It is an essential part of the protocol. Examinations can be carried out using videoscopy or recorded using videography. It plays a crucial role.
But any interpretation of the results can not be considered reliable unless the practitioner rigorously sticks to a straightforward methodology. Furthermore, it can be replicated within the same session (which is not the case for caloric testing). It allows one to carry out a study under physiological conditions (can anyone name another single daily movement that does not question both vestibules?) of the whole set of vestibular channels, from the periphery to the centre. For all the above, and in order to keep things clear, we will look at the origins, the physiology and the conditions for carrying it out in a step-by-step fashion and the interpretation techniques of this now essential test.
The practical side of carrying out this test is straightforward. The subject, with their head still, sitting down in a rotatory chair is subjected to a clockwise followed by a counter-clockwise rotation of 180° in 9 seconds. During the rotation, the nystagmic tremors are counted (or recorded). Each rotation is followed by a full 10-seconds stop in order to observe any nystagmic eye movements.
For one of us, the development of the test is part of a series of daily findings and undeniable truths:
- the harsh discovery that there is no correlation between signs and symptoms;
- the incredible disappointment of a patient who has gone through reeducation, is getting back to normal life and who after a follow-up ENG is told: "There is no change!"
Among the multitude of balance evaluation tools available, the first and only one that allows us to quantify the use of input and monitors the correlation of compensation and the fading away of the subject’s symptoms is the Equitest™. But it must also be acknowledged that its price severely limits its general distribution...
Knowing that balance can not exist without stabilising the gaze, that walking naturally can not happen without the ability to anticipate with the gaze the direction one is travelling in, and that balance is not an ascending system that starts at the feet and makes its way up to the head, but rather a descending system that moves downwards from the head, we wanted to see how it was possible not only to measure VOR but also to ensure the monitoring of the recovery or compensation.
As is always the case, the discovery was made by chance. E. Ulmer was kind enough to invite one of us to participate in the building of a prototype of his “video goggles”. The daily use of these, in order to familiarise ourselves with them, on each patient coming to undergo reeducation gave rise to an initial observation: the vertigal nystagmus were considerable while with the Frenzel goggles, nothing at all was observed.
We were able to note that there were many more cases of vertical nystagmus when the moment of putting on the goggles was close to the moment when the patient was sitting on the chair. Conversely, if the patient continued to be seated for a certain time, there were no more vertical movements than expected.
Faced with this finding, the important thing was to check to see if the nystagmus faded away and if this was the case, after how long. Indeed, the time it took to fade away varied greatly, anywhere from a few seconds to a full minute.
This finding enabled us to understand the importance of darkness for the non-inhibition of the nystagmus caused by the vertical downward movement of the body as the patient sat down.
If there was such persistence and such a sensitivity to vertical movements, why no employ an identical paradigm for the horizontal canals. For the latter, we use a rotatory chair. When rotating the subject, a nystagmus is noted once the chair comes to a stop. This is quite normal and logical given that although at a different speed, this is something that we are used to observing in vestibular reeducation and which we use to lower the VOR time constant on the healthy side.
Another finding was that the post-rotational response with the chair set to a low speed (that used for testing) did not necessarily correlate to the same observations, on the same subject, at high speed, such as those used for vestibular reeducation at 500°/s. Instead of asymmetrical responses and therefore a directional nystagmic preponderance, we were able to observe either post-rotational symmetrical responses or complete asymmetry of the per- and post-rotational responses.
Could the speed be the issue? To get a better understanding, we have to test a population of normal subjects. The results of the study were as follows:
- with rotation speeds of below 30° + /- 2°/s, 65% of the subjects did not present with a nystagmus once the chair stopped. The others had either one or two nystagmic saccadic eye movements.
- Lowering the rotation speed to 20°/s +/- 2°/s, the number of normal subjects, without presenting with a nystagmus once the chair stopped, was significantly higher whereas there was no significant change in the number of nystagmic saccadic eye movements during rotation.
Schematically, this can be summarised in the following way:
- with a normal subject, one observes nystagmic saccadic eye movement during rotation and there are no (or few) once the chair comes to a stop,
- subjects with a unilateral vestibular deficit present with a post-rotational nystagmus following ipsilateral rotation beating towards the unhealthy side.
In a second session, after a sufficiently long practical session to be able to verify replication of the test by a number of users, it was time to move ahead with normalisation.
Normalisation was carried out based on standard, rigorous statistical criteria, whereby numerous teams took part following the examination protocol to the letter. We therefore tested 100 normal subjects (the average age being 46), with no cochlear or vestibular otological disorders.
The next issue was how to present the results of testing in a graphic format. Conventionally, taking into account the direction of the endolymphatic flow, the per-rotational nystagmus beats to the side of rotation of the chair and therefore the side of the ipsilateral ear during rotation. Post-rotational pathological nystagmus beat to the opposite side of the direction of rotation, therefore to the side of the contralateral ear to the direction of rotation. We therefore find ourselves faced with a test that is comparable to binaural and bithermic caloric testing. The per-rotational nystagmus corresponds to the warm test, the post-rotational nystagmus, to the cold test.
It was therefore possible to adapt the design of “Professor Freyss’ butterfly” to these responses. However, we chose not to do this so as not to lead to confusion. Furthermore, in the case of a normal subject, and for a high percentage of subjects, the cold test would give a null response, in other words a bilateral areflexia!
The person who designed the graph therefore chose a dynamic representation, a curved format, to remind us that it depicts rotational testing. As with Professor Freyss’ “butterfly”, it is convenient to join together the per- and post-rotational responses for each rotation, the same way as one would link together the warm and cold responses during caloric testing.
The two lines cross each other as normal within the “circle of trust” of the graph. In a pathological situation, the crossing point of the two lines is to the side with the highest number of responses, contrary to the reading from caloric testing where this would be to the side with the deficit.
Using videography, this can be expressed as a representation of the speed values of the eventual slow phase per- and post-rotational nystagmus. This level of precision allows us to analyse the results in even finer detail.
This test, although perhaps still in its youth, is already widely used. It is nevertheless certain that formulae that enable us to measure directional preponderance and relative reflectivity shall be perfectly suitable.
With this clinical data, in order to better understand the pathological modifications, it was important to explain the underlying physiological mechanism of this test.
This test is broken down into two repeated sessions, as there is both a clockwise and a counter-clockwise 180° rotation over 9 seconds, each followed by a 10 second rest time.
Rotation is carried out at a constant speed and can be broken down as follows: an acceleration phase (25°/s2) which lasts one second, a constant speed stage lasting 7 seconds, a deceleration phase that is identical to the initial acceleration phase, therefore also lasting one second. There is no sudden stop in this test.
After one rotation of the head in the upright position, the cupula of the ipsilateral semi-circular canal in the direction of the rotation moves towards the utricle. If the rotation continues at a constant speed, the cupula will take 25 seconds +/- 2 to return to its initial position. This length of time, labelled "decay", equates to the time of discharge of the velocity storage mechanism, which is longer that the mechanical return of the cupula to its neutral position. It is during this phase that we are able to measure the time constant of the canalo-ocular system. Said time constant is, for a normal subject undergoing this stimulus, 7.4 seconds +/- 1.
It is worth noting that this value is very close to the time constant of the cupula.
Thus, during the initial phase of RCIT (in other words, acceleration at a rate of 25°/sec), the cupula moves towards the utricle, reaching maximum deviation at 7 sec. +/-1 sec. At this point, "decay" begins but one second later, deceleration begins. This in fact will bring the cupula back to it original position. When the chair comes to a stop, there will be no visible nystagmus. The two eyes jerks that are sometimes noted, falling within the realms of reality, are due to normal variability of these time constants.
One might think, taking into account our knowledge, from our observations and from our work, that the RCIT protocol is based on a pivotal value, the threshold beyond which gain no longer changes, regardless of the frequency of stimulation.
This test is ideally situated, a compromise between acceleration, duration and speed so that the cupular dynamic is respected, both in its provoked deviation and in its return to zero, also provoked.
Henceforth, it takes no effort to understand that in the case of a subject presenting with a unilateral deficit, the value of the time constant on the damaged side will fall while that of the healthy side remains the same. Hence, following rotation of the damaged side, there is a noticeable fall in response, and when the chair comes to a stop, the deceleration will activate the opposite side whose response will not be cancelled out by the return to its original position of the cupula on the damaged side. Therefore a certain number of nystagmic saccadic eye movements will be observed, beating towards the opposite direction of the rotation and provoked by the response of the contralateral side of the direction of rotation. When rotation is ipsilateral to the healthy side, the movement of the cupula is normal and therefore there is no response from the damaged side.
We have synthesized, albeit in a simplified manner, these different situations on the graphics below:
Mechanisms involved during clockwise rotation for a normal subject Mechanisms involved when rotation comes to a stop for a normal subject
Modifications to per-rotational mechanisms for a subject suffering from a unilateral deficit
Modifications to post-rotational mechanisms for a subject suffering from a unilateral deficit
From this data, the reading and plotting of the results onto the graph are easily done, with a simple glance once this becomes part of the daily clinical practice.
Main results on these graphics
Standard and Symmetrical Rotatory Chair Impulse Testing (RCIT)
Abnormal RCIT: Asymmetric with Abnormal Right Deficit
RCIT: Asymmetric with Left Deficit
Abnormal RCIT: Symmetrical but the level of per and post-rotational responses are high compared to values "within the normal range"
Abnormal RCIT: Completely asymmetric with Left-side Areflexia
Abnormal RCIT: Completely asymmetric with Right-side Areflexia
Abnormal RCIT: Graphic complete for bilateral hyper-reflectivity
Lastly, it would be of interest to carry out a correlation between this test and the rather more classic caloric stimulation. The different tests were entrusted to a number of examiners who work using the same equipment and more importantly, rigorously followed an identical protocol. We believed these to be the minimum conditions to have any chance of achieving reliable and replicable results.
VNG allows for a manual review of the results of all the tests. This therefore means it is possible to correct for the inevitable shortcomings of a single automatic treatment. We believe in fact, that a practitioner should not hand over full responsibility for the treatment and analysis of the registered results to a single machine (regardless of how powerful it might be).
We are in a position to confirm that of over 100 cases, there was no single normal RCIT with a pathological caloric test. In fact, the opposite was true. In all situations, there was evidence of imaging anomalies. This therefore clearly shows the sensitivity of the test.