New for 3/17/2020: Suggest criterion for oVEMP abnormality should be at least 25 rather than 20.
VEMP stands for vestibular evoked myogenic potentials, and SCD stands for Superior canal dehiscence. The VEMP test is very useful in diagnosing SCD.
There are several variants or "flavors" of VEMP testing. The input for the VEMP may be air-conducted sound (AC), bone-conducted sound (BC), vibration, or electrical (galvanic). The muscle being measured may be the sternocleidomastoid ("cervical"), the extraocular muscles ("ocular"), or any number of limb muscles (i.e. triceps, etc).We will use the terminology "cVEMP" to denote vestibular evoked myogenic potentials elicited from the sternocleidomastoid muscle. When we use the terms "oVEMP" or tVEMP or whatever, the small letter indicates that a muscle other than the SCM is being monitored - - such as ocular or triceps. When we use the unqualified "VEMP", we mean any vestibular evoked myogenic potential (i.e. cVEMP, oVEMP, tVEMP, etc).
Thus by combining one of the four inputs (AC, BC, vibration, galvanic), and one of the 2 main output muscles (cervical, ocular), we have the possibility of 8 different variants of VEMPs. For example, a bone-conducted ocular VEMP would be a "BC oVEMP".
|Figure left: AC cVEMP obtained in an individual with left sided superior canal dehiscence, using a Bio-Logic Navigator Pro. Right -- threshold AC cVEMP in same person. See the SCD page for his CT scan.|
In our practice in Chicago, as of mid 2019, we have done 5691 VEMPs (both cVEMP and oVEMP), and thus our experience level is very high.
cVEMPs so far have been mainly useful in documenting abnormally low thresholds in persons which largely occurs in persons with fistula or Superior Canal Dehiscence syndrome (SCD) (Brandtberg et al, 1999). Low threshold to sound, clinically, is called the "Tullio" effect. If one does not do thresholds, there nearly always is an amplitude asymmetry in this syndrome, as well as a very large VEMP in an ear with a air-bone gap. VEMP's normalize after surgery to plug the superior canal (Welgampola et al, 2008). This can be interpreted in several ways -- the saccule may be less stimulated after canal plugging, or the canal may be less stimulated after canal plugging.
The essential bits of information that might be useful are: 1). is the VEMP present at abnormally low threshold on either or both sides ? and 2). Is the VEMP absent on one side at a high threshold ? These two bits of information tell one whether there is Tullio's, and also whether there may be damage to the saccule, inferior vestibular nerve or it's projections.The presence of cVEMPs in a person with an air-bone gap (see hearing testing page) is also suggestive of SCD.
Fife et al (2017) reviewed several techniques of doing cVEMPs and their usefulness in diagnosing SCD. They reported that thresholds are the most useful to diagnose SCD. cVEMP raw amplitudes are not effective to detect SCD. cVEMP corrected amplitudes "may also be used to distinguish SCDS from controls". Thus according to this rather extensive review, either thresholds or corrected amplitudes can be used. We ourselves favor thresholds, and given that a VEMP is present at high intensity, we will always combine it with a low intensity stimulus to see whether or not the person's thresholds are low. We no longer use raw amplitudes as a diagnostic method for SCD. We tried and discarded using "corrected" VEMPs, because our feeling was that they added noise.
Noij, Herrmann, Gunan and Rauch (2018) reported that 2000-Hz tone bursts improve the detection of SCD. These conclusions were reached using research grade instrumentation, and whether or not this method is useful in the clinical world remains to be seen.
cVEMP's don't do so well in a few situations:
cVEMPs (using bilateral, binaural method), using amplitude criteria, are not always successful in detecting bilateral SCD. For this, one needs either a temporal bone CT or threshold VEMPs. We recommend doing a threshold cVEMP in any person with a complaint of dizziness induced by sound (Tullio's), should their regular VEMP be normal. We have stopped doing binaural VEMPs entirely.
We have also encountered a few patients with very low threshold cVEMPS (i.e. 65 dB) on both sides, who do NOT have SCD on CT scan of the temporal bone. These are generally adolescents, and we think that one should be very slow at getting CT scans of the temporal bone when one encounters this situation. See also the comment below about low threshold oVEMPs.
An example of a positive oVEMP in SCD.
|oVEMP obtained in a patient using sound. The right ear has clear SCD on temporal bone CT scan.|
Concerning the utility of oVEMP in SCD, our current strong impression is that oVEMPs are much more sensitive than cVEMPS for detecting SCD. We prefer the absolute amplitude criterion, supplemented by the threshold criterion. In other words, we like to see both high amplitude (e.g. 25+) and low threshold. We use the 500 Hz auditory input, and we gave up on doing oVEMPS using bone stimuli as it was clumsy and unnecessary. We do think that the high amplitude should probably be age adjusted. Amplitudes vary roughly 3 fold between the young and the old. (Piker et al, 2013). There are some youngsters that meet the amplitude and threshold criteria, but don't have SCD on CT scan. Below, we have indented the papers which which we don't agree, for various reasons. Sadly, this is most of them.
oVEMP amplitudes in persons at Chicago Dizziness and Hearing, excluding those with diagnosis of SCD (n=1902 ears). Date of data set is 8-2019
Finding of large oVEMPs in dizzy patents is rare.
The main problem with oVEMPs for diagnosis of SCD is false positives. More about this later.
Janky et al (2013) reported that "OVEMPs in response to air conduction stimulation (click and 500 Hz tone burst) provide the best separation between SCDS and healthy controls and are therefore the best single-step screening test for SCDS. " We agree but we think that setting the criteria for abnormality at 20 is too low. We think it should be higher than 20 because of false positives are just too common. We have encountered many situations where the oVEMP is > 20, but the CT scan temporal bone is absolutely normal (mainly in younger people). We are not sure where it should be, but tentatively 25 uV. Perhaps it should be on a sliding scale for age.
Fife et al (2017) reviewed 5 oVEMP studies in SCD, and reported a rather amazing variety in amplitudes. Criteria proposed for diagnosis of SCD based on amplitude included 1.5uV, "2.5 SD above mean", 9.3uV, and 8.25 uV. These criteria are smaller than normal oVEMPs reported by Piker et al (2013). It appears to us that technique must be wildly variable among labs. We ourselves use a much higher criterion -- 25 uV for SCD. We rather routinely get normal amplitudes of 5. Fife and associates said "oVEMP testing using either specific thresholds or amplitudes may be used in patients to aid in making an SCDS diagnosis". We think that the operational word here is "may".
Interestingly, the "ceiling" for oVEMPs seems to be much larger (compared to the mean) than cVEMPs. In other words, it is not all that unusual to see an oVEMP of 40 (as above) in SCD, in spite of the average oVEMP in our hands being only about 5. We have seen some as high as 120 ! This is the real reason why oVEMPs are superior to cVEMPs for diagnosis of SCD.
Fife et al (2017) also reported on two different threshold studies with oVEMPs, suggesting that thresholds of 102 or 99 dB were fairly sensitive. In our clinic, we look for much lower thresholds, about 65. We use the Bio-Logic NavPro, and perhaps thresholds vary according to equipment.
Manzari et al ( 2012b) suggested that "oVEMP testing with 500 Hz Fz BCV allows very simple, very fast identification of a probable unilateral SSCD." In 2013, Manzari and others suggested that oVEMP being present at 4000 hz is even better. Again, we would like to see more data (Manzari et al, 2013) . The 4000 hz oVEMP described here is a frequency threshold test. While this result is plausible, we think more data is needed, using standard commercial equipment, and a larger number of subjects than 26. According to Noij et al (2018), the 4000 hz stimulus can be uncomfortable which may limit the use of this method.
Verrechchia et al (2016) suggested that oVEMPS were larger on the affected ear, using 125Hz single cycle vertex vibration. We are a bit dubious about the frequency content of this signal, and wonder why it is different than Manzari's report.
Govender et al (2016) reported that in 13 SCD patients, abnormal thresholds were found in 85% of air conduction cVEMPs and 62% of oVEMPs. Bone conduction brought the oVEMP sensitivity up to 83%. cVEMP's showed evidence for saturation with larger responses at smaller intensity, while oVEMPs did not. The implication of this paper is that cVEMPs are likely better than oVEMPS for diagnosis of SCD. We don't agree with this (see Janky paper above).
Verrecchia et al (2019) reported on oVEMPs in 10 patients with SCD, and compared them to 135 with no SCD (the diagnosis of SCD was partially based on their clinical impression -- we have no gold standard here). They suggested that the limits could be lower than 20 (such as 16) for a positive oVEMP. We think this is highly unlikely to be true, and we have not changed our clinic limits of 25.
We presume that oVEMPS that are present, in spite of a 30 dB conductive loss at 500 hZ, strongly support SCD, but we have no proof of this idea. Practically, there are not very many people with conductive hearing patterns in SCD, but quite a few people with large oVEMPs, so this observation is not very useful.
No test is perfect, and oVEMPs can have both false positives and false negatives. The false positives are more obvious as usually people with very abnormal oVEMPs get temporal bone CTs. Presumably the false negatives are usually missed.
False negative oVEMPsA pair of hypotheses for false negative oVEMPs:
There are several reasons why an oVEMP might have normal amplitude and latency, in spite of gaping holes in the top of the superior canal. The pair of images above show a gigantic hole in the top of the superior canal (on the left), but also there is one on the right, accompanied by absolultely normal oVEMPs, with amplitudes of about 5 (far less than 20). In other words, this patient shows that normal oVEMPs do not rule out radiologically positive CT scans for SCD.
One reason for a false negative oVEMP is poor technique.
- oVEMPs are trickier to do than cVEMPs, as the muscles are much smaller. If the electrodes are positioned wrongly, there may be no response at all. Or perhaps just small responses. We think this is somewhat unlikely.
- Lack of cooperation on the part of the patient might prevent a good recording. If the patient pretends to look upward, but actually looked straight ahead, this would reduce the response.
- In theory, there might be "autoplugging" of the superior canal. In this instance, one would expect no oVEMP at all.
The technical issues are best tested by doing it over again with a different provider.
- An interesting interaction between technique and sensitivity is that oVEMPs are done upright, and cVEMPs (at least in our setting) are done supine. This means that in patients with a large dehiscence, there might be plugging of the open canal by dura while upright, and opening up while supine. No oVEMP, but perhaps a cVEMP (as the tests are often done upright for oVEMP and supine for cVEMP). This hypothesis has yet to be tested.
- In theory, some people's dura might be more distensible than others (e.g. those with EHD, for example). This is probably not enough to affect VEMPs however.
A false positive oVEMP for SCD is defined by a patient with a positive oVEMP (i.e. amplitude > 20, or 17 according to some) , but a negative CT scan of the temporal bone. For example, the CT scan of the patient above has oVEMPs of about 40 on each side. They clearly do not have SCD. . This particular patient was in their mid-40's.
Hassannia et al (2021) recently wrote about this, and reported 26 patients (out of about 2000 tested), with both oVEMPs greater than 17, and negative CT scans of the temporal bone. While there was a lot left out in this paper (such as for example, the total # of oVEMPs greater than 17.1%), and one cannot calculate sensitivity and specificity, it does show that these cases are found in other dizziness clinics, and that they are not especially rare.
Even if the amplitude of the oVEMP is larger than normal, it does not necessarily mean that they have SCD. Another example follows:
|oVEMP suggesting SCD (amplitudes of 21 and 23, low threshold on L).|
|CT scan of temporal bone showing normal SCC on both sides.|
The case illustrated above had a very large oVEMP on both sides as well as a low threshold on the left (at 60). Nevertheless, the CT-temporal bone did NOT show SCD as is clearly shown above, on either side. This patient meets criteria for Meniere's disease, and perhaps has a dilated utricle.
Much more study needs to be done of these cases.
We would like to know sensitivity/specificity of oVEMPs (this would be difficult to figure out as it would require a non-VEMP procedure to determine how many were missed). Our guess right now is that about 5% of oVEMPs are false positives.
|Dizzy person with low threshold on R, but normal amplitudes (of about 5).|
Occasionally one runs into the situation where the threshold is low (such as 65), but the amplitude is normal. Generally speaking this only occurs in young persons. Thresholds are more reliable than amplitudes, so here we have a highly reliable parameter conflicting with a moderately reliable parameter (absolute amplitude). Our position on this is that one should not necessarily get a CT of the temporal bone in individuals like this, as these procedures produce high radiation, which is especially to be avoided in younger persons. Rather we think it is best to watch them with noninvasive activities such as physical exam and perhaps another VEMP test on follow-up.
As SCD gets worse, oVEMPs can get much bigger. Because oVEMPs can grow proportionally so much more than cVEMPs, they are a far better screening tool than cVEMPs or audiograms. It would be irrational to use CT-temporal bone for screening of oVEMPS due to the high radiation load.
oVEMPs increasing with time First oVEMP done in 2016, 48.45 uV Second oVEMP done in 2019 (top half only as wouldn't fit on page), response of 126.
This is an example of a patient being followed with oVEMPs for 3 years -- the response just keeps growing. Note that the average oVEMP is about 10, and a response of 120 is thus about 12 times the average. This is far more growth in response than might occur with a cVEMP.
Oddly, Aetna medical insurance does not cover VEMP testing for SCD as of 2021. Other insurance companies generally cover VEMP testing. Aetna, on the other hand, covers CT scans of the temporal bone, which are both more expensive, and more dangerous to the patient. We see Aetna as being very out of step with the rest of the insurance world, but perhaps it saves them money. Go figure.