Article Text

Download PDFPDF

Panic and fear induced by deep brain stimulation
  1. N A Shapira1,
  2. M S Okun2,
  3. D Wint5,
  4. K D Foote3,
  5. J A Byars1,
  6. D Bowers4,
  7. U S Springer4,
  8. P J Lang4,
  9. B D Greenberg6,
  10. S N Haber7,
  11. W K Goodman1
  1. 1Department of Psychiatry, University of Florida, Gainesville, Florida, USA
  2. 2Departments of Neurology and Neurosurgery and Psychiatry, University of Florida, Gainesville
  3. 3Department of Neurosurgery, University of Florida, Gainesville
  4. 4Department of Clinical and Health Psychology, University of Florida, Gainesville
  5. 5Section on Integrative Neuroimaging, National Institute of Mental Health, Bethesda, Maryland, USA
  6. 6Department of Psychiatry and Human Behavior, Brown University, Providence, Rhode Island, USA
  7. 7Department of Pharmacology and Physiology, University of Rochester, New York, USA
  1. Correspondence to:
 Dr Michael S Okun
 University of Florida Brain Institute, Department of Neurology, 100 S Newell Drive, Room L3-100, Gainesville, FL 32610, USA; okun{at}


  • Competing interests: none declared

View Full Text

Statistics from

The neural circuitry underlying panic and fear is a topic of active investigation.1–5 Some proposed mechanisms have focused on the amygdala, believed to carry out emotional processing; the frontal cortex, which may modulate or inhibit panic; and the hypothalamus, which may mediate the relation between subjective panic and the autonomic system.1,2,5 There is a paucity of data examining the role of circuits involving the nucleus accumbens region in producing panic. In this report we discuss the phenomenon of de novo panic induced by both right and left deep brain stimulation (DBS) in the region of nucleus accumbens.


A 52 year old right handed man with a 32 year history of severe treatment refractory and crippling obsessive-compulsive disorder (OCD) presented for a surgical evaluation. He had obsessions (symmetry and exactness) and compulsions (ritualised handwashing and checking). He also suffered from symptoms of “incompletion” which compelled him to repeat behaviours until they felt “right”. He did not meet criteria for other anxiety disorders and had never experienced a panic attack. He had poor responses to multiple appropriate drugs and three courses of intensive behavioural therapy. At initial presentation, he also met DSM-IV diagnostic criteria for moderate recurrent major depressive disorder with melancholic features and was on escitalopram 30 mg daily. Neurological examination was normal.

The patient understood the risks and benefits of DBS and signed informed consent to participate in a DBS protocol for treatment refractory OCD (NIMH R21-MH64161). Using MRI/CT fusion based stereotactic guidance, DBS electrodes were inserted bilaterally through the anterior limb of the internal capsule into the nucleus accumbens near the anterior commissure (fig 1). Computed tomographic/magnetic resonance imaging (CT/MRI) fusion was undertaken for electrode localisation using a method previously published by us.6

Figure 1

 Anatomical localisation of the DBS electrode. The views presented include (1) axial, (2) sagittal, (3) coronal, and (4) a non-orthogonal view along the cross section of the electrode. Note that on both sides the deepest contact lies within the nucleus accumbens, and the more superficial contacts span the anterior limb of the internal capsule region. The mid-commissural point (MCP) was used as the centre of an orthogonal Cartesian coordinate system for all spatial relations. Computed tomographic images were fused to preoperative high resolution magnetic resonance images in order to definitively localise the electrode positions. The measured Cartesian coordinates of the distal end of the deepest contact relative to the mid-commissural point (MCP) were as follows. Left: 6.1 mm lateral to midline, 12.4 mm anterior to MCP, 3.5 mm inferior to the anterior commisure–posterior commissure (AC–PC) plane. Right: 4.5 mm lateral to midline, 11.3 mm anterior to MCP, 4.5 mm inferior to the AC–PC plane (AC–PC line is 25.1). The left electrode was placed at the posterior border of AC, and the right slightly posterior to AC. Each DBS electrode (3387 IES) contained four 3 mm contacts, spaced 4 mm apart. The length of the AC–PC line was 25 mm.

Brief stimulation paradigms were used to evaluate effects of right and left DBS in the region of nucleus accumbens and internal capsule. The intraoperative evaluation of the DBS electrodes was carried out using bipolar stimulation at each contact. Pulse width and stimulation frequency were fixed at 210 μs and 135 Hz, respectively. Voltage was varied between 0, 2, 4, 6, and 8 volts. The patient was blind to parameter changes, and the voltage was returned to 0 before testing at each contact. Testing occurred for approximately two to four minutes at each setting.

The results of intraoperative macrostimulation on the right are summarised in table 1. Most notably, the patient experienced reproducible sensations of flushing, fear, and “panic” when the most ventral contact (0) was stimulated (with bipolar stimulation). He described the sensation as feeling hot, flushed, fearful, and “panicky.” He could feel palpitations in his chest, and when asked indicated he had an impending sense of doom. The feelings were coincident and continuous with the stimulator “on” setting and they rapidly dissipated when switched off. With stimulation on the right, the feeling of panic was accompanied by tachycardia (heart rate increase from 53 to 119 beats/min when “on”, returning to 74 beats/min after stimulation was discontinued). The patient experienced a similar transient increase in his pulse (from 69 to 82 beats/min, subsequently returning to 67 beats/min) when contact 0 was activated on the left side, despite administration of a β blocker one hour before stimulation (three doses of 20 mg labetolol intravenously over 30 minutes). The β blocker was given by the anaesthesia team following right DBS as prophylaxis against tachycardia during intraoperative stimulation on the left side.

Table 1

 Results of right intraoperative deep brain stimulation testing

Approximately one month after implantation, the patient underwent sham testing of his DBS. He was asked to describe changes in emotional state in response to the DBS; no current was applied to any electrodes during this session. The patient was blinded to DBS settings by a curtain that separated the patient and examiner from the DBS programming nurse. The examiner communicated with the programmer by displaying written instructions. The patient was asked at regular intervals how he felt, regardless of whether stimulation parameters had changed. During this sham testing, the patient denied any anxiety or other significant changes in his emotional state. A planned second outpatient testing session confirmed the panic/fear response at the ventral contact.

The patient’s current stimulator settings are with contacts 1, 2, and 3 (anode), case (cathode), pulse width 210 μs, 130 Hz, 3.5 volts (18 months post-implantation). He has not had any panic with chronic DBS. He is currently taking 10 mg of escitalopram daily.


The findings in this case report show that stimulation of a circumscribed area in the region of the nucleus accumbens in an OCD patient can result in panic. This response may provide a glimpse of the potential brain circuitry underlying this phenomenon. The finding that the ventral contact elicited the strongest response may provide useful information about the potential circuits responsible for fear and panic. It is likely that within that circuitry the hypothalamic or autonomic fibres, or both, in this region may mediate the increase in heart rate, flushing, and heat sensation. The region of the anterior hypothalamus is very close to the active contact. Further, the main inputs to the accumbens (or ventral striatum) are the anterior cingulate and orbitofrontal cortices, the amygdala, and the midbrain dopamine neurones.7–9 The findings in this paper should be interpreted with caution as the patient involved in the study had OCD, and therefore we cannot be certain if they are generalisable. Additionally, an important consideration is that response of panic could have been secondary to physiological changes induced by stimulation, and this point will need to be clarified in future studies.

Amygdalofugal and hypothalamic pathways, as well as amygdalar connections with the locus ceruleus, and the periaqueductal grey matter, thalamus, and frontal cortex, may produce hormonal, autonomic, and behavioural changes characteristic of panic, although this circuitry also awaits further elucidation.10 Other triggers that can activate the panic response include fear provoking memories from the hippocampus1 and projections from the noradrenergic locus ceruleus.11,12 The locus ceruleus, the periaqueductal grey matter, and the nucleus paragigantocellularis project to the amygdala and may produce panic, possibly when stimulated by autonomic disturbances.1 The amygdala, orbitofrontal, and anterior cingulate circuit has been shown to be involved in the fear response. Interestingly, there is a recent report of panic with autonomic disturbances occurring in a patient who was receiving hypothalamic DBS for cluster headaches,13 and also for DBS in the subthalamic nucleus region.14

We believe it plausible that in this case DBS activated accumbens region outflow and that this led to the dramatic clinical effect of panic. This region has many potential structures that may play a role in this response, including accumbens, internal capsule, amygdala/stria terminalis, anterior hypothalamus, and medial olfactory stria. We have now observed this response in three patients implanted at this target. The mechanisms of DBS are complex15,16 and it is likely that both inhibition and excitation of cells and fibre bundles may play a role in the panic response.

Recent data indicate that subthalamic nucleus (STN) DBS can both impair fear recognition17 and induce fear and panic.14 Although in this case we stimulated in the nucleus accumbens circuitry, we may have activated the amygdala. The activation may have occurred in a manner similar to STN region DBS, suggesting that the panic response is the result of efferent fibres that may emanate from many different regions of brain, but that all connect to amygdala and ultimately frontal cortex.

The finding that stimulation of the nucleus accumbens/hypothalamic region can provoke panic may reveal a new aspect of the underlying neural circuitry of fear and panic. These observations will need to be followed with functional imaging to assist in clarifying our understanding of the neural circuitry underlying panic.


View Abstract


  • Competing interests: none declared

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.