PATHOPHYSIOLOGY

Under normal physiologic conditions, the cupula has the same density as the surrounding endolymph. The SCCs are normally not sensitive to linear acceleration (e.g., gravity). A fixed cupular deposit would render the posterior canal crista sensitive to gravity, however.¹⁶ Rotation of the canal in the earth’s vertical plane during the Dix-Hallpike maneuver would produce cupular displacement through the gravitational pull on the deposit, resulting in nystagmus and vertigo. This condition, so-called cupulolithiasis, may represent the extremely rare, more chronic form of this disorder. Cupulolithiasis also seems to be more prevalent in lateral canal BPPV.¹⁵

The more common, self-limited variant and the variant with remissions and recurrences likely have a different pathophysiologic mechanism. Free-floating endolymph particles within the posterior canal produce a Dix-Hallpike response identical to that of a fixed cupular deposit.¹⁷ Because the posterior canal is the most gravity-dependent part of the vestibular labyrinth, free-floating endolymph particles have a predilection for settling in the posterior canal endolymph.

With the head upright, the most dependent part of the canal is the area just posterior and inferior to the ampulla on the side of the cupula opposite the utricle. As the posterior canal rotates during the Dix-Hallpike maneuver, the particles initially rotate upward because of their inertia. After a short latent period, gravity pulls them down and away from the cupula (utriculofugal) to a more dependent position. Their hydrodynamic drag creates an endolymph current in the same direction, displacing the cupula away from the utricle. Utriculofugal displacement of the posterior canal cupula increases the resting discharge rate of the hair cells and first-order vestibular neurons. As known from previous animal studies, this action produces excitation of the ipsilateral superior oblique and contralateral inferior rectus muscles,⁹ which causes counterclockwise eye rotation with stimulation of the left posterior crista and clockwise rotation with right-sided stimulation. The compensatory fast component of the induced nystagmus is in the opposite direction, however, corresponding with the clinical findings of typical BPPV.

These free-floating posterior canal particles have been identified in vivo in patients undergoing surgery for BPPV.^18,19 This theory of free-floating particles, also referred to as canalithiasis, is an important concept because it relates to the treatment of this condition.

PREOPERATIVE PATIENT COUNSELING AND CONSERVATIVE MANAGEMENT

First, the patient must be reassured that BPPV is an inner ear disorder that is relatively benign and most often self-limited. Medical management for BPPV is mostly ineffective.²⁰ The most efficacious means of vertigo control is avoidance of the specific provocative head movements that induce the attacks. Most patients already use this approach by not lying on the affected side and by not extending the neck to look upward. Patients who stringently avoid these movements may have more prolonged courses because the absence of provocative movements prevents movement within and subsequent dispersement of the particles from the canal.

Most cases of BPPV resolve spontaneously over weeks to months without any treatment. Brandt and Daroff²¹ recommended a rigorous course of physiotherapy under heavy sedation during several days of hospitalization. They believed that the exercises shook free the otolithic debris from the cupula. Other clinicians could not reproduce their good results. Semont and associates²² reported excellent results using a technique termed the liberatory maneuver. They theorized that this technique liberated deposits from the cupula and reported a 92% success rate following two maneuvers.

The liberatory maneuver is difficult to perform in elderly, frail patients. The particle repositioning maneuver^23–25 provides the same benefits as the liberatory maneuver, but is simpler to accomplish. It is based on the free-floating particle (canalithiasis) pathophysiologic theory of BPPV and is adapted from Epley’s canalith repositioning procedure.²⁶ For the purpose of this discussion, it is important to remember that the cupula forms a complete barrier across the ampullated end of the canal that is impermeable to endolymph and free-floating particles. Free-floating posterior canal endolymph particles can enter and exit the canal only through the common crus.

The current particle repositioning technique (Fig. 38-1) begins with the patient seated lengthwise on the examining table. The first part of the Dix-Hallpike maneuver is performed by rotation of the posterior SCC of the undermost (affected) ear in the earth’s vertical axis (see Fig. 38-1B). The examiner should observe the classic nystagmus response, which confirms the diagnosis, and then reassure the patient as the vertigo subsides. The patient maintains this position for 30 to 60 seconds after resolution of the nystagmus, allowing the particles to settle in their new dependent position closer to the common crus. In the second stage, the patient rolls laterally through position C into position D onto the opposite side with the head turned 45 degrees downward. This stage is performed in a smooth, continuous motion, and the neck is kept extended throughout. This method rotates the affected posterior canal 180 degrees in the plane of gravity, allowing the free-floating particles to follow the natural curve of the canal and continue their relative course through the common crus into the utricle, presumably from whence they came.

FIGURE 38-1 Particle-repositioning maneuver, four positions, right ear. Schematic representation of patient and concurrent movement of labyrinth, specifically posterior and superior semicircular canals. In each position, dark oval represents the new position of the particle conglomerate in the most dependent part of posterior canal, and open oval represents the previous position. A, Patient seated. B, Patient in Dix-Hallpike head position. Particles gravitate in ampullofugal direction, causing counterclockwise rotatory nystagmus (right ear). Position is maintained for 2 to 3 minutes. C, Midposition. D, Final position of second stage of maneuver performed in one steady continuous motion. Particles continue gravitating in ampullofugal direction through common crus into utricle. Eyes are observed for nystagmus response. Position is maintained for 1 to 2 minutes, then patient sits up. D, direction of view of labyrinth.

While the examiner supports the patient’s head in position D, a secondary nystagmus response is usually noted, again following a short latent period. A nystagmus response that replicates the initial nystagmus during the Dix-Hallpike maneuver can result only from further passage of the particles in the same ampullofugal direction. Such passage would lead the particles through the common crus into the utricle, where they would no longer induce a pathologic response.

Conversely, a secondary nystagmus that reverses direction from that of the initial head-hanging position (position B) may occur through two possible mechanisms. In one, the particles reverse their direction of movement because of an improperly performed maneuver, resulting in a utriculopetal endolymph current. This reversal usually occurs when the neck is not hyperextended enough during the roll. Cupulolithiasis is the other possible mechanism underlying reversal nystagmus. The gravitational effect on a fixed cupular deposit results in utriculofugal cupular deflection during the Dix-Hallpike maneuver (see Fig. 38-1B), as would be seen with free-floating particles. The position assumed during the second stage of the particle-repositioning maneuver effectively rotates the posterior canal 180 degrees in the earth’s vertical plane (see Fig. 38-1D). This action serves to flip the whole posterior canal and its cupula upside down, reversing the gravitational pull on the cupula and resulting in utriculopetal cupular displacement and a reversal of the nystagmus response.

After another 30 to 60 seconds, the patient is brought back up to the sitting position, and the eyes are observed for nystagmus. With a successful maneuver, there should be no nystagmus when the patient returns to the upright position because the particles would have been removed from the posterior canal. This result is in contrast with that of the conventional Dix-Hallpike maneuver, in which one notes a reversal of the nystagmus when the patient sits back up. No specific postmaneuver instructions to limit head and neck movements are necessary.¹⁵ Patients should be reassessed 1 to 4 weeks after the maneuver, and if the Dix-Hallpike test is still positive, the repositioning maneuver should be repeated. When appropriately administered, repositioning maneuvers should alleviate BPPV in greater than 95% of cases.²⁷

Several other clinical findings help support the free-floating particle theory as the mechanism underlying most cases of BPPV. In some patients with classic histories of BPPV, an initial Dix-Hallpike maneuver often fails to induce a positive response. In many of these patients, a vigorous head shake or application of a skull oscillator may often elicit a latent response. In these instances, it is postulated that the vibration overcomes the particle conglomerate’s inertia or its minor adherence to the membranous canal wall, and allows for its mobilization.

Free-floating particles may also explain the fatigability of a conventional Dix-Hallpike maneuver. Each maneuver likely causes increased endolymph dispersion of the particle conglomerate, resulting in a smaller mass effect with each subsequent maneuver and reducing the degree of hydrodynamic drag and endolymph current. In several patients, a repeat Dix-Hallpike maneuver 30 to 60 minutes after a fatigued response often elicits the same maximal response seen during the initial Dix-Hallpike maneuver. Theoretically, the particles have had time to reassemble into the large conglomerate mass within the endolymph of the posterior canal.

SURGICAL PATIENT SELECTION

Operative intervention is offered for intractable cases in which symptoms are severe enough to affect the patient’s occupation or lifestyle significantly, and when the disorder fails to respond to repeated repositioning maneuvers. Rarely, surgery is offered to patients with severe and frequent recurrences. Until the advent of posterior canal occlusion, transection of the posterior ampullary nerve (singular neurectomy) was the gold standard of operative treatment.^28–30 Singular neurectomy is technically difficult, is performed by very few surgeons, and yields variable rates of failure and sensorineural hearing loss.^31,32 Ohmichi and colleagues³³ showed that the singular nerve would be inaccessible through a tympanotomy approach in 14% of human temporal bones. In a 2007 literature review and temporal bone dissection study comparing posterior canal occlusion with transection of the posterior ampullary nerve, Leveque and associates³⁴ found that posterior canal occlusion presented far fewer technical issues and a far lower risk of hearing loss compared with singular neurectomy.

Posterior SCC occlusion evolved to circumvent the technical challenges and high risk of hearing loss associated with transection of the posterior ampullary nerve. Money and Scott³⁵ initially used this technique in feline vestibular physiology experiments. Plugging individual SCCs blocked their receptivity to angular acceleration without influencing the responses of the other ipsilateral vestibular receptors. Although posterior canal occlusion was at first a theoretical remedy for BPPV, the main concern in applying this technique to humans was its possible detrimental effect on hearing. This problem was not addressed in the original cat studies. Parnes and McClure³⁶ carried out a study in guinea pigs to measure the effect of canal occlusion on hearing using brainstem auditory evoked responses. The hearing responses remained unchanged during follow-up periods lasting 6 months.

The hypothesis that canal occlusion would abolish BPPV was tested when two patients presented with intractable BPPV in ears with coexisting profound sensorineural hearing losses. With no hearing to lose, both patients agreed to undergo what at that time was an experimental procedure. Both patients were relieved of their BPPV and remained symptom-free for at least 4 years, when they were lost to follow-up.^37,38 In addition, both patients maintained postoperative lateral SCC function as measured by caloric responses. This important finding indicated that the procedure’s success resulted from the isolated defunctioning of the posterior canal and not from a generalized destructive process of the vestibular labyrinth.

The theoretical intent of the procedure is to compress the membranous labyrinth closed against the opposite bony wall, creating a closed, fluid-filled (endolymph) space between the plug and cupula, both of which are impermeable to endolymph. Because fluid cannot expand or compress without a change in temperature, this action eliminates all endolymph movement within the posterior canal, effectively fixing the cupula. The canal’s end organ receptor no longer responds to free-floating endolymph particles even if they remain in the partitioned posterior canal (canalithiasis); it likewise does not respond to a fixed cupular deposit (cupulolithiasis), permanently eliminating the BPPV. In addition, the canal no longer responds to physiologic angular acceleration. Because the deficit is constant and permanent, however, gradual compensation occurs through central adaptation, while the corresponding contralateral anterior SCC assumes control of response detection in that particular plane.

Singular neurectomy eliminates the resting discharge from the posterior SCC crista, creating a static vestibular asymmetry³⁹ between the two posterior canals. This effect results in immediate postoperative vertigo at rest and spontaneous rotatory nystagmus. Posterior canal occlusion does not disturb the resting neuronal discharge from the occluded canal, however; most patients do not have spontaneous postoperative vertigo or nystagmus at rest, unless complicated by other factors. Singular neurectomy and posterior canal occlusion result in a dynamic vestibular asymmetry,³⁹ itself resulting in motion sensitivity. The dynamic asymmetry gradually resolves with central adaptation, which can be hastened with vestibular physiotherapy.

PREOPERATIVE EVALUATION

Preoperative evaluation includes a routine audiogram. The procedure is not recommended in an only or significantly better hearing ear. Preoperative electronystagmography or videonystagmography helps to ensure there is a normal vestibular response from the contralateral ear. A high-resolution computed tomography (CT) scan of the temporal bone defines the anatomy and ensures that the posterior canal is accessible through a transmastoid approach. Preoperative imaging with either CT or magnetic resonance imaging (MRI) rules out central lesions that may mimic BPPV, a very rare occurrence.⁴⁰

SURGICAL TECHNIQUE

The use of perioperative broad-spectrum antibiotic coverage is debatable, but antibiotics should be used in ears with a past history of otitis media. The procedure is contraindicated during acute or subacute episodes of otitis media. The procedure is also relatively contraindicated in much better hearing or vestibular functioning ears. Under general anesthesia, the patient is placed in the supine position with the head turned 45 degrees toward the opposite side. The surgical site preparation and draping are performed in a routine fashion. Intraoperative facial nerve monitoring is standard as in all mastoid surgery, but auditory nerve monitoring is not required.

A limited mastoidectomy with a fluted ball drill and suction-irrigation is performed through a standard postauricular incision (Fig. 38-2A). The antrum is opened, providing exposure of the lateral SCC. In most instances, identification of the tegmen and digastric ridge is unnecessary. The bony sigmoid sinus prominence is identified, but in most cases it is unnecessary to bare the sigmoid sinus dura. Bone removal proceeds anteriorly from the sigmoid sinus along the cerebellar plate toward the posterior canal. When the posterior canal otic capsule is identified, the bone is blue-lined (thinned down so much that the transmitted light through the canal’s perilymph creates a dark appearance) with progressively smaller diamond burrs and copious suction irrigation. The target zone for the occlusion is the area at, or just inferior to, a line extending posteriorly from the lateral SCC (Donaldson’s line). Sometimes there are abundant air cells between the posterior canal otic capsule bone and the dura, and sometimes the bony canal directly contacts the dura. The preoperative CT scan helps with navigation through this area.

FIGURE 38-2 A, Exposing right posterior semicircular canal otic capsule. B, Creating the 1 × 3 mm endosteal island with small diamond burr. C, Cross-sectional view.

FIGURE 38-3. A, Lifting out endosteal island with a fine 90 degree pick. B, Magnified lateral view.

FIGURE 38-4. Creating plug with two-component fibrinogen glue and mastoid cortex bone chips.

FIGURE 38-5. A, Tamping plug through fenestra into canal. B, Cross section of canal shows intact but occluded membranous canal.

FIGURE 38-6. Covering fenestra and surrounding bone with fascia and glue.

After blue-lining the canal, a 1 mm diamond burr is used to skeletonize a 3 mm segment of the canal 180 degrees around the outer circumference down to endosteum, creating a 1 × 3 mm endosteal island (Fig. 38-2B). Bone removal should proceed evenly along the circumference so that when the endosteum is violated and perilymph is exposed, all drilling can cease. The endosteal island is removed with a fine 90 degree pick to expose the perilymph (Fig. 38-3). Great care must be taken not to suction directly on the perilymph and especially the membranous labyrinth. At this stage, the exact outline and limits of the membranous labyrinth are usually not clearly discernible. Although not essential, perilymph may be gently “wicked” away with a Cottonoid, or gently suctioned with a No. 24 suction with the thumb off the hole to expose the membranous labyrinth, at which time the membranous duct is seen to collapse.

Various materials have been used to occlude the canals, including bone dust alone, bone wax, fascia or other tissue, and bone pâté made from either bone dust and blood or bone dust and fibrin glue. I use dry bone chips gathered from the mastoidectomy mixed with a couple of drops of two-component, fast-acting human fibrinogen glue. Once set (about 30 seconds), it forms an easily workable, but malleable plug with a firm consistency (Fig. 38-4). The plug is gently and firmly inserted through the fenestra with the intention of completely filling the canal lumen and compressing the membranous labyrinth closed (Fig. 38-5). A blunt 45 degree probe works well to pack the plug in the canal tightly. The membranous labyrinth is resistant to shearing and tearing. As has been learned from partial labyrinthectomies in skull base surgery (see next section), controlled resection of the membranous duct is not risky or dangerous to the other inner ear receptors (i.e., hearing), unless the membranous labyrinth is roughly avulsed. The bone chips within the plug lead to intracanal ossification resulting in complete permanent occlusion of the canal.

After completing the plug insertion, the fenestra and surrounding bone are covered with a piece of temporalis fascia, which is kept in place by several more drops of fibrinogen glue (Fig. 38-6). A good tissue seal is necessary to prevent a postoperative perilymph fistula.

The incision is closed in two layers; a standard mastoid dressing may be applied. A drain is unnecessary. Patients are admitted to the hospital typically for 1 to 3 days, mostly because of the postoperative dysequilibrium and motion sensitivity, the recovery of which can be hastened with vestibular physiotherapy.

In variations of this technique, Anthony⁴¹ successfully treated BPPV by applying an HGM argon laser to the blue-lined posterior canal. The laser burns were purported to create fibrous bands within the canal, leading to obstruction of the membranous duct. Kartush and Sargent⁴² used a CO₂ laser–assisted occlusion technique.

RESULTS

To date, 97 posterior canal occlusions have been reported in the literature. Many more have been done and gone unreported, such as in one busy Australian clinic that has successfully done more than 70 cases (Pohl DV, personal communication). To date 2008, I have performed 55 posterior canal occlusions in 52 patients; 3 patients underwent sequential bilateral occlusions. The average age at surgery was 61 years (range 25 to 83 years). There were 43 women and 12 men, and 34 left ears and 21 right ears. The duration of symptoms ranged from 1 to 35 years, with a mean of 6.8 years and median of 3 years. The follow-up times range from 1 to 20 years.

In approximately 30% of ears, after the canal has been fenestrated, and the membranous labyrinth has been exposed, particles can be seen freely floating within the endolymph.¹⁹ In one previously reported case, the membranous labyrinth was resected with the particles and was subjected to scanning electron microscopy. In this particular case, the particles proved to be degenerating otoconia.¹⁹

All 55 ears remain completely free of BPPV. Four ears had profound preoperative deafness. Of the 51 ears with normal preoperative hearing, most had an initial postoperative mixed hearing loss that in all cases recovered to the preoperative level. One patient had a delayed sudden hearing loss accompanied by a 7 day vertigo spell during the 4th postoperative month. The symptoms were preceded by an intense 2 day headache. This patient did not return for follow-up until 1 month later, at which time her audiogram showed a 70 dB sensorineural hearing loss with only 32% word recognition. This hearing loss persisted on subsequent audiograms. The working diagnosis was labyrinthitis, but the exact causal relationship to the canal occlusion was unknown. The patient had undergone two previous unsuccessful attempts at singular neurectomy in the same ear. One other patient had a 20 dB decrease in bone conduction levels at her 1 year follow-up audiogram and no change in word recognition score.

All 55 cases were followed by an initial 1 to 4 week period of imbalance, giddiness, and motion sensitivity. Six cases had more protracted courses, and five were thought to have had labyrinthitis. Four of these were thought to have had aseptic labyrinthitis; the other was thought to have postoperative otitis media/mastoiditis. This patient had a history of chronic otitis media and had undergone a prior cortical mastoidectomy; her sensorineural hearing loss recovered after antibiotic therapy. The sixth patient developed BPPV in the contralateral ear, but there also seemed to be some functional overlay.

Four other cases warrant brief mention. As noted previously, three patients underwent sequential bilateral posterior canal occlusions, ranging from 1 to 8 years between procedures All three patients remain symptom-free with no other adverse effects from the surgery. As expected, the intact contralateral superior SCC provided the complementary vestibular input of the posterior canals that were rendered nonfunctional by surgical occlusion. The other notable patient had her canal occluded under local anesthesia because of multiple other medical problems. This patient had no intraoperative complaints of vertigo or dizziness.

The postoperative hospitalization was 1 to 5 days in uncomplicated cases. As expected, older patients had longer recovery periods and tended to require longer hospital stays. The degree of postoperative motion sensitivity usually determined the length of stay. Although postoperative vestibular suppressants may provide early short-term relief, their use was discouraged because they tend to prolong the overall recovery. All patients were provided with early postoperative vestibular physiotherapy.

FURTHER APPLICATIONS OF CANAL OCCLUSION SURGERY

Several modifications of posterior canal occlusion have ensued for various other inner ear and skull base conditions. There are only two reported cases of canal occlusion for intractable BPPV of the other canals: one of a single lateral canal occlusion using an argon laser on a blue-lined canal,⁴³ and another case report of an anterior canal plugging.⁴⁴ McElveen and colleagues⁴⁵ described three patients who each underwent a successful hearing-preserving modified translabyrinthine acoustic neuroma resection in which all three SCCs were occluded. Molony and associates⁴⁶ reported two patients who underwent extended middle fossa approaches and one who underwent an extended suboccipital approach for acoustic tumor removal. Hearing was preserved in all three patients after occluding and resecting the superior SCC and posterior SCC.

Hirsch and colleagues⁴⁷ first reported on the partial labyrinthectomy, in which the superior and posterior canals of the same labyrinth were occluded and resected for access to various skull base lesions, with consistent hearing preservation. Minor and coworkers⁴⁸ published their seminal work on the superior SCC dehiscence syndrome in 1998, in which several patients underwent superior SCC occlusion through the middle fossa approach. Several subsequent reports have ensued from their center and other centers showing good hearing preservation rates. Further positive reports on partial labyrinthectomy by Sekhar and associates⁴⁹ and Kaylie and colleagues⁵⁰ substantiated the utility and success of labyrinthine partition with multiple canal occlusions, with predictably good hearing preservation rates.

Dehiscence of the posterior canal can also occur (rarely), and it too can be treated successfully with canal occlusion.⁵¹ Agrawal and Parnes⁵² reported their successful results using the transmastoid approach for superior SCC occlusion in three cases of superior SCC dehiscence. In one patient, hearing remained unchanged, whereas in the other two patients hearing improved with the resolution of the conductive component of the hearing loss.

SUMMARY

Most typical BPPV cases seem to result from free-floating posterior SCC endolymph particles. Most of these patients may be cured with the particle-repositioning maneuver. Intractable cases are rare. For this small subset of patients, posterior SCC occlusion is a safe, curative procedure. Occlusion of all three ipsilateral SCCs alone or in combination can be safely performed for various other disorders, with predictably good hearing results.