VEMP testing in Vestibular Nerve Disorders

Timothy C. Hain, MD • Page last modified: March 7, 2021 Return to testing index

See also: oVEMP testing • and VEMP testing.

Herein we use the terminology of "cVEMP" for cervical VEMP, and "oVEMP" for ocular VEMP.

The vestibular nerve has two main divisions -- inferior and superior. If the superior vestibular nerve is affected, such as in vestibular neuritis (or neuronitis), oVEMPs should be absent on that side, while cVEMPs should be unaffected. If the entire vestibular nerve is affected, such as in acoustic neurinoma, in theory anyway, both should be sensitive and be diminished by the lesion.

There is an odd pattern in the literature to attribute abnormal oVEMPs to "inferior vestibular neuritis" (e.g. Murofushi et al, 1996). In our opinion, this is circular reasoning at worst, and putting the cart before the horse at best. We would like to see autopsy confirmation that this entity exists and correlates with oVEMP testing -- this may take a long time.

Vestibular neuritis (neuronitis) (VN)

VEMP obtained in an individual with left sided Ramsay Hunt syndrome, using a Bio-Logic Navigator Pro. Ramsay Hunt is a facial weakness, sometimes combined with 8th nerve neuritis, due to a recurrence of Herpes Zoster (the chicken-pox virus).

In a recent review, Fife et al (2017) stated that in regard to vestibular neuritis and other vestibular nerve disorders, "Evidence is insufficient to determine whether cVEMP/oVEMP use would clarify which vestibular structures are affected in VN".

Suggesting that cVEMPs are sensitive to saccule pathway disease, Ochi and associates (2003) reported use of cVEMPs to diagnose vestibular neuritis involving the inferior division of the vestibular nerve. Because the saccule is supplied by the inferior division, cVEMPs should be absent in this situation. cVEMP does not distinguish between the saccule and inferior vestibular nerve, and available techniques seem unlikely to be able to resolve between these two. It seems to us to be a bit of a stretch to infer that cVEMPs are diagnosing inferior neuritis -- why not just saccule damage ?

Shin et al (2011) atempted to differente VN into two types -- "superior and inferior", using VEMPs, and suggested that oVEMPs were different than cVEMPs. Again, the logic here seems a bit backwards -- one should start with pathology, and correlate physiology. These authors inferred pathology from physiology.

As most types of vestibular neuritis spare the inferior vestibular nerve, cVEMPs would be expected to generally be normal in vestibular neuritis. In fact, in our clinical practice in Chicago, we sometimes use the pattern of very abnormal VENG, accompanied by very normal cVEMP to strongly suggest Vestibular neuritis -- i.e. we use it to exclude total vestibular nerve disease rather than to detect vestibular nerve disease. We also use similar logic to diagnose bilateral vestibular neuritis. When cVEMP's are abnormal in vestibular neuritis, they recover more rapidly than canal related tests (Kim et al, 2008). Adamec et al (2014) reported general improvement in oVEMP, but no change in cVEMP during the follow-up of vestibular neuritis. This might suggest that as vestibular neuritis is commonly confined to the superior vestibular nerve, oVEMPs which are also driven by this nerve improve with follow-up.

Galvanic VEMP stimulation stimulates the entire vestibular nerve and accordingly would be expected to be present even if the inferior vestibular nerve were damaged. Thus an absent sound-VEMP and present galvanic-VEMP would not differentiate between a saccule lesion and an inferior vestibular nerve lesion.

Prolonged latency of cVEMPs has recently been suggested to be a sign of a retrocochlear (vestibular nerve) lesion, such as is found in vestibular neuritis. We are somewhat dubious about this and think that in this case the first wave of the cVEMP may simply be missing, but later waves related to cochlear function are preserved, causing the appearance of a prolonged latency. Abnormal cVEMPs (asymmetrical or long latency) are reported in about 25% of persons diagnosed with vestibular neuritis (Murofuschi et al, 1996).  Prolonged latency also occurs in persons with long necks (of course).

In Vestibular neuritis, cVEMP's reportedly normalize more rapidly than canal related tests (Kim et al, 2008). However, this conclusion is suspect as the VEMP technique used in this study (head-turning) is intrinsically flawed (see discussion on main VEMP page).


Acoustic Neurinoma and VEMP testing.

cVEMPs are generally absent or reduced in persons with acoustic neuroma scheduled for surgery (Murofushi et al, 1998). This of course implies that these patients had large tumors, and does not bear on whether cVEMPs might be a useful screening test for acoustics.

cVEMP asymmetry in patient with small intracanalicular acoustic neuroma, without any hearing deficit.

Taylor et al (2015) reported on oVEMPS in 50 patients with acoustic neuroma. They reported a sensitivity ranging between 36 and 61% for a battery of VEMP and VHIT. They suggested that dysfunction of superior and inferior vestibular nerve function occurs in parallel for most patients. The 36-61% sensitivity does not compare very well with contrast MRI, and thus it would not seem that these tests are a substitute.

Ogawa et al (2018) reported that bone conducted cVEMPS do not work in acoustic neuromas. Presumably this is because sound goes to both ears. They said "BCV cVEMP did not correlate with ACS cVEMP. Therefore, BCV cVEMP cannot be used as a substitute for ACS cVEMP."

Failed vestibular nerve section: VEMPS are not very likely to work

Vestibular nerve sections fail to control intractable vertigo due to Meniere's disease in about 5% of patients. When they fail, the question can arise whether the nerve section was incomplete. VEMPs might, in theory, be useful for detecting residual function in the inferior vestibular nerve, as this branch of the vestibular nerve is sometimes intermingled with cochlear fibers (Lehnen et al, 2004). To do this, one wants a very sensitive test and specific test for vestibular damage. However, as cVEMPs require a very strong stimulus, it seems unlikely that they would be very sensitive. Head impulse tests for the posterior canal reveal residual function more more frequently than do cVEMPs (Lehnen et al, 2004).

VEMPs are just not the best choice -- there is ambiguity regarding their source (auditory, vestibular, other inputs), as well as a tendency for them to vanish with age (i.e. they are insensitive).

That being said, cVEMPs should be and nearly always are absent in persons with vestibular nerve section, and if one gets a VEMP in that situation, there is likely something remaining. The odds of this happening is tiny however.

Expected absent VEMP in person with Right sided VNS. See this page for more about this patient.

A potential pitfall in doing cVEMP's in persons with VNS are contralateral responses and extraneous responses (mainly posterior auricular muscle responses). You should NOT do bilateral stim cVEMP's in persons with VNS. You should also check for the PAM (by doing a test with the head resting on the table) if there is something that appears to be a response on the sectioned side.

Generalized polyneuropathy

There are occasional reports of delayed VEMP responses in patients with generalized polyneuropathy. Poretti et al (2013) reported reduction of VHIT gain and increased cVEMP latencies in 15 patients with CMT. This supports the conjecture that CMT is a cause of bilateral vestibular impairment. This may be related to the underlying pathophysiology of demyelination. We are dubious that delayed cVEMPs are generally diagnostic of neuropathic vestibular damage, as the delay is not large and these potentials are a bit noisy, but in the right context (known CMT in a young individual), they might be useful.

Bottom lines

Of course, there are much easier ways to detect unilateral vestibular loss than oVEMPs. For example, just examining the patient with video Frenzel goggles and checking for spontaneous nystagmus, vibration nystagmus and head-shaking nystagmus. The curious thing might be to find an oVEMP when one is not expecting one.