration perception in the roll plane before and after unilateral vestibular neurectomy

The ability of 33 patients to perceive the direction, relative to the body long axis, of a linear acceleration vector acting in the coronal plane, rolltilt perception, was studied at various times, before and from 1 week to 6 months after unilateral, selective vestibular neurectomy for Meniere's disease, acoustic neuroma or intractable paroxysmal vertigo. The results of these patients were compared with the results of 31 normal subjects and two control patients who had both vestibular nerves surgically sectioned. Rotating on a fixed-chair centrifuge in an otherwise darkened room, each observer was required to indicate his perception of the direction of the resultant gravito-inertial vector by setting a small, motor-driven, illuminated bar, attached to the chair but rotatable in the frontoparallel plane, to the perceived gravitational horizontal. Normal subjects accurately align the bar with respect to the gravito-inertial resultant vector which, in the dark, they assume to be the gravitational vertical. This percept has been called the oculogravic illusion. Accurate roll-tilt perception is due to vestibular (probably mainly otolithic) sensory information since patients with bilateral vestibular neurectomies do not perceive the resultant vector accurately. Whereas normal subjects perceive resultant vectors directed to the right and to the left equally accurately, roll-tilt perception was invariably asymmetrical one week after unilateral vestibular neurectomy. Even at rest there was an asymmetry in the baseline settings, so that patients set the bar down on the side of the operated ear, in order for it to appear gravitationally horizontal: if a patient had a right vestibular nerve section then he set the bar clockwise (from the patient's view) below the true gravitational horizontal. With increasing gravitoinertial resultant angles there was an increasing asymmetry of roll-tilt perception due both to decreased sensitivity to roll-tilt stimuli directed towards the operated ear and to transiently increased sensitivity to roll-tilt stimuli directed towards the intact ear. A progressive decrease in both perceptual asymmetries followed, rapidly in the first 3 weeks, more slowly in the next 6 months. Based on these results, which are consistent with what is known about the responses of primary and secondary otolithic neurons to linear acceleration, we propose: (1) that the asymmetric roll-tilt perceptual response following unilateral vestibular neurectomy is an otolithic analogue of Ewald's second law; (2) that the perceptual asymmetries may be due to decreased spontaneous activity in the deafferented lateral vestibular nucleus; (3) that the progressive recovery of roll-tilt perceptual symmetry after vestibular neurectomy may be part of the otolithic component of the total recovery phenomenon known as vestibular compensation; (4) that ocular torsion caused by the unilateral vestibular neurectomy is a major factor contributing to the systematic errors in baseline settings to the gravitational horizontal one week after operation.