Introduction
Ménière's disease, named after Prosper Ménière, is an inner-ear disorder characterised by vertigo attacks with combinations of fluctuating hearing, tinnitus and aural fullness in the affected ear.
In the nineteenth century, common notions of the functions of inner-ear organs were: the cochlea was responsible for mediating the nature and pitch of sound, the otolith organs for the perception of loudness, and the semicircular canals for sound direction. In 1824, Marie Jean Pierre Flourens, professor of physiology in Paris, lesioned pigeon semicircular canals, causing violent head movements in the plane of each canal. Although not his interpretation, he is credited with the first observation that the semicircular canals are involved in balance function.
Prosper Ménière, medical director of the Institute for Deaf Mutes in Paris, was aware of Flourens’ experiments. In a lecture to the Imperial Academy of Medicine, in 1861, he stated: ‘An auditory apparatus, until then perfectly healthy, can become, all of a sudden, the seat of functional troubles, … repeated cerebral accidents such as giddiness, uncertain walking, spinning and falling, … nausea, vomiting, … soon followed by the most profound deafness…. There is every reason to believe that the material lesion which is the cause of these functional troubles … lies in the semicircular canals’.Reference Meniere1 At least in some respects he may have been correct.
After 156 years, the true cause of Ménière's disease remains unproven. It has been likened to Winston Churchill's wartime comment on Russia as its being ‘a riddle, wrapped in a mystery inside an enigma; but perhaps there is a key’. PubMed lists 7820 publications on Ménière's disease since 1883. In the nineteenth century, vertigo as now recognised was classified as ‘apoplexy’, along with stroke and epilepsy, and was believed to be due to impaired blood supply to the brain. The earliest PubMed article on Ménière's disease, by Edward Woakes, extended this notion as a defect of blood supply to the inner ear.Reference Woakes2 Subsequent aetiological theories are too numerous to discuss in detail, but include: a variation in the size or position of the endolymphatic sac and duct, an autoimmune cause, allergy, viral infection, or (assuming a provable family history) a genetically determined abnormality of endolymph control.Reference Chiarella, Petrolo and Cassandro3
Histological theories
The advent of cadaver temporal bone pathology allowed an internal view of the inner ear. In 1938, Hallpike and Cairns noted that the characteristic feature of Ménière's disease is an excess of endolymph called endolymphatic hydrops.Reference Hallpike and Cairns4 At the time, endolymph was presumed to be produced in the cochlea and absorbed in the endolymphatic duct by longitudinal flow.
Schuknecht established the largest and one of the few surviving human temporal bone laboratories, in Boston, USA. Experiments blocking the endolymphatic duct or destroying the endolymphatic sac in guinea pigs, cats and rabbits consistently resulted in endolymphatic hydrops and degenerative changes in the organs of Corti and cochlear neurons, but never vertigo attacks. This led Schuknecht to state that ‘the most plausible explanation is that Ménière's disease is caused by functional failure of the endolymphatic sac’.Reference Schuknecht5 In ears with Ménière's disease, he had frequently observed ruptures of Reissner's membrane and stated that ‘the episodes of vertigo and fluctuating hearing could be accounted for by the toxic effect of potassium on the sensory and neural structures which are normally bathed in endolymph’.Reference Schuknecht5 This did not explain why vertigo attacks do not occur in non-human animals, but the rupture and potassium intoxication theory has been the predominant theory for Ménière's vertigo attacks since the 1960s.
There are many arguments against the rupture theory. The strongest is the question of how a rupture in the cochlea could cause vertigo generated by distant vestibular organs. Another is why hearing losses do not occur at discrete frequency sites rather than the characteristic initial low frequency fluctuation, or that a total loss of hearing does not occur.
Nystagmus during attacks
Observations on nystagmus direction during an attack have been used both to support and disprove the potassium intoxication theory. Dohlman theorised that an initial ipsilateral nystagmus during an attack would be consistent with an increased potassium concentration surrounding afferent nerves.Reference Dohlman6
McClure et al. performed electronystagmography on eight patients at differing times during Ménière's attacks.Reference McClure, Copp and Lycett7 In all cases, the initial observed nystagmus direction was contralateral to the assumed Ménière's ear, often reversing to the ipsilateral side over hours, and in one case over seconds. Others have observed the opposite pattern (initial ipsilateral) at the onset of attacks. McClure and colleagues also perfused the perilymphatic space in guinea pigs, with a high potassium solution. This produced an initial ipsilateral nystagmus followed by reversal in all animals. Nevertheless, the authors rejected that the observations supported the potassium intoxication theory, but instead considered them to support the alternative theory of raised endolymphatic pressure.Reference Brown, McClure and Downar-Zapolski8 Tonndorf had argued that raised endolymphatic pressure could explain the aural symptoms of Ménière's disease.Reference Tonndorf9 McClure and colleagues speculated that such a sudden rise in endolymphatic pressure could result in mechanical distortion of the semicircular canal ampullae, which for the horizontal canal would rotate the crest of the crista and turn the base of the cupula away from the vestibule (utriculofugal), causing a contralateral nystagmus.Reference Brown, McClure and Downar-Zapolski8
Electrophysiological and video head impulse testing
Newer approaches have improved on the limitations of conventional animal and human post-mortem histology studies, and have advanced interpretations of observations of nystagmus during attacks.
Gibson performed transtympanic electrocochleography recordings in ears with Ménière's disease and normal ears using tone burst stimuli.Reference Gibson10 In ears affected by Ménière's disease, a 1 kHz tone burst produces a measurable large negative summating potential, reflecting basilar membrane distortion from endolymphatic hydrops. From hundreds of recordings, 87 patients could be identified as having an attack during the test, at 1–24 hours before the attack and up to 72 hours after the attack. All patients had an abnormally enlarged summating potential, which was greatest 1–24 hours prior to the attack, dramatically decreasing during the attack and then increasing over the next 72 hours. This implies that prior to the attack, the hydrops increases suddenly and then decreases over 72 hours, and that the action potential measuring cochlear nerve function, which might be expected to be reduced or lost because of a rupture, is preserved. McNeil and colleagues allowed Ménière's patients to test their own hearing daily with a programmable hearing aid.Reference McNeil, Cohen and Gibson11 In six patients who measured their own hearing during an attack, five had no change of hearing before, during or after an attack, and one had a probable change in threshold prior to but not during an attack. These observations indicate that a rupture and potassium intoxication are unlikely.
Brown and colleagues injected 3–4 µl of artificial endolymph into the inner ears of guinea pigs, and measured vestibular-evoked short latency potentials, which are assumed to be from utricular neurons.Reference Brown, Chihara, Curthoys, Wang and Bos12, Reference Brown, Chihara and Wang13 Recovery of cochlear function was often followed by a transient increase or decrease in utricular sensitivity (similarly if injected directly into the utricle), suggesting there may be a sudden opening of the urticulo-saccular duct to alleviate hydrostatic pressure, resulting in a change of utricular function due to an increase in its volume. Micro computed tomography (CT) on these ears did not show ruptures.
Vestibular-evoked myogenic potentials have been studied in terms of their utility to diagnose hydrops in the vestibule. In static studies on ears between attacks, there appears to be no single reliable vestibular-evoked myogenic potential measure.Reference Johnson, O'Beirne, Lin, Gourley and Hornibrook14 Manzari and colleagues, in a unique dynamic study on 15 patients during an attack, found a significant increase of the n10 of the contralateral ocular vestibular-evoked myogenic potential and a significant decrease in the ipsilateral cervical vestibular-evoked myogenic potential.Reference Manzari, Tedesco, Burgess and Curthoys15 This implies a decrease in saccular function followed by an increase in utricular function, consistent with a surge of endolymph towards the utricle. Although not often appreciated, isolated otolith stimulation or otolith paralysis can elicit nystagmus and vertigo.Reference Manzari and Curthoys16
The video head impulse test is a new technique to objectively measure semicircular canal function both statically and dynamically. Manzari and colleagues conducted a dynamic video head impulse test study on horizontal canal function in six ears with Ménière's disease. They found that eye velocity (gain) was enhanced at quiescence and during an attack, and was reduced after the attack, likely due to mechanical changes in the horizontal canal ampulla.Reference Manzari, Burgess, MacDougall, Bradshaw and Curthoys17 Yacovino and colleagues reported similar changes in the horizontal vestibulo-ocular reflex in one Ménière's patient during an attack, but they attributed it to perilymph potassium intoxication.Reference Yacovino, Hain and Musazzi18
Gibson–Arenberg theory
Experiments by Salt and DeMott have challenged the normal longitudinal flow of endolymph by injecting chemical markers into guinea pig cochleas without volume disturbance.Reference Salt and DeMott19 In normal volumes, marker diffusion was static (radial) and not longitudinal, suggesting that the latter occurs only when there is abnormal volume. The endolymphatic sac contains hydrophilic proteins.Reference Rask-Andersen and Danckwardt-Lilliestrom20 It also contains natriuretic peptides, which may have a role in controlling cochlear endolymph.Reference Moller, Kirkeby, Vikesa, Nielsen and Caye-Thomasen21
In combination, some of these findings have been the basis of the drainage theory by Gibson and Arenberg.Reference Gibson and Arenberg22, Reference Gibson23 Mild hydrops is cleared by radial flow, but if there is excessive hydrops, longitudinal flow is initiated by the endolymphatic sac. The sac may clear it, but if the endolymphatic duct is dysfunctional or blocked, obstruction to flow results in an endolymph build-up in the sinus of the endolymphatic duct. Excessive endolymph refluxes through the utricular valve of Bast and into the ampullae of the semicircular canals. In a magnetic resonance imaging inner-ear study, Gurkov and colleagues found hydrops only in the horizontal canals of three patients, suggesting it may be the predominant canal affected.Reference Gurkov, Flatz, Ertl-Wagner and Krause24
Overall, there is strong evidence from non-histological studies to indicate that ruptures leading to potassium intoxication of the perilymph are not the cause of Ménière's vertigo attacks. Other evidence supports a hydrostatic process, wherein an obstructed endolymphatic sac attempts to clear the hydrops. The variable nystagmus observed can most likely be explained by endolymph moving in one direction and then in the opposite direction with (possibly alternating) displacement of the horizontal canal receptor and stimulation of the utricle.
Otoconia
In vitro experiments on mouse otoconia in endolymph have indicated that they usually dissolve in 20 hours,Reference Zucca, Valli, Valli, Perin and Mira25 or more slowly if the calcium concentration is raised. It is likely that the same occurs in human ears. It is now well accepted that benign paroxysmal positional vertigo (BPPV) is caused by detached non-dissolved utricular otoconia entering a semicircular canal.Reference Parnes and McClure26 However, it is seldom asked what is the fate of the detached saccular otoconia that do not dissolve? Most human histological studies on otoconia are confounded by their dissolution by the temporal bone decalcification process, but that is not a limitation for studies on embryos.
The endolymphatic sac has histological features suggesting a phagocytic role in removing endolymph debris, including otoconia.Reference Lundquist27–Reference Rask-Andersen, Friberg and Bagger-Sjöbäck29 Otoconia have been found in the endolymphatic sac of fetal guinea pigs.Reference Imoto, Rask-Andersen and Bagger-Sjöbäck30 Yamane and colleagues injected fetal and adult guinea pigs with streptomycin. Electron microscopy of endolymphatic sacs revealed otoconia in the sacs of 30-day old fetuses but none in the adult sacs, suggesting that this is possibly a preparation effect.Reference Yamane, Imoto, Nakaio, Igarashi and Rask-Andersen31 Ohashi and Igarashi studied dislodged otoconia from streptomycin and intense linear acceleration in squirrel monkeys and found phagocytosed otoconia in the endolymphatic sac.Reference Ohashi and Igarashi32
Kao, Parnes and Chole recently undertook a scanning electron microscopic analysis of material intra-operatively removed from the endolymphatic compartment of the posterior canals of two patients with BPPV.Reference Kao, Parnes and Chole33 The findings revealed broken fragments of utricular macula membrane, with attached and detached otoconia. In one patient, there was similar particulate matter in the endolymphatic duct. Overall, there is evidence that detached otoconia can reach the endolymphatic sac, and their most likely origin is the saccule.
A multi-institutional temporal bone study concluded that Ménière's disease always begins in the cochlea.Reference Pender34 The cause of the initial cochlear hydrops is unknown. Kumara, Schuknecht and colleagues surgically blocked the reuniting duct in guinea pigs; the majority of the animals showed cochlear hydrops, saccule collapse and an abnormal utricle.Reference Kumara, Schuknecht, Ota and Jones35 Yamane and colleagues, using micro three-dimensional CT scans, suggested that, in ears affected by Ménière's disease, the reuniting duct may have become obstructed by detached saccular otoconia.Reference Yamame, Takayama, Sunami, Sakamoto and Imoto36.Reference Yamane, Sunami, Iguchi, Sakamoto, Imoto and Rask-Andersen37 In a further study, the vestibular aqueduct and proximal endolymphatic sac were not visible in 60 per cent of ears with Ménière's disease, compared with 32 per cent of the opposite healthy ears and 4 per cent of healthy control ears.Reference Yamane, Konsihi, Sakamaoto, Yamamoto, Matsuhita and Oishi38 These findings suggest that obliteration is a characteristic finding in Ménière's disease.
Combination of two theories
A proposed comprehensive mechanism is the following. It appears that the likely fate of detached otoconia is to dissolve in endolymph. Detached utricular otoconia that do not dissolve cause BPPV. The normal passage of non-dissolved saccular otoconia is to be phagocytosed in the endolymphatic sac. However, if otoconia obstruct the reuniting duct, they may initiate cochlear hydrops. The hydrops moves into the saccule. Otoconia from the damaged saccule move towards the endolymphatic duct and sac, as the sac attempts to clear the hydrops and the otoconia. If the hydrops refluxes through the valve of Bast past the utricle, the damaged utricle releases more otoconia.Reference Calzada, Lopez, Ishiyhama and Ishiyama39 The sac may eventually clear the otoconia and the hydrops as the endolymph moves in the opposite direction. As the sac becomes more dysfunctional and as more attacks occur, the hydrops eventually involves the whole inner ear.
Onset age
New data on the age of onset of Ménière's disease show a similarity to BPPV.Reference Hornibrook and Bird40 This is based on two studies with identical criteria, in which the age of onset was defined as age at the first vertigo attack and clinical certainty was achieved with tone burst electrocochleographic proof of endolymphatic hydrops.Reference Hornibrook41 Both conditions are rare in children and occur occasionally in teenagers. Their incidence increases in patients in their twenties, with most patients aged 30–60 years. Approximately one-third of affected patients are aged over 60 years at disease onset.
Conclusion
The common association of Ménière's disease and BPPV has been recognised. Other authors have postulated an otoconial cause for Ménière's disease, including speculation that successful treatments with a micropressure device might be explained by a massage effect on free otoconia.Reference Phillips and Prinsley42, Reference Jahn43
The remarkable similarity in age of onset for BPPV and Ménière's disease raises the possibility that the two conditions have the same fundamental cause.Reference Hornibrook and Bird40
Further progress on elucidating the mechanism of Ménière's attacks and other inner-ear disorders is likely to await significant advances in the resolution clarity of CT inner-ear imaging in human ears.
Competing interests
None declared
