SEARCH

SEARCH BY CITATION

Keywords:

  • Musicogenic;
  • Therapy;
  • Carbamazepine;
  • Temporal lobectomy

Summary

  1. Top of page
  2. Summary
  3. Musical Processing and the Human Brain
  4. Music and Epilepsy
  5. Music Therapy for Epilepsy
  6. Mechanisms for the Proconvulsant and Anticonvulsant Effects of Music in Epilepsy
  7. The Effect of Antiepileptic Therapies on Musicality
  8. Preoperative and Postoperative Evaluation of Musical Functioning in Patients with Epilepsy
  9. Conclusions
  10. Acknowledgments
  11. Disclosure
  12. References

The effect of music on patients with epileptic seizures is complex and at present poorly understood. Clinical studies suggest that the processing of music within the human brain involves numerous cortical areas, extending beyond Heschl’s gyrus and working within connected networks. These networks could be recruited during a seizure manifesting as musical phenomena. Similarly, if certain areas within the network are hyperexcitable, then there is a potential that particular sounds or certain music could act as epileptogenic triggers. This occurs in the case of musicogenic epilepsy, whereby seizures are triggered by music. Although it appears that this condition is rare, the exact prevalence is unknown, as often patients do not implicate music as an epileptogenic trigger and routine electroencephalography does not use sound in seizure provocation. Music therapy for refractory epilepsy remains controversial, and further research is needed to explore the potential anticonvulsant role of music. Dopaminergic system modulation and the ambivalent action of cognitive and sensory input in ictogenesis may provide possible theories for the dichotomous proconvulsant and anticonvulsant role of music in epilepsy. The effect of antiepileptic drugs and surgery on musicality should not be underestimated. Altered pitch perception in relation to carbamazepine is rare, but health care professionals should discuss this risk or consider alternative medication particularly if the patient is a professional musician or native-born Japanese. Studies observing the effect of epilepsy surgery on musicality suggest a risk with right temporal lobectomy, although the extent of this risk and correlation to size and area of resection need further delineation. This potential risk may bring into question whether tests on musical perception and memory should form part of the preoperative neuropsychological workup for patients embarking on surgery, particularly that of the right temporal lobe.

Music is an integral part of everyday life and culture. For most people, listening or playing music is a pleasurable experience that may evoke a memory or emotion. Advances in technology have improved our access to all kinds of music, with many people now downloading and sharing digitized music files and using portable devices to play and store large collections of music.

Although the environmental link between photosensitivity and epilepsy is well known, the interaction between music and epilepsy is less well perceived or understood. It appears some patients with seizures might gain benefit from musical exposure, whereas other patients may experience an exacerbation of their seizures. The dichotomous effect of music on epilepsy is an intriguing yet poorly understood phenomenon and the subject of ongoing research and debate. In addition, the effects of medications and surgical procedures used in epilepsy may have the potential to alter a patient’s musicality, which could have disastrous consequences.

This review aims to summarize the current body of evidence on music and its association with epilepsy. It outlines our current understanding of musical processing in the human brain, and how this process could be disrupted in the production and propagation of seizures. Also discussed are ictal musical phenomena, music therapy in refractory epilepsy, and the effect of antiepileptic medication and epilepsy surgery on musicality.

Musical Processing and the Human Brain

  1. Top of page
  2. Summary
  3. Musical Processing and the Human Brain
  4. Music and Epilepsy
  5. Music Therapy for Epilepsy
  6. Mechanisms for the Proconvulsant and Anticonvulsant Effects of Music in Epilepsy
  7. The Effect of Antiepileptic Therapies on Musicality
  8. Preoperative and Postoperative Evaluation of Musical Functioning in Patients with Epilepsy
  9. Conclusions
  10. Acknowledgments
  11. Disclosure
  12. References

The relation between music and the brain has been extensively researched over the past century, from the authoritative work published by Critchley and Henson (1977) titled “Music and the Brain” to a modern update of the literature from intervening decades by Stewart et al. (2006).

During the late 19th century, German researchers published numerous studies analyzing the disturbance of musical functioning in patients with brain damage, observing how focal lesions affect musical activities. In 1888, Knoblauch introduced the term “amusia” meaning impaired musical capabilities. He described a sensory (receptive) amusia whereby affected patients cannot hear, read, or understand music, and motor (expressive) amusia whereby patients have an inability to sing, write, or play music (Knoblauch, 1888). Several other authors explored lesional consequences on music perception (Head, 1926; Kleist, 1962; Luria et al., 1965), and a plethora of interesting discoveries were made. Although many cases of amusia were related to an abnormality within the right temporal lobe, this was by no means true in all reported cases and some lesions lateralized to the left hemisphere. Thalamic lesions in association with a hemiparesis were reported to cause a peculiar alteration in musical perception on the affected side. One patient seemed unable to tolerate hymns in church on his affected side, causing him to rub the affected hand (Head, 1920). Although these studies provided some evidence on the anatomic components of musical perception, the cases were confined to patients with damaged brains, and it is possible that the underlying pathologic process may have affected the results.

Our modern understanding of the mechanisms involved in musical processing has been aided by advances in functional imaging studies, using positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) techniques. These techniques display the changes in hemodynamic response to mean synaptic firing rates in the brain, and have allowed researchers to delineate the anatomic areas pertinent to the processing of music. Electroencephalography (EEG) and magnetoencephalography (MEG) studies have also been employed to record postsynaptic potentials during listening to music. Recent studies have also identified the structural and functional organization of the brain in musicians.

Musical processing encompasses brain mechanisms in musical perception, recognition, and emotion. Musical perception requires the decoding of a musical stimulus within the primary auditory cortex in Heschl’s gyrus and the association cortex in the superior temporal gyrus (planum temporale). The primary auditory cortex is thought to receive thalamic afferents from the medial geniculate nucleus, which in turn connect through networks to the association cortex, mesolimbic systems, and other multisensory cortices (Stewart et al., 2006). The primary auditory cortex appears especially sensitive to tone, whereas the auditory association cortex is thought to perceive pitch (Penagos et al., 2004) and perform more complex musical processing tasks relating to linear stimuli, for example, melodies (Liegeois-Chauvel et al., 1998) and nonlinear stimuli, for example, chords and consonances. Similar areas within the secondary auditory cortex are also activated in speech (Price et al., 2005). The perception of rhythms, with no particular melodic content, is thought to involve activation of the cerebellum and basal ganglia as well as the superior temporal lobes, suggesting a motor aspect to rhythm perception.

Musical recognition and emotion are thought to involve orbitofrontal areas and the limbic system, which may serve to store past auditory memory and emotional evaluation of a musical stimulus (Dellacherie et al., 2009). For example, chord dissonance per se may be detected via the auditory association cortex, but the evoked unpleasant emotional response appears to be mediated via limbic structures.

An extensive review by Stewart et al. (2006)) analyzed 38 case reports and 27 case series reporting on symptomatic musical listening deficits, with associated structural imaging findings. The authors summarized anatomic areas that are pertinent to pitch processing, temporal processing, musical memory, and emotional responses to music using five cartoons. Anatomic areas implicated in 50% or more of the studies for a particular function were mapped out using colored dots, with the size of the dot representing the proportion of studies (Fig. 1). The review ascertained that the anatomic areas implicated in musical listening deficits were central, affecting a number of regions beyond Heschl’s gyrus, with right-sided predominance. Half of cases were associated with coexisting problems of speech perception.

image

Figure 1.   Critical brain substrates for musical listening disorders according to pitch processing, temporal processing, musical memory, and emotional responses to music. The presence of a colored circle corresponding to a particular function in a region indicates that at least 50% of studies of the function implicate that region. The size of each circle is scaled according to the proportion of studies of the function implicating that region. Meter is not represented, as no brain area was implicated in 50% or more of cases. amyg, amygdala; aSTG, anterior superior-temporal gyrus; bg, basal ganglia; cc, corpus callosum; fr, frontal; hc, hippocampal; HG, Heschl’s gyrus; ic, inferior colliculi; i, inferior; ins, insula; l, lateral; m, medial; thal, thalamus; PT, planum temporale; TG, temporal gyrus. Reprinted from Stewart et al. (2006) with permission from Oxford Journals.

Download figure to PowerPoint

Determinants of musicality have also been explored over the years by examining environmental and genetic factors. Musicality is likely to be a polygenic trait and dependent on home environment and parental attitudes to the development of musical talent in children.

Studies have also identified structural differences in the brains of musicians versus nonmusicians in a number of anatomic areas including auditory (Schneider et al., 2002), motor (Amunts et al., 1997), somatosensory, superior parietal (Gaser & Schlaug, 2003), callosal (Schlaug et al., 1995), and cerebellar (Hutchinson et al., 2003) areas. The larger size of the corpus callosum in musicians may indicate increased cross-communication between both sides of the brain, merging spatial-emotional processing of the right brain with linguistic analytical processing of the left brain. People with absolute pitch engage left dorsofrontal regions, whereas those lacking absolute pitch appear to have a so-called working memory for pitch located within inferior frontal areas. The age of onset of musical training, that is, younger than 4 years old, appears to be a key predictor of absolute pitch, suggesting that this ability develops during a critical period for imprinting (Sergeant & Roche, 1973).

Music and Epilepsy

  1. Top of page
  2. Summary
  3. Musical Processing and the Human Brain
  4. Music and Epilepsy
  5. Music Therapy for Epilepsy
  6. Mechanisms for the Proconvulsant and Anticonvulsant Effects of Music in Epilepsy
  7. The Effect of Antiepileptic Therapies on Musicality
  8. Preoperative and Postoperative Evaluation of Musical Functioning in Patients with Epilepsy
  9. Conclusions
  10. Acknowledgments
  11. Disclosure
  12. References

It is possible that brain mechanisms involved in musical processing may be involved in the generation and propagation of seizures, manifesting with musical semiology. Hyperexcitable cortical areas may also become sensitized to specific musical triggers and may explain the basis of musicogenic epilepsy.

Musicogenic epilepsy

Musicogenic epilepsy was first coined by Critchley, 1937, who described a rare form of epilepsy in which seizures, typically simple or complex partial types, are triggered by music. However, descriptions of possible music-induced seizures had been reported in the literature as early as 1841. The Chinese poet Jichin Kyo, quoted by Fujinawa et al. (1977) stated:

Since my remote boyhood I have always been absent minded while hearing the sound of a street vendor’s flute. I fall sick when I hear the sound of the flute in the evening sun, although I do not know the reason.

Musicogenic epilepsy has an estimated prevalence of one case per 10,000,000 population. It is classified as a rare form of complex reflex epilepsy by the International League Against Epilepsy (ILAE) (Berg et al., 2010), with seizures induced by listening to music in most cases, but playing, thinking, or dreaming of music have all been cited. In reported cases, seizures following a musical stimulus can often be delayed by several minutes. During this latent period, patients may experience distress, agitation, tachycardia, and rapid breathing building up to the seizure, although this is inconsistently reported. The seizures may not all be exclusively stimulated by music, and there is reported variability in the form of musical stimulus. For example some patients report seizures according to type (jazz, classical, choral, popular), instrument (organ, flute, piano), emotional content of the music (sad, sentimental, upbeat), or even composer (Wagner, Beethoven, Beatles). Some patients have also found strategies that enable them to abort a seizure. These features have led one author to question whether musicogenic epilepsy is a true form of reflex epilepsy (Vizioli, 1989).

In many cases, however, the seizure has been reported to occur within seconds and following a very specific stimulus, for example church bells (Poskanzer et al., 1962) or the melody of the Marseillaise (Vercelletto, 1953). Other reports cite the incorrect positioning of the larynx leading to a husky or “metallic” quality to singing causing seizures (Brien & Murray, 1984), whereas others cite cognitive processing (Ogunyemi & Breen, 1993), playing a certain hymn (Sutherling et al., 1980), or improvisation (Le Chevalier et al., 1985) as triggering events.

A recent review reported on 110 cases of music-evoked seizures published between 1884 and 2007 (Pittau et al., 2008). The mean age of onset of musicogenic seizures was 28 years, with a female predominance. Patients more often than not had high musicality. Seizures induced exclusively by music occurred in 34 patients, with the majority also reporting spontaneous seizures. Autonomic auras were common with oroalimentary automatisms occurring in more than one fourth of cases. In 60 patients with ictal electroencephalography (EEG), the focus of activity localized to the temporal lobe with right-sided predominance. Ictal single photon emission computed tomography (SPECT) in six cases demonstrated abnormalities within the right temporal lobe involving mesial structures in four and within the left temporal lobe in two cases. The musical stimulus varied greatly, with a small proportion identifying certain tones or sounds causing seizures and others reporting that any form of music triggered events. The review also reported an fMRI study on a patient with typical music-induced seizures played “neutral” music and “emotionally charged melody” music. It revealed that during the playing of neutral music, only acoustic areas were highlighted, but during emotional melodies, the activity also involved frontooccipital areas. This broadly correlates with findings by Mórocz et al. (2003), who demonstrated early activation of the right frontoorbital area prior to seizure activity from the left temporal lobe. However, Mórocz et al. (2003) suggested that activity provided evidence of emotional arousal and memory related to the music rather than to seizure activity per se, whereas in the review, frontal cortical activity was thought to represent the beginning of paroxysmal discharges undetected by scalp EEG.

Invasive monitoring of seizure propagation in musicogenic seizures has been reported in three studies. Subdural recordings from one patient revealed cortical dysplasia within the superior temporal gyrus (Trevathan et al., 1999). One study using a subdural grid showed foci in the right lateral temporal lobe, right mesial temporal lobe, and bilateral temporal lobes (Tayah et al., 2006). A third study used fluorodeoxyglucose (FDG)–PET with invasive techniques (subdural grid and depth electrodes) and identified a focus within the right mesial temporal lobe with propagation to the lateral temporal cortex, Heschl’s gyrus, insula, and frontal lobes (Mehta et al., 2009).

These clinical studies suggest an affective role in musicogenic seizures, as evidenced by mesial temporal activity. Authors speculate that emotion triggered by music is the causal factor rather than the auditory content of the music per se. Merely thinking about music has been reported as a trigger, suggesting the importance of musical memory and emotion.

Responses to limbic stimulation in subjects with epilepsy are thought to depend on widespread neuronal matrices linked through connections that have become strengthened through repeated use (Gloor, 1990). This is of interest given that the onset of seizures often predates the sensitivity to music.

However, the imaging studies provide less obvious clues as to why a subset of patients may develop seizures on exposure to a simple sound or tone, where there may be a less obvious emotional component (Wieser et al., 1997). In addition, musicogenic seizures have since been reported in children as young as 6 months old (Lin et al., 2003), suggesting musicogenic seizures are likely to be epileptiform and not functional in nature.

Treatment of musicogenic seizures usually comprises avoidance of the musical triggers together with antiepileptic medication. In the past, various behavioral therapies have been tried in patients with high emotional states. Joynt et al. (1962) suggested the use of sensory extinction, by playing innocuous music or small snatches of noxious music prior to playing the seizure-producing music, the idea being that cells typically initiating seizure activity would be set into activity by similar combinations of tones and chords, rendering them less susceptible to the actual seizure-triggering melody. Thereby, excitable nerve cells are “usurped” by a competing stimulus. Other methods of psychotherapy and deconditioning techniques have been used as treatment in musicogenic epilepsy patients (Daly & Barry, 1957; Forster et al., 1965).

At present the exact pathophysiology of musicogenic seizures is undetermined. Questions around what make a certain type or facet of music epileptogenic to a particular person remain unanswered, but it is possible that hyperexcitable cortical areas could be stimulated to different degrees and extents by different musical stimuli. These questions remain an intriguing subject of ongoing research.

Musical ictal phenomena

Music may manifest as part of ictal semiology in association with temporal lobe epilepsy. The phenomena could be either positive, as in musical auditory hallucinations, musicophilia, ictal singing or whistling, or negative as in ictal aprosody and amusia.

Musical hallucinations

Some patients may develop musical auditory hallucinations as part of temporal lobe epilepsy. In comparison to musicogenic epilepsy, there is a less clear association with high musicality, and the emotional associates of temporal lobe attacks with musical components are varied including anger, fear, and panic. In a cohort of 666 patients with temporal lobe epilepsy, around 16% had auditory hallucinations that took the form of sounds such as ticking or banging or more complex forms such as a specific melody or orchestral piece (Currie et al., 1971). These seizures were associated with activation of the superior temporal gyrus with right-sided predominance. In addition, some patients who experienced voices rather than music as a feature of their seizures had stimulation within the same cortical regions as those triggered by a musical stimulus. More recent research in subjects with musical hallucinations implicated an autonomous network of nonprimary auditory areas in generating abnormal neural activity that sustains musical hallucinations (Griffiths, 2000).

Musicophilia

Musicophilia or musical craving is an extremely rare phenomenon. It has been reported in a patient with a stroke causing damage to the left hemisphere (Jacome, 1984) and also in frontotemporal dementia (Boeve & Geda, 2001). Rohrer et al. (2006) described an interesting case of a 65-year-old woman with right-sided temporal lobe epilepsy who developed musical craving following commencement of lamotrigine for complex partial seizures. Before introduction of treatment she had been indifferent to music, avoiding music where possible and never attending concerts. She had no musical training. Within weeks of treatment she was actively listening to classical music stations for several hours a day, and demanding to attend musical concerts. She described listening to music as an intensely pleasurable and highly emotional experience. There were no other changes to her behavior or personality.

In this case the musicophilia was driven by the pleasurable emotional response derived from the music. The authors speculated that emotional responses to music might have been altered by functional reorganization of neocorticolimbic interactions as a consequence of long-standing seizures. Treatment might enhance these altered emotional responses by restoring flow within the reorganized sensory-limbic networks. This theory relates to the described phenomenon of “forced normalization,” whereby patients may develop behavioral changes, for example depression in the context of improved seizure control. This is thought to occur due to electrophysiologic and neurochemical alterations involving limbic circuitry (Krishnamoorthy et al., 2002).

There have been other fascinating accounts of patients reporting musicophilia and musical inspiration in association with altered emotional perception: following electrocution by lightning and following surgery for a right temporal oligodendroglioma (Sacks, 2008). The underlying neural mechanisms of musicophilia in these cases remain unexplained.

Ictal singing, humming, and whistling

The act of singing involves multiple cortical areas including left superior temporal and parietal regions as well as both left and right premotor cortex, anterior superior temporal gyrus, and planum polare (Callan et al., 2006).

Singing as an ictal phenomena is rare. There have been nine cases of ictal singing reported in the literature (Enatsu et al., 2011). Four had frontal lobe epilepsy, three had temporal lobe epilepsy, and in two cases the epileptogenic zone could not be localized. There has also been variability as to whether the ictal focus emanates from a dominant or nondominant focus. In one case the authors were able to distinguish different localizations depending on whether the patient was humming or singing. Humming appeared to occur due to a temporal focus, whereas singing occurred due to a focus within the frontal lobe, specifically within the right prefrontal cortex (Bartolomei et al., 2007).

Authors have speculated that ictal singing is a form of automatism that occurs either from a release phenomenon, or by activation of learned motor patterns or memories (Meierkord & Shorvon, 1991). Because the action of singing involves multiple regions, it suggests that ictal singing may occur through recruitment of a propagation network mimicking musical action via numerous cortical areas, rather than it occurring via a specific cortical focus.

There have been five cases of ictal whistling reported within the literature. (Lazzarino & Valassi, 1982; Tan et al., 1990; Loring et al., 1994; Raghavendra et al., 2010). Lateralization has not been consistent among cases. One report suggested a frontal focus, whereas the remaining cases reported this rare automatism with temporal lobe epilepsy. In two cases, ictal whistling was abolished by temporal lobe surgery. The act of whistling requires the utilization of perioral, oral, and respiratory muscles and the recruitment of a complex neuronal network. Functional imaging has demonstrated multiple areas involved in whistling including inferior rolandic cortex, cingulate cortex, basal ganglia, amygdala, thalamus, and cerebellum (Dresel et al., 2005).

Ictal aprosody and amusia

Disorders of prosody refer to abnormalities in the affective components of speech including intonation, melody, pitch, and gestures. This abnormality is considered either expressive or receptive and may occur due to lesions in the nondominant hemisphere. Amusia, as described previously, and aprosody are considered to be negative phenomena often occurring as a result of stroke. However, there has been one reported case of expressive amusia and aprosody manifesting as an ictal feature of seizures emanating from a right temporooccipital lobe focus (Bautista & Ciampetti, 2003). On commencing antiepileptic treatment, the clinical events ceased and the patient’s musical capabilities were restored. Negative ictal phenomena are poorly understood and can include ictal blindness, paralysis, speech arrest, and loss of hearing. They are thought to occur either via stimulation of inhibitory cortical regions, dampening of receptive abilities of sensory regions, or inhibition of spinal motor neurons.

Music Therapy for Epilepsy

  1. Top of page
  2. Summary
  3. Musical Processing and the Human Brain
  4. Music and Epilepsy
  5. Music Therapy for Epilepsy
  6. Mechanisms for the Proconvulsant and Anticonvulsant Effects of Music in Epilepsy
  7. The Effect of Antiepileptic Therapies on Musicality
  8. Preoperative and Postoperative Evaluation of Musical Functioning in Patients with Epilepsy
  9. Conclusions
  10. Acknowledgments
  11. Disclosure
  12. References

The therapeutic potential of music has largely been explored in cognitive science. The cognitive effects of music are well documented in the literature, although these effects have been subject to scrutiny. Rauscher et al. (1993) observed an immediate enhancement in spatial-temporal reasoning in college students exposed to 10 minutes of the Mozart Sonata K448 (Rauscher et al., 1993). The authors subsequently coined the term the “Mozart effect.” Further studies in Parkinson’s disease, senile dementia, and attention-deficit/hyperactivity disorder also described cognitive benefits from listening to music (Pacchetti et al., 2000; Foster & Valentine, 2001; Rickson & Watkins, 2003; Turner, 2004).

Rauscher et al. went on to devise an animal model of the “Mozart effect” and demonstrated that rats exposed to long-term Mozart in utero and also postpartum performed better in negotiating a T maze than control rats that had been exposed to silence, white noise, or the music of Philip Glass. The effect lasted for at least 4 h after months of exposure (Rauscher et al., 1998). Further studies in mice showed similar enhanced maze skills for up to 24 h having been exposed to 12 h of the Mozart sonata but not with mice exposed to Beethoven (Aoun et al., 2005). Animal studies have also shown that music exposure may enhance dendritic branching, cell proliferation, and neurogenesis in the hippocampus and amygdala (Kim et al., 2006).

The reported musical effects on cognition have been viewed with scepticism among neuroscientists. Concerns have arisen from the lack of matched musical controls within experiments. It could be argued that Hayden is a better musical match to Mozart in terms of structure than Beethoven or Glass. In addition, the short duration of effect and difficulty in controlling for affective responses to music in experiments has attracted criticism.

The evidence for music as an anticonvulsant in human epilepsy is limited as is our understanding of the brain mechanisms involved. The clinical evidence is confined mainly to Mozart music and in Taiwanese children with generalized seizures. Two small studies by the same author reported on the effect of Mozart K448 on epileptiform discharges and seizure frequencies in Taiwanese children. The first study included 58 Taiwanese children with partial epilepsy. Continuous EEG monitoring occurred before, during, and after eight minutes of exposure to the Sonata. In 81% of patients there was a reduction in interictal discharges by an average of 33%, with the greatest reduction in those patients with generalized discharges. However, around 20% showed an increase in interictal discharges by an average of 14%. No significant differences were found according to gender, IQ, or number of antiepileptic drugs and response to music. The reduction in interictal discharges was not dependent on the level of alertness or specific emotional response (Lin et al., 2010). The second study included 11 Taiwanese children aged 2–14 years old with refractory epilepsy. Two-thirds had generalized seizures of symptomatic causes, and the majority (70%) had learning difficulties. Seizure frequency was observed for 6 months before music and during 6 months of Mozart K448 exposure. The study found that 73% of patients had a 50% or greater reduction in their seizure frequency, with two patients becoming seizure free during exposure to music. There were no significant differences in response status according to seizure type, IQ, etiology, or gender (Lin et al., 2011).

There have also been reports of a reduction in interictal discharges when exposed to Mozart K448 in patients with rolandic types of seizures (Turner, 2004). Reductions in ictal spiking have been reported in subjects in coma and refractory nonconvulsive status following exposure to Mozart Sonata K448 (Hughes et al., 1998; Lahiri & Duncan, 2007; Kuester et al., 2010), and Johann Sebastian Bach (Miranda et al., 2010). Authors suggest a direct cortical response to Mozart as opposed to responses relying on emotion or level of awareness.

Neuroscientists have also published on possible theories for the anticonvulsant effect of music. Hughes et al. (1998) discussed the Trion model and that exposure to a highly patterned or superorganized stimulus (in space and time) may lead to learning of innate memory patterns, which in turn may decrease the excitability of an epileptogenic focus.

Other authors have implicated mirror neurons in mediating the anticonvulsant effect. Mirror neurons discharge or are modified when an individual is performing an action while exposed to musical or visual stimulation (Molnar-Szakacs & Overy, 2006). Musical activities, such as dancing to music, playing musical instruments, and moving mouths and larynxes to sing are examples in which music and motor functioning are connected. Authors speculate that mirror neurons mediate neuronal activity by linking auditory stimulation directly to the motor cortex. Motor system modulation in transcranial magnetic stimulation and in behavioral studies has been observed to change during auditory stimulation (Buccino et al., 2005).

A further theory proposes modification of the dopamine neurotransmitter pathways in the effect of Mozart K.448 on epilepsy. Exposure to music is known to increase the expression of dopamine in the brain (Sutoo & Akiyama, 2004). In recent years, dopamine has been reported to have an important role in the pathophysiology of epilepsy. The reduced binding capacity of dopamine receptors in the basal ganglia has been hypothesized to contribute to seizures in autosomal dominant frontal lobe epilepsy (ADNFLE) (Fedi et al., 2008) and juvenile myoclonic epilepsy (JME) (Landvogt et al., 2010). An FDG-PET study in seven patients with mesial temporal lobe epilepsy identified reduced D2/D3 receptor binding (Werhahn et al., 2006). In a study in animals, pilocarpine-induced seizures altered the binding capacity of dopaminergic receptors in striatal and hippocampal areas, and facilitated the propagation and maintenance of seizures (Mendes de Freitas et al., 2005). It is possible that listening to music modifies dopaminergic pathways within subcortical structures to influence thalamocortical projections.

In summary, further research is needed to explore the potential anticonvulsant role of music both at the basic science and clinical level. The research to date has focused principally on Mozart music without a good understanding of the basic science behind the effect. There is limited evidence of an effect with Mozart on seizures in the clinical setting. It may be that some aspects of musical stimuli will ultimately prove to have therapeutic potential, but for the moment these remain largely unknown and unproven and our understanding of the brain mechanisms involved are limited.

Mechanisms for the Proconvulsant and Anticonvulsant Effects of Music in Epilepsy

  1. Top of page
  2. Summary
  3. Musical Processing and the Human Brain
  4. Music and Epilepsy
  5. Music Therapy for Epilepsy
  6. Mechanisms for the Proconvulsant and Anticonvulsant Effects of Music in Epilepsy
  7. The Effect of Antiepileptic Therapies on Musicality
  8. Preoperative and Postoperative Evaluation of Musical Functioning in Patients with Epilepsy
  9. Conclusions
  10. Acknowledgments
  11. Disclosure
  12. References

To evaluate the therapeutic potential of music we need to try and understand why music might behave as anticonvulsant in some patients with epilepsy yet proconvulsant in others. Understanding why this paradox occurs may be explained through the dichotomous effect of dopamine on receptors in the brain. Previous studies have shown that the anticonvulsant action of dopamine has been attributed to D2 receptor stimulation in the forebrain, whereas selective D1 receptor activation appears to lower the seizure threshold both clinically and in animal models (Al-Tajir & Starr, 1990; Starr, 1996). Studies in mesial temporal lobe epilepsy have identified reduced D2/D3 binding capacity in the striatum. This may be a consequence of receptor downregulation, due to ongoing seizures (Starr, 1996). Striatal dopamine receptor downregulation might disinhibit thalamocortical connections by means of inhibition of the substantia nigra, thereby further enhancing cortical hyperexcitability. This has been explored in animal models of absence epilepsy (Deransart et al., 2000) A further study has ascertained that signaling through D2 receptors could play a neuroprotective role against pathologies involving glutamate-induced neurodegeneration (Bozzi et al., 2000). Playing music may promote dopamine release thereby flooding the dopaminergic systems and upregulating D2 receptors. In patients with temporal lobe epilepsy, dopamine flooding may potentially behave in an anticonvulsant way through D2 receptor activation. In musicogenic epilepsy, it is possible that the emotional effect of the musical trigger could lead to increased dopamine within the medial prefrontal cortex (PFC) (Kaneyuki et al., 1991). This increase in dopaminergic activity within the PFC could in turn suppress the limbic dopaminergic response and result in the propagation of seizures. The limbic-PFC circuit has also been implicated in an animal model of anxiety (Bishop, 2007). This is interesting given that anxiety correlates with a lack of seizure control in JME, and a risk of >20 lifetime generalized tonic–clonic seizures (De Araujo Filho et al., 2006). Further research is needed into the role of dopamine pathways in epilepsy and the interaction with music.

Another interesting area of research is the ambivalent action of sensory and cognitive input in ictogenesis. This is relevant to understanding the effect of music in epilepsy given that limbic and multisensory cortices are involved in musical processing.

A study by Guaranha et al. (2009) observed the effect of cognitive tasks on epileptiform discharges in 76 patients, aged 12–53 years old with JME. The study used a video-EEG protocol recording brain activity during action programming tasks, for example, spatial construction, and thinking tasks, for example, mental calculation. The recording was performed over 4–6 h, and EEG discharge rates during specific tasks were compared to a baseline awake EEG recording. The study observed a provocative effect on discharges during cognitive tasks in 29 patients and an inhibitory effect in 28 patients. In 30 patients there was no effect on discharges during cognitive tasks. Action programming tasks, particularly manual praxis, were the most provocative, whereas thinking tasks, in particular mental calculation, were the most inhibitory. The effect of cognitive tasks on epileptiform discharges appeared independent of drug treatment, seizure control, or age of patient. These findings broadly correlate with other studies (Matsuoka et al., 2002; Matsuoka et al., 2005). The authors speculated that cognitive tasks recruiting motor pathways were more epileptogenic, whereas spatial thinking involving activation of parietal regions without motor recruitment may act to dampen activity in neighboring motor cortices and thus seizure activity. Music may act as anticonvulsant by activating and enhancing areas of cortex involved in spatial cognitive processes that produce greater inhibition, either directly on surrounding motor areas or via inhibitory corticothalamic feedback loops. Similarly music may act as proconvulsant in musicogenic epilepsy where anticipation of music and memory responses may enhance activity in the PFC and associated frontostriatal connections (Leaver et al., 2009).

The Effect of Antiepileptic Therapies on Musicality

  1. Top of page
  2. Summary
  3. Musical Processing and the Human Brain
  4. Music and Epilepsy
  5. Music Therapy for Epilepsy
  6. Mechanisms for the Proconvulsant and Anticonvulsant Effects of Music in Epilepsy
  7. The Effect of Antiepileptic Therapies on Musicality
  8. Preoperative and Postoperative Evaluation of Musical Functioning in Patients with Epilepsy
  9. Conclusions
  10. Acknowledgments
  11. Disclosure
  12. References

Music exposure may thus influence seizures, but the treatment of seizures may affect musicality. This could have potentially disastrous effects, particularly if the patient relies on his or her musical skills for employment.

Antiepileptic drugs

Carbamazepine has been associated with a reversible disturbance of pitch perception. Since 1993, there have been 26 reported cases of this adverse effect reported in the literature (Tateno et al., 2006). Of interest is that all but one case have involved Japanese patients. Patients were aged between 4 and 42 years old with a mean age of 19 years and most were female. Twenty of 26 patients were prescribed carbamazepine for epilepsy. Symptoms of altered pitch perception seem to develop 2 h to 2 weeks after the administration of carbamazepine. On reduction in dose or discontinuation of carbamazepine, the abnormality appeared to fully reverse. Patients perceived a lower pitch than usual and in the majority of cases by a semitone. Many of the patients were musical and had studied an instrument.

The mechanism of pitch disturbance with carbamazepine is unknown. Authors speculate that carbamazepine acts at a local level, either by changing the mechanics of the organ of Corti (Chaloupka et al., 1994), or by affecting the sarcolemma of the stapedius muscle and altering the tension on the tympanic membrane receiving sound. Alternatively carbamazepine may act at a central level. Carbamazepine-treated patients have been found to have prolonged I–III and I–V latencies of the auditory brainstem response (Medaglini et al., 1988). However, auditory brainstem responses in patients with abnormal pitch perception have been reported to be normal. Acoustic effects of lamotrigine in pediatric patients with epilepsy have been assessed in a recent study. The authors examined speaking rate, pitch, and articulation rate, finding no evidence of adverse effect with lamotrigine (Yun et al., 2011).There have not been any reports of studies on other antiepileptic drugs and effect on musicality.

Temporal lobe surgery

The effect of temporal lobe surgery for intractable epilepsy on musical processing is an important issue and the consequences for patients could be devastating. Tables 1 and 2 highlight published studies examining the effect of temporal lobe surgery on musical emotion, recognition, and perception. Most of these studies have involved small numbers of surgical patients, and the methods used for assessing musical processing have varied considerably.

Table 1.   Summary of studies investigating the effect of temporal lobe surgery for refractory epilepsy on musical emotion and recognition
Author and yearResectionNoAge, years (gender)Musical educationTestKey finding
  1. M, mean value; STG, superior temporal gyrus; HG, Heschl’s gyrus; R, right; L, left; C, controls.

Musical emotion      
 Dellacherie et al. (2011)Controls (C)1236 (m)All groups had limited or no formal musical trainingMultidimensional scaling analysis on emotional dissimilarityNo difference in ability to assess emotional dissimilarity in resection groups versus (C)
Medial temporal, sparing STG (L)1038 (m)
Medial temporal, sparing STG (R)937 (m)
 Gosselin (2005)Controls (C)1627–47Patient group: None Control group: Some 5 years trainingEmotional task Error detection taskImpaired rating of fearful music in both patient groups. Ability to recognize happy and sad music not impaired
Medial temporal, sparing STG (L)820–60
Medial temporal, sparing STG (R)820–60
 Khalfa et al. (2008)Controls (C)6032 (m)30% control group and 50–60% of patient groups with 5 years musical trainingMusical excerpts from Montreal Battery of Evaluation of AmusiaImpaired sadness recognition and dissonance in both resection groups versus (C) Impaired happiness recognition in (L) versus (R) and (C)
Anterior temporal including STG (L)1229 (m)
Anterior temporal including STG (R)1434 (m)
Musical recognition      
 Samson and Zatorre (1991)  (Experiment 1)Controls (C)2031 (m)30% of each group with musical trainingSong excerpts (Serafine et al., 1984)Impaired recognition of previously presented tunes in patient groups
Anterior temporal (including HG) (L)2227 (m)
Anterior temporal (including HG) (R)2130 (m)
 Samson and Zatorre (1991)  (Experiment 2)Controls (C)1028 (m)UnknownSong excerpts (Serafine et al., 1984)Impaired recognition of previously presented tunes without lyrics in (R) group
Anterior temporal (including HG) (L)1230 (m)
Anterior temporal (including HG) (R)1229 (m)
 Samson and Zatorre (1992)Controls (C)1528 (m)10% of each group with musical training112 melodic patternsImpaired recognition in determining which tune had previously been presented in both patient groups compared to (C)
Anterior temporal (including HG) (L)2030 (m)
Anterior temporal (including HG) (R)2030 (m)
 Shankweiler (1966)Controls (C)20UnknownDichotic presentation of tunes pre- and postsurgeryImpaired recognition of previously presented tunes postsurgery for (R) group versus (C) and (L)
Anterior temporal (including HG) (L)20
Anterior temporal (including HG) (R)20
Table 2.   Summary of studies investigating the effect of temporal lobe surgery for epilepsy on musical perception
Author and YearResectionNoAge (years)Musical educationTestKey finding
  1. M, mean value; STG, superior temporal gyrus; HG, Heschl’s gyrus; R, right; L, left; C, controls.

Musical perception      
 Johnsrude et al. (2000)Controls (C)1422–52UnknownPitch discrimination and direction tasksImpaired pitch direction task in right temporal resection including HG group
Anterior temporal sparing HG (L)1222–53
Anterior temporal sparing HG (R)522–53
Anterior temporal including HG (L)622–53
Anterior temporal including HG (R)822–53
 Kester et al. (1991)Controls (C)1227 (m)100% of study participants <1 year of trainingMusical aptitude profile Seashore tonal memory testPitch pattern and meter impaired in right temporal lobectomy versus controls and left temporal lobectomy
Anterior temporal, sparing HG (L)927 (m)
Anterior temporal, spaing HG (R)1232 (m)
 Koike et al. (1996)Temporal, including STG (R)1028 (m)UnknownSeashore Measures of Musical TalentsNo impairment of any of the six subsets in any of the resection groups
Temporal, sparing STG (R)1028 (m)
Temporal, sparing STG (L)937 (m)
 Liegeois-Chauvel et al. (1998)Controls (C)2432 (m)10% of each group with musical trainingTemporal organization taskPitch interval impaired in all resective groups particularly posterior STG group versus controls Rhythm impaired only in posterior STG group Meter impaired in anterior STG resections
Temporal, sparing STG (L)1431 (m)
Temporal, sparing STG (R)1930 (m)
Anterior STG (L)539 (m)
Anterior STG (R)830 (m)
Including posterior STG (L aid R)8
 Milner (1962)Temporal, including HG (L)16UnknownSeashore Measures of Musical Talent pre- and postsurgeryPitch pattern, timbre, loudness, and time subtests impaired in (R)
Temporal, including HG (R)11
 Samson and Zatorre (1988)Controls (C)2029 (m)25% of each group with musical trainingMelodic and chord discrimination tasksImpaired pitch pattern in right temporal lobectomy resections including HG versus (C) and (L) No difference in tonal structure versus controls in all resection groups
Anterior temporal including HG (L)2828 (m)
Anterior temporal including HG (R)2629 (m)
Frontal (R)1431 (m)
Frontotemporal (R)925 (m)
 Samson and Zatorre (1994)Controls (C)1526 (m)25% of each group with musical training48 pairs of digtized sounds (see paper)Impaired timbral change in tone pairs in right temporal lobectomy group versus (C) and (L)
Anterior temporal including HG (L)1528 (m)
Anterior temporal including HG (R)1530 (m)
 Samson et al. (2001)Controls (C)14UnknownPsychophysical task Tempora vaidion task (see paper)Impaired detection of temporal irregularity in familiar tunes in (L) versus (C) and (R)
Anterior temporal (L)11
Anterior temporal (R)11
 Samson et al. (2002)Controls (C)1526 (m)25% of each group with musical training36 stimulus pairs played on different timbresImpared judgment in timbral dissmilarity in pairs of melodies in both resective groups versus (C)
Anterior temporal (L)1528 (m)
Anterior temporal including HG (R)1530 (m)
 Warrier and Zatorre (2004)Controls (C)1236 (m)20% of each group with musical trainingSee paper for methodsImpaired judgment of final note tuning in right temporal resection group versus (C)
Anterior temporal including HG (L)1835 (m)
Anterior temporal including HG (R)1837 (m)
 Zatorre (1985)Controls (C)2032 (m)UnknownSix tonal melodies altered by contour or scaleImpaired discrimination of pitch direction in right temporal lobectomy including HG versus (L) and (C)
Anterior temporal including HG (L)2827 (m)
Anterior temporal including HG (R)3029 (m)
 Zatorre (1988)Controls (C)1825 (m)UnknownPairs of tones with altered pitch directionImpaired discrimination of pitch direction in right temporal lobectomy including HG
Anterior temporal sparing HG (L)1527 (m)
Anterior temporal sparing HG (R)1525 (m)
Anterior temporal including HG (L)1630 (m)
Anterior temporal including HG (R)1831 (m)
 Zatorre and Halpern (1993)Controls (C)1431 (m)A few participants in each group with musical trainingSeven songs used to assess pitch heightPitch height of two lyrics within familiar songs impaired in (R) versus (C)
Anterior temporal (L)1433 (m)
Anterior temporal including HG (R)1432 (m)
 Zatorre and Samson (1991)Controls (C)1827 (m)No professional musicians includedPitch discrimination assessed by target tone variation and distracter tonesImpaired discrimination of pitch difference between two tones with intervening distracter tones in right temporal lobectomy group
Anterior temporal including HG (L)2628 (m)
Anterior temporal including HG (R)2629 (m)
Frontal (R)1331 (m)
Frontotemporal (R)726 (m)

Studies investigating the effect of surgery on musical emotion report conflicting results. Some studies report impaired recognition of scary music following anteromedial temporal lobe resection (Khalfa et al., 2008; Gosselin et al., 2011), whereas another study found that patients who underwent unilateral medial temporal lobe surgery were able to judge emotional dissimilarities in music as well as healthy participants (Dellacherie et al., 2011). These discrepancies may relate to the amount of surgical resection, the methodologic differences in assessing emotional responses, or the cognitive processes underlying the emotional responses.

Studies examining musical recognition broadly agree and report impairment in recognizing previously presented tunes following temporal lobe surgery but with a right-sided predominance (Shankweiler, 1966; Samson & Zatorre, 1991, 1992). The outcomes suggest a bitemporal role in musical recognition and possibly for emotion as well.

Studies examining pitch processing report impairments in pitch direction, pitch pattern, timbre, and tonal structure in patients who have undergone right temporal lobectomy including Heschl’s gyrus (Milner, 1962; Zatorre, 1985; Samson & Zatorre, 1988; Zatorre, 1988; Zatorre & Samson, 1991; Kester et al., 1991; Zatorre & Halpern, 1993; Samson & Zatorre, 1994; Johnsrude et al., 2000; Warrier & Zatorre, 2004). Similarly, studies exploring the effect on rhythm and time intervals show impairment in patients who have undergone right temporal lobectomy, including the superior temporal gyrus (Milner, 1962; Liegeois-Chauvel et al., 1998). It appears that patients undergoing right temporal lobectomy are more at risk than those undergoing left temporal lobectomy in acquiring postoperative problems with musical perception.

Preoperative and Postoperative Evaluation of Musical Functioning in Patients with Epilepsy

  1. Top of page
  2. Summary
  3. Musical Processing and the Human Brain
  4. Music and Epilepsy
  5. Music Therapy for Epilepsy
  6. Mechanisms for the Proconvulsant and Anticonvulsant Effects of Music in Epilepsy
  7. The Effect of Antiepileptic Therapies on Musicality
  8. Preoperative and Postoperative Evaluation of Musical Functioning in Patients with Epilepsy
  9. Conclusions
  10. Acknowledgments
  11. Disclosure
  12. References

Published studies suggest a bitemporal role for musical recognition and possibly emotion, but a right temporal role for musical perception. The potential risks to musical processing should be discussed with patients, particularly those undergoing right temporal lobectomy. It would be interesting to ascertain current practices in discussing musical implications in neurosurgical clinics.

Preoperative neuropsychological evaluation of musical processing should also be considered in temporal lobe resective patients. There have been many tests devised over the years, the detail of which is beyond the scope of this review. However, two internationally recognized assessment tools are mentioned briefly here.

The Seashore Measures of Musical Talent was a musical aptitude test developed in 1919 and modified in 1960 to assess musical abilities in preschool children (Seashore et al., 1960). The test consists of six measures: pitch, loudness, rhythm, time, timbre, and tonal memory. The test requires a person to compare a pair or short series of tones produced by a beat frequency oscillator. This test is based on the psychophysical measures of auditory perception with little recognizable musical content and has been reported to be sensitive to right temporal lobe dysfunction (Milner, 1962). However, the test has been criticized for being “atomistic” and “unmusically orientated” (Mursell, 1937). Other authors have criticized its localizing value (Steinmeyer, 1984; Boone & Rausch, 1989). The tool was also used to assess professional musicians. The musicians scored higher on only three of the six measures compared to population scores, suggesting the Seashore assumptions about the requirements for high musical talent may be incorrect (Henson & Wyke, 1984). These more recent studies cast uncertainty on the validity of this test in the preoperative setting.

The Montreal Battery for the Evaluation of Amusia (Peretz et al., 2003) is a more recent test that uses musical material as the stimulus. The test assesses melodic and temporal processing of musical perception and memory via six tests on contour, interval, scale, meter, and memory using thirty musical phrases. The musical excerpts can be found at http://www.brams.umontreal.ca/plab/research. The test has a high sensitivity even on retesting at 4 months. The test correlated well with another widely used aptitude test, Gordon’s musical aptitude profile (Gordon, 1965). The authors stated that the advantage of this tool over the Seashore measures or Gordon’s musical aptitude profile was that having been more recently devised it better reflected the current concepts of musical perception and memory. The test was more likely to pinpoint specific deficiencies in components of musical perception compared to Gordon’s musical aptitude profile, which was designed to assess musical aptitude in children and tests structural components all at once. The test better reflected melody discrimination and ordinary musical ability compared to the Seashore measures. This tool therefore may be more suited compared to previously devised tools in diagnostic work looking at baseline musical functioning and evaluation of deficits. The website above has a tool that can be used to evaluate musical emotion. In summary this test may be more suitable in the presurgical evaluation setting, although further work is required in this area.

Further research is needed into the neuropsychological consequences of temporal lobe surgery on musical processing. This could be achieved by collaboration of professionals within the cognitive sciences and neurosurgeons specializing in epilepsy surgery. The trend for referring increasing numbers of refractory epilepsy patients for surgery could potentially expose increasing numbers of individuals to this problem. Understanding the consequences of surgery on musical processing is therefore paramount.

Conclusions

  1. Top of page
  2. Summary
  3. Musical Processing and the Human Brain
  4. Music and Epilepsy
  5. Music Therapy for Epilepsy
  6. Mechanisms for the Proconvulsant and Anticonvulsant Effects of Music in Epilepsy
  7. The Effect of Antiepileptic Therapies on Musicality
  8. Preoperative and Postoperative Evaluation of Musical Functioning in Patients with Epilepsy
  9. Conclusions
  10. Acknowledgments
  11. Disclosure
  12. References

The dichotomous proconvulsant and anticonvulsant role of music in epilepsy is a fascinating and at present poorly understood phenomenon. Dopaminergic system modulation may provide one possible explanation, but this theory needs further exploration. It is clear from clinical studies that musical perception occurs via a complex network, which if recruited or hyperexcitable may manifest as seizures. What makes a particular sound or sequence of sounds epileptogenic may be explained by the stimulation to varying effect of different areas involved within the music network. Musicogenic epilepsy may be more common than we suspect as routine EEG does not use sound in seizure provocation, and there may be a delay in linking a particular musical trigger to seizures.

There is limited evidence of an anticonvulsant effect with music in epilepsy. The therapeutic potential remains for the moment largely unknown and unproven and our understanding of the brain mechanisms involved are very limited.

Altered pitch perception in relation to carbamazepine is rare, but health care professionals should discuss this risk or consider alternative medication, particularly if the patient is a professional musician or native-born Japanese. The effect of epilepsy surgery on musicality is a potential and significant issue especially with right temporal lobectomy, although the extent of this risk and any correlation to size and area of resection need further delineation. This potential risk may bring into question whether musical perception, and memory testing should form part of the preoperative neuropsychological workup for patients embarking on surgery, particularly on the right temporal lobe.

Acknowledgments

  1. Top of page
  2. Summary
  3. Musical Processing and the Human Brain
  4. Music and Epilepsy
  5. Music Therapy for Epilepsy
  6. Mechanisms for the Proconvulsant and Anticonvulsant Effects of Music in Epilepsy
  7. The Effect of Antiepileptic Therapies on Musicality
  8. Preoperative and Postoperative Evaluation of Musical Functioning in Patients with Epilepsy
  9. Conclusions
  10. Acknowledgments
  11. Disclosure
  12. References

I would like to thank Dr Michael Johnson, Consultant Neurologist at Leeds General Infirmary for his kind donation of Critchley & Henson’s Music and the Brain, a fascinating book for all who are working within the neurologic sciences.

References

  1. Top of page
  2. Summary
  3. Musical Processing and the Human Brain
  4. Music and Epilepsy
  5. Music Therapy for Epilepsy
  6. Mechanisms for the Proconvulsant and Anticonvulsant Effects of Music in Epilepsy
  7. The Effect of Antiepileptic Therapies on Musicality
  8. Preoperative and Postoperative Evaluation of Musical Functioning in Patients with Epilepsy
  9. Conclusions
  10. Acknowledgments
  11. Disclosure
  12. References