Capgras Syndrome in Advanced Parkinson’s Disease

Psychosis is common in Parkinson’s disease (PD), especially in advanced disease. Visual hallucinations are the most common manifestation of psychosis in PD, but a number of other psychotic symptoms can occur, including delusions. One type of delusion that may occur in PD is Capgras syndrome (CS), a delusional misidentification syndrome.¹ In CS, the affected individual believes that a closely related person (or sometimes inanimate object or animal) has been replaced by an impostor or duplicate. CS has been reported in a small number of PD patients with advanced disease.^1–8 There is one reported case of CS in PD occurring after deep brain stimulation (DBS) targeting the subthalamic nucleus.⁵ The aim of the present study was to report on three PD patients who developed CS following DBS surgery and to investigate the frequency of CS in PD patients who have undergone DBS.

Methods

Using a clinical database, we identified 115 PD patients who underwent unilateral or bilateral DBS targeting either the subthalamic nucleus (STN) or globus pallidus intera (GPi) between January 1, 2010, and December 31, 2015, at the University of Virginia Medical Center (Charlottesville, Va.). Only patients receiving a first neurosurgical procedure for PD were included. Patients with diagnoses of concomitant essential tremor and PD were excluded. Following approval by the University of Virginia Institutional Review Board for Health Sciences Research, we performed a retrospective review of the electronic medical record to collect clinical information from the presurgical evaluation, surgical inpatient admission, and postoperative follow-up visit notes.

To identify those with CS, we searched each subject’s record for at least the 6 months following DBS surgery using the following search terms: “Capgras,” “delusion,” psychosis,” “psychotic,” “imposter,” and “misidentif*” (for misidentify or misidentification). Clinical follow-up for most subjects following surgery was longer than 6 months. When search terms were present in a subject’s record, clinical documentation was reviewed to determine whether the subject experienced CS or had other forms of delusions. Detailed histories were collected for the three subjects identified to have CS following DBS surgery.

Results

Characteristics of the entire PD DBS population (N=115) and three subjects with CS are summarized in Table 1. As this was a clinical cohort, clinical follow-up after surgery varied. However, neither age at the time of surgery (Spearman’s correlation: p=0.68) nor sex (Mann-Whitney U test: p=0.79) were significantly associated with the duration of clinical follow-up. Case descriptions of the three subjects with CS are presented below. In addition to these three cases, there were three other patients who developed delusions after DBS placement. These were delusions related to infidelity, religion, and persecution.

Case 1

The patient is a right-handed woman with a history of depression and anxiety, with onset in childhood, who first noted difficulty with handwriting at age 50. Over the following year, she developed a rest tremor in her right-upper and lower extremities. At age 51, she was diagnosed with PD. Her motor symptoms were well controlled with oral medications for 8 years until she developed worsening dyskinesias. Over the subsequent year, she had progressive gait changes with imbalance. At age 61, she underwent neuropsychological testing in the context of subjective declines in working memory, short-term memory, and word finding. Around that time, she also started experiencing amotivation and instances of feeling like her deceased parents were in her house. Testing showed subtle declines in processing speed and visuospatial abilities relative to her estimated superior baseline intellect and a weakness in executive functioning. She was diagnosed with PD-mild cognitive impairment. Over the next year, she developed well-formed visual hallucinations of her deceased family members in her home, which were not distressful, as she retained insight. Repeat neuropsychological testing at age 62 prior to DBS surgery revealed modest declines in processing speed and attention but largely stable abilities. Her score on the Montreal Cognitive Assessment (MoCA) was 24/30. Her Unified Parkinson Disease Rating Scale (UPDRS) total score (parts I-III) off medications prior to surgery was 46. Her preoperative MRI brain was unremarkable.

The patient underwent bilateral GPi placement at age 63 and had improvement in her motor symptom control. At the time of surgery, she was taking one tablet of carbidopa/levodopa (25 mg/100 mg) four times daily and 1.5 tablets once daily. She was also taking carbidopa/levodopa controlled release (25 mg/100 mg) at bedtime and pramipexole (1 mg) three times daily. However, her postsurgical course was complicated by wound breakdown and infection, and the entire system was explanted 1 month later. MRI brain following explantation showed only minimal T₂-hyperintensity along the prior deep brain stimulator tracts.

At follow-up 5 months after explantation, the patient’s PD motor symptoms were stable, but she complained of worsening cognition, concentration, and difficulty sleeping. She also began experiencing periods when she was convinced that she had two identical husbands, one of whom was real and one who was a duplicate and not real. She would confront her husband to ask whether he was her real husband or not. A repeat neuropsychological evaluation in December 2014 revealed a stable MoCA score of 24/30 but significant declines in cognitive processing speed, letter fluency, and ability to encode and later recognize verbal information. At that time, she was taking one tablet of carbidopa/levodopa (25 mg/100 mg) five times daily and 1.5 tablets once daily. She was also taking carbidopa/levodopa controlled release (25 mg/100 mg) twice a day and pramipexole (0.5 mg) three times daily. She was started on quetiapine (12.5 mg) at bedtime, and her pramipexole dose was reduced, which led to resolution of her CS symptoms and hallucinations.

Case 2

A left-handed woman reported onset of right-hand tremor at age 58. She was diagnosed with PD and had an excellent response to levodopa. By age 62, she developed significant peak-dose dyskinesias and anxiety, for which she was prescribed clonazepam. She also developed symptoms of depressed mood, including amotivation and crying spells. Her sleep quality declined and was marked by episodes of shouting in her sleep. On neuropsychological testing prior to DBS surgery, her score on the Mini-Mental Status Examination was 28/30, but reductions in attention, with significantly impaired basic visual attention, modest difficulties encoding new information, and below-expectation problem-solving abilities were evident. She had placement of bilateral STN DBS at age 63. Prior to surgery, she was taking carbidopa/levodopa (25 mg/100 mg) five times daily and subcutaneous apomorphine as needed for unpredictable off-periods. Her preoperative UPDRS total score while off medication was 49. Her preoperative MRI brain scan was unremarkable, and her postoperative MRI showed only expected changes related to DBS placement.

At age 67, the patient was noted to have more confusion. She would repeat questions and occasionally ask her husband who he was. She subsequently had episodes when she did not recognize her husband as her “real husband.” She would refer to her husband in the third person when talking to him, as she did not recognize him as her “real husband.” She contended that her true husband was in their home or would return shortly. When her husband tried to confront her about this, she was dismissive. At that time, she was taking carbidopa/levodopa (25 mg/100 mg) five to six times per day. She was started on quetiapine (12.5 mg), and the misidentification improved, but she was noted to still have rare delusions of a similar nature when seen a year after onset of these symptoms.

Case 3

A left-handed man developed left-arm stiffness at age 61. Two years later, he presented for evaluation and was diagnosed with PD. He was started on levodopa and had significant improvement. His PD progressed over the next several years, with development of peak-dose dyskinesias. He was referred for DBS surgery. At that time, he described cognitive slowing, modest declines in multitasking ability and memory, and mild anxiety. On preoperative neuropsychological testing, his MoCA score was 23/30. More extensive testing showed cognitive slowing, reduced efficiency in his ability to encode new information, particularly visual information, and signs of executive dysfunction, including impaired set-switching, cognitive perseverations, and reduction in abstract reasoning. He was diagnosed with PD-mild cognitive impairment. His preoperative total UDPRS score while off medication was 61.

At age 71, he had bilateral STN DBS placed. At that time, he was taking one tablet of carbidopa/levodopa (25 mg/100 mg) four times daily and one tablet of carbidopa/levodopa controlled release (25 mg/100 mg) four times daily. He reported increased confusion after the surgery but also improvement in motor symptoms. His cognitive problems persisted, and brief neuropsychological testing 3 months after surgery showed significant cognitive deficits and declines in functioning, including impairments in cognitive processing speed and verbal fluency. His MoCA score declined to 13/30.

At follow-up 8 months after surgery, the patient reported continued worsening of his cognitive function and stated that he had two wives. After his wife made a change to her hairstyle, he began believing that she had an identical twin living in the house and that both had the same name and looked very similar. He reported feeling guilty that he had another woman in his life, and he called his wife and the identical twin “wife 1” and “wife 2.” At times he recognized one as his wife and the other as an imposter. At that time, he was taking one tablet of carbidopa/levodopa (25 mg/100 mg) four times daily and one tablet of carbidopa/levodopa controlled release (25 mg/100 mg) four times daily. His MRI brain scan at the time of surgery showed mild global atrophy with chronic microvascular ischemic changes. Also noted was a small left-parietal extra-axial lesion, like a calcified meningioma. Only typical postoperative changes were seen following DBS placement. For the delusions, he was started on quetiapine (25 mg) nightly, which, as observed during follow-up visits, only provided limited improvement.

Discussion

In our cohort of 115 advanced PD patients who underwent DBS for management of motor symptoms, we have reported on three case subjects who developed a delusional misidentification syndrome consistent with CS. As we do not have long-term follow-up data for all patients in this series and case subjects were ascertained retrospectively, this may have resulted in underestimation of the occurrence of CS in our cohort of PD patients who underwent DBS. CS in PD is certainly a rare phenomenon: a search of medical records for CS over an 11-year period for an entire health system uncovered 10 cases of CS in PD.⁹ Additionally, there are no studies that have assessed CS prospectively in PD. Thus, what is known about CS in PD comes from a number of case reports and chart reviews.^1–9

CS is known to occur more often in PD patients with dementia and other psychotic symptoms, such as visual hallucinations.⁹ All of our case subjects had cognitive decline, including two with mild cognitive impairment documented prior to DBS. All of our case subjects also had premorbid psychiatric symptoms, including anxiety and, in one case, visual hallucinations. In our three case subjects, the time period between DBS placement and CS onset was 6 months or longer. The delay between surgery and CS, combined with the fact that CS was observed following DBS targeting both the STN and GPi and that simple medication changes ameliorated the symptoms of CS, argues against any lesional or stimulation-induced effect. Moreover, in one of our patients, CS emerged months after removal of the DBS electrodes. From the data available, the age at surgery and duration of disease do not seem to influence the development of CS.

In our three case subjects, delusional misidentification was accompanied by duplication. This phenomenon has previously been reported in PD and may be a common characteristic of CS in PD.^5–7 This feature was noted in 1923 in the seminal paper by Joseph Capgras when he first described a 53-year-old woman with a 10-year history of a delusion of misidentification.¹⁰ In describing her condition, the patient herself stated that, “Doubles are people who resemble each other.” The patient expressed that her own husband and daughter, among others, were imposters or doubles due to the belief that the true persons had died or had been abducted, respectively. She was unable to correctly identify her family; however, the true persons did in fact exist.

Others have also explored the potential association between DBS and psychosis, given case reports of its occurrence.¹¹ However, a prospective case series in older patients, although showing an increased likelihood to develop psychosis following DBS, found a similar rate of incident psychosis in each of the 5 years following DBS and no difference in psychosis risk based on electrode depth or stimulation parameters.¹² Therefore, these findings are consistent with the idea that DBS itself does not increase risk for CS.

The etiology of CS and its relation to PD remains unclear. CS has been associated with lesions of the right hemisphere, bilateral frontal lobes, and temporal lobes. These lesions may contribute to CS by disconnecting limbic circuits associated with emotional context from areas of the occipito-temporal cortex involved in facial recognition.¹³ Disruptions in fronto-temporal or fronto-limbic pathways have also been proposed to cause CS. Therefore, patients with PD-associated frontal atrophy and impaired visuospatial processing may be at increased risk for development of CS.⁷

A limitation of this study is the lack of a non-DBS control PD population. Thus, we are unable to state definitively whether the three patients represent an increase in occurrence of CS compared with a similarly advanced PD cohort without DBS. Another limitation is how CS was assessed. This retrospective cohort study relied on description of symptoms present in clinical notes. If a validated measure of CS were available, prospective assessment of CS would be preferable. Nevertheless, even in advanced PD patients undergoing DBS, subsequent development of CS is a rare occurrence and does not appear to be associated with DBS placement. However, given the distressing nature of the condition, patients with advanced PD who undergo DBS should be regularly screened for symptoms of psychosis with awareness of CS as a potential form.