A Clinical Sign of Canal Paresis
G. Michael Halmagyi, MB, BS, Ian S. Curthoys, PhD
\s=b\ Unilateral loss of horizontal semicir-
cular canal function, termed canal pare-
sis, is an important finding in dizzy
patients. To our knowledge, apart from
head-shaking nystagmus, no clinical sign
of canal paresis has yet been described
and the term derives from the characteris-
tic finding on caloric tests: little or no
nystagmus evoked by either hot or cold
irrigation of the affected ear. We describe
a simple and reliable clinical sign of total
unilateral loss of horizontal semicircular
canal function: one large or several small
oppositely directed, compensatory, refix-
ation saccades elicited by rapid horizon-
tal head rotation toward the lesioned
side. Using magnetic search coils to mea-
sure head and eye movement, we have
validated this sign in 12 patients who had
undergone unilateral vestibular neurecto-
my.
(Arch Neurol 1988;45:737-739)
"^^"hile impairment
function of one
of vestibular
ear compared
with the other is the pathophysiologic
basis of most vertigo, clinical exami¬
nation of patients complaining of ver¬
tigo is generally unrewarding. We
Accepted publication Nov 17, 1987.
for
From the Eye and Ear Unit, Department of
Neurology, Royal Prince Alfred Hospital (Dr
Halmagyi), and the Department of Psychology,
University of Sydney (Dr Curthoys), Sydney,
New South Wales, Australia.
Reprint requests to Department of Neurology,
Royal Prince Alfred Hospital, Camperdown, Syd-
ney, New South Wales 2050, Australia (Dr Hal-
magyi).
Downloaded From: http://archneur.jamanetwork.com/ by a GLASGOW UNIVERSITY LIBRARY User on 08/26/2013
describe a clinical sign that we have
found in patients who have lost all
vestibular function in one ear: com¬
pensatory refixation saccades elicited
only by rapid head movements to¬
ward the affected side and not by
head movements toward the intact
side. We have verified this sign by
objective head and eye movement
recordings and suggest that it may be
useful in the clinical evaluation of
dizzy patients.
PATIENTS AND METHODS
Patients
Twelve patients were examined before
and from one week to one year after unilat¬
eral vestibular neurectomy for acoustic
neuroma or intractable vertigo. All 12 had
some caloric responses from the affected
ear before surgery but no caloric
responses, even with 0°C irrigation, after
surgery. Each patient and 12 control sub¬
jects gave informed consent for these tests
and the protocols had been approved by the
hospital's Human Ethics Committee.
Clinical Examination
Patients would sit upright and fix their
gaze on a target about 3 m away. The
examiner would sit facing the patient and
give the following instructions: "Please
keep looking carefully at the target while I
turn your head from one side to the other"
(Fig 1, A). When the patients had accus¬
tomed themselves to the test and could
relax their neck muscles, the examiner
would try to turn the patient's head as
quickly as possible to one side. It was
easiest to do this if the head was already
positioned 20° or so away from the side to
which it was about to be turned. The
examiner could then observe that during
such rapid head rotations, a normal sub¬
ject would not make any saccades, indicat¬
ing that the subject's gaze had remained
fixed on target. Patients with a total uni¬
lateral canal paresis, on the other hand,
could only keep their gazes fixed on target
when their heads were turned away from
the lesioned side, ie, toward the intact side
(Fig 1); when their heads were turned
toward the lesioned side, they would
invariably make one or more, clinically
evident, compensatory, refixation saccades
in the opposite direction to the head
motion (Fig 3). The examiner can readily
observe compensatory saccades if he keeps
reminding the patient to (1) try to fix on
target; (2) avoid making amiicompensatory
saccades (ie, saccades in the direction of
head movement); and (3) avoid blinking.
With a little perseverance, all of our
patients were able to cooperate with these
instructions.
Head and Eye Movement Examination
Horizontal displacement of the head and
of the left eye were recorded, using mag¬
netic search coils.1 The test paradigm was
identical to the clinical paradigm, except
that the fixation point was 1 m, not 3 m,
away and the examiner turned the
patient's head from behind and not from in
front.
Caloric Tests
Each ear was irrigated in turn with
water at 30°C and 44°C from a closed-loop
balloon system. For irrigation at 0°C fol¬
lowing vestibular neurectomy, the ear was
syringed with 50 mL of ice water. Vestibu¬
lar nystagmus was recorded, using an
infrared scierai reflection system and
monitored with infrared, closed circuit
television.
 Fig 1.—Normal gaze fixation during rapid head turn toward intact side. A and B, With her face turned a little to the right and
her eyes fixed on a distant target, patient (professional model) waits for her head to be moved rapidly to left by examiner. C,
After leftward head movement, gaze is still fixed
Controls
Twelve normal subjects without evi¬
dence of eye, ear, or brain disease were
examined clinically and five of them were
also tested with the magnetic search
coils.
RESULTS
When the head was rapidly rotated
toward the lesioned side, all 12
patients who had undergone unilater¬
al vestibular neurectomy made clini¬
cally evident, oppositely directed,
compensatory, refixation saccades.
During rightward head rotations,
patients who had undergone right ves¬
tibular neurectomy would make left¬
ward saccades; during leftward head
rotations, patients who had under¬
gone left vestibular neurectomy would
make rightward saccades; when the
head was rotated away from the
lesioned side, the patient would make
only smooth, compensatory eye move¬
ments, and no saccades were clinically
evident. These findings were con¬
firmed by search coil recordings:
whereas during head rotation away
from the lesioned side, gaze was
almost as stable as in the control
subjects (Fig 2), during head rotation
toward the lesioned side, one large or
several smaller, compensatory, refixa¬
tion saccades would invariably occur
(Fig 4). During head rotation toward
the lesioned side, the total saccadic
eye displacement in the opposite
direction side could in acute cases
approach the total head displacement,
indicating that almost no slow, com¬
pensatory eye movements had been
generated. The peak head acceleration
in these tests was about 3000c/s2 and
peak head velocity about 300°/s. Com¬
pensatory saccades were no less evi¬
dent one year after vestibular neurec¬
tomy than they were one week after¬
ward.
None of the control subjects showed
this sign clinically. With search coil
recordings, some gaze instabilities
Downloaded From: http://archneur.jamanetwork.com/ by a GLASGOW UNIVERSITY LIBRARY User on 08/26/2013
on target so that no refixation saccades are required.
were shown, but these were symmet¬
rical and any compensatory saccades
were less than 2° in amplitude.
COMMENT 10°
These results suggest that if during
rapid rightward head movements, a
patient makes leftward saccades, then
he has a severe right canal paresis;
conversely, if during leftward head
movements, he makes rightward sac-
cades, then he has a left canal paresis.
In general terms then, oppositely
directed saccades, ie, compensatory
refixation saccades during rapid hori¬
zontal head rotations, indicate a
severe horizontal semicircular canal
lesion on the side to which the head is
100°s-'
being rotated. This simple and reli¬
able clinical sign depends on the rela¬
tive inefficacy of contralateral or
ampullofugal stimulation of the nor¬
mal horizontal semicircular canal
compared with ipsilateral or ampullo- 50 ° s-
Intact Side
petal stimulation, first described by
Ewald almost 100 years ago.2 He Fig 2.—Search coil recording of head and eye
evoked more vigorous nystagmus movement during rapid head turn toward
intact side. Eye position (e) mirrors head
from the horizontal semicircular
canal when he caused the endolymph position (h) and eye velocity (è) mirrors head
velocity (li); neither gaze position (g) nor gaze
to move toward the ampulla and so velocity (g) is perturbed enough to require
displace the cúpula toward the utricle refixation saccade. Calibrations are as fol¬
(ampullopetal flow) than when he lows: position, 0° to 10° or ±5°; velocity, 0°
caused endolymph to move away from to 100°s-' or ±50°s-'; time, 250 ms.
the ampulla and so displace the cúpu¬
la away from the utricle (ampullofu¬
gal flow). From this observation, ie,
that the semicircular canal could
respond to stimulation in either direc¬
tion, Ewald correctly concluded that
there must be some sort of vestibular zero with sufficiently rapid contralat¬
"tone," identified as the resting dis¬ eral, ampullofugal-flow-producing ac¬
charge rate of primary vestibular celerations, with ipsilateral, ampullo-
afférents by Lowenstein and Sand3 petal-flow-producing accelerations,
nearly 50 years later. no matter how rapid, the discharge
The cause of Ewald's ampullofugal- rate could not be saturated. In other
ampullopetal asymmetry was estab¬ words, the firing rate of primary ves¬
lished by Goldberg and Fernandez" in tibular afférents kept increasing with
1971. They found that while the dis¬ ipsilateral acceleration but could only
charge rate of some primary vestibu¬ decrease to zero with increasing con¬
lar afférents could be driven from tralateral acceleration.
about 90 spikes per second at rest to Subsequently, Baloh and others5·6
 Fig 3.—Clinical sign of right canal paresis: abnormal gaze fixation during rapid head turn toward lesioned side. A, With her
face turned a little to left and with her eyes fixed on a distant target, patient (professional model) waits for her head to be
moved rapidly to the right. B, Following rightward head turn, it becomes evident that gaze has shitted during head turn with
head to right. C, Leftward or compensatory saccade is now required to refix gaze.
250 ms
Fig 4.—Search coil recording of head and eye
movement during rapid head turn toward
lesioned side: peak smooth eye velocity (è) is
only about half of peak head velocity (h) so
that gaze (g) is displaced 10° to right during
20° rightward head turn, so that leftward
saccades (arrows) are required to refix gaze.
These saccades are obvious to an observer
and constitute clinical sign of right canal
paresis. Calibrations are as follows: position,
0" to 10° or ±5°; velocity, 0° to 100°s-' or
±50°s"'; and time, 250 ms. e indicates eye
position, h, head position, and g, gaze posi¬
tion.
measured the horizontal vestibulo-
ocular reflex of patients with total
unilateral canal paresis and con¬
firmed these findings, adding that the
ampullopetal-ampullofugal asymme¬
try was only evident with head accel¬
erations above lOOVs2. From our work
also, it appears that rapid head accel¬
erations are required to make the
ampullopetal-ampullofugal asymme-
Downloaded From: http://archneur.jamanetwork.com/ by a GLASGOW UNIVERSITY LIBRARY User on 08/26/2013
try of the horizontal semicircular
canal clinically obvious in patients
who have undergone vestibular neur¬
ectomy. More recently, Hain et al7
confirmed that head-shaking nystag¬
mus is also a sign of unilateral canal
paresis and that it also depends at
least in part on the operation of
Ewald's Second Law.
If the heads of normal subjects
staring at a distant target are quickly
turned sideways, their eyes will invol¬
untarily turn in the opposite direction
in their heads. Since their eye move¬
ments will be just as smooth and just
as quick as their head movements,
their gaze will remain unperturbed
and fixed on target. Gaze is normally
fixed by three distinct groups of ocu¬
lomotor reflexes: the cervico-ocular
reflex, the optokinetic-pursuit reflex,
and the vestibulo-ocular reflex. That
only one of these three, the vestibulo-
ocular reflex, is quick enough to keep
up with natural head movements is
illustrated by the predicament of the
bilaterally labyrinthine defective pa¬
tients.7 If they only turn their heads
slowly, the optokinetic-pursuit reflex
and perhaps the cervico-ocular reflex
can all generate the smooth compen¬
satory eye movements required to sta¬
bilize their gaze. If, however, they
turn their heads quickly (say faster
than 1 Hz or 100°/s peak velocity,
lOOOVs2 peak acceleration), then the
optokinetic-pursuit reflex and the cer¬
vico-ocular reflex cannot keep up and
their compensatory smooth eye move¬
ments will be too slow; they will have
to make compensatory fast eye move¬
ments, ie, saccades, to stay on target;
their gaze will become unstable and
they may even experience oscillopsia.
The gaze of
patients with bilateral
labyrinthine defects shifts with their
head, because their smooth compensa¬
tory eye movements are too slow; gaze
error accumulates during head move¬
ment and is finally corrected by oppo¬
sitely directed, compensatory saccade.
The gaze of patients with a unilateral
labyrinthine defect shifts only when
their heads move quickly toward the
lesioned side and then the accumu¬
lated gaze error can only be corrected
by the characteristic, oppositely
directed, compensatory saccades.
Gauthier and Robinson9 measured
the horizontal vestibulo-ocular reflex
similarly to the way we have mea¬
sured it; they, however, took the
amplitude of the refixation saccade,
which occurred after a rapid head
rotation, to represent the error in the
vestibulo-ocular reflex during the
head rotation.
This investigation was supported by the
National Health and Medical Research Council.
Carl Bento took the photographs of model Kate
Weiler.
References
1. Collewijn H, Van Der Mark F, Jansen TC: A
precise recording of human eye movements.
Vision Res 1975;15:447-450.
2. Ewald EJR, cited by Hallpike CS: On the
case for repeal of Ewald's second law: Some
introductory remarks. Acta Otolaryngol
1961;159(suppl):7-14.
3. Lowenstein 0, Sand A: The individual and
integrated activity of the semicircular canals of
the elasmobranch labyrinth. J Physiol 1940;
99:89-101.
4. Goldberg JM, Fernandez C: Physiology of
peripheral neurons innervating semicircular
canals of the squirrel monkey: I. Resting dis-
charge and response to constant angular acceler
ations. J Neurophysiol 1971;34:635-660.
5. Baloh RW, Honrubia V, Konrad HR:
Ewald's second law re-evaluated. Acta Otolaryn-
gol 1977;83:475-479.
6. Honrubia V, Jenkins HA, Baloh RW, et al:
Evaluation of rotatory vestibular tests in periph-
eral labyrinthine lesions, in Honrubia V, Brazier
MAB (eds): Nystagmus and Vertigo: Clinical
Approaches to the Patient With Dizziness. Orlan-
do, Fla, Academic Press Inc, 1982, pp 57-80.
7. Hain TC, Fetter M, Zee DS: Head-shaking
nystagmus in patients with unilateral peripheral
vestibular lesions. Am J Otolaryngol 1987;8:36\x=req-\
47.
8. Gresty MA, Hess K, Leech J: Disorders of
the vestibulo-ocular reflex producing oscillopsia
and mechanisms compensating for loss of laby-
rinthine function. Brain 1977;100:693-716.
9. Gauthier GM, Robinson DA: Adaptation of
the human vestibulo-ocular reflex to magnifying
lenses. Brain Res 1975;92:331-335.