ENDOLYMPHATIC ANATOMY AND EMBRYOLOGY

Scarpa (1752-1832) discovered and described endolymph and the membranous labyrinth in 1789,³ 72 years before Méniere⁴ ascribed symptoms of the clinical syndrome to the inner ear. Understanding the embryology of the endolymphatic system may provide some clues to understanding Meniere’s disease. During the fourth week of embryogenesis, three buds of the primordial otocyst appear representing the pars superior, pars inferior, and endolymphatic duct, which eventually develop into the utricle and the semicircular canals, the saccule and the cochlear duct, and the endolymphatic duct and sac.⁵ The endolymphatic duct leaves the medial vestibule and courses dorsally through the bony vestibular aqueduct to terminate on the posterior surface of the temporal bone, enveloped within dural folds of the posterior fossa. The short, straight endolymphatic duct acquires its mature hook shape configuration by the fourth year of life.^6,⁷ Histologic and functional maturation of sac elements may predate complete anatomic maturity. Potassium-rich endolymph fills the endolymphatic sac and duct, the saccule and utricle, the membranous semicircular canals, and the cochlear duct or scala media. These structures are interconnected by the smaller utricular duct, saccular duct, and ductus reuniens. The membranous endolymphatic structures are surrounded by the sodium-rich perilymph that fills the periotic spaces within the bony labyrinth.⁸

The position of the endolymphatic sac along the posterior fossa dura is relatively constant, but its size and the amount of bony covering by the operculum are variable.⁹ In most patients, 50% of the sac lies outside of the temporal bone, and 50% of the sac is intraosseous. Approximately 10% of sacs are completely extraosseous along the posterior fossa dura.¹⁰ The sac can extend posterolaterally to cover the lateral sinus.¹¹ The morphology of the sac is also variable. The distal sac has a smooth open lumen within the dura mater. Its cuboidal epithelium contains light and dark cells.¹² The lining of the intermediate portion of the sac shows more complex epithelial folds forming papillae and crypts of tall columnar light and dark cells. Cells of the intraosseous proximal rugose sac are intermediate between the taller, more distal cells and the squamous-to-cuboidal cells of the duct. The duct narrows at its isthmus to 0.1 to 0.2 mm in diameter. Luminal folding and transversely oriented tubules make the endolymphatic sac a more complex structure than it otherwise outwardly appears.¹³

The normal bony vestibular aqueduct is readily apparent on high-resolution computed tomography (CT) scanning of the temporal bone. It is funnel-shaped or tubular with the width of its external aperture averaging 6 mm.^14,¹⁵ Radiographic observation of the affected ear in patients with Meniere’s disease showed a filiform narrowing of the external aperture averaging 2.2 mm.¹⁴ The amount of narrowing of the external aperture was also shown to be correlated with an increasing percentage of positive electrophysiologic measures in the affected ears of patients with Meniere’s disease.¹⁵ Statistically significant differences in the percentage of patients with enlarged summating potential-to-action potential (SP:AP) ratios by transtympanic electrocochleography were seen when correlated with the size of the external aperture of the vestibular aqueduct. An increased SP:AP ratio was noticed in 95% of ears with nonvisible external apertures, 91% when the aperture was less than 5 mm, 58% when the aperture was 5 to 7 mm, and 29% when the aperture was greater than 7 mm.

The endolymphatic duct and sac can be seen on high-resolution fast spin echo magnetic resonance imaging (MRI). The endolymphatic sac and duct were seen on MRI of 20 temporal bones in healthy subjects using strongly T2-weighted sequences and postprocessing software.¹⁶ A retrospective review of 42 ears with MRI data that underwent endolymphatic sac surgery correlated surgical findings with the ability to image the endolymphatic sac and duct.¹⁷ Surgical findings were classified as normoplastic in 17, atrophic in 14, and invisible in 11. Proton density imaging and T2 sequencing positively identified the endolymphatic duct and sac in 14 patients. The endolymphatic sac and duct were shown by proton density imaging alone in 14; neither proton density imaging nor T2 sequencing showed images in the remaining 14 ears. Findings at surgery showed statistically significant correlation with the ability to identify structures on imaging. Normoplastic surgical anatomy was identified on both imaging modalities; however, atrophic sacs were rarely seen on T2 imaging.

In another study using submillimeter MRI, the endolymphatic ducts were visualized in 29% of patients with Meniere’s disease and in 91% of healthy individuals. Temporal bone measurements between the posterior semicircular canal and the subarachnoid space, and the vestibule and the subarachnoid space were shorter in Meniere’s disease patients than in healthy individuals. It was noted anecdotally that endolymphatic shunt surgery was more effective in the few patients with visualized ducts compared with patients with nonvisualized ducts.¹⁸ Currently, the practicality of imaging in Meniere’s disease or for endolymphatic surgery lies primarily in ruling out retrocochlear pathology.

From a surgical perspective, the endolymphatic sac is generally approached through the mastoid bone, and is isolated on the lateral surface of the posterior fossa dura. When performing a suboccipital craniotomy, the sac can often be identified on the posterior petrous aspect of the temporal bone. Discussing the location of the endolymphatic sac, Gibson ¹⁹ noted that the extraosseous portion is difficult to define surgically, appearing only as a thickened area of dura. He also noted that after splitting the layers of the endolymphatic sac no endolymph is seen, and usually no significant electrophysiologic changes occur. Huang²⁰ attributed higher success in endolymphatic sac surgery to definite identification of the sac, with entry into the true sac lumen, and preservation of the sac anatomy. Amiratti and colleagues²¹ advocated preserving the integrity of the endolymphatic system, suggesting severe audiovestibular disturbances that may follow sac disturbance; however, our experience is that patients tolerate complete excision of the sac without hearing loss, whether incidental in other cranial base procedures, or intentional for complete endolymphatic sac ablation.

Important topographic landmarks for identifying the endolymphatic sac exist from the transmastoid or posterior fossa approach. The transmastoid extradural landmarks for localization of the endolymphatic sac and for preservation of labyrinthine structures include Donaldson’s line, which is an imaginary line drawn posteriorly through the plane of the horizontal semicircular canal (see Fig. 34-1), and measurements delineating the hard angle. The endolymphatic sac is generally found along the posterior fossa dura inferior to Donaldson’s line. Caution must be used when identifying the sac to avoid damage to the facial nerve and the posterior semicircular canal. Anatomic variants of normal temporal bone anatomy have been associated with Meniere’s disease. An understanding of potentially altered anatomy is important for surgical planning of sac procedures.²² Hypoplasia of the mastoid air-cell system, hypocellularity of periaqueductal cells around the endolymphatic duct and sac, reduction of the aditus ad antrum, and hypoplasia of the facial recess all have been described.

FIGURE 34-1 Completed mastoidectomy showing the head of the incus (1), second genu of facial nerve (2), horizontal semicircular canal and plane of Donaldson’s line and triangle outlining the hard angle (3), position of endolymphatic sac (4), sigmoid sinus (5), retrofacial air cell tract (6), posterior semicircular canal (7), and posterior fossa dural plate (8).

Intradural identification of the endolymphatic sac in relation to anatomic structures of the posterior fossa places the sac 10 to 15 mm lateral to the internal auditory meatus, and 11 to 17 mm posterosuperiorly to the eleventh cranial nerve in the jugular foramen.²¹ Typically, the thickening of the dura and the bony ledge of the operculum pinpoint the location of the sac.

ENDOLYMPHATIC SAC PHYSIOLOGY

Surgical shunting of the endolymphatic sac was proposed soon after the initial anatomic observations of hydrops in the endolymphatic compartments to alleviate inferred dysfunction.^1,²³ The presumed longitudinal flow of endolymph from the stria vascularis to the endolymphatic sac and the role of the sac as a primary resorptive organ have long been assumed.²⁴ Radial flow of endolymph has also been shown, however.²⁵ Endolymph is produced by dark cells located largely in the stria vascularis, but also in vestibular ampullae, within the maculae of the saccule and utricle, and along the endolymphatic duct. Maintenance of a potassium-rich endolymph produces the endocochlear potential, a DC voltage gradient, to drive the transduction process important in the detection of sound, motion, and position.^26,²⁷ This is a pH-sensitive process and is based on an active transport system in the vestibular dark cells by a sodium–hydrogen ion exchange system.^27,²⁸ Local production concentrated within strial dark cells and radial movement of endolymph with local chemical exchange throughout its course maintain a chemical balance and gradient that promote physiologic endolymphatic function, and may promote a slow linear flow toward the sac.²⁵

Other theories of endolymphatic fluid homeostasis exist. Salt ²⁹ stated that direct measurements of the dispersal of markers in endolymph fail to support dynamic flow theories, and suggested that, in the normal state, there is negligible flow. The ionic component of endolymph is maintained through single cell transport of ions. This local control theory is overridden when endolymph volume is abnormally high or low, with the endolymphatic sac acting as a regulator of bidirectional flow in response to volume needs. The endolymphatic sac functions as the master volume regulator by numerous observed characteristics. In contrast to endolymph throughout the inner ear, a gradient exists along the duct leaving the electrolyte state in the endolymphatic sac high in sodium and low in potassium.³⁰ An active equilibration of ions creates an osmotic potential that may influence the transepithelial flow of fluids. Higher concentrations of Na⁺,K⁺-ATPase are seen in the endolymphatic sac, but diminish proximally along the endolymphatic duct.³¹

Several other findings suggest an active role of the endolymphatic sac on endolymph fluid homeostasis.³²^–³⁴ Aquaporin 2, vasopressin type 2 receptor, and transient receptor potential channel vanilloid (TRPCV), subfamily type 1 and 4, were found in the epithelial lining of the endolymphatic sac, but not in other extracellular tissues, although TRPCV 1 was seen in the surrounding vasculature. Similar findings are seen within the kidneys, suggesting a parallel role in fluid filtration and resorption.³² A unique protein, saccin, secreted by the endolymphatic sac, acts within the kidney as an endogenous inhibitor of sodium reabsorption. Intracellular morphology of the endolymphatic sac chief cells possesses organelles capable of endocrine function, and cellular ultrastructure consistent with merocrine activity also is observed.^33,³⁴

Biochemical and cellular findings within the endolymphatic sac support its functional role in phagocytic, immune, and allergic responses of the inner ear.³⁵^–³⁹ Volume may also be regulated by the functional mechanical entity of an extracellular matrix of interstitial cells that support the endolymphatic duct, and take part in the control of inner ear fluid dynamics and endolymph resorption.⁴⁰ As new techniques are developed to detect and monitor minute fluid volumes and changes, and to evaluate the chemical composition and cellular characteristics of the endolymphatic sac, a better understanding of the physiologic role and pathophysiologic state of the endolymphatic system in Meniere’s disease will be achieved.^26,⁴¹

ENDOLYMPHATIC HYDROPS: PATHOPHYSIOLOGY

It was not until the histopathologic observation of Hallpike and Cairns in 1938 ⁴² that a proposed malfunction of the endolymphatic system was correlated with the clinical syndrome. These authors showed dilated endolymphatic spaces in temporal bone specimens from two patients with the clinical symptoms of Meniere’s disease who died after neurotomy of CN VIII. These findings showed end organ changes of the inner ear in patients with hearing and balance symptoms. The hydropic state of the inner ear has been confirmed in other temporal bone studies and described as the primary pathologic correlate of Meniere’s disease.⁴³^–⁴⁵

The underlying cause of the hydropic state is at present unknown, although many theories exist. Hydrops is seen more often in the cochlea and saccule, structures derived from the later developing pars inferior.⁴³ Congenital insults or developmental aberrations later in the course of embryogenesis could presumptively account for this difference. A familial connection is seen in the history of 20% of patients clinically diagnosed with Meniere’s disease.⁴⁴ This connection, in consideration of the embryologic and anatomic findings associated with Meniere’s disease, suggests a multifactorial predisposition to developing endolymphatic hydrops. Whether this precondition is genetic or related to shared environmental factors or insults is yet to be determined.

The pathophysiologic state of the endolymphatic system has been partially modeled in mice by experimental destruction and obstruction of the sac or duct in attempts to pinpoint the underlying mechanism of the hydropic condition.^46,⁴⁷ Although these experiments were able to reproduce hydrops and audiometric findings similar to Meniere’s disease, vestibular dysfunction or vertigo was noticed only after placing the animals in a head-down position theoretically by inducing additional pressure within inner ear fluids.⁴⁸

The predominant theories that could explain the symptom complex of Meniere’s disease are based on the observed pathologic and induced experimental evidence of hydrops. The various temporal bone findings are summarized succinctly by Costa and associates,⁴⁹ and include ruptures of the membranous labyrinth, fistulas of the membranous labyrinth, collapse of the membranous labyrinth, obstruction of longitudinal flow, vestibular fibrosis, sensory lesions, and neural lesions. Schuknecht⁵⁰ proposed the rupture theory. He reasoned that distention of the endolymphatic space with eventual membrane rupture could cross-contaminate perilymphatic spaces and toxify delicate sensory hair cells with the potassium-rich endolymph. In the distention theory, Paparella⁵¹ described decompensation of radial flow as the perilymphatic spaces ebb, leading to largely longitudinal flow along the hydropic membranous pathway with the saccule acting as a reservoir for the excess endolymph. As the dilated saccule encroaches on the confines of the vestibule, mechanical interference of cochlear and vestibular function occurs by inhibition of traveling waves and physical contact with the crista ampullaris.

The drainage theory, as presented by Gibson and Arenberg,¹⁰ suggests that obstruction of a narrowed endolymphatic duct divests the sac of endolymph. The endolymphatic sac responds by secreting glycoproteins and saccin. Glycoproteins act osmotically by “pulling” endolymph toward the sac, and saccin stimulates secretion of endolymph from dark cells that distend the endolymphatic spaces “pushing” against the obstruction toward the sac. The obstructing debris ultimately and suddenly passes, and the resultant rapid flow of endolymph purportedly brings on an acute vertiginous episode. Gibson and Arenberg suggested that patients with Meniere’s disease with larger vestibular aqueducts could experience resolution of auditory symptoms as in Lermoyez’s syndrome after clearance of the obstructing debris. Patients with Meniere’s disease have been shown to have widening of the vestibular aqueduct aperture, but, although enticing to establish a mechanism of pathology, patients with Lermoyez’s syndrome have not been shown to have wider vestibular aqueducts than other patients with Meniere’s disease. Gibson and Arenberg also theorized that Tumarkin crises could be the effect of a membrane rupture in the overdistended endolymphatic space.

Although Meniere’s disease has no known etiologic cause by definition, there are several etiologies of secondary Meniere’s syndromes. Meniere’s syndrome can be mimicked by disease processes in all categories—traumatic (inner ear fracture, perilymphatic fistula), infectious (viral, syphilis), inflammatory (autoimmune, allergy, Cogan’s syndrome, sarcoidosis), metabolic (diabetes), congenital/genetic (inner ear malformations, enlarged vestibular aqueduct), neoplastic (vestibular schwannoma), vascular (migraine, hemorrhage), and iatrogenic (stapes/mastoid surgery). These disease processes may lead to secondary endolymphatic hydrops by causing scarring within the labyrinth; activating inflammatory, cytokine, and complement pathways, yielding edema and fibrosis and altered extracellular matrices; altering hemodynamics that may affect transcellular ionic exchange and alter endolymph homeostasis; creating obstructive immune complexes or cellular debris; and activating cellular responses through humoral messengers. The end result is distention or obstruction of the endolymphatic flow disturbing sensation of cochleovestibular signals and sending distorted messages of sound, station, and motion.

MENIERE’S DISEASE

Epidemiology

Arenberg and colleagues ⁵² have suggested that incidence and prevalence estimations reported in Meniere’s disease are inaccurately low by not recognizing the early or atypical cases, or cases misdiagnosed by lengthy remissions drawing out the episodic nature. The definition of Meniere’s disease may alter prevalence and incidence numbers. The 1995 AAOHNS Committee on Hearing and Equilibrium recognized that the 1985 diagnostic criteria for Meniere’s disease were rigid to the point of precluding patients who were most likely Meniere’s cases. Current criteria are listed at the beginning of this chapter.² These gradations may alter the epidemiologic accounting of Meniere’s disease because they allow for inclusion of patients who may be early in the course of their disease, and patients who may have milder symptoms of the disease. Several retrospective reviews have shown the prevalence of Meniere’s disease to be 10 to 20 per 100,000 in various populations around the world.⁴⁹ Using 1995 committee criteria for diagnosis, a more recent study in the Finnish population showed 43 cases per 100,000 population, and an annual incidence of 4.3 per 100,000.⁵³

Costa and associates ⁴⁹ described in detail the epidemiology of Meniere’s disease. It generally manifests in the fifth decade of life. The incidence in childhood is thought to be 1% to 7% of all Meniere’s cases.⁵⁴ Of 14 children diagnosed with definite Meniere’s disease, 5 were shown to have secondary disease manifesting 5 to 11 years after a history of Haemophilus influenzae meningitis, mumps, temporal bone fracture, and congenital and embryopathic complications.⁵⁵ Nine of the 14 children had idiopathic disease. These 14 children represented 1% of the combined Meniere’s population of four neurotologic clinics. There is a slight female preponderance. In women with definite Meniere’s disease who were pregnant, a clear decline in symptoms was associated with delivery.⁴⁹ No socioeconomic, occupational, or racial effect has been consistently shown, although there are statistically increased numbers in married individuals and anxious individuals, and decreased numbers in obese individuals.

Diagnosis

A detailed history is the essence of diagnosis in Meniere’s disease. Classic symptoms were reaffirmed by the AAOHNS Committee on Hearing and Equilibrium in 1995. The symptoms include recurrent spontaneous episodic vertigo, definitive spells including spontaneous rotational vertigo lasting 20 minutes to 24 hours, often prostrating, accompanied by dysequilibrium that may last several days; nausea, commonly with vomiting or retching; but significantly no loss of consciousness. Horizontal or rotary nystagmus is always present during an acute episode. Auditory symptoms include hearing loss (fluctuating or not fluctuating) and aural fullness, tinnitus, or both. Patients who give a clear history of two or more definitive episodes of vertigo as detailed earlier with hearing loss and tinnitus or aural fullness without evidence of inciting illness, injury, or other inner ear pathology meet the diagnosis of definite Meniere’s disease.

An audiometric evaluation is necessary, and may specify the side of disease. No other testing is necessary for diagnosis, although ancillary testing may help determine the side of disease and the extent of pathology, and help to identify primary causes of Meniere’s- like, symptoms such as syndromes, or other disease processes if suspected.

Electronystagmography can establish vestibular dysfunction, and may delineate the side of the affected ear, but may also be normal early in the course of Meniere’s disease. MRI based on asymmetric hearing loss helps rule out retrocochlear pathology. Brainstem auditory evoked potentials in lieu of MRI are useful for patients with contraindications to or intolerance of MRI. Serology to detect otosyphilis or autoimmune disease is helpful as indicated. Allergy testing for evaluation of patients with seasonal or trigger-induced symptoms can be profitable. Empirical treatment or neurologic consultation for suspected vertebrobasilar or atypical migraine disease may be warranted, especially in patients with atypical symptoms. In children, CT scanning to evaluate inner ear anatomy is advisable. Posturography does not have diagnostic value in the diagnosis of Meniere’s disease, but may be useful in documenting rehabilitation progress.

Electrophysiologic tests, including electrocochleography, cochlear microphonics, and vestibular evoked myogenic potentials, have been used as adjuncts in diagnosis, and to monitor the efficacy of treatment in Meniere’s disease.^44,⁵⁶ In 2002, Ge and Shea⁵⁷ reported a 10-year experience using transtympanic electrocochleography. Transtympanic electrocochleography was performed in 2421 ears of 2140 patients with Meniere’s disease. These authors concluded that electrocochleography is a reliable test to detect the presence of endolymphatic hydrops in Meniere’s disease using parameters of an enlarged SP:AP ratio greater than 0.4, a broadened action potential waveform (>3 ms), and a prolonged action potential latency (>0.2 ms). Combined click and tone burst responses yielded an enlarged SP:AP ratio in 81.7%, and a prolonged AP latency was found in 62.2% of ears with Meniere’s disease. An enlarged SP:AP ratio significantly correlated with stage and duration of disease. The SP:AP ratio was found to be elevated in 71% of stage 1 Meniere’s disease, 82% of stage 2, 85% of stage 3, and 90% of stage 4. The SP:AP ratio was elevated in 43% of patients during their first year of diagnosis and in 100% of patients with the disease for more than 30 years.

Cochlear microphonics were also used to assess the presence of hair cell survival, and were found to be present in 69% of ears with pure tone averages greater than 40 dB. It has been suggested that large cochlear microphonics in patients with Meniere’s disease indicate hearing loss resulting from altered cochlear mechanics, whereas severe hearing loss with small cochlear microphonics represents a hearing deficit resulting from hair cell loss.⁵⁸ Electrocochleography has been shown to have a low sensitivity (57%), but is specific for endolymphatic hydrops (94%).⁴⁴ In patients with an elevated SP:AP ratio in the immediate preoperative period, a statistically significant intraoperative reduction of the SP:AP ratio has been reported after endolymphatic sac incision and drainage.⁵⁹

Vestibular evoked myogenic potential is used to assess the vestibulocollic or sacculocollic reflex. This test selectively assesses saccular function and integrity of the inferior vestibular nerve. It is being explored in Meniere’s disease for its potential in identifying endolymphatic or saccular hydrops.⁶⁰ Vestibular evoked myogenic potential testing may aid in identification of active Meniere’s disease,⁶⁰ and may help identify ears more prone to develop contralateral Meniere’s disease.^62,⁶³ Vestibular evoked myogenic potential may also offer information complementary to electronystagmography,^64,⁶⁵ and provide a way to measure severe saccular dysfunction, a finding associated with Tumarkin crisis.⁶⁶ These tests show promising diagnostic potential. Further testing experience and validation are necessary.