Patients with obsessive-compulsive disorder (OCD) may present with a wide range of different obsessions and compulsions. Although checking and washing compulsions are the most frequent,1 a variety of less common symptoms have been described, including intrusive aggressive, sexual, and blasphemous thoughts and images.
Despite the heterogeneity of OCD symptoms, researchers have found at least some evidence that specific neurobiological dysfunctions mediate the disorder. The current literature emphasizes the role of frontal—basal ganglia circuits2 and of the serotonin neurotransmitter system.3
One of the possibly less common symptoms in OCD is that of intrusive musical tunes. In this report we present 2 cases in which this symptom was the presenting problem and in which functional brain imaging was obtained. We also briefly discuss the relevant literature on obsessive musical tunes and possibly related phenomena such as musical hallucinations and musicogenic epilepsy.
Case 1. Ms. A. is a 59-year-old woman who developed obsessive musical tunes in the context of a major depression following her husband's death 20 years previously. Her depression responded to treatment with an antidepressant, but the tunes continued.
She described the tunes as intrusive and unwanted, and she attempted to get rid of them by substituting other thoughts or tunes in her mind. The tunes led to significant interference in her life, disrupting her ability to work and to socialize.
[99mTc]HMPAO SPECT of the brain was performed. Prior to injection of the radiopharmaceutical, the patient was asked to bring her musical tunes to mind, as if to mimic the state in which she usually had these obsessions. SPECT demonstrated decreased flow in the left anteromedial and medial temporal and right anteromedial lobes, with smaller perfusion defects also in the left and right anterolateral frontal lobes and parieto-occipital cortex. Cerebellar blood flow was asymmetric, with lower flow on the left compared with the right side. Electroencephalography was normal.
The patient did not respond to adequate trials of several serotonin reuptake inhibitors (SRIs), including fluoxetine, paroxetine, citalopram, and clomipramine. She unfortunately also failed to respond to risperidone or gabapentin augmentation of an SRI and was unable to tolerate therapeutic doses of carbamazepine or valproate.
Case 2. Ms. B. is a 29-year-old woman who sustained a closed head injury in a motor vehicle accident and whose case has been described elsewhere.4 Loss of consciousness was present for 4 days and posttraumatic amnesia for 1 month. However, a computed tomography scan of the brain was normal.
After the accident, personality changes manifested, including aggressiveness and apathy. At that point, there was no evidence of mood or anxiety symptoms. The patient showed little insight into these changes and maintained that she had not sustained brain damage.
However, 6 months after the head injury, Ms. B. presented with a 1-month history of an intrusive musical tune that constantly "played" in her mind. She found the tune intrusive and irritating, and she attempted to suppress it by occupying her mind with other thoughts.
SPECT scan of the brain with the patient at rest showed decreased blood flow in the left temporal lobe and right anteromedial temporal lobe and decreased blood flow in both frontal lobes (left greater than right), particularly in the inferior posterolateral frontal cortices bilaterally.
Medication treatment was discussed with the patient, but to date she has declined this intervention.
Both of our patients presented with musical obsessions. Even though one patient appeared to have idiopathic OCD and one had acquired OCD, the phenomenology of the symptoms was remarkably similar. Musical hallucinations by definition have "the compelling sense of reality of a true perception"5 whose source is presumably often experienced as outside the head; however, in our patients the music was experienced as an internally generated cognitive product that was nevertheless intrusive and inappropriate (analogous to an obsessive visual image). On SPECT scanning, both patients demonstrated prominent decreases of blood flow in the temporal lobes as well as frontal perfusion defects.
The majority of work on functional brain imaging in OCD has emphasized frontal—basal ganglia circuitry. Thus, a recent review of this area notes that neutral-state studies of OCD most consistently implicate hyperactivity in prefrontal cortex, with some studies demonstrating cingulate and striatal involvement; that pre-/post-treatment studies show reduced activity of orbitofrontal cortex, caudate nucleus, or cingulate cortex after effective treatment regardless of modality; and that symptom provocation studies show activation of the ventral corticostriatal system (anterior/lateral orbitofrontal cortex and caudate) and of a paralimbic belt (posterior orbitofrontal cortex and anterior cingulate).2 Similarly, a previous case report notes repetitive musical intrusions secondary to basal ganglia pathology.6
Nevertheless, there is also a literature on temporal lobe involvement in OCD. Jenike and Brotman,7 for example, described temporal EEG abnormalities in a subset of OCD patients and reviewed the literature on obsessive-compulsive symptoms in the context of seizure disorder. Similarly, in their review of OCD secondary to frontal lesions, Ames et al.8 note the presence of temporal lesions in many patients. Furthermore, Berthier et al.9 found a variety of lesions on magnetic resonance imaging in acquired OCD, including temporal involvement in 6 of 13 patients. Interestingly, patients with Tourette's syndrome, which may overlap neurobiologically with OCD, demonstrate abnormal brain activation on music perception in temporal and parietal regions.10
From an extensive case series and literature review of musical hallucinations, Berrios11 concluded that these were frequently associated with neurological lesions (tumors, seizures, stroke) in the temporal lobe, particularly the right temporal lobe. Similarly, Wieser et al.12 recently reviewed the literature on seizures triggered by music (musicogenic epilepsy) and concluded that this is strongly correlated with the temporal lobe, again predominantly right-sided. These reviews are consistent with many studies on the brain processing of music, ranging from an early experiment by Penfield and Perot13 demonstrating that stimulation of the superior temporal lobe could elicit the perception of music, to more recent neuropsychological14 and neuropsychiatric15 studies.
There are few reports of functional imaging in patients with musical hallucinations or seizures. Erkwoh et al.16 reported one patient with increased metabolic activity in superior temporal regions (left greater than right). However, a number of functional imaging studies of patients with auditory hallucinations in schizophrenia have been published.17 There is some evidence that auditory hallucinations may activate the right temporal cortex in these patients.17
It is interesting to note that SSRIs were effective for depression but not for musical obsessions in our first patient. There are, however, previous case reports of obsessive musical symptoms in idiopathic18 and acquired19 OCD responding to clomipramine. Our patient unfortunately was unable to tolerate carbamazepine, a medication that may be useful in OCD patients with possible temporal lobe dysfunction20 and in musical hallucinations secondary to temporal lobe abnormalities.21
Much remains to be understood about obsessive musical symptoms. These symptoms deserve additional attention from both clinicians and researchers. The current study suggests that musical obsessions may be mediated by temporal areas in addition to other areas believed to be involved in OCD. Certainly these areas of the brain are thought to be involved in normal musical processing and in musical and auditory hallucinations.
This study was supported by the Medical Research Council Research Unit on Anxiety and Stress Disorders (South Africa).