Statistics from Altmetric.com
Brasić and Perry1 convincingly describe a boy with unilateral otopathic auditory hallucinations. However, their literature review is very misleading, postulating dubious CNS causes for auditory hallucinations other than from a hyperactive ear. They also do not cite a relevant prior case.
Their paper starts: “Unilateral auditory hallucinations...are associated with contralateral CNS lesions”. Their only supporting reference is Toulouse (1892), who reviewed four adults with neurological disease and unilateral auditory hallucinations, allegedly of cortical origin. These turned out to be anything but. One case (Regis, 1881) had voices in his left ear aggravated by alcohol, sounds of a bell and water in his right, impaired hearing in his left ear (thought by Toulouse to be of cortical origin!); the 1888 case had bilateral auditory hallucinations, worse on the left, abolished by blocking off the ear which was full of pus; the 1890 case, an alcoholic wine merchant, had bilateral noises (bells, whistles, rattles, etc) which alternated and changed into voices in his right ear, but there was no examination of his ears or hearing; his 1892 case, also an alcoholic wine merchant, heard voices in her left ear, and had bilateral deafness from chronic otitis media. Whatever other lesions these four patients had, there were peripheral lesions sufficient to trigger tinnitus and auditory hallucinations, although in alcoholic patients these can probably be triggered from ears with relatively normal hearing.2 By contrast with these totally unconvincing cases of non-otological cortical origin, Toulouse reviewed seven cases of unilateral auditory hallucinations from homolateral ear disease, four of which were particularly convincing as the auditory hallucinations vanished (three cases) or waned (one case) with effective ear treatment. Toulouse also noted cases with quite different auditory hallucinations in opposite ears simultaneously; if of cortical origin, a quite implausible degree of functional hemispheric independence is shown.
Their next three references supposedly show that auditory hallucinations are caused by stimulatory phenomena in the CNS—namely, epilepsy (Keshavan et al, 1992), schizophrenia (Silbersweig et al, 1995), and drugs (Ketter et al, 1996).
Keshavan et al reviewed musical hallucinations. In epilepsy, they noted six cases from Hecaen and Ropert in whom music occurred as part of an epileptic aura, four of whom had concomitant ear disease; three case reports without structural brain lesions, two with pronounced deafness, the third with attacks of nausea and rotatory vertigo but no otological investigations; three of Penfield’s cases with brain tumours but no ear or hearing examination. In one of these last three, seizures were of a “new and curious type”, including deafness, unilateral voices, and music, starting two years after tumour removal. Far from implicating the brain, this review of musical hallucinations strongly implicates the ear. Even if epileptic patients are not known to have ear disease, this should be suspected, as Jackson and Gowers established last century that epilepsy can arise from the ear.
In their PET study on hallucinating schizophrenic patients, Silbersweiget al found increased blood flow in the thalamus and not the neocortex. This is not evidence that auditory hallucinations are generated in the brain; instead, it is consistent with peripheral impulses funnelling up via the thalamus.
The volunteers of Ketter et al “consistently (29 out of 32 subjects) reported procaine induced auditory hallucinations (unformed buzzing, ringing, or electronic sounds)”. They considered procaine a selective limbic activator, even though no change in cerebral blood flow corresponded to the “auditory hallucinations”. In an ear, nose, and throat clinic the above noises would definitely be labelled tinnitus, and an obvious peripheral cause usually found. In line with this, procaine causes cochlear hyperactivation (increased wave I and II amplitude in dogs)3; in view of the link between epilepsy and the ear, it is noteworthy that the EEG changed from an awake pattern to seizure activity.
The1968 review of sensory deprivation by Petrella et alwas cited as an example of auditory hallucinations from sensory disinhibiton. I reviewed some of this literature,4concluding that for musical hallucinations there is as much evidence for ear disease and labyrinthine hyperactivity as when they occur in all other conditions and diseases. Deprivation is a misnomer; in many experimental situations white noise is used to mask environmental sounds, whereas if all background noises are reduced, normal subjects will start to have tinnitus. I recently tested a 16 year old dyslexic patient who had never known silence (“silence has a permanent noise running through it”), yet had never complained of tinnitus. Sensory depravation is more likely than deprivation, as in fact misprinted!1
A similar case to that of Brasić and Perry1 throws considerable light on pathophysiological processes involved. Both were tormented by voices of devils while having symptoms of unilateral ear disease; saw devils and animals interchanging; had fiery visions (fire surrounded by cockroaches, devil transforming into burning wisp of straw); possible fluid in one middle ear (blocked grommet plus air-bone gap; filthy discharge from one ear). Conspicuously omitted1 was any mention of tinnitus, the most likely generator of the auditory hallucinations. Fortunately, the other case was Martin Luther,5 who clearly described ringing in his ears, unendurable buzzing, thundering, cracks, thumps, etc. Once, he had a musical hallucination (bells of specific churches) while awake in bed because of noises in his head. Curiously, Luther did not have a simple demonic or religious explanation for his torments. Instead he blamed Satan for his Meniere’s symptoms (headache, episodic vertigo, tinnitus), which in turn he recognised caused his hallucinations. In fact his symptoms were typical of otosyphilis, and there was good evidence that he had “French disease”.
If Brasić and Perry still assert a CNS origin for auditory hallucinations they need an original case report, not reviews or secondary sources. Repeated appeals for non-otological neurological musical hallucinations have failed5 (suggested cases with brainstem lesions were also deaf). I would now like to broaden the challenge to cover auditory hallucinations as well. Unless someone can come up quickly with a case of auditory hallucination due to a clear neurological lesion in someone with normal ears and hearing, the only proved cause of auditory hallucinations is otological.
Brasić and Perry reply:
Gordon conjectures that otological pathology is the necessary and sufficient condition for auditory hallucinations. We disagree. We hypothesise that auditory hallucinations have many aetiologies which can be classified as otological, neurological, neuropsychiatric, and combined. Auditory hallucinations may result from the multiple effects of otopathology, such as altered signal transduction in hair cells.1-1 For example, in response to minimal environmental stimuli, diseased cochlear hair cells may generate random frequencies producing white noise perceived as tinnitus in some persons. Auditory hallucinations may also result from neurological illnesses,1-2 including after right temporal lobectomy for intracerebral hemorrhage without seizures.1-3 We are preparing a manuscript concerning auditory hallucinations in neurological disorders. Auditory hallucinations due to neuropsychiatric disorders are being studied, particularly in schizophrenia. On functional MRI, two patients with schizophrenia experiencing auditory hallucinations showed reduced responses of the temporal cortex to external auditory stimuli. Therefore, auditory hallucinations in some patients with schizophrenia may correspond with maximal activation of the auditory association cortex.1-4 The physiology of thinking in words was assessed utilising PET in six persons with schizophrenia who experienced auditory hallucinations, six persons with schizophrenia who did not experience auditory hallucinations, and six normal controls. Imagining sentences spoken by another person activated the left middle temporal gyrus and rostral supplementary motor area of normal controls and schizophrenic patients without auditory hallucinations, but not schizophrenic patients with auditory hallucinations. Thus, some people with schizophrenia with auditory hallucinations seem to lack activation of portions of the brain associated with the monitoring of inner speech.1-5 These reports suggest that auditory hallucinations in schizophrenia are correlated with physiological abnormalities of regional cerebral blood flow in the left auditory association cortex and rostral supplementary motor area. Auditory hallucinations also are associated with combinations of otological, neurological, and neuropsychiatric disorders. For example, our patient had conductive hearing loss, bilateral tinnitus, and psychosis.1-2 We agree with Gordon that a thorough otological history and examination including audiology is a necessary component of the assessment of a person with auditory hallucinations.