"Ms. A," a 55 year-old woman, was seen with a history of severe major depression. She had a family history significant for major depression in her brother, father, and paternal grandfather. Her own depressive history began at age 9 and continued intermittently during her adolescence. The condition became more severe at age 17, and she continued to have fluctuating symptoms from that time forward. The diagnosis of depression was made by her referring psychiatrist, based on DSM clinical criteria (Structured Clinical Interview for DSM-IV [SCID, 9]: major depressive disorder, chronic, severe, without psychotic features). Her psychiatrist had been treating her since 1981. Upon referral to our center, she underwent a standardized diagnostic interview in the form of the SCID. Her condition had been characterized by feelings of misery and by somatic complaints of low energy, fatigability, and muscle tension in the context of her depression. She was not determined to have a somatization disorder. She had a remission taking bupropion from ages 36 to 48, at which time she worked as a letter carrier for the postal service. From that time forward, she experienced severe depression, which had not remitted for the 7 years prior to her referral. She was incapacitated by her illness and unable to work for that time. She had been treated and had limited clinical benefit from trials adequate in dose and duration of each of the following antidepressants: nomifensine, maprotiline, amitriptyline, clomipramine, desipramine, trimipramine, trazodone, phenelzine, moclobemide, nefazodone, citalopram, escitalopram, fluoxetine, fluvoxamine, and bupropion. She had also been treated with trials of adequate dose and duration of the following augmentation and adjunctive strategies: lithium carbonate, l-tryptophan, tri-iodothyronine, methylphenidate, buspirone, amoxapine, flupenthixol, carbamazepine, topiramate, and valproic acid. All had failed to provide any sustained benefit. Ms. A had also tried venlafaxine and pramipexole but was unable to tolerate these medications before conventional therapeutic doses were reached. Additionally, she had undergone a course of eight bilateral ECT treatments without benefit.

Ms. A was referred for surgical management of her depression. Her 17-item Hamilton Depression Rating Scale (17-item HAM-D) score at that time was 22. Her case was reviewed by a multidisciplinary committee, which meets regularly for the express purpose of overseeing the surgical management of psychiatric disease. The conference includes individuals from neurosurgery, psychiatry, neurology, and neuropsychology at the University of Toronto. This group reviews and discusses the merits of each case being considered for surgical treatment of psychiatric disease. Ms. A was deemed a good candidate for either cingulotomy or for participation in a novel trial of deep brain stimulation of the subgenual cingulate. She was offered these options. After discussion with her original psychiatrist, she declined the untested deep brain stimulation therapy in favor of a lesional cingulotomy. This procedure was approved by the same committee.

Ms. A underwent a bilateral stereotactic ablative cingulotomy. The procedure, which is described below, was performed without complication. Afterward she reported subjective improvement in her depression and improvements in her symptoms of fatigue and hypersomnia. She stated that she felt better than she had in years. Her 17-item HAM-D score improved to 8 one month after surgery and further to 5 when she was tested 6 weeks after surgery. In the weeks following her surgery, her psychiatrist was able to decrease her bupropion dose considerably. Ms. A remained depression free for a period of 6 months with a sustained 17-item HAM-D score of 5. She gradually noted a worsening of her depression, both in her symptoms of melancholia and in her fatigue. Her antidepressant doses, previously reduced, were restored to her presurgery doses. Despite this, her 17-item HAM-D score increased to 9 that month, 15 the next month, and was 19 two months later. A magnetic resonance imaging (MRI) scan was obtained to review the extent of her prior cingulotomies. The lesions were felt to be appropriate in size and location. The option of a repeat cingulotomy was considered less promising for this reason.

Ms. A was again presented at the multidisciplinary case conference. At this time, our center had just completed the pilot study assessing the benefit of deep brain stimulation of the Cg25 region for refractory depression. It was felt that this procedure might offer a greater probability of benefit to her refractory condition. Ms. A was sent for additional neuropsychiatric testing to ensure that her prior cingulotomy had not diminished her cognitive reserve in a way that might confound interpretation of the effects of the deep brain stimulation procedure. After an extensive discussion with her psychiatrist, the conference was convened and recommended that the patient be offered deep brain stimulation surgery to stimulate the subgenual cingulate. The option of vagal nerve stimulation was not seriously considered for several reasons. The definitive blind study on vagal nerve stimulation for depression was not published until September 2005, 4 months after Ms. A underwent deep brain stimulation surgery (10, 11). The procedure did not receive Food and Drug Administration approval until July 2005, and we did not have an ongoing vagal nerve stimulation trial at our center at the time Ms. A was treated. The initial open study of vagal nerve stimulation therapy showed that ECT response predicted vagal nerve stimulation outcome (12). Ms. A had failed to respond to ECT so, based on the best available evidence, we would have been unlikely to recommend vagal nerve stimulation therapy had it been available.

The deep brain stimulation procedure was scheduled, based on compassionate use of the new procedure. Just over a year after Ms. A’s cingulotomy, placement of deep brain stimulators in the subgenual cingulate was performed. Her 17-item HAM-D score just before surgery was 19 and had been stable for 4 months. The surgical procedure is described in greater detail below. Intraoperative assessment was conducted as done with the deep brain stimulation pilot study with patient self-report, analogue mood rating, and changes in the Positive and Negative Affect Schedule (13) to assess the effect of stimulation at the various contacts of the deep brain stimulation electrode. Notably, during intraoperative stimulation, Ms. A reported alleviation of the ever-present sensation of heaviness described as a surrounding "black cloud." She also expressed immediate interest in returning to activities avoided since her relapse. This effect was contact-specific and reversible.

Stimulation was begun on the day after surgery. Initial screens with the Positive and Negative Affect Schedule were performed to determine the most effective contacts for stimulation. Parameters of stimulation were a rate of 130 Hz, a pulse width of 60, and an amplitude of 4.5 V. Slight adjustments were made over the next several months to maximize Ms. A’s response. At the 1-month evaluation, she reported a subjective improvement in her mood. Her 17-item HAM-D score had decreased to 11 when she was tested 3 months after surgery. Her level of activity increased, and she was able to resume volunteer work, similar to what had occurred after her cingulotomy. Subsequent 17-item HAM-D scores were 8 at 6 months and 7 at 1 year. She continues to have a sustained remission to ongoing Cg25 stimulation, with a 17-item HAM-D score of 7 at 30 months after her deep brain stimulation surgery (Figure 1).

Since her deep brain stimulation procedure, Ms. A reports marked and sustained amelioration of the following symptoms: suicidal ideation, depressed mood, guilty ruminations, appetite, anxiety symptoms, and initiating and maintaining sleep. Lesser improvements have been reported in energy, libido, and concentration. She has also been noted on assessment to have a bright affect and an increased range of emotional expression. She is no longer demonstrating signs of psychomotor retardation. No signs or symptoms of mania or psychosis are present. She continues to work full-time 6 days a week as a volunteer driving patients to and from a local hospital. She has not missed a day since her return after deep brain stimulation surgery. The relative degree and longevity of benefit that the patient received from the cingulotomy and deep brain stimulation procedures are compared in Table 1.

A Leksell G stereotactic head frame (Elekta, Inc., Atlanta) was applied under local anesthesia and an MRI was obtained. Images were fused in an imaging workstation (FrameLink 4.1 software, StealthStation, Medtronic SNT, Minneapolis) and used for the stereotactic targeting. A coronal slice 2 cm to 2.5 cm posterior to the anterior tip of the frontal horn was identified, and a point was selected in the center (in the coronal plane) of the cingulate gyrus on this image. The stereotactic coordinates of this point were calculated, and microelectrodes were passed on a trajectory through this point.

The details of the microelectrode recording mapping procedures and equipment used in our service have been previously published (4, 14). Microelectrode recording mapping procedures were used to delineate cell-rich areas interpreted as the gray matter of the upper and lower banks of the cingulate gyrus. Areas devoid of somatodendritic action potentials were electrophysiologically silent and corresponded to the cingulate bundle and the corpus callosum. After verifying the depth of the corpus callosum, the microelectrodes were removed, and a radiofrequency electrode (1.2 mm wide with a 5-mm exposed tip) (Owl Universal RF system, Diros Technology, Toronto) was inserted just above the corpus callosum. Test stimulation (RF Generator/Stimulator Model URF-1 or RFS-1, Diros Technology) was performed at 100 Hz and up to 5 V through the macroelectrode. Because no adverse effects were noticed, three stacked lesions were created by applying radiofrequency current to the electrode to a temperature of approximately 90°C for 90 seconds. The electrode was withdrawn 5 mm, and after another test stimulation, the process was repeated with the same parameters. The electrode was withdrawn another 5 mm, and the procedure was repeated a third time. Then the electrode was withdrawn completely, and three lesions were created 1 cm anterior to this track with a similar process. The same process was repeated in the contralateral hemisphere (Figure 2).

The placement of the stereotactic frame and the imaging procedures for the insertion of deep brain stimulation electrodes were similar to that employed for cingulotomy. However, rather than targeting the cingulate gyrus above the body of the corpus callosum, we targeted the subgenual cingulum. For that, a line was traced from the most anterior aspect (genu) of the corpus callosum to the anterior commissure and the midpoint was selected (5). The coronal section correspondent to that plane was identified, and the coordinates of the transition between the gray and white matters of area 25 were calculated. Microelectrode recording mapping procedures were conducted in this case to identify the location of the area of transition between gray and white matter of area 25, characterized, respectively, by the recording of neuronal activity and cell-sparse areas. Thereafter, deep brain stimulation quadripolar electrodes (Medtronic 3387, Medtronic, Inc., Minneapolis) were implanted bilaterally such that the tip of the electrode array sat at the inferior gray matter-white matter junction, and the four contacts spanned the full gyrus. Each of the four electrode contacts was tested for adverse effects and clinical benefits. The electrodes were then connected to a pulse generator (Kinetra, Medtronic, Inc.) that was implanted in the infraclavicular region under general anesthesia.

After several adjustments, stimulation parameters were selected with monopolar stimulation from contacts 2 and 5 (third contact from the bottom on the right and second contact from the bottom on the left). The anterior commissure-posterior commissure coordinates of these contacts are the following—right side: x=5.1, y=33.7, z=—4.4; and left side: x=3.8, y=34.3, z=—4.9. Stimulation was performed at a rate of 130 Hz, a pulse width of 60 Hz, and an amplitude of 4.5 V. These settings have not been substantially changed during Ms. A’s subsequent therapy.