Third Window Syndrome: Doctor-Patient Communication


TMWD (Third Mobile Window Disorders) are defined as a group of disorders, among which Superior Semicircular Canal Dehiscence Syndrome is the most recognized. They have a common clinical presentation including a combination of Tullio’s phenomenon, pressure/strain-induced vertigo and/or autophony, but there are certainly other associated symptoms. The “third window” effect results from the altered inner ear mechanics due to a defect in the inner ear or an aberration of its structural integrity. Initially thought of as rare when the first paper was published in 1998, TMWD are in fact very much more common than realized. If doctors elicit a full medical history from their patients, including any head trauma, otic barotrauma resulting from air travel etc., the chance of a wrong diagnosis being handed out will be greatly reduced.  The first textbook is now available as a single reference source of information for TMWD, and there follows excerpts from Chapter 29 Doctor-Patient Communication – the patient being Philippa Thomson who has had surgeries for bilateral Superior Semicircular Canal Dehiscence Syndrome, and has co-edited the book.

Doctor-Patient Communication

Effective and collaborative doctor-patient communication is central to building a therapeutic relationship between the two parties, and it enables the patient to make well-informed decisions about their treatment as well as to set expectations that can be met. Positive exchanges of this kind can be a source of motivation and support, helping both sides achieve their agreed-upon goals.

Philippa Thomson, a patient with bilateral SCDS, outlines a series of questions that arise regularly for TMWS patients and they are answered by Dr Gerard Gianoli. The direct questions seek specific information related to the topics that concern most patients who are affected by this disorder.


Q. Is there an association between a patient’s symptoms and the size of their dehiscence?

A. Size does seem to play a part in symptoms associated with Superior Semicircular Canal Dehiscence (SCD). The size of the dehiscence has been correlated with the size of the air-bone gap on audiometry. (1) The longer the dehiscence, the larger the air-bone gap noted on audiometry, in general. Specifically, a 3 mm dehiscence has been demonstrated to be a potential “critical point” at which defects shorter than 3 mm don’t tend to have conductive gaps and those greater than 3 mm do, with an increasing gap for sizes greater than 3 mm. (2, 3) The expected symptoms associated with this would be more difficulty with hearing loss and greater degree in autophony. However, the perception of symptoms and how they are expressed by individual patients varies greatly.

While VEMP testing is not a symptom, the size of dehiscence has demonstrated a positive correlation between the length of SSCD and the cVEMP and oVEMP (500 Hz) thresholds and cVEMP amplitude.(4) Similar to an increasing conductive gap, one would expect a lower VEMP threshold and greater VEMP amplitude to generally result in greater sound-induced symptoms, specifically Tullio phenomenon and sound sensitivity. However, vestibular symptoms have not been correlated with the length of dehiscence.

Concurrent tegmen defects (5) have been shown to correlate with increasing SSCD length. These, of course, do not cause any symptoms, unless there is an associated encephalocele or CSF leak.

Q. Do the majority of patients with TMWS hear their eyes move? Those who do not are sometimes told, wrongly, that they cannot therefore be suffering from that condition.

A. My experience is that most patients do not hear their eyes move. The symptoms associated with TMWS are varied. It has been called the great Otologic Mimicker due to its varied presentation and its ability to present like many other otologic conditions. Not all patients with TMWD have autophony or bone conduction hyperacusis and not all patients with autophony or bone conduction hyperacusis have TMWD. As mentioned above, patulous eustachian tube is a condition with the prominent symptom of hearing one’s own voice and breathing. Autophony to eye movement is more specific to TMWS. The Barany Society, in an attempt to organize a diagnostic criterion for SSCD listed several symptoms that are needed for diagnosis (6):

At least 1 of the following symptoms consistent with the presence of a ‘third mobile window’ in the inner ear:

      • Bone conduction hyperacusis1
      • Sound-induced vertigo and/or oscillopsia time-locked to the stimulus2
      • Pressure-induced vertigo and/or oscillopsia time-locked to the stimulus3
      • Pulsatile tinnitus

Bone conduction hyperacusis is defined as: “Symptoms can include hearing one’s voice loudly or distorted in the affected ear (auto-phony), abnormal perception of one’s own internal body sounds like hearing loudly one’s eye movements or blinking, borborygmi, crepitus from jaw or neck movements, and footfalls.”

According to these criteria, not only is “hearing your eyes move” not necessary to make the diagnosis, bone conduction hyperacusis (of any internal noises) is not necessary to make the diagnosis.

Q. Patients of TMWS often suffer from neck stiffness or pain, without realising that symptom is connected to their inner ear disorder. Can you explain why this happens – does the vestibular system partly control the neck muscles?

A. While the vestibulo-ocular reflex is the most well-known of the vestibular reflexes, there are several others. Abnormal vestibular stimulation will result in symptoms associated with these reflexes when provoked. The vestibulocollic reflex (VCR) is a well described reflex to the neck muscles whose purpose is to stabilize the head. Labyrinthectomized animals will demonstrate the classic head tilt reaction (head tilted to the labyrinthectomized side) due to abnormal stimulation of the VCR. In humans, the head tilt reaction is muted by comparison to lower animals. The main symptom from abnormal VCR stimulation is neck pain/stiffness. The most common location for this discomfort is the upper part of the neck near the insertion into the occiput.

Q. Many also complain of more generalized head pressure – is there anything in particular that might be contributing to that, or are various factors involved?

A. Generalized head pressure is a symptom of elevated intracranial pressure (ICP). In my practice, we routinely measure ICP. Some patients with TMWS are found to have Pseudotumor Cerebri and will benefit greatly from efforts to lower ICP. There are other patients with TMWD who do not have Pseudotumor Cerebri but appear to be much more sensitive to mildly elevated ICP. These patients also note significant benefit from lowering ICP and specifically the symptom of head pressure improves.

Q. In connection to the previous question, low barometric pressure has adverse consequences for a lot of patients. Is the biggest issue the change in pressure differential at the round and oval windows, with the inner ear reflecting the intracranial pressure and the other side the barometric pressure?

A. While it has been observed that low barometric pressure can cause exacerbation of otologic symptoms in TMWD patients, there have been no studies regarding this topic. That said, the only place the inner ear interfaces with atmospheric pressure is through the oval and round windows. Consequently, it would seem logical that the oval and/or round window (OW/RW) would be involved somehow in barometric pressure affecting the inner ear in TMWD. Most likely the change in the pressure across the OW/RW would cause a shift in pressure across the inner ear due to the relatively higher intracranial pressure (compared to higher barometric pressure). We know that increased pressure into the middle ear (via Nasal Valsalva) and at the OW (via Fistula Testing) can provoke vestibular stimulation in TMWS patients. It would seem likely that a similar mechanism occurs with a change to lower barometric pressure.

Alternatively, low barometric pressure has been seen as an exacerbating of Migraine, and the co-occurrence of Migraine with TMWS would seem to be assured given the prevalence of Migraine.

Q. Have you ever encountered a patient whose dehiscence has closed up on its own, or who has improved dramatically without surgery?

A. While I have never encountered a patient whose dehiscence has closed spontaneously, recently Bhatt et al. (7) reported five cases of superior canal bone growth after cartilage capping procedures. These were patients who had preoperative CT scans demonstrating SSCD and postoperative CT scans demonstrating growth of bone covering the prior SSCD. There was one additional patient who had ectopic bone growth adjacent to the dehiscence.

Symptom resolution or improvement without surgery is not unusual. It has been observed in the past that symptoms can wax and wane with long periods of remission. It appears that much of the symptom improvement (or worsening) is correlated with patient activity – the more strenuous activity, the more symptoms. Some patients recognize this and avoid triggers thus minimizing vestibular stimulation, while other patients fail to recognize this. We have advocated a number of medical measures to minimize TMWS (see Chapter 14 on Medical Therapy) as an alternative to surgical intervention, while leaving surgery for those who fail to have adequate resolution with such measures.

Q. Is it the case that TMWS symptoms quite often trigger migraine symptoms? And can treatment of the migraine sometimes give a patient sufficient relief to avoid surgery?

A. Migraine is a very common disorder and concomitant Migraine in TMWS occurs. It has been observed that Migraine patients have a prolonged recovery and somewhat less successful outcome from surgical repair than those without Migraine. Because Migraine and TMWS have overlapping symptoms, it would seem prudent to maximize therapy for Migraine before recommending surgery for TMWS. There is also a subgroup of patients with TMWS and headache that will have resolution of headache after surgery. Whether this represents a Migraine resolved by treatment of TMWS or a different headache disorder entirely is up for debate.


Q. Could you please put to rest a matter that has often delayed an accurate diagnosis for TMWS patients. If a VEMP test result is normal, can a person still potentially be suffering from the disorder?

A. VEMP testing includes both oVEMP (testing principally the utricle) and cVEMP (testing principally the saccule). cVEMP testing is the most widely available and the test most commonly referred to as “VEMP testing”. While it is often abnormally present at higher amplitude responses and lower threshold levels in SSCD, it is a test dependent on the patient’s volitional contraction of their sternocleidomastoid muscle. Inadequate muscle contraction, or inadequate muscle size can result in less responsiveness unrelated to SSCD. So, a patient who does not contract the muscle (or does not adequately contract the muscle) can have an absent cVEMP response, regardless of the status of their saccule or SSCD. Further, any labyrinthine damage can cause damage to the saccule and reduce its responsiveness to sound stimulation. Many TMWS patients have evidence of inner ear damage, and concomitant damage to the saccule cannot be ruled out. oVEMP testing is less commonly employed but appears to be more sensitive and specific to TMWS. oVEMP is less plagued with the problem with volitional muscle contraction since the only instruction for the patients is to keep their line of site elevated. However, the utricle can be damaged by the same processes that cause caloric weakness (horizontal canal dysfunction). While abnormal cVEMP or oVEMP results support the diagnosis of TMWS, normal results do not rule out the diagnosis.

Q. The diagnosis of Vestibular Migraine is possibly overused, and some might argue it has become the ‘trash can’ diagnosis when a physician doesn’t know what is wrong with a patient. Is there something specific that could be done to prevent this happening?

A. I do believe the diagnosis of Migraine is often used as a default diagnosis when a physician is unsure of the diagnosis. That said, a trial of anti-migraine therapy is very reasonable as an initial treatment strategy. However, if a patient does not respond to conventional anti-migraine therapy, the clinician should be prompted to look for other diagnoses. Unfortunately, in medicine patients often get a label that becomes difficult for doctors to look beyond, despite evidence to the contrary and despite failed therapy directed towards that diagnosis.

Q. How do you think hypothyroidism in a patient may contribute towards TMWS symptoms? Do you rule it in or out with all your vestibular patients, and how do you treat it in the positive cases?

A. Hypothyroidism has long been associated with Ménière’s Disease and treatment of Hypothyroidism has been associated with improvement of Ménière’s Disease. (8) Given that there is a significant overlap in symptoms for Ménière’s and TMWS, as well as the recent studies demonstrating a high incidence of endolymphatic hydrops among SSCD patients, it would seem prudent to screen for and treat hypothyroidism. Besides the benefit of relief from symptoms of hypothyroidism, I have witnessed significant improvement in TMWS symptoms among patients I have treated. So, from a pragmatic as well as a possible etiologic standpoint, it seems to make sense to screen for hypothyroidism. My experience has demonstrated that the most common patient that benefits from this is the patient who was already diagnosed with hypothyroidism many years earlier and is currently getting treatment. However, their current treatment has not been assessed in quite some time and, when assessed, is found to be subtherapeutic.


Q. I think there may be a certain amount of misunderstanding among the patient population about what a plugging surgery does. Plugging a semicircular canal decreases its function, rather than completely destroying it – is that correct? And the function of the canal may even have been diminished before surgery if the overlying dura was compressing it – does that happen often?

A. The concept behind plugging of the superior semicircular canal dehiscence is to prevent endolymphatic motion through the superior canal. Theoretically, this would in fact completely remove superior canal function from the plugged canal and prevent otolithic stimulation. In reality, there is still some residual movement and consequently function of the superior canal, albeit quite diminished from preoperatively. In contrast, resurfacing or capping procedures at least theoretically, preserve or improve superior canal function. In practical application, however, sometimes these approaches result in partial occlusion and reduction in superior canal function.

Regarding superior canal function, our only clinically available means for testing is vHIT. Keep in mind that vHIT only assesses the very high frequency of vestibular response, and we have no means to assess the rest of the vestibular response frequency spectrum such as we have for the horizontal semicircular canal. Regarding reduced vHIT responses in SSCD preoperatively, it has been noted to be a frequent finding but, interestingly, vHIT of the other canals (posterior and horizontal) have also demonstrated reduced response on the side affected by SSCD. (9) There does appear to be a size effect that is associated with reduced function. This has been ascribed to the possibility of herniated dura plugging a portion of the superior canal. Postoperatively for SSCD plugging, vHIT has demonstrated reduced VOR response for the surgically plugged superior canal, but also for the ipsilateral horizontal and posterior canals. (10) This has been attributed to a possible inflammatory response in the postoperative labyrinth.

Q. Occlusions/plugging surgeries have been known to fail after being initially successful. Have you ever known a successful resurfacing surgery suddenly fail?

A. I have certainly encountered failed capping and resurfacing procedures – both my own and other surgeons’. However, the more common history of failed surgery is either 1.) failure to relieve symptoms, noted almost immediately postoperative or 2.) slow return of symptoms. The typical reason for failed resurfacing surgery is failure of the resurfacing material to provide a solid repair. In this scenario, the patient may have a short reprieve from symptoms due to swelling in the area and reduced transmission of pressure from the dura. However, once this has resolved, the symptoms return to become almost identical to preoperative. In surgical exploration, these cases usually demonstrate “soft” or semisolid HA cement, or no residual resurfacing material, suggesting that it had been completely resorbed. The cases of failed capping procedures are usually due to either improper placement or slippage of the capping material. These cases usually have a similar clinical course as the resurfacing failures with symptoms returning shortly after surgery. By contrast, failure of occlusion surgeries can sometimes be delayed by a year postoperatively. This is likely due to scar contracture of the plugging material, eventually reaching a critical point where the plug in no longer of sufficient size to prevent endolymphatic movement.

Q. What if hearing loss is increasing over time, and tinnitus is getting worse? In those circumstances might a surgery prevent a patient’s hearing from deteriorating even further?

A. In general, I do not advocate SSCD surgery for hearing loss or tinnitus due to poor reliability of outcome. My main indications for surgical intervention are vestibular symptoms, autophony and noise intolerance since these symptoms are reliably resolved with surgical intervention.

While theoretically surgery to prevent hearing loss/tinnitus progressions seems to make sense, currently there does not appear to be any literature to support “prophylactive” surgery in TMWS. Having done my first SSCD surgery in January 1998, I have had the opportunity to follow postoperative patients for more than two decades. Even in successful surgeries with patients who have no TMWS complaints, there are some in whom I have witnessed their hearing to slowly decline. The cause for this may be related to the previously repaired SSCD or it may be a completely unrelated etiology. However, this observation would seem to indicate that hearing may decline despite successful surgery in some patients.


Q. Are there factors in a patient’s past history that will determine the extent and speed of their recovery from surgery? Will a ballet dancer with their excellent poise, for example, generally do better than a rather sedentary office worker?

A. Multiple factors play into recovery from surgery – some within our control and some not. It has long been known that recovery from a vestibular deafferentation procedure correlates with the process of central vestibular compensation. The requirement for central compensation is brain plasticity and an active patient – both of which tend to decline with age. Vestibular exercises enhance this process. Vestibular surgery of any kind will change the vestibular function to some degree and require central compensation. The more drastic the change in function, the more prolonged the recovery. The surgical procedure employed and the preoperative residual vestibular function both factor into time for central compensation. A patient with no measurable vestibular function will not see a dramatic change with a labyrinthectomy, and hence will not have near as prolonged recovery as a patient with normal preoperative vestibular function. So, in this example the patient with normal preoperative vestibular function will generally have the longer and more difficult recovery. One wild card with SSCD surgery is the possibility of an inflammatory reaction that can alter vestibular function much more than the intent of the surgeon. These patients will take much longer to recover in general than those who do not have an inflammatory response. However, if everything else were equal, a ballet dancer will recover much more quickly than a sedentary office worker.

Q. Could you explain what post-operative symptoms, such as autophony and ear fullness, are caused by fluid, blood and swelling in the ear, and how long they may be expected to last.

A. The most common reason for autophony is a conductive hearing loss – much more common than TMWS. Postoperatively, virtually all patients will have a conductive hearing loss due to hemotympanum and associated swelling. Patients who have window reinforcement surgery will also have some additional temporary conductive hearing loss due to the repair material utilized. How quickly the middle ear aerates, and the conductive gap resolves will vary to a great degree for individual patients, often depending on whether there is any eustachian tube dysfunction. The resolution of the above middle ear problems correlates with reduction of the autophony and fullness sensation postoperatively.

Q. Post-surgery vertigo and disequilibrium can cause patients anxiety about whether or not their operation has been successful. Is BPPV often at the root of this, and needs to be addressed with Epley maneuvers? Can any damage be caused by doing these maneuvers when the problem is actually not BPPV?

A. Postoperative BPPV is so common after surgery for TMWS, that I tell patients to expect it, especially the first week. It usually resolves without treatment but will respond to canalith repositioning. Performing canalith repositioning in the immediate postoperative period is not deleterious but I am reticent to use mastoid oscillation in the early weeks after surgery. I am unaware of any potential damage that can occur in performing an Epley maneuver when BPPV is not the cause of the patient’s vertigo. I have performed Epley maneuvers on patients where I was not sure if they had BPPV simply because it has such a favorable risk/benefit ratio. Another cause for postoperative vertigo is edema and inflammation, both of which will resolve with time.

Q. How important is post-surgery vestibular therapy? Would a patient have an equally successful long-term outcome without it, and it just speeds up that recovery process?

A. Vestibular exercises are routinely employed after surgery and many studies support their benefits in hastening a more complete recovery. The exact form of vestibular therapy is not yet defined and is being actively investigated. Are there some patients who would have successful long-term outcome without vestibular therapy? Almost certainly, but there are many who would struggle without it. It could also be argued that those who “do well” without vestibular therapy may have had a better outcome with it.

Q. When it comes to a revision surgery, are there increased risks, such as hearing loss being more likely? Could you please summarise what is known about a revision surgery’s success rate being lower?

A. As with most revision surgeries, complications tend to be higher, and success tends to be lower. Specific to SSCD surgery, the literature suggests a higher rate of hearing loss and a higher risk of CSF leak. Scarring and distorted landmarks are inevitably encountered in revision surgery and may be the cause for some of this. Success in revision SSCD surgery has been reported to be lower than primary surgery. I suspect the reason for this may have more to do with diagnosis than surgical technique. The current diagnostic criteria for SSCD includes 1.) history compatible with SSCD, 2.) testing compatible with SSCD, and 3.) CT findings of SSCD. However, just arriving at the diagnosis of SSCD is not enough. Some of these patients undergoing revision surgery will have other problems not addressed with the surgery. Some of the more common concomitant problems are second dehiscent sites, elevated intracranial pressure, perilymphatic fistula, endolymphatic hydrops, and migraine. A thorough evaluation of the patient from the start will help obviate the need for revision surgery and the very unfortunate event of failed revision surgery.


Excerpt from “Third Window Syndrome of the Inner Ear” by Dr. Gerard Gianoli and Philippa Thomson.