Timothy C. Hain, MD • Page last modified: September 4, 2022The main role of last resort treatment is to manage intractable vertigo. For the most part these are surgical. Surgical treatments have not been shown to preserve hearing to any greater extent than medical treatments (Kinney et al, 1996). A review of surgical treatment has recently been published by LaRouere (1996).
Briefly, a decision is made whether or not medical management has failed and whether or not more vigorous attempts at management are indicated.
For intractable unilateral Menieres disease cases, we often presently advise a low dose gentamicin protocol. Gentamicin is administered through the ear drum once a month for a total of one or two administrations. We have had very good results with this procedure, and minimal hearing loss. If gentamicin fails, then if symptoms are severe we may follow with a labyrinthectomy/BAHA combination, or in patients with good hearing and good health, vestibular neurectomy. The jury is still out concerning whether endolymphatic duct clipping (e.g. Saliba et al, 2015) is an effective treatment.
Persons who have intractable bilateral Meniere's disease are offered a trial of steroids with a safer immunosuppressant should this be effective, a modified gentamicin protocol, and very occasionally systemic aminoglycosides to deaden both ears.
There are certain procedures that we only rarely recommend, and for this reason we have left them out of the algorithm- -these include endolymphatic shunt procedures and "sac" procedures. The endolymphatic shunt, like all "sac" surgery, is probably a slightly effective treatment. We rarely recommend a shunt or sac procedure in our (non-surgical) practice -- we prefer low-dose gentamicin. There is a large literature about the effect (or lack of effect) of this surgical treatment. Again, overall it seems to be a "slightly effective" procedure.
A systemic review comparing intratympanic treatments (ITG -- gentamicin, ITC -- corticosteroids) was performed by Yaz et al (2020). They reported that "The result of this systematic review shows some benefit of ITG over ITC for subjective outcomes and no difference regarding objective outcomes or complication rate. However, this superiority of ITG is rather weak. Both interventions can be effectively and safely used in controlling Meniere's disease in acute situations." So it would seem that either gentamicin or steroids can be reasonably offered.
The following text discusses the major branch points in the algorithm above:
|Trans (intra) tympanic injection of Gentamicin for control of intractable Meniere's disease. This method can also be used to inject steroids (see below).|
In severe cases of Meniere's disease, treatments that deaden the inner ear such as gentamicin injections or surgery may be considered. This is a last resort for persons who have severe disabling attacks. At present, we favor gentamicin for most instances where destructive treatments are being considered. Injections of gentamicin are given through the ear drum, through a small hole or through a small tube. This procedure allows one treat one side alone, without affecting the other.
Note that this is not the "high-dose" gentamicin protocol and also that there is no reasonable expectation of hearing loss from this procedure either. We have encountered many situations where otologic surgeons propose highly invasive destructive procedures, and claim that all procedures (including low dose genta micin) will take out hearing. If you encounter this claim, we strongly suggest getting another opinion from an expert.
We, as do others (Cohen-Kerem et al, 2004), generally obtain very good results with just one or two injections. This "low-dose" gentamicin protocol has become part of our standard staged management. The risk to hearing and balance is minimal for this procedure (although there is substantial risk from older versions of this protocol, where 4 doses were administered over a month).
Dizziness may reoccur one year later, requiring another series. This is because the gentamicin may only temporarily damage the hair cells, which can recover. This happens in about 10%.
Nerve section for unilateral Meniere's that has not responded to low or high-dose gentamicin.
Alternatively, a surgical treatment is used in which the vestibular nerve is clipped. This operation, called a vestibular neurectomy or vestibular nerve section is very effective in eliminating vertigo. Oddly enough, some authors put nerve section into the same category as low-dose gentamicin (e.g. Wegampola et al, 2017). We think that this is unfortunate, we see this as analogous to cutting off one's hand when there is a cancer on one's index finger. Yes it works, but one can get the same result from a far less invasive method.
While very effective, vestibular nerve section, especially the hearing-sparing variant (done via the middle-fossa approach), is presently generally felt to be of much higher risk than gentamicin injection. We presently do not favor vestibular nerve section, except in situations where gentamicin injection has failed.
Miyazaki et al (2017) reported on a new procedure called "minimally invasive vestibular neurotomy". While it would seem likely that this procedure might "work", the issue here is risk. We would find it very difficult to understand how any procedure involving cutting the vestibular nerve could be superior in the risk profile to low-dose gentamicin. At this writing -- it seems to us to be a more "blind" method of doing a vestibular neurectomy, and best avoided unless there are special circumstances -- failed gentamicin, and high risk surgical candidate.
Another operation, called a labyrinthectomy is recommended in persons who have lost all usable hearing or in whom vestibular nerve section is considered too dangerous, or where hearing has been already lost so there is less to lose. Again, this procedure seems ONLY applicable to situations where gentamicin has failed and hearing is already gone. In persons with unilateral disease, one can combine a labyrinthectomy with a BAHA hearing device, and end up with more usable hearing after the labyrinthectomy.
Another rarely used surgical procedure, with even less morbidity but less assurance that it will be effective, is a "tack" procedure. We think it is best to avoid doctors who advice the "Cody Tack".
Intramuscular streptomycin for bilateral Meniere's disease that has not responded to low dose gentamicin.
For bilateral Meniere's disease, when the patient is incapacitated and it cannot be determined which ear is causing the dizziness, intramuscular streptomycin (1 gm twice/day) can be given on an outpatient basis until the first sign of ototoxicity develops (generally about a month). This treatment can generally reduce or eliminate vertigo spells without affecting hearing. This treatment however damages the inner ear and causes bilateral vestibular paresis, which has it's own set of symptoms and disability. This treatment is extremely rare as well as rather arduous. Basically, nobody wants to destroy the balance system.
Occasionally though, one may approach the problem similarly by using gentamicin injections on both ears. This is possible but rare as well.
This very aggressive treatment was recently reported by Mackeith et al (2014). The basic idea is to eliminate all inner ear function with bilateral labyrinthectomies, and use a cochlear implant to provide "bionic" hearing. We are unenthused about this idea for several reasons: First, surgical removal of the inner ear would make it impossible for hair cell regeneration to every be useful. This treatment is currently in clinical trials, and we expect that it will be eventually, be able to repair some of the "original equipment" in Meniere's disease. Second, bilateral labyrinthectomy is an irrational procedure. One can use streptomycin (see above), or just bilateral gentamicin, to eliminate bilateral vestibular function without affecting hearing. One could then still implant one ear, leaving both vestibular systems potential candidates for hair cell regeneration in the future.
Intratympanic steroid injections have been used to treat Meniere's (e.g. Shea and Ge in Otol Clin NA, 1996, 29:353-8, also Am J Otol,18:4,1997). The core methodology is to inject in enough dexamethasone solution (24 mg/ml) to fill the middle ear, on either a one-time basis or over several sessions separated by weeks or months. Thus it involves both making a hole in the ear drum (with the needle), as well as administering a drug to the inner ear (by diffusion across the round window membrane).
Use by otologic surgeons of this methodology is growing and papers supporting its use are proliferating very rapidly . Not all methods seem to work -- some comments about variant methods follows. This page -- intratympanic steroid injections -- has more detail.
Studies using a brief period of administration, suggest that it is no better than placebo (Silverstein et al, 1998). Wildly different to this idea, Patel et al (2016) reported that in 30 patients with Meniere's, two injections of methylprednisolone was as effective as gentamicin injections (about 90% in each group of 30). The same fortunate patients were followed up in a subsequent paper (Harcourt et al, 2019), and they continued to do well. As steroids are neither curative nor long-lasting agents, brief intratympanic steroid treatments are obviously not able to cure Meniere's disease (Barrs, 2004; Dodson et al, 2004). Okuda et al (2021) suggested that treatment response was variable. They stated "Satisfactory control of vertigo for 1 year after the first round of injection was found in 18 patients (56.3%; the response group). However, the injections failed to control vertigo in the other 14 patients (43.8%; the non-response group)". So roughly a 50-50 response rate. These patients typically got 3 injections.
Our view is that one-time steroid treatment (defined as a brief burst) has not yet been proven effective in Meniere's disease or to have a reasonable scientific basis either (Doyle et al, 2005). The core scientific difficulty is that steroids don't make any permanent alterations to the inner ear, and that they are gone from the ear in a short period of time (Harugnani et al, 2006). Meniere's is a chronic ear disorder. The hypothesis that one or two injections of steroids can "cure" a chronic ear disorder is very hard to understand. In another study, when compared head-head, intratympanic Dexamethasone was far worse than intratympanic gentamicin (Casani et al, 2011). In our opinion, we think that the Patel (2016) trial and the Harcourt et al (2019) followup of the same patients was likely just underpowered (i.e. they studied some fortunate patients). We do not think it is wise to rely on an irrational treatment for Meniere's, especially in dangerous situation such as drop attacks or unpredictable spells while driving.
On the other hand, if one has Meniere's disease in both ears (bilateral Meniere's disease), and one is not certain which one is causing dizzy attacks, it would seem reasonable to us to first try steroids. If the steroids work for 3 months, then one might reasonably move on to gentamicin. If they don't work, then either the choice of ear was wrong, both ears are active at the same time, or the diagnosis is wrong (migraine is the usual alternative confounding diagnosis in bilateral Menieres).
IT Steroids vs no treatment at 6 months and 2 years Attrache et al (2016) -- triangles are active treatment with IT dexamethosone, circles are no treatment. Both groups had 24 subjects so this is not a large study.
Studies using a long term protocol, or an "as needed" methodology (no more than 4 injections/year) suggest that their results are very good (Sennaroglu et al, 2001; Boleas-Aguirre et al, 2007; 2008, Leng et al, 2016). Boleas-Aguirre et al reported "acceptable" vertigo control in 117/129. Here there were sufficient patients, but of course, there is some fuzziness to what one considers as "acceptable".
Attrache et al (2016) reported good improvements with several injections over no treatment by 0-6 months, but similar results to both injections and no treatment at all at 18-24 months. Note that the control group here was not a placebo injection, but rather was just no treatment. To emphasize, at 2 years, doing nothing at all is almost as good as injections of steroids through the ear drum. This is similar to observations about the natural history of improvement in Meniere's made in the past by Torok and others. It presents a rather high bar for treatment studies. Still there seems to be a role for IT steroids here buying time for the natural history to kick in.
Otonomy undertook a trial involving a steroid gel that slowly dissolves in the middle ear. This is still in progress, but we would expect similar results to other steroid regimens. Variants of long term treatment that extend the duration make more sense to us than single injections, but likely will have their own problems If this methodology does indeed prove out, then the main issue would be to determine where it belongs in the staged management of Menieres. Perhaps, after medical treatment has failed, prior to gentamicin treatment, based on the characteristics of the patient.
The problems with multiple steroid injections as well as presumably other long term steroid methods are substantial - -there is a higher prevalence of holes in the ear drum (as steroids impair wound healing). Regarding 3 months of steroids, -- think what might happen if your skin was exposed to a steroid gel continuously for 3 months. Will there be more middle ear infections ? Steroids are not a "durable" or a "sure" treatment, and anyone treated with steroids might have a relapse. Thus steroids would not seem a very good method of controlling Meniere's in persons who have function safely for large periods of time -- such as someone who drives to work. Steroids could possibly be done every 3 months - -one patient we encountered had it done 40 times. This could cost quite a bit compared to the "one-shot" protocol with gentamicin.
Other variant methods of administration of steroids.
Intratympanic steroids combined with a viscous osmotic agent (hyaluronic acid) that reduces endolymph volume, was reported in an uncontrolled study to be effective (Selivanova et al, 2005) in persons who have failed intravenous steroids and vasodilators. They used a protocol of every other day injections, with a maximum of 7 total, combining 8mg of dexamethasone and a 0.2 mg/ml solution of hyaluronic acid. In our view, this is simply the combination of two short-lived treatments, but because hyaluronic acid is very viscous, the combination may keep the steroids in the ear for longer periods than simple injections of an aqueous solution, and might have better results due to this effect. A comparison between steroids injected with and without hyaluronic acid would be a reasonable next step.
Betahistine is a medication commonly used for Meniere's disease, with rather poor data establishing efficacy, but nevertheless huge enthusiasm for its use world-wide . Recently, led by Strupp, it has been suggested that betahistine can/should be used in much higher doses for Meniere's. Supporting this train of thought, Albu et al (2014) recently reported that high-dose betahistine (i.e. 144 mg/day) achieved similar results as intratympanic dexamethasone. This was randomized and double-blind. There was no statistical difference between the two groups in vertigo control. This doesn't say much for IT dexamethasone (no surprise there).
While we are uncertain about the validity of this result, we do typically "double up" on betahistine when there is a Meniere's flare. Most often, we go from 32 mg/day to 64 mg/day.
Betahistine can be combined with the MAO-B inhibitor selegiline to boost the bioavailability of betahistine(Strupp et al, 2018). We have had no experience with this approach. Strupp described using 2.5 mg once or twice/day of selegiline with 3 24 mg tablets of betahistine/day. In the US, as betahistine in very large doses can be very expensive, and the cost is generally "out of pocket". Thus this might be a method of keeping high dose betahistine within reason. Selegiline use comes with some caution about drug interactions.