Surgery for Psychiatric Disorders

Neurosurg Clin N Am 14 (2003) xiii–xiv Preface Surgery for psychiatric disorders

Ali R. Rezai, MD Steven A. Rasmussen, MD Benjamin D. Greenberg, MD, PhD Guest Editors

Surgery for psychiatric disorders has intrigued negative effects on marital, parental, social, voca- the psychiatric, neurological, and neurosurgical tional, and other life functions. communities as well as the lay public for the past Advances in the efficacy, safety, and tolerabi- 70 years. Few other therapies have generated such lity of treatments for OCD and depression have controversy, enthusiasm, misconception, and, at been made in the past 20 years. However, a signif- times, indiscriminate use. Recent advances in image icant number of patients with OCD and major guidance, devices, and stereotactic procedures, as depression refractory to nonsurgical treatments well as rapid expansion of our knowledge of the remain severely ill. These individuals, who are liv- functional neuroanatomy of the major psychiatric ing lives of hopeless desperation, may ultimately disorders has led to renewed interest in neurosurgi- face the tragic end of suicide. The hope that sur- cal approaches to these conditions. This issue of gery can offer relief to these patients from agony the Neurosurgery Clinics of North America was and suffering is the fundamental imperative that envisioned with the purpose of compiling the demands ongoing research in this area. perspectives of experts on the cutting edge of It is clear that psychiatric neurosurgery research in the neurosurgical treatment of psychi- will undergo a rapid evolution over the next atric disorders. decade. Today, dedicated multidisciplinary teams Currently, the two most common psychiatric specializing in treating psychiatric patients have disorders amenable to surgical intervention are at their disposal the entire spectrum of modern obsessive-compulsive disorder (OCD) and major functional neurosurgical techniques, including depression. OCD affects 4 to 7 million people in lesioning, radiosurgery, and neurostimulation. the United States at an annual cost of $8 billion Advances in our understanding of the functional per year and was among the 10 leading medical neuroanatomy and electrophysiology of the rele- or psychiatric causes of disability in developed vant circuitry underlying these conditions will countries (1998 World Health Organization increasingly guide placement of lesions or deep study). Conservative estimates of the lifetime brain stimulation (DBS) electrodes. To promote prevalence of major depression are from 2.6% further growth in this field, centers with psychi- to 5.5% in men and 6.0% to 11.8% in women. atric and surgical expertise must design studies Approximately 50% to 85% of patients with major that are prospective and randomized, using reli- depression experience recurrent episodes of illness. able and validated diagnostic and evaluation tools In addition to subjective distress, the disorder can that measure changes in symptoms and quality of be a cause of profound disability, with pervasively life.

To critically examine neurosurgery for psychia- Given the past history of the field, it is particu- tric disorders, we must start by carefully examining larly important that research in this area move the past. Kopell et al provide a historical perspec- forward on a solid scientific and ethical founda- tive that reminds us of the complexities of psycho- tion. Collaborative efforts between centers with surgery and an enduring caution for the future. dedicated multidisciplinary teams of psychiatrists, These historical endeavors were limited by a lack neurosurgeons, and basic scientists are needed to of scientific rationale and other limitations such develop treatments with proven efficacy and safety as surgical techniques and devices, vague patient before the premature introduction of these proce- selection, indiscriminate use, and nonstandardized dures into routine clinical use. In this context, we evaluations. In spite of the decline of surgery for have included an editorial reprinted from a recent psychiatric disorders since the 1950s, early studies publication of the OCD-DBS collaborative group of stereotactic lesioning demonstrated significant in neurosurgery that focuses on recommendations improvements in symptoms of patients with major for future research efforts in this area. Finally, any depression and OCD with few cognitive or neuro- publication focusing on surgery for psychiatric logical side effects. Current stereotactic methods, disorders must include a thoughtful discussion using considerably smaller and more precisely of the ethical implications of these procedures. located targets, have much lower morbidity. Neu- Fins et al offer a careful and considered review roimaging research has focused attention on the of the ethical issues in this field, including issues relationship between activity in specific neuroana- of informed consent. tomical networks and psychiatric symptoms and Together, these articles provide a comprehen- on changes after effective treatment. The articles sive overview of the current state of surgery for by Rauch and Greenberg et al provide an overview psychiatric disorders and the complex issues sur- of the scientific foundation and the emerging thera- rounding the re-emergence of this field. They peutic strategies being explored today. represent a window into what promises to be The articles by Greenberg, Cosgrove, and an exciting future for the development of novel Chang et al consider the evidence for the safety neurosurgical approaches to psychiatric disorders and efficacy of lesioning procedures. The evidence that are based on the neurobiologic basis of these has definite limitations but nevertheless sheds light conditions. on critical issues in assessing the long-term effectiveness and morbidity associated with anterior Ali R. Rezai, MD cingulotomy, anterior capsulotomy, subcaudate Section of Stereotactic and tractotomy, and limbic leucotomy. Jeanmonod Functional Neurosurgery et al describe a novel lesioning approach to neuro- Department of Neurosurgery psychiatric disorders based on recent evidence The Cleveland Clinic Foundation related to thalamocortical dysrhythmias that their Cleveland, OH, USA group have shown are present in these conditions. Nuttin et al present recent data on the use of DBS Steven A. Rasmussen, MD technology for the treatment of OCD. Butler Hospital Two additional novel somatic treatments for Brown University refractory depression have also been the subject Providence, RI, USA of systematic study over the past decade. George et al review the neurobiologic rationale and sum- Benjamin D.Greenberg, MD, PhD marize treatment outcome data for transcranial Department of Psychiatry and Human Behavior magnetic stimulation, while Carpenter et al review Brown University similar evidence for vagal nerve stimulation. Providence, RI, USA Neurosurg Clin N Am 14 (2003) xv–xvi Letter to the editor Deep brain stimulation for psychiatric disorders

There is increasing interest in the use of deep have ongoing oversight of the project should brain stimulation (DBS) as a treatment for pa- approve the investigational protocol. tients with psychiatric conditions. For example, 2. A patient assessment committee should eval- the preliminary results from the first trials, uate each patient as a possible candidate for reported by Nuttin et al [1], have shown that inclusion in the protocol. The role of this DBS may have beneficial effects in patients with committee is to ensure that potential candi- severe obsessive-compulsive disorder (OCD) that dates meet certain medical and psychiatric is refractory to treatment. These preliminary re- criteria and are appropriate for inclusion in sults suggest that DBS may have a role in the the study and to monitor the adequacy of the treatment of patients with intractable OCD. consent process. Patient assessment commit- Treatment of psychiatric patients with DBS re- tees should be constituted broadly to achieve mains investigational, however, and is not con- an ethically valid consensus, and they should sidered standard therapy. have the opportunity to obtain independent Concern regarding the use of neurosurgery for capacity assessments when indicated. the treatment of patients with psychiatric illnesses 3. Candidates for DBS surgery should meet is attributable largely to the indiscriminate and defined criteria for severity, chronicity, dis- widespread application of extensive, destructive ability, and treatment refractoriness. procedures before the stereotactic era. The tragic 4. The use of DBS should be limited solely to example of frontal lobotomy, which was per- those patients with decision-making capacity formed many times before adequate long-term who are able to provide their own informed safety data were available, remains an enduring consent. Patient consent should be main- caution. tained and monitored throughout the process, Given this history, it is incumbent on the sci- and patients should be free to halt their entific community to respect the dignity of pa- participation voluntarily. tients who may be included in studies by ensuring 5. Patient selection, surgical treatment, device adequate protection while providing access to programming, and comprehensive, regular potential therapeutic innovations. We recommend psychiatric follow-up should be conducted at that all investigators adhere to comprehensive or supervised by a clinical research center. standards that protect this potentially vulnerable 6. The investigative team should include special- population while they pursue valuable clinical re- ists from the following disciplines, and they search. Toward that end, we urge that investigators should work in close collaboration: who are engaged in this research establish multi- a. A functional neurosurgical team with es- disciplinary, multicenter teams to systematically tablished experience in DBS. investigate DBS in OCD. On the basis of our ex- b. A team of psychiatrists with extensive experience, we recommend that studies aimed at in- perience in the psychiatric condition under vestigating the use of DBS to treat patients with investigation. psychiatric illnesses meet the following minimum c. Preferably, both of the preceding groups requirements. should have some experience in neurosurgical treatment for psychiatric disorders. If not, 1. An ethics committee (eg, the Institutional close consultation with experienced centers is Review Board in the United States) that will indicated.

7. Investigators must disclose potential conflicts Paul Cosyns of interest to regulatory bodies such as ethics Loes Gabriels committees or institutional review boards and Antwerp, Belgium to potential enrollees during the informed consent process. Bjorn Meyerson 8. The surgery should be performed only to Sergej Andreewitch restore normal function and relieve patients’ Stockholm, Sweden distress and suffering. 9. The procedure should be performed to im- Steven A. Rasmussen prove patients’ lives and never for political, Benjamin Greenberg law enforcement, or social purposes. Gerhard Friehs Butler Hospital In our experience, embarking on this type of Brown University research requires a major commitment of time, 345 Blackstone Boulevard energy, and resources across disciplines before Providence, RI 02906, USA and after device implantation. DBS has the potential to offer hope for severely ill patients, but Ali R. Rezai investigations in this area should proceed cau- Erwin Montgomery tiously to maintain the public trust necessary for Don Malone scientific progress. Cleveland, OH, USA

Acknowledgments Joseph J. Fins We acknowledge the Flemish Scientiﬁc Foun- New York, NY, USA dation (project G.0273.97.N), the Research Coun- cil of the Katholicke Universiteit Leuven (project References OT-98-31), and the National Alliance for Re- search on Schizophrenia and Depression for ﬁ- [1] Nuttin B, Cosyns P, Demeulemeester H, Gybels J, nancial support. We acknowledge the Medtronic, Meyerson B. Electrical stimulation in anterior limbs Inc. QUEST program (L1170) for providing the of internal capsules in patients with obsessive- electrical stimulation devices. compulsive disorder [letter]. Lancet 1999;354:1526.

Bart Nuttin Jan Gybels Leuven, Belgium Neurosurg Clin N Am 14 (2003) 181–197

Psychiatric neurosurgery: a historical perspective Brian Harris Kopell, MDa, Ali R. Rezai, MDb,* aDepartment of Neurosurgery, New York University Medical Center, Suite 8R, 560 First Avenue, New York, NY 10016, USA bSection of Stereotactic and Functional Neurosurgery, Department of Neurosurgery, S-80, Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, OH 44195, USA

Why intervene surgically to treat psychiatric of localization that allowed the surgical interven- disorders? Certainly, past efforts have been tion on neural function. It was the first step at marked with ambiguous results at best and have unlocking the ‘‘black box.’’ been associated with ignoble circumstances at Medical scientists and clinicians soon seized on worst. Yet, the impetus to explore the surgical the idea of localization in the treatment of men- options in psychiatric treatment has once again tal disorders. In the late 1800s, Friedrich Goltz arisen. performed experiments on dogs in which removal of the temporal lobes resulted in animals that were more tame and calmer than the ones not operated The legacy of Moniz, Fulton, and Freeman on [1]. In 1891, drawing from these animal studies, The birth, flourishing, and demise of psycho- a Swiss psychiatrist named Gottlieb Burckhardt surgery in the early twentieth century may serve to (Fig. 3) reported the results a series of surgical caution our current endeavor to intervene surgi- procedures in which he drilled holes in the cally in psychiatric patients. In the hands of the heads of six severely agitated psychiatric patients most prominent physicians and Nobel laureates and extracted sections of their frontal lobes [2]. of the age, despite the best of intentions, such Although in Burckhardt’s series of six patients, intervention led to virtual worldwide banning of three were considered ‘‘successes’’ and two ‘‘par- a seemingly promising surgical venue. This is the tial successes,’’ pressure from his colleagues led tale of a treatment whose promise remains largely him to abandon his efforts. unknown and unfulfilled even today. Why the It might seem surprising, almost inconceivable, descent from mainstream ‘‘miracle’’ to medical that such a rapid leap from animal experiments in anathema? the laboratory to clinical practice on human When Paul Broca presented his case of patient beings could take place; however, it is this phe- M. Leborgne (‘‘Tan’’) in front of the Parisian nomenon that is fundamental to understanding Anthropological Society meeting in 1861, he the rise and ultimate fall of psychosurgery in the ushered in the age of localization (Fig. 1). Lo- twentieth century. Indeed Burckhardt’s report it- calization married the concepts of behavior and self ends with a bold challenge to the fundamental function with a precisely defined anatomic neural practice of medicine: substrate. Before this, Franz Joseph Gall’s phrenology (Fig. 2) tried to localize function based on Doctors are different by nature. One stands fast in the old principle: ‘‘primum non cranial landmarks, but it was Broca’s work that nocere’’; the other states: ‘‘melius anceps looked specifically to the brain. It was the concept remedium quam nullum.’’ I belong naturally to the second category...Every new surgical approach must first seek its special indications and contraindications and methods, * Corresponding author. and every path that leads to new victories is E-mail address: [email protected] (A.R. Rezai). lined with the crosses of the dead. I do not

involving perception, learning, memory, and other higher intellectual faculties, objective data are far more difﬁcult to obtain.’’ [4] This discovery would echo the troubles clinicians would have a decade later as the proliferation of lobotomy raged out of control. In the short run, however, it led Fulton to collaboration with a recently hired psychologist, Carlyle Jacobsen. Then, in 1935, another fateful meeting occurred. At the Second World Congress of Neurol- ogy in London, Fulton and Jacobsen presented their work showing behavioral changes in chimpanzees after ablation of frontal lobe areas (Fig. 5) [5]. Fulton and Jacobsen made the observation that frontal lobe ablation could result in the lessening of ‘‘anxiety states’’ in chimpanzees [6]. In attendance at that meeting was a Portuguese neurologist by the name of Egas Moniz and an American neuropsychiatrist by the name of Walter Freeman (Figs. 6 and 7). Drawing from Fulton and Jacobsen’s data as well as synthesizing case reports of neurosurgeons operating on various frontal lobe lesions at the time, Egas Moniz made this observation in 1935:

It is necessary to alter these synapse adjustments and change the paths chosen Fig. 1. Paul Broca. (Available at: www.uic.edu/depts/ by the impulses in their constant passage mcne/founders/page0013.html.) so as to modify corresponding ideas and force thoughts into different channels... By upsetting the existing adjustments and believe that we should allow this to hold us setting in movement in other [connections], back... [3] I [expect] to be able to transform the psychic reactions and to relieve the patient ‘‘First, do no harm ( primum non nocere)’’ thereby [7]. versus ‘‘Better an unknown cure than nothing at all (melius anceps remedium quam nullum).’’ The Moniz made the critical analytic jump linking battle line for psychosurgery in the early twentieth the seemingly irrational behaviors and thoughts of century was drawn. It is this fundamental struggle psychiatric patients with an anatomically ‘‘nor- between the obligation of the physician to remain mal’’ but disordered neural substrate—a concrete cautious and the desire to help those in need that neural substrate which, when altered in a surgi- would define the efforts to intervene surgically in cal fashion, would result in definitive change of the mentally ill. the seemingly ethereal entities of thought and In 1929, at the age of 30 years, John F. Fulton mind itself. His plan was to sever the white mat- became the chairman of Yale University’s De- ter bundles connecting frontal lobe regions with partment of Physiology (Fig. 4). His research the rest of the brain, the frontal leukotomy. He interest was an extension of Sherrington’s thought: convinced a young neurosurgeon in Lisbon, Al- to reveal the relations among neural systems, it meida Lima (Fig. 8), to undertake the procedure, was necessary to cut the connecting pathways and a series of 20 patients commenced. Despite in progressively higher neuroanatomic levels of this bold leap into human clinical practice, Moniz the system. The lower level thus becomes discon- realized that his inspiration from animal data was nected or ‘‘disinhibited’’ and reveals itself through a tenuous one: ‘‘...it is not possible to obtain changes in neurologic behavior. Fulton applied experimental subjects among animals... [There this approach to frontal lobe function in chim- simply existed too] great a difference between the panzees, yet he found that ‘‘turning to questions psychic life of man and that of animals... [8]’’ B.H. Kopell, A.R. Rezai / Neurosurg Clin N Am 14 (2003) 181–197 183

Fig. 2. Phrenology localization scheme. (Available at: www.artsci.wustl.edu/Àmjstrube/psy315/phrenology.jpg.)

In 1936, Walter Freeman chanced on Moniz’s inspiration and his intention to apply Moniz’s initial communications regarding frontal leuko- procedure in the United States. tomy in the obscure journal Lisboa Medicina. His Why would scientists of such renown as Moniz contemporaries who treated psychiatric maladies and of such promise as Freeman be so hasty in with psychotherapy, useless prattle by his ac- applying such a radical and virtually unfounded count, had frustrated Freeman, a neuropsychia- remedy? Why did melius anceps remedium quam trist of the ‘‘organicist’’ school, who held that nullum become so much more important and more psychiatric illness had a tangible basis in neuro- seductive than primum non nocere? Why did ‘‘do pathophysiology. Freeman, too, recognized that something’’ become more imperative than ‘‘ﬁrst, the scientiﬁc basis for the procedure was ‘‘naive’’; do no harm’’? nevertheless, here was ‘‘something tangible, some- Medicine is an art that is not practiced in the thing that an organicist like myself could un- idealized bubble of science. Science is the realm of derstand and appreciate’’ [9]. For Freeman, a theory, hypothesis, and the experimental model. therapy that was untried and possibly dangerous It deals with the currency of truth and hard was better than the status quo. Freeman wrote fact. Medicine is where science meets all the to Moniz in May of 1936, telling him of his human foibles, hopes, prejudices, and desires to 184 B.H. Kopell, A.R. Rezai / Neurosurg Clin N Am 14 (2003) 181–197

Fig. 4. John F. Fulton. (Available at: www.epub.org.br/ cm/n02/istoria/lobotomy.htm). Fig. 3. Gottlieb Burckhardt. (From Feldman RP, Goodrich JT. Psychosurgery: a historical overview. because of their clinical ‘‘effectiveness,’’ in these Neurosurgery 2001;48:647–59; with permission.) early versions, shock therapies were dangerous and difficult to manage. Insulin-therapy was ‘‘do something.’’ ‘‘[Medicine] is more wrapped up cost- and labor-intensive, because these patients in human, time-dependent concerns than is gene- required constant vigilance from nurses and medi- rally admitted’’ [10]. cal staff. Furthermore, patients were terrified of Nowhere was this desire to ‘‘do something’’ entering into the near-death states of insulin stronger than in the field of psychiatry at the dawn coma. The seizures of metrazol and electrocon- of the twentieth century. Indeed, there were no vulsive therapies, unmodified by anesthesia or drugs or medical procedures available that were muscle relaxants, often had the byproduct of used to treat the specific symptoms of mental spinal and long bone fractures. Yet, these un- illness. Even with the advent of powerful drugs, fortunate side effects were tolerated because of the such as barbiturates, there was little ground ability of these ‘‘somatic’’ treatments to alter the gained in the ability to change the emotional and clinical course of a patient’s mental disease and, mental dynamics of psychiatric disease. ‘‘Thera- more importantly, increase the number of dis- peutic nihilism’’ was the philosophy of treatment charges from inpatient facilities. among psychiatrists at this time, because pa- The economic environment of inpatient psy- tients were allowed by default to languish accord- chiatric care in the nascent twentieth century is ing to the natural history of their disease. Then, also central to understanding the widespread beginning in the 1930s, a wave of ‘‘somatic’’ adoption of such a risky and uncertain treatment therapies began to ‘‘revolutionize’’ psychiatric as psychosurgery by the medical community. The practice (Fig. 9). The most important of these inpatient psychiatric institutions of the early ‘‘shock’’ treatments were injections of metrazol twentieth century had become a vast network of (camphor), insulin-induced hypoglycemic comas, juggernaut institutions, some containing as many and electroconvulsive therapy, all designed to as 10,000 patients. Inpatient admissions were trigger convulsions and states of unconsciousness climbing at an alarming rate. By 1940, there were in the patient. Although extremely welcomed 480,000 psychiatric inpatients, equal to the total B.H. Kopell, A.R. Rezai / Neurosurg Clin N Am 14 (2003) 181–197 185

Fig. 5. Cortical areas ablated in Fulton’s experiments. (From Egas Moniz centenary. Scientific reports. Lisbon; 1977. p. 146; with permission.) number of beds in all nonpsychiatric hospitals brought the miracle of Moniz to the United States combined [11]. As the number of admissions in 1936. increased, the discharge rate decreased. With the Freeman, having convinced a young neurosur- lack of effective therapies, the rate of recoveries geon by the name of James Watts to work with dropped. Evidence mounted that any patient him (himself a former student of Fulton’s), began hospitalized for longer than 2 years would stay a series of patients in September of 1936. They until death. The mathematics of this situation led considered their initial patient, A.H., a resound- to vast overcrowding and inhuman conditions ing success. In their initial communication they (Fig. 10). It was not uncommon for institutions to noted: ‘‘She was well dressed, talked in a low, house twice the number of patients for which they natural tone...showed excellent appreciation of were designed. Scenes of 100 cots placed side by her changed condition. Her husband asserts that side in a ward built for 25 leading to elderly she is more normal than she has ever been’’ [14]. patients climbing over one another to use the This was a middle-aged woman who suffered toilet pervaded the inpatient psychiatric landscape from ‘‘agitated depression,’’ having had a habit of [12]. Albert Maisel’s pictorial essay in the May wild outbursts in which ‘‘she exposed herself 1946 edition of Life magazine described the before the window and urinated upon the floor’’ inpatient mental health system as ‘‘little more [15]. Emboldened by their success, they rapidly than concentration camps on the Belsen pattern.’’ proceeded with an increasing number of patients. Even the psychiatrists themselves began to lose They also modified the Moniz procedure, aban- hope. In a communication that appeared in doning the Moniz leukotome for a dull flat knife Psychiatric Quarterly in 1945, one psychiatrist known as a bistoury (Fig. 11) and approaching equated inpatient psychiatric admission as being from the side as opposed to the top (Figs. 12 and ‘‘identical with doom’’ [13]. Therapeutic nihilism 13). The Moniz procedure, frontal leukotomy, coupled with an overcrowded and economically became the Freeman-Watts technique, the pre- overwhelmed inpatient psychiatric system was frontal or standard lobotomy. fertile ground for an untested, untried, and largely Within a year, Freeman and Watts were eager scientifically unsupported ‘‘miracle.’’ Freeman to share the results of their successes with the 186 B.H. Kopell, A.R. Rezai / Neurosurg Clin N Am 14 (2003) 181–197

Fig. 6. Egas Moniz. (Available at: www.epub.org.br/cm/ n02/historia/lobotomy.htm.) Fig. 7. Walter Freeman. (Available at: www.epub.org. br/cm/n02/historia/lobotomy.htm.) professional community. Freeman would note that the choosing of patients based on disease type was critical to the procedure’s success, with community proceeding with the operations but depressive and obsessional types faring best predicated this support on the concept of carefully and schizophrenics faring worst (in fact, five of designed clinical trials in elite academic institu- six schizophrenics in his series showed no im- tions [18]. Fulton won support for the Freeman, provement) [16]. The initial response from the Watts, and Moniz procedure, but his warnings fell professional community, especially among psy- on deaf ears. The procedure was soon to spread chiatrists, was hostile. They believed that the pro- throughout the medical community. cedure was too radical and too unfounded in its By the end of 1937, James Lyerly, a well- scientific basis. One psychiatrist wrote: ‘‘[It is like] respected community neurosurgeon and a found- burning down the house to roast a pig ... What ing member of the Harvey Cushing Society, began has Moniz accomplished?... No one knows... applying the Freeman-Watts technique in Florida. least of all Moniz’’ [17]. He modified the Freeman-Watts procedure that Initially, Fulton rose to defend the work of he believed was a ‘‘blind’’ one by using a brain Freeman and Watts. Although he acknowledged speculum so that the white matter to be cut was the tenuous scientific link between animal studies exposed to visual inspection. In 1 year, Lyerly had and clinical practice, he thought that there was operated on 19 patients and presented his results rich opportunity for research and significant pro- at the meeting of the Florida Medical Association. mise for the procedure. Using his professional Four previously chronically institutionalized pa- reputation among neurologists and psychiatrists tients had already been discharged. None died. as well as his ties to neurosurgeons (indeed, he was None had any ‘‘serious complications.’’ Lyerly the Harvey Cushing Society’s second president), called the procedure, ‘‘nothing less than miracu- he echoed the need for a procedure to relieve the lous’’ [19]. Other neurosurgeons took notice. overburdening of the national psychiatric hospi- Francis Grant, Chief of the Neurosurgical Service tals. He was in solid support of the medical at the Hospital of the University of Pennsylvania, B.H. Kopell, A.R. Rezai / Neurosurg Clin N Am 14 (2003) 181–197 187

Fig. 9. A patient undergoing insulin shock therapy. Fig. 8. Almeida Lima. (From Feldman RP, Goodrich (From Pressman J. Last resort. Psychosurgery and the JT. Psychosurgery: a historical overview. Neurosurgery limits of medicine. Cambridge: Cambridge University 2001;48:647–59; with permission.) Press; 1998. p. 159; with permission.) and W. Jason Mixter at the Massachusetts General The lay public was especially taken with Hospital began performing lobotomies in 1938. By Freeman and Watts’ book. One state hospital 1939, J.G. Love of the Mayo Clinic traveled to superintendent wrote to Freeman that his 17-year- Florida to learn the procedure from Lyerly so that old daughter read it and told him: ‘‘Daddy, this is he could start a series of lobotomies. the most interesting medical book that I have ever With the publication of Freeman and Watts’ seen. Even I can read it and understand what I am Psychosurgery in 1942, the fervor over the success reading about’’ [21]. Soon, lay publications, such of the lobotomy spread from the professional as The New York Times, Time, Life, Newsweek, and community to the lay community. Using his flair Reader’s Digest, were sensationalizing the suc- for the dramatic, Freeman dispensed with the cesses of the prefrontal lobotomy. customary scientific format and instead presented Once both the profession and the public recog- what was equivalent to pulp nonfiction: nized lobotomy as a treatment of vast potential, Surgeon: Well, the operation is over now its fate ceased to be in the hands of Fulton, anyway, isn’t it? Freeman, or any one person. Psychosurgery’s Patient: Yes, I’m glad it is. You know I fragile connection to the laboratory and the sci- wasn’t dreading this particularly. I’m glad entific community would begin to grow weaker you did it under local because I wanted to and weaker. Fulton once envisioned the psycho- see what it was all about. surgery effort as an experimental one, the task of Surgeon: Did you feel anything particu- rigorously controlled academic institutions and larly during the operation? off limits to the community practitioner. Now, Patient: No, there wasn’t any real pain every aspect of the medical community was par- except when the first needle went in. Drilling through the skull was rather ticipating in the psychosurgery ‘‘miracle.’’ A peculiar, but it didn’t hurt at all. Now, it’s ‘‘clinical drift’’ effect was evident as the proce- just as if that vague unformed apprehension dure became free from any constraints. Initially, that has been with me all these years Freeman and other fervent supporters of loboto- suddenly cleared up [20]. my warned that the procedure should be used 188 B.H. Kopell, A.R. Rezai / Neurosurg Clin N Am 14 (2003) 181–197

Fig. 10. The Incontinent Ward, Byberry State Hospital. (From Pressman J. Last resort. Psychosurgery and the limits of medicine. Cambridge: Cambridge University Press; 1998. p. 149; with permission.)

sparingly in cases of schizophrenia because of its poor clinical eﬀect compared with other psychiatric diseases; however, they were ignored in the desire to ‘‘do something’’ for a group of patients who had no alternative form of therapy available.

Fig. 12. Demonstration of the approach of the Moniz procedure (frontal leukotomy). (From Hitchcock E, Laitinen L, Vaernet K. Psychosurgery. Proceedings of Fig. 11. Examples of bistoury knives. (Available at: the Second International Conference on Psychosurgery. www.gggodwin.com/17-18m.jpg.) Charles C. Thomas: Springﬁeld; 1972; with permission.) B.H. Kopell, A.R. Rezai / Neurosurg Clin N Am 14 (2003) 181–197 189

Fig. 13. Demonstration of the approach in the Freeman-Watts procedure (prefrontal or standard lobotomy). (From Freeman W, Watts J. Psychosurgery: in the treatment of mental disorders and intractable pain. 2nd edition. Springﬁeld: Charles C Thomas; 1950; with permission.)

Psychosurgery’s most ardent proponent ironi- with Freeman. Like other neurosurgeons, he was cally exemplified this clinical drift. At the end of horrified at the ghastly treatment that the patient World War II, prefrontal lobotomy was entering received under Freeman’s new procedure. Patients its heyday. Freeman, however, was looking to were not draped in sterile linen, and there was no expand the use of his procedure. Gone were the surgical backup if a hemorrhage did occur. Yet, days of ‘‘surgery of last resort’’; Freeman wished ‘‘free’’ from his restrictive association with Watts, to operate on a ‘‘better grade’’ of patient, in- Freeman operated in earnest. Freeman and Watts cluding those recently institutionalized as well as recorded 625 operations between 1936 and 1948. those not hospitalized. Freeman was frustrated By 1957, Freeman had lobotomized another 2400 with Watts’ veto power over patients who he patients [24]. In one 12-day period, he operated on believed were not disabled enough by their illness. 225 patients. Time magazine heralded the age of Freeman thought that the rate-limiting step to ‘‘mass lobotomies’’ [25]. psychosurgery’s dissemination was, in fact, the Fulton was appalled at Freeman’s new course neurosurgeon, being too few in number and too of action. ‘‘What are these terrible things I hear expensive in cost. Thus, with an ice pick, Freeman about you doing lobotomies in your office with an severed psychosurgery’s final ties with its roots, ice pick?...Why not use a shot gun’’ [26], Fulton’s those of neurosurgery itself. incredulous missive to Freeman seethed. Freeman Drawing from an obscure report of an Italian did not relent, claiming that the transorbital psychiatrist, Amarro Fiamberti, Freeman devel- procedure was ‘‘much less traumatizing than a oped the transorbital lobotomy, a ghastly pro- shotgun and almost as quick’’ [26]. Their conflict cedure in which the frontal white matter is cut by exemplified their personal philosophies regarding a metal spike inserted through the thin bony orbit the leap from laboratory to clinic and the growing above the eye [22]. Freeman’s initial choice to chasm between psychiatric neurosurgery and its accomplish this procedure was the common ice physiologic roots. Fulton believed that the most pick (Fig. 14). Although Freeman refined the reliable and safe medical knowledge came from common house tool into what he called a ‘‘trans- a combination of basic animal research coupled to orbital leucotome,’’ he envisioned the procedure a few rigorously designed and executed clinical being able to be performed by any surgically trials. Fulton realized that the complete lack of untrained physician after the most minimal in- reliable objective testing methods available to struction, with ‘‘every physician his own loboto- psychiatry at the time made it impossible to track mist’’ [23]. By 1948, Freeman, with ice pick in lobotomy’s efficacy from any scientific point of hand, traveled across the country to fulfill what view. Conversely, the most important question to he considered was an unanswered need (Fig. 15). Freeman was, ‘‘Did it work?’’ [27]. What counted It did not take long for Watts to sever his ties to Freeman were not intelligence tests but the 190 B.H. Kopell, A.R. Rezai / Neurosurg Clin N Am 14 (2003) 181–197

Fig. 14. Freeman transorbital (‘‘ice pick’’) procedure. (From Freeman W, Watts J. Psychosurgery: in the treatment of mental disorders and intractable pain. 2nd edition. Springfield: Charles C Thomas; 1950; with permission.) successful return of patients to everyday life. violent behavior could now make contact with Primum non nocere versus Melius anceps remedium others and be left without supervision. quam nullum. Yet, even as psychosurgery’s most In 1954, the U.S. Food and Drug Administra- ardent scientific proponents began to question the tion approved chlorpromazine, and it took the widespread use of lobotomy, the procedure had field of psychiatry by storm. The death knell become as mainstream a treatment as appendec- for psychiatric neurosurgery was sounded; yet, tomy. The awarding of the Nobel Prize to Egas in Freeman’s case, it fell on deaf ears. Freeman Moniz in 1949 typified the simultaneous legitimi- carried on his one-man war. He was relentless in his zation of lobotomy amid growing concerns regard- attempt to convince his colleagues to perform more ing its efficacy and side effects. From 1936 to the of his procedures. On occasion, Freeman would mid-1950s, approximately 20,000 psychosurgical dump shoe boxes crammed with letters from procedures, virtually all frontal lobotomies of some ‘‘grateful’’ lobotomized patients onto the desks of form, were performed in the United States [28]. skeptical colleagues. They remained unconvinced. Then, it was over. Ironically, it was a surgeon Freeman died in 1972 at the age of 76 years [29]. who helped end the ‘‘golden age’’ of psychosur- Yet, as this interest waned, technologic devel- gery and usher in the pharmacologic age of psy- opments, especially in the realm of stereotaxis, chiatry. In 1952, Henri Laborit, a surgeon in ushered in a second wave of psychiatric neuro- Paris, was looking for a way to reduce surgical surgery. shock in his patients. Having tried antihistamines, generally used to fight allergies, he noticed that The stereotactic era when he gave a strong dose to his patients, their mental state changed. No longer did they seem Indeed, the first modern neurosurgical pro- anxious about their upcoming surgery; in fact, cedure for psychiatric disease, the frontal leukot- they were rather indifferent. A fellow surgeon omy, sought to interrupt white matter tracts passed the word to his brother-in-law, a psychia- associated with the frontal lobes. This procedure trist named Pierre Deniker. Deniker’s interest was started as a rather extensive one and became more piqued, and he ordered some chlorpromazine to refined as the volume of brain in the surgical try on his most agitated and uncontrollable target became smaller. As the experience with patients. frontal lobotomy increased, there was some The results were stunning. Patients who had suggestion that minimizing cortical damage and stood in one spot without moving for weeks focusing on the white matter tracts could retain and patients who had to be restrained because of efficacy while minimizing side effects. This trend B.H. Kopell, A.R. Rezai / Neurosurg Clin N Am 14 (2003) 181–197 191

bioethicists and neurosurgeons is assembled to make sure the patient in question is both refractory and appropriate. Although many of the following studies of psychiatric neurosurgery have had significant flaws, most notably, the inherent bias of a nonrandomized non-double-blind study and the lack of objective functional imaging techniques, they do suggest a viable means of treatment for a subset of patients who may have no other options. The procedures themselves are described below. Their safety and efficacy for their main current indications, intractable OCD and depression, are considered at length in the article by Greenberg and his colleagues in this issue.

Cingulotomy Fig. 15. Freeman in a Winnebago. Freeman often traveled the country, performing his transorbital pro- Surgery on the cingulate gyrus dates back to cedure. (From Pressman J. Last resort. Psychosurgery observations in the 1940s that severing fibers from and the limits of medicine. Cambridge: Cambridge the cingulate gyrus led to a decrease in anxiety University Press; 1998. p. 406; with permission.) type states [32]. In 1952, Whitty et al [33] reported their cingulectomy, in which a 4-cm Â 1-cm sec- toward increasingly discrete subcortical lesions tion of cingulate gyrus was resected bilaterally. In culminated in the application of stereotaxis on 1967, Ballantine et al [34] introduced the modern psychiatric neurosurgical procedures. The second stereotactic procedure in which a lesion, localized wave of psychosurgery was born. by air ventriculography and made using thermo- In 1947, Wycis and Spiegel introduced the coagulation, was made bilaterally in the anterior dorsomedial thalamotomy [30,31], the first sub- cingulate. The lesion is typically made bilaterally cortical stereotactic neurosurgical procedure per- 2 to 2.5 cm from the tip of the frontal horns, 7 mm formed on human beings and the model on which lateral from the midline, and 1 mm above the all modern psychiatric neurosurgical procedures roof of the ventricles (Fig. 16). The procedure per- are based. formed today has been refined using the latest in The four psychiatric neurosurgical procedures stereotactic equipment and imaging techniques. currently in use are cingulotomy, capsulotomy, Stereotactic cingulotomy is the most frequently subcaudate tractotomy, and limbic leukotomy, reported neurosurgical procedure for psychiatric which are all stereotactic interventions. These disease in the United States and Canada. procedures are typically performed on the severe and refractory psychiatric patient. First, a patient Capsulotomy must meet Diagnostic and Statistical Manual of Mental Disorders (DSM IV) criteria for a partic- Developed in Sweden by Lars Leksell and by ular psychiatric disease such as obsessive-compul- Talairach in France, anterior capsulotomy has sive disorder (OCD) or affective disorder. Next, a been in use for refractory psychiatric illness since patient must fail several rounds of treatment with 1949. There are two forms of this procedure, both multiple psychotropic medications combined with of which are stereotactic operations. One tech- appropriate psychotherapy before he/she is con- nique involves the use of radiofrequency, and the sidered for surgical treatment. Therapeutic failure other uses gamma radiation to make the lesion. In is determined by quantitative analysis using the either case, the target area is between the anterior most appropriate and accurate psychiatric batter- third and middle third of the anterior limb of the ies of tests available, such as the Yale-Brown internal capsule at the approximate level of the obsessive-compulsive scale (Y-BOCS), the Clini- foramen of Monro. Specifically, the ideal target cal Global Impression (CGI), and the Hamilton lays 17 mm from the midline, 10 mm rostral to Depression scale for depression (HAM-D). When the anterior commissure, and 8 mm above the surgical treatment is ultimately considered, a mul- intercommissural line. The lesion is approximately tidisciplinary team consisting of psychiatrists, 15 to 18 mm in length and 4 to 5 mm in width neuropsychologists, neurologists, lawyers, clergy, (Fig. 17) [35,36]. 192 B.H. Kopell, A.R. Rezai / Neurosurg Clin N Am 14 (2003) 181–197

Fig. 16. Typical cingulotomy lesions. (From Laitinen L, Livingston K. Surgical approaches in psychiatry. Proceedings of the Third International Congress of Psychosurgery. Lancaster: MTP; 1973; with permission.)

Fig. 17. Capsulotomy lesion sites. (From Hitchcock E, Ballantine H, Meyerson B. Modern concepts in psychiatric surgery. Proceedings of the Fifth World Congress of Psychiatric Surgery. Elsevier/North-Holland Biomedical Press; 1979; with permission.) B.H. Kopell, A.R. Rezai / Neurosurg Clin N Am 14 (2003) 181–197 193

Subcaudate tractotomy Another stereotactic procedure geared toward interrupting fibers from the orbitofrontal cortex to the thalamus is subcaudate tractotomy (in- nominatomy). Developed by Knight in 1965 in London, the operation was designed to relieve depressive, anxiety, and obsession symptoms while minimizing postoperative epilepsy and cognitive/personality deficits [37,38]. The lesion is created by multiple 1-mm Â 7-mm rods of ytr- rium-90, a beta-emitter that releases lethal radiation to tissue within 2 mm. These rods have a half- life of 68 hours, after which they become inert. The target site, a region of white matter localized beneath the head of the caudate known as the substantia innominata, has traditionally been localized by ventriculogram. A stereotactic apparatus places the rods after bilateral burr holes are made just above the frontal sinus and 15 mm from the midline. The lesion itself lays at the anteroposterior level of the planum sphenoidale, extending from 6 to 18 mm from the midline and being 20 mm long in an anteroposterior direction (Fig. 18). Initially, placing two rows of four rods each made the lesion. Later studies, having refined the technique, have created the lesion by radiofrequency thermocoagulation [39]. Fig. 18. Sites of subcaudate (basofrontal) tractotomy lesions. (From Laitinen L, Livingston K. Surgical ap- Limbic leukotomy proaches in psychiatry. Proceedings of the Third Inter- national Congress of Psychosurgery. Lancaster: MTP; Although the other three aforementioned pro- 1973; with permission.) cedures each target a single anatomic substrate, a fourth procedure is designed to interrupt fibers at two separate areas, one involving a fronto- continued to grow. Like the human behavioral thalamic loop and the other involving an area of example of Newton’s third law of motion, the the Papez circuit (Fig. 19). Called limbic leu- fervor with which psychosurgery was welcomed in kotomy, the procedure was developed in England the first half of the twentieth century became an by Desmond Kelly and Alan Richardson in the ardent backlash against what was regarded as early 1970s. The operation itself consists of three a reckless and sinister procedure. In the United 6-mm thermocoagulative or cryogenic lesions in States, charges of abuse and allegations of surgery the lower medial quadrant of each frontal lobe (to for social control culminated in the establishment interrupt frontothalamic connections) and two of a National Commission in 1977. This board 6-mm lesions in each cingulum (Fig. 20). examined all the neurosurgical procedures performed in the United States from the freehand frontal lobotomies to the stereotactic lesioning procedures. Careful emphasis was taken to review What now? the efficacy and safety of these procedures. The Despite the fact that neurosurgery for psychi- National Commission’s findings were surprising, atric illness decreased in frequency in the latter as the Chairman reported in his review: ‘‘We half of the twentieth century and that neurosur- looked at the data and so they did not support our gical technologic developments like stereotaxis prejudices. I, for one, did not expect to come out vastly improved the quality of the surgical inter- in favor of psychosurgery. But we saw that some vention itself, the negative bias toward its practice very sick people had been helped by it’’ [40]. 194 B.H. Kopell, A.R. Rezai / Neurosurg Clin N Am 14 (2003) 181–197

Fig. 19. Papez circuit. (From Laitinen L, Livingston K. Surgical approaches in psychiatry. Proceedings of the Third International Congress of Psychosurgery. Lancaster: MTP; 1973; with permission.)

The National Commission was so impressed number of these refractory psychiatric patients. by the potential benefit of psychiatric neurosur- Treatment-resistant mental disorders, in any of gery that it recommended a review board be their forms, must be looked on as degenerative formed to study these procedures in a more and potentially fatal neurologic illnesses. scientific manner. This review board was never The promise of a modern undertaking of formed, however. psychiatric neurosurgery is that it is no longer There are several reasons to continue to equivalent to lobotomy. We benefit from the tech- evaluate the role of neurosurgery in the treatment nologic leaps that have occurred over the past 50 of psychiatric disease. Despite adherence to years. One of the challenges in treating psychi- therapeutic guidelines and conscientious compli- atric disease during the zenith of lobotomy was ance, there still exists a population of psychiatric the quantitative analysis of patients throughout patients who are refractory to aggressive conven- the course of treatment. Modern psychiatric test- tional treatment with medication and empirically ing batteries, such as the Y-BOCS, CGI, and proven psychotherapies. A particularly intrac- HAM-D, allow for a more accurate and objective table population exists among OCD patients. By evaluation of patients undergoing psychiatric most reviews of current treatment strategies, ap- neurosurgical procedures in ways that were not proximately 10% of all OCD patients show an previously available. unrelenting downward course despite all psycho- Neurosurgical practice has also grown tremen- therapeutic and pharmacologic treatments [41,42]. dously since the heyday of frontal lobotomy. Because it is estimated that 3% of the world’s Today’s neurosurgeon, bolstered by technologic population suffers from OCD, this alone is a advancements in stereotactic equipment coupled substantial patient population. Affective disorder, with the advent of intraoperative microelectrode which includes major depression and bipolar recording, can hit physiologic targets with a degree disorder, similarly has a treatment-resistant subset of accuracy never before possible. Furthermore, of patients [43,44]. For some of these patients, techniques have been developed that allow the surgery may still be a viable treatment alternative neurosurgeon to modulate neurologic function in the psychiatrists’ armamentarium. while minimizing the risk to the patient. Stereo- It should be made clear that the end point for tactic radiosurgery, exemplified by such devices untreated psychiatric disease is dismal. Setting as the gamma knife, enables lesions to be made aside the loss of productivity on both a personal without any upfront risk to the patient. Perhaps and professional level, self-mutilation and suicide the neurosurgical technical development that can be the final tragic outcome for a significant holds the most significance for psychiatric neuro- B.H. Kopell, A.R. Rezai / Neurosurg Clin N Am 14 (2003) 181–197 195

Fig. 20. Limbic leukotomy lesion sites. (From Laitinen L, Livingston K. Surgical approaches in psychiatry. Proceedings of the Third International Congress of Psychosurgery. Lancaster: MTP; 1973; with permission.) surgery is deep brain stimulation (DBS). With its the experimental animal model has bolstered ability to modulate neuronal function without the movement disorder surgery, the inherent lack of necessity of a permanent lesion, DBS has already valid animal models in the latter has hindered become the mainstay of the neurosurgical treat- efforts. This intermediate link between animal ment of movement disorders. Because previous physiologic models and clinical practice is a fun- efforts to prove the efficacy of surgical interven- damental paradigm shift with respect to the initial tion in psychiatric disease have been hampered by efforts with lobotomy. As our knowledge of the the lack of double-blind studies, DBS, with its circuitry underlying psychiatric disease increases, inherent ability to be turned ‘‘on’’ and ‘‘off,’’ better surgical targets will become apparent. Such might prove particularly useful. a link between functional imaging and surgical Our knowledge of the pathophysiology un- practice has already been demonstrated in the derlying psychiatric disease has also grown. endeavors in chronic pain and movement disor- Today, state-of-the art imaging techniques, such ders. Together, these tools can help to eliminate as positron emission tomography (PET), func- some of the shortcomings of past studies of tional magnetic resonance imaging (fMRI), and psychiatric neurosurgical procedures. magnetoencephalography (MEG), allow the clin- As we re-explore surgery’s role in psychiatric ical investigator a noninvasive method of directly treatment, we have learned from the example of and precisely localizing brain function and anat- lobotomy that we must tread carefully. Our initial omy. Although experimental, these imaging tech- efforts, of course, must remain firmly in the realm of niques allow us to evaluate human brain function a limited number of academic institutions. Ulti- directly without having the potential misdirection mately, because of the vast resources implicit in the that animal models may entail. This highlights the multidisciplinary approach, psychiatric neurosur- difference between the surgical efforts for move- gery may never enter community practice. Despite ment disorders and psychiatric disease: although all these improvements and the heeding of past 196 B.H. Kopell, A.R. Rezai / Neurosurg Clin N Am 14 (2003) 181–197 mistakes, patients and clinicians alike must always [18] Pressman J. Last resort. Psychosurgery and the lim- be cognizant of the potential dangers in this task. its of medicine. Cambridge: Cambridge University Nevertheless, past efforts suggest that a real poten- Press; 1998. p. 146. tial to benefit some extremely sick people may exist. [19] Pressman J. Last resort. Psychosurgery and the lim- It is up to future multidisciplinary endeavors its of medicine. Cambridge: Cambridge University Press; 1998. p. 104–5. conducted in carefully controlled centers of aca- [20] Freeman W, Watts J. Psychosurgery. Springfield: demic expertise to determine whether this potential Charles C. Thomas; 1942. p. 110. is either real or folly. [21] Pressman J. Last resort. Psychosurgery and the limits of medicine. Cambridge: Cambridge University References Press; 1998. p. 138. [22] Fiamberti A. Proposta di una tecnica operatoria [1] Goltz F. Ueber die Verrichtungen des Grosshirns. modificata e semplificata per gli interventi alla Arch f d ges Physiol 1991;26:1–49. Moniz sui lobi frontali in malati di mente. Rass [2] Burckhardt G. Uber Rindenexcisionen, als Beitrag Studi Psichiatr 1937;26:797. zur operativen Therapie der Psychosen. Allege- [23] Pressman J. Last resort. Psychosurgery and the lim- meine Z Psychiatr 1891;47:463–548. its of medicine. Cambridge: Cambridge University [3] Freeman W, Watts J. Psychosurgery. Springfield: Press; 1998. p. 341. Charles C. Thomas; 1942. [24] Pressman J. Last resort. Psychosurgery and the lim- [4] Pressman J. Last resort. Psychosurgery and the its of medicine. Cambridge: Cambridge University limits of medicine. Cambridge: Cambridge Univer- Press; 1998. p. 337–40. sity Press; 1998. p. 61–73. [25] Pressman J. Last resort. Psychosurgery and the lim- [5] Fulton JF, Jacobsen CF. The functions of the frontal its of medicine. Cambridge: Cambridge University lobes: a comparative study in monkeys, chimpanzees, Press; 1998. p. 340, (also see Time. September 15, and man. In: Abstracts of the Second International 1952. p. 86). Neurological Congress; 1935. p. 70–1. [26] Pressman J. Last resort. Psychosurgery and the lim- [6] Jacobsen CF, Wolfe JB, Jackson TA. An experi- its of medicine. Cambridge: Cambridge University mental analysis of the functions of the frontal as- Press; 1998. p. 342. sociation areas in primates. J Nerv Ment Dis 1935; [27] Pressman J. Last resort. Psychosurgery and the lim- 82:1–14. its of medicine. Cambridge: Cambridge University [7] Moniz E. How I came to perform leucotomy. Psy- Press; 1998. p. 345. chosurgery. Lisbon; 1948. p. 11–4. [28] Kramer M. In: Overholser W, editor. The 1951 sur- [8] Valenstein. Great and desperate cures. 97–100. vey of the use of psychosurgery. Proceedings of the [9] Pressman J. Last resort. Psychosurgery and the lim- Third Research Conference on Psychosurgery. Wash- its of medicine. Cambridge: Cambridge University ington, DC: USPHS Publication221; 1954. p. 159–68. Press; 1998. p. 77. [29] Pressman J. Last resort. Psychosurgery and the lim- [10] Pressman J. Last resort. Psychosurgery and the limits of medicine. Cambridge: Cambridge University its of medicine. Cambridge: Cambridge University Press; 1998. p. 407. Press; 1998. p. 198. [30] Spigel EA, Wycis HT, Marks M, et al. Stereotac- [11] Pressman J. Last resort. Psychosurgery and the lim- tic apparatus for operations on the human brain. its of medicine. Cambridge: Cambridge University Science 1947;106:349. Press; 1998. p. 152. [31] Spiegel EA, Wycis HT. Physiological and psycho- [12] Pressman J. Last resort. Psychosurgery and the lim- logical results of thalamotomy. Proc R Soc Med its of medicine. Cambridge: Cambridge University 1949;42:89–93. Press; 1998. p. 151. [32] Freeman W, Watts J. Psychosurgery: intelligence, [13] Eliasberg W. Institutionalizing the obsessive psy- emotion and social behavior following prefrontal chopath. Psychiatr Q 1945;19:697–701. lobotomy for mental disorders. Springfield: Charles [14] Pressman J. Last resort. Psychosurgery and the lim- C. Thomas; 1942. its of medicine. Cambridge: Cambridge University [33] Whitty CWM, Duffield JE, Tow PM. Anterior Press; 1998. p. 79. cingulectomy in the treatment of mental disease. [15] Pressman J. Last resort. Psychosurgery and the lim- Lancet 1952;1:475–81. its of medicine. Cambridge: Cambridge University [34] Ballantine HT Jr., Cassidy WL, Flanagan NB, et al. Press; 1998. p. 77. Stereotactic anterior cingulotomy for neuropsychi- [16] Pressman J. Last resort. Psychosurgery and the lim- atric illness and intractable pain. J Neurosurg 1967; its of medicine. Cambridge: Cambridge University 26:488–95. Press; 1998. p. 81. [35] Mindus P, Meyerson BA. In: Schmidek HH, Sweet [17] Pressman J. Last resort. Psychosurgery and the lim- WH, editors. Operative neurosurgical techniques. its of medicine. Cambridge: Cambridge University Indications, methods and results. 3rd edition. Phila- Press; 1998. p. 83–4. delphia: WB Saunders; 1995. p. 1443–54. B.H. Kopell, A.R. Rezai / Neurosurg Clin N Am 14 (2003) 181–197 197

[36] Kihlstrom L, Hindmarsh T, Lax I, et al. Radio- Report and recommendations. Washington, DC: surgical lesions in the normal human brain 17 years DHEW Publication (05)77–0001, 1977. after gamma knife capsulotomy. Neurosurgery 1997; [41] Rasmussen SA, Eisen J. Treatment strategies for 4:396–402. chronic and refractory obsessive-compulsive dis- [37] Knight K. Stereotactic tractotomy in the surgical order. J Clin Psychiatry 1997;58(Suppl 13):9–13. treatment of mental illness. J Neurol Neurosurg [42] Jenike MA. An update on obsessive-compulsive Psychiatry 1965;28:304–10. disorder. Bull Menninger Clin 2001;65:4–25. [38] Kartsounis LD, Poynton A, Bridges PK, et al. [43] Fava M. New approaches to the treatment of Neuropsychological correlates of stereotactic sub- refractory depression. J Clin Psychiatry 2000; caudate tractotomy. Brain 1991;114:2657–73. 61(Suppl 1):26–32. [39] Malhi GS, Bartlett JR. Depression: a role for [44] Frye MA, Ketter TA, Leverich GS, et al. The neurosurgery? Br J Neurosurg 2000;14:415–22. increasing use of polypharmacotherapy for refrac- [40] National Commission for the Protection of Human tory mood disorders: 22 years of study. J Clin Subjects of Biomedical and Behavioral Research. Psychiatry 2000;61:9–15. Neurosurg Clin N Am 14 (2003) 199–212

Neurosurgery for intractable obsessive-compulsive disorder and depression: critical issues Benjamin D. Greenberg, MD, PhDa,*, Lawrence H. Price, MDa, Scott L. Rauch, MDc, Gerhard Friehs, MDb, Georg Noren, MDb, Donald Malone, MDe, Linda L. Carpenter, MDa, Ali R. Rezai, MDd, Steven A. Rasmussen, MDa aDepartment of Psychiatry and Human Behavior, Brown Medical School, Butler Hospital, 345 Blackstone Boulevard, Providence, RI 02906, USA bDepartment of Clinical Neuroscience, Brown Medical School, Providence, RI, USA cDepartment of Psychiatry, Massachusetts General Hospital, Boston, MA, USA dDepartment of Neurosurgery, The Cleveland Clinic Foundation, Cleveland, OH, USA eDepartment of Psychiatry, The Cleveland Clinic Foundation, Cleveland, OH, USA

The effectiveness and tolerability of treatments surgery may be of benefit to patients who fail to for obsessive-compulsive disorder (OCD) and improve with the best available conventional depression have seen impressive improvements treatment. This article considers the efficacy and over past decades. Nonetheless, some patients safety of lesion procedures. The evidence has with either disorder continue to manifest severe important limitations but sheds light on critical chronic illness that is refractory to treatment. For issues in assessing the long-term effectiveness and them, modern neurosurgical procedures remain morbidity associated with existing procedures, a therapeutic option. Psychiatric neurosurgery including anterior cingulotomy, anterior capsu- remains controversial, largely because indiscrim- lotomy, subcaudate tractotomy, and limbic leu- inate use of prefrontal lobotomy in the middle of kotomy. The same methodologic issues arise when the twentieth century frequently produced signif- considering the newer nondestructive techniques icant deficits in emotional responsiveness and that are currently in development for the treat- motivation, sometimes with little or no therapeu- ment of intractable psychiatric illness, including tic benefit. Although the historical experience deep brain stimulation. Determining the effective- remains an enduring caution, current stereotactic ness and side effect burden of neurosurgery for methods using considerably smaller and more intractable psychiatric illness is a task primarily precisely located targets have much lower mor- for psychiatrists, collaborating closely with neuro- bidity. Moreover, an increasingly specific neurosurgeons, neurologists, and neuropsychologists in biologic rationale for psychiatric neurosurgery specialized multidisciplinary teams. is being developed. Neuroimaging research has Treatments for OCD and depression have focused attention on the relations between activity come a long way in recent decades. Even so, in specific neuroanatomic networks and psychiat- some patients with either disorder have severe ric symptoms and on changes in such relations chronic illness that the best conventional methods after effective treatment. A small number of are unable to improve. For this reason, there has prospective studies support the view that neuro- been continuing work aimed at developing what might be termed neuroanatomically-based treatments. For example, there has been considerable * Corresponding author. interest in transcranial magnetic stimulation, a E-mail address: [email protected] (B.D. Greenberg). noninvasive method for stimulating the cerebral

1042-3680/03/$ - see front matter Ó 2003, Elsevier Inc. All rights reserved. doi:10.1016/S1042-3680(03)00005-6 200 B.D. Greenberg et al / Neurosurg Clin N Am 14 (2003) 199–212 cortex, as a treatment for refractory depression (see at $8 billion per year in 1990, including $2.1 billion the article by George et al in this issue). Most in direct costs and $5.9 billion in indirect costs recently, vagus nerve stimulation (VNS) has been related to lost productivity [5]. OCD was among the proposed as a novel surgical approach for re- 10 leading medical or psychiatric causes of disabil- fractory depression (see the article by Carpenter ity in developed countries in a 1998 World Health et al in this issue). Organization study [6]. The disorder tends to be Even as research on those techniques contin- familial, with the risk 1.5 to 3 times higher in indi- ues, modern neurosurgical procedures in use for viduals with an affected first-degree relative [4]. the last four decades have remained treatments of Specific treatments for OCD have been de- last resort for patients with otherwise intractable veloped over the past decades, significantly im- illness. This article considers the evidence for the proving patients’ lives. Current behavioral and safety and effectiveness of existing lesion proce- pharmacologic therapies are far from universally dures. This literature offers a degree of hope to effective, however. It is estimated that they pro- affected individuals and their families. The evi- vide substantial benefit to only 50 to 70% of pa- dence also suffers from limitations, however, and tients seeking treatment [1]. Selective serotonin leaves some important questions unanswered. We reuptake inhibitors (SSRIs), such as fluoxetine, conclude by focusing on those unresolved issues sertraline, fluvoxamine, citalopram and paroxe- and suggest ways of answering them in systematic tine, are the mainstay of pharmacologic treat- prospective research, both on existing lesion pro- ment. Clomipramine, a tricyclic antidepressant and cedures and on the newer nondestructive thera- a potent but non-serotonin reuptake inhibitor, is a pies of deep brain stimulation targeting the same second-line agent. Even when effective, side effects circuitry. may substantially limit the treatment adherence Evaluating the benefits and side effect burdens usually necessary for ongoing symptom relief. For of psychiatric neurosurgery will require the example, one study found that more than half of dedicated work of interdisciplinary teams. Psy- patients who responded to an adequate trial of chiatrists will play the central role, in close a serotonin reuptake inhibitor stop taking the collaboration with neurosurgeons, neurologists, medicine within 2 years because of sexual dys- neuropsychologists, and bioethicists. function, weight gain, or sedation [7]. Medication To begin, the epidemiology, symptoms, and cur- augmentation strategies can benefit patients who rent treatment of OCD and depression are briefly fail to respond to SSRI monotherapy [8,9]. considered. The discussion then focuses on current Typical and atypical neuroleptics are effective in views of neurosurgery for intractable OCD and controlled studies but impose an additional side depression among psychiatrists, the historical effect burden. These include sedation, weight gain, background, the current procedures, and evidence extrapyramidal reactions, and, especially for for safety and effectiveness. The article concludes typical neuroleptics, a risk of tardive dyskinesia with a discussion of the limitations of the evidence or tardive dystonia over time. Clonazepam also and considers some central unanswered questions. seems to be partially effective as an augmenting agent but, as with other benzodiazepines, is sedating and may induce dependence. Obsessive-compulsive disorder Behavior therapy is particularly useful for OCD is common, affecting from 2% to 3% of OCD and holds a central place in effective the population, or 4 to 7 million people in the treatment regimens. After the symptoms and their United States [1,2]. Prevalence seems to be similar environmental triggers are characterized in an in other countries and cultures. The illness is charac- individual patient, the therapy involves deliberate terized by recurrent thoughts, images, feelings, or provocation of increases in anxiety by exposure behaviors that persist against the patient’s attempts to those triggers while encouraging resistance to to eliminate them and which are accompanied compulsive rituals. The anxiety and compulsive by marked and often overwhelming anxiety. OCD urges typically become progressively more tran- typically begins in childhood or adolescence. sient and less intense as therapy continues. Behav- Symptoms, and the distress resulting from them, ior therapy is effective in 70% to 80% of patients are usually chronic [3,4]. OCD is associated with completing a course of treatment. A substantial significant, and often dramatic, impairment in number of OCD sufferers find that the distress social and occupational functioning [1]. Total costs induced as a necessary part of the therapy is of the disorder in the United States were estimated intolerable, however, and either refuse to start or B.D. Greenberg et al / Neurosurg Clin N Am 14 (2003) 199–212 201 fail to complete it. Particularly for seriously ill these various approaches are frequently employed patients, it seems that the most effective treatment in clinical practice, particularly in the manage- is a combination of expert medication management of complicated or refractory cases [16]. More ment and intensive behavior therapy offered by severe forms of depression are often treated with experts at specialized treatment centers. Unfortu- electroconvulsive therapy (ECT), which has been nately, such optimal therapy is not widely avail- in use for this indication for nearly 70 years and is able and is also expensive, limiting its application still considered a gold standard of antidepressant to only a few individuals presenting for treatment. treatment. ECT can be associated with significant Even when the best available medication and adverse effects, however, most prominently mem- behavior therapies are applied, an estimated 10% ory loss. Moreover, ECT is often viewed by the of patients remain severely affected. This group lay public (and many nonpsychiatric physicians) suffers from intractable OCD, with tremendous as primitive, punitive, and potentially neurotoxic. suffering and overall functional impairment. Despite the availability of effective treatments, a substantial proportion of patients fail to recover from episodes of depression. Fava and Davidson Depression [17] estimated that 29% to 46% of depressed pa- Major depression is characterized by depressed tients fail to respond fully to an antidepressant mood, apathy, anhedonia, appetite and weight trial in which adequate dosing and duration have disturbance, sleep disruption, psychomotor ab- occurred. Using rigorous operational criteria, normalities, fatigue, guilt, impaired concentra- Little et al [18] observed an 18.9% rate of refrac- tion, and suicidal ideation and behavior [10]. toriness in depressed geriatric patients treated in Delusions, hallucinations, and catatonia are a university tertiary care setting. sometimes present. Depression is quite common; in the United States, conservative estimates place Current awareness of psychiatric neurosurgery its prevalence at 2.6% to 5.5% in men and 6.0% to 11.8% in women [11]. The disorder tends to Neurosurgical treatment for intractable psy- be familial, with the risk 1.5 to 3 times higher in chiatric illness has gradually become more visible individuals with an affected first-degree relative to the general public, in part because of recent [10]. The average age of onset is the late 20s, but it attention paid by the media to the therapeutic may develop at any age. Approximately 50% to potential of deep brain stimulation in psychiatry. 85% of patients with major depression experience Most psychiatrists, in contrast, have been aware recurrent episodes of illness [12]. In addition to that modern lesion procedures offered a last subjective distress, the disorder can be a cause of avenue of hope for patients with intractable profound disability, with pervasively negative OCD and depression for the last several decades. effects on marital, parental, social, vocational, This awareness is based on past retrospective and academic role function [13]. The recent studies and, more recently, on a small number of Global Burden of Disease Study identified de- prospective investigations. For example, a survey pression as the fourth leading cause of disability in published in 1984 found that 78% of adult the world and the leading cause of disability in psychiatrists in the United Kingdom had referred adults [6]. Death from suicide is a major direct patients for subcaudate tractotomy, mainly for complication of the illness, but depression can refractory depression and OCD [19]. Later data also exacerbate the course of other psychiatric and indicate that referrals for psychiatric neurosur- medical conditions, with increases in both mor- gery in Britain continued at roughly the same rate bidity and mortality often reported. from 1979 to 1993, although the proportion of pa- More than 20 drugs are approved or com- tients accepted for surgery was reduced from 50% monly used to treat depression in the United to 60% to about 20% during that period. This de- States [14]. Efficacy for these agents is well estab- cline resulted from the institution of high-dose lished [15]. Available antidepressant drugs can be medication trials as a screening criterion, and more grouped into four major categories: tricyclics and patients responded to such trials before surgery [20]. tetracyclics, SSRIs, monoamine oxidase inhibi- A later survey of psychiatrists in the United tors (MAOIs), and other drugs acting on biogenic States, published in 1999 [21], likewise showed amine systems. Less severe forms of the illness can widespread awareness of psychiatric neurosurgery be treated effectively with certain forms of psycho- and a willingness to consider it for selected pa- therapy and light therapy [14]. Combinations of tients. Eighty-three percent of a random sample 202 B.D. Greenberg et al / Neurosurg Clin N Am 14 (2003) 199–212 of American Psychiatric Association members Still, neurosurgical treatment of patients with replied that they knew about neurosurgical treat- OCD, major affective illness, and other psychiatric ment for intractable OCD; 74% of psychiatrists conditions has a long and controversial past. In the in the same survey indicated that they would con- midst of gradually emerging newer information sider referring appropriate patients. So, despite and cautious optimism about the potential for the advances in conventional behavior therapy newer procedures, the adverse consequences of and medication treatments that had occurred by psychiatric neurosurgery in the middle of the the end of the twentieth century, there remained twentieth century remain in the minds of physi- a recognition that neurosurgery might be appro- cians, psychotherapists, patients, and society at priate for a small group of patients with otherwise large. The powerful social and scientific legacies of intractable illness. Anecdotal experience of one of the past indiscriminant use of prefrontal lobotomy us found a similar high degree of awareness of are considered at length in the article by Koppell neurosurgery for intractable OCD recently. At and Rezai in this issue. Later in this issue, Fins a 2002 psychopharmacology review course, 85% considers ethical implications emerging from this of an audience of 124 psychiatrists responded that history in light of present developments. That they knew of psychiatric neurosurgery for OCD. article, and an editorial statement by the OCD- In this unsystematic sample, the majority (68%) DBS Collaborative Group [26] reprinted in this indicated that they would consider referring pa- issue, focus especially on recommendations for tients with intractable illness (B. Greenberg, un- future research in this area. To introduce some key published observations). issues here, a few features of the early experience Importantly, acceptance for surgical treatment and literature are briefly mentioned. by the few centers with specialized teams of psychiatric and neurosurgical experts requires Some lessons of history that patients meet rigorous entry criteria. Patients must be capable of fully informed consent. Care- Reports of the deleterious sequelae of radical ful multidisciplinary review is undertaken to estab- destructive operations have had a lasting impact. lish accuracy of diagnoses and that the illness is A negative perspective persists, overshadowing treatment refractory. For OCD, this includes, in any benefits that accrued. Irreversible and some- part, establishing that definitive medication trials times devastating adverse effects were common and behavior therapy conducted by clinicians after lobotomy. Nevertheless, a careful reading expert in treating refractory illness have failed to of that early literature also leads to the conclusion provide adequate benefit [22,23]. Similarly, a de- that some patients were helped by these pro- tailed review of the response to prior treatment cedures: trials, and the adequacy of those trials, is made A woman, who for some thirty years had for patients with intractable depression. There are suffered from obsessive fear of contamina- important absolute and relative contraindication and who had scrubbed not only the tions and procedural safeguards. Prospective data toilet seat but the whole bathroom for an on potential adverse effects on cognition and hour or so before using the toilet and then personality are systematically collected before for an hour or more afterward in an effort to spare others from the danger of con- and for years after the procedures. These issues tamination... Following her operation, for are discussed in more detail at the end of this a long period this woman manifested the article. same tendency toward compulsive cleaning Awareness of current neurosurgical procedures of the bathroom before and after evaluation among psychiatrists occurs against a background even though she admitted that she did not of advances in our understanding of the neuro- feel the same anxiety and fear of contam- anatomic bases of OCD and, increasingly, de- inating others that had previously been pression (see the article by Rauch in this issue) present. The compulsive activity gradually [24,25]. Recent US Food and Drug Administra- disappeared during the ensuing years. Free- tion (FDA) approvals of deep brain stimulation man and Watts, 1950 [27]. for treatment-refractory tremor and Parkinson This case vignette illustrates several important disease and increasing therapeutic use of these issues. One is that OCD, particularly when severe, techniques in other countries have also enhanced is generally a chronic illness. Another point is that consideration of the potential of neurosurgery for the first effect of the operation was a reduction intractable psychiatric illness. in anxiety, even though compulsive behavior B.D. Greenberg et al / Neurosurg Clin N Am 14 (2003) 199–212 203 continued. The response to treatment was usually Development of modern lesion procedures slow, especially in severely affected individuals. In Speigel and Wycis, who began stereotactic this example, the compulsive cleaning gradually neurosurgery in patients, were the first to report abated over a period of years, despite the radical that dorsomedial thalamotomies improved obses- nature of that operation. For the much more focal sive-compulsive symptoms. Their stereotactic pro- lesion procedures used currently, the best evidence cedure was much less radical than lobotomy but is also that maximal improvement takes months still dangerous. Because the lesioning electrodes to years (see the article by Cosgrove et al in this were placed in a highly vascular structure without issue) [28]. The reasons for this are poorly modern imaging guidance, hemorrhages were understood and likely multifactorial. Research frequent and 10% of the patients died, mainly on neurosurgical treatments for intractable OCD for this reason [30]. Later, during the 1950s and should take into account that the course of re- 1960s, several groups of neurosurgeons and sponse is likely to be prolonged and may depend psychiatrists, mainly in Europe, explored the on the availability of, and adherence to, post- therapeutic effects of selective lesions made under surgical behavior therapy. stereotactic guidance. The development of these Freeman and Watts depicted responses to techniques was informed at first not by specific lobotomy like that cited previously in terms of empiric evidence but by the general ideas that the dominant psychodynamic model of the day: frontosubcortical (and particularly frontothala- We have compared the emotion to the mic) connections were important in higher brain fixing agent that prevents a photographic function and that limbic networks modulate image from fading back into obscurity. emotion [22,31–34]. Remove the emotion and the image grad- The aim was to sever connections between ually fades. Prefrontal lobotomy bleaches the affect attached to the ego. Freeman and subcortical structures and the frontal lobes. Watts, 1950 [27]. The procedures developed most successfully (Fig. 1) were subcaudate tractotomy, anterior Although not intended by the authors, this capsulotomy, cingulotomy, and limbic leucotomy understanding also hinted at the possibility that (a combination of subcaudate tractotomy and emotional blunting after neurosurgery could go cingulotomy). Subcaudate tractotomy and anterior well beyond a reduction in distress caused by capsulotomy in particular interrupt connections of OCD symptoms. In 1947, Rylander [29] gave a orbital and medial prefrontal cortex to the thala- compelling glimpse of this and other adverse mus. The target of anterior cingulotomy, the most effects after the Freeman and Watts prefrontal widely performed and recognized procedure in the lobotomy procedure. He studied changes in United States, is within cingulate cortex itself. personality and cognition in great detail over time All these operations remain in use for a small in a series of eight patients, in some cases, by number of patients with intractable neuropsychi- having a patient join his own household. In one atric disorders. Their primary indications are inpatient, a 28-year-old woman suffering from tractable OCD and major depression. Effects of ‘‘anxiety periods, with compulsive and hysterical these procedures on a small number of patients fits,’’ the resulting catastrophically diminished with severe non-OCD anxiety disorders have also emotional capacity after lobotomy is vividly been reported. Although developed before the era described by her mother: of functional and modern structural neuroimag- She is my daughter but yet a different ing, each of these techniques would be expected person. She is with me in body but her soul to affect activity within networks suggested by is in some way lost. Those deep feelings, neuroimaging studies to be important in OCD those tendernesses, are gone. G. Rylander, and depression (see the article by Rauch in this 1947 [29]. issue). It is because of such adverse effects that What follows is a relatively brief description of lobotomy was permanently abandoned. One of these four procedures. The evidence of their safety the crucial lessons of this history is that research and efficacy is presented in subsequent sections. on psychiatric neurosurgery must systematically The reader is referred to the article by Cosgrove assess patients for potential changes in personal- and colleagues in this issue for recent perspectives ity, including emotional responsiveness and moti- based on their experience in Boston with anterior vation. cingulotomy. 204 B.D. Greenberg et al / Neurosurg Clin N Am 14 (2003) 199–212

operations were performed in the United King- dom from the period of its ﬁrst development until the early 1990s. There is a more recent report of a modiﬁed technique of subcaudate tractotomy [36].

Anterior capsulotomy This procedure targets the fiber bundles in the anterior limb of the internal capsule connecting the frontal lobes and thalamus [37,38]. Talairach and co-workers were the first to make selective lesions there. Although therapeutic effects in schizophrenia were considered unsatisfactory, results in patients with severe anxiety were better. Capsulotomy was further developed and used in a large series of patients by Lars Leksell and colleagues at the Karolinska Institute in Sweden, starting in the 1950s. After craniotomy, thermocoagulation lesions were made bilaterally using bipolar electrodes placed in the anterior third of the capsule. This procedure is now called open capsulotomy or thermocapsulotomy, in contrast to the newer technique of gamma knife capsulotomy. The gamma knife procedure has been the focus of ongoing research in Providence, Rhode Island, over the past decade. The particular intent of gamma knife capsulotomy in the United States has been to target connections between dorsomedial thalamus and orbital and medial prefrontal cortex. In the United States, gamma knife anterior capsulotomy has been used almost exclusively for intractable OCD. As discussed later in this article, trials of deep brain stimulation at the capsulotomy target site for intractable OCD are underway.

Anterior cingulotomy

Fig. 1. Neurosurgical procedures for intractable obses- Originally conceived by Fulton as a treatment sive-compulsive disorder and depression. for psychiatric disorders, cingulotomy’s first use was actually for intractable pain. The procedure was later applied to psychiatric disorders when Subcaudate tractotomy mood and anxiety symptoms were found to In this procedure, lesions intended to interrupt improve in pain patients. Whitty et al [39] first orbitofrontal-subcortical connections are made reported the effects of cingulotomy in OCD, under the head of the caudate nucleus in the followed by Kullberg [40] and his colleagues. It substantia innominata [35]. This approach was is the work of Ballantine and his colleagues at the developed by Geoffrey Knight in the United Massachusetts General Hospital in Boston, how- Kingdom in 1964 as an attempt to limit adverse ever, that is responsible for cingulotomy being the effects by restricting lesion size. Radioactive best known and most widely used procedure for yttrium-90 seeds were placed at targets under the intractable psychiatric illness in North America. head of the caudate nucleus. A total lesion volume Beginning in 1962, this group demonstrated that of approximately 4 mL resulted. In addition to anterior cingulotomy had a favorable safety pro- intractable OCD and depression, subcaudate file. This investigative team has performed approxi- tractotomy has also been used for other severe mately 1000 cingulotomies, studying its efficacy anxiety disorders. More than 1300 of these for a range of psychiatric indications. Current B.D. Greenberg et al / Neurosurg Clin N Am 14 (2003) 199–212 205 indications for anterior cingulotomy include in- measures taken after surgery, several independent tractable pain, depression, and OCD. The targets groups of investigators have evaluated the effects of for this procedure are located in the anterior cin- capsulotomy, cingulotomy, subcaudate tracto- gulate cortex (Brodmann areas 24 and 32) adja- tomy, and limbic leucotomy. Persistent deteriora- cent to the underlying fibers of the cingulum tion of intellectual function has been relatively rare bundle. Under local anesthesia, thermocoagula- in patients with severe OCD and depression who tion electrodes are used to make lesions on each underwent these procedures [23,45,46]. Moreover, side through bilateral burr holes. Initially, ven- in several instances, improved performance on triculography was used to guide lesion placement. cognitive measures has been documented, pre- In 1991, this was replaced by MRI guidance. Two sumably as a result of symptomatic improvement. or three sets of bilateral lesions are made using radiofrequency electrodes (see the article by Cos- Subcaudate tractotomy grove and Rauch in this issue). Because the intent Adverse effects is to produce the smallest effective lesion, the pro- Postoperative side effects were reported to cedure is often done in stages. About 40% of include headache, confusion, or somnolence, patients return months later for a lesion-extending typically lasting 1 week at most, in a 1975 study second operation to enhance efficacy. As currently of 208 patients who underwent subcaudate practiced, the resulting total lesion volume is in the tractotomy, mainly for depression but including range of 4 to 6 mL [41]. some OCD patients. Individuals were followed for Limbic leucotomy a mean of 2.5 years. Transient disinhibition after surgery was described as common. Longer term Kelley and colleagues [42] developed this multi- adverse effects included mild untoward personal- target procedure in the 1970s, which, in essence, ity change in 6.7% of patients. Seizures were combines the bilateral lesions of cingulotomy with reported in 2.2%. One death occurred as a direct those of subcaudate tractotomy. The latter set of complication of surgery, resulting from migration lesions may have a more anterior placement than is of one of the yttrium seeds. Three of the 208 typical for subcaudate tractotomy, however. In- patients died by suicide during the follow-up terestingly, the effects of intraoperative electric period. A later review reported on 1300 patients stimulation have been used to identify the surgical who had subcaudate tractotomy up until 1993 target for this procedure. Lesions at sites where [20]. Again, intractable depression was the most stimulation induced marked autonomic changes common diagnosis; a smaller number of patients were believed to be the most effective [43]. Thermo- had the surgery for intractable OCD. The rate of coagulation or cryoprobes are used. Indications seizures was similar to that reported previously. In for limbic leucotomy have historically been in- contrast, persistent adverse personality changes tractable depression, OCD, and some other severe were not found. There were no deaths as anxiety disorders. Recent evidence suggests that a complication of surgery. Compared with a com- limbic leucotomy may also be of benefit to patients parison group of patients with major affective with severe repetitive self-injurious behaviors disorders, the subcaudate tractotomy group occurring in the setting of severe tic disorders [44]. showed a markedly lower suicide rate: 1% after surgery compared with 15% in patients with Safety affective illness who were not treated surgically. The adverse effect profiles of the more focal Anterior capsulotomy surgical interventions of the past 40 years have been notably more benign than that of lobotomy. The Adverse effects major operative complications of the open neuro- Adverse effects of open anterior capsulotomy surgical approaches have included infection, hem- in the initial series of 116 patients described by orrhage, seizures, and weight gain. Such side effects Leksell and colleagues [47] included postoperative have been relatively rare. The risk of postoperative headache, incontinence, or confusion. The dura- epilepsy has been estimated at less than 1%. The tion of confusion, which frequently lasted as long risk of changes in cognitive function and person- as 1 week, influenced the length of the hospital stay. ality after these procedures has been carefully In a later group of 24 patients with OCD studied. Using comprehensive batteries of meas- followed prospectively, side effects of thermocap- ures collected before surgery and compared with sulotomy included one intraoperative hemorrhage 206 B.D. Greenberg et al / Neurosurg Clin N Am 14 (2003) 199–212 without neurologic sequelae and 1 patient who secondary to intraoperative hemorrhage in the era developed seizures. There were transient episodes before image-guided surgery was estimated to be of confusion during the first week in 19 of 22 0.03%. There has been only one stroke in the era patients available for follow-up as well as occa- of MRI-guided cingulotomy. Side effects in the sional nocturnal incontinence. Fatigue was pres- immediate postoperative period include headache, ent in 7 patients (29%), 4 (17%) described poor nausea, and difficulty with urination, usually re- memory, and 2 patients (8%) had ‘‘slovenliness.’’ solving within days. Seizure incidence has ranged One patient committed suicide in the postopera- from 1% to 5%. This effect was seen particularly tive phase, and 8 patients suffered from depres- in patients with a prior seizure history. sion severe enough to require treatment. Excessive In the most recent series [28], 9 of 44 patients fatigue was a complaint in 7 patients, and 4 had (20%) had at least one adverse effect after cin- poor memory. Weight gain was common after gulotomy. In two cases, sequelae were enduring: open capsulotomy, with an average increase of seizures responsive to ongoing anticonvulsant about 10% in patients in this sample [38]. treatment and worsening of preexisting urinary In another report, no significant cognitive incontinence as a result of prostate cancer. Another dysfunction or adverse personality changes were patient developed edema and hydrocephalus re- found on a psychometric test battery administered quiring ventriculostomy. In addition, 2 patients to a sample of 200 capsulotomy patients [46]. reported worsened memory, and 1 patient de- In contrast, a small study found perseverative scribed apathy and decreased energy. Those be- responses to be more common after thermocap- havioral symptoms resolved within 1 year after sulotomy in 5 patients with a severe, treatment- surgery. refractory, non-OCD anxiety disorder [48]. Although transient memory problems were Differences in rates of adverse effects across reported by up to 5% of patients overall, an studies seem due, at least partly, to differences in independent analysis concluded that no significant the volume of tissue lesioned, although this is not cognitive or behavioral impairments occurred in fully clear in published reports. The same seems a series of 34 patients undergoing cingulotomy for to be true for procedure efficacy; that is, the intractable psychiatric illness. A later study of 57 effectiveness and the side effect burden of the open additional patients reached the same conclusion. procedure both appeared to increase with greater [41,49–51]. In fact, in these reports, patients are lesion volume [41]. noted to exhibit improved cognitive function, Most recently, the Providence group has found perhaps because symptom reduction after cingu- that gamma capsulotomy was generally well lotomy facilitated test performance. One study tolerated and effective for patients with otherwise [52] described subtle impairment of attention, intractable OCD. The lesions resulting from the however, and another [53] described mild alter- gamma capsulotomy procedure are generally ations of intention and self-initiated action after smaller than those produced by open thermocap- cingulotomy for chronic pain. It has been specu- sulotomy. Adverse events included transient cere- lated [28] that cingulotomy may be less likely to bral edema and headache (in 6 of 31 patients produce cognitive deficits in OCD or major de- [20%]), small asymptomatic caudate infarctions pression because anterior cingulate dysfunction (3 of 31 patients [10%]), and possible exacerba- may already be intrinsic to those disorders. tion of preexisting bipolar mania (2 of 31 patients In that series [28], one patient (2%) committed [6%]). No group decrements were observed on suicide approximately 6 years after cingulotomy. cognitive or personality testing. Nevertheless, 1 of This patient’s OCD symptoms had improved after 31 patients (3%) developed a persistent mild frontal surgery. Before cingulotomy, the patient had lobe syndrome, including apathy and amotivation a long history of severe depression, with more (S. Ramussen et al, manuscript in preparation). than 8 years of nearly continuous suicidal thinking and a prior suicide attempt. Anterior cingulotomy Limbic leucotomy Adverse effects In the Massachusetts General Hospital experi- Adverse effects ence of approximately 1000 anterior cinguloto- Kelly and his collaborators have described the mies, there have been no deaths from the surgical effects of limbic leucotomy in several reports. procedure itself [41]. The incidence of hemiplegia In an initial prospective study of 66 patients, no B.D. Greenberg et al / Neurosurg Clin N Am 14 (2003) 199–212 207 seizures and no deaths resulted from the pro- Subcaudate tractotomy cedure. Early postoperative side effects included Effectiveness headache, lethargy, apathy, and incontinence Depression has been the most common diagno- lasting from days to weeks. Postoperative confu- sis for patients undergoing this procedure. In an sion lasted for days at least; patients frequently early report, Strom-Olsen and Carlisle [56] de- remained hospitalized for more than 1 week for scribed beneficial effects in depressed patients who this reason. Patients were subsequently followed underwent stereotactic subcaudate tractotomy. for an average of 16 months. One patient had A subsequent report from this group, in which severe memory impairment attributable to im- structured interviews were used, described a 55% proper lesion placement. Twelve percent of pa- response rate in a total of 96 depressed patients tients had persistent lethargy in that series. IQ operated on through 1973 and followed for 2.5 testing showed a slight improvement for this years on average [35]. More recently, in their group after limbic leucotomy [54]. review of the same group’s experience from 1979 A recent report of 21 patients undergoing to 1991, Hodgkiss et al [57] classified 34% of limbic leucotomy for OCD or depression at depressed patients as ‘‘recovered’’ or ‘‘well’’ and Massachusetts General Hospital [55] found the 32% as ‘‘improved’’ of a total 183 such patients adverse effects noted in previous reports, includ- who had undergone the surgery. Malizia [58] ing apathy, urinary incontinence, and memory found similar rates of response. impairment. These side effects were infrequent OCD patients also reportedly benefit from and transient. subcaudate tractotomy. Strom-Olsen and Carlisle [56] reported that of 20 OCD patients, 10 were either fully recovered or with only slight residual Effectiveness symptoms 3 months after subcaudate tractotomy. Four of these patients subsequently relapsed over Reports on efficacy of current neurosurgical a 2-year follow-up period. Goktepe et al [35] procedures for intractable psychiatric illness span subsequently described a response rate of 50% in a period of more than 40 years. Psychiatric and a second sample of OCD patients after subcau- neurosurgical methods used have therefore varied date tractotomy. A response rate of 62.5% in over time. Several limitations result. Illness patients with severe non-OCD anxiety disorders definitions were not necessarily consistent across was also noted in that sample. In contrast, efficacy sites or over time. Outcome measures have also was poor for patients diagnosed with schizophre- differed. Furthermore, effective medication and nia, substance abuse, or personality disorders. cognitive-behavioral treatments only appeared The most comprehensive review of responses after the earliest reports. Because it was later to subcaudate tractotomy included 1300 cases. required that these be systematically tried before Published in 1994, it concluded that subcaudate patients would be eligible for surgery, only the tractotomy enabled 40% to 60% of patients to more recent studies have enrolled patients most lead normal or near-normal lives [20]. As noted similar to those who would be potential candi- previously, compared with a suicide rate of 15% dates for such procedures today. Furthermore, in patients with similar major affective disorders, earlier reports of the effectiveness of these proce- 1% of patients committed suicide after subcau- dures were retrospective and typically described date tractotomy. Those investigators suggested relatively small patient samples. More recent in- that several clinical features were predictors of vestigations of these treatments are prospective, positive response. These included (1) major de- with generally larger sample sizes. pression, (2) an onset that is sudden or occurs in Direct comparisons of any two procedures midlife or the peripartum period, (3) a positive at the same center, for example, the study by family history, and (4) a prior response to ECT Kullberg [40], are rare, and randomized controlled treatment. trials are nonexistent. One reason for this is that sham procedures involving craniotomy have Anterior capsulotomy generally been considered unethical. It is important in this regard that the newer procedures of Effectiveness gamma knife capsulotomy and deep brain stim- Leksell’s group initially reported that half of ulation lend themselves more easily to controlled those patients with obsessional neurosis and just trials. less than half (48%) of depressed patients (of 208 B.D. Greenberg et al / Neurosurg Clin N Am 14 (2003) 199–212 a total sample of 116, which also included patients 3-year follow-up, 10 of 16 (62%) met this response with schizophrenia and nonobsessional anxiety) criterion. Anecdotally, after surgery, adherence to had satisfactory outcomes. Although modern and success of behavioral therapy seemed much rating scales were not used, criteria for judging enhanced in responders. Overall, improvement, improvement were strict. Only patients who were judged by conservative criteria, was faster and free of symptoms or much improved were judged occurred in more patients after the one-stage, to be responders. Herner [47] reported that the double, bilateral gamma capsulotomy procedure. obsessional patients in this series benefited most The therapeutic response, once achieved, was after capsulotomy. After follow-up ranging from generally stable. 2 to 6 years, outcome was ‘‘good’’ or ‘‘fair’’ in 14 of 18 patients (78%) who underwent open capsu- Anterior cingulotomy lotomy in the 1950s. A study by Bingley et al [37] found that 25 of 35 patients (71%) were either Effectiveness symptom-free or much improved an average of 35 Whitty and colleagues [39] were the first to months after thermocapsulotomy. Twenty-four of report the effects of cingulotomy in OCD. Four of these patients had been unable to work before the initial sample of five patients showed signif- surgery because of OCD symptoms; 20 resumed icant improvement in symptoms and function. work after surgery. Later, Kullberg [40] reported that 4 of 13 (31%) Mindus and Jenike [59] retrospectively re- OCD patients improved significantly after cingu- viewed all cases of capsulotomy reported by the lotomy. In a larger sample, Ballantine and early 1990s. They judged that 64% of 213 patients colleagues [62] reported marked improvement in for whom adequate information was available 17 of 32 (35%) obsessional patients after cingu- could be considered responders. Response could lotomy. Of those, 8 were almost symptom-free not be determined in 149 of the total of 362 and 9 others had marked symptom reduction. patients, however, and it is not fully clear if the Including these individuals and other patients patients for whom response could be determined judged to have significant but less marked benefit, were representative of the entire sample. a total of 56% of OCD patients noted significant A later prospective study of capsulotomy for improvement an average of 8.6 years after the intractable OCD found that 16 of 35 patients procedure [62]. A later study using more current (46%) were judged symptom-free by independent methods of diagnosis and severity assessment was psychiatrists and that 9 more were much im- reported by Jenike and colleagues [63]. They proved, giving a response rate of 70% overall [60]. retrospectively evaluated 33 patients meeting The effectiveness of lesion enlargement in OCD criteria for OCD who had undergone cingulotomy patients unresponsive to an initial thermocapsu- at the Massachusetts General Hospital over a 25- lotomy was addressed by Burzaco [61]. He year period. Using the Y-BOCS and the Clinical reported that in 17 OCD patients of a total of Global Improvement Scale, the authors estimated 85 who did not respond to the first procedure, half that 25% to 30% of the patients benefited were judged to have a satisfactory outcome after substantially from the operation. In a subsequent reoperation. prospective study of 18 severely ill treatment- Preliminary findings from an ongoing study of refractory OCD patients, they found that 5 of 18 anterior capsulotomy performed using the gamma (28%) met conservative criteria for being much or knife also find evidence of efficacy (S.A. Rasmussen very much improved. Three of the 18 patients had et al, manuscript in preparation). In the first series lesser but still notable benefit, giving an overall of 15 patients, single bilateral lesions in the anterior rate of significant improvement of 44% [63]. capsule lacked therapeutic effects. After placement In the most recent prospective study, 44 OCD of a second set of bilateral lesions, improvement patients were studied using current methodolo- began to occur in some patients. Using a conserva- gies, including rigorous screening and review. tive definition of therapeutic response, 4 of 15 Empirically validated symptom and quality of life patients had at least a 35% drop on the Yale-Brown assessments were used. Thus, the diagnosis was Obsessive-Compulsive Scale (Y-BOCS) plus a min- certain to be intractable OCD (ie, severe OCD imum 15-point improvement on the Global As- refractory to adequate trials of available treat- sessment Scale on 5-year follow-up. ments). At a mean of 32 months after one or more A subsequent group of 16 patients received two cingulotomies, 14 patients (32%) met conservative pairs of bilateral lesions during one session. At the criteria for treatment response and 6 others (14%) B.D. Greenberg et al / Neurosurg Clin N Am 14 (2003) 199–212 209 were partial responders. Thus, 20 patients (45%) about 35%–70%) of patients with intractable were at least partial responders at long-term OCD and depression. The literature also suggests follow-up after one or more cingulotomies [28]. that although serious long-term adverse events For depression, Ballantine et al [62] reported have occurred, these are relatively infrequent over- that of 118 depressed patients treated with all. Methodologic limitations of the earlier reports stereotactic cingulotomy, 42% were ‘‘recovered’’ on any of these procedures were described pre- or ‘‘well’’ and 24% were ‘‘improved.’’ A later viously in this article. The major academic centers review of 34 patients who underwent cingulotomy conducting this work have since been obtaining in the era of MRI guidance found that 60% of systematic prospective data using modern assess- patients with unipolar depression had favorable ment tools. Nevertheless, even with improved responses. methodologies, more recent studies confront some remaining issues that have been difficult to over- Limbic Leucotomy come fully. First, the number of patients who have re- Effectiveness ceived any one procedure has been relatively An initial report from Mitchell-Heggs and small, constraining statistical power. This limits colleagues [54] described 89% of 27 OCD patients the ability of researchers to enhance patient as improved at a postoperative follow-up of 16 selection based on clinical characteristics. This is months. Bartlett and Bridges [64] later disputed important, because patients with intractable OCD these results, noting that patients with significant and depression referred for neurosurgery have residual symptoms were included in the 89% high rates of comorbid Axis I diagnoses, person- improvement rate. Kelly [65] then extended his ality disorders, and functional impairments, which group’s original findings and reported on 49 may have value in predicting response. Other patients, including the group of 27 described in features, such as age of onset, chronicity, and the original report by Mitchell-Heggs et al [54]. symptom subtypes, may be likewise useful. He found that 84% of the patients were improved Another key factor in response may be post- at a mean follow-up of 20 months [65]. operative management, which has varied most In their initial report, Mitchell-Heggs et al [54] over time but also across patients enrolled in also described a 78% response in patients with trials. As noted previously, randomized controlled major depression. The sample sizes were relatively trials of neurosurgical treatment for intractable small, however, and the patients were less well psychiatric illness have not been reported, al- characterized than in some studies. though one has been proposed for gamma knife Using modern diagnostic and symptom assess- capsulotomy in intractable OCD [23]. The de- ment procedures, Montoya and colleagues [55] velopment of deep brain stimulation has also found positive responses (conservatively defined) made sham-controlled studies possible and also in 36 or 50% of 21 patients who underwent limbic allows within-patient designs to be considered. leucotomy for OCD or depression at the Massa- Bearing these problems in mind, the literature chusetts General Hospital after a mean follow-up does provide important guidance on a number of of 26 months. key points, including approaches to referral, Price and colleagues [44] recently reported patient selection, and the need for long-term beneficial responses to limbic leucotomy in five prospective follow-up and postoperative manage- patients with otherwise intractable severe repeti- ment. Nevertheless, important gaps in knowledge tive self-mutilation occurring in the setting of tic- remain in all these areas. Research is expected to like behaviors. narrow these gaps in a number of ways, including patient selection, optimizing the procedures themselves, and understanding the mechanisms of Summary therapeutic action. Neuroimaging studies will play Intractable OCD and depression cause tremen- a key role in achieving these aims (see the article by dous suffering in those affected and in their Rauch in this issue). So will cross-species trans- families. The impaired ability to function of those lational research on the anatomy and physiology of affected imposes a heavy burden on society as the pathways implicated in the pathophysiology a whole. and response to treatment in these disorders. Existing data suggest that lesion procedures Future research in psychiatric neurosurgery offer benefit to a large proportion (ranging from must proceed cautiously. A recent editorial 210 B.D. Greenberg et al / Neurosurg Clin N Am 14 (2003) 199–212 statement of the OCD-DBS Collaborative Group during 15 years of observational follow-up. Am J [26] recommends a minimum set of standards for Psychiatry 1999;156:1000–6. any multidisciplinary teams contemplating work [13] Klerman GL, Weissman MM. The course, morbid- in this domain. The rationale for those standards ity, and costs of depression. Arch Gen Psychiatry is found throughout this issue and is especially 1992;49:831–4. [14] American Psychiatric Association. Practice guide- developed in the article by Fins. The need for safe line for the treatment of patients with major depres- and effective therapeutic options for people sive disorder. Am J Psychiatry 2000;157(Suppl):1–5. suffering with these severe illnesses is just as clear. [15] Agency for Healthcare Policy Research. Evidence The experience over the last several decades report on treatment for depression—newer phar- provides grounds for careful optimism that re- macotherapies. Evidence-Based Practice Centers: fined lesion procedures or reversible deep brain San Antonio, TX; 1999. stimulation may relieve suffering and improve the [16] Nelson JC. Treatment of refractory depression. lives of people with these devastating disorders. Depress Anxiety 1997;5:165–74. [17] Fava M, Davidson KG. Definition and epidemiology of treatment-resistant depression. Psychiatr References Clin North Am 1996;19:179–200. [18] Little JT, Reynolds CF III, Dew MA, et al. How [1] Rasmussen SA, Eisen JL. The epidemiology and common is resistance to treatment in recurrent, clinical features of obsessive compulsive disorder. nonpsychotic geriatric depression? Am J Psychiatry Psychiatr Clin North Am 1992;15:743–58. 1998;155:1035–8. [2] Narrow WE, Rae DS, Robins LN, Regier DA. [19] Snaith RP, Price DJ, Wright JF. Psychiatrists’ Revised prevalence estimates of mental disorders in attitudes to psychosurgery. Proposals for the the United States: using a clinical significance cri- organization of a psychosurgical service in York- terion to reconcile 2 surveys’ estimates. Arch Gen shire. Br J Psychiatry 1984;144:293–7. Psychiatry 2002;59:115–23. [20] Bridges PK, Bartlett JR, Hale AS, Poynton AM, [3] Skoog G, Skoog I. A 40-year follow-up of patients Malizia AL, Hodgkiss AD. Psychosurgery: stereo- with obsessive-compulsive disorder [see comments]. tactic subcaudate tractotomy. An indispensable Arch Gen Psychiatry 1999;56:121–7. treatment. Br J Psychiatry 1994;165:599–612. [4] Nestadt G, Samuels J, Riddle M, et al. A family [21] Mathew SJ, Yudofsky SC, McCullough LB, study of obsessive-compulsive disorder. Arch Gen Teasdale TA, Jankovic J. Attitudes toward neuro- Psychiatry 2000;57:358–63. surgical procedures for Parkinson’s disease and [5] Dupont R, Rice D, Shiraki S. The economic burden obsessive-compulsive disorder. J Neuropsychiatry of obsessive compulsive disorder. Pharmacoeco- Clin Neurosci 1999;11:259–67. nomics Med Interface 1995;8:102–9. [22] Sachdev P, Sachdev J. Sixty years of psychosurgery: [6] Murray CJ, Lopez AD. Global mortality, disability, its present status and its future [see comments]. Aust and the contribution of risk factors: Global Burden NZ J Psychiatry 1997;31:457–64. of Disease Study. Lancet 1997;349:1436–42. [23] Jenike MA, Rauch SL, Baer L, Rasmussen SA. [7] Eisen JL, Goodman WK, Keller MB, et al. Patterns Neurosurgical treatment of obsessive-compulsive of remission and relapse in obsessive-compulsive disorder. In: Jenike MA, Baer L, Minichiello WE, disorder: a 2-year prospective study. J Clin Psychi- editors. Obsessive-compulsive disorders: practical atry 1999;60:346–52. management. St. Louis: Mosby; 1998. p. 592–610. [8] Rasmussen SA, Eisen JL. Treatment strategies for [24] Mayberg HS, Liotti M, Brannan SK, et al. Re- chronic and refractory obsessive-compulsive dis- ciprocal limbic-cortical function and negative order. J Clin Psychiatry 1997;58(Suppl 13):9–13. mood: converging PET findings in depression and [9] Hollander E, Bienstock CA, Koran LM, et al. normal sadness. Am J Psychiatry 1999;156:675–82. Refractory obsessive-compulsive disorder: state-of- [25] Drevets WC. Neuroimaging studies of mood the-art treatment. J Clin Psychiatry 2002;63(Suppl disorders. Biol Psychiatry 2000;48:813–29. 6):20–9. [26] OCD-DBS Collaborative Group. Deep brain stimu- [10] American Psychiatric Association Committee lation for psychiatric disorders. Neurosurgery 2002; on Nomenclature and Statistics. Diagnostic and 51:519. statistical manual of mental disorders. 4th edition. [27] Freeman W, Watts JW. Psychosurgery in the American Psychiatric Association: Washington, treatment of mental disorders and intractable pain. DC; 1994. Charles C. Thomas: Springfield; 1950. [11] Lehtinen V, Joukamaa M. Epidemiology of de- [28] Dougherty DD, Baer L, Cosgrove GR, et al. pression: prevalence, risk factors and treatment sit- Prospective long-term follow-up of 44 patients uation. Acta Psychiatr Scand Suppl 1994;377:7–10. who received cingulotomy for treatment-refractory [12] Mueller TI, Leon AC, Keller MB, et al. Recurrence obsessive-compulsive disorder. Am J Psychiatry after recovery from major depressive disorder 2002;159:269–75. B.D. Greenberg et al / Neurosurg Clin N Am 14 (2003) 199–212 211

[29] Rylander G. Personality analysis before and after [45] Greenberg BD, Murphy DL, Rasmussen SA. frontal lobotomy: the frontal lobes. Archives of Neuroanatomically based approaches to obsessive- Nervous and Mental Disease 1948;27:691–705. compulsive disorder. Neurosurgery and transcranial [30] Spiegel EA, Wycis HT, Freed MD, Orchinik MA. magnetic stimulation. Psychiatr Clin North Am The central mechanism of the emotions. Am J 2000;23:671–86. Psychiatry 1951;121:426–32. [46] Nyman H, Andreewitch S, Lundback E, Mindus P. [31] MacLean PD. The triune brain in evolution: role in Executive and cognitive functions in patients with paleocerebral functions. Plenum Press: New York; extreme obsessive-compulsive disorder treated by 1990. p. 580–635. capsulotomy. Appl Neuropsychol 2001;8:91–8. [32] MacLean PD, Kral VA. A triune concept of the [47] Herner T. Treatment of mental disorders with brain and behaviour. Toronto: University of frontal stereotactic thermo-lesions: a follow-up of Toronto Press; 1973. 116 cases. Acta Psychiatr Scand Suppl 1961;36: [33] Marino R, Cosgrove GR. Neurosurgical treatment 941–67. of neuropsychiatric illness. Psychiatr Clin North [48] Nyman H, Mindus P. Neuropsychological corre- Am 1997;20:933–4. lates of intractable anxiety disorder before and [34] Papez JW. A proposed mechanism of emotion. after capsulotomy. Acta Psychiatr Scand 1995;91: Neuropsychiatry Clin Neurosci 1937;7:103–12. 23–31. [35] Goktepe EO, Young LB, Bridges PK. A further [49] Ballantine HT Jr. Neurosurgery for behavioral review of the results of sterotactic subcaudate disorders. In: Rengachary S, editors. Neurosurgery. tractotomy. Br J Psychiatry 1975;126:270–80. New York: Elsevier/North Holland Biomedical [36] Malhi GS, Bartlett JR. A new lesion for the Press; 1985. p. 2527–37. psychosurgical operation of stereotactic subcaudate [50] Corkin S, Twitchell T, Sullivan E. Safety and tractotomy (SST). Br J Neurosurg 1998;12:335–9. efficacy of cingulotomy for pain and psychiatric [37] Bingley T, Leksell L, Meyerson BA, et al. Long disorders. In: Hitchcock E, Ballantine H, editors. term results of stereotactic capsulotomy in chronic Modern concepts in psychiatric surgery. New York: obsessive-compulsive neurosis. In: Sweet WH, Elsevier/North Holland Biomedical Press; 1979. Obrador Alcalde S, Martıń-Rodrı´guez JG, editors. [51] Corkin S. A prospective study of cingulotomy. In: Neurosurgical treatment in psychiatry, pain, and Valenstein ES, editor. The psychosurgery debate. epilepsy: proceedings of the Fourth World Congress San Francisco: WH Freeman & Company; 1980. p. of Psychiatric Surgery, September 7–10, 1975, 164–204. Madrid, Spain. Baltimore: University Park Press; [52] Cohen RA, Kaplan RF, Moser DJ, Jenkins MA, 1977. p. 287. Wilkinson H. Impairments of attention after [38] Mindus P. Capsulotomy in anxiety disorders: a cingulotomy. Neurology 1999;53:819–24. multidisciplinary study [thesis]. Karolinska Insti- [53] Cohen RA, Kaplan RF, Zuffante P, et al. Al- tute: Stockholm; 1991. teration of intention and self-initiated action [39] Whitty CM, Duffield JE, Tow PM. Anterior associated with bilateral anterior cingulotomy. cingulotomy in the treatment of mental disease. J Neuropsychiatry Clin Neurosci 1999;11:444–53. Lancet 1952;1:475–81. [54] Mitchell-Heggs N, Kelly D, Richardson A. Stereo- [40] Kullberg G. Differences in effects of capsulotomy tactic limbic leucotomy: a follow-up at 16 months. and cingulotomy. In: Sweet WH, Obrador WS, Br J Psychiatry 1976;128:226–40. Martıń-Rodrı´guez JG, editors. Neurosurgical treat- [55] Montoya A, Weiss AP, Price BH, et al. Magnetic ment in psychiatry, pain and epilepsy. Baltimore: resonance imaging-guided stereotactic limbic leu- University Park Press; 1977. p. 301–8. kotomy for treatment of intractable psychiatric [41] Rauch SL, Greenberg BD, Cosgrove GR. Neuro- disease. Neurosurgery 2002;50:1043–52. surgical treatments and deep brain stimulation. In: [56] Strom-Olsen R, Carlisle S. Bi-frontal stereotactic Sadock VA, editor. Kaplan and Sadock’s compre- tractotomy. A follow-up study of its effects on 210 hensive textbook of psychiatry. 8th edition. Phila- patients. Br J Psychiatry 1971;118:141–54. delphia: Lippincott Williams & Wilkins, in press [57] Hodgkiss AD, Malizia AL, Bartlett JR, Bridges [42] Kelly D, Richardson A, Mitchell-Heggs N, Green- PK. Outcome after the psychosurgical operation up J, Chen C, Hafner RJ. Stereotactic limbic of stereotactic subcaudate tractotomy, 1979–1991. leucotomy: a preliminary report on forty patients. J Neuropsychiatry Clin Neurosci 1995;7:230–4. Br J Psychiatry 1973;123:141–8. [58] Malizia AL. The place of stereotactic tractotomy [43] Kelly D. Physiological changes during operations in the management of severe psychiatric disorder. on the limbic system in man. Cond Reflex 1972; In: Katona C, Montgomery S, Sensky T, editors. 7:127–38. Psychiatry in Europe: directions and developments. [44] Price BH, Baral I, Cosgrove GR, et al. Improve- London: Gaskell; 1994. p. 85–94. ment in severe self-mutilation following limbic [59] Mindus P, Jenike MA. Neurosurgical treatment of leucotomy: a series of 5 consecutive cases. J Clin malignant obsessive compulsive disorder. Psychiatr Psychiatry 2001;62:925–32. Clin North Am 1992;15:921–38. 212 B.D. Greenberg et al / Neurosurg Clin N Am 14 (2003) 199–212

[60] Mindus P, Rasmussen SA, Lindquist C. Neurosur- [63] Jenike MA, Baer L, Ballantine T, et al. Cin- gical treatment for refractory obsessive-compulsive gulotomy for refractory obsessive-compulsive disorder: implications for understanding frontal disorder. A long-term follow-up of 33 patients lobe function. J Neuropsychiatry Clin Neurosci [see comments]. Arch Gen Psychiatry 1991;48: 1994;6:467–77. 548–55. [61] Burzaco J. Stereotactic surgery in the treatment of [64] Bartlett JR, Bridges PK. The extended subcaudate obsessive-compulsive neurosis. In: Perris C, Strowe, tractotomy lesion. In: Martı´n-Rodrı´guez JG, editor. G, Jannson B, editors. Biological Psychiatry. Neurosurgical treatment in psychiatry, pain and Amsterdam: Elsevier/North Holland Biomedical epilepsy. Baltimore: University Park Press; 1977, p. Press. p. 1103. 387–98. [62] Ballantine HT Jr, Bouckoms AJ, Thomas EK, Giriunas [65] Kelly D. Anxiety and emotions. Physiological basis IE. Treatment of psychiatric illness by stereotactic and treatment. Charles C. Thomas: Springﬁeld; cingulotomy. Biol Psychiatry 1987;22:807–19. 1980. Neurosurg Clin N Am 14 (2003) 213–223

Neuroimaging and neurocircuitry models pertaining to the neurosurgical treatment of psychiatric disorders Scott L. Rauch, MD Psychiatric Neuroscience Program, Massachusetts General Hospital–East, Bldg. 149, 13th Street, Room 9130, Charlestown, MA 02129, USA

Contemporary neuroimaging methods have for adverse effects is still considerable. Thus, it been influential in advancing neurobiologic mod- would be of great value to find ways to improve els of psychiatric disorders. Providing means for outcome. One goal is to refine neurosurgical measuring indices of human brain structure and treatments for OCD and MD so that they are function in vivo, these tools enable investigators safer and more effective; another goal is to iden- to test hypotheses about pathophysiology, the tify selection criteria, based on predictors of changes associated with treatment, and predictors response, that would enable patients who are of treatment response [1]. In general, it is anti- unlikely to have good outcomes forego the risks cipated that neuroimaging research will sub- and costs of neurosurgical treatment. stantially contribute to the ultimate achievement In fact, despite a resurgence of interest and of a pathophysiology-based diagnostic scheme in accelerated research in this field, little is known psychiatry and also help to elucidate the mecha- regarding the mechanisms by which neurosurgical nisms of action by which treatments have their treatments for OCD and MD have their beneficial effects. Moreover, as the neuroscience of psychi- effects. Given recent advances in neuroimaging atry evolves, it should be possible to rationally and neurocircuitry models of these disorders, develop new and superior treatments as well as however, the time seems ripe for a fruitful syn- to predict treatment outcomes in a manner that thesis of psychiatric neuroimaging and neuro- could optimally guide selection among alternative surgical data. I would propose that there is great treatments for individual patients in the clinical opportunity for synergy at this interface: neuro- setting. imaging and neurocircuitry models of disease can As reviewed in this issue, a variety of neuro- help to guide practical progress in the domain surgical treatments have been developed in an of neurosurgical treatment, and, reciprocally, data effort to help people suffering from severe and from neurosurgical treatment research can pro- otherwise treatment-refractory psychiatric dis- vide feedback to inform evolving models of orders—principally obsessive-compulsive disorder psychiatric disease. (OCD) or major depression (MD). To date, Toward that end, in the current article, I clinical reports suggest that the effectiveness of describe neurocircuitry models of OCD and MD, contemporary neurosurgical treatment for these focusing on relevant neuroimaging data. I then indications is only modest, whereas the potential review the neuroanatomy of psychiatric neurosurgical procedures and related neuroimaging findings. Finally, I present anticipated future E-mail address: [email protected]. directions of research in this field. This article Dr Rauch is supported in part by RO1MH60219 necessarily extends previous reviews that I have and the Massachusetts General Hospital Psychosurgery written, together with my colleagues, on related Fund. topics [2–6].

Neuroimaging and neurocircuitry models motor behavior. This raises the possibility that of obsessive compulsive disorder OCD (as well as other OC spectrum disorders) might not reflect dysfunction within a single seg- Neurocircuitry model of obsessive-compulsive regated CSTC circuit but rather represent a fail- disorder and related disorders ure in the smooth cascade of information across Contemporary neurobiologic models of OCD the various CSTC circuits. For instance, in the and related disorders have focused on cortico- case of OCD, cognitions and motivations to act striato-thalamo-cortical (CSTC) circuitry. In fact, seem to persist (as obsessions with attendant an- the influential reviews by Alexander and col- xiety, respectively) such that motor output fails leagues [7,8] describing segregated parallel CSTC to reset these thoughts and motivations, hence circuits provided an important framework for driving stereotyped motor repetition (compul- subsequent theories of pathophysiology in OCD. sions). Specifically, it has been proposed that the CSTC It is evident that neurobiologic models of OCD circuits involving orbitofrontal cortex (OFC), and related disorders have evolved in parallel with anterior cingulate cortex (ACC), and the caudate our understanding of the relevant basic neuro- nucleus are central to the pathophysiology of science as well as with accrual of data from pa- OCD [6,9]. Further, there is a convergence of tient studies. In this regard, it is useful to review evidence to suggest that some primary pathologic the findings of pertinent psychiatric neuroimaging process within the striatum might underlie the research. CSTC dysfunction in OCD. The prevailing theory suggests that a relative imbalance favoring the Neuroimaging studies of obsessive- direct versus indirect pathways within this cir- compulsive disorder cuitry lead to overactivity (ie, amplification) within OFC and ACC, the caudate nucleus, and There is now an impressive array of different the thalamus resonant with failed striato-thalamic neuroimaging techniques that can be used in inhibition (ie, filtration) within this same circuitry. a complementary fashion to advance psychiatric This basic scheme has been extended to provide neuroscience. Here, I systematically review how a comprehensive model for a group of purportedly these various methods have been employed to related disorders called ‘‘obsessive compulsive develop a cohesive neurobiologic model of OCD. spectrum disorders,’’ encompassing Tourette syn- Morphometric MRI studies have been per- drome (TS), trichotillomania, and body dysmor- formed to test hypotheses regarding regional phic disorder as well as OCD. The ‘‘striatal brain volumes in OCD. In adults with OCD topography model’’ of OC spectrum disorders versus healthy comparison subjects, findings of suggests that these diseases share the attribute of subtle volumetric abnormalities involving the CSTC dysfunction due to primary striatal pathol- striatum have predominated [12–14]. Moreover, ogy and that the clinical picture in each case reflects whereas the caudate has been principally impli- the topography of pathology within the striatum cated in OCD, consistent with the striatal topo- and hence the constellation of dysfunction across graphy model of OC spectrum disorders, studies CSTC circuits [6,10]. To elaborate, the notion is of TS [15,16] and trichotillomania [17] have that OCD and body dysmorphic disorder, the OC principally implicated the putamen or lenticulate, spectrum disorders characterized by intrusive cog- although an initial study of body dysmorphic nitive and visuospatial symptoms, involve caudate disorder has also found a subtle abnormality pathology, whereas TS and trichotillomania, involving the caudate [18]. principally characterized by intrusive sensorimotor Such morphometric methods have only modest symptoms, involve pathology within the putamen sensitivity, and some studies have failed to find and dysfunction of sensorimotor CSTC circuitry. striatal abnormalities in OCD [19]. Magnetic Most recently, pioneering neuroanatomic re- resonance spectroscopy (MRS) provides a means search by Haber and colleagues [11] has provided for measuring the relative concentration of cer- a scheme for considering CSTC function that tain endogenous chemical compounds in vivo. emphasizes a cascading spiral interaction, rather In particular, N-acetyl aspartate (NAA) is an than segregation, across CSTC circuits. This MRS-visible compound thought to serve as model of normal CSTC function suggests a flow a marker for healthy neuronal density. Thus, of information from motivation to cognition to reduced levels of NAA can be detected as S.L. Rauch / Neurosurg Clin N Am 14 (2003) 213–223 215 a possibly more sensitive indication of regional Moreover, pharmacologic induction of anxiety neuronal degeneration or maldevelopment char- attacks in healthy subjects has likewise been acterized by lower neuronal density or sparse associated with increased activity within anterior arborization. Interestingly, a series of MRS paralimbic regions but not in anterolateral OFC studies of OCD have found reduced striatal or the caudate nucleus [37]. NAA [20,21] but normal NAA levels within the Pre- and posttreatment neuroimaging stud- lenticulate [22]. These findings converge well with ies are conducted to test hypotheses regarding morphometric MRI results and are also consistent changes in brain activity profiles associated with with neuroimmunologic research, which indicates successful reduction in symptoms. Several such that some cases of OCD and related disorders can studies have been performed to investigate the occur as a consequence of autoimmune-mediated neural correlates of OCD symptom reduction striatal degeneration [23]. after treatment with serotonergic reuptake inhib- There have been several isolated reports of itors as well as behavior therapies. A series of regional volumetric abnormalities beyond the studies have demonstrated a reduction of activity striatum in OCD. In adults, these have included within OFC, ACC, and the caudate nucleus after gross increases in white matter volume [12,24] and successful treatment of OCD [38–41]. Further, reduced volumes of OFC and amygdala [25]; in in one of these studies, an interregional cor- children, these have included abnormal thalamic relation analysis indicated that at the pretreat- and putamen volumes [26,27]. Similarly, a single ment time point, there was an abnormal positive report of reduced NAA within the ACC in OCD correlation between right OFC and the right has been published [21]. caudate, which was neutralized after successful Functional neuroimaging studies of OCD have treatment [40]. likewise produced convergent results implicating Pretreatment neuroimaging data can also be OFC, ACC, and the caudate. Neutral state (in- used to test hypotheses regarding predictors of cluding nominal resting state) studies have been treatment response. A series of studies of this type performed to test hypotheses regarding baseline in OCD have indicated that pretreatment activity differences in regional brain activity between within OFC predicts subsequent treatment re- patients with OCD and healthy comparison sub- sponse [41–44]. To elaborate, in patients with jects. Across numerous such studies, the most OCD, higher levels of OFC hyperactivity were consistent findings in OCD have been relative associated with poorer subsequent responses to hyperactivity of OFC, ACC, and, to a lesser treatment with serotonergic reuptake inhibitors extent, the caudate nucleus [28–30]. [41–44]. Interestingly, in contrast, higher levels of Symptom provocation methods have been OFC hyperactivity were associated with better sub- used in conjunction with functional imaging to sequent responses to treatment with behavior ther- test hypotheses regarding the brain regions that apy [42]. These findings are consistent with the idea exhibit changes in activity level in association that serotonergic reuptake inhibitors might have with OCD patients experiencing obsessions their antiobsessional effects via their action within and the urge to perform compulsions. Again, several OFC [45] and that behavioral therapy may have studies have yielded convergent findings that most its beneficial antiobsessional effects via extinction consistently indicate increases in activity within mediated by frontoamygdala interactions [6]. OFC, ACC, and the caudate nucleus during the Cognitive activation studies have been per- OCD symptomatic state versus a control state formed in conjunction with functional neuro- [31–34]. Given that such studies are confounded imaging methods to test hypotheses regarding the by nonspecific anxiety, it has been instructive to functional integrity of striato-thalamic circuitry in contrast the brain activity pattern associated with the context of an implicit sequence learning task. OCD symptoms with the patterns reported for Implicit (ie, nonconscious, automatic) sequence other anxiety disorders. In this context, symptom learning is known to be normally mediated by provocation studies of other anxiety disorders, CSTC circuitry, and this function may be right- such as posttraumatic stress disorder and specific lateralized to some extent [46–48]. Interestingly, phobias [35,36], have frequently found activation however, when patients with OCD were studied of ACC and other anterior paralimbic regions, while performing such a task, although they exhi- whereas the recruitment of the caudate nucleus bited a capacity to learn the sequence information and anterolateral OFC seems to be somewhat that was comparable to normal subjects, they more specific to the symptomatic state in OCD. failed to recruit the right striatum in a normal 216 S.L. Rauch / Neurosurg Clin N Am 14 (2003) 213–223 fashion [49,50]. Moreover, OCD subjects showed circuit, prevailing models of MD have focused on aberrant bilateral medial temporal activation (not other critical elements of the limbic system, seen in normal subjects during implicit sequence namely, the amygdala and hippocampus as well learning) in a manner similar to that evidenced as the hypothalamic-pituitary-adrenal (HPA) axis when normal subjects are learning information [2,55–61]. In general, it is useful to consider the consciously [49–51]. These findings in OCD have clinical characteristics of MD as they relate to been replicated and seem to indicate that patients different functional domains, which, in turn, map with OCD are deficient at recruiting the striatum onto a corresponding functional anatomy [57]. in the service of thalamic gating [49,50,52]. MD episodes (both in unipolar MD and bipolar This may explain why patients with OCD have disorder) are characterized by cognitive, motor, intrusions of information into consciousness that and neuroendocrinologic as well as affective dis- might otherwise be processed nonconsciously by turbances. At the cortical level, the cognitive and individuals without OCD. motor deficits of MD may be explained by dys- Taken together, neuroimaging studies of OCD function within a ‘‘dorsal compartment,’’ includ- support a cohesive model. There is hyperactivity ing anterior, dorsal and lateral prefrontal cortex; at rest within the OFC-caudate CSTC circuit, dorsal ACC; and parietal cortex as well as which is exaggerated during symptom provoca- premotor cortex. The affective symptoms of MD tion and attenuated after successful treatment. A may be related to dysfunction within a paralimbic similar profile is present within ACC, although ‘‘ventral compartment,’’ including subgenual this seems to be a more nonspecific finding across ACC, OFC, and anterior insular cortex. These anxiety states. Of clinical relevance, activity within dorsal and ventral compartments communicate OFC predicts subsequent response to treatment with their striatal counterparts; the dorsal com- with medication or behavior therapy. Further, partment is linked to the dorsal (cognitive/motor) MRI and MRS studies suggest striatal pathology, striatum, and the ventral compartment is linked which is consistent with cognitive activation to the ventral (limbic) striatum. Interestingly, the studies that indicate deficits in striatal recruitment dorsal and ventral compartments seem to be and thalamic gating as well as possible compen- reciprocally inhibitory [36,62–64]. Thus, in MD, sation via hippocampal activation. grossly, there appears to be hypoactivity within Of note, patients with OCD plus comorbid the dorsal compartment and hyperactivity within MD exhibit a profile that differs from that of the ventral compartment. patients with OCD alone [53]. This is particularly Importantly, a triad of areas seem to play germane to the focus of this review, because most a critical role in mediating the balance of activity patients with severe treatment-refractory OCD between the ventral and dorsal compartments who receive neurosurgical treatment also suffer both in health and disease. The amygdala is posi- from comorbid MD. Resting state neuroimag- tioned to assess the reward and threat value of ing data suggest that patients with OCD plus external stimuli and has the capacity to drive the MD have lower metabolism within the caudate, balance of activity toward the ventral compart- thalamus, and hippocampus than patients with ment. The pregenual ACC has the capacity to OCD alone [53]. Further, functional imaging data facilitate the restoration of dynamic equilibrium from before and after treatment with paroxetine between the compartments via its inhibitory in- suggest that patients with OCD plus MD exhibit fluence over both dorsal and ventral elements increased activity within the striatum after suc- [36,62–64]. Finally, the hippocampus, in addition cessful treatment, whereas patients with OCD to its role in cognition, has reciprocal connections alone exhibit decreased activity within the stria- with the amygdala and projects to the hypothal- tum after treatment [54]. amus to influence the HPA axis as well as other functions that are disturbed in depression, such as sleep and appetite. Therefore, it is proposed that Neuroimaging and neurocircuitry models amygdala hyperactivity and hippocampal ineffi- of major depression cacy may be central to the pathophysiology of MD. Of note, it has been proposed that exposure to Neurocircuitry model of major depression stress during early development or chronically Similar to OCD, neuroimaging-motivated neu- could represent a risk factor for the evolution of robiologic models of MD have also involved such a profile [65]. Thus, successful treatment of CSTC circuitry. In addition to the limbic CSTC MD (via any of a number of modalities) may S.L. Rauch / Neurosurg Clin N Am 14 (2003) 213–223 217 rely on some combination of deactivation of the context of successful antidepressant treatment ventral compartment, inhibition of the amygdala, [59,86,88,90]. Of note, however, activity within stimulation (or protection) of the hippocampus, OFC seems to be inversely correlated with MD and/or enhanced efficacy of pregenual ACC. symptom severity (as well as with anxiety in some cases), suggesting that this region may be re- cruited in a compensatory fashion [34,55,86]. Neuroimaging studies of major depression Elements of the dorsal compartment, including In comparison to OCD, the neuroimaging anterior and dorsolateral prefrontal cortex as well literature on MD is more abundant and more as dorsal ACC, have been found to exhibit difficult to interpret parsimoniously. The com- reduced activity during MD episodes [91]. Like- plexity of this area has been exacerbated by wise, reduction in activity within the dorsal com- clinical heterogeneity; MD studies have included partment has been observed during transient mixed cohorts as well as more homogeneous induction of sad mood [59]. Conversely, hypo- cohorts of patients with unipolar MD and bipolar activity in these regions returns toward normal MD as well as MD in the elderly and in the after successful antidepressant treatment of pa- context of primary neurologic disorders. In addi- tients with MD [59,92]. Moreover, cognitive tion, early disparities in how various neuro- activation studies have indicated an attenuated anatomic terms were applied contributed to capacity for patients with MD to recruit dorsal confusion over how apparently discrepant find- ACC successfully [93]. ings could be reconciled. Most recently, careful The amygdala has been found to be hyper- neuroanatomic studies along with functional active in neutral state studies of patients with MD neuroimaging studies, including those investigat- [86,94], and there is some suggestion that this ing mood states in normal subjects and studies of finding might be left-lateralized [86]. The severity changes associated with various antidepressant of MD symptoms is correlated with amygdala treatments, have begun to yield a more cohesive activity [86]. Although hyperactivity within this picture of MD pathophysiology and its resolu- region is attenuated with successful treatment, tion. some residual hyperactivity may persist even Structural neuroimaging studies of MD using during extended remission of MD [86]. Functional morphometric MRI segmentation methods have imaging studies using pictures of human faces shown reduced volumes of the hippocampus displaying various emotional expressions have [60,66–71] and the striatum [72,73]. Of note, find- provided a means for assessing amygdala respon- ings within the striatum have been less consistent sivity to social cues of varying arousal value and [74,75], and studies that segmented the hippo- valence [95,96]. Further, by presenting such face campus and amygdala conjointly have often been stimuli beneath the level of conscious awareness, it negative [76]. MRI studies have also found has been possible to probe automatic aspects of reduced volumes in prefrontal cortex [73,77], and such amygdala responses relatively dissociated cortical parcellation methods have been applied from the top-down influences of frontal cortex to discover reduced volume in subgenual cortex [97]. Using this approach in conjunction with [56,78] and OFC [79]. These structural imaging functional MRI, it has been found that patients findings resonate with postmortem studies of MD, with posttraumatic stress disorder exhibit ex- which have revealed glial cell loss in subgenual aggerated responsivity within the right amygdala cortex [80] as well as reduced glia and smaller to threat-related stimuli [98]. Interestingly, a neurons within OFC [81] and cell atrophy within subsequent analogous study of patients with MD dorsolateral prefrontal cortex [82]. revealed exaggerated amygdala responses within Neutral state functional imaging studies have the left amygdala; moreover, the exaggerated demonstrated hyperactivity within OFC and ante- response in the left amygdala was attenuated rior insular cortex during MD episodes [83–87]. toward normal after successful treatment with Although initial studies of MD indicated reduced antidepressant medication [99]. Importantly, re- activity within subgenual cortex [56,88], sub- cent functional imaging studies of healthy subjects sequent correction for reduced volume in that seem to suggest that the right amygdala may be area has revealed relative hyperactivity [56]. more temporally dynamic in its response, exhibit- Further, these same areas seem to exhibit elevated ing prompt habituation effects, whereas the left activity during induced states of transient sadness amygdala seems to exhibit a more temporally [59,89] and reduction of hyperactivity in the stable response that is highly valence dependent 218 S.L. Rauch / Neurosurg Clin N Am 14 (2003) 213–223

[100]. This provides a framework for understand- ciprocal connections between the ACC and ing lateralized differences in amygdala function several other structures, including OFC, the across mood and anxiety disorders. To elaborate, amygdala, the hippocampus, or posterior cingu- the prevailing negative bias that is characteristic late cortex [105]. In fact, comparisons of presur- of MD could be associated with left amygdala gical with postsurgical MRI data indicate that dysfunction, and the exaggerated response and volume may be reduced within the caudate failure to habituate to threat-related stimuli that is nucleus and posterior cingulate cortex by 6 to 12 characteristic of selected anxiety disorders could months after anterior cingulotomy [104,106]. be associated with right amygdala dysfunction Given the prevailing neurocircuitry model of [101]. This could also help to explain the high OCD, these are all potential sites of therapeutic comorbidity that is observed across mood and action. Of note, posterior cingulate cortex is well anxiety disorders, given that various etiologies of positioned to modulate activity within the OFC- amygdala dysfunction might be agnostic to the caudate CSTC circuit [107–110]. Interestingly, laterality of pathology. a recent functional neuroimaging study of OCD Although studies of MD have repeatedly noted demonstrated that presurgical activity within volumetric reductions in the hippocampus, resting posterior cingulate cortex correlated with sub- state functional neuroimaging studies have rarely sequent response after anterior cingulotomy [111]. found hypoactivity in this region [53,102]. Rather, Given the prevailing neurocircuitry model of during the course of a successful trial of anti- MD, it might be more appealing to consider that depressant medication, MD patients seemed to lesions of dorsal ACC might produce disinhibition exhibit increased activity within the hippocampus of pregenual ACC, which, in turn, might render at week 1 and subsequent decreases in hippocam- patients more responsive to antidepressant phar- pal activity by week 6 [58]. Likewise, although macotherapy after surgery. Alternatively, lesions resting state functional imaging studies of MD of the cingulum might interrupt ascending influ- have not typically indicated baseline abnormali- ences of the amygdala on the dorsal compartment. ties in pregenual ACC, pretreatment levels of activ- In the case of subcaudate tractotomy as well as ity within this area have been found to predict bilateral orbitomedial leukotomy, the lesions are subsequent response to antidepressant medication purportedly placed so as to interrupt fibers of [103]. Specifically, individuals with higher levels passage connecting OFC and subgenual ACC to of pregenual ACC activity before treatment the thalamus. For orbitomedial leukotomy, the exhibited a superior antidepressant treatment lesions might also disrupt amygdalofugal fibers response. to OFC and subgenual ACC [112]. In OCD, interruption of reciprocal projections between OFC and the thalamus would theoretically decrease reverberating (amplified) activity in the OFC- Anatomy and imaging of neurosurgical caudate CSTC, leading to reduced OCD symp- treatments for major depression and toms. Likewise, in MD, reducing activity within obsessive-compulsive disorder the ventral compartment, such as directly lesion- As outlined in detail throughout this issue, ing subgenual ACC or OFC, would be hypothe- ablative neurosurgical treatments for OCD and sized to reduce symptoms. A single case study of MD include anterior cingulotomy, subcaudate bilateral orbitomedial leukotomy has demon- tractotomy, limbic leukotomy, and anterior strated reduced activity within OFC, ACC, the capsulotomy. In fact, the neuroanatomic ramifi- caudate, and the thalamus in a woman with OCD cations of these ablative procedures remain in- when comparing postsurgical with presurgical completely understood. regional cerebral metabolism [113]. Such findings In the case of anterior cingulotomy, the lesions are confounded by symptomatic improvement of are placed within dorsal ACC and typically OCD, however, which characteristically yields this impinge on the cingulum bundle [104]. Thus, in profile of attenuated hyperactivity throughout the addition to reducing cortical mass and activity circuit. within dorsal ACC, it is likely that these lesions In the case of limbic leukotomy, lesions similar modify cingulo-striatal projections and disinhibit to those of anterior cingulotomy and subcaudate pregenual ACC. Given the composition of the tractotomy are combined. Hence, this multisite cingulum bundle, it is also possible that its operation would presumably combine the benefits disruption in cingulotomy could influence re- (as well as the potential adverse effects) of the two S.L. Rauch / Neurosurg Clin N Am 14 (2003) 213–223 219 aforementioned procedures. A single case study surgery has its effects. Furthermore, advancing of limbic leukotomy has demonstrated reduced science in this domain is essential to optimizing activity within the caudate nuclei in a patient with neurosurgical treatment for psychiatric disorders OCD and TS when comparing postsurgical with in the future. It is foreseeable that progress in this presurgical regional cerebral oxygen metabolism area could lead to the identification of superior [114]. Clinically, the patient exhibited improve- targets for ablative procedures, more effective ment in both OCD and TS symptoms after limbic selection of appropriate surgical candidates, leukotomy. and/or individualized interventions (eg, whereby In the case of anterior capsulotomy, lesions of targets are individually determined for each the ventral portion of the anterior limb of the patient). anterior capsule are likewise purported to in- With the advent of deep brain stimulation, terrupt OFC/subgenual ACC–thalamic connec- there are now opportunities to more flexibly (and tions. Moreover, the placement of these lesions reversibly) modulate activity within neural circuits may also compromise adjacent territories of the of interest [117]. This promises the potential to striatum. This can occur if the lesions interrupt advance our understanding of OCD and MD fronto-striatal projections, if the lesions themselves pathophysiology while systematically progressing impinge on the striatum, or if infiltration of edema toward improved neurosurgical treatments for surrounding the lesions encroaches on the stria- the most severe and treatment-refractory forms tum itself or on fronto-striatal projections. Again, of these diseases. Controlled treatment trials in for OCD, disruption of pathologic CSTC circuitry this area should include a parallel neuroimaging at the level of OFC-caudate or reciprocal OFC- component to maximize the clinical and scientific thalamic communications could underlie the benefits derived from these efforts. therapeutic effects of anterior capsulotomy. For MD, deactivation of subgenual ACC or disruption of interconnections among the elements of References the ventral compartment is a plausible mode of therapeutic action for anterior capsulotomy. [1] Dougherty DD, Rauch SL, editors. Psychiatric Interestingly, an MRI study of anterior cap- neuroimaging research: contemporary strategies. sulotomy for OCD and other anxiety disorders Washington. DC: American Psychiatric Publish- ing; 2001. indicated that appropriate placement of lesions [2] Dougherty DD, Rauch SL. Neuroimaging and within the right anterior capsule was critical to neurobiological models of depression. Harvard subsequent therapeutic response [115]. Further- Rev Psychiatry 1997;5:138–59. more, functional imaging data from a small [3] Rauch SL, Baxter LR. Neuroimaging of OCD cohort of patients with severe anxiety disorders and related disorders. In: Jenike MA, Baer L, undergoing anterior capsulotomy demonstrated Minichiello WE, editors. Obsessive-compulsive reductions in activity within orbitomedial frontal disorders: practical management. Boston: Mosby; cortex from presurgical to postsurgical scans 1998. p. 289–317. [116]. [4] Rauch SL, Cora-Locatelli G, Greenberg BD. Pathogenesis of OCD. In: Stein DJ, Hollander E, editors. Textbook of anxiety disorders. Wash- ington, DC: American Psychiatric Press; 2002. p. Summary and future directions 191–206. [5] Rauch SL, Cosgrove GR. Neurosurgical treat- Neuroimaging research has helped to provide ments for psychiatric diseases. In: Kaplan HI, a progressively sound basis for constructing Sadock BJ, editors. Comprehensive textbook of neurocircuitry models of OCD and MD. Still, psychiatry. 7th edition. Baltimore: Williams & our understanding of the pathophysiologic under- Wilkins; 1999. p. 2516–21. pinnings of these disorders remains only tentative. [6] Rauch SL, Whalen PJ, Dougherty DD, Jenike Similarly, current knowledge regarding the neuro- MA. Neurobiological models of obsessive compulsive disorders. In: Jenike MA, Baer L, biologic consequences of contemporary psychi- Minichiello WE, editors. Obsessive-compulsive atric neurosurgical procedures is quite limited. disorders: practical management. Boston: Mosby; Although extrapolation from nonhuman primate 1998. p. 222–53. data can be informative, obtaining additional data [7] Alexander GE, Crutcher MD, DeLong MR. Basal from human subjects is essential to understanding ganglia-thalamocortical circuits: parallel sub- the mechanisms by which psychiatric neuro- strates for motor, oculomotor, ‘‘prefrontal’’ and 220 S.L. Rauch / Neurosurg Clin N Am 14 (2003) 213–223

‘‘limbic’’ functions. Prog Brain Res 1990;85: [21] Ebert D, Speck O, Konig A, et al. H-magnetic 119–46. resonance spectroscopy in obsessive-compulsive [8] Alexander GE, DeLong MR, Strick PL. Parallel disorder: evidence for neuronal loss in the cingu- organization of functionally segregated circuits late gyrus and the right striatum. Psychiatry Res linking basal ganglia and cortex. Annu Rev Neuroimaging 1996;74:173–6. Neurosci 1986;9:357–81. [22] Ohara K, Isoda H, Suzuki Y, et al. Proton [9] Saxena S, Brody AL, Schwartz JM, Baxter LR. magnetic resonance spectroscopy of lenticular Neuroimaging and frontal-subcortical circuitry in nuclei in obsessive-compulsive disorder. Psychiatry obsessive-compulsive disorder. Br J Psychiatry Res 1999;92:83–91. 1998;35:26–37. [23] Swedo SE, Leonard HL, Garvey M, et al. Pediatric [10] Baxter LR, Schwartz JM, Guze BH, et al. Neuro- autoimmune neuropsychiatric disorders associated imaging in obsessive-compulsive disorder: seeking with streptococcal infections: clinical description the mediating neuroanatomy. In: Jenike MA, Baer of the ﬁrst 50 cases. Am J Psychiatry 1998; L, Minichiello WE, editors. Obsessive compulsive 155:264–71. disorder: theory and management. 2nd edition. [24] Breiter HC, Filipek PA, Kennedy DN, et al. Chicago: Year Book Medical Publishers; 1990. Retrocallosal white matter abnormalities in pa- p. 167–88. tients with obsessive-compulsive disorder. Arch [11] Haber SN, Fudge JL, McFarland NR. Striatoni- Gen Psychiatry 1994;51:663–4. grostriatal pathways in primates form an ascend- [25] Szeszko PR, Robinson D, Alvir JM, et al. Orbital ing spiral from the shell to the dorsolateral frontal and amygdala volume reductions in obses- striatum. J Neurosci 2000;20:2369–82. sive-compulsive disorder. Arch Gen Psychiatry [12] Jenike MA, Breiter HC, Baer L, et al. Cerebral 1999;56:913–9. structural abnormalities in obsessive-compulsive [26] Gilbert AR, Moore GJ, Keshavan MS, et al. disorder: a quantitative morphometric magnetic Decrease in thalamic volumes of pediatric pa- resonance imaging study. Arch Gen Psychiatry tients with obsessive-compulsive disorder who are 1996;53:625–32. taking paroxetine. Arch Gen Psychiatry 2000; [13] Robinson D, Wu H, Munne RA, et al. Reduced 57:449–56. caudate nucleus volume in obsessive-compulsive [27] Rosenberg DR, Keshavan MS, O’Hearn KM, disorder. Arch Gen Psychiatry 1995;52:393–8. et al. Frontostriatal measurement in treatment- [14] Scarone S, Colombo C, Livian S, et al. Increased naive children with obsessive-compulsive disorder. right caudate nucleus size in obsessive compulsive Arch Gen Psychiatry 1997;54:824–30. disorder: detection with magnetic resonance imag- [28] Baxter LR, Schwartz JM, Mazziotta JC, et al. ing. Psychiatry Res Neuroimaging 1992;45:115–21. Cerebral glucose metabolic rates in non-depressed [15] Peterson B, Riddle MA, Cohen DJ, et al. Reduced obsessive-compulsives. Am J Psychiatry 1988;145: basal ganglia volumes in Tourette’s syndrome 1560–1563. using three-dimensional reconstruction techniques [29] Nordahl TE, Benkelfat C, Semple WE, et al. from magnetic resonance images. Neurology 1993; Cerebral glucose metabolic rates in obsessive- 43:941–9. compulsive disorder. Neuropsychopharmacology [16] Singer HS, Reiss AL, Brown JE, et al. Volumetric 1989;2:23–8. MRI changes in basal ganglia of children with [30] Swedo SE, Schapiro MG, Grady CL, et al. Tourette’s syndrome. Neurology 1993;43:950–6. Cerebral glucose metabolism in childhood onset [17] O’Sullivan R, Rauch SL, Breiter HC, et al. obsessive-compulsive disorder. Arch Gen Psychia- Reduced basal ganglia volumes in trichotillomania try 1989;46:518–23. by morphometric MRI. Biol Psychiatry 1997;42: [31] Breiter HC, Rauch SL, Kwong KK, et al. 39–45. Functional magnetic resonance imaging of symp- [18] Rauch SL, Phillips KP, Segal E, et al. A pre- tom provocation in obsessive-compulsive disorder. liminary morphometric magnetic resonance imag- Arch Gen Psychiatry 1996;49:595–606. ing study of regional brain volumes in body [32] Cottraux J, Gerard D, Cinotti L, et al. A dysmorphic disorder. Psychiatry Res Neuroimag- controlled positron emission tomography study ing [in press]. of obsessive and neutral auditory stimulation in [19] Aylward EH, Harris GJ, Hoehn-Saric R, et al. obsessive-compulsive disorder with checking rit- Normal caudate nucleus in obsessive-compulsive uals. Psychiatry Res 1996;60:101–12. disorder assessed by quantitative neuroimaging. [33] McGuire PK, Bench CJ, Frith CD, et al. Arch Gen Psychiatry 1996;53:577–84. Functional anatomy of obsessive-compulsive phe- [20] Bartha R, Stein MB, Williamson PC, et al. A short nomena. Br J Psychiatry 1994;164:459–68. echo 1H spectroscopy and volumetric MRI study [34] Rauch SL, Jenike MA, Alpert NM, et al. Regional of the corpus striatum in patients with obsessive- cerebral blood ﬂow measured during symptom compulsive disorder and comparison subjects. Am provocation in obsessive-compulsive disorder us- J Psychiatry 1998;155:1584–91. ing oxygen 15-labeled carbon dioxide and positron S.L. Rauch / Neurosurg Clin N Am 14 (2003) 213–223 221

emission tomography. Arch Gen Psychiatry 1994; [48] Rauch SL, Whalen PJ, Savage CR, et al. Striatal 51:62–70. recruitment during an implicit sequence learning [35] Rauch SL, Savage CR, Alpert NM, et al. A task as measured by functional magnetic reso- positron emission tomographic study of simple nance imaging. Hum Brain Mapping 1997;5: phobic symptom provocation. Arch Gen Psychia- 124–32. try 1995;52:20–8. [49] Rauch SL, Savage CR, Alpert NM, et al. Probing [36] Rauch SL, van der Kolk BA, Fisler RE, et al. A striatal function in obsessive compulsive disorder: symptom provocation study of posttraumatic a PET study of implicit sequence learning. J stress disorder using positron emission tomogra- Neuropsychiatry 1997;9:568–73. phy and script-driven imagery. Arch Gen Psychi- [50] Rauch SL, Whalen PJ, Curran T, et al. Probing atry 1996;53:380–7. striato-thalamic function in OCD and TS using [37] Benkelfat C, Bradwejn J, Meyer E, et al. Func- neuroimaging methods. In: Cohen DJ, Jankovic J, tional neuroanatomy of CCK4-induced anxiety in Goetz CG, editors. Tourette syndrome. Advances normal healthy volunteers. Am J Psychiatry 1995; in neurology. Philadelphia: Lippincott, Williams & 152:1180–4. Wilkins; 2001. p. 207–24. [38] Baxter LR, Schwartz JM, Bergman KS, et al. [51] Schacter DL, Alpert NM, Savage CR, Rauch SL, Caudate glucose metabolic rate changes with Albert MS. Conscious recollection and the human both drug and behavior therapy for obsessive- hippocampal formation: evidence from positron compulsive disorder. Arch Gen Psychiatry 1992; emission tomography. Proc Natl Acad Sci USA 49:681–9. 1996;93:321–5. [39] Benkelfat C, Nordahl TE, Semple WE, et al. Local [52] Rauch SL, Whalen PJ, Curran T, et al. Thalamic cerebral glucose metabolic rates in obsessive- deactivation during early implicit sequence learn- compulsive disorder: patients treated with clomi- ing: a functional MRI study. Neuroreport 1998; pramine. Arch Gen Psychiatry 1990;47:840–8. 9:865–70. [40] Schwartz JM, Stoessel PW, Baxter LR, et al. [53] Saxena S, Brody AL, Ho ML, et al. Cerebral Systematic changes in cerebral glucose metabolic metabolism in major depression and obsessive- rate after successful behavior modification treat- compulsive disorder occurring separately and ment of obsessive-compulsive disorder. Arch Gen concurrently. Biol Psychiatry 2001;50:159–70. Psychiatry 1996;53:109–13. [54] Saxena S, Brody AL, Ho ML, et al. Differential [41] Swedo SE, Pietrini P, Leonard HL, et al. Cerebral cerebral metabolic changes with paroxetine treat- glucose metabolism in childhood-onset obsessive- ment of obsessive-compulsive disorder vs major compulsive disorder: revisualization during phar- depression. Arch Gen Psychiatry 2002;59:250–61. macotherapy. Arch Gen Psychiatry 1992;49:690–4. [55] Drevets WC. Neuroimaging studies of mood [42] Brody AL, Saxena S, Schwartz JM, et al. FDG- disorders. Biol Psychiatry 2000;48:813–29. PET predictors of response to behavioral therapy [56] Drevets WC, Price JL, Simpson JR, et al. Sub- versus pharmacotherapy in obsessive-compulsive genual prefrontal cortex abnormalities in mood disorder. Psychiatry Res Neuroimaging 1998;84: disorders. Nature 1997;386:824–27. 1–6. [57] Mayberg HS. Limbic-cortical dysregulation: a pro- [43] Rauch SL, Shin LM, Dougherty DD, et al. posed model of depression. J Neuropsychiatry Clin Predictors of fluvoxamine response in contamina- Neurosci 1997;9:471–81. tion-related obsessive compulsive disorder: a PET [58] Mayberg HS, Brannan SK, Tekell JL, et al. symptom provocation study. Neuropsychophar- Regional metabolic effects of fluoxetine in major macology 2002;27:782–91. depression: serial changes and relationship to [44] Saxena S, Brody AL, Maidment KM, et al. clinical response. Biol Psychiatry 2000;48:830–43. Localized orbitofrontal and subcortical metabolic [59] Mayberg HS, Liotti M, Brannan S, et al. Re- changes and predictors of response to paroxetine ciprocal limbic-cortical function and negative treatment in obsessive-compulsive disorder. Neu- mood: converging PET findings in depression ropsychopharmacology 1999;21:683–93. and normal sadness. Am J Psychiatry 1999; [45] Mansari ME, Bouchard C, Blier P. Alteration of 156:675–82. serotonin release in the guinea pig orbito-frontal [60] Sheline YI. 3D MRI studies of neuroanatomic cortex by selective serotonin reuptake inhibitors. changes in unipolar major depression: the role of Neuropsychopharmacology 1995;13:117–27. stress and medical comorbidity. Biol Psychiatry [46] Doyon J, Owen AM, Petrides M, et al. Functional 2000;48:791–800. anatomy of visuomotor skill learning in human [61] Swerdlow NR, Koob GF. Dopamine, schizophre- subjects examined with positron emission tomo- nia, mania and depression: toward a unified graphy. Eur J Neurosci 1996;8:637–48. hypothesis of cortico-striato-pallido-thalamic func- [47] Rauch SL, Savage CR, Brown HD, et al. A PET tion. Behav Brain Sci 1987;10:197–245. investigation of implicit and explicit sequence [62] Drevets WC, Raichle ME. Reciprocal suppression learning. Hum Brain Mapping 1995;3:271–86. of regional cerebral blood flow during emotional 222 S.L. Rauch / Neurosurg Clin N Am 14 (2003) 213–223

versus higher cognitive processes. Implications for [77] Coffey CE, Wilkinson WE, Weiner RD, et al. interactions between emotion and cognition. Cogn Quantitative cerebral anatomy in depression. A Emotion 1998;12:353–85. controlled magnetic resonance imaging study. [63] Shulman GL, Corbetta M, Buckner RL, et al. Arch Gen Psychiatry 1993;50:7–16. Common blood ﬂow changes across visual tasks:. [78] Botteron KN, Raichle ME, Drevets WC, et al. II. Decrease in cerebral cortex. J Cogn Neurosci Volumetric reduction in left subgenual prefrontal 1997;9:647–62. cortex in early onset depression. Biol Psychiatry [64] Whalen PJ, Bush G, McNally RJ, et al. The 2002;51:342–4. emotional counting Stroop paradigm: an fMRI [79] Bremner JD, Vythilingam M, Vermetten E, et al. probe of the anterior cingulate affective division. Reduced volume of orbitofrontal cortex in major Biol Psychiatry 1998;44:1219–28. depression. Biol Psychiatry 2002;51:273–9. [65] Kaufman J, Plotsky PM, Nemeroff CB, Charney [80] Ongur D, Drevets WC, Price JL. Glial reduction in DS. Effects of early adverse experiences on brain the subgenual prefrontal cortex in mood disorders. structure and function: clinical implications. Biol Proc Natl Acad Sci USA 1998;95:13290–5. Psychiatry 2000;48:778–90. [81] Rajkowska G, Miguel-Hidalgo JJ, Wei J, et al. [66] Bremner JD, Narayan M, Anderson ER, et al. Morphometric evidence for neuronal and glial Hippocampal volume reduction in major depres- prefrontal cell pathology in major depression. Biol sion. Am J Psychiatry 2000;157:115–8. Psychiatry 1999;45:1085–98. [67] Mervaala E, Fohr J, Kononen M, et al. Quanti- [82] Rajkowska G. Postmortem studies in mood tative MRI of the hippocampus and amygdala in disorders indicate altered numbers of neurons severe depression. Psychol Med 2000;30:117–25. and glial cells. Biol Psychiatry 2000;48:766–77. [68] Sheline YI, Sanghavi M, Mintun MA, Gado MH. [83] Baxter LR Jr, Phelps ME, Mazziotta JC, et al. Depression duration but not age predicts hippo- Local cerebral glucose metabolic rates in obses- campal volume loss in medically healthy women sive-compulsive disorder. A comparison with rates with recurrent major depression. J Neurosci in unipolar depression and in normal controls. 1999;19:5034–43. Arch Gen Psychiatry 1987;44:211–8. [69] Sheline YI, Wang PW, Gado MH, et al. Hippo- [84] Biver F, Goldman S, Delvenne V, et al. Frontal campal atrophy in recurrent major depression. and parietal metabolic disturbances in unipolar Proc Natl Acad Sci USA 1996;93:3908–13. depression. Biol Psychiatry 1994;36:381–8. [70] Steffens DC, Byrum CE, McQuoid DR, et al. [85] Cohen RM, Gross M, Nordahl TE, et al. Pre- Hippocampal volume in geriatric depression. Biol liminary data on the metabolic brain pattern of Psychiatry 2000;48:301–9. patients with winter seasonal affective disorder. [71] Vakilli K, Pillay SS, Lafer B, et al. Hippocampal Arch Gen Psychiatry 1992;49:545–52. volume in primary unipolar depression: a magnetic [86] Drevets WC, Videen TO, Price JL, et al. A resonance imaging study. Biol Psychiatry 2000;47: functional anatomical study of unipolar depres- 1087–90. sion. J Neurosci 1992;12:3628–41. [72] Husain MM, McDonald WM, Doraiswamy PM, [87] Ebert D, Feistel H, Barocka A. Effects of sleep et al. A magnetic resonance imaging study of deprivation on the limbic system and the frontal putamen nuclei in major depression. Psychiatry lobes in affective disorders: a study with Tc-99m- Res 1991;40:95–9. HMPAO SPECT. Psychiatry Res 1991;40:247–51. [73] Krishnan KR, McDonald WM, Escalona PR, [88] Buchsbaum MS, Wu J, Siegel BV, et al. Effect of et al. Magnetic resonance imaging of the caudate sertraline on regional metabolic rate in patients nuclei in depression. Preliminary observations. with affective disorder. Biol Psychiatry 1997; Arch Gen Psychiatry 1992;49:553–7. 41:15–22. [74] Lenze EJ, Sheline YI. Absence of striatal volume [89] George MS, Ketter TA, Parekh PI, et al. Brain differences between depressed subjects with no activity during transient sadness and happiness comorbid medical illness and matched comparison in healthy women. Am J Psychiatry 1995;152: subjects. Am J Psychiatry 1999;156:1989–91. 341–351. [75] Pillay SS, Renshaw PF, Bonello CM, et al. A [90] Brody AL, Saxena S, Silverman DH, et al. Brain quantitative magnetic resonance imaging study of metabolic changes in major depressive disorder caudate and lenticular nucleus gray matter volume from pre- to post-treatment with paroxetine. in primary unipolar major depression: relationship Psychiatry Res 1999;91:127–39. to treatment response and clinical severity. Psy- [91] Baxter LR Jr, Schwartz JM, Phelps ME, et al. chiatry Res 1998;84:61–74. Reduction of prefrontal cortex glucose metabolism [76] Dupont RM, Jernigan TL, Heindel W, et al. common to three types of depression. Arch Gen Magnetic resonance imaging and mood disorders. Psychiatry 1989;46:243–50. Localization of white matter and other subcortical [92] Bench CJ, Frackowiak RS, Dolan RJ. Changes in abnormalities. Arch Gen Psychiatry 1995;52: regional cerebral blood ﬂow on recovery from 747–55. depression. Psychol Med 1995;25:247–61. S.L. Rauch / Neurosurg Clin N Am 14 (2003) 213–223 223

[93] George MS, Ketter TA, Parekh PI, et al. Blunted bundle in the rhesus monkey. J Comp Neurol 1984; left cingulate activation in mood disorder subjects 225:31–43. during a response interference task (the Stroop). [106] Rauch SL, Makris N, Cosgrove GR, et al. A J Neuropsychiatry Clin Neurosci 1997;9:55–63. magnetic resonance imaging study of regional [94] Wu JC, Gillin JC, Buchsbaum MS, et al. Effect cortical volumes following stereotactic anterior of sleep deprivation on brain metabolism of cingulotomy. CNS Spectrums 2001;6:214–22. depressed patients. Am J Psychiatry 1992;149: [107] Cavada C, Company T, Tejedor J, et al. The 538–43. anatomical connections of the macaque monkey [95] Breiter HC, Etcoff NL, Whalen PJ, et al. Response orbitofrontal cortex. Cereb Cortex 2000;10: and habituation of the human amygdala during 220–42. visual processing of facial expression. Neuron [108] Kemp JM, Powell TPS. The cortico-striate projec- 1996;17:875–87. tions in the monkey. Brain 1970;93:525–46. [96] Morris JS, Frith CD, Perrett DI, et al. A [109] Vogt BA, Pandya DN. Cingulate cortex of the differential neural response in the human amyg- rhesus monkey; II. Cortical afferents. J Comp dala to fearful and happy facial expressions. Neurol 1987;262:271–89. Nature 1996;383:812–5. [110] Zald DH, Kim SW. Anatomy and function of the [97] Whalen PJ, Rauch SL, Etcoff NL, et al. Masked orbital frontal cortex, I: anatomy, neurocircuitry, presentations of emotional facial expressions and obsessive-compulsive disorder. J Neuropsy- modulate amygdala activity without explicit chiatry Clin Neurosci 1996;8:125–38. knowledge. J Neurosci 1998;18:411–8. [111] Rauch SL, Dougherty DD, Cosgrove GR, et al. [98] Rauch SL, Whalen PJ, Shin LM, et al. Exagger- Cerebral metabolic correlates as potential predicated amygdala responses to masked facial stimuli tors of response to anterior cingulotomy for obses- in posttraumatic stress disorder: a functional MRI sive compulsive disorder. Biol Psychiatry 2001;50: study. Biol Psychiatry 2000;47:769–76. 659–667. [99] Sheline YI, Barch DM, Donnelly JM, et al. [112] Amaral DG, Price JL. Amygdalo-cortical projec- Increased amygdala response to masked emotional tions in the monkey. J Comp Neurol 1984; faces in depressed subjects resolves with antide- 230:465–96. pressant treatment: an fMRI study. Biol Psychia- [113] Sachdev P, Trollor J, Walker A, et al. Bilateral try 2001;50:651–8. orbitomedial leucotomy for obsessive-compulsive [100] Wright CI, Fischer H, Whalen PJ, et al. Differen- disorder: a single-case study using positron emis- tial habituation in the prefrontal cortex and sion tomography. Aust NZ J Psychiatry 2001; amygdala to repeatedly presented emotional facial 35:684–90. stimuli. Neuroreport 2001;12:379–83. [114] Sawle GV, Lees AJ, Hymas NF, et al. The [101] Rauch SL, Shin LM, Wright CI. Neuroimaging metabolic effects of limbic leucotomy in Gilles studies of amygdala function in anxiety disorders. de la Tourette syndrome. J Neurol Neurosurg Ann NY Acad Sci [in press]. Psychiatry 1993;56:1016–9. [102] Cohen RM, Semple WE, Gross M, et al. Evidence [115] Lippitz BE, Mindus P, Meyerson BA, et al. Lesion for common alterations in cerebral glucose metab- topography and outcome after thermocap- olism in major affective disorders and schizophre- sulotomy or gamma knife capsulotomy for obses- nia. Neuropsychopharmacology 1989;2:241–54. sive-compulsive disorder: relevance of the right [103] Mayberg HS, Brannan SK, Mahurin RK, et al. hemisphere. Neurosurgery 1999;144:452–8. Cingulate function in depression: a potential pre- [116] Mindus P, Ericson K, Greitz T, et al. Regional dictor of treatment response. Neuroreport 1997; cerebral glucose metabolism in anxiety disorders 8:1057–61. studied with positron emission tomography [104] Rauch SL, Kim H, Makris N, et al. Volume before and after psychosurgical intervention. reduction in caudate nucleus following stereotactic A preliminary report. Acta Radiol Suppl 1986; lesions of anterior cingulate cortex in humans: 369:444–8. a morphometric magnetic resonance imaging [117] Nuttin B, Cosyns P, Demeulemeester H, et al. study. J Neurosurg 2000;93:1019–25. Electrical stimulation in anterior limbs of internal [105] Mufson EJ, Pandya DN. Some observations on capsules in patients with obsessive-compulsive the course and composition of the cingulum disorder. Lancet 1999;354:1526. Neurosurg Clin N Am 14 (2003) 225–235

Stereotactic cingulotomy G. Rees Cosgrove, MD, FRCS(C)a,*, Scott L. Rauch, MDb aDepartment of Neurosurgery, Massachusetts General Hospital and Harvard Medical School, 15 Parkman Street, ACC Suite 331, Boston, MA, USA bDepartment of Psychiatry, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA

Since its introduction in 1936, surgery for was initially carried out as an open procedure psychiatric illness has had a long and controver- [6,7]. Foltz and White [8] reported their experience sial history for various scientific, moral, and with stereotactic cingulotomy for intractable pain ethical reasons. Many different neurosurgical pro- and noted that the best results were in those cedures have been performed on a variety of cor- patients with concurrent anxiety-depressive states. tical and subcortical targets. Initially, Moniz [1] Ballantine and Giriunas [9] subsequently demon- performed prefrontal leukotomies by injection strated the safety and efficacy of cingulotomy in of absolute alcohol into the frontal lobes and a large number of patients with psychiatric illness, reported ‘‘worthwhile’’ improvement in 14 of 20 and it has been the surgical procedure of choice in patients. Moniz [2] later described his experience North America over the last 30 years. with larger numbers of patients after more specific Currently, the accepted therapeutic approach and restricted frontal lobe lesions were created to most psychiatric disease involves a combina- using a leukotome. Freeman and Watts [3] sub- tion of well-supervised pharmacologic, behavioral, sequently described their prefrontal lobotomy, and, in some instances, electroconvulsive thera- which was performed with a specially designed pies. Not all patients respond to these modern calibrated instrument that was inserted blindly to treatment methods, however, and many remain the midline and swept back and forth to interrupt severely disabled. Some of these patients might surgically the white matter tracts in the frontal be considered appropriate candidates for surgi- lobes. The authors noted that to achieve the best cal intervention if the therapeutic result and results, the lesion had to involve the medial overall level of functioning could be improved. frontal lobes and, by definition, the cingulate gyri. In this article, we explore the anatomic and These early operations were associated with physiologic basis for cingulotomy and the indica- significant mortality and morbidity. Tooth and tions for surgery. Guidelines for the appropriate Newton [4] reviewed 10,365 standard prefrontal selection of surgical candidates are presented lobotomy operations performed between 1943 along with details of the operative technique. and 1954 and confirmed that the rate of ‘‘im- Finally, the results and complications of cingu- provement’’ was approximately 70% but also lotomy are reviewed and compared with those of reported a 6% mortality, 1% epilepsy rate, and other common psychosurgical procedures. 1.5% marked disinhibition. These complications underscored the need for a less radical and more specific approach to the surgery. Anatomic and physiologic rationale Fulton [5] was the first to suggest that the for cingulotomy anterior cingulum would be an appropriate target for psychosurgical intervention, and cingulotomy In 1937, the same year that Moniz [1] reported his initial experience with prefrontal lobotomy, Papez [10] postulated that a reverberating circuit * Corresponding author. in the human brain might be responsible for E-mail address: [email protected] emotion, anxiety, and memory. The components (G.R. Cosgrove). of this rudimentary limbic system included the

1042-3680/03/$ - see front matter Ó 2003, Elsevier Science (USA). All rights reserved. doi:10.1016/S1042-3680(02)00115-8 226 G.R. Cosgrove, S.L. Rauch / Neurosurg Clin N Am 14 (2003) 225–235 hypothalamus, septal nuclei, hippocampi, mamil- lesions in the frontal, temporal, and cingulate lary bodies, anterior thalamic nuclei, cingulate areas in some cases of new-onset OCD [19]. gyri, and their interconnections. It was sub- Detailed morphometric analysis of MRI scans in sequently expanded by McLean [11] in 1952 OCD patients has also suggested focal abnormal- to incorporate paralimbic structures, including ities in striatal areas with subtle volumetric orbital frontal, insular, and anterior temporal abnormalities involving caudate nuclei [20]. cortices; the amygdala; and dorsomedial thalamic Positron emission tomography (PET) studies nuclei. have provided perhaps the most compelling Although the exact mechanisms are unknown, evidence for implicating orbitofrontal cortex, there is mounting evidence that the limbic system, cingulate cortex, and basal ganglia dysfunction including the cingulate gyrus and its interconnec- in OCD. PET 18F-fluorodeoxyglucose studies in tions, plays a central role in the pathophysiology children and adults have consistently reported of major depressive disorder (MDD), obsessive- significant elevation of absolute glucose metabolic compulsive disorder (OCD), and other anxiety rates for the cerebral hemispheres and orbital gyri disorders. Electric stimulation of the anterior and somewhat less consistent elevation for the cingulum and subcaudate region has been shown caudate nucleus in OCD patients as compared to to alter both autonomic responses and anxiety normal controls [21–23]. In one study of clomi- levels in human beings [12]. Stimulation of the pramine treatment in childhood-onset OCD, 6 hypothalamus in animals produces autonomic, patients who failed to respond had significantly endocrine, and complex motor effects, which higher right anterior cingulate and right orbital suggests that the hypothalamus integrates and co- metabolism than did 11 drug-responsive patients ordinates the behavioral expression of emotional [24]. Two reports have found regional decreases states [13]. The limbic system, which includes the in metabolic activity correlating with a decrease cingulate gyrus, has direct input to the hypo- in severity of OCD symptoms as measured by thalamus and seems to be strategically located the Yale-Brown Obsessive-Compulsive Scale to mediate and interconnect somatic and visceral (YBOCS) after successful pharmacologic or be- stimuli with higher cortical functions. Therefore, havioral treatment; one reported decreased cau- it is likely that certain psychiatric disorders date activity, and the other reported decreased (ie, MDD, OCD, other anxiety disorders) may right orbitofrontal metabolism [21,25]. In a small reflect a final common pathway of limbic system series of patients with chronic anxiety disorder dysregulation. and severe phobias, activation PET studies Data from clinical and neuroimaging studies performed as the patients were presented with have also converged to implicate the cortico- stimuli to recreate their fears demonstrated striatothalamic circuits in the pathophysiology of consistently increased regional cerebral blood flow OCD [14–16]. The frontal-striatal-pallidothala- (rCBF) in the anterior cingulate cortices, orbito- mic-frontal loop, which has been so well char- frontal cortex, left thalamus, and right caudate acterized for its control of motor function in nucleus [26]. Parkinson’s disease, may also explain some fea- Currently, one can only speculate how disrup- tures of OCD. From a clinical perspective, rare tion of different pathways in the limbic system or neurologic movement disorders like Von Economo’s cingulate gyri might normalize activity, leading to encephalitis and Sydenham’s chorea are known symptom improvement. Recently, Rauch et al [27] to affect the basal ganglia and have been asso- reported atrophy in the caudate body in subjects ciated with obsessive and compulsive symptoms who had undergone one or more cingulotomies [17]. Many patients with Tourette’s syndrome approximately 6 months before morphometric (TS), another disorder of the basal ganglia MRI studies. Such investigations are beginning to characterized by coprolalia and motor tics, demonstrate the functional connectivity of the have significant OCD symptoms throughout their cingulate gyri with subcortical nuclei. This may lives [18]. Orbitofrontal and cingulate cortex has explain how lesions in one anatomic area might also been implicated in OCD, because the cog- affect the integrity and function of other brain nitive and behavioral features associated with regions. Clinical observations suggest that OCD lesions in this area, such as decreased response and MDD patients do not improve immediately inhibition, inflexibility, and overattention to after psychosurgery but that several weeks to irrelevant details, are reminiscent of OCD symp- months are required for positive clinical effects to toms. MRI has also demonstrated specific brain be fully manifest. Thus, it is likely that secondary G.R. Cosgrove, S.L. Rauch / Neurosurg Clin N Am 14 (2003) 225–235 227 neural degeneration or metabolic alterations in enduring symptoms without significant remission, brain areas other than the region where the lesions although practically speaking, confirmation of are actually made may be involved in the thera- treatment refractoriness usually requires more peutic effect. than 5 years of illness before surgery. Severity is Neurochemical models suggest that the affec- usually measured using validated clinical research tive and anxiety disorders may be mediated via instruments corresponding to specific indicators, monoaminergic systems. In particular, the sero- such as a YBOCS score of greater than 20 for tonergic system has received emphasis with re- OCD or a Beck Depression Inventory (BDI) score spect to OCD. Because of the diffuse nature of the greater than 30. Disability may be reflected, for monoaminergic projections and their role as instance, by a Global Assessment of Function neuromodulators, however, these models are not (GAF) score of less than 50. particularly instructive in terms of the functional The major psychiatric diagnostic groups as neuroanatomy relevant to different neurosurgical defined by the Diagnostic and Statistical Manual of treatments as they are currently employed. Never- Mental Disorders, 3rd edition, revised (DSM-III- theless, lesions in the cingulate cortex or other R) that might benefit from surgical intervention targets in the limbic system may ultimately modify include OCD and major affective disorder (ie, these diffuse monoaminergic systems to exert a unipolar major depression or bipolar disorder) beneficial effect. [28]. In many instances, patients present with Although the exact neuroanatomic and neu- mixed disorders combining symptoms of anxiety, rochemical mechanisms underlying depression, depression, and OCD, and these patients remain OCD, and other anxiety states remain unclear, it candidates for surgery. Schizophrenia is not is clear that the basal ganglia, limbic system, and currently considered an indication for surgery. A cingulate cortex in particular play a principal role history of personality disorder, substance abuse, in the pathophysiology of these diseases. Simi- or other significant axis II symptomatology is larly, lesions that affect these target areas might often a relative contraindication to surgery. be expected to modulate neuropsychiatric dysfunc- To determine that their psychiatric illness is tion (the reader is referred to the article by Rauch refractory to treatment despite appropriate care, et al in this issue for a more comprehensive pre- all patients must be referred for surgical inter- sentation of these issues). vention by their treating psychiatrist. The referring psychiatrist must demonstrate an ongoing commitment to the patient and the evaluation process and must also agree to be responsible for Patient selection postoperative management. Detailed question- Only patients with severe, chronic, disabling, naires that document the extent and severity of and treatment-refractory psychiatric illness should the illness as well as a thorough account of the be considered for cingulotomy. Chronicity in this diagnostic and therapeutic history must be pro- context refers to the enduring nature of the illness vided by the psychiatrist. The specifics of phar- without extended periods of symptomatic relief macologic trials should include the agents used, and may be less important than the severity of the dose, duration, response, and reason for discon- illness. The severity of the patient’s illness must tinuation for any suboptimal trial. Adequate trials be manifest both in terms of subjective distress of electroconvulsive therapy or behavioral ther- and a decrement in psychosocial functioning. The apy when clinically appropriate must also be illness must prove to be refractory to systematic demonstrated. trials of pharmacologic, psychologic, and, when The patient and family must also agree to appropriate, electroconvulsive therapy before participate completely in the evaluation process as considering neurosurgical intervention. As in all well as in the postoperative psychiatric treatment medical decisions, the potential benefit from such program. In general, only adult patients (older an intervention must be balanced against the risks than 18 years) who are able to render informed imposed by surgery. consent and who express a genuine desire and Thoughtful assessment of psychosurgical can- commitment to proceed with surgery are accepted. didacy requires that criteria for severity, chronic- If the patient meets these criteria, at our ity, disability, and treatment refractoriness be institution, he or she would undergo a more de- operationalized to form guidelines. In this regard, tailed presurgical screening evaluation by an ex- chronicity would require at least 1 year of perienced multidisciplinary group of psychiatrists, 228 G.R. Cosgrove, S.L. Rauch / Neurosurg Clin N Am 14 (2003) 225–235 neurosurgeons, and neurologists (Cingulotomy 1.5 to 2.0 cm from the midline. After dural Assessment Committee). A thorough review of opening and cauterization of the pia, a standard the medical record is carried out to ensure that thermistor-equipped thermocoagulation electrode the illness is indeed refractory to an exhaustive (Radionics, Burlington, MA) with a 10-mm non- array of conventional therapies. The Massachu- insulated tip is introduced to the target and heated setts General Hospital (MGH) evaluation also to 85°C for 90 seconds. If necessary, the electrode includes an electroencephalogram (EEG), brain can be withdrawn 5 to 10 mm, and an additional MRI, neuropsychologic testing, and indepen- lesion is made superiorly using the same param- dently conducted clinical examinations by a psy- eters to ensure complete destruction of the chiatrist, neurologist, and neurosurgeon in the cingulum. The resulting lesion is approximately outpatient setting. An electrocardiogram and 15 to 20 mm in height and 10 mm in diameter and appropriate blood tests are obtained to assess should encompass the cingulum from the roof of medical risks and to exclude organic etiologies the ventricle to the cingulate sulcus. The pro- for mental status abnormalities. Validated clinical cedure is then carried out in an identical fashion research instruments (eg, YBOCS, BDI, GAF, on the opposite side. Intraoperative stimulation is Minnesota Mutiphasic Personality Inventory not performed routinely, but neurologic testing is [MMPI]) are employed to quantify psychiatric carried out during lesioning to ensure that no symptom severity and outcomes. There must be impairment of motor or sensory function, espe- unanimous agreement that the patient satisfies cially in the lower extremities, is incurred. On the selection criteria, that the surgery is indicated, day after surgery, a postoperative MRI scan is and that the requirements for informed consent obtained to document the placement and extent of are fulfilled. A family member or close relative the lesions (Fig. 1). must also understand the evaluation process and Although the patient may experience an the indications for, risks of, and alternatives to immediate reduction in anxiety, there is generally surgery and must agree to be available to pro- a delay to the onset of beneficial effect on de- vide emotional support for the patient during the pression and OCD. This latency may be as long hospitalization. as 6 to 12 weeks and must be clearly explained to the patient and referring psychiatrist. If there has been no response to the initial cingulotomy Surgical technique after 3 to 6 months, reoperation and enlarge- Cingulotomy was initially performed using ment of the cingulotomy lesion are considered ventriculography for surgical guidance, but over (Fig. 2). the past decade, this has been replaced by MRI– Using the surgical techniques described pre- guided stereotactic techniques. This allows for viously, additional lesions can be placed anterior more accurate placement of the lesions and direct to the initial cingulotomy. The MGH experience visualization of individual differences in cingulate suggests that 40% to 50% of patients require and ventricular anatomy. repeat surgery and that better long-term outcomes The procedure is carried out under mild are achieved with repeat surgery and multiple sedation and local anesthesia. A T1-weighted lesions in the cingulate gyrus. Therefore, over the stereotactic MRI scan is obtained (echo time past several years, three separate lesions have been [TE]=10 milliseconds, repetition time [TR]=600 placed at the initial operation, incorporating milliseconds), and using the midsagittal scan as approximately 2.0 to 2.5 cm of anterior cingulate a reference, oblique coronal sections (4-mm thick, cortex (Fig. 3). If this operation does not provide 1-mm interval) are selected in a plane parallel to satisfactory improvement in the patient’s symp- the proposed trajectory of the thermocoagulation toms, the cingulotomy can be converted to a probe and spanning the entire anterior cingulum limbic leukotomy at a later date by placing a and frontal horns of the lateral ventricles. Target lesion in the subcaudate region bilaterally. coordinates are calculated for a point 2 to 5 mm above the roof of the lateral ventricle, 7 mm from Results the midline, and 20 to 25 mm posterior to the tip of the frontal horns [29]. After calculation of the The results of bilateral cingulotomy in 198 pa- stereotactic target coordinates, a limited bicoronal tients suffering from a variety of psychiatric dis- scalp incision is made, and burr holes are placed orders were reported retrospectively by Ballantine bilaterally just anterior to the coronal suture and et al [30] in 1987. With a mean follow-up of G.R. Cosgrove, S.L. Rauch / Neurosurg Clin N Am 14 (2003) 225–235 229

Fig. 1. Midsagittal (A) and axial (B) T1-weighted MRI scans obtained 48 hours after a single anterior cingulotomy lesion. The acute hemorrhagic lesion undergoes a dramatic reduction in size as the edema resolves.

8.6 years, 62% of patients with severe affective inhibitors (SSRIs). It was thought that many of disorder were found to have had worthwhile these patients might have responded positively to improvement. Similarly, in patients with OCD, the SSRIs and thus were not truly refractory to approximately 56% were found to have under- treatment. gone worthwhile improvement. In 14 patients A prospective study was therefore undertaken suffering from nonobsessive anxiety disorders, using independent observers and clinically vali- 50% were found to be functionally well and dated outcome rating scales. Of 18 OCD patients 29% were found to have shown marked improve- in this study who underwent cingulotomy at the ment. Valid criticism of these results included the MGH after failing all modern pharmacotherapies, concern that many of these patients might not including the SSRIs, 5 met conservative criteria have met modern DSM-IV-R criteria for their as treatment responders (>35% improvement in psychiatric disorder and that the assessment of their YBOCS and a Clinical Global Improvement outcome was performed by biased observers, [CGI] outcome of very much improved or much namely, the referring psychiatrists, using sub- improved) and 2 others were considered to be jective rating scales. possible responders (>35% improvement in their A retrospective study was therefore devised to YBOCS or a CGI outcome of very much evaluate cingulotomy in the treatment of OCD improved or much improved), for an overall using independent observers and strict outcome response rate of 28% to 40% [32]. Average criteria. Of the 35 patients who were operated on duration of follow-up was longer than 2 years. for refractory OCD, 33 met modern DSM-IV-R Overall, the entire group improved significantly in criteria for OCD. In addition, at least 25% to terms of functional status, and no serious adverse 30% of patients demonstrated substantial benefit effects were found. This was the first study to from the procedure as defined by a 50% improve- demonstrate in a prospective fashion that cingu- ment in their YBOCS [31]. Concerns remained lotomy is an effective intervention in OCD about the validity of these conclusions, because patients as measured by standardized psychiatric many of the patients were treated before the rating scales and independent observers. Recently, introduction of modern psychopharmacology, another prospective study with long-term follow- specifically the selective serotonergic reuptake up by the same group of investigators reported 230 G.R. Cosgrove, S.L. Rauch / Neurosurg Clin N Am 14 (2003) 225–235

similar findings in 44 treatment-refractory OCD patients [33]. Between 32% and 45% of patients were judged to be responders, with a mean follow- up of almost 3 years. In more than 800 cingulotomies performed at the MGH since 1962, there have been no deaths and only two infections. Two acute subdural hematomas occurred early on in the series secondary to laceration of a cortical artery at the time of introduction of ventricular needles, but only 1 patient suffered permanent neurologic impairment. One patient suffered a delayed in- tracerebral hematoma after a fall 4 weeks after surgery but suffered no long-term neurologic deficits. An independent analysis of 34 patients who underwent cingulotomy demonstrated no significant behavioral or intellectual deficits as a result of the cingulate lesions themselves [34]. A subsequent evaluation of 57 patients before and after cingulotomy found no evidence of lasting Fig. 2. Axial T1-weighted MRI scans obtained 48 hours neurologic or behavioral deficits after surgery. A after repeat anterior cingulotomy. Note the original comparison of preoperative and postoperative lesion placed 12 months earlier and the acute lesion Weschler IQ scores demonstrated significant gains anteriorly. after surgery. This improvement was greatest in

Fig. 3. Midsagittal (A) and axial (B) T1-weighted MRI scans obtained 48 hours after three anterior cingulotomy lesions ablating approximately 2.5 cm of cingulate cortex. G.R. Cosgrove, S.L. Rauch / Neurosurg Clin N Am 14 (2003) 225–235 231 patients with chronic pain and depression but but the issue of case selection remains a major negligible in those with the diagnosis of schizo- consideration. Cingulotomy has been the pre- phrenia [35]. ferred surgical intervention in our experience Overall, it appears that cingulotomy is a useful because it provides substantial relief to a signifi- intervention in many patients with severe treat- cant number of patients and is associated with ment-refractory major depression or OCD. Using a low incidence of complications or adverse events subjective rating scales, approximately 60% to [29]. Many other centers believe that alternative 70% of patients show significant improvement. surgical interventions, such as subcaudate tractot- When more objective clinically validated rating omy, limbic leukotomy, or anterior capsulotomy, scales are used to assess outcome, approximately can provide better results in a similar patient 30% to 45% of patients who undergo cingulot- population. omy are considered to be responders. Although Subcaudate tractotomy was introduced by this response rate may appear low at first glance, Knight [37] in Great Britain in 1964 as one of one must remember that all patients were com- the first attempts to restrict the size of the surgical pletely refractory to every available treatment lesion and therefore minimize the side effects seen and that any measurable improvement might thus with standard prefrontal lobotomy. The aim was be considered salutary. to interrupt white matter tracts between orbital cortex and subcortical structures by placing a lesion in the region of the substantia innominata Discussion just below the head of the caudate nucleus. Although any patient with a severe psychiatric Surgical indications included major depressive illness was once considered a candidate for illness, OCD, and anxiety states as well as a variety surgical intervention, it is now clear that the of other psychiatric diagnoses. The original prudent indications for psychosurgery are more surgical procedure was performed using stereo- restrictive. There is general agreement among tactically implanted radioactive yttrium 90 seeds, centers that patients with major affective disorder, which yielded large lesional volumes of approxi- chronic anxiety states, and OCD are the best mately 2000 mm2. Currently, smaller lesions in are candidates for surgery. It can be safely concluded created by thermocoagulation with MRI stereo- that schizophrenia is not an indication for tactic guidance. psychosurgery, although patients with concomi- In patients with depression and OCD, total tant psychotic disorders and depression might still improvement or improvement with minimal be helped with surgery and should not be ex- symptoms was clinically observed in two thirds cluded. Personality disorders and psychoactive of the patients. The best review of the surgical substance use disorder are also significant relative results for subcaudate tractotomy was presented contraindications to surgery. by Goktepe et al [38] in 1975. Using a five-point Appropriate selection of patients for surgery global scale and rating scales for depression and remains the major responsibility of the psychia- anxiety, they reviewed 208 patients with a mean trist, guided by the informed and expert opinions follow-up of 2.5 years. Of the 134 patients of the other members of the psychosurgical team. available for structured interview, good results Advanced neuroimaging data may ultimately aid were seen in 68% of patients suffering from in optimal patient selection, and preliminary depression, 62.5% of patents with anxiety states, PET studies of patients with OCD indicate that and 50% of patients with obsessive neurosis. presurgical cerebral metabolic rates within a ter- Patients with schizophrenia, personality disorder, ritory of posterior cingulate cortex predict sub- drug abuse, or alcohol abuse did poorly. Some sequent outcome after cingulotomy [36]. If patients who had only temporary benefit from the preoperative evaluation could select those patients initial lesion had second lesions created lateral to with a better chance of responding favorably to the first, with good results seen in about half of cingulotomy, overall success rates could be greatly these individuals. improved. The incidence of complications was small but Much of the controversy surrounding the use included postoperative seizures in 2.2% of pa- of psychosurgery may be attributed to the in- tients and undesirable personality traits in 6.7%. discriminate application and high morbidity seen Transient disinhibition was common. Of the with the early surgical procedures. Stereotactic 25 patients who had died at the time of review, techniques have certainly minimized side effects, 3 had committed suicide. One patient died from 232 G.R. Cosgrove, S.L. Rauch / Neurosurg Clin N Am 14 (2003) 225–235 inadvertent destruction of the hypothalamus interrupt presumed frontothalamic connections in when a yttrium seed migrated off target. the anterior limb of the internal capsule, where Limbic leukotomy was introduced by Kelly they pass between the head of the caudate nucleus et al [39] in 1973 and combines subcaudate tractot- and the putamen. Clinical indications initially omy with anterior cingulotomy. This procedure included schizophrenia, depression, chronic anx- was designed to disconnect orbital-frontal-thala- iety states, and obsessional neurosis but are mic pathways with the former lesion and to currently almost entirely limited to OCD. interrupt an important portion of Papez’s circuit The target coordinates as described by Leksell with the latter. Kelly et al [39] reasoned that these are in the anterior one third of the anterior limb of two lesions might lead to a better result for the the internal capsule 5 mm behind the tip of the symptoms of OCD than either lesion alone. frontal horns and 20 mm lateral to the midline Indications for surgical intervention included at the level of the intercomissural plane. Lesions obsessional neurosis, anxiety states, depression, were created by thermocoagulation using a bipolar and a variety of other psychiatric diagnoses. This electrode system and were typically approximately stereotactic procedure placed three small (6-mm 15 mm in height and 4 to 5 mm in diameter. diameter) lesions in the lower medial quadrant of In the first 116 patients operated on by Leksell, each frontal lobe and two lesions in each cingulate 50% of patients with obsessional neurosis and gyrus. 48% of depressed patients had a satisfactory Using the same five-point scale described in the response [44]. Only 20% of patients with anxiety study of Goktepe [38], 66 patients were assessed neurosis and 14% of patients with schizophrenia before and after surgery (mean follow-up of 16 were improved. In this classification system, only months). In patients with obsessional neurosis, patients who were free of symptoms or markedly 89% were clinically improved; in chronic anxiety, improved were judged as having a satisfactory 66% were improved; in depression, 78% were im- response. Of the patients who were rated as worse proved; and in a small number of schizophrenics, after capsulotomy, 9 were schizophrenics, 4 were more than 80% were improved [40]. Kelly et al depressives, and 3 were obsessives. In another [41] later reported in 49 patients with OCD that series of 35 patients with OCD who underwent 84% were improved 20 months after surgery. capsulotomy and were followed prospectively by They too noted that postoperative symptom im- independent psychiatrists, 16 were rated as free of provement was not immediate, with a fluctuating symptoms and 9 were much improved, for an but progressive reduction of symptoms over the overall satisfactory result of 70% [45]. In a review first postoperative year. Although many patients of all cases of capsulotomy previously reported in complained of lethargy, confusion, and lack the literature, Mindus et al [46] found sufficient of sphincter control in the early postoperative data to categorize outcome in 213 of 362 patients. period, persistent complications were rare. No pa- Of these, 137 [64%] were deemed to have a tients developed seizures after surgery, 1 patient satisfactory result. suffered severe memory loss because of improper More recently, Mindus et al [47] followed 24 lesion placement, and 12% of patients complained patients prospectively with standardized rating of persistent lethargy. Measurements of IQ showed scales. Complications of the surgery included slight improvement after surgery. transient episodes of confusion during the first Recently, the results of a more modern series week in 19 of 22 patients available for follow-up, of MRI-guided stereotactic limbic leukotomy with occasional nocturnal incontinence. One were reported in 21 patients [42]. Mean follow-up patient was noted to have an intracranial hemor- was longer than 2 years, and 35% to 50% of pa- rhage without neurologic sequelae, and 1 patient tients were considered to be treatment responders suffered seizures. One patient committed suicide using clinically validated rating scales. Transient in the postoperative phase, and 8 patients suffered side effects were more common than after either from depression requiring treatment. Excessive cingulotomy or subcaudate tractotomy alone, with fatigue was a complaint in 7 patients, and 4 had permanent minor memory loss or urinary difficul- poor memory. Two patients showed slovenliness. ties observed in approximately 10% of patients. Weight gain is common after capsulotomy, with Although Talairach et al [43] were the first to an overall mean weight gain of about 10% in all describe anterior capsulotomy, Leksell popular- patients. No evidence of cognitive dysfunction has ized the procedure for patients with a variety been reported in 200 capsulotomy patients studied of psychiatric disorders [44]. The aim was to using a variety of psychometric tests. Reoperation G.R. Cosgrove, S.L. Rauch / Neurosurg Clin N Am 14 (2003) 225–235 233 was required in 2 patients who did not achieve scales (eg, YBOCS, BDI). Based on current meth- a satisfactory result, with only 1 improving after ods of comparison, the clinical superiority of the second operation. Burzaco [48] subjected 17 any one procedure is not convincing. Although of his 85 patients to a second procedure during many centers claim advantages for their specific which the lesions were enlarged, and half of these surgical intervention, we are unable to determine reoperations yielded satisfactory results. whether one of the four major psychosurgical In 1977, Kullberg [49] attempted to compare procedures is superior to the others at this point. cingulotomy and capsulotomy in the treatment of Cingulotomy is more commonly performed in the 26 patients in a randomized fashion. Six of 13 United States, whereas in Europe, capsulotomy capsulotomy patients and 3 of 13 cingulotomy and limbic leukotomy are more prevalent. They patients were better, but transient deterioration all seem to be roughly equivalent from a thera- in mental status was much more marked after peutic point of view, but in terms of unwanted capsulotomy than after cingulotomy. side effects, cingulotomy seems to be the safest of With currently available data, it is impossible all procedures currently performed. to determine whether there is one optimal surgical Regardless of the choice of procedure, surgical technique or strategy. The obstacles that prevent failures should be investigated, and if the lesion a direct comparison of results across centers size or location is suboptimal, consideration include diagnostic inaccuracies, nonstandardized should be given to another procedure. At least presurgical evaluation tools, center bias, and 45% of patients undergoing cingulotomy require varied outcome assessment scales. In virtually all repeat operation, with good results being salvaged published reports, however, some modification of in half [29]. Because of this high rate of reopera- the Pippard Postoperative Rating Scale, CGI tion in cingulotomy, we now recommend place- scale, or equivalent has been used to determine ment of three lesions in each cingulate gyrus at clinical outcome [50]. The Pippard scale rates the first operation to ablate approximately 2.5 cm outcome in five categories as follows: A=very of the gyrus. The exact size or volume of tissue much improved, B=much improved, C=slightly required for an effective outcome at each of the improved, D=unchanged, and E=worse. The target sites has yet to be determined. CGI expands the scale and rates outcome as Although controversy exists regarding the exact follows: 1=very much improved, 2=much im- choice of surgical procedure to be employed, there proved, 3=slightly improved, 4=unchanged, is unanimous agreement that the presurgical eva- 5=slightly worse, 6=much worse, and 7=very luation be performed by committed multidisci- much worse. plinary teams with expertise and experience in the Although comparisons are imperfect and the surgical treatment of psychiatric illness. Diagnosis rating of outcome is subjective, these scales do based on the DSM-III-R or DSM-N-R classifica- seem to have some clinical validity. If a Pippard tion scheme is encouraged, and although it is im- outcome of A and B or a CGI outcome of 1 or 2 possible to mandate uniformly across all centers, is considered satisfactory, cingulotomy was effec- prospective trials employing standardized clinical tive in 56% of patients with OCD, subcaudate instruments with long-term follow-up are needed. tractotomy in 50%, limbic leukotomy in 61%, Comparisons of preoperative and postoperative and capsulotomy in 67%. In patients with major functional status remain an important parameter affective disorder, cingulotomy was effective in in addition to targeting psychiatric symptoms in 65%, subcaudate tractotomy in 68%, limbic characterizing outcome. All centers with experience leukotomy in 78%, and capsulotomy in 55% [51]. emphasize the importance of rehabilitation after The difficulty with these results is that they surgery and the need for ongoing psychiatric were obtained many years ago before the avail- follow-up. The operation is not a panacea and ability of modern psychopharmacologic treat- should be considered as only one aspect in the ments. Patients who undergo surgery today have overall management of these patients. a much wider range of drug therapies and tend to be much more resistant to treatment. Comparing current outcomes with historical results can Summary therefore be misleading. Direct comparisons of the relative efficacy of the various modern surgical Cingulotomy can be helpful in certain patients options will only be possible when centers report with severe, disabling, and treatment-refractory their experience using validated objective rating major affective disorders, OCD, and chronic 234 G.R. Cosgrove, S.L. Rauch / Neurosurg Clin N Am 14 (2003) 225–235 anxiety states. This form of psychosurgical [15] Rauch SL, Jenike MA. Neurobiological models of treatment should only be carried out by an expert obsessive compulsive disorder. Psychosomatics 1993; multidisciplinary team with experience in these 34:20–32. disorders. Cingulotomy should be considered as [16] Weilburg JB, Mesulam MM, Weintraub S, et al. one part of an entire treatment plan and must be Focal striatal abnormalities in a patient with obsessive compulsive disorder. Arch Neurol 1989; followed by an appropriate psychiatric rehabili- 46:233–6. tation program. Many patients are greatly im- [17] LaPlane E, Levasseur M, Pillon B, et al. Obses- proved after cingulotomy, and the complications sions-compulsions and behavioral changes with or side effects are few. Cingulotomy remains an bilateral basal ganglia lesions: a neuropsychological, important therapeutic option for disabling psy- magnetic resonance imaging and positron tomog- chiatric disease and is probably underutilized. raphy study. Brain 1989;112:699–725. [18] Baxter LR, Schwartz JM, Guze BH, et al. Neuro- References imaging in obsessive compulsive disorder: seeking the mediating neuroanatomy. In: Jenike MA, Baer [1] Moniz E. Prefrontal leucotomy in the treatment of L, Minichiello WE, editors. Obsessive compulsive mental disorders. Am J Psychiatry 1937;93:1379–85. disorder: theory and management. 2nd edition. [2] Valenstein ES. Great and desperate cures. The rise Chicago: Year Book Medical Publishers; 1990. and fall of psycosurgery and other radical treat- p. 167–88. ments for mental illness. New York: Basic Books, [19] Berthier ML, Kulisevsky J, Gironell A, et al. Inc; 1986. Obsessive compulsive disorder associated with brain [3] Freeman W, Watts JW. Prefrontal lobotomy in the lesions: clinical phenomenology, cognitive func- treatment of mental disorders. South Med J tion, and anatomic correlates. Neurology 1996; 1937;30:23–31. 47:353–61. [4] Tooth JC, Newton MP. Leucotomy in England and [20] Jenike MA, Breiter HCR, Baer L, et al. Cerebral Wales 1942–1954. Reports on public health and structural abnormalities in patients with obsessive- medical subjects, no. 104. London: Her Majesty’s compulsive disorder: a quantitative morphometric Stationary Office; 1961. magnetic resonance imaging study. Arch Gen [5] Fulton JE. Frontal lobotomy and affective behav- Psychiatry 1996;53:625–32. ior: a neurophysiological analysis. New York: WW [21] Baxter LR Jr, Schwartz JM, Bergman KS, et al. Norton; 1951. Caudate glucose metabolic rate changes with both [6] Le Beau J. Anterior cingulectomy in man. J Neuro- drug and behavior therapy for obsessive compulsive surg 1954;11:268–76. disorder. Arch Gen Psychiatry 1992;49:681–9. [7] Whitty CWM, Duffield JE, Tow PM, et al. Anterior [22] Nordahl TE, Benkelfat C, Semple WE, et al. cingulectomy in the treatment of mental disease. Cerebral glucose metabolic rates in obsessive-com- Lancet 1952;1:475–81. pulsive disorder. Neuropsychopharmacology 1989; [8] Foltz EL, White LE Jr. Pain relief by frontal 2:23–8. cingulotomy. J Neurosurg 1962;19:89–94. [23] Saxena S, Brody AL, Maidment KM, et al. [9] Ballantine HT, Giriunas IE. Treatment of intracta- Localized orbitofrontal and subcortical metabolic ble psychiatric illness and chronic pain by stereo- changes and predictors of response to paroxetine tactic cingulotomy. In: Schmidek HH, Sweet WH, treatment. I: obsessive-compulsive disorder. Neuro- editors. Operative neurosurgical techniques. New psychopharmacology 1999;21:683–93. York: Grune & Stratton; 1982. p. 1069–75. [24] Swedo SE, Pietrii P, Leonard HL, et al. Cerebral [10] Papez JW. A proposed mechanism of emotion. glucose metabolism in childhood-onset obsessive- Arch Neurol Psychiatry 1937;38:725–43. compulsive disorder. Revisualization during phar- [11] McLean PD. Some psychiatric implications of macotherapy. Arch Gen Psychiatry 1992;49:690–4. physiologic studies on the frontotemporal portion [25] Swedo SE, Schapiro MB, Grady CL, et al. Cerebral of limbic system. Electroencephalogr Clin Neuro- glucose metabolism in childhood-onset obsessive- physiol 1952;4:407–18. compulsive disorder. Arch Gen Psychiatry 1989; [12] Laitinen LV. Emotional responses to subcortical 46:518–23. electrical stimulation in psychiatric patients. Clin [26] Rauch SL, Jenike MA, Alpert NM, et al. Regional Neurol Neurosurg 1979;81:148–57. cerebral blood flow measured during symptom [13] Ranson SW. The hypothalamus: its significance for provocation in obsessive-compulsive disorder using visceral innervation and emotional expression. Trans 15O-labeled CO2 and positron emission tomogra- Stud Coll Physicians Philadelphia 1934;2:222–42. phy. Arch Gen Psychiatry 1994;51:62–70. [14] Insel TR. Toward a neuroanatomy of obsessive- [27] Rauch SL, Kim H, Makris N, et al. Volumetric compulsive disorder. Arch Gen Psychiatry 1992; reduction in caudate nucleus following stereotactic 49:739–44. lesions of anterior cingulate cortex in humans: G.R. Cosgrove, S.L. Rauch / Neurosurg Clin N Am 14 (2003) 225–235 235

a morphometric magnetic resonance imaging study. [40] Mitchell-Heggs N, Kelley D, Richardson A. Stereo- J Neurosurg 2000;93:1019–25. tactic limbic leucotomy—a follow-up at 16 months. [28] American Psychiatric Association. Diagnostic and Br J Psychiatry 1976;128:226–40. statistical manual of mental disorders. 3rd edition, [41] Mitchell-Heggs N, Kelley D, Richardson A, revised. Washington, DC: American Psychiatric McLeigh J. Further exploration of Limbic Leucot- Association; 1987. omy. In: Hitchcook ER, Ballantine, HT, Meyerson [29] Spangler W, Cosgrove GR, Ballantine HT, et al. BA, editors. Modern concepts in Pschosurgery. MR-guided stereotactic cingulotomy for intractable Amsterdam: Elsevier/North Holland Biomedical psychiatric disease. Neurosurgery 1996;38:1071–8. Press; 1979. [30] Ballantine HT, Bouckoms AJ, Thomas EK, et al. [42] Montoya A, Weiss AP, Price BH, et al. Magnetic Treatment of psychiatric illness by stereotactic resonance image-guided stereotactic limbic leuco- cingulotomy. Biol Psychiatry 1987;22:807–19. tomy for intractable psychiatric disease. Neuro- [31] Jenike MA, Baer L, Ballantine HT, et al. Cingulot- surgery 2002;50:1043–52. omy for refractory obsessive compulsive disorder. [43] Talairach J, Hecaen H, David M. Lobotomie A long term follow-up of 33 patients. Arch Gen prefrontale limitee par electrocoagulation des fibres Psychiatry 1991;48:548–55. thalamo-frontalis leur emergence du bras anterior [32] Baer L, Rauch SL, Ballantine HT, et al. Cingulot- de la capsule interne. In: Proceedings of the Fourth omy for treatment of refractory obsessive-compul- Congress Neurologique Internationale. Paris: sive disorder: prospective long-term follow-up of 18 Masson; 1949. p. 141. patients. Arch Gen Psychiatry 1995;52:384–92. [44] Herner T. Treatment of mental disorders with [33] Dougherty DD, Baer L, Cosgrove GR, et al. frontal stereotactic thermal lesions. A follow-up Update on cingulotomy for intractable obsessive- study of 116 cases. Acta Psychiatr Scand Suppl compulsive disorder: prospective long-term follow- 1961;36:1–40. up of 44 patients. Am J Psychiatry 2002;159:269–75. [45] Mindus P. Capsulotomy in anxiety disorders. A [34] Corkin S, Twitchell TE, Sullivan EV. Safety and multidisciplinary study [thesis]. Stockholm: Karo- efficacy of cingulotomy for pain and psychiatric linska Institute; 1991. disorders. In: Hitchcock ER, Ballantine HT, Myer- [46] Mindus P, Rasmussen SA, Lindquist C. Neurosur- son BA, editors. Modern concepts in psychiatric gical treatment for refractory obsessive-compulsive surgery, Amsterdam: Elsevier; 1979. p. 253–72. disorder: implications for understanding frontal [35] Corkin S. A prospective study of cingulotomy. In: lobe function. J Neuropsychiatry 1994;6:467–77. Valenstein ES, editor. The psychosurgery debate. San [47] Mindus P, Edman G, Andreewitch S. A pro- Francisco: WH Freeman & Company; 1980. p. 264. spective, long-term study of personality traits in [36] Rauch SL, Dougherty DD, Cosgrove GR, et al. patients with intractable obsessional illness treated Cerebral metabolic correlates as potential predic- by capsulotomy. Acta Psychiatr Scand 1999;99: tors of response to anterior cingulotomy for ob- 40–50. sessive compulsive disorder. Biol Psychiatry 2001; [48] Burzaco J. Stereotactic surgery in the treatment of 50:659–67. obsessive compulsive neurosis. In: Perris C, Struwe [37] Knight GC. The orbital cortex as an objective in the G, Janssen B, editors. Biological psychiatry. surgical treatment of mental illness. The develop- Amsterdam: Elsevier; 1981. p. 1108–9. ment of the stereotactic approach. Br J Surg 1964; [49] Kullberg G. Differences in effect of capsulotomy 51:114–24. and cingulotomy. In: Sweet WH, Obrador S, [38] Goktepe EO, Young LB, Bridges PK. A further Martin-Rodriguez JG, editors. Neurosurgical treat- review of the results of stereotactic subcaudate ment in psychiatry, pain and epilepsy. Baltimore: tractotomy. Br J Psychiatry 1975;126:270–80. University Park Press; 1977. p. 301–8. [39] Kelly D, Richardson A, Mitchell-Heggs N. Tech- [50] Pippard J. Rostral leucotomy: a report on 240 cases nique and assessment of limbic leucotomy. In: personally followed up after one and one half to five Laitinen LV, Livingston KE, editors. Surgical years. J Ment Psychiatry 1955;101:756–73. approaches in psychiatry. Baltimore: University [51] Cosgrove GR, Rauch SL. Psychosurgery. Neuro- Park Press; 1973. p. 165–73. surg Clin North Am 1995;6:167–76. Neurosurg Clin N Am 14 (2003) 237–250

Single photon emission computed tomography imaging in obsessive-compulsive disorder and for stereotactic bilateral anterior cingulotomy Jin Woo Chang, MDa,d,*, Chan Hyung Kim, MDb, Jong Doo Lee, MDa,c, Sang Sup Chung, MD, PhDd aBK21 Project for Medical Science, Yonsei University College of Medicine, Seoul, Korea bDepartment of Psychiatry, Yonsei University College of Medicine, Seoul, Korea cDivision of Nuclear Medicine, Department of Diagnostic Radiology, Yonsei University College of Medicine, Seoul, Korea dDepartment of Neurosurgery, Yonsei University College of Medicine, CPO BOX 8044, Seoul, 120-752, Korea

Obsessive-compulsive disorder (OCD) is a com- may be a number of reasons for this, including mon neuropsychiatric disorder involving intrinsic inconsistency of SPECT data analysis, low spatial and repetitive thoughts as well as ritualistic and scanner resolution, different scanning techniques, irrational behavior that causes marked distress subject variation, and intraindividual variation [1,2]. Kodala et al [3] have reported the lifetime because of the presence or absence of OCD and 6-month prevalence rates to be 2.9% and symptoms at the time of injection. In particular, 1.6%, respectively. Recently, the neurobiologic semiquantitative measurements of the radioactivity basis of OCD has received much attention, with within regions of interest (ROIs) after normaliza- many modalities of evidence suggesting a neuro- tion to the cerebellar uptake are currently used as logic basis for OCD. Some of the most important an analytic method; however, the ROI method is information has resulted from functional neuro- subjective to operator bias in that large ROIs may imaging studies using functional MRI (fMRI), dilute small activation sites but small ROIs may single photon emission tomography (SPECT), and not differentiate activation sites from noise. positron emission tomography (PET) [4–10]. Recently, voxel-based analysis using statisti- Neuroimaging studies provide evidence for cal parametric mapping (SPM) has become avail- a role in OCD of the specific frontal-subcortical able, and in this system, an activated region is brain circuit connecting the orbitofrontal cortex, compared with a control area by statistical pro- the anterior cingulate gyrus, and elements of the cesses after assigning some threshold values and basal ganglia and thalamus [10,11]. These defined a probability value [12]. The resultant SPM data regions in the pathogenesis of OCD were pre- are more accurate than conventional semiquanti- dominantly determined based on PET studies. tative measurements [13]. Using this sophisticated Compared with the consistent metabolic alteration analysis method, many functional neuroimaging recorded on PET, the published data on blood flow studies have been performed with PET; however, studies with SPECT have been inconsistent and few SPECT data have been reported. variable in determining cortical and subcortical In this report, we evaluate the role of SPECT structures with abnormal perfusion patterns. There in OCD diagnosis by comparing the brain SPECT images of drug-free OCD patients with those of * Corresponding author. normal controls and drug-naive schizophrenia pa- E-mail address: [email protected] tients. We then explore the role of SPECT in sur- (J.W. Chang). gical outcome after bilateral anterior cingulotomy

1042-3680/03/$ - see front matter Ó 2003, Elsevier Inc. All rights reserved. doi:10.1016/S1042-3680(03)00006-8 238 J.W. Chang et al / Neurosurg Clin N Am 14 (2003) 237–250 by comparing pre- and postoperative brain Single photon emission computed SPECT imaging. tomography and surgical outcome Fifteen refractory OCD patients underwent Single photon emission computed tomography for bilateral stereotactic anterior cingulotomy using obsessive-compulsive disorder diagnosis MRI guidance. Thirteen patients in this group were followed up for more than a year (mean Seven patients with severe primary OCD (six follow-up of 22.4 months) (Table 2). Criteria for men and one woman) were studied. The mean age surgical candidacy included the following: (SD) was 25.4 4.7 years. All patients met the 1. The patient fulfills current diagnostic criteria Diagnostic and Statistical Manual for Mental for OCD. Disorder, 4th Edition (DSM-IV), (American Psy- 2. The duration of illness exceeds 3 years. chiatric Association, 1994) criteria for OCD as 3. The disorder is causing substantial suffering evaluated by two independent psychiatrists and as evidenced by ratings. a neurosurgeon who is experienced in psychosur- 4. The disorder is causing substantial reduction gery. Their mean onset age of symptoms was in the patient’s psychosocial functioning as 19.3 4.7 years, and the duration of symptoms evidenced by ratings. was 6.3 5.5 years (range: 2–27 years). No pa- 5. Current and up-to-date treatment options tients had any psychotropic medications in the 4 have been tried systemically for at least 3 weeks before SPECT study. The main compulsive years without appreciable effect on the symp- symptoms were washing, checking, and hoarding. toms or have had to be discontinued because Exclusion criteria included histories of central of intolerable side effects. neurologic disorders, tics, or substance abuse. The 6. If a comorbid psychiatric condition is present, severity of OCD behavior was quantified by the this disorder must also have been thoroughly Yale-Brown Obsessive-Compulsive Scale (Y- addressed with appropriate trials of first-line BOCS) [14], revealing an overall total score of treatments. 28.9 3.23. The depression and anxiety scales 7. The prognosis, without neurosurgical inter- were also evaluated using the Hamilton Rating vention, is considered poor. Scales for Depression (HAM-D) and Anxiety 8. The patient gives informed consent. (HAM-A) [15,16], revealing results of 13.6 4.1 9. The patient agrees to participate in the pre- and 12.8 2.6, respectively (Table 1). operative evaluation program and postoper- The control group consisted of seven normal ative rehabilitation program. volunteers (four men and three women) with a mean age of 30.6 6.8 years. These healthy sub- Also excluded from consideration for surgery jects had no psychiatric or neurologic illnesses. were patients younger than 18 years of age or Nine drug-naive cases of schizophrenia (three older than 60 years of age, a complicating axis I undifferentiated and six paranoid types) were also diagnosis (eg, organic brain syndrome; delusional included for comparison purposes. The mean age disorder; manifest abuse of alcohol, sedatives, or of the schizophrenia group was 28.4 8.2 years. illicit drugs); a complicating current axis II

Table 1 Clinical ﬁndings of obsessive-compulsive disorder patients for study I Sex/age Onset age Symptom (years) (years) duration (years) Y-BOCS HAM-D HAM-A Main symptoms Case 1 M/24 15 10 27 19 10 H, C Case 2 M/22 20 2 28 20 15 W, C Case 3 F/29 27 2 35 26 10 W, C Case 4 M/29 24 5 25 18 9 W, C Case 5 M/20 15 5 30 17 10 C, R Case 6 M/33 16 17 27 25 14 C, R Case 7 M/21 18 3 30 15 15 O, C Abbreviations: M, male; F, female; Y-BOCS, Yale-Brown Obsessive-Compulsive Scale; HAM-D, Hamilton Depres- sion Scale; HAM-A, Hamilton Anxiety Scale; W, washing; C, checking; O, obsession; H, hoarding; R, ordering. J.W. Chang et al / Neurosurg Clin N Am 14 (2003) 237–250 239

Table 2 Clinical findings of obsessive-compulsive disorder patients for study II Sex/Age Onset age Symptom (years) (years) duration (years) Y-BOCS CGI-S HAM-D HAM-A Main symptoms Case 1 M/25 15 10 29 7 19 10 H, C Case 2 M/34 16 17 36 7 25 13 C, W Case 3 F/53 44 8 40 7 50 32 C, W Case 4 M/41 36 4 34 7 47 25 C, W Case 5 M/45 39 5 37 7 30 24 C, R Case 6 M/26 15 10 38 7 21 8 C, W Case 7 M/40 18 21 38 7 21 8 S, C Case 8 F/36 20 15 39 7 41 17 C, W Case 9 F/31 24 7 35 7 20 10 W Case 10 M/18 14 4 28 7 13 7 C Case 11 F/38 27 11 38 7 15 9 W Case 12 M/48 33 15 30 6 22 20 C Case 13 M/30 20 10 32 7 29 22 S Abbreviations: M, male; F, female; Y-BOCS, Yale-Brown Obsessive-Compulsive Scale; HAM-D, Hamilton Depres- sion Scale; HAM-A, Hamilton Anxiety Scale; CGI-S, Clinical Global Improvement of Severity Scale; W, washing; C, checking; O, obsession; H, hoarding; R, ordering; S, sin. diagnosis from clusters A (eg, paranoid personality posterior to the frontal horn of the lateral disorder) or B (eg, antisocial or histrionic per- ventricle, 2 mm above the roof of the ventricle, sonality disorder); and a complicating current axis and 7 mm lateral to the midline. The electrode III diagnosis with brain pathologic findings (eg, was then withdrawn 8 mm to produce the fourth atrophy, tumors). lesion. The result was an elliptocylindrically The mean age of the patients (nine men and shaped lesion approximately 18 mm high, 13 mm four women) was 35.8 9.9 years. The average in anteroposterior (AP) dimension, and 6 mm symptom onset age was 24.7 10.2 years, and in lateral dimension (Figs. 1 and 2). Fig. 1 de- symptom duration was 10.5 5.3 years. No pa- monstrates the diffuse perilesional edema along tients had taken any psychotropic medications the radiofrequency lesions immediately after in the 4 weeks before SPECT study (before and cingulotomy. This edema resolved gradually, after surgery). however. Fig. 2 demonstrates the resolved perilesional edema along the radiofrequency lesions 3 Surgical procedure months after cingulotomy. Lesions were confined to the anterior cingulate gyrus. Patients were placed under local anesthesia, and bilateral burr holes were made in front of the Imaging procedures coronal suture 20 mm laterally from the midline. Stereotactic localization of the targets was ac- The SPECT procedure was performed for hieved using the MRI-compatible Leksell ste- patients and controls after an intravenous in- reotactic frame (Elekta Instruments, Atlanta, GA) jection of 740 MBq of Tc 99m ethyl cysteinate on a General Electric Signa 1.5-T unit (Milwau- (ECD). Brain images were obtained using a brain- kee, WI). A 1.8-mm electrode with a 10-mm bare dedicated annular crystal gamma camera (RAS- tip (Radionics, Burlington, MA) was inserted into PECT; Digital Scintigraphic Incorporation, the target to create radiofrequency thermocoagu- (Waltham, MA) equipped with low-energy, high- lation lesions at 85C for 90 seconds using a lesion resolution, parallel-hole collimators. The SPECT generator (RFG-3C; Radionics). Four lesions study was acquired in a 128 128 matrix with 3 along two tracks were created on either side of of angular increment for 20 minutes. Transaxial the anterior cingulate gyrus. The first lesion was images were obtained by the filtered-back pro- made 15 mm posterior to the frontal horn of the jection method using a Butterworth filter (cutoff lateral vehicle, 2 mm above the roof of the ven- frequency of 1.1 cycle per centimeter, order no. tricle, and 7 mm lateral to the midline. The elec- 10). Attenuation correction was performed by trode was then withdrawn 8 mm to produce the Chang’s method, and coronal and sagittal images second lesion. The third lesion was made 22 mm were calculated. 240 J.W. Chang et al / Neurosurg Clin N Am 14 (2003) 237–250

Fig. 1. T2-weighted MRI 3 days after cingulotomy demonstrated the diﬀuse perilesional edema along the radiofrequency lesions. (A) The axial image. (B) The coronal image.

don, UK) software [18]. After attenuation and Statistical parametric mapping scatter correction, the reconstructed SPECT data Analysis of data was performed on a SUN were reformatted into Analyze (Mayo Foundation, UltraSpare 10 workstation (Sun Microsystems, Baltimore, MD) header format consisting of 4096 Silicon Valley, CA) using automatic image regis- bytes of header, 1.67 mm of x and y pixel size, tration (AIR) [17] and SPM (SPM96; Institute of 3.34 mm of slice thickness, and 2 bytes of signed Neurology, University College of London, Lon- integer of pixel values. The SPECT images of the

Fig. 2. T1- and T2-weighted MRI 3 months after cingulotomy demonstrated the resolved perilesional edema along the radiofrequency lesions. Lesions were confined in the anterior cingulate gyrus. (A) T2-weighted axial image. (B) T1- weighted axial image. (C) The coronal image. J.W. Chang et al / Neurosurg Clin N Am 14 (2003) 237–250 241 normal and patient groups were separately Neuropsychologic evaluation coregistered to remove variations resulting from Patients were evaluated using a battery of neu- differences in the size and shape of individual rocognitive/neuropsychologic evaluation methods brains. The parameters for co-registration were: designed to measure the functional domains of intra-modality, 5000 pixel value for thresholds to language, visual integration, learning and mem- be used for organ of interest in both the standard ory, executive control, attention, motor control, and the reslice images, linear algorithm, 12 affine and general intellectual function. The neurocog- parameter model for controlling the number of nitive battery of tests to assess each domain degrees of freedom used in registration, and tri- consisted of the following standardized tests with linear interpolation. The data were then normali- well-established reliability and validity: intel- zed to a PET template (MNI template; Montreal lectual functioning by the Korean Wechsler Adult Neurological Institute, Montreal, Quebec, Cana- Intelligence Scale (K-WAIS); verbal memory, da) [2,18,19] and smoothed with 8-mm Full Width including immediate and delayed memory, by Half Maximum (FWHM) before SPM statistical Hopkins Verbal Learning Test (HVLT); visuospa- analysis. The statistical analysis was performed to tial construction and memory, including immedi- compare the mean SPECT images of the OCD ate and delayed memory, by the Rey-Osterrieth group with those of the schizophrenia and normal Complex Figure Test (RCFT); attention and control groups and the postoperative images with executive functions by the Wisconsin Card Sort- preoperative images in the surgical group. The ing Test (WCST) and Stroop test; and verbal flu- resulting t statistic, SPM{t}, was transformed to ency by Controlled Oral Word Association Test SPM{Z} with a threshold of 3.09 or 1.64. The (COWA). statistical results were displayed by rendering on In addition to this battery of standard neuro- the reference three-dimensional MRI images with cognitive tests, we evaluated all patients on a probability value of 0.001 or 0.05 and a corrected postoperative day 14 with the Korean version of extent threshold probability value of 0.05 for the Short Blessed Test (SBT-K) [21], which was SPM{Z}. Individual OCD images were also com- adapted from a six-item Orientation-Memory- pared with the normal group by shifting the SPM Concentration Test [22]. The optimal cutoff point parameters to a probability value of 0.005 and to differentiate patients with dementia from con- a Z value of 2.58. For graphic presentation of the trol subjects was 10/11 on the SBT-K. results, sections were displayed as transverse, sagittal, and coronal slices with hot color maps. Results of single photon emission computed tomography for obsessive-compulsive Clinical assessment and follow-up disorder diagnosis Clinical findings, complications, and patient The mean image of the OCD group at progress were prospectively recorded. All patients a threshold probability value of 0.001, with a Z were evaluated at or as near to 6-month intervals as value of 3.09 and a corrected extent threshold possible after surgery in our hospital. Evaluation probability value of 0.05, showed increased per- by the neurocognitive battery and by postoperative fusion within the anterior cingulate gyrus, the SPECT was performed between 6 and 12 months left basal ganglia, and the front lobe (Fig. 3). after surgery. All patients were followed up with Hyperperfusion within the orbitofrontal cortex or a personal interview conducted by the psychiatrists. the caudate nucleus was not evident; however, The severity of OCD behaviors was quantified by diffusely decreased perfusion within the bilateral the Y-BOCS [14]. The surgical outcome was also cerebral and cerebellar cortices, including the assessed/evaluated by the results of Clinical Global inferior frontal gyrus, was seen in the OCD group Improvement Scale (CGI) as well as the HAM-D (Fig. 4). When individual SPECT images of the and the HAM-A tests [15,16,20]. Statistical analy- OCD patients were compared with the mean ses were conducted to examine the effects of images of the normal control group, increased cingulotomy. The differences in these scales be- perfusion was seen within the anterior cingulate tween the preoperative state and postoperative gyrus in five patients and the frontal lobe in two. state (6 months, 12 months) were assessed using Increased perfusion within the basal ganglia, the a Wilcoxon signed rank test. A probability of less thalamus, and the orbitofrontal cortex was seen in than 0.01 was considered significant. only one patient (Fig. 5). 242 J.W. Chang et al / Neurosurg Clin N Am 14 (2003) 237–250

Fig. 3. Group comparison of obsessive-compulsive disorder (OCD) patients with healthy control subjects at the threshold of P < 0.001. (A) Statistical parametric mapping (SPM) results showing areas of hyperperfusion in OCD patients. (B) SPM results rendered on the reference three-dimensional MRI revealed signiﬁcantly increased perfusion of the anterior cingulate cortex on the reference three-dimensional MRI.

When the SPM parameters were shifted to the mean CGI-Severity score was also reduced sig- threshold probability value of 0.005, with a Z nificantly from 6.9 before surgery to 4.8 and 4.2 value of 2.58 and a corrected extent threshold at 6 and 12 months, respectively (P < 0.01). The probability value of 0.05, hyperperfusion was seen mean CGI score at 12 months was 2.1. All OCD within the anterior cingulate gyrus in all patients, symptoms seemed to respond to surgery, except within the thalamus in five, within the basal hoarding behavior. ganglia in four, within the caudate nucleus in Most patients experienced an immediate re- three, and within the orbitofrontal cortex in three duction in depression and anxiety after cingulot- (Table 3). omy. The mean HAM-D score was also reduced The OCD group showed increased perfusion significantly from 27.5 11.8 before surgery to within the anterior cingulate gyrus, the bilateral 16.2 10.6 and 10.0 6.0 at 6 and 12 months frontal, and the temporo-occipital areas compared after cingulotomy, respectively (P < 0.01). Also with the schizophrenia group (Fig. 6). reduced significantly was the mean HAM-A score from 16.1 7.8 before surgery to 9.2 7.1 and 6.3 5.4 at 6 and 12 months after cingulotomy, Outcome of anterior bilateral cingulotomy respectively (P < 0.01). The mean postoperative All patients demonstrated improvement in SBT-K score was 6.9 2.1. All patients scored their Y-BOCS score after cingulotomy (Table 4). below the optimal cutoff point for differentiation The mean Y-BOCS score was 34.9 4.0 before of dementia (10/11) on the SBT-K. surgery, whereas those at 6 and 12 months after The patients tolerated the procedure well, and cingulotomy were reduced significantly to 24.3 none experienced serious permanent complica- 7.4 and 21.5 6.6, respectively (P < 0.01). The tions related to the procedure. Three patients J.W. Chang et al / Neurosurg Clin N Am 14 (2003) 237–250 243

Fig. 4. Group comparison of healthy control subjects with obsessive-compulsive disorder (OCD) patients at the threshold of P<0.001. (A) Statistical parametric mapping (SPM) results showing areas of decreased perfusion in OCD patients (increased perfusion in healthy control subjects). (B) SPM results rendered on the reference three-dimensional MRI showing decreased perfusion within bilateral parieto-occipital, inferior frontal, and cerebellar areas. showed transient disorientation during the post- tive function. On the contrary, we observed an operative period. One case of wound dehiscence improvement of intellectual function after cingu- occurred from scratching of the wound as a result lotomy in several patients, although it was not of compulsive behavior; however, the wound statistically significant (mean preoperative IQ ¼ healed well after reclosure. Several patients exper- 99, postoperative [Ed-corrected] IQ ¼ 105). ienced minor interference with verbal fluency and minor personality changes; however, these patients Single photon emission computed tomography recovered from their minor problems within the and anterior bilateral cingulotomy first operative month. The neuropsychologic testing demonstrated We performed postoperative SPECT in eight that cingulotomy did not cause any significant patients after cingulotomy; the other five patients cognitive problems in areas like intelligence, refused for various reasons. After surgery, the language, visuospatial skills, and frontal execu- OCD patients demonstrated reduced perfusion 244 J.W. Chang et al / Neurosurg Clin N Am 14 (2003) 237–250

Fig. 5. A single case analysis of an obsessive-compulsive disorder patient (case 6) at the threshold of P < 0.005 revealed hyperperfusion within the anterior cingulate cortex, a bilateral lentiform nucleus, the left head of the caudate nucleus, and the thalamus. (A) Statistical parametric mapping (SPM) results rendered on the reference three-dimensional MRI. (B) SPM results rendered on the sagittal single photon emission computed tomography image. within the anterior cingulate gyrus and the right significant change in neuropsychologic testing orbitofrontal cortex in the summated and sub- after stereotactic bilateral anterior cingulotomy. tracted postoperative SPECT images compared In general, bilateral anterior cingulotomy is with the preoperative SPECT images (P < 0.01) regarded as an important therapeutic option for (Fig. 7). intractable OCD [23,24]. Studies on the effects of cingulotomy for OCD during the last decade have demonstrated that about one third of patients Discussion were improved after cingulotomy [25]. As reported previously [23,24], there was a general de- Outcome of bilateral anterior cingulotomy for lay in the onset of improvements of symptoms. obsessive-compulsive disorder In our series, all patients showed gradual im- There was an overall significant improvement provement after surgery, and this delayed process in the Y-BOCS and HAM-D scores, with no may suggest that effects are related not only to J.W. Chang et al / Neurosurg Clin N Am 14 (2003) 237–250 245

interruption but to reorganization of neural pathways after surgery. Prior reports have

0.005 claimed that only 39% of patients were considered Occipital Temporal Occipital Temporal < responders or possible responders on the basis of Occipital Temporal Temporal Frontal P demonstrating an improvement on the Y-BOCS of more than 35% or a CGI score of 1 or 2 [24]. Our entire group improved signiﬁcantly in terms of functional status (P < 0.01), and no 0.001

< significant adverse effects were encountered after P cingulotomy. The responder and possible responder ratios were 30.8% and 30.7%, respectively, on the basis of an improvement on the Y-BOCS 0.005 of more than 35% or a CGI score of 1 or 2. Our < Lt Lt Rt Frontal (Bilat) Temporal P Lt study featured more possible responders than those of previous reports [20,24]. One possible explanation for this is that our lesion is larger 0.001 than those used in other studies. We did expe- < P rience minor and transient mental deterioration (subsided within 1 week), described as a nonre- sponsive/flat affect, minor interference with 0.005 verbal fluency, lack of spontaneity, and minor < P Bilat Lt personality changes. This transient postoperative complication seemed to occur as a result of temporary perilesional brain edema around the le-

0.001 sions, as evident in Fig. 1. Our neurocognitive/ < neuropsychologic battery of tests in these patients P demonstrated that cingulotomy did not cause any signiﬁcant cognitive problems in areas like in-

0.005 telligence, language, visuospatial skills, and fron- < tal executive function. Bilat P Rt For patients undergoing stereotactic bilateral anterior cingulotomies, it becomes important to

0.001 have the proper target location and the optimal

< lesion volume in the anterior cingulate gyrus. We P are not recommending that our positive results justify general adoption of the procedure as a

0.005 routine treatment for OCD, however. As Cohen

< et al [26] reported, subtle personality and func- P Lt Lt Bilat Bilat Bilat Lttional Lt changes Lt couldremain. Lt We Frontal (Rt) believe Frontal that these minor neurobehavioral sequelae do not hinder the performance of cingulotomy as a treatment 0.001

< method for a certain portion of intractable OCD P patients so as to improve their quality of life.

0.005 Single photon emission computed tomography

< and obsessive-compulsive disorder + P + A number of neuroimaging studies using PET

: OFC, orbitofrontal cortex; Bilat, bilateral; Lt, left; Rt, right. have determined abnormal activity in the orbito- 0.001 frontal cortex, the anterior cingulate gyrus, and < Anterior cingulateP Caudate Basal ganglia OFC Thalamus Other areas the head of the caudate nucleus. These areas are interconnected anatomically and are among

Abbreviations the regions implicated in the pathophysiology Case 7 Table 3 Statistical parametric mapping data of obsessive-compulsive disorder versus healthy control group Case 6 Case 5 + + Case 4 + + Case 3 + + Case 2 + + Case 1 + + of OCD. 246 J.W. Chang et al / Neurosurg Clin N Am 14 (2003) 237–250

Fig. 6. Group comparison of obsessive-compulsive disorder (OCD) patients with drug-naive schizophrenia at the threshold of P < 0.001. (A) Statistical parametric mapping (SPM) results showing areas of hyperperfusion in OCD patients. (B) SPM results rendered on the reference three-dimensional MRI showing hyperperfusion within a bilateral anterior cingulate cortex and the frontal and parieto-occipital areas.

Our brain perfusion study of OCD patients ized tests. SPECT results demonstrated reduced using high-resolution SPECT showed consistently perfusion within the anterior cingulate gyrus and increased perfusion within the anterior cingulate the right orbitofrontal cortex after surgery. gyrus compared with that of normal controls and Functional imaging of OCD with 18F-ﬂuoro- other psychiatric patients (drug-naive schizophre- deoxyglucose (18F-FDG) PET at resting state nia). Hyperperfusion within the caudate nucleus demonstrated the hyperactivity of various ana- or the orbitofrontal cortex was less frequently tomic structures, such as the basal ganglia, the observed. Our results have shown outcomes using thalamus, and, in particular, the orbitofrontal SPECT for the ﬁrst time and have correlated it to cortex. In addition to resting PET, comparative clinical improvements as determined by standard- studies of the OCD patients before and after

Table 4 Surgical outcome after cingulotomy for study II Evaluation method Preoperative Postoperative 6-month Postoperative 12-month Y-BOCS (mean SD) 34.9 4.01 24.3 7.431* 21.5 6.612* CGI-S (mean) 6.9 4.83* 4.24* CGI 2.4 2.1 HAM-D (mean SD) 27.5 11.75 16.2 10.621* 10.0 5.982* HAM-A (mean SD) 16.1 7.88 9.2 7.123* 6.3 5.414* Abbreviations: Y-BOCS, Yale-Brown Obsessive-Compulsive Scale; HAM-D, Hamilton Depression Scale; HAM-A, Hamilton Anxiety Scale; CGI-S, Clinical Global Improvement of Severity Scale; CGI, Clinical Global Improvement. * P < 0.01 (Wilcoxon signed rank test). J.W. Chang et al / Neurosurg Clin N Am 14 (2003) 237–250 247

Fig. 7. Group comparison of preoperative images with postoperative images at the threshold of P < 0.05. (A) Statistical parametric mapping (SPM) results showing decreased perfusion at the anterior cingulate gyrus and the right orbitofrontal cortex. (B) SPM results rendered on the sagittal single photon emission computed tomography image. treatment have shown a decrease in cerebral The difference between PET and SPECT may glucose metabolism within the caudate nucleus, stem from several factors. First, the PET studies the orbitofrontal cortex [4,5,10,11], and the were conducted using 18F-FDG, which requires 40 anterior cingulate gyrus [27]. This decrease in to 60 minutes between injection and imaging. metabolism was correlated with a significant im- 99mTc EDC for brain SPECT, however, is rapidly provement in OCD symptoms. Despite these taken up by the neurons in the brain within a few findings, however, there are several inconsistencies minutes, proportional to the cerebral blood flow in the published data. For example, Swedo et al [32]. Therefore, the emotional state and presenting [28,29] showed increased metabolic rates within symptoms at the time of injection are important the orbitofrontal cortex and the anterior cingulate determinants in brain perfusion SPECT but are gyrus, whereas Baxter et al [4,5] found increased less critical in PET. In fact, regional cerebral metabolism in the whole brain, especially in the blood flow measured during symptom provoca- orbitofrontal cortex and the caudate nucleus but tion using oxygen-15–labeled carbon dioxide PET not in the anterior cingulate gyrus. Perani et al showed increased blood flow within the caudate [27] described hypermetabolism within the ante- nucleus, the anterior cingulate gyrus, and the rior cingulate gyrus, thalamus, and basal ganglia orbitofrontal cortex [33], although resting brain but not in the orbitofrontal cortex. Nordahl et al SPECT has demonstrated contradictory findings, [30] found no difference in activity within the such as decreased perfusion in the orbitofrontal caudate nucleus. Mindus and Jenike [31] found cortex [34] and the caudate nucleus [8,35]. There- that orbital and caudate glucose metabolic rates fore, a SPECT study performed during symp- decreased significantly 1 year after bilateral an- tom provocation would be better than resting terior capsulotomy compared with those before SPECT to evaluate the pathophysiologic mecha- surgery. nism of OCD. Regardless of the specific treatment or type of The spatial resolution of current cameras used imaging modality used, it is apparent that studies may be another important factor, because the of pretreatment and posttreatment OCD have orbitofrontal cortex and caudate nucleus are most consistently shown findings of decreased relatively smaller and thinner than the anterior orbitofrontal cortex and caudate activity after cingulate gyrus, especially on transaxial images. treatment. Therefore, small activation sites within these small 248 J.W. Chang et al / Neurosurg Clin N Am 14 (2003) 237–250 areas might be obscured and diluted by the glucose metabolism in the orbitofrontal cortex surrounding normal cells or background activity after OCD treatment [4,11,36]. Those reports also as a result of the poor spatial resolution of SPECT showed an increased glucose metabolism in the cameras. caudate nucleus and the orbitofrontal cortex The analytic method of measuring radioactiv- before the treatment. In our study, however, there ity in terms of ROI or calculation of hemispheric was no observation of any hyperperfusion in the ratio may contribute to false-negative results as orbitofrontal cortex before the treatment. The well as false-positive results as previously de- only equivalent observation was in the anterior scribed. In this study, data were analyzed using an cingulate gyrus. SPM method in which relative activity distribu- We suggest two potential explanations for this tion between the subject and control groups was divergence of results. First, as previously dis- compared and transformed to the unit of nor- cussed, SPECT has many weaknesses, including mal distribution in terms of z values and then poor spatial resolution. The caudate nucleus and displayed in coronal, axial, and sagittal views the orbitofrontal cortex are relatively smaller and showing only those voxels that reach a statistical thinner than the anterior cingulate gyrus, espe- significance of P < 0.001 or P < 0.005. The SPM cially on transaxial images, allowing the possibil- analysis is known to be more accurate and reliable ity for small activation sites within these small than semiquantitative measurement [13]. In this areas to be obscured and diluted by the surround- study, although five of the seven patients showed ing normal cells or background activity as a result hyperperfusion within the anterior cingulate cortex of poor spatial resolution. Second, the disparity on higher z and probability values (3.09 and between various studies, including PET and 0.001, respectively), all patients showed anterior SPECT images, in terms of secondary change of cingulate hyperperfusion on lower thresholds (z perfusion according to cingulotomy has not been value of 2.58 and P < 0.005). In addition, hyper- fully elucidated. The SPECT data may correctly perfusion within the caudate nucleus (n ¼ 3), the demonstrate the major changes of perfusion after orbitofrontal cortex (n ¼ 3), and the thalamus cingulotomy. Thus, the anterior cingulate gyrus (n ¼ 5) was frequently observed. Therefore, in may be the key center for the genesis of OCD, and SPECT or even PET studies without SPM, the improvement of OCD symptoms may be analysis may be inconsistent and conflicting in related to change of the orbitofrontal cortex. cases with slightly increased radioactivity within Of special note was the observed decrease of an activation site. perfusion in the right side of the orbitofrontal Although not all studies agree, a review of the cortex. Several studies comparing therapy-related OCD functional brain imaging literature reveals regional cerebral metabolic effects in patients with a remarkable amount of data suggesting the OCD have demonstrated orbitofrontal hyper- presence of abnormalities in the orbitofrontal metabolism on the right side before treatment cortex, the anterior cingulate gyrus, the caudate, [28,29,37]. Right-sided orbitofrontal metabolic and the thalamus, all of which are structures reduction was noted in the same reports. Lippitz linked by well-described neuroanatomic circuits et al [38] also suggested the crucial role of the right [10]. hemisphere in OCD. In our study, the OCD patients exhibited The observation that obsessive-compulsive increased perfusion in the anterior cingulate gyrus behaviors can be controlled by lesioning at several compared with both the healthy controls and the targets, including the orbitofrontal cortex, inter- schizophrenic patients (see Figs. 2 and 3). A more nal capsule, and anterior cingulate gyrus, supports interesting finding was the observation of post- the hypothesis that these structures are closely operative hypoperfusion in the anterior cingulate related and involved with the neuronal circuitry gyrus and the right orbitofrontal cortex in the of OCD. OCD patients. We assumed that the perfusion in Neuroanatomic studies have also shown that the anterior cingulate gyrus was caused by the the striatum contains a complex system—neu- cingulotomy itself; however, the reduction of rochemically specialized zones called striosomes perfusion in the right orbitofrontal cortex was dispersed within a larger compartment called the an interesting result. In the literature, there are matrix [7,39]. The striosomes of the caudate many reports concerning the role of the orbito- nucleus receive inputs from the orbitofrontal frontal cortex in the genesis of OCD and many cortex and the anterior cingulate gyrus, and the PET studies demonstrating a decrease in cerebral caudate processes the cortical information in J.W. Chang et al / Neurosurg Clin N Am 14 (2003) 237–250 249 preparation for an initiation of behavioral re- [3] Kolada JL, Bland RC, Newman SC. Epidemiology sponses [10,11]. The imbalance of this neuronal of psychiatric disorders in Edmonton. Obsessive- circuit is important for the generation of OCD compulsive disorder. Acta Psychiatr Scand Suppl symptoms. Our findings of reduced perfusion in the 1994;376:24–35. right orbitofrontal cortex after cingulotomy may [4] Baxter LR, Schwartz JM, Bergman KS, et al. Caudate glucose metabolic rate changes with both offer an explanation for the improvement of OCD drug and behavioral therapy for obsessive-compul- symptoms being related not only to the anterior sive disorder. Arch Gen Psychiatry 1992;49:681–9. cingulate gyrus but to the orbitofrontal cortex. [5] Baxter LR, Schwartz JM, Mazziotta JC, et al. Over the last several decades, many attempts Cerebral glucose metabolic rates in nondepressed have been made to investigate the outcome of patients with obsessive-compulsive disorder. Am J surgery prospectively. Although it is difficult to Psychiatry 1988;145:1560–3. compare the results because of lack of control [6] Breiter HC, Rauch SL, Kwong KK, et al. Functional groups, the OCD perfusion patterns were distin- magnetic resonance imaging of symptom provo- guished from those of schizophrenia in this study. cation in obsessive-compulsive disorder. Arch Gen Because a certain population (reported to be 26% Psychiatry 1996;53:595–606. [7] Insel TR. Toward a neuroanatomy of obsessive- in Porto’s analysis of schizophrenia) has comor- compulsive disorder. Arch Gen Psychiatry 1992; bid OCD, and given that schizophrenia also 49:739–44. involves the frontostriatal-thalamic circuit [3], we [8] Rubin RT, Ananth J, Vilaneueva-Meyer J, et al. performed a group comparison of SPECT images Regional 133Xenon cerebral blood flow and cerebral between OCD and drug-naive cases of schizo- 99mTc-HMPAO uptake in patients with obsessive- phrenia using SPM. The results demonstrated that compulsive disorder before and during treatment. the OCD patient group had a significant increase Biol Psychiatry 1995;38:429–37. of blood flow in the anterior cingulate gyrus as [9] Sachdev P, Trollor J, Walker A, et al. Bilateral well as in the temporo-occipital area. This finding orbitomedial leucotomy for obsessive-compulsive is consistent with the neuropathologic analysis of disorder: a single-case study using positron emission tomography. Aust NZ J Psychiatry 2001;35:684–90. schizophrenia, which demonstrated cytoarchitec- [10] Saxena S, Rauch SL. Functional neuroimaging and tural alteration of the anterior cingulate gyrus and the neuroanatomy of obsessive-compulsive disor- the superior temporal gyrus as well as decreased der. Psychiatr Clin North Am 2000;23:563–86. volume and cell number of the mediodorsal [11] Schwartz JM. Neuroanatomical aspects of cogni- thalamic nucleus [40]. tive-behavioural therapy response in obsessive- compulsive disorder. An evolving perspective on brain and behaviour. Br J Psychiatry 1998; Summary 173(Suppl 35):38–44. In conclusion, the present study suggests that [12] Friston KJ, Holmes A, Polin JB, et al. Detecting activations in PET and fMRI: levels of inference SPECT, using a sophisticated SPM analysis and power. Neuroimaging 1996;40:223–35. method, may be useful as a potential diagnostic [13] Acton PD, Friston KJ. Statistical parametric tool for OCD and a possible predictor of mapping in functional neuroimaging: beyond PET treatment outcome for OCD patients undergoing and fMRI activation studies. Eur J Nucl Med 1998; bilateral anterior cingulotomy. The anterior cin- 25:663–7. gulate gyrus seems to be an important structure in [14] Goodman WK, Price LH, Rasmussen SA, et al. The the pathogenesis of OCD symptoms. Further- Yale-Brown Obsessive Compulsive Scale (Y- more, our operative technique of anterior cingu- BOCS). Part I: development, use, and reliability. lotomy, featuring a larger lesion, seems to be Arch Gen Psychiatry 1989;46:1006–11. effective in ameliorating the symptoms of OCD [15] Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry 1960;23:164–73. without causing any serious complications. [16] Hamilton M. The assessment of anxiety states by rating. Br J Med Psychol 1959;32:50–5. References [17] Woods RP, Gafton ST, Watson JDG, et al. Automated image registration II. Intersubject val- [1] Attiullah N, Eisen JL, Rasmussen SA. Clinical idation of linear and nonlinear models. J Comput features of obsessive-compulsive disorder. Psychiatr Assist Tomogr 1998;22:153–65. Clin North Am 2000;23:469–91. [18] Friston KJ, Holmes AP, Worsley KJ, et al. Statis- [2] Friston KJ, Ashburner CD, Frith CD, et al. Spatial tical parametric maps in functional imaging: a gen- registration and normalization of images. Hum Brain eral linear approach. Hum Brain Mapp 1995;2: Mapp 1995;3:165–89. 189–210. 250 J.W. Chang et al / Neurosurg Clin N Am 14 (2003) 237–250

[19] Friston KJ, Frith CD, Liddle PF, et al. Plastic [31] Mindus P, Jenike MA. Neurosurgical treatment of transformation of PET images. J Comput Assist malignant obsessive-compulsive disorder. Psychiatr Tomogr 1991;15:634–9. Clin North Am 1992;15:921–38. [20] Jenike MA, Baer L, Ballantine T, et al. Cingulot- [32] Holman BL, Hellman RS, Goldsmith SJ, et al. omy for refractory obsessive-compulsive disorder. Biodistribution, dosimetry, and clinical evaluation A long-term follow-up of 33 patients. Arch Gen of technetium-99m-ethyl cysteinate dimer in normal Psychiatry 1991;48:548–55. subjects and in patients with chronic cerebral in- [21] Lee DY, Yoon JC, Lee KU, et al. Reliability farction. J Nucl Med 1989;30:1018–24. and validity of the Korean version of Short Blessed [33] Rauch SL, Jenike MA, Alpert NM, et al. Regional Test (SBT-K) as a dementia screening instrument. cerebral blood flow measured during symptom J Korean Neuropsychiatr Assoc 1999;38:1365–75. provocation in obsessive-compulsive disorder using [22] Katzman R, Brown T, Fuld P, et al. Validation of oxygen 15-labeled carbon dioxide and positron a short orientation-memory-concentration test of emission tomography. Arch Gen Psychiatry 1994; cognitive impairment. Am J Psychiatry 1983;140: 51:62–70. 734–9. [34] Crespo-Facorro B, Cabranes JA, Lopez-Ibor AMI, [23] Ballantine HT, Bouckoms AJ, Thomas EK, et al. Paya B, et al. Regional cerebral blood flow in Treatment of psychiatric illness by stereotactic obsessive-compulsive patients with and without cingulotomy. Biol Psychiatry 1987;22:807–19. a chronic tic disorder. A SPECT study. Eur Arch [24] Cosgrove GR, Ballantine HT. Cingulotomy in psy- Psychiatry Clin Neurosci 1999;249:156–61. chosurgery. In: Gildenberg PL, Tasker RR, editors. [35] Lucey JV, Costa DC, Busatto G, et al. Caudate Textbook of stereotactic and functional neurosur- regional cerebral blood flow in obsessive-compul- gery. New York: McGraw-Hill; 1998. p. 1965–70. sive disorder, panic disorder and healthy controls [25] Feldman RP, Alterman RL, Goodrich JT. Con- on single photon emission computerized tomogra- temporary psychosurgery and a look to the future. phy. Psychiatry Res Neuroimaging 1997;74:25–33. J Neurosurg 2001;95:944–56. [36] Saxena S, Brody AL, Schwartz JM, et al. Neuro- [26] Cohen RA, Kaplan RF, Moser DJ, et al. Impair- imaging and fronto-subcortical circuitry in obses- ments of attention after cingulotomy. Neurology sive-compulsive disorder. Br J Psychiatry 1998; 1999;53:819–24. 173(Suppl 35):26–37. [27] Perani D, Colombo C, Bressi S, et al. [18F] FDG [37] Simpson S, Baldwin B. Neuropsychiatry and PET study in obsessive-compulsive disorder. A SPECT in an acute obsessive compulsive syndrome clinical/metabolic correlation study after treatment. patient. Br J Psychiatry 1995;166:390–2. Br J Psychiatry 1995;166:244–50. [38] Lippitz BE, Mindus P, Meyerson BA, et al. Lesion [28] Swedo SE, Schapiro MB, Grady CL, et al. Cerebral topography and outcome after thermocapsulotomy glucose metabolism in childhood-onset obsessive- or Gamma Knife capsulotomy for obsessive-com- compulsive disorder. Arch Gen Psychiatry 1989; pulsive disorder: relevance of the right hemisphere. 46:518–23. Neurosurgery 1999;44:452–60. [29] Swedo SE, Pietrini P, Leonard HL, et al. Cerebral [39] Graybiel AM, Aosaki T, Flaherty AW, et al. The glucose metabolism in childhood-onset obsessive basal ganglia and adaptive motor control. Science compulsive disorder: revisualization during phar- 1994;265:1826–31. macotherapy. Arch Gen Psychiatry 1992;49:690–4. [40] Heckers S. Neuropathology of schizophrenia: cor- [30] Nordahl TE, Benkelfat C, Semple WE, et al. Cerebral tex, thalamus, basal ganglia, and neurotransmitter- glucose metabolic rates in obsessive compulsive specific projection system. Schizophr Bull 1997;23: disorder. Neuropsychopharmacology 1989;2:23–8. 403–21. Neurosurg Clin N Am 14 (2003) 251–265

Neuropsychiatric thalamocortical dysrhythmia: surgical implications D. Jeanmonod, MDa, J. Schulman, BSb, R. Ramirez, BSb, R. Cancro,MDc, M. Lanz, MDd, A. Morel, PhDa, M. Magnin, PhDa, M. Siegemund, MDa, E. Kronberg, PhDb, U. Ribary, PhDb, R. Llinas, MD, PhDb,* aNeurosurgical Clinic, Laboratory for Functional Neurosurgery, University Hospital, Sternwartstrasse 6, 8091 Zurich, Switzerland bDepartment of Physiology and Neuroscience, New York University School of Medicine, 550 First Avenue, MSB-442, New York, NY 10016, USA cDepartment of Psychiatry, New York University School of Medicine, New York, NY, USA dPsychiatry Center Hard, Zurich, Switzerland

Neuropsychiatric surgery has had a long and and anxiety disorders as well as obsessive-com- complex history with examples of less than optimal pulsive disorder (OCD) and impulse control dis- surgical procedures implemented in wrong settings. order (ICD). Considering the central role of Such past errors have raised important philosophic recurrent oscillatory thalamocortical properties in and ethical issues that remain with us for good the generation of normal hemispheric functions, we reasons. Nevertheless, the existence of enormous propose a surgical approach that provides re- suffering as a result of chronic therapy-resistant establishment of normal thalamocortical oscilla- disabling neuropsychiatric disorders compels a tions without reduction of cortical tissue and its search for alternative surgical approaches based specific thalamic connectivity. It consists of small, on a sound understanding of the underlying strategically placed, pallidal and medial thalamic physiopathologic mechanisms. We bring evidence lesions that serve to make subcritical the increased from single cell physiology and magnetoencepha- low-frequency thalamocortical recurrent network lography for the existence of a set of neuropsychi- activity. This result is attained via reduction of atric disorders characterized by localized and both thalamic overinhibition and low-frequency protracted low-frequency spontaneous recurrent oversynchronization. Thalamic disinhibition is activation of the thalamocortical system. This con- obtained by a lesion in the anterior medial dition, labeled thalamocortical dysrhythmia (TCD), paralimbic pallidum. The medial thalamic lesion underlies certain chronic psychotic, affective, is localized in the posterior part of the central lateral nucleus (CLN), where most cells have been shown to be locked in low-frequency production and to Reprinted from: Thalamus and Related Systems 2003; have lost their normal activation patterns. We 2(2). * Corresponding author. present here our experience with 11 patients, E-mail address: [email protected] including clinical follow-ups and pre- and post- (R. Llinas). surgical magnetoencephalographic (MEG) studies. This work was supported by the University of Zurich, The evidence speaks for a benign and efficient University Hospital of Zurich, Swiss National Science surgical approach and for the relevance of the Foundation (grants 31-36330.92, 31-47238.96, and 31- patient’s presurgical cognitive and social settings, 54179.98), NIH/NINDS 1 F31 NS42973-01, Charles A. making them more or less prone to postoperative Dana Foundation, and the New York University psychoreactive manifestations on rekindling of Medical Center General Clinical Research Center personal goals and social re-entry. (NYU.GCRC NIH NCRR M01RR00096).

After the questionable and inordinately wide hallucinatory/delusional disorder (4 patients), (2) application of the prefrontal lobotomy [1,2] in OCD (8 patients), (3) major (or endogenous) the first half of the twentieth century, many groups depression (7 patients) or bipolar mood disorder realized the necessity of confining surgical in- (2 patients), (4) anxiety disorder (7 patients), and terventions to the paralimbic (or mesocortical) (5) ICD (4 patients). We believe that the presence domain [3]. This gave rise to the three main of a high percentage of complex atypical neuro- stereotactic operations still in use today in view of psychiatric syndromes is a result of the fact that their efficiency and limited side effects: (1) the these patients exhibited particularly resistant anterior cingulotomy [4], (2) the subcaudate trac- disease forms related to widespread and strong totomy [5], and (3) the anterior capsulotomy [6]. TCD mechanisms and were thus at the forefront in These three procedures entail an interruption of the terms of surgical indication. Available drug treat- thalamocortical paralimbic frontal network. Such ments were used without success in all of them, surgical approaches have been efficient against and electroconvulsive therapy (ECT) was applied major depression and OCD but not against in four patients before they were referred to us. In psychosis. Early on, however, Spiegel and Wycis addition, three patients came to us after prior [7] explored the possibilities of stereotactic inter- unsuccessful interventions. One of them had ventions in the mediodorsal nucleus of the thal- a gamma knife capsulotomy, the second had a amus and described positive results in psychotic gamma knife thalamotomy, and the third had a patients. vagal nerve stimulator. The patients, seven men and There are both older and more recent data four women, ranged in age between 21 and 59 demonstrating the physiopathologic and histo- years (with a mean age of 35 years) at the time of logic involvement of paralimbic cortical areas and the first surgery. The duration of the disease before their corresponding specific thalamic partner, the the first surgery ranged between 6 and 21 years, with mediodorsal nucleus, in the generation of the a mean of 15 years. Postoperative relief percent- neuropsychiatric disorders. Indeed, electroenceph- ages were given by the patients themselves. alographic (EEG) [8–12], positron emission tomographic (PET) [13,14], histologic [15,16], and Surgical strategy radiologic [16] studies indicate spiking activities, increased low frequencies, and hypometabolism as We have applied the following criteria for sur- well as histopathologic and MRI changes in these gical indication: (1) a protracted evolution of the domains. disease (many years), (2) resistance to pharmaco- Our surgical approach is based on two logic and other conservative therapies, and (3) a premises: the central role of the oscillatory thal- major impact of the symptoms on the patient’s amocortical properties in the generation of normal quality of life. hemispheric functions [17] and the existence Two surgical targets have been developed, the as a physiopathologic basis to neuropsychiatric central lateral thalamotomy (CLT) and the ante- disorders of a TCD as evidenced by thalamic sin- rior medial pallidotomy (AMP). The CLT targets gle cell recordings [18,19] and magnetoencephalog- the posterior part of the CLN, and its stereo- raphy [20–23] and characterized by increased tactic coordinates are located anteroposteriorly low-frequency generation. In this context, we 2 mm posterior to the posterior commissure, medio- elected to normalize low-frequency production laterally 6 mm lateral to the border of the third without affecting the anatomic integrity of the ventricle, and dorsoventrally at the level of the functional cognitive thalamocortical network. intercommissural plane. The target is reached using The goal of this article is to describe this ap- an anteroposterior angle of 60 and a mediolateral proach at the surgical, clinical, and MEG levels. angle of 5 to 10. The CLT lesion measures 4 mm in diameter over 12 to 14 mm in length. The AMP, as its name implies, targets the Methods paralimbic anterior and medial pallidum, and its stereotactic coordinates are located anteroposte- Patient group riorly 4 mm posterior to the anterior commissure, All patients (n=11) suffered from chronic mediolaterally 12 mm lateral to the ventricular therapy-resistant neuropsychiatric disorders, in- border, and dorsoventrally 2 mm ventral to the cluding at least one, and often several, of the intercommissural plane. It is reached using an following clinical manifestations: (1) psychotic anteroposterior angle of 55 and a mediolateral D. Jeanmonod et al / Neurosurg Clin N Am 14 (2003) 251–265 253 angle of 20. The AMP lesion measures 4 mm in contrary, considered fully treated), two patients diameter over 6 mm in length. partly operated on (the first with a left CLT/AMP Physiologic confirmation of target localiza- and the other with a bilateral CLT), and two tions within given geometric confines is provided parkinsonian patients displaying endogenous by microelectrode recording (Fig. 1) as well as by neuropsychiatric manifestations (anxiodepressive macrostimulation. Figure 2 displays the CLT and episodes and ICD plus depression) and treated by AMP lesions as seen on our stereotactic atlas [24] a unilateral AMP. and MRI. The general surgical goal was a bilateral Magnetoencephalography coupling of CLT and AMP. We report on five patients with such full surgical treatment, two Pre- and postoperative MEG recordings were patients operated on only on one side on the basis performed in three patients. Analyses included of the discovery of a unilateral thalamic causal power spectra and coherence studies [20,21] as lesion (and, in the absence of evidence to the well as magnetic source imaging.

Fig. 1. Example of a unit activity recorded in the posterior part of the central lateral nucleus of the thalamus. Upper row: the discharge is composed of a series of recurring bursts of action potentials at a frequency of 3.3 Hz. Two of these bursts are shown using a faster time scale, revealing their intrinsic characteristics. Note that (1) the interspike interval (ISI) within a burst increases with each successive interval and that (2) the shorter the first ISI is in a burst, the larger is the number of spikes within this burst. These features are quantified in the two lower histograms. On the left, plotting of the duration of ISIs as a function of their position within the burst illustrates the progressive lengthening of successive intervals within bursts composed of an increasing number of action potentials. On the right, the duration of the first ISI in a burst shows a tight inverse relation to the number of following spikes within this burst (fitting with a logarithmic function). All these properties indicate that these bursts are the consequence of the deinactivation of calcium T-channels as a result of cell membrane hyperpolarization, the so-called low-threshold calcium spike bursts. 254 D. Jeanmonod et al / Neurosurg Clin N Am 14 (2003) 251–265

Fig. 2. Sagittal atlas projections (left column) and 2-day postoperative T1-weighted MRI images (right column) of central lateral thalamotomy (A) and anterior medial pallidotomy (B) lesions. The coordinates of atlas sections are 7.2 mm (A) and 12 mm (B) lateral to the ventricular border, and the crosses indicate the position of the posterior (A) and anterior (B) commissures. In both series, sections are oriented with posterior to the left and the intercommissural plane (represented by an interrupted line in the left panels) horizontal. The lesions represented in atlas sections do not include the edematous area seen on MRI. Stippled areas in the central lateral nucleus represent densocellular clusters. ac=anterior commissure; AV=anteroventral nucleus; Cd=caudate nucleus; CL=central lateral nucleus; CM=centre me´dian nucleus; LD=lateral dorsal nucleus; Li=limitans nucleus; GPi,e=internal and external pallidum; MDpc=mediodorsal nucleus, parvocellular division; Pf=parafascicular nucleus; PuM=medial pulvinar; PuT=putamen; R=reticular nucleus; STh=subthalamic nucleus; VA=ventral anterior nucleus; Vamc=magnocellular division of VA; VLp(d,v)=ventral lateral posterior nucleus, dorsal and ventral divisions; Vla=ventral lateral anterior nucleus; VM=ventral medial nucleus; VPMpc=ventral posterior medial nucleus, parvocellular division; ZI=zona incerta. Scale bars: 2 mm in atlas and 10 mm in MRI.

A whole-head, 148-channel, MEG system (4D was performed with in-house software and com- Neuroimaging, San Diego, CA) was used for all mercial Matlab 6 (Mathworks, Natick, MA) on patients. Spontaneous brain activity was contin- a Linux (Red Hat, Raleigh, NC) cluster computer uously recorded for 5 minutes while the patient system. Artifacts emanating from cardiac and reclined and was asked to stay alert with eyes other distant sources were removed by the soft- closed or eyes open. The bandpass was 0.1 to 100 ware in a channel-specific manner. Hz, and the sample rate was 508 Hz. The As for magnetic source imaging, the Fourier- electrocardiogram was simultaneously recorded transformed signals obtained with the multitaper digitally for off-line heartbeat artifact rejection. method were expanded with the singular value Spectral analysis of 5-second windows using the decomposition for frequencies of interest. A multitaper technique and cross-correlation be- complex space mode for each frequency was tween spectral amplitudes at different frequencies composed using the first five singular vectors. A D. Jeanmonod et al / Neurosurg Clin N Am 14 (2003) 251–265 255 volumetric source space was constructed using episodes of agitation, anxiety, confusion, and the averaged MRI MNI-152 [25] provided in depression resulting from personal and social SPM99. Voxels with a gray matter probability stressors, with activation of powerful guilt and higher than 0.4 were selected. This probabilistic feelings of self-insufficiency. Medication is being source space was transformed onto a head-cen- tapered slowly in view of its 20-year use. Figures 3 tered coordinate system and scaled. The lead field and 4 display the patient’s MEG findings before operator was computed using a spherically sym- and 3 months after surgery. metric volume conductor model. Inverse source Patient 3, born in 1956, suffered from a 14- imaging of the space modes at frequencies of year syndrome characterized by the coupling interest were computed using a recursively weight- of an OCD with major depression in addition to ed minimum-norm algorithm [26]. Noise reduc- anxiety and psychotic elements. Six years after tion was performed with Tikhonov regularization the bilateral CLT/AMP, the patient presents with using the generalized cross-validation criterion. complete relief of the OCD, anxiety, major de- The current density solutions were smoothed with pressive, and psychotic manifestations and has a three-dimensional gaussian kernel. For visuali- redeveloped a social and professional life to zation, smoothed solutions were linearly interpo- a degree inconceivable before surgery. She still lated onto the tessellated cortical surfaces of experiences moderate to strong episodes of a sealed MRI previously segmented using Free reactive depression based on marked insufficiency Surfer [27]. feelings, frustrating/disappointing confrontations, and unfulfilled personal wishes. Patient 4, born in 1980, was affected for 14 Results years by a complex syndrome with OCD and ICD as well as anxiety and psychotic disorders. Patient descriptions At the follow-up 6 months after the bilateral Patient 1, born in 1970, experienced a 10-year CLT/AMP and with a significant drug reduc- history of a complex neuropsychiatric syndrome, tion, there is relief of the psychotic elements, at including schizoaffective (delusions and bipolar least a 50% reduction of the OCD, and only disorder) and anxiety disorders as well as ICD and moderate improvement of the anxiety and ICD. OCD. He underwent surgery on the left side only Although the patient could never conduct social on the basis of the discovery of a small, vascular, interaction for more than a few minutes before probably developmental, inactive abnormality in surgery, after surgery, he can have hour-long his left mediodorsal thalamic nucleus. At a follow- discussions with members of the team. Anxiety up of 1 year after the left-sided CLT/AMP, he and impulse control problems are centered on presents with 100% relief of the schizoaffective, year-long difficulties in social and familial OCD, and ICD symptomatology but with only interactions. 50% relief of his anxiety and hyperactivity. His Patient 5, born in 1971, suffered from an OCD global improvement estimation is 60%. His drug associated with outspoken vegetative manifesta- intake could be markedly reduced. He experienced tions (ie, sweating, daily vomiting) and anxiety for only one reactive depressive episode in the post- 6 years before surgery. At a follow-up 2 years operative phase, caused by the concurrence of after the bilateral CLT/AMP, he enjoys complete multiple acute mental stressors, from which he relief of his OCD symptoms. In the context of fully recovered over a few weeks. Independence a preexistent substance use disorder, the patient is level and daily activities have increased signifi- currently in a withdrawal process from benzodi- cantly. His clinical improvements correlate with azepine dependence marked by anxiety. the MEG recordings shown in Figure 3 before Patient 6, born in 1973, suffered from a 21-year and 1 year after surgery. OCD coupled with major depression. Immedi- Patient 2, born in 1957, had a 20-year history ately after the second CLT/AMP, she described of a schizoaffective disorder characterized by complete relief of her obsessive manifestations delusional, hallucinatory, and affective bipolar and was shaken by the emptiness left by their manifestations with a few ICD elements. Over disappearance. One and a half year later, she a few weeks after the second CLT/AMP, he ex- complains about obsessive ideas that she has great perienced progressive and then complete symptom difficulty in describing. She shows signs of an relief, which is still present 2 years after surgery. anxiodepressive state and, only recently, motiva- In the interim period, he presented with two tion for psychotherapeutic treatment. The facts 256 D. Jeanmonod et al / Neurosurg Clin N Am 14 (2003) 251–265

Fig. 3. Magnetoencephalographic data from three patients in this series. Power spectra (four top panels) and coherence plots (three bottom panels) are displayed. The top left panel shows the power spectra from controls (blue) and thalamocortical dysrhythmia (TCD) patients, including neuropsychiatric as well as parkinsonian, neurogenic pain, and tinnitus patients. A peak shift into the theta domain and power increase in the theta and beta bands are demonstrated for TCD patients in comparison to controls. The top right panel presents the power spectrum of Patient 2, with a postoperative (blue) curve demonstrating the reappearance of the alpha peak and the power decrease in the theta and beta bands. The postoperative power spectra of Patient 7 (middle left) and Patient 1 (middle right) are mainly characterized by a reduction in theta and beta power, allowing reappearance (Patient 7) or better visualization (Patient 1) of the alpha peak. Coherence was analyzed applying a cross-correlation analysis of the variation along time of the spectral power for frequencies between 0 and 40 Hz. The bottom left panel shows such coherence for controls, the middle panel shows coherence for Patient 2 before surgery, and the right panel shows coherence for the same patient after surgery. (See also Color Plate 1.) D. Jeanmonod et al / Neurosurg Clin N Am 14 (2003) 251–265 257 that she has no memories, even as a child, of significant improvements in daily activities and obsession-free moments and that several first- independence. A second CLT/AMP is being degree relatives also suffer from neuropsychiatric considered. disorders provide a formidable barrier to adjust- Patient 9, born in 1971, suffered from a 13-year ment of her perspective toward a disease-free OCD associated with major depressive episodes. existence. Given this context, a drug reduction has During the 2 years after the bilateral CLT, there not yet been possible. have been fluctuating degrees of response of the Patient 7, born in 1970, suffered from a 20-year OCD symptoms, ranging between 50% and 100% neuropsychiatric syndrome characterized by ob- improvement. There were no postoperative major sessive, anxiety, and depressive disorders associ- depressive episodes. The acute postoperative ated with anger, pain, tinnitus, and dyskinetic evolution was characterized by total control of elements. MRI examinations showed the presence the OCD symptomatology, resulting in a state of of an inactive, probably old, vascular anomaly massive anxiety as a result of the emptiness after in the left mediodorsal nucleus, which was sub- suppression of the constant obsessive ideas. The mitted to gamma knife surgery with no change patient’s spontaneous interpretation concerning in symptomatology. Multiple ECT sessions were the fluctuating reappearance of obsessions was applied without results, and a vagal nerve that her anxiety was so severe that she herself stimulator brought no relief but only an increase reactivated some of them to relieve her inner in anxiety. A unilateral left-sided CLT/AMP tension. This psychodynamic context was compli- (considered in principle to be sufficient, as for cated by the presence of an excess in self-demands Patient 1) was performed. One year and a half so as to comply with a strict and perfectionist later, the patient describes an at least 50% picture of herself. At the moment, we await com- improvement of his OCD, 90% relief of anxiety, plementary evidence to ascertain better the neces- 70% relief of depression and pain, 20% relief of sity of a bilateral AMP. tinnitus, 80% relief of dyskinesias, but only Patient 10, born in 1937, suffered in the context a slight (30%) improvement of anger. The global of a parkinsonian disease from a 16-year neuro- improvement rate is 70%. The patient’s evolution psychiatric depressive disorder and ICD. For 5 is colored by (1) the fear that something still may years after a right-sided AMP coupled to a sub- progress in his brain and jeopardize his improve- thalamotomy to alleviate his motor syndrome, the ments and future, and (2) the mismatch between patient presented with complete relief of both what is/has been and what should be, associated motor and neuropsychiatric symptoms. Recently, with marked feelings of self-insufficiency. There is disease progression at both motor and neuropsy- a clear-cut increase of independence and 60% chiatric levels poses the question of surgical improvement of daily activities (the patient restabilization of the other hemisphere. traveled two times on his own from the Midwest Patient 11, born in 1944, suffered in the context to New York City and found the office of one of of a parkinsonian disease from a 13-year major the authors, an activity that would have been depressive disorder characterized by unmotivated impossible before surgery without his parents). He abrupt anxiodepressive episodes. He presented is presently on one antidepressant drug at a low 80% to100% relief of this disorder during a year dose only and is undergoing psychotherapy. His and a half after a left-sided CLT/AMP coupled to MEG recordings before and a year and a half a pallidothalamic tractotomy [19]. As for the pre- after surgery correlate well with these clinical vious patient, disease progression at both motor results and are shown in Figure 3. and neuropsychiatric levels raises the possibility Patient 8, born in 1970, suffered from an 18- of a right-sided intervention. year OCD associated with an anxiety disorder and Summarizing (Table 1), all patients demon- depression. Repeated ECT and a gamma knife strated significant to complete relief from their bilateral capsulotomy did not improve the symp- neuropsychiatric ailments. We observed the fol- tomatology and may have triggered anxiodepres- lowing postoperative reactive manifestations: (1) sive manifestations. At a follow-up of 9 months anxiety (1 patient), (2) anxiodepressive state (3 after the first left-sided CLT/AMP, a differentiated patients), (3) anxiety and nonacceptance/frus- result is observed, with strong improvement of tration (3 patients), and (4) frustration and depres- one OCD element (obsessive social component) sion (1 patient). A conceptual rigidity with lack of but lack of improvement of another (obsessive adaptation to the new situation was seen in two cleanliness). There have been variable but overall patients. 258 .Jamnde al et Jeanmonod D. Table 1 Overview of patient data Disease Relief (%) Age duration Causal Patient (years) (years) ECT GK VNS lesion Symptoms Operation PSY DEP/BIP OCD ICD ANX

1 31 10 L Thal/MD PSY/BIP/ANX/OCD/ICD L CLT/AMP 100 100 100 100 50 / 2 43 20 + PSY/BIP/ICD Bilateral CLT/AMP 100 100 100 251–265 (2003) 14 Am N Clin Neurosurg 3 36 14 + OCD/DEP/ANX/PSY Bilateral CLT/AMP 100 100 100 100 4 21 14 OCD/ICD/ANX/PSY Bilateral CLT/AMP 100 50 <50 <50 5 28 6 OCD/ANX Bilateral CLT/AMP 100 0 6 27 21 OCD/DEP Bilateral CLT/AMP 100 0 7 30 20 + + + L Thal/MD OCD/ANX/DEP L CLT/AMP 70 >50 90 8 31 18 + + OCD/ANX/DEP L CLT/AMP <50 9 29 13 OCD/DEP Bilateral CLT 100 75 10 59 16 DEP/ICD R AMP 100 100 11 50 13 ANX/DEP L CLT/AMP 90 90 Abbreviations: ECT, electroconvulsive theraphy; GK, gamma knife; VNS, vagal nerve stimulator; L, left; Thal, thalamus; MD, mediodorsal nucleus of the thalamus; R, right; CLT/AMP, central lateral thalamotomy and anterior medial pallidotomy; PSY, psychotic disorder; BIP, bipolar affective disorder; ANX, anxiety disorder; OCD, obsessive-compulsive disorder; ICD, impulse control disorder; DEP, major depression; F, frustration. Percentages correspond to relief estimation by the patients. D. Jeanmonod et al / Neurosurg Clin N Am 14 (2003) 251–265 259

Magnetoencephalographic data network (comprising conceptual, emotional, mnestic, and attentional functions) is dis- Preoperative MEG recordings demonstrated cussed later in this article. a marked increase of power in the theta domain 2. This hyperpolarized state is the source of (4–8 Hz), with the presence of a variable number calcium T-channel deinactivation [31], caus- of identifiable peaks (see Fig. 3). In addition, ing the production of low-threshold calcium there was an increase of beta power and of the spike (LTS) bursts by thalamic (see Fig. 1) coherence between the theta domain and the and/or reticular neurones. alpha and beta (9–13 Hz and 13–30 Hz) ranges. 3. Neurones in such a state impose a slow After surgery, we observed a clear-cut decrease rhythmicity to the thalamocortical loops they in theta power down to values comparable with are part of, being locked in the theta low- those of controls and the reappearance of a nor- frequency domain by their ionic properties. mal alpha peak at or above 10 Hz. There was also Recurrent divergent corticothalamic and retic- a reduction of beta power. Interfrequency co- ulothalamic projections back to the thalamus herence came down close to that of controls. In provide the necessary coherent diffusion terms of source localization, Fig. 4 displays in of these frequencies to various related cor- Patient 2 the presence of a domain of increased tical areas. Our MEG recordings (see Fig. 3) theta activity (selected according to his power indeed demonstrate increased theta power. Its spectrum between 4 and 10 Hz) in the right-sided existence has also been revealed by EEG medial and lateral temporopolar, anterior para- studies [9,12] and may be directly correlated hippocampal, orbitofrontal, and basal medial with cortical [14] and thalamic [13] hypome- prefrontal cortices. After surgery, this paralimbic tabolism in PET studies. theta focus disappeared. There was maintenance 4. The final step in the description of this of well-defined bilateral mu Rolandic activity as syndrome is the proposed existence of acti- well as discrete temporal and occipital, probably vation of high-frequency (beta and gamma) normal, slow rhythmicity. The Rolandic mu cortical domains as the result of an asymmet- rhythm, which is considered a normal EEG com- ric corticocortical GABA-ergic collateral in- ponent, was also described in individuals with hibition [20,21]. The proposed mechanism, an mild to moderate anxiety, irritability, and emo- ‘‘edge effect’’ as observed in the retina as tional instability [28]. a result of lateral inhibition, would result from the asymmetric inhibition between Discussion a low-frequency cortical area and neighboring high-frequency domains, providing a ring of The neuropsychiatric thalamocortical dysrhythmia reduced inhibition onto, and thus activation of, the cortex surrounding this low-frequency Evidence is presented both at the single cell area. Our coherence studies [20,21] support and MEG levels that a TCD, as previously defined the proposition of such an edge effect, as [18–23], underlies certain of the main neuropsy- evidenced by an increased multifrequency chiatric disorders, namely, psychotic and affective coherence between theta and beta domains, disorders as well as ICD and ICD. This TCD is which is an event seen with much less characterized by the following sequential set of prominence in the normal brain. This activa- events (Fig. 5): tion of high-frequency areas might express 1. Hyperpolarization through disfacilitation itself through abnormal EEG spiking activity, and/or overinhibition of thalamic relay and/ as demonstrated in psychotic patients [8,11]. or reticular cells by the disease source. In psychotic disorders, such a source anomaly Surgical control of the thalamocortical may be found in the paralimbic cortical dysrhythmia domain or in the paralimbic striatum [16,29,30], with the cortical anomaly pro- Considering the central role of resonant viding corticothalamic disfacilitation and the oscillatory thalamocortical properties in the gen- striatal anomaly providing pallidothalamic eration of normal hemispheric functions, we overinhibition. The possibility of the alterna- propose a surgical approach that does not imply tive triggering of a neuropsychiatric disorder a reduction of functional thalamocortical loops by a chronic dysfunction of the cognitive and is based on regulation toward normality of 260 D. Jeanmonod et al / Neurosurg Clin N Am 14 (2003) 251–265 D. Jeanmonod et al / Neurosurg Clin N Am 14 (2003) 251–265 261

Fig. 5. Schematic diagram of the thalamocortical circuits that support the thalamocortical dysrhythmia mechanisms. Three thalamocortical modules are shown, each with its specific (yellow) and nonspecific (green) thalamic relay cell projecting to the cortex and reticular nucleus, one (blue) pyramidal cell with its corticothalamic and corticoreticular output, and two reticular cells (red) with their projections back to thalamic relay cells. Three thalamocorticothalamic loops are thus displayed, with their feed-forward thalamocortical activation and their respective recurrent feedback corticothalamic and corticoreticulothalamic activation inhibition. The conjunction of the specific and nonspecific loops is proposed to generate increased cortical activation through temporal coincidence. Either thalamic cell disfacilitation or overinhibition (central module) hyperpolarizes thalamic cell membranes. This allows the deinactivation of calcium T- channels and the generation of low-threshold calcium spike bursts, thus resulting in low-frequency thalamocortical oscillation. Divergent corticothalamic and reticulothalamic projections (from central module to left module) provide the substrate for low-frequency coherent discharge of an increasing number of thalamocortical modules. At the cortical level, low-frequency activation of corticocortical inhibitory interneurones (red), by reducing lateral inhibitory drive, results in high-frequency coherent activation of the neighboring (right) cortical module (edge effect). (See also Color Plate 3.)

dysrhythmic thalamocortical oscillations as docu- and oversynchronization, which can be accom- mented by our MEG data. For this purpose, we plished by a carefully placed and restricted lesion use small pallidal and medial thalamic lesions, the in the medial thalamus. A similar if less well- goal of which is to make subcritical the increased defined procedure has been known as medial low-frequency thalamocortical generation, in thalamotomy since the dawn of stereotactic other words, to move the dynamic properties of neurosurgery [32]. The rationale and general the system away from the increased low-frequency results of this operation are in full accordance coherent activity. The CLT has as a goal a with long-standing electrophysiologic studies [33] reduction of low-frequency overamplification demonstrating the progressive spread of the b Fig. 4. Magnetoencephalographic source localization for Patient 2. Projection of 4- to 10-Hz activity onto the patient’s MRI examination before (eight top images) and after (eight bottom images) surgery. The thalamocortical dysrhythmia of this patient is localized in the right-sided paralimbic domain comprising the temporopolar, anterior parahippocampal, orbitofrontal, and basal medial prefrontal areas. This low-frequency focus disappears after surgery. (See also Color Plate 2.) 262 D. Jeanmonod et al / Neurosurg Clin N Am 14 (2003) 251–265 cortical recruiting response during low-frequency quality of CLT has been documented in a large stimulation of the medial thalamus. In addition, group of patients suffering from other TCDs the thalamic nonspecific system has been shown to (manuscript in preparation). A future systematic serve as a temporal coincidence activator when study will address this issue concerning the summed with the thalamic specific input to the coupling of CLT and AMP lesions. The observed same cortical site [34]. We have taken advantage patient histories already dispel the possibility of of physiopathologic data to focus our medial adverse postoperative prefrontal abulic manifes- thalamotomy target on the posterior part of the tations or personality changes, however, and CLN, where more than 95% of the cells produced neither patients nor families complained about spontaneous LTS bursts (see Fig. 1) and/or were a personality reduction or alteration. unresponsive to stimuli [18]. By doing so, we restrict the lesion to the affected site and spare the Thalamocortical slow burn other medial thalamic subnuclei, which, over time, Evidence of progressive cortical [36] and seem to have taken over the function originally thalamic [16] atrophy has been documented in fulfilled by the CLT. This is a reasonable conjec- patients suffering from neuropsychiatric disor- ture given the fact that patients experience no ders. This phenomenon leads to thalamocortical obvious reduction in sensory, motor, or cognitive self-destruction. One likely mechanism for such abilities after CLT. The second target, the AMP, morbidity is a persistent increase in calcium entry addresses the issue of reducing thalamic excess into the thalamic cells generating LTS bursts, inhibition. In the case of neuropsychiatric disor- which may cause long-term deleterious effects (eg, ders, this reduction in inhibition (or disinhibition) via calcium-triggered apoptosis). Furthermore, is attained by targeting the paralimbic anterior the continuous high-frequency cortical activation internal pallidum [30]. caused by the edge effect provides a framework The results presented here for 11 patients for the development of excitotoxicity of cortical unequivocally demonstrate that this surgical cells. Under these conditions, the system may fall approach has significant therapeutic value against into a self-reduction mode, a sort of ‘‘slow burn’’ symptoms pertaining to psychotic, major depres- with loss of neuronal substrate. This process sive, and bipolar disorders as well as OCD and serves to reinforce the TCD, however, as the level ICD. There is, however, a clear-cut limit concern- of thalamocortical and corticothalamic activation ing reactive, as opposed to endogenous, symp- decreases and exposes thalamic cells to more toms in the domains of anxiety, depression, and disfacilitation and, consequently, to more LTS frustration. Postoperative reactive decompensa- burst generation and more low-frequency pro- tions are understandable by the fact that the duction. If this is confirmed, the surgical treat- disinhibitory approach presented here may result ment of a TCD would not only provide symptom in a momentary hyperactive phase during the relief and disease control but would additionally acute postoperative period. This condition serves acquire a protective role. to emphasize the fundamental difference between the procedure described here and the other avail- The cognitive factor able surgical approaches, such as capsulotomies, cingulotomies, and subcaudate tractotomies, in By cognitive, we mean conceptual, emotional, which the goal is a reduction of cortical activity mnestic, and attentional functions, which are through thalamocortical disconnection. supported by the activation of the widespread As expected given the nature of the lesions, we paralimbic (or mesocortical) and association net- did not observe permanent classical postoperative works [3] located in dorsolateral and medial prefrontal deficits. Patient 6 presented with neuro- prefrontal, orbitofrontal, cingulate, posterior pari- psychologic impairments in the executive and etal, insular, and medial temporal areas. Accumu- memory domains, which were fully compatible lating evidence from EEG and MEG studies with her anxiodepressive postoperative reactive underscores the fact that conceptual and mnestic profile [35] and happened in the context of [10,37–39] as well as emotional [40] activation in a maintained general cognitive performance and human beings increases low-frequency theta activ- improved general IQ. Patient 3 presented with ity. A low-frequency increase may thus arise either clear-cut neuropsychologic improvements in a large on the basis of a disease-related or endogenous number of domains, including prefrontal and per- abnormal input to the thalamus (as the result of ceptual functions. The neuropsychologic-sparing a micro- or macroscopic brain anomaly) or via D. Jeanmonod et al / Neurosurg Clin N Am 14 (2003) 251–265 263 a ‘‘top-down’’ mechanism driven by mental activity related manifestations, the question may be raised and generating low frequencies on a reactive basis. as to their origin and their resistance to surgery in This finding provides a substratum for (1) the view of the fact that anxious and depressive appearance of reactive phenomena in the post- symptoms on the other side may respond to operative period, (2) the genesis of chronic reactive surgery. One possibility is that these manifesta- psychiatric disorders with a long-term unsolved tions may be responsive to surgery in the measure mental conflict at their source, and (3) the grouping in which they are endogenous. Another way of of many if not all neuropsychiatric disorders into viewing this is that whereas all anxiodepressive a dynamic realm of thalamocortical dysfunction. and frustration-related disorders are of reactive Indeed, whether their triggering mechanism is origin, some of them may be self-entertained, thus endogenous or reactive, they may be regarded as presenting the observed resistance to surgery. mirroring an uninterrupted continuum of uncon- Along this line, postoperative anxiety reactions trollable and thus disturbing low-frequency dis- may be initially strong but amenable to external tortions of the cognitive network. In this sense, the and internal relieving factors and may thus loss of control that may occur in any healthy human subside relatively quickly and easily (Patients 1 brain during strong stimulus-bound transitory and 2). To the contrary, frustration/anger and emotional reactions may be viewed as a short-lived conceptual rigidity do not cause direct suffering, and in this sense not unhealthy (although some- and frustration is even self-entertained. Non- times quite undesirable) TCD phenomenon. acceptance of a given situation recedes only when Our clinical observations show that surgery the person has decided that it should indeed can provide marked reduction or even suppres- recede. This mental profile may thus represent sion of the disease-related TCD but does not a stronger challenge for the relevant network and address the cognitive reactive manifestations of correlate with more difficult and protracted post- the patient to her/his new postoperative situation. operative evolutions (Patients 3, 4, 6, and 7). Fear, expectations, despair, and frustration may even increase for a time after surgery, representing Summary a powerful cognitive source for increased low- frequency activity necessitating intensive psycho- Clearly, more clinical experience must be therapeutic support. Our experience has shown amassed to define in detail the possibilities of this that this situation provides a new chance to surgical approach in disabling neuropsychiatric approach and solve long-standing resistant emo- disorders. We propose, however, that the evidence tional conflicts by psychotherapy, however. for benign and efficient surgical intervention Such postoperative reactive manifestations are against the neuropsychiatric TCD syndrome is represented by anxious or anxiodepressive states already compelling. The potential appearance of and are also characterized by nonacceptance/ strong postoperative reactive manifestations re- frustration postures possibly leading to depression quires a close association between surgery and or by lack of adaptation to the new situation. psychotherapy, with the latter providing support Distinctive characteristics indicating a reactive for the integration of the new situation as well as basis to these observations are (1) symptomatology the resolution of old unresolved issues. not present before surgery in relation to the new postoperative situation (eg, Patient 9), (2) large variations across time and according to situations Acknowledgments (eg, Patients 4, 6, and 7), (3) difficulties in describing The authors thank V. Streit for histologic the symptoms (eg, Patient 6), and (4) symptoms work, C. Lebzelter for psychotherapeutic support fitting well with the personal internal and situa- to the patients, J. Dodd for instrumentation, and tional profile (eg, Patients 1, 2, 3, 4, 6, and 7). V. Bu¨gler for secretarial help. Psychotic and affective (bipolar and major depressive) disorders described are viewed here as having an endogenous origin and are well References controlled by the CLT AMP. OCD and ICD / [1] Freeman W. Transorbital leucotomy. Lancet 1948; manifestations also receded after surgery, but the 2:371–3. issue of the endogenous or reactive origin of these [2] Moniz E. Les premie` res tentatives ope´ratoires dans le disease entities remains as yet unresolved. As for traitement de certaines psychoses. Encephale 1936; the postoperative anxiodepressive and frustration- 31:1–29. 264 D. Jeanmonod et al / Neurosurg Clin N Am 14 (2003) 251–265

[3] Mesulam MM. Principles of behavioral neurology. [20] Llinas R, Ribary U, Jeanmonod D, et al. Thala- Philadelphia: FA Davis; 1985. mocortical dysrhythmia. I. Functional and imaging [4] Ballantine HT, Levy BS, Dagi TF, Giriunas IB. Cin- aspects. Thalamus 2001;1:237–44. gulotomy for psychiatric illness: report of 13 years’ [21] Llinas R, Ribary U, Jeanmonod D, Kronberg E, experience. In: Sweet WH, Obrador S, Martin- Mitra PP. Thalamocortical dysrhythmia: a neuro- Rodriguez JG, editors. Neurosurgical treatment in logical and neuropsychiatric syndrome character- psychiatry, pain and epilepsy. Baltimore: University ized by magneto-encephalography. Proc Natl Acad Park Press; 1977. p. 333–53. Sci USA 1999;96:15222–7. [5] Knight G. Further observations from am experience [22] Ribary U, Jeanmonod D, Kronberg E, et al. of 660 cases of stereotactic tractotomy. Postgrad Thalamo-cortical spatio-temporal dynamics and its Med J 1973;49:845–54. alterations in human brain pathology. Neuroimage [6] Bingley T, Leksell L, Meyerson BA, Rylander G. 2000;11:S68. Stereotactic anterior capsulotomy in anxiety and [23] Schulman J, Horenstein C, Ribary U, et al. Thala- obsessive-compulsive states. In: Laitinen LV, Liv- mocortical dysrhythmia in depression and obsessive- ingston KE, editors. Surgical approaches in psychi- compulsive disorder. Neuroimage 2001;13(6):S1004. atry. Lancaster: Medical and Technical Publishing; [24] Morel A, Magnin M, Jeanmonod D. Multiarchitec- 1973. p. 159–64. tonic and stereotactic atlas of the human thalamus. [7] Spiegel EA, Wycis HT. Stereoencephalotomy, parts J Comp Neurol 1997;387:588–630. I/II. New York: Grune & Stratton; 1952/1962. [25] Mazziotta JC, Toga AW, Evans AC, Fox P, [8] Heath RG. Studies in schizophrenia. Cambridge: Lancaster J. A probabilistic atlas of the human Harvard University Press; 1954. brain: theory and rationale for its development. [9] Pascual-Marqui RD, Lehmann D, Koenig T, et al. Neuroimage 1995;2:89–101. Low resolution brain electromagnetic tomography [26] Ramirez R, Horenstein C, Kronberg E, Ribary U, (LORETA) functional imaging in acute, neuro- Llinas R. Anatomically constrained iterative algo- leptic-naive, first-episode, productive schizophrenia. rithm for MEG-based mapping of electrical activity Psychiatry Res Neuroimaging 1999;90:169–79. in the time and frequency domains. Neuroimage [10] Prichep LS, John ER. QEEG profiles of psychiatric 2001;13(6):S227. disorders. Brain Topogr 1992;4:249–57. [27] Dale AM, Fischl B, Sereno MI. Cortical surface- [11] Spiegel EA, Wycis HT. Thalamic recordings in man based analysis. I: segmentation and surface recon- with special reference to seizure discharges. Electro- struction. Neuroimage 1999;9(2):179–94. encephalogr Clin Neurophysiol 1950;2:23–7. [28] Niedermeyer E. The normal EEG of the waking [12] Sponheim SR, Clementz BA, Iacono WG, Beiser adult. In: Niedermeyer E, Lopes Da Silva F, M. Clinical and biological concomitants of resting editors. Electroencephalography. Basic principles, state EEG power abnormalities in schizophrenia. clinical applications and related fields. 4th edition. Biol Psychiatry 2000;48:1088–97. Baltimore: Williams & Wilkins; 1999. p. 149–73. [13] Buchsbaum MS, Nenadic I, Hazlett EA, et al. [29] Kandel ER. Disorders of thought and volition: Differential metabolic rates in prefrontal and tem- schizophrenia. In: Kandel ER, Schwartz JH, Jessel poral Brodmann areas in schizophrenia and schizo- TM, editors. Principles of neural science. 4th typal personality disorder. Schizophr Res 2002;54: edition. New York: McGraw-Hill; 2000. p. 1188– 141–50. 208. [14] Hazlett EA, Buchsbaum MS, Byne W, et al. Three- [30] Morel A, Loup F, Magnin M, Jeanmonod D. Neu- dimensional analysis with MRI and PET of the size, rochemical organization of the human basal ganglia: shape, and function of the thalamus in the schizo- anatomofunctional territories defined by the dis- phrenia spectrum. Am J Psychiatry 1999;156:1190–9. tributions of calcium-binding proteins and SMI-32. [15] Bogerts B. Neuropathology of schizophrenias. J Comp Neurol 2002;443:86–103. Fortschr Neurol Psychiatr 1984;52:428–37. [31] Llinas R, Jahnsen H. Electrophysiology of mam- [16] Shenton ME, Dickey CC, Frumin M, McCarley malian thalamic neurons in vitro. Nature 1982;297: RW. A review of MRI findings in schizophrenia. 406–8. Schizophr Res 2001;49:1–52. [32] Hećaen H, Talairach T, David M, Dell MB. Coagu- [17] Llinas R, Ribary U, Contreras D, Pedroarena C. lations limiteés du thalamus dans les algies du The neuronal basis for consciousness. Philos Trans syndrome thalamique. Rev Neurol 1949;81:917–31. R Soc Lond B Biol Sci 1998;353:1841–9. [33] Morison R, Dempsey E. A study of thalamocortical [18] Jeanmonod D, Magnin M, Morel A. Low-threshold relations. Am J Physiol 1942;135:281–92. calcium spike bursts in the human thalamus: com- [34] Cummings JL. Clinical neuropsychiatry. New York: mon physiopathology for sensory, motor and limbic Grune & Stratton; 1985. positive symptoms. Brain 1996;119:363–75. [35] Corvit A, Wolf OT, de Leone MJ, et al. Volumetric [19] Jeanmonod D, Magnin M, Morel A, et al. Thal- analysis of the prefrontal regions: findings in aging amocortical dysrhythmia. II. Clinical and surgical and schizophrenia. Psychiatry Res 2001;107: aspects. Thalamus 2001;1:245–54. 61–73. D. Jeanmonod et al / Neurosurg Clin N Am 14 (2003) 251–265 265

[36] Llinas R, Leznik E, Urbano F. Temporal binding ciation cortex in calculating and thinking. Neurosci via cortical coincidence detection of specific and Res 1994;19:229–33. non-specific thalamocortical inputs: a voltage de- [39] von Stein A, Sarnthein J. Different frequencies for pendent dye imaging study in mouse brain slices. different scales of cortical integration: from local Proc Natl Acad Sci USA 2002;99:449–52. gamma to long range alpha/theta synchronization. [37] Klimesch W. EEG alpha and theta oscillations Int J Psychophysiol 2000;38:301–13. reflect cognitive and memory performance: a review [40] Machleidt W, Gutjahr L, Mu¨gge A. Grundgefu¨hle: and analysis. Brain Res Rev 1999;29:169–95. Pha¨nomenologie, Psychodynamik, EEG-Spektrala- [38] Sasaki K, Tsujimoto T, Nambu A, Matsuzaki R, nalytik. Berlin: Springer Verlag; 1989. Kyuhou S. Dynamic activities of the frontal asso- Neurosurg Clin N Am 14 (2003) 267–274

Electrical stimulation of the anterior limbs of the internal capsules in patients with severe obsessive-compulsive disorder: anecdotal reports Bart J. Nuttin, MD, PhDa,*, Loes Gabriels, MD, MScEngb, Kris van Kuyck, Drsc, Paul Cosyns, MD, PhDd aDepartment of Neurosurgery, Katholieke Universiteit Leuven and University Hospitals Leuven, Herestraat 49, B-3000, Leuven, Belgium bDepartment of Psychiatry, UZA, Antwerp, Belgium cDepartment of Neurosciences and Psychiatry, Katholieke Universiteit Leuven, Leuven, Belgium dDepartment of Psychiatry, UZA, Antwerp, Belgium

Electric stimulation of the anterior limbs of the Materials and methods internal capsule was introduced as a treatment for In six patients with severe incapacitating patients with severe incapacitating and treatment- and treatment-resistant OCD, two quadripolar resistant obsessive-compulsive disorder (OCD) by electrodes (Model 3887 [4-mm contact spacing, two groups that collaborated closely: the Belgian 3-mm contact length], Pisces Quad Compact; Med- group, which consisted of a psychiatric branch tronic) (Medtronic Inc., Minneapolis, MN) were at the University of Antwerp and a neurosurgical implanted into the anterior limbs of the internal branch at the Katholieke Universiteit Leuven, and capsules by one of the authors (B.N.). Using strict the Swedish group at the Karolinska Institute selection criteria, the patients were screened by in Stockholm [1,2]. Although it was clearly shown a committee for neurosurgery for psychiatric dis- that anterior capsular stimulation induces benefi- orders, an ethical review board, and an expert cial effects in patients with severe OCD, some un- committee [3]. The investigational treatment was published psychiatric and neurosurgical clinical approved by the Leuven and Antwerp local observations may enlighten and document certain hospital ethical standards committees on human effects of this kind of brain modulation. experimentation and was in accordance with the declaration of Helsinki of 1975 (1983 revision and subsequent revisions). Patients gave written and witnessed informed consent. The stimulation targets were similar to those aimed for in the * Corresponding author. anterior capsulotomy [1,3]. The electrodes were E-mail address: [email protected] placed 3 mm anterior to the anterior border of (B.J. Nuttin). the anterior commissure and entered the brain The authors acknowledge the FWO (project via precoronal burr holes. The three most ventral G.0273.97.N) and Research Council of the Katholieke stimulating contacts were placed in the internal Universiteit Leuven (project OT-98-31 and VIS-project capsule, and the fourth one was sited dorsal to ZKB1159) for financial support, and the Medtronic the internal capsule. Although psychiatrists and QUEST program (L1170) for providing the stimulating devices. psychologists blindly evaluated obsessions, com- The authors (B.N., L.G., and P.C.) have been pre- pulsions, anxiety, depression, and several other viously involved in research and/or education that was relevant parameters using validated psychiatric in part sponsored by Medtronic. One author (B.N.) is and neuropsychologic scales, we only report here also filing a patent together with Medtronic. on some observations that interested the authors

1042-3680/03/$ - see front matter Ó 2003, Elsevier Science (USA). All rights reserved. doi:10.1016/S1042-3680(02)00117-1 268 B.J. Nuttin et al / Neurosurg Clin N Am 14 (2003) 267–274 and on responses to stimulation subjectively re- harmful household product. He had strong ideas ported by the patients. Unless otherwise defined that a road accident or drowning could have hap- in the text, one author (B.N.) changed the stim- pened to someone and that he would not have ulation parameters (amplitude varying between given the necessary assistance or would not have 0 and 10.5 V, frequency varying between 5 and alerted the emergency services. He contacted his 130 Hz, and pulse width varying between 60 and sister (or the police) several times a day to check 450 microseconds) and the stimulation contacts that no accidents had taken place where he had of the quadripolar electrodes. Two other authors passed. He could not take a bus, and when in (L.G. and P.C.) evaluated the patients. During a car, he laid on the back seat with his eyes closed all evaluation sessions, these authors as well and a towel covering his face. as the patients and their families were blinded Patient 5, a 40-year-old woman, had severe for stimulation condition. In some conditions, the contamination obsessions and was overly con- patients may have guessed that the stimulator cerned with bodily waste and secretions. Going to was on because of the simulation-induced effects. the toilet took an hour (two or three times per If the patients received a programming device, day) and required assistance from her mother they were not blinded. In that condition, how- both day and night. Washing herself had become ever, the patients still did not know when the impossible, and a nurse came daily to help her stimulator was automatically switched on or off. with this routine. She had excessive and intrusive sexual thoughts about masturbating. She had an urgent need to ask for reassurance and to Subjects repeat sentences and statements for herself. She The first patient (Patient 1) was a 35-year-old also hoarded old magazines and empty cigarette man who suffered from many different obsessions packs. and compulsions, including washing, counting, Patient 6 was a 37-year-old man who suffered and rereading-rewriting-recalculating rituals as from mostly aggressive obsessions. He had con- well as concerns with dirt, germs, and household tinuous intrusive thoughts and flashing images items (white-spirit). His ‘‘major’’ current obsession of harming others by poisoning, strangling, was ‘‘the sound of silence.’’ He avoided being in stabbing, or drowning. He also had a fear of a quiet environment and listened to loud beat writing compromising things. He avoided having music whenever he could. He had an exces- pens and paper in the same room and could only sive fear of hair growth and detested aging. A read the newspaper when no ballpoint pen was comorbid undifferentiated somatoform disorder near. He had excessive checking compulsions and was present with persistent complaints of vertigo, cognitive escapes (rationalizing) to counter these hypersomnolence, and loss of energy. aggressive thoughts. He could not drive a car Patient 2, a 52-year-old woman, suffered from anymore, because his habit of always checking in excessive concerns with urine and feces and ob- the back mirror made driving dangerous. sessive thoughts about harming others after having touched herbs and plants. Compulsive hand washing, checking, and a compelling urge to ask Observations and subjective reports questions were omnipresent. Acute changes with stimulation in the early Patient 3 was a 38-year-old woman who had postoperative period excessive toilet, washing, and counting rituals; buying compulsions; and the compelling urge Bilateral square wave pulse stimulation was to touch. The thought that persons, objects, and started immediately after surgery. Effects that ap- things might not be real preoccupied her, and peared immediately after starting anterior capsu- she developed a whole range of compulsions to lar stimulation are presented in the following prove their existence and reality. Her parents list. We never started chronic stimulation if always had to stay within eyesight or, when this these effects were interpreted as side effects and if was not possible, had to talk to her continuously they persisted longer than 1 minute. to reassure her. Subjective feelings reported by the patient Patient 4 was a 35-year-old man who had strong fears of harming others by not being Anxiety: the patient may be tense or may be careful enough and obsessions about poisoning so anxious that he/she experiences an im- others by contamination with nicotine, tar, or any pression of death B.J. Nuttin et al / Neurosurg Clin N Am 14 (2003) 267–274 269

Relief of anxiety The list includes subjective feelings and objec- Feeling of happiness tive signs that appeared immediately after start- Sadness ing anterior capsular stimulation and may have Nausea disappeared within 10 to 30 seconds, although General unpleasant feeling some of these effects remained during continuous Strange feeling stimulation. These effects were sometimes repro- Feeling of fainting ducible with similar amplitudes and sometimes Feels something in the head not reproducible. Apart from the feelings and Sensation of warmth (in the face, chest, signs mentioned here, there were also occasions throat, feet, and/or legs) when the patient did not report anything when Tingling sensation (in the face or chest) being stimulated at 10.5 V, 200 microseconds, and Feels something in the throat 100 Hz or that the reaction was unclear. Difficulty in swallowing Patient 1 reported paresthesias in his face. At Headache higher amplitudes, he felt uncomfortable, as if he Sleepiness was going to pass out. An improvement in mood Peculiar thoughts (eg, patient wants french was induced by using the three most ventral con- fries, patient feels that the stimulator is tacts as cathodes and the uppermost contact as working like a compact disk that is turning, an anode with an amplitude of 3 V and with the patient thinks about tennis) pulse lasting for 210 microseconds at a frequency Visual symptoms and visual hallucinations of 100 Hz. The patient himself did not feel any (eg, patient sees black pieces of dust every- improvements in his obsessive thoughts during where, patient has a black visual field on the this first stimulation episode, but a friend and his left side and a white one on the right side, mother both remarked that he seemed less tense. everything becomes darker, the wall seems When the stimulator was switched off for 3 days, to come closer, everything becomes smaller, Patient 1 started to comment on his noise ob- patient has an impression that the room session and felt worse. moves, patient sees fog, patient sees a certain At first, the therapeutic effects were not at all person he/she has heard about but has pronounced in either Patient 1 or Patient 2. It was never seen sitting beside her/him, patient only in Patient 3 that the authors saw marked has an impression of being in a tunnel, acute changes as have been described elsewhere patient has an impression that the ceiling is [1]. She was stimulated with either the three most dirty) ventral contacts programmed as cathodes and the Feels difficulty in breathing uppermost contact as an anode or with all four Becomes nervous and has the impression contacts programmed as cathodes and the stim- that time passes quickly ulator as an anode at 4 V using pulses of 210 Feeling of heavy legs microseconds at a frequency of 100 Hz. Clear Frontal pain improvements in obsession, compulsion, depres- One patient reported vaginal contractions on sion, and anxiety were noted. Gradually, the three occasions amplitude necessary to induce those beneficial effects rose to 6.5 V. Similar high voltages applied Objective signs noted by the observer in Patient 1 and Patient 2 also induced more Patient makes a deep sigh pronounced beneficial effects in those patients. Skin of the face turns red The same stimulation parameters were used in Sweating in the face and hands Patient 4, Patient 5, and Patient 6 with similar Patient says meaningless words during 20 beneficial effects as in Patient 2 and Patient 3. seconds Immediately after stimulation with these pa- Patient repeats words or sentences continu- rameters, Patient 4 and Patient 6 uttered verbal ously perseverations for about 10 to 15 seconds (per- Repetitive movements with the hands haps the equivalent of recurrent obsessive thoughts Laughing or crying and repetitive compulsive acts). Patient 4 reported Yawning feeling happy and inclined to laugh, although he Patient speaks louder could state no reason for this cheerfulness. He Orofacial or total body dyskinesias once reported seeing a beautiful woman exactly Dysarthria at the moment the stimulator was switched on. 270 B.J. Nuttin et al / Neurosurg Clin N Am 14 (2003) 267–274

When prompted to elaborate on this, he said it without the patient knowing. Both patients awoke was only a brief flash in his mind that had im- at night about 1 minute after the stimulator was mediately disappeared. switched off; could not continue sleeping; were Patient 5’s sister kept a diary during the pa- anxious, tense and depressed; and were trapped tient’s postsurgical hospitalization. ‘‘No change’’ in obsessional thoughts and mental compulsions was noted during visits when stimulators were (constructing and repeating sentences or count- turned off. On days with stimulation, she noted ing to decrease restlessness) as severe as before that Patient 5 was in the best of spirits, had more surgery. They could only continue sleeping after energy, talked, and laughed a lot. She demanded 6:00 AM, when battery power resumed. During much less reassurance than usual. the day, they felt better, as with the continuous When stimulation was first turned on, Patient stimulation regimen, but were tired as a result of 6 heaved several deep sighs before starting his the loss of sleep at night. They both asked to get perseverative sentences. He felt more light-hearted another stimulation paradigm after some days. and stated that he did not see a reason why he Programming the stimulators for 1 minute on would make his life so difficult. He felt casual but and for 1 minute off induced a happy face for not indifferent. He talked a lot more and faster 1 minute with a smile and a sad face during the than before. With higher amplitudes, he felt rest- following minute, and this was tested for 30 less and worked up, he felt his heart beating, and minutes in two patients. The patients did not like he had difficulty in falling asleep. this kind of stimulation algorithm because they never felt continued relief of anxiety and the sudden mood switches made them feel completely Manipulation of stimulation parameters unnatural. Unilateral right-sided stimulation for The amplitudes necessary to induce symptom 2 weeks with the stimulator on during 1 minute reduction rose to between 6.5 and 10.5 V in all six and off during the following minute did not im- patients 3 to 5 months after the intervention. To prove symptoms as compared to the preoperative reduce battery consumption, a programming de- condition. vice was given to Patient 2 and Patient 3 as one of Many contact combinations were tried out. As the strategies. The patients themselves were given an example, using the three most ventral contacts the capacity to use high amplitudes and pulse as cathodes and the uppermost contact as an widths only when they felt they needed it. They anode, the symptoms were reduced in Patient 1 were allowed to change amplitude and pulse and Patient 3. Conversely, doing the opposite in width, whereas frequency was kept constant at the same patients, programming the three most 100 Hz. The upper amplitude and pulse width caudal contacts as anodes and the uppermost thresholds were determined by one of the authors contact as a cathode, symptoms were comparable (B.N.) after ensuring that these parameters did to the prestimulation level. not cause side effects in an acute setting. Patient Stimulation at 100 Hz or 130 Hz and stimu- 3 used 6.5 V when she was at home, 7.5 V when lation at higher pulse widths (210 microseconds she visited someone, and 8.5 V when she was in up to 450 microseconds) induced better symptom a crowd. She told us that if she programmed the reduction than stimulation at 5 Hz or at a pulse amplitude to 9 V, she would start dancing on the width of 60 microseconds. table. This higher amplitude immediately induced palpitations and an excessive need for sleep, Effects of continuous stimulation a condition that she disliked. Decreasing the amplitude to 6.5 V for the sake of saving battery We evaluated the six patients on a regular time made her a bit more anxious and worried, basis and asked peers and family about each a feeling she described ‘‘as if she loses something.’’ patient’s condition. Respective follow-up times She chose 330 microseconds as the optimal pulse are 50 months (Patient 1), 49 months (Patient width and has been using this pulse width for 2), 47 months (Patient 3), 28 months (Patient 4), more than 3 years now. Patient 2 never wanted to 9 months (Patient 5), and 3 months (Patient 6). use the programming device. Patients 2 through 6 reported spontaneously that To decrease battery consumption even more, several symptoms had improved during the period the stimulator was programmed so that it was that stimulation actually took place. Patient 1 automatically switched off at night and switched stated that he experienced only marginal or on at 6:00 AM in Patient 2 and Patient 3, again temporal benefit, but both his friend and the B.J. Nuttin et al / Neurosurg Clin N Am 14 (2003) 267–274 271 authors observed a slight but definite improve- enjoyed going with his children to a playground or ment in mood. His friend also noted that he amusement park, where he had formerly dreaded seemed more peaceful and less preoccupied by his the thought of having strangled or hurt one of the obsessions. He certainly talked less about them other children. He joined his family in a barbeque and was less self-centered, as if the noise obsession party without fear of burning others. He readily was still there but had become less intrusive. finished his ‘‘to do’’ list of chores that he had Patient 2 asked fewer questions and started to always postponed before surgery and felt the need do part of the housekeeping (cleaning, laundry, for new interests. ironing, and cooking) again. She still felt the urge It is clear that anterior capsular stimulation to ask for reassurance but could control it most of may improve some of the symptoms of OCD the time. She could again enter rooms at home patients. Validated tests using strict research where she did not dare to go before surgery. paradigms have been submitted for publication Patient 3 reported a 90% decrease in obsessions elsewhere. A disequilibrium in the systemic and compulsions, felt much happier, was much less environment of the patient in reaction to the anxious, and was not depressed. She started changes attributable to stimulation has to be listening to music again and enjoyed watching taken into account, however. When Patient 2 television, whereas before stimulation, she com- became able to do some of the housekeeping plained loudly each time someone switched the again, she was expected to take up part of the television on. She could re-enter her former responsibilities and regain some independence. bedroom, which had been impossible previously She found it difficult to develop interests of her because she became obsessed with thoughts of own and longed for the continuous closeness of death and mortality. She used clothes she had not her husband. He had been taking care of his wife dared to wear for many years, and she also dared and children for so long that he felt he deserved to touch a cleaning towel, which was previously some relief and the right to engage in activities he impossible. She was a lot more dynamic, attended had had to give up because of her disorder. cultural events, and regained her interest in art Patient 3 became more independent and started history. She actually started to read the books she traveling alone and abroad. She continued to live formerly collected and stored in the cellophane with her parents because this was comfortable, packing in which they came. She stopped hiding but she sometimes wondered if she wanted to care for family meetings. She accompanied her mother for them when they became older and in need of to the cemetery on All Souls’ Day. This was help. She was also sad because she realized that unthinkable before, because she would have she had missed a large part of her young life as become trapped in ideas of nonexistence. a result of the continuous severity of her OCD. Patient 4 reported being more relaxed, was Always being amid older people, she also had happier, and controlled less. He realized that he trouble in building meaningful relationships with could resist the urge to control for accidents, that people of her own age. Patient 5 felt much more his obsessions and compulsions were only prod- self-confident and assertive during stimulation. ucts of his mind, and that there was no real need She had been subdued and lived with her parents, to act on them. He could postpone checking the who, together with her two sisters, had taken parking lot and make short walks. care of her for many years. After stimulation, Patient 5 was less anxious and less tense. Going she wanted to decide for herself at what time she to the toilet was less time-consuming, and she did would come home at night. She also could not not insist on her mother’s assistance during the stand the fact that her sisters were sharing their night anymore. After some weeks of stimulation, deepest feelings with each other and wanted to the nurse who came to wash her before surgery take part in this. She no longer accepted any only paid short reassuring visits but did not need authoritarian behavior from her father or domi- to assist her anymore. She experienced overall a nance by her sisters. She started to go to school 50% reduction in contamination fears and wash- and wanted to live on her own. Conversely, ing compulsions. She continued to ask many she was not symptom-free and thus claimed questions, but she took more initiative, read news- that she could not accept full responsibility for papers, and watched television more frequently. herself and still needed help. She was living She was in a better mood. on disability benefits and controlled her own Patient 6, in whom follow-up is only 3 months, finances. She realized that living in her own felt a lot more relaxed with stimulation. He again apartment was financially unacceptable for the 272 B.J. Nuttin et al / Neurosurg Clin N Am 14 (2003) 267–274 moment. This caused numerous discussions with- the first year of stimulation. No additional in the family. psychotherapy specifically aimed at improving A similar form of increased independence as OCD (eg, exposure and response prevention a result of stimulation could be noted in Patient 1. [ERP]) was given during this period. Patients He was a successful artist. After he started being were asked to remain under the care of their stimulated, he dared to try out new ways of treating psychiatrist for treatment ‘‘as usual’’ and painting, although he never thought about this to come to our centers only as part of the research before surgery. He painted more of his own ideas evaluation. It seems likely, although not yet stud- and developed a style of his own. He received ied in detail, that the stimulated patients re- critical acclaim, and art connoisseurs stated that spond better to drug treatment and behavioral he painted better and more balanced than before therapy than when not stimulated. Only the first implantation of the electrodes. During stimula- four patients have completed their first year of tion, he felt lots of energy, which made him stimulation to date. Patient 1 is still taking the produce many paintings per day. His success same medication, but the dose of benzodiazepine placed social demands on him, and he tended to (BZD) could be slightly reduced for Patient experience more fear of failure and an extreme 2 and Patient 3. For Patient 4, medication was need for endorsement. completely changed 2 years after surgery by the treating psychiatrist. None of the patients engage Impact of stimulation on quality of life in ERP treatment, but they avoid less and seem to involve themselves more in everyday life. Quality of life (QoL) in patients considering neurosurgery for treatment-refractory OCD is low. Severe OCD interferes with the ability to Replacement of batteries work and to do the housekeeping and with When batteries wore out, patients started to education, family relationships, and social life; feel worse again and OCD symptoms resumed at beyond that, physical and psychologic well-being their former intensity over the course of a few is low. After 1 year of stimulation, all six patients days. All patients immediately contacted one of remain defective in several domains of health- the authors (B.N.) when they thought the battery related QoL. Patient 1 was more successful in his was empty, although only two of them had work. Patients 2 through 5 were unemployed at a programmer by which they could really check the time of surgery and did not start to work whether the battery had run low. Stimulation afterward. Patient 2 became more active in was always performed with a symmetric voltage housekeeping tasks, whereas she had been hospi- supply, but because impedances differed some- talized almost continuously for several years what between the left and right sides, the battery before surgery. Patient 3 was much less dependent of one stimulator was usually empty earlier than on her parents, and her family relationships that of the other. Usually, only one of the stim- improved tremendously. Patient 4 went through ulators (either right or left side) had run low a divorce at the time of surgery. He has recently when patients requested a check on battery status started to go out again on Saturday nights and is for sudden worsening, implying that unilateral meeting new people. He also started some stimulation is less effective than bilateral stimula- volunteer work at a library (part-time). Patient 5 tion. Similarly, in observing acute reactions started adult education classes and devotes serious (minutes) when changing stimulation parameters, energy to her studies. it was almost always clearly noticeable that the effects of bilateral stimulation were more pro- Medication, additional treatment, and counseling nounced than those of unilateral stimulation, In the opinion of the authors, the follow-up of whether it was right-sided or left-sided stimula- electrically stimulated OCD patients is compara- tion. The patients’ urge to ask for renewal of the ble to the follow-up of electrically stimulated stimulators was comparable to that of some parkinsonian patients. As parkinsonian patients parkinsonian patients, who are electrically stimu- are still followed by a neurologist, OCD patients lated in either subthalamic nuclei or ventral need follow-up with their psychiatrist. Until now intermediate nuclei, to request this when their and for the sake of scientific methodology, battery runs low. Their condition changed dra- medication was tapered off to a minimum before matically after stimulator renewal, even on the surgical intervention and kept stable throughout operating table, straightaway after the stimulators B.J. Nuttin et al / Neurosurg Clin N Am 14 (2003) 267–274 273 were switched on. In Patient 2, the degree of this men, to migrate more caudally, which induced change fluctuated. The beneficial effects were traction on the extension cable and thus on the sometimes only seen after several days, but the electrode. This traction led to fracture of the left husband once said that we had suddenly altered electrode and worsening of the OCD symptoms 44 his wife into a lovely young flower. He was not months after surgery. The electrode was replaced, used to this appeal, and he felt uneasy about it. and the patient improved again. These sudden changes required adaptations in In Patient 5, the impedance of the uppermost their relationship that were not easy to attain. All contact increased to more than 4000 shortly after patients reacted strongly when the batteries ran implantation, which suggests a broken contact. down. They felt bad again, and although after Therefore, we could only stimulate with three of the a few battery replacements, they knew that these four contacts on the left side, but clear beneficial bad feelings were caused by technical reasons, it effects on OCD symptoms were still seen. was difficult to accept this and to put it into perspective. Patient 4 twice experienced a brief Side effects hypomanic state (2–3 days) after placement of new batteries. His family reported that he was Possible side effects were a major concern when overactive and could not sleep at all for 36 hours. starting this new kind of therapy. There were no He tried to impress others and even threatened intracranial hemorrhages, epileptic fits, or infec- violence if others did not accept him as ‘‘the man tions, although there was a more than usual who could do it all.’’ He felt good about himself swelling of the face in Patient 5 during the first and regretted that his brother (who also had 4 days after surgery, which was prophylactically severe OCD and had committed suicide several treated with antibiotics for 1 week even though years before) could not have benefited from stim- cultures of aspirated subcutaneous fluid remained ulation. He also made some sexual insinuations sterile. and passes directed at some acquaintances, and he On some days, Patient 1 felt the leads and the suddenly demanded a paternity test for his 8-year- stimulator and wanted to take the system out. old son. When questioned about these episodes in During most of the time, however, he wanted a psychiatric interview some time later, he re- to continue the stimulation. He found the ad- membered the hyperactive state but not the show- vantages of the capsular stimulation not comparing off and swaggering behavior. He explained able to the disadvantage of needing to have the that during an argument years earlier, his former stimulators replaced every 8 months. Apart from wife had told him he could never be sure that he was severe OCD, he also suffered from undifferenti- his son’s father and that this had come into his mind ated somatoform disorder. For those reasons, and obsessed him for a few days. He did not persist his electrodes were removed, and he underwent in demanding the paternity test. a bilateral anterior capsulotomy. Other patients complained less about the presence of the stimulators. Complications Fatigue lasting for several months was a major complaint of Patient 3. When she was given the Technical problems and hardware failures option of decreasing the amplitude to get imme- The most important inconvenience was defi- diate relief of her fatigue, she chose to continue nitely the high current amplitude needed to stimulation because she appreciated the fact that maintain acceptable results in the years after the the benefits of the stimulation were so much surgery. Stimulators had to be replaced every 5 to greater than the disadvantage of the fatigue. 12 months in all patients. This makes the therapy Paradoxically, she showed more energy during difficult to use as standard therapy, but one can the daytime and took much more initiative than think of strategies either to improve battery before surgery. All patients reported better sleep technology (eg, rechargeable batteries) or to re- at night. duce battery consumption. The stimulators could Although Patient 3 once reported that one of be replaced under local anesthesia and in an am- the authors resembled her previous lover, no bulatory setting. frontal signs were noted during subsequent neuro- Severe weight loss as a result of severe psychologic testing. Also, no permanent patho- pyelonephritis in Patient 2 caused the left stimu- logic frontal signs were observed in the other lator, which was implanted in the left hemiabdo- patients. Patient 3 complained of a worsening of 274 B.J. Nuttin et al / Neurosurg Clin N Am 14 (2003) 267–274 memory, but neuropsychologic tests before and same technique, similar effects were seen in the 1 year after surgery could not document any patients. The authors still want to stress that difference. Patient 2 was treated for pyelonephri- anterior capsular stimulation remains investiga- tis, which was interpreted as being unrelated to tional and needs optimization, especially to try to the stimulation. Patient 2 lost weight, and Patients solve the problem of the short battery life of the 3 through 5 gained weight to different degrees. stimulators.

References Summary [1] Nuttin B, Cosyns P, Demeulemeester H, et al. Anterior capsular stimulation induces some Electrical stimulation in anterior limbs of internal improvement in severe treatment-resistant OCD capsules in patients with obsessive-compulsive dis- patients. At this stage, not all stimulation-induced order. Lancet 1999;354:1526. eﬀects can be explained. The eﬀects are a valuable [2] Nuttin B, Gabriels L, Cosyns P, et al. Electrical stimulation of the brain for psychiatric disorders. source for further neurophysiologic and neuroan- CNS Spectrums 2000;5:35–9. atomic research. It was reassuring that when the [3] Cosyns P, Caemaert J, Haaijman W, et al. group of Drs Rasmussen, Greenberg, and Friehs Functional stereotactic neurosurgery for psychi- in Providence and the group of Drs Rezai, atric disorders: an experience in Belgium and The Montgomery, and Malone in Cleveland started Netherlands. Adv Tech Stand Neurosurg 1994;21: to operate on OCD patients using exactly the 239–279. Neurosurg Clin N Am 14 (2003) 275–282

Cervical vagus nerve stimulation for treatment-resistant depression Linda L. Carpenter, MDa,b,*, Gerhard M. Friehs, MDb, Lawrence H. Price, MDa aDepartment of Psychiatry, Brown University Medical School and Butler Hospital, 345 Blackstone Boulevard, Providence, RI 02906, USA bDepartment of Clinical Neurosciences and Neurosurgery, Brown University Medical School, 100 Butler Drive, Providence, RI 02906, USA

Intermittent electric stimulation of cranial Approximately 80% of the vagus is sensory nerve X (called the ‘‘vagus,’’ Latin for ‘‘wander- afferent fibers, carrying information to the brain ing’’) via a surgically implanted programmable from the head, neck, thorax, and abdomen [1]. prosthesis was approved by the US Food and Through sensory afferent connections in the Drug Administration (FDA) for medically re- nucleus tractus solitarius (NTS), the vagus has fractory partial-onset seizures in 1997. Only 4 extensive projections to brain regions that are years later, and without the benefit of controlled thought to modulate activity in the limbic system data to establish its efficacy, the VNS Therapy and higher cortex [2–4]. Pathways connecting the system (Cyberonics, Houston, TX) for delivery of NTS with the parabrachial nucleus and the locus cervical vagus nerve stimulation (VNS) had be- ceruleus (LC) carry afferent vagal input to nor- come an approved treatment for medication-re- epinephrine-containing neurons, which reach the sistant or -intolerant depression or bipolar disorder amygdala, hypothalamus, insula, thalamus, orbito- in Europe and Canada. That VNS has come to be frontal cortex, and other limbic structures [5]. regarded as one of the most promising new forms of therapeutic brain stimulation reflects a tremendous need for better long-term treatments of disabling Preclinical investigations of vagus nerve depression as well as great expectations for the stimulation application of new technologies in treating mental illness. The history of its development and The first published report suggesting that VNS practical considerations for its application in psy- directly affected central function appeared in chiatric disorders are reviewed in this article. 1938, when Bailey and Bremner [2] described synchronized activity of the orbital cortex produced by VNS in cats. Slow-wave response in the Anatomy of the vagus nerve anterior rhinal sulcus and amygdala to VNS was The vagus nerve, best known for its para- noted in awake cats with high cervical spinal sympathetic efferent functions, such as autonomic section [3]. Further evidence of VNS effects on the control and regulation of heart and gut viscera, is basal limbic structures, thalamus, and cingulate actually a mixed sensory and motor nerve. was generated by a study in monkeys [4]. A syn- thesis of this preclinical literature led Zabara [6, The authors have received grant/research support 7] to hypothesize and further investigate in dogs from Cyberonics. the notion that VNS would have anticonvulsant * Corresponding author. action. Zabara observed VNS-induced cortical E-mail address: [email protected] electroencephalographic (EEG) changes and sei- (L.L. Carpenter). zure cessation in dogs, leading him to postulate

1042-3680/03/$ - see front matter Ó 2003, Elsevier Science (USA). All rights reserved. doi:10.1016/S1042-3680(02)00121-3 276 L.L. Carpenter et al / Neurosurg Clin N Am 14 (2003) 275–282 that the antiepileptic mechanisms of action of Europe to investigate its role in medically re- VNS would involve direct termination of an on- fractory epilepsy [11]. Results from three open going seizure as well as seizure prevention [8]. In trials and two double-blind studies [11,12] in- preclinical investigations, access to the vagus was dicated a 25% to 30% mean decline in seizure achieved through an abdominal or diaphragmatic frequency and suggested a tendency for anticon- approach. vulsant benefits to be sustained or improved over time with continuing VNS [13,14]. Collective data from the epilepsy trials showed that after 2 years Clinical trials of vagus nerve stimulation of continuous VNS, the categorical response rate for epilepsy (defined as a 50% or greater reduction in seizure In 1988, the first human implants were frequency) reached 43% and was maintained at performed for a pilot study of cervical VNS in the 3-year mark [13]. Although few epilepsy patients with medication-resistant epilepsy who patients receiving VNS have achieved full re- were not candidates for neurosurgery [9]. The mission and become free of anticonvulsant medi- VNS Therapy system commercially manufactured cations, data from these studies supported the use by Cyberonics includes a pocket watch-sized gen- of cervical VNS as a safe and effective adjunct erator that is implanted subcutaneously in the treatment for difficult-to-treat epilepsy [15]. Per- left chest wall in a fashion similar to placement haps more important to the future development of of a cardiac pacemaker [10]. Bipolar electrode clinical VNS applications, the epilepsy studies coils are wrapped around the left vagus nerve near demonstrated that cervical VNS was well toler- the carotid artery in a separate neck incision ated, with adverse events (AEs) rarely leading (Fig. 1). The leads are subsequently tunneled to discontinuation of VNS therapy [16]. Pooled under the skin for connection with the program- data from epilepsy clinical studies (n=454) show mable generator. A telemetric wand connected minimal surgical complications associated with to a portable computer is held to the chest (over implantation: infection without explantation of the patient’s clothing) to assess and control the the device (1.8%), infection with subsequent ex- stimulation parameters in a noninvasive office plantation (1.1%), hoarseness or temporary vocal procedure. cord paralysis (0.7%), and hypesthesia or lower By 1990, the first multicenter pivotal VNS trial left facial paresis (0.7%) [17]. Side effects related was being conducted in the United States and to the intermittent stimulation itself (ie, voice alteration or hoarseness, cough, paresthesia, dys- pepsia) were judged to be mild or moderate most of the time and were noted to decrease over time with ongoing VNS at the same ‘‘dosage.’’ Stim- ulation-related AEs could be diminished by re- programming the device to deliver a lower level of output current. Alternatively, a patient can completely abort a stimulation-induced AE by holding or taping a small magnet over the pulse generator. High continuation rates (72% at 3 years), lack of compliance issues or drug- interactions, and favorable practice and reimbursement economics have all contributed to the success of VNS in the treatment of patients with severe epilepsy. At the time of this publication, more than 20,000 epilepsy patients worldwide have received cervical VNS (S. Perkins, Cyberonics, personal communication 2003), and the treatment has been judged by the American Academy of Neurology’s Technology and Thera- Fig. 1. Schematic representation of placement of the peutics Committee as having ‘‘sufficient evi- NeuroCybernetic Prosthesis manufactured by Cyber- dence ...to rank VNS for epilepsy as effective onics, Inc. for therapeutic left cervical vagus nerve and safe, based on a preponderance of Class I stimulation. evidence’’ [15]. L.L. Carpenter et al / Neurosurg Clin N Am 14 (2003) 275–282 277

Anticonvulsant mechanism of action research assessments during epilepsy trials [24] suggested that VNS was associated with a reduction in de- Ascending projections from the NTS to the pressive symptoms, even in the absence of improve- midline raphe and LC are hypothesized to be the ment in seizures. pathways through which VNS exerts antiseizure Incidental findings of psychiatric improve- and neuropsychiatric effects. Studies in rats ments during clinical trials of anticonvulsant during VNS reveal increases in cellular activity therapies have provided the rationale for further as measured through the oncogene C-fos in the investigation into the potential utility of other amygdala, cingulate, LC, and hypothalamus [18]. drugs like carbamazepine in the management Further support for a role of noradrenergic neuro- of bipolar disorder [25]. In light of the growing transmission comes from a report of suppres- number of modern anticonvulsant agents (eg, car- sion of the antiseizure effects of VNS in animals bamazepine, valproic acid, lamotrigine) that have after lesions of the LC [19]. Although the basic demonstrated beneficial effects in mood-dis- mechanisms of action are unknown, a theoretic ordered patients, a possible role for VNS in the VNS-induced increase in NTS concentration of treatment of depression seemed worthy of in- c-aminobutyric acid (GABA) and/or a decrease in vestigation. The similarities between VNS and NTS glutamate level could explain the antiseizure electroconvulsive therapy (ECT), considered to be activity of VNS [20]. Consistent with this hypoth- the most effective available antidepressant treat- esis, a study of lumbar cerebrospinal fluid (CSF) ment, also supported the hypothesis that VNS may components in epilepsy patients sampled before have primary antidepressant properties. These and after 3 months of VNS showed significant considerations, together with the known anatomic increases in CSF concentrations of GABA and projections of the vagus to regions of the brain trend-level decreases in glutamate [21]. Other involved in mood regulation and the preclini- provocative findings from the CSF study were cal and human CSF data suggesting that VNS trends toward VNS-induced increases in levels of altered major neurotransmitter systems implicated the major metabolite of dopamine, homovanillic in depression, provided a rationale for studying acid (HVA), and the major metabolite of seroto- VNS in a new population of subjects. nin, 5-hydroxyindoleacetic acid (5-HIAA). A positron emission tomography (PET) study in epilepsy patients demonstrated significant VNS-induced modulation of blood flow in key brain struc- Open-label study of vagus nerve stimulation for tures thought to be involved in seizures and emo- treatment-resistant depression (D-01) tion regulation [22]. Blood flow increases were The safety and efficacy of VNS for non- seen in the rostral medulla, thalamus, hypothala- psychotic, chronic, or recurrent treatment-resis- mus, insula, and postcentral gyrus, all with greater tant depression was first studied in an open-label activation on the right side. Bilateral decreases fashion in 30 patients at four academic sites in were seen in the hippocampus, amygdala, and the United States [26]. Patients enrolled in the cingulate gyrus. study, called ‘‘D-01,’’ had either bipolar or unipolar forms of affective illness and continued to be severely symptomatic despite exposure to an Vagus nerve stimulation as a potential average of 16 different clinically administered treatment for depression treatments (including an average of five strictly During the early epilepsy trials of VNS, pa- defined adequate antidepressant trials) for the tients frequently stayed in the same hotel in current major depressive episode. The majority Gainesville, Florida, during follow-up visits at (66%) had undergone a course of ECT for the the study site. An astute observation made by index depressive episode before enrolling in the a hotel clerk and reported to VNS investigator VNS study and undergoing surgical implantation B.J. Wilder was that the VNS patients seemed to of the VNS Therapy system. After a 2-week be in better spirits as time passed. Anecdotal single-blind period for recovery from surgery, the reports of mood improvements apparently un- patients began active VNS with an initial 2-week related to reduction in seizure frequency further period of stimulation adjustment (consistently inspired the VNS investigators to systematically programmed to take place in cycles of 30 seconds assess mood and anxiety symptoms [11,12]. Both ‘‘on’’ and 5 minutes ‘‘off’’ but variable in intensity retrospective data analysis [23] and prospective of output current) and a subsequent 8-week 278 L.L. Carpenter et al / Neurosurg Clin N Am 14 (2003) 275–282 period of fixed-dose VNS. All patients continued the entire sample (n=59) participating in the on stable psychotropic medication regimens. The open-label (D-01) study was followed to the 1- categoric response rate (50% or greater decrease year mark, the response rate was 45% and the in total score on the 28-item Hamilton Rating remission rate was 27%. For those who had Scale for Depression [HRSD]) after 10 weeks of reached the 2-year mark, there continued to be adjunctive VNS was 40% for the first 30 patients. evidence of sustained or even enhanced response Substantial functional improvement was also to VNS (13/24 responders [54%]) [29]. It is worth seen, suggesting acute antidepressant efficacy in noting that changes in dose or type of psychotro- a difficult-to-treat depressed population. A second pic medication and VNS stimulation parameters cohort of 30 patients with treatment-resistant were not controlled after exit from the acute- depression meeting similar inclusion and exclu- phase study, introducing the possibility that the sion criteria was added to the open-label study. A observed improvements were not entirely attri- somewhat less robust response rate of 21% was butable to the ongoing VNS. Nevertheless, the seen at the end of the ‘‘acute phase’’ (10 weeks of association of adjunct VNS with sustained de- active VNS) for this second cohort. The overall pressive symptom reduction and improved func- acute response of 30.5% for the combined cohorts tional status after 2 years is suggestive of (n=60), with 15% meeting criteria for full re- antidepressant efficacy in a naturalistic setting. mission of the depressive episode, was still much higher than what might be expected in such a Pivotal, placebo-controlled study of vagus nerve severely ill and treatment-resistant population, stimulation for depression (D-02) however. Quality of life assessments suggested that VNS Encouraging results from the open-label stud- was associated with improvements in vitality, ies clearly indicated a need for a placebo-con- social function, and mental health domains even trolled investigation of the antidepressant efficacy among patients who were considered VNS acute- of VNS. A large-scale pivotal study was thus phase nonresponders [27]. Safety data from the designed to be almost identical to the open-label D-01 open-label depression study mirrored that of study that preceded it, with the exception of the epilepsy trials, as most side effects were double-blind randomization to either active VNS experienced only during stimulation and were or a sham condition. Additionally, the protocol considered mild. No dose-response relationship exclusion criteria were revised to exclude those was detected when final output current data were with the highest levels of treatment resistance (six examined. Neuropsychologic tests indicated neu- or more failed adequate trials in the index de- rocognitive improvements after VNS relative pressive episode), because they did not appear to baseline, especially in those who experienced to benefit from VNS in the open-label study. After decreased depressive symptoms [27]. The only completing the 12-week acute study, patients in identifiable predictor of response was degree of the sham condition who continued depressed treatment resistance as measured by the number would cross over to active stimulation, and long- of failed antidepressant trials in the index de- term data would be collected on all patients. pressive episode. More severely refractory patients Results of the randomized controlled study, experienced poorer responses to the 10-week VNS called ‘‘D-02,’’ were released to the press in early therapy. 2002. Two hundred twenty-five patients received VNS (2 weeks of VNS stimulation parameter adjustments and 8 weeks of fixed-dose VNS) in an Longer term follow-up results: open-label study ‘‘add-on’’ study design. Acute safety and tolera- of vagus nerve stimulation for depression (D-01) bility of VNS were demonstrated. The placebo Although the short-term results were consid- (sham) response rate was low at 10%, but only ered encouraging, the longer term data generated 15% responded to active VNS, which failed to even more enthusiasm for VNS in treatment- confirm the short-term antidepressant efficacy of resistant depression. After 1 year of VNS, 91% the adjunct therapy statistically [30]. (10/11) of the first cohort of acute-phase study Speculation about why the D-02 treatment responders had maintained their response and group response was inconsistent with the open- 18% (3/17) of the initial nonresponders had label study results included hypotheses about achieved a reduction of depressive symptoms inadequate dosing of VNS stimulation. A pre- sufficient to meet ‘‘responder’’ criteria [28]. When liminary comparison of the output current L.L. Carpenter et al / Neurosurg Clin N Am 14 (2003) 275–282 279 delivered in the D-02 depression study with that more placebo-controlled trials of VNS will be used in the initial open-label depression study and needed or that the present data set will be ade- in epilepsy studies suggests that stimulation set quate to demonstrate an adequate level of anti- at 1.0 mA or higher is associated with higher depressant efficacy. rates of clinical response. Stimulation parameters were at lower settings in the D-02 study compared ‘‘Mechanisms of action’’ research in depression with those used in the initial open-label D-01 depression studies and epilepsy trials. At present, One of the most compelling aspects of the VNS systematic efforts are underway to ‘‘ramp up’’ or development ‘‘story’’ is the growing body of increase the dose of VNS in pivotal study patients clinical neurobiologic research findings that speak who have not yet remitted in hopes of generating to its direct actions on the brain and central data that will better address the possibility of a nervous system. Because antidepressant action of dose-response relationship for VNS in depression. VNS has yet to be empirically established, it is not appropriate to interpret dynamic brain imaging results and other biologic correlates of VNS in Longer term results in the pivotal study (D-02) depressed patients as evidence of its mechanism of In light of the apparent gradual accumulation antidepressant action. That qualification notwith- of more VNS responders over time in the D-01 standing, such data seem to provide converging open-label study, there has been considerable lines of evidence that VNS exerts measurable interest in the clinical course and longer term effects in brain regions and neurotransmitter sys- outcomes experienced by the depressed patients tems implicated in mood disorders. who continue to receive VNS after finishing the A single-photon emission computerized tomo- acute-phase trial. In a preliminary look at data graphy (SPECT) imaging study conducted in six from the first 36 patients to have completed a full depressed patients receiving VNS in the open- year of VNS in the D-02 pivotal study [30], the label study found that compared with normal response rate at the 1-year mark was 44%, which controls, patients had reduced regional cerebral is consistent with 1-year outcomes measured from blood flow (rCBF) to left dorsolateral prefrontal, subjects in the open-label study. Although this anterolateral temporal, and perisylvian temporal finding raises the provocative possibility that VNS structures, including the posterior insula, at base- simply takes longer (ie, longer than the 10 weeks line [31]. After 10 weeks of VNS, these depressed of acute-phase therapy) to exert its antidepressant patients showed increased rCBF in the superior effects, the longer term data do not reflect res- frontal gyrus and right mesial (posterior hippo- ponse under continued controlled conditions. It is campus) and lateral temporal cortex. The 10- possible that changes in antidepressant therapies week trial of VNS seemed to produce resolution or VNS stimulation parameters account for the of classic rCBF abnormalities in depressed pa- growing percentage of patients meeting categori- tients, especially among those showing a favor- cal criteria for antidepressant response over time. able clinical response. More extensive evaluation of follow-up data, with A method for synchronized blood oxygenation larger sample sizes and over longer periods of level-dependent (BOLD) functional MRI (fMRI) exposure to VNS, will eventually be available. was developed to detect signal from the implanted Nevertheless, it will remain difficult, if not im- device and link it to fMRI image acquisition [32]. possible, to parse out the effects of antidepres- In depressed patients, a BOLD fMRI response to sant medications or spontaneous remissions in VNS was shown in areas regulated by the vagus this large cohort and to demonstrate antidepres- nerve: orbitofrontal and parieto-occipital cortex sant effects that can be clearly attributed to bilaterally, left temporal cortex, hypothalamus, adjunct VNS. and left amygdala. This newly developed fMRI Interestingly, recent device regulatory prece- technique was also used to examine whether dents suggest that statistically significant long- BOLD signal changes depend on the frequency term longitudinal results may be appropriate to of VNS [33]. Results confirmed that acute im- support FDA approval, particularly in light of the mediate regional brain activity changes vary pressing need for safe, tolerable, and effective with the frequency or total dose of stimulation. maintenance treatments for patients with severe Additionally, results suggested that VNS exerts a chronic and/or recurrent depressions. In sum- dose-dependent modulatory effect on other brain mary, it seems premature to conclude either that activities, such as hearing a tone. 280 L.L. Carpenter et al / Neurosurg Clin N Am 14 (2003) 275–282

Sleep EEG studies have also measured ap- for ‘‘application of pulsed electrical signal’’ in parent VNS-induced improvement in indices of the treatment of movement disorders, eating sleep macro- and microarchitecture in patients disorders/obesity, anxiety disorders, dementia and with treatment-resistant depression [34]. After 10 Alzheimer’s disease, chronic pain, migraine head- weeks of VNS, the amplitude of sleep EEG ache, and cardiac disease. Several of these areas are rhythms was restored to near-normal levels, and under current clinical investigation, including an patients manifested significantly less awake time open-label study of bilateral diaphragmatic VNS and ‘‘light sleep’’ and significantly more stage 2 (using higher output currents than used in (deep) sleep. cervical VNS applications) for morbid obesity. Lumbar CSF samples were collected in 18 pa- Multicenter trials of cervical VNS for panic tients both before and after VNS therapy, in the disorder, posttraumatic stress disorder (PTSD), pivotal study (D-02) [35]. Consistent with CSF obsessive-compulsive disorder (OCD), and rapid- findings reported for a group of seizure patients cycling bipolar disorder are also underway. receiving 3 months of VNS, the depressed group showed a significant VNS-associated increase in CSF concentrations of HVA, the major dopamine Practical considerations in the use of vagus nerve metabolite. Unlike the CSF results obtained stimulation for psychiatric disorders during the epilepsy trials, however, no changes in CSF GABA were detected. Although categor- Surgical implantation of the VNS Therapy ical clinical response rates were low, the data system is considered a procedure of low technical suggested that an increase in CSF level of complexity for a surgeon with experience in the norepinephrine during VNS was correlated with head and neck area. The surgery typically takes better clinical outcome. This norepinephrine from 30 minutes to 1 hour in the operating room finding is of interest in light of known anatomic and is usually performed on an outpatient or day connections and animal studies demonstrating surgery basis, without subsequent admission to that VNS exerts effects on higher brain regions via the hospital for routine postsurgical recovery or actions at the LC, the major noradrenergic monitoring. General anesthesia is used in most nucleus in the brain. cases, but regional or local anesthesia has been Several other types of biologic measures of used in some centers. A device ‘‘programmer’’ VNS are currently being investigated to evaluate who is knowledgeable about the operation of the the effects of VNS in treatment-resistant depres- VNS Therapy system must be present but not sion, including evoked response potential (ERP) sterile in the operating room to perform lead recordings and provocative neuroendocrine chal- testing before surgical incisions are closed. A 4- to lenge responses. The biologic data thus far suggest 5-cm incision is made in the neck for placement of that VNS has acute and chronic effects on various the two stimulating electrodes, which coil around key indices of brain activity and neurotransmitter the vagus nerve (see Fig. 1). A second incision regulation that have also been implicated in mood similar in size is made in the chest wall (or regulation. As is still the case with most available alternatively in the axillary region for a superior pharmacotherapies, however, the direct mech- cosmetic result) for insertion of the pacemaker- anism of therapeutic action of VNS is still not like pulse generator. A tunneling tool subcutane- known. ously passes the electrodes to the site connection with the pulse generator. The battery life of the currently available pulse generator model is approximately 8 to 12 years. A single surgical Other potential clinical applications of vagus incision is needed in the chest wall to replace the nerve stimulation entire pulse generator once the battery has In addition to the use of VNS for treating expired. medication-refractory epilepsy and depression, Because of the potential for heating of the a number of potential new indications for VNS electric leads, whole-body MRI is contraindicated are being explored. Based on clinical observa- in patients who have the VNS pulse generator tions from epilepsy and depression studies, results implanted. Special ‘‘send-receive coils’’ have been from preclinical and clinical mechanisms of used to concentrate magnetic fields away from the action research, neuroanatomic knowledge, and neck area when MRI of the brain is necessary. market opportunities, Cyberonics, Inc. has patents Patients with the VNS Therapy system are asked L.L. Carpenter et al / Neurosurg Clin N Am 14 (2003) 275–282 281 to carry identification cards and are educated Summary about risks related to being in close proximity to Therapeutic brain stimulation through left cer- strong magnetic fields. vical VNS now has established safety and efficacy The cost of the VNS Therapy system and as a long-term adjunct treatment for medication- surgical implantation for cervical VNS is approx- resistant epilepsy. There is considerable evidence imately $20,000, making it to roughly comparable from both animal and human studies that the to the cost of a course of ECT for depression in an vagus nerve carries afferent signals to limbic and inpatient setting. If the generator battery life is 10 higher cortical brain regions, providing a ratio- years, VNS costs can be calculated at about $2000 nale for its possible role in the treatment of per year. Early success in establishing adequate psychiatric disorders. Open-label studies in pa- terms of coverage and reimbursement by third- tients with treatment-resistant depression have party payers has contributed to the wide-scale produced promising results, especially when availability of VNS for patients with epilepsy response rates at longer term (1 year and 2 years) in the United States. At present, VNS as a treat- follow-up time points are considered. Short-term ment for depression remains investigational in the (10 weeks) treatment with VNS failed to demon- United States; as such, it can be offered only at strate statistical superiority over sham treatment academic centers conducting approved research in a recently completed double-blind study, so protocols. Data regarding the optimal stimula- antidepressant efficacy has not yet been established. tion parameters for antidepressant effects are Longer term data on VNS in depressed patients as extremely limited. Because there is much yet to be well as further information regarding the possible learned about the effects of various stimulation dose-response relation will help to determine the ‘‘doses’’ and patterns of electric pulse delivery, off- place of VNS in the armament of therapeutic label use of VNS is discouraged at this time. Al- modalities available for major depression. though it is tempting to imagine that VNS may replace psychotropic medications and the many undesirable side effects that accompany them, it References is important to bear in mind that VNS has been investigated as an adjunct therapy rather than as [1] Foley JO, Dubois F. Quantitative studies of the vagus nerve in the cat, I: the ratio of sensory and a monotherapy in most cases to date. Patients’ motor studies. J Comp Neurol 1937;67:49–67. expectations for dramatic symptom recovery or [2] Bailey P, Bremer F. A sensory cortical represen- even cure from severe psychiatric illness may be tation of the vagus nerve. J Neurophysiol 1938;1: fueled by the introduction of new technology and 405–12. the highly interventional nature of the device [3] Dell P, Olson R. Projections ‘‘secondaries’’ mesen- implantation surgery. Management of such ex- cephaliques, diencephaliques et amygdaliennes des pectations should be undertaken with great care, afferences viscerales vagales. C R Soc Biol 1951; particularly in depressed patients, who are at 145:1088–91. heightened risk for acting impulsively and self- [4] MacLean PD. Triune brain in evolution: role in pa- destructively on feelings of disappointment and leocerebral functions. New York: Plenum Press; 1990. [5] Van Bockstaele EJ, Peoples J, Valentino RJ. hopelessness. Anatomic basis for differential regulation of the With those caveats in mind, it is useful to rostrolateral peri-locus coeruleus region by limbic consider the relation of VNS to other somatic afferents. Biol Psychiatry 1999;6:1352–63. methods of therapeutic brain stimulation, such as [6] Zabara J. Peripheral control of hypersynchronous ablative neurosurgery, gamma knife neurosur- discharge in epilepsy. Electroencephalogr Clin gery, deep brain stimulation (DBS), ECT, MR Neurophysiol 1985a;6(Suppl):S162. spectroscopy (MRS), and transcranial magnetic [7] Zabara J. Time course of seizure control to brief, stimulation (TMS). On a spectrum of relative inva- repetitive stimuli [abstract]. Epilepsia 1985;26:518. siveness of the procedure, with ablative surgery [8] Zabara J. Inhibition of experimental seizures in at one end and TMS at the other, VNS might be canines by repetitive vagal stimulation. Epilepsia 1992;33:1005–12. ranked in the middle [36]. Future applications [9] Penry JK, Dean JC. Prevention of intractable of VNS may include combining it with brain partial seizures by intermittent vagal stimulation imaging techniques to evaluate immediate dy- in humans: preliminary results. Epilepsia 1990; namic effects and, possibly, to influence regional 31(Suppl 2):S40–3. focus through adjustment of the stimulation [10] Amar AP, Heck CN, Levy ML, et al. An in- parameters [36]. stitutional experience with cervical vagus nerve 282 L.L. Carpenter et al / Neurosurg Clin N Am 14 (2003) 275–282

trunk stimulation for medically refractory epilepsy: patient series. J Clin Neurophysiol 2001;18: rationale, technique and outcome. Neurosurgery 408–14. 1998;43:1265–80. [24] Harden CL, Pulver MC, Nikolov B, et al. Effect of [11] Ben-Menacham E, Manon-Espaillat R, Ristanovic vagus nerve stimulation on mood in adult epilepsy R, et al. Vagus nerve stimulation for treatment of patients [abstract]. Neurology 1999;52(Suppl):A238– partial seizures, I: a controlled study of effect on P03122. seizures. Epilepsia 1994;35:616–26. [25] Ballenger JC. Post RM. Carbamazepine (Tegretol) [12] Handforth A, DeGiorgio CM, Schachter SC, et al. in manic-depressive illness: a new treatment. Am J Vagus nerve stimulation for treatment of partial- Psychiatry 1980;137:782–90. onset seizures: a randomized active-control trial. [26] Rush AJ, George MS, Sackeim HA, et al. Vagus Neurology 1998;51:48–55. nerve stimulation (VNS) for treatment-resistant de- [13] Morris GL, Mueller WM. E01–E05 VNSSG: long- pression: a multicenter study. Biol Psychiatry 2000; term treatment with vagus nerve stimulation in pa- 47:276–86. tients with refractory epilepsy. Neurology 1999;53: [27] Sackeim HA, Rush AJ, George MS, et al. Vagus 1731–5. nerve stimulation (VNS) for treatment-resistant [14] Salinsky MC, Uthman BM, Ristanovic RK, et al. depression: efficacy, side effects, and predictors Vagus nerve stimulation for the treatment of of outcome. Neuropsychopharmacology 2001;25: medically intractable seizures: results of a 1-year 713–28. open-extension trial. Arch Neurol 1996;53:1176–80. [28] Marangell LB, Rush AJ, George MS, et al. Vagus [15] Fisher RS, Handforth A. Reassessment: vagus nerve stimulation (VNS) for major depressive epi- nerve stimulation for epilepsy: a report of the sodes: One year outcomes. Biol Psychiatry 2002;51: Therapeutics and Technology Assessment Subcom- 280–7. mittee for the American Academy of Neurology. [29] Martinez JM, George MS, Rush AJ, et al. Vagus Neurology 1999;53:666–9. nerve stimulation shows benefits in treatment- [16] Schachter SC, Saper CB. Vagus nerve stimulation resistant depression for up to two years [abstract (progress in epilepsy research). Epilepsia 1998;39: NR21]. In: APA 2002 Annual Meeting New 677–86. Research Program. American Psychiatric Associa- [17] Bruce DA. Surgical complications. Epilepsia 1998; tion. Washington DC:2002. p. 6. 39(Suppl 6):92–3. [30] Marangell LB. Does vagus nerve stimulation change [18] Naritoku DK, Terry WJ, Helfert RH. Regional the long term course of mood disorders [abstract induction of fos immunoreactivity in the brain by 16E]? In: APA 2002 Annual Meeting Syllabus and anticonvulsant stimulation of the vagus nerve. Proceedings Summary. American Psychiatric Asso- Epilepsy Res 1995;22:53–62. ciation. Washington DC:2002. p. 263. [19] Krahl SE, Clark KB, Smith DC, et al. Locus [31] Devous MD, Husain M, Harris TS, et al. The coeruleus lesions suppress the seizure attenuating effects of VNS on regional cerebral blood flow in effects of vagus nerve stimulation. Epilepsia 1998; depressed subjects. Biol Psychiatry 2002;51(8):152S. 39:709–14. [32] Bohning DE, Lomarev MP, Denslow S, et al. [20] Walker BR, Easton A, Gale K. Regulation of Feasibility of vagus nerve stimulation-synchronized limbic motor seizures by GABA and glutamate blood oxygenation level-dependent functional MRI. transmission in nucleus tractus solitarius. Epilepsia Invest Radiol 2001;36:470–9. 1999;40:1051–7. [33] Nahas Z, Lomarev MP, Denslow S, et al. Synchro- [21] Ben-Menacham E, Hamberger A, Hedner T, et al. nized BOLD fMRI [abstract]. Biological Psychiatry Effects of vagus nerve stimulation on amino acids 2002;51:(8)152S. and other metabolites in the CSF of patients with [34] Armitage R, Husain R, Hoffman R, et al. The partial seizures. Epilepsy Res 1995;20:221–7. effects of VNS on sleep in depression [abstract]. [22] Henry TR, Bakay RAE, Votaw JR, et al. Brain Biological Psychiatry 2002;51(8)152S. blood flow alterations induced by therapeutic vagus [35] Carpenter LL, Moreno F, Kling M. The effects of nerve stimulation in partial epilepsy: acute effects VNS on CSF components in depressed subjects at high and low levels of stimulation. Epilepsia [abstract]. Biological Psychiatry 2002;51(8)152–3S. 1998;39:983–90. [36] George MS, Sackeim HA, Rush AJ, et al. Vagus [23] Scherrmann J, Hoppe C, Kral T, et al. Vagus nerve stimulation: a new tool for brain research and nerve stimulation: clinical experience in a large therapy. Biol Psychiatry 2000;47:287–95. Neurosurg Clin N Am 14 (2003) 283–301

Transcranial magnetic stimulation Mark S. George, MDa–e,*, Ziad Nahas, MDa,d,e, Sarah H. Lisanby, MDf, Thomas Schlaepfer, MDg,h, F. Andrew Kozela,d,e, Benjamin D. Greenberg, MD, PhDi aDepartment of Psychiatry, 502 North, Institute of Psychiatry, Medical University of South Carolina, 67 President Street, Charleston, SC, USA bDepartment of Radiology, Medical University of South Carolina, Charleston, SC, USA cDepartment of Neurology, Medical University of South Carolina, Charleston, SC, USA dBrain Stimulation Laboratory, Medical University of South Carolina, Charleston, SC, USA eCenter for Advanced Imaging Research, Medical University of South Carolina, Charleston, SC, USA fDepartment of Psychiatry, Columbia University, 1051 Riverside Drive, Unit 126, New York State Psychiatric Institute, New York, NY, USA gDepartment of Psychiatry and Mental Hygiene, University of Berne, University Hospital, 3010 Berne, Switzerland hJohns Hopkins University, Department of Psychiatry, Baltimore, MD, USA iDepartment of Psychiatry and Human Behavior, Brown University, 345 Blackstone Boulevard, Providence, RI 02906, USA

The ability to stimulate the brain in awake standing of the brain [1,2]. Their results were con- alert adults without the need for anesthesia or fined to epilepsy patients, however, and the testing open craniotomy is a significant advance that has was done during open craniotomy. long been a dream of clinicians and neuroscien- Transcranial magnetic stimulation (TMS) is tists. The pioneering work of the neurosurgeons the answer to this dream of noninvasive stimu- Penfield and Perot demonstrated the behavioral lation in awake individuals. TMS produces non- results of direct electric stimulation of epilepsy pa- invasive direct cortical brain stimulation by tients during open craniotomy. This work demon- creating a powerful transient magnetic field [3,4], strated how powerful and important direct brain which then induces electric currents in the brain. electric stimulation might be for advancing under- Because of its noninvasiveness, TMS has considerable promise as a research tool to understand brain- * Corresponding author. behavior relations. Although TMS is not limited to E-mail address: [email protected] (M.S. George). use by neurosurgeons and does not involve surgery, Drs George, Nahas, and Kozel’s work with tran- it is important for neurosurgeons to be familiar scranial magnetic stimulation (TMS) has been sup- with this most powerful new tool for causing ported in part by research grants from National ‘‘electrodeless electric stimulation.’’ Advances Alliance for Research on Schizophrenia and Depression made with TMS promise to set the stage for other (NARSAD), the Stanley Foundation, the Borderline brain stimulation advances in neurosurgery. More- Personality Disorders Research Foundation (BPDRF), over, information gleaned from TMS studies will NINDS grant RO1-AG40956, and the Defense likely inform us about the mechanisms of action of Advanced Projects Agency (DARPA). classic neurosurgical techniques like deep brain Drs George, Nahas, and Kozel hold several TMS- related patents either alone or in combination. These stimulation (DBS). are not in the area of TMS therapeutics but are for Despite TMS’s promise and widespread use, new TMS machine designs as well as combining TMS there is still inadequate understanding of the exact with MRI. mechanisms by which TMS affects the brain and

1042-3680/03/$ - see front matter Ó 2003, Elsevier Science (USA). All rights reserved. doi:10.1016/S1042-3680(02)00120-1 284 M.S. George et al / Neurosurg Clin N Am 14 (2003) 283–301 how these effects change as a function of dif- important to realize that TMS, which produces ferent TMS use parameters, such as intensity powerful but brief magnetic fields that, in turn, and frequency. Progress in this area is occurring induce electric currents in the brain, radically by performing TMS in animal models and by differs from the currently popular use of low-level combining TMS with functional neuroimaging. static magnetic fields as alternative therapies. There has been much interest in using repeated Constant exposure to static magnetic fields can daily administration of TMS as a potential treat- have biologic effects [6]. TMS does not produce ment in a variety of neuropsychiatric disorders. magnetic fields for long (microseconds), however, The largest body of work has been done in the and they are relatively weak, except directly under treatment of major depression. Although several the TMS coil. It is thus assumed that TMS pro- meta-analyses of these antidepressant studies show duces its behavioral effects solely through the that TMS has statistically significant effects greater production of electric currents in the cortex of the than those of placebo, larger sample studies are brain. This assumption has not been proved, still needed to clarify its clinical role. Additionally, however. If TMS pulses are delivered repetitively TMS as a treatment for depression has not yet and rhythmically, the process is called repetitive been approved by the US Food and Drug Admin- TMS (rTMS) [7]. rTMS can be modified by the istration (FDA), although pivotal trials geared for term fast if the frequency is greater than 1 Hz [8]. eventual FDA approval are currently underway. Fast rTMS is currently limited to brief runs of This article succinctly summarizes basic TMS approximately 25 to 30 Hz. Stimulation frequen- physics and mechanisms and then critically re- cies faster than this have an increased seizure risk, views the potential clinical uses of TMS, particu- and most modern capacitors cannot keep de- larly in the treatment of depression and other livering the needed energy before depleting. Thus, neuropsychiatric conditions. fast TMS is performed at frequencies that would be considered slow for DBS, where stimulation frequencies sometimes exceed 150 Hz. Description of transcranial magnetic stimulation The magnetic field induced by TMS declines and mechanisms of action rapidly with distance away from the coil. Thus, TMS uses a powerful handheld magnet to with current technology, TMS coils are directly and create a time-varying magnetic field, where electrically only able to stimulate the superficial a localized pulsed magnetic field over the surface cortex and are not able to produce direct electric of the head depolarizes underlying superficial stimulation deep in the brain [9]. Although this neurons [3–5]. High-intensity current is rapidly limited depth of penetration is a limitation of the turned on and off in the electromagnetic coil present technology, deeper brain structures can be through the discharge of capacitors (Fig. 1). It is influenced by cortical TMS because of the cortex’s

Fig. 1. Diagram of events leading to transcranial magnetic stimulation excitation. (From George MS, Belmaker RH. Transcranial magnetic stimulation in neuropsychiatry. Washington (DC): American Psychiatric Press; 2000; with permission). M.S. George et al / Neurosurg Clin N Am 14 (2003) 283–301 285 massive interconnections and redundant cortical- effects. For example, repeated stimulation of a subcortical loops [10]. Moreover, there are several single neuron at low frequency in culture produces groups working on novel TMS coil designs that long-lasting inhibition of cell-cell communications might be able to reach deeper into the brain without (called long-term depression [LTD]) [21–24]. overwhelming superficial cortical structures. Conversely, repeated high-frequency stimulation The amount of electricity needed to cause can improve cell-cell communication (called long- changes in the cortex varies from person to person term potentiation [LTP]) [25]. With these cellular and also from one brain region to the next [11]. data on LTD and LTP as a background, there has One commonly used method for standardizing been much interest in whether TMS, exciting and adjusting the amount of electricity delivered hundreds or thousands of neurons in a pulse, can and induced by TMS across different individuals produce sustained inhibitory or excitatory effects is to determine each person’s motor threshold [26,27]. Several studies have now shown that (MT). The MT is commonly defined as the min- chronic low-frequency stimulation of the motor imum amount of electricity needed to produce cortex can produce inhibitory intermediate-term movement in the contralateral thumb when the effects (lasting for several minutes) after stimula- coil is placed optimally over the primary motor tion [28,29]. There is also some evidence that cortex. MT can be determined either by using high-frequency stimulation can produce interme- electromyographic (EMG) recordings [12] or, with diate-term excitatory effects [30]. One of the most less precision, by using visible movement [13]. easily demonstrated immediate effects of TMS is TMS has been shown to produce immediate speech arrest, where high-frequency TMS placed effects (within seconds), such as the movement precisely over the Broca’s area can immediately of the thumb or direct inhibition of another TMS and transiently block fluent speech [31]. TMS used pulse followed shortly in time. These immediate in studies like this can produce what are sometimes effects are thought to result from direct excitation referred to as ‘‘virtual lesions.’’ Importantly, none of inhibitory or excitatory neurons. There is some of these temporary lesion effects were demon- evidence to suggest that TMS at different intensi- strated to persist beyond the time of active TMS ties, frequencies, and coil angles excites different administration. Thus, the lesions are truly virtual elements (eg, cell bodies, axons) of different neuro- and temporary. This contrasts with observations nal groups (eg, interneurons, neurons projecting that a train of repeated high-frequency stimulation to other parts of the cortex, U fibers) [8,14–17]. can excite the brain so much that it results in This is further complicated by the complex six- seizure activity [32]. With knowledge of the ap- layer arrangement of human neocortex along with propriate limits of stimulation, seizures can be the varying gyral folds, which places some aspects avoided safely. of the brain close to the surface and others far away in sulcal folds. Safety issues Another example of immediate TMS effects is called paired-pulse TMS. This technique involves Although there is minimal risk of a seizure when delivering two TMS pulses to the same region TMS is performed within the published safety with varying interpulse intervals (usually milli- guidelines, the most well-known critical safety seconds long) and intensities [12]. Depending on concern with TMS is inadvertently causing a seizure the relative strength of the first pulse to the second [7]. It is important to realize that this TMS safety and the interpulse interval, the first pulse can either table was developed in a small subject sample inhibit or enhance the second pulse. Paired-pulse using a surrogate end point for a seizure— TMS over the motor area can be used to assess spread of TMS-induced motor-evoked potentials natural brain inhibitory and excitatory systems at (MEPs) beyond the target area of stimulation. rest in individuals with different disorders [18,19] Thus, the safety table exists only for stimulation of as well as after the administration of different motor cortex and cannot readily be applied to using centrally active compounds or other treatments TMS over other brain regions. Finally, although [20]. The different TMS use parameters (eg, fre- the intensity and frequency of stimulation were quency, intensity, length of stimulation, intertrain examined, the intertrain interval was not. One of interval) are most certainly all biologically relevant the inadvertent seizures was induced with stimula- and likely important. Particular attention has been tion trains that were within the safety guidelines but focused on whether and to what degree different that were administered with an excessively short frequencies of TMS might have divergent biologic intertrain interval [33]. A general rule of thumb is 286 M.S. George et al / Neurosurg Clin N Am 14 (2003) 283–301 that one should have an intertrain interval at least Interestingly, in light of the initial pioneering as long as the period of stimulation. A known results of Penfield, stimulation with TMS has seizure disorder, history of epilepsy, or intracranial never evoked the type of complex behavioral symp- abnormality, such as a prior stroke, can all increase toms reported with direct electric stimulation in the risk of a TMS-induced seizure [34]. Although an human beings. TMS has never been reported to inadvertent seizure is the main safety hazard provoke a memory, smell, or song as reported by associated with TMS, there have been only 12 Penfield [1,2,51]. There may be several explana- reported cases since 1985 when cranial TMS began. tions for this difference between TMS and direct It is, in fact, not easy to intentionally use TMS to cortical stimulation. First, most TMS studies are produce a seizure, even in patients with epilepsy in healthy subjects, whereas the Penfield results [35]. For example, an attempt to use TMS to were from patients with refractory epilepsy produce a seizure intentionally in a patient with requiring surgery. Complex behaviors have not a focal epilepsy was not successful [35]. In addition, been provoked by TMS, even when it has been in a study exploring rTMS as a method to induce used in epilepsy patients [35,52]. This area has therapeutic seizures, stimulation parameters far not been systematically explored, however. above the published safety thresholds had to be Second, Penfield and Perot used high-intensity used to reliably induce seizures [36]. electric stimulation that even caused seizures in A muscle tension type headache and dis- some instances. In other words, the initial comfort at the site of stimulation are less serious electric stimulation likely spread to other regions but relatively common side effects of TMS. In through pathologically kindled neuronal circuits. contrast to electroconvulsive therapy (ECT), no Stimulation of these circuits led to the expres- deleterious cognitive effects of 2 weeks of slow or sion of auras. Thus, most importantly, all the fast rTMS have been found [37,38]. This is not complex behaviors reported by Penfield and surprising, because ECT induces a generalized Perot were actually auras commonly experienced seizure, whereas rTMS is being used at subcon- by the patients. This fact is underappreciated vulsive levels. Like MRI, TMS could cause the but was recently confirmed by Perot (P. Perot, movement of paramagnetic objects in or around personal communication, 2002). Thus, it is likely the head. For this reason, subjects with para- that TMS does not produce complex behaviors magnetic metal objects in the head or eye are like historic accounts of direct cortical electric generally excluded from TMS studies. TMS can stimulation because TMS has not been delivered cause heating of metallic implants and the in- in intensities high enough to cause spreading of activation of a pacemaker, medication pumps, or a seizure discharge and elicitation of auras. cochlear devices. In the United States, rTMS is an experimental procedure that requires an investi- Use of transcranial magnetic stimulation gational device exemption (IDE) from the FDA and speech arrest to determine eloquent cortex for research. It should be kept in mind that modern TMS did not begin until 1985 [39] and TMS delivered over the motor speech area can that the total number of subjects or patients to produce transient speech arrest. There was initial receive TMS is likely still less than 10,000. interest in whether TMS might be used to determine Nevertheless, substantial experience to date sug- eloquent cortex. This information could be use- gests that at least in the short term (<10 years), ful in patients about to undergo surgical resection TMS at moderate intensity has no other evident of tumors or seizure foci. Initial reports found lasting adverse effects in adults. that TMS needed to be delivered at frequencies around 20 Hz, which was painful and might induce a seizure in a patient with a focal lesion or Overview of research uses relevant epilepsy [52–54]. A later study showed that TMS to neurosurgery could be performed at lower frequencies, some- As a research tool, TMS has been used to times as low as 4 Hz [31]. There has not been a influence many brain functions, including move- formal comparison of TMS speech arrest with the ment [40], visual perception [41,42], memory Wada test or functional MRI (fMRI) presurgical [43,44], attention, speech [31,45], and mood [46– mapping. Some studies have found that the TMS- 48]. A full review is beyond the scope of this article indicated region for motor cortex correlates well [49,50]. We discuss below the TMS uses germane with fMRI-predicted motor cortex and later di- to neurosurgery. rect neurosurgical stimulation [55]. Further work M.S. George et al / Neurosurg Clin N Am 14 (2003) 283–301 287 using TMS to help with presurgical planning for effects of different TMS use parameters on brain motor or speech areas is warranted [56]. function. Because it seems that TMS at different frequencies has divergent effects on brain activity, Use parameter effects in animal models combining TMS with functional brain imaging will better delineate not only the behavioral neuro- Although TMS has been used quite effectively psychology of various psychiatric syndromes but as a tool to investigate normal and pathologic some of the pathophysiologic circuits in the brain. brain function, its full and proper uses as a research In contrast to imaging studies with ECT, which tool and clinical treatment are still hampered by have found that ECT shuts off global and regional incomplete knowledge of the neurobiologic cas- activity [75], most studies using serial scans in cade of events triggered by TMS at different depressed patients undergoing TMS have found settings. Numerous animal studies have been im- increased activity in the cingulate and other limbic portant in trying to bridge this knowledge gap regions [76–78]. However, two studies have now and improve our understanding of the modes of found divergent effects of TMS on regional activity action of TMS. in depressed patients, determined both by the TMS studies with intracranial electrodes in frequency of stimulation and the baseline state of rhesus monkeys have provided information about the patient [79,80]. In other words, for patients with the nature and spatial extent of the rTMS-induced global or focal hypometabolism, high-frequency electric field [57,58]. Corticospinal tract develop- prefrontal stimulation has been found to increase ment, aspects of motor control, and medication brain activity over time, with the opposite happen- effects on corticospinal excitability have been ing as well. Conversely, patients with focal hyper- studied fairly extensively in nonhuman primates activity have been shown to have reduced activity using single-pulse TMS [59–67]. Such work has over time after chronic daily low-frequency stimu- yielded information about TMS neurophysiologic lation. These two small sample studies have numer- effects, such as the observation that TMS-evoked ous flaws, however. They simultaneously show the motor responses result from direct excitation of potential and the complexity surrounding the issue corticospinal neurons at or close to the axon of how to use TMS to change activity in defined hillock [67]. circuits. They also point out an obvious difference Rodent rTMS studies have reported antide- from ECT, where the net effect of the ECT is to pressant-like behavioral and neurochemical ef- decrease prefrontal and global activity [75]. fects. In particular, rTMS enhances apomorphine- Over the past year, several studies combining induced stereotypy and reduces immobility in the TMS with other neurophysiologic and neuroimag- Porsolt swim test [68,69]. rTMS has been reported ing techniques have helped to elucidate how TMS to induce electroconvulsive shock (ECS)–like achieves its effects. The Medical University of changes in rodent brain monoamines, b-adrener- South Carolina (MUSC) group has pioneered and gic receptor binding, and immediate early gene perfected the technique of interleaving TMS with induction [69–71]. The effects of rTMS on seizure blood oxygen level–dependent (BOLD) fMRI, threshold are variable and may depend on the allowing for direct imaging of TMS effects with parameters and chronicity of stimulation [26,72]. high spatial (1–2 mm) and temporal (2–3 seconds) Within the past year, Pope and Keck [73] have resolution [81–83] (Fig. 2). Another group in completed a series of studies using more focal Germany has now succeeded in interleaving TMS TMS in rat models. They have largely replicated and fMRI in this manner, replicating some of the earlier TMS animal studies using less focal coils. earlier MUSC work [84]. Work with this technol- Even with the attempt at focal rat stimulation, the ogy has shown that prefrontal TMS at 80% MT effects involve an entire hemisphere and cannot produces much less local and remote blood flow readily be extrapolated to what is happening in changes than does 120% MT TMS [85]. Strafella et human TMS using focal coils [74]. al [86] used positron emission tomography (PET) to show that prefrontal cortex TMS causes dopamine Combining transcranial magnetic stimulation release in the caudate nucleus, and Paus et al [87] with functional imaging used the same modality to show that prefrontal A critically important area that will ultimately cortex TMS has reciprocal activity with the anterior guide clinical parameters is to combine TMS with cingulate gyrus. Our group at the MUSC [76,77,88] functional imaging to monitor TMS effects on the as well as investigators in Scotland (K. Ebmeier, brain directly and to thus understand the varying personal communication 2002) and Australia [80] 288 M.S. George et al / Neurosurg Clin N Am 14 (2003) 283–301

Fig. 2. Image of blood oxygen level–dependent (BOLD) functional MRI (fMRI) data from healthy adults, where transcranial magnetic stimulation (TMS) has been applied at 120% motor threshold over the prefrontal cortex. Note that although TMS initially affects only superficial cortex, cortical-subcortical connections cause changes in deeper limbic regions (8 individuals, P < 0.001, extent P < 0.05 for display on Talairach normalized MRI template). (From Nahas Z, Lomarev M, Roberts DR, et al. Unilateral left prefrontal transcranial magnetic stimulation (TMS) produces intensity-dependent bilateral effects as measured by interleaved BOLD fMRI. Biol Psychiatry 2001;50:712–720; with permission). have all shown that lateral prefrontal TMS can lamotrigine, their TMS MT was elevated by 15% cause changes in the anterior cingulate gyrus and compared with the day they received placebo. other limbic regions in depressed patients. This demonstrates that lamotrigine has an inhib- It is thus clear that TMS delivered over the itory effect on motor cortex. TMS/fMRI images prefrontal cortex has immediate effects in impor- showed that lamotrigine markedly reduced the tant subcortical limbic regions. The initial TMS secondary effects of TMS in other cortical regions. effect on cortex and the secondary synaptic In direct contrast, however, lamotrigine increased changes in other regions likely differ as a function the TMS-induced effects in limbic regions. This of mood state, cortical excitability, and other proof of concept study highlights several key factors that would change resting brain activity. points in understanding TMS brain effects. First, The MUSC group thus wondered whether these it is abundantly clear that prefrontal TMS has TMS-induced limbic effects might be modified by direct modulating effects on limbic structures. medications that are known to treat or stabilize Second, these TMS effects are easily seen with the mood. Would a medication that inhibits cortical interleaved TMS/fMRI and other imaging techni- spreading of electric stimulation change the TMS- ques. Most importantly, factors like medications induced limbic effects? To answer this, we recently and probably diseases like depression modify the designed and completed a study in 12 healthy TMS-induced limbic effects. young men, where we measured the TMS MT as The interleaved TMS/fMRI technique shows well as performed TMS/fMRI on 2 separate days promise as a useful new method for understanding [89]. On the first day, subjects received one single the regional brain effects of CNS active com- oral dose of 325 mg of lamotrigine or placebo. pounds, particularly those with inhibitory mech- On the other day 1 week later, in a randomized anisms. Further studies are needed to determine fashion, they received whatever they did not if this technique is useful in new compound de- receive during the first session. Lamotrigine is velopment or in monitoring or predicting clinical a use-dependent sodium channel inhibitor with efficacy. broad-spectrum anticonvulsant, antidepressant, and mood-stabilizing effects. As with many central nervous system (CNS) active compounds Uses of transcranial magnetic stimulation in which much is known about the pharmacology, as therapy little is known about how lamotrigine works Depression within brain circuits to achieve its clinical effects, particularly its mood-stabilizing and antidepres- Although the functional anatomy of mood sant effects. On the day that subjects received regulation is not nearly as well understood as the M.S. George et al / Neurosurg Clin N Am 14 (2003) 283–301 289 circuitry of the visual or motor systems, most An initial study from Spain from a group not scientists agree that certain brain regions are involving a psychiatrist and without prior treat- consistently affected in depression and, to a lesser ment trial experience in depression found pro- extent, mania. Although there is controversy and found antidepressant effects in psychotically much more work is needed, certain regions have depressed patients with only 1 week of left consistently been implicated in the pathogenesis prefrontal treatment [106]. The design of this trial of depression and mood regulation [90–99]. was unorthodox for studying depression, involv- These include the medial and dorsolateral prefron- ing 1 week of therapy per month repeated over 6 tal cortex, the cingulate gyrus, and other regions months. Nevertheless, it generated much interest commonly referred to as limbic (ie, amygdala, in the field. The findings of rapid response after hippocampus, parahippocampus, septum, hypo- only 1 week have not been replicated despite thalamus, limbic thalamus, insula) and paralimbic numerous attempts, and many early TMS trials (ie, anterior temporal pole, orbitofrontal cortex). were designed using this study to determine the The notion of using something like TMS as an sample size and treatment algorithms, leading to antidepressant dates back at least to the turn of small underpowered studies using an inadequate the century, when a patent was filed in Vienna in stimulation intensity (80% of MT). Since then, 1902 [100]. In more modern times, there were two most of approximately 20 studies have found open studies in Europe in the early 1990s using modest antidepressant effects that take several nonfocal round coils centered over the vertex to weeks to build. Not all TMS antidepressant deliver TMS to broad frontal and parietal regions treatment studies have been positive, however in an attempt to treat depression [101,102]. [107]. Reasoning that prefrontal and limbic regions were more important for mood regulation than Meta-analyses of transcranial the brain regions near the vertex and that theories magnetic stimulation antidepressant effect of ECT action emphasize the role of prefrontal cortex effects [103], one of us (M.S.G.) performed There have now been five independent meta- the first open trial of prefrontal TMS as an anti- analyses of the published or public TMS antidepressant in 1995 [104], followed immediately depressant literature (Table 1). Each of these by a cross-over double-blind study [105]. The meta-analyses has used different methods of se- theory behind this work was that chronic, lecting studies as well as different methods of per- frequent, subconvulsive stimulation of the pre- forming the statistical analysis of the literature. frontal cortex over several weeks might initiate Their different approaches are summarized in a therapeutic cascade of events in the prefrontal Table 1. Their results are the same: daily prefrontal cortex as well as in connected limbic regions. TMS delivered over several weeks has antidepres- Thus, beginning with these prefrontal studies, sant effects greater than sham treatment. An initial modern TMS was specifically designed as a focal, meta-analysis by McNamara et al [108] of 5 sham- nonconvulsive, circuit-based approach to therapy. controlled studies found that TMS was statisti- TMS was conceived of and launched to serve as cally significantly superior to sham TMS and that a bridge from functional neuroimaging advances this difference was robust. In a different meta- in circuit knowledge to the bedside as a focal non- analysis, Burt et al [109] examined 23 published invasive treatment. comparisons for controlled TMS prefrontal anti- Since the initial studies, there has been depressant trials and found that TMS had a com- continued high interest in TMS as an antidepres- bined effect size of 0.67, indicating a moderate sant. Multiple trials have been conducted by to large antidepressant effect. A subanalysis was researchers around the world. In general, there is done on those studies directly comparing TMS not a large industry sponsoring or promoting with ECT. The effect size for TMS in these studies TMS as an antidepressant (or therapy for other was greater than in the studies comparing TMS disorders), and the funding for these trials has with sham, perhaps reflecting subject selection largely come from foundations and governments. bias. The authors suggested that perhaps TMS The samples sizes in these antidepressant trials are works best in patients who are also clinical thus small (in all, less than 100 subjects per trial) candidates for ECT. Holtzheimer et al [110] compared with industry-sponsored pharmaceutic analyzed both published and some unpublished trials of antidepressants. A thorough review of all sham-controlled studies and concluded that, over- of these trials is beyond the scope of this article. all, prefrontal TMS was superior to sham and that 290 M.S. George et al / Neurosurg Clin N Am 14 (2003) 283–301

Table 1 Meta-analysis of transcranial magnetic stimulation antidepressant literature No. No. Study Selection method studies subjects TMS effect size Remarks McNamara et al, R and L PFCX, DB 5 81 Active 43% greater 2001 than sham Holtzheimer et al, R and L PFCX, 12 210 Overall, Cohen’s 2001 DB, published D=0.81 and unpublished Left, Cohen’s D=0.9, right=0.7 Burt et al, 2001 Open published and 9 248 Cohen’s D=1.37 unpublished DB, published and 16 432 Cohen’s D=0.67 unpublished Compare with ECT, 3 112 Cohen’s D=0.21 published favoring ECT Kozel and George, Left PFCX, DB, data 12 230 Hedge’s D=0.53 Funnel plot shows 2002 available publication bias, 25–50 negative studies needed to change results Martin et al, 2002 DB, published and 14 overall unpublished Left PFCX 9 197 2 weeks, TMS WMD- Does not allow for 0.35 greater than sham adjustment for difference Right PFCX 1 2 weeks, TMS WMD- 6.0 greater than sham Compare ECT 1 40 2 weeks, ECT WMD 1.7 greater than TMS Abbreviations: TMS, transcranial magnetic stimulation; DB, double-blind; PFCX, prefrontal cortex; R, right; L, left; ECT, electroconvulsive therapy; WMD, weighted mean difference. left-sided treatment had a nonsignificantly greater by chance. These authors then used techniques to effect size. Yet another meta-analysis was recently determine how large this publication bias would conducted by Kozel and George [111]. This have to be to change the results of the meta- analysis was confined to published double-blind analysis. The fail-safe results indicated that there studies with individual data using TMS over the would have to be 56 nonsignificant unpublished left prefrontal cortex. The summary analysis using studies of approximately the same average sample all 10 studies that met criteria revealed a cumulative size as the published studies to change the effect size of 0.53 (Cohen’s D) (range: 0.31–0.97), cumulative meta-analysis effect to a nonsignificant with the total number of subjects studied being 230. result (56 studies with Rosenthal’s method, 22 The authors used a funnel plot technique to assess studies with Orwin’s method). The most critical whether there was a publication bias in the meta-analysis of the TMS antidepressant field was literature to date and whether this bias might recently conducted using the guidelines put forth in affect the results of the meta-analysis. This the Cochrane Library [112]. This study cannot be technique assumes that with small sample studies, compared directly with the other meta-analyses there is a large chance of both erroneous positive because it looks at the field from a completely and negative results. As the sample size of studies different angle. Cochrane reviews typically try to increases, the effect sizes should begin to converge, establish the clinical value of a given therapy using resembling a funnel. The funnel plot (Fig. 3) stringent guidelines. The method is not well suited indicates that a publication bias is likely and that to look at small effects, which are scientifically there are more positive small sample studies in the important but are of lesser value in establishing TMS antidepressant literature than should occur clinical significance. In addition, Cochrane study M.S. George et al / Neurosurg Clin N Am 14 (2003) 283–301 291

Fig. 3. Summary of transcranial magnetic stimulation (TMS) sham-controlled antidepressant studies, with effect sizes (A) and funnel plot (B). Note that there are insufficient small sample negative studies (shaded oval in B), which would be predicted by chance. This indicates there is likely a publication bias in the current TMS antidepressant literature. groups try to include unpublished negative data to cussed previously concur on an effect size of correct, at least in part, for publication bias. Cohen’s D of 0.65, which is a moderate effect in Moreover, the Cochrane method does not allow the same range as the effects of antidepressant for within-study adjustments for between-group medications. For example, small to medium effect differences in illness baseline and instead uses raw sizes (0.31–0.40) are common in randomized con- illness variables. Thus, unequal initial distribution trolled trials of novel antidepressants [113,114]. of illness severity, especially in small studies, can Thus, with respect to whether or not TMS has introduce significant bias into the results and clinical significance, an important clinical issue is conclusions. Even this stringent meta-analysis in- whether TMS would be clinically effective in cluded 14 trials suitable for analysis, however [112]. patients referred for ECT. This question has been Importantly, these investigators found ‘‘No differ- addressed in a series of studies in which ECT ence...between rTMS and sham TMS using the referrals were randomized to receive either ECT or Beck Depression Inventory or the Hamilton De- rTMS. In an initial study, Grunhaus et al [115] pression Rating Scale, except for one time period compared 40 patients who presented for ECT (after 2 weeks of treatment) for left dorsolateral treatment and were randomized to receive either prefrontal cortex and high frequency; and also for ECT or TMS. ECT was superior to TMS in patients right dorsolateral prefrontal cortex and low fre- with psychotic depression, but the two treatments quency, both in favor of rTMS and both using the were not statistically different in patients without Hamilton scale’’ [112]. In summary, all five meta- psychotic depression. This same group recently analyses of the TMS published literature concur replicated this finding in a larger and independent that repeated daily prefrontal TMS for 2 weeks has cohort with an improved design (L. Grunhaus, antidepressant effects greater than those of sham. personal communication, 2001). Recently, Janicak Although there is general consensus that TMS and colleagues [116] reported a similar small series has statistically significant antidepressant effects, finding near equivalence between TMS and ECT. a more important question is whether these effects The major differences between these studies and the are clinically significant. The meta-analyses dis- rest of the controlled studies of TMS efficacy are the 292 M.S. George et al / Neurosurg Clin N Am 14 (2003) 283–301 patient selection (suitable for ECT), length of Maintenance transcranial magnetic stimulation treatment (3–4 weeks), lack of blinding, and lack to prevent depression relapse after recovery of a sham control. Unfortunately, none of the Because of its noninvasiveness and positive studies explicitly measured differences in cognitive safety and cognitive profile, TMS is potentially side effects, although, presumably, TMS has no attractive as a maintenance treatment. At MUSC measurable cognitive side effects, whereas ECT has recently, seven treatment-resistant bipolar de- several. In a similar but slightly modified design, pressed patients who had responded to an acute Pridmore [117] recently reported on a study com- TMS trial were offered admission into a 1-year paring the antidepressant effects of standard ECT maintenance therapy of weekly TMS [124]. TMS (three times per week) and ECT given one time per was performed 1 day per week over the left week followed by TMS on the other 4 weekdays. At prefrontal cortex at 110% MT and 5 Hz for 8 3 weeks, he found that both regimens produced seconds for 40 trains. During this follow-up similar antidepressant effects. Unfortunately, de- period, four subjects dropped out of the mainte- tailed neuropsychologic testing was not performed, nance study and were labeled nonresponders but one would assume that the TMS and ECT (average of 25 weeks of treatment). Three subjects group had less cognitive side effects than the pure completed 1 full year of weekly TMS without ECT group. Finally, an Israeli group recently a depression relapse. These data suggest that TMS published their finding that relapse rates in the 6 might eventually be used as a maintenance tool in months after ECT or rTMS were similar [118]. For depression and that one treatment per week might both treatments to maintain maximal benefit, some be a good first attempt at a maintenance schedule. form of maintenance therapy is recommended. In Much more work is needed, however. sum, TMS clinical antidepressant effects are in the range of those of other antidepressants and persist as long as the clinical effects after ECT. Transcranial magnetic stimulation to treat mania Although the literature suggests that prefrontal Grisaru and his colleagues in Israel [125,126] TMS has an antidepressant effect greater than reported on an interesting study using either right that of sham and that the magnitude of this effect or left prefrontal TMS in bipolar affective is at least as large as that of other antidepressants, disorder (BPAD) manic patients admitted to their many issues are not resolved. For example, it is hospital for mania. TMS was given daily in addi- unclear how best to deliver TMS. Most but not all tion to the standard treatment for mania. After 2 [119] studies have used focal coils positioned over weeks, the group receiving right-sided TMS was the left prefrontal cortex. It is still not known significantly more improved than the group that whether TMS over one hemisphere is better than had received left-sided TMS. The authors con- that over another hemisphere or whether there are cluded that TMS might be useful as an antimanic better methods for placing the coil. For the most agent. This same group has attempted to replicate part, the coil has been positioned using a rule- these findings but has not found similar results (N. based algorithm to find the prefrontal cortex, Grisaru, personal communication, 2001). Addi- which was adopted in the early studies [46,104– tionally, although subjects were assigned to the 106]. This method was shown to be imprecise in two groups at random, the left-sided group was the particular prefrontal regions stimulated di- more ill than the right-sided group on several rectly underneath the coil, depending largely on measures. Further work is needed. the subject’s head size [120]. Additionally, most studies have stimulated with the intensity needed to cause movement in the thumb (MT). There is Current state of transcranial magnetic now increasing recognition that higher intensities stimulation clinical practice for depression of stimulation might be needed to reach the prefrontal cortex, especially in elderly patients, In summary, TMS is a promising tool for where prefrontal atrophy may outpace that of treating depression acutely. It likely can also induce motor cortex, where MTs are measured [121–123]. mania or hypomania in BPAD patients or suscep- There are also emerging data that TMS therapeu- tible patients. Its antimanic properties remain to be tic effects likely take several weeks to build. explored. Although it is approved in Canada and Consequently, many of the initial trials, which Israel as a treatment, it is still considered in- lasted only 1 to 2 weeks, were likely too brief to vestigational in the United States by the FDA. generate maximum clinical antidepressant effects. Much work remains to understand the optimum M.S. George et al / Neurosurg Clin N Am 14 (2003) 283–301 293 dosing strategy for the antidepressant effect of Three other recent studies [131–133], however, as TMS. It is unlikely that the initial combinations well as a study from Japan using TMS over the of intensity, frequency, coil shape, scalp location, prefrontal cortex at low frequencies and doses number of stimuli, or dosing strategy (daily, twice [134,135] report that TMS may improve effects in daily) are the most effective for treating depression. Parkinson’s disease. Further studies are needed. It Some US and European psychiatrists are using should be remembered that only a small portion of TMS in clinical practice to treat depression under the combinations of use parameters, brain regions, their general license to practice. and dosing schedules have been tried. There are two small positive studies showing Important unanswered clinical questions that TMS can benefit writer’s cramp, a form of focal dystonia [136]. After publication of a positive small It is not clear which, if any, medications work abstract, two groups have used TMS to investigate well with TMS or interfere with its therapeutic and possibly treat Gilles de la Tourette syndrome effects. Despite these major unanswered questions, [137]. One study found modest and transient ben- since the first use of prefrontal TMS as an anti- eficial effects on tics when applied over prefrontal depressant in 1995, this tool has clearly opened cortex (M. Trimble, personal communication 2002). up new possibilities for clinical exploration and Another study at MUSC also found positive effects treatment of depression. Many parameters, such on tics and obsessive-compulsive disorder (OCD) as intensity, location, frequency, pulse width, symptoms [137]. Further work is needed in this intertrain interval, coil type, duration, numbers promising area. of sessions, interval between sessions, and time of The TMS MT is reduced in patients with day, remain to be systematically explored. Al- untreated epilepsy [138], hinting at widespread though there are suggestions of antidepressant problems in cortical excitability. Therapeutically, effects of TMS, there are questions about how it there is one report of potential beneficial effects of might be used in treatment algorithms. It will slow rTMS in action myoclonus [139]. Addition- perhaps always be easier to see a clinician occa- ally, TMS has been used to examine cortical sionally and take daily medication rather than excitability and inhibition in Tourette’s syndrome, traveling to a treatment facility for TMS on a daily dystonia, and OCD [18,19]. Reduced intracortical basis. Thus, the ultimate clinical role of TMS in inhibition has been reported in all three illnesses. treating depression may be in medication-refractory cases, those who would otherwise receive Schizophrenia ECT, or in patients who are unable to tolerate systemic therapy because of pregnancy [127] or Several studies have used TMS to investigate a medical condition. schizophrenia without consistent replication of early findings, which were compounded by medication issues [140,141]. A 1-day prefrontal TMS Other conditions challenge study by Nahas and colleagues [142] at TMS has also been investigated as a possible MUSC failed to find significant effects on negati- treatment for a variety of neuropsychiatric dis- ve symptoms. Hoffman and colleagues [143,144] orders. In general, the published literature on have used low-frequency TMS over the temporal these conditions is much less extensive than for lobes to treat hallucinations in patients with TMS as an antidepressant; therefore, conclusions schizophrenia. Although they have replicated their about the clinical significance of effects must earlier study, another group has tried to replicate remain tentative until large sample studies are this effect without success (K. Ebmeier, personal conducted. communication 2002).

Movement disorders Anxiety disorders Some initial studies found positive effects in In a randomized trial of left and right pre- Parkinson’s disease [128]; however, one of these frontal and midoccipital 20-Hz stimulation in 12 early results could not be replicated [129], and some patients with OCD, Greenberg et al [145] found of the methods described were actually not credible. that a single session of right prefrontal rTMS Moreover, a recent study found that TMS delivered decreased compulsive urges for 8 hours. Mood was over the supplementary motor area (SMA) actually also transiently improved, but there was no effect worsened Parkinson’s disease symptoms [130]. on anxiety or obsessions. Using TMS probes, the 294 M.S. George et al / Neurosurg Clin N Am 14 (2003) 283–301 same group reported decreased intracortical in- Lisanby et al [36] used an enhanced device with hibition in patients with OCD [146], which has also four times the usual number of charging modules been noted in patients with Tourette’s syndrome to induce seizures in depressed patients referred [19]. Somewhat surprisingly, OCD patients had for ECT. Further clinical and preclinical work a lowered MEP threshold in one study [147], which with this exciting technique called magnetic was unrelated to intracortical inhibition and seems seizure therapy (MST) has proceeded. An initial to replicate (E.M. Wassermann, personal commu- safety study found that MST seizures were briefer nication 2002). Only two other studies have ex- in duration than ECT seizures, that patients awoke amined possible therapeutic effects of rTMS in from anesthesia much faster with MST, and that OCD. A double-blind study using right prefrontal their acute cognitive side effects were much less slow (1 Hz) rTMS and a less focal coil failed to find with MST [153]. Further work is underway to statistically significant effects greater than those of determine whether this technique has antidepres- sham [148]. In contrast, a recent open study in sant effects. Because MST induces a seizure, it still a group of 12 OCD patients refractory to standard requires repeated episodes of general anesthesia. treatments, who were randomly assigned to right The long-held dogma in the field of convulsive or left prefrontal fast rTMS, found that clinically therapy was that a seizure was necessary and significant and sustained improvement was ob- sufficient to constitute an effective treatment for served in one third of patients [149]. Clearly, depression. Recent research in ECT and new work further work is warranted testing TMS as a poten- in TMS have challenged both aspects of that tial treatment for OCD. theory with respect to the antidepressant potential McCann et al [150] reported that 2 patients of brain stimulation. First, it was demonstrated with posttraumatic stress disorder (PTSD) im- that it was possible to generate seizures with ECT proved during open treatment with 1 Hz of rTMS that lacked efficacy; thus, a seizure cannot be over the right frontal cortex. Grisaru et al [151] sufficient [154,155]. What we have learned about similarly stimulated 10 PTSD patients over motor the features that distinguish effective from in- cortex and found decreased anxiety. Grisaru and effective seizures has guided the development of colleagues also reported a positive TMS study in MST as a more targeted way of producing an PTSD patients (N. Grisaru, personal communi- effective treatment that should be less contami- cation, 2001). Further work is needed. nated by the consequences of generalization to regions of the brain linked to cognitive side effects (medial temporal structures). Because MST offers A potential new way of using transcranial a focal means of inducing seizures, it is possible to magnetic stimulation–magnetic seizure therapy use MST to examine the roles of the location of as an antidepressant seizure onset and patterns of seizure spread in the The discussion throughout this article has antidepressant effects and cognitive side effects focused on using TMS to change brain function more precisely than one could do with ECT. without inadvertently causing a seizure. As men- In addition to not being sufficient, recent work tioned previously, TMS at high frequencies and with TMS has challenged the theory that a seizure intensities can cause seizures. ECT produces a is necessary at all [156]. This radical notion was seizure through direct electric stimulation, under not easily accepted by the field as recently as 8 anesthesia, of the scalp and skull. Although years ago, because it challenged the then widely ECT is the most effective antidepressant, it has held dogma that a seizure was needed for ECT, cognitive side effects and does not work in up to and by extension, any form of transcranial half of treatment-resistant patients. If one used stimulation to be effective in treating depression. TMS to induce an ECT-like seizure, one might be The recent suggestions of antidepressant effects of able to focus the point of origin of the seizure and vagus nerve stimulation (VNS) in treatment- thus spare some brain regions from unnecessary resistant depressed patients [157–159] as well as exposure to electric currents and to seizure spread. in comorbid epilepsy and depression patients This is possible with TMS, because magnetic fields [160,161] add weight to the notion that somatic pass through the scalp and skull unimpeded, treatments can improve mood without causing whereas the direct application of electricity to the a seizure (VNS does not cause seizures and is, in scalp with ECT loses focality and power as a result fact, an anticonvulsant treatment method). Addi- of the impedance of the overlying tissue. After a tionally, the recent reports of mood effects with proof of concept demonstration in primates [152], DBS also underline the point that nonconvulsive M.S. George et al / Neurosurg Clin N Am 14 (2003) 283–301 295 stimulation can powerfully change mood [162]. Acknowledgment The recent studies with MST, although perhaps a radical improvement over the current methods The authors dedicate this article to Marty of inducing seizures, still work within the para- Szuba, MD, from the University of Pennsylvania. digm that a seizure is needed for antidepressant effects. If MST has clinically significant antide- References pressant effects that are not associated with side effects, one could imagine eventually doing a direct [1] Penfield W, Perot P. The brain’s record of auditory and visual experience: a final summary and dis- comparison of MST versus TMS over the same cussion. Brain 1963;86:595. region and finally settling the argument over the [2] Penfield W, Jasper H. Epilepsy and the functional necessity of the seizure to treat depression. While anatomy of the human brain. Boston: Little, Brown debating the theory can be useful in refining our and Company; 1954. perspective on the modes of action of antidepres- [3] George MS, Belmaker RH. Transcranial magnetic sant treatment, it is important to recognize that stimulation in neuropsychiatry. Washington (DC): these various brain stimulation techniques for American Psychiatric Press; 2000. depression are not in competition; rather, each [4] George MS, Lisanby SH, Sackeim HA. Trans- may be helpful for specific subpopulations of cranial magnetic stimulation: applications in neu- depressed patients. Demonstrating that a subcon- ropsychiatry. Arch Gen Psychiatry 1999;56:300. [5] Barker AT, Freeston IL, Jarratt JA, Jalinous vulsive form of stimulation can be effective in R. Magnetic stimulation of the human nervous depression does not necessarily make convulsive system: an introduction and basic principles. In: forms of treatment obsolete. The goal has been to Chokroverty S, editor. Magnetic stimulation in expand our therapeutic options for severely ill and clinical neurophysiology. Boston: Butterworth; resistant patients and, in the process, to illuminate 1989. p. 55. the brain circuitry and common mechanisms [6] National Research Council. Possible health effects underlying antidepressant action. of exposure to residential electric and magnetic fields. Washington (DC): National Academy Press; 1996. [7] Wassermann EM. Risk and safety of repetitive Summary transcranial magnetic stimulation: report and suggested guidelines from the International Workshop TMS is a powerful new tool with extremely on the Safety of Repetitive Transcranial Magnetic interesting research and therapeutic potentials. Stimulation, June 5–7, 1996. Electroencephalogr Further understanding of the ways by which TMS Clin Neurophysiol 1998;108:1. changes neuronal function, especially as a function [8] Roth BJ, Saypol JM, Hallett M, Cohen LG. A of its use parameters, will improve its ability to theoretical calculation of the electric field induced answer neuroscience questions as well as to treat in the cortex during magnetic stimulation. Electro- diseases. Because of its noninvasiveness, it does encephalogr Clin Neurophysiol 1991;81:47. not readily fit under the umbrella of neurosurgery. [9] Bohning DE. Introduction and overview of TMS Nevertheless, it is important for neurosurgeons physics. In: George MS, Belmaker RH, editors. Transcranial magnetic stimulation in neuropsy- to be aware of TMS, because findings from TMS chiatry; Washington (DC): American Psychiatric studies will have implications for neurosurgical Press; 2000. p. 13. approaches like DBS and VNS. Indeed, it is pos- [10] Alexander GE, DeLong MR, Strick PL. Parallel sible to think of using TMS as a potential noninva- organization of functionally segregated circuits link- sive initial screening tool to identify whether ing basal ganglia and cortex. Annu Rev Neurosci perturbation of a circuit has short-term clinical 1986;9:357. effects. In the example of chronic refractory de- [11] Stewart LM, Walsh V, Rothwell JC. Motor and pression or OCD, which is generally a chronic ill- phosphene thresholds: a transcranial magnetic ness, it might then follow that rather than having stimulation correlation study. Neuropsychologia daily or weekly TMS for the rest of their lives, 2001;39:415. [12] Ziemann U, Hallett M. Basic neurophysiological patients would have DBS electrodes implanted in studies with, TMS. In: George MS, Belmaker RH, the same circuit. editors. Transcranial magnetic stimulation in Whatever road the future takes, TMS is an neuropsychiatry. Washington (DC): American Psy- important new tool that will likely be of interest to chiatric Press; 2000. p. 45. neurosurgeons over the next 20 years and perhaps [13] Pridmore S, Filho JAF, Nahas Z, Liberatos C, even longer. George MS. Motor threshold in transcranial 296 M.S. George et al / Neurosurg Clin N Am 14 (2003) 283–301

magnetic stimulation: a comparison of a neuro- cortex excitability by 1 Hz transcranial magnetic physiological and a visualization of movement stimulation. Neurosci Lett 1998;250:141. method. J ECT 1998;14:25. [30] Wu T, Sommer M, Tergau F, Paulus W. Lasting [14] Amassian VE, Eberle L, Maccabee PJ, Cracco RQ. influence of repetitive transcranial magnetic stim- Modelling magnetic coil excitation of human cere- ulation on intracortical excitability in human bral cortex with a peripheral nerve immersed in subjects. Neurosci Lett 2000;287:37. a brain-shaped volume conductor: the significance [31] Epstein CM, Lah JJ, Meador K, Weissman JD, of fiber bending in excitation. Electroencephalogr Gaitan LE, Dihenia B. Optimum stimulus param- Clin Neurophysiol 1992;85:291. eters for lateralized suppression of speech with [15] Cohen LG, Roth BJ, Nilsson J, et al. Effects of coil magnetic brain stimulation. Neurology 1996;47: design on delivery of focal magnetic stimulation. 1590. Technical considerations. Electroencephalogr Clin [32] Wassermann EM. Report on risk and safety of re- Neurophysiol 1990;75:350. petitive transcranial magnetic stimulation (rTMS): [16] Saypol JM, Roth BJ, Cohen LG, Hallett M. A the- suggested guidelines from the International oretical comparison of electric and magnetic stimula- Workshop on Risk and Safety of rTMS (June tion of the brain. Ann Biomed Eng 1991;19:317. 1996). Electroencephalogr Clin Neurophysiol [17] Davey KR, Cheng CH, Epstein CM. Prediction of 1997;108:1. magnetically induced electric fields in biologic tis- [33] Wassermann EM, Cohen LG, Flitman SS, Chen sue. IEEE Trans Biomed Eng 1991;38:418. R, Hallett M. Seizures in healthy people with [18] Greenberg BD, Ziemann U, Harmon A, Murphy repeated safe trains of transcranial magnetic stim- DL, Wassermann EM. Reduced intracortical in- uli. Lancet 1996;347:825. hibition in obsessive-compulsive disorder on trans- [34] Lorberbaum JP, Wassermann EM. Safety con- cranial magnetic stimulation. Lancet 1998;352:881. cerns of transcranial magnetic stimulation. In: [19] Ziemann U, Paulus W, Rothenberger A. De- George MS, Belmaker RH, editors. Transcranial creased motor inhibition in Tourette’s disorder: evi- magnetic stimulation in neuropsychiatry. Wash- dence from transcranial magnetic stimulation. Am ington (DC): American Psychiatric Press; 2000. J Psychiatry 1997;154:1277. p. 141. [20] Ziemann U, Steinhoff BJ, Tergau F, Paulus W. [35] Jennum P, Winkel H. Transcranial magnetic Transcranial magnetic stimulation: its current role stimulation. Its role in the evaluation of patients in epilepsy research. Epilepsy Res 1998;30:11–30. with partial epilepsy. Acta Neurol Scand Suppl [21] Bear MF. Homosynaptic long-term depression: 1994;152:93. a mechanism for memory? Proc Natl Acad Sci [36] Lisanby SH, Schlaepfer TE, Fisch HU, Sackeim USA 1999;96:9457. HA. Magnetic seizure therapy for major depres- [22] Linden DJ. The return of the spike: postsynaptic sion. Arch Gen Psychiatry 2001;58:303. action potentials an the induction of LTP and [37] Little JT, Kimbrell TA, Wassermann EM, et al. LTD. Neuron 1999;22:661. Cognitive effects of 1 and 20 Hz repetitive trans- [23] Artola A, Brocher S, Singer W. Different voltage cranial magnetic stimulation in depression: pre- dependent thresholds for inducing long-term de- liminary report. Neuropsychiatry Neuropsychol pression and long-term potentiation in slices of rat Behav Neurol 2000;13:119. visual cortex. Nature 1990;347:69. [38] Nahas Z, DeBrux C, Lorberbaum JP, et al. Safety [24] Stanton PK, Sejnowsky TJ. Associative long-term of rTMS: MRI scans before and after 2 weeks of depression in the hippocampus induced by hebbian daily left prefrontal rTMS for the treatment of covariance. Nature 1989;339:215. depression. J ECT 2000;16:380. [25] Malenka RC, Nicoll RA. Long-term potentiation: [39] Barker AT, Jalinous R, Freeston IL. Non-invasive a decade of progress? Science 1999;285:1870. magnetic stimulation of the human motor cortex. [26] Wang H, Wang X, Scheich H. LTD and LTP Lancet 1985;1:1106. induced by transcranial magnetic stimulation in [40] Desmurget M, Epstein CM, Turner RS, Prablanc auditory cortex. Neuroreport 1996;7:521. C, Alexander GE, Grafton ST. Role of the [27] Post RM, Kimbrell TA, McCann UD, Osuch EA, posterior parietal cortex in updating reaching Speer AM, Weiss SR. Repetitive transcranial mag- movements to a visual target. Nat Neurosci 1999; netic stimulation as a neuropsychiatric probe 2:563. tool: present status and future potential. J ECT [41] Epstein CM, Verson R, Zangaladze A. Magnetic 1999;15:39–59. coil stimulation suppresses visual perception at an [28] Chen R, Classen J, Gerloff C, et al. Depression of extra-calcarine site. J Clin Neurophysiol 1996;13: motor cortex excitability by low-frequency trans- 247. cranial magnetic stimulation. Neurology 1997;48: [42] Ray PG, Meador KJ, Epstein CM, Loring DW, 1398. Day LJ. Magnetic stimulation of visual cortex: [29] Wassermann EM, Wedegaertner FR, Ziemann U, factors influencing the perception of phosphenes. George MS, Chen R. Crossed reduction of motor J Clin Neurophysio 1998;4:351. M.S. George et al / Neurosurg Clin N Am 14 (2003) 283–301 297

[43] Grafman J, Wassermann E. Transcranial magnetic induced voltage in vivo. Biol Psychiatry 1998; stimulation can measure and modulate learning 43:100s. and memory. Neuropsychologia 1999;37:159. [59] Amassian VE, Quirk GJ, Stewart M. A compar- [44] Boroojerdi B, Phipps M, Kopylev L, Wharton ison of corticospinal activation by magnetic coil CM, Cohen LG, Grafman J. Enhancing analogic and electrical stimulation of monkey motor cor- reasoning with rTMS over the left prefrontal tex. Electroencephalogr Clin Neurophysiol 1990; cortex. Neurology 2001;56:526. 77:390. [45] Flitman SS, Grafman J, Wassermann EM, et al. [60] Olivier E, Edgley SA, Armand J, Lemon RN. Linguistic processing during repetitive transcranial An electrophysiological study of the postnatal magnetic stimulation. Neurology 1998;50:175. development of the corticospinal system in the [46] George MS, Wassermann EM, Williams W, et al. macaque monkey. J Neurosci 1997;17:267. Changes in mood and hormone levels after rapid- [61] Flament D, Hall EJ, Lemon RN. The development rate transcranial magnetic stimulation of the pre- of cortico-motoneuronal projections investigated frontal cortex. J Neuropsychiatry Clin Neurosci using magnetic brain stimulation in the infant 1996;8:172. macaque. J Physiol 1992;447:755. [47] Martin JD, George MS, Greenberg BD, et al. [62] Edgley SA, Eyre JA, Lemon RN, Miller S. Mood effects of prefrontal repetitive high-fre- Excitation of the corticospinal tract by electro- quency TMS in healthy volunteers. CNS Spectrums magnetic and electrical stimulation of the scalp in 1997;2:53. the macaque monkey. J Physiol 1990;425:301. [48] Mosimann UP, Rihs TA, Engeler J, Fisch HU, [63] Lemon RN, Johansson RS, Wrestling G. Modu- Schlaepfer TE. Mood effects of repetitive trans- lation of corticospinal influence over hand muscles cranial magnetic stimulation of left prefrontal during gripping tasks in man and monkey. Can J cortex in healthy volunteers. Psychiatry Res Physiol Pharmacol 1996;74:547. 2000;94:251. [64] Baker SN, Olivier E, Lemon RN. Task-related [49] Grafman J. TMS as a primary brain mapping tool. variation in corticospinal output evoked by trans- In: George MS, Belmaker RH, editors. Trans- cranial magnetic stimulation in the macaque cranial magnetic stimulation (TMS) in neuropsy- monkey. J Physiol 1995;488:795. chiatry. Washington (DC): American Psychiatric [65] Ghaly RF, Stone JL, Aldrete A, Levy W. Effects Press; 2000. p. 115. of incremental ketamine hydrochloride doses on [50] Walsh V, Cowey A. Transcranial magnetic stimu- motor evoked potentials (MEPs) following trans- lation and cognitive neuroscience. Nat Rev Neuro- cranial magnetic stimulation: a primate study. sci 2000;1:73. J Neurosurg Anesthesiol 1990;2:79. [51] Penfield W. The mystery of the mind. A critical [66] Stone JL, Ghaly RF, Levy WJ, Kartha R, Krinsky study of consciousness and the human brain. L, Roccaforte P. A comparative analysis of en- Princeton: Princeton University Press; 1975. flurane anesthesia on primate motor and somato- [52] Jennum P, Friberg L, Fuglsang-Fredericksen A, sensory evoked potentials. Electroencephalogr Dam M. Speech localization using repetitive trans- Clin Neurophysiol 1992;84:180. cranial magnetic stimulation. Neurology 1994; [67] Baker SN, Olivier E, Lemon RN. Recording an 44:269. identified pyramidal volley evoked by transcranial [53] Pascual-Leone A, Gates JR, Dhuna A. Induction magnetic stimulation in a conscious macaque of speech arrest and counting errors with rapid- monkey. Exp Brain Res 1995;99:529. rate transcranial magnetic stimulation. Neurology [68] Fleischmann A, Steppel J, Leon A, Belmaker RH. 1991;41:697. The effect of transcranial magnetic stimulation [54] Michelucci R, Valzania F, Passarelli D, et al. compared with electroconvulsive shock on rat Rapid-rate transcranial magnetic stimulation and apomorphine-induced stereotypy. Eur Neuropsy- hemispheric language dominance: usefulness and chopharmacol 1994;4:449. safety in epilepsy. Neurology 1994;44:1697. [69] Fleischmann A, Sternheim A, Etgen AM, Li C, [55] Roberts DR, Vincent DJ, Speer AM, et al. Multi- Grisaru N, Belmaker RH. Transcranial magnetic modality mapping of motor cortex: comparing stimulation downregulates beta-adrenoreceptors in echoplanar BOLD fMRI and transcranial mag- rat cortex. J Neural Transm 1996;103:1361. netic stimulation. J Neural Transm 1997;104:833. [70] Ben-Sachar D, Belmaker RH, Grisaru N, Klein E. [56] Stewart L, Walsh V, Frith U, Rothwell JC. TMS Transcranial magnetic stimulation induces alter- produces two dissociable types of speech disrup- ations in brain monoamines. J Neural Transm tion. Neuroimage 2001;13:472. 1997;104:191. [57] Lisanby SH, Luber B, Finck D, et al. Primate [71] Fujiki M, Steward O. High frequency trans- models of transcranial magnetic stimulation cranial magnetic stimulation mimics the effects [abstract]. Biol Psychiatry 1998;41:76s. of ECS in upregulating astroglial gene expression [58] Lisanby SH, Luber B, Schroeder C, et al. In- in the murine CNS. Mol Brain Res 1997;44: tracerebral measurement of rTMS and ECS 301. 298 M.S. George et al / Neurosurg Clin N Am 14 (2003) 283–301

[72] Jennum P, Klitgaard H. Effect of acute and [86] Strafella AP, Paus T, Barrett J, Dagher A. chronic stimulations on pentylenetetrazole-induced Repetitive transcranial magnetic stimulation of clonic seizures. Epilepsy Res 1996;23:115. the human prefrontal cortex induces dopamine [73] Pope A, Keck ME. TMS as a therapeutic tool in release in the caudate nucleus. J Neurosci 2001; psychiatry: what do we know about neurobiolog- 21:RC157. ical mechanisms? J Psychiatr Res 2001;35:193. [87] Paus T, Castro-Alamancos MA, Petrides M. [74] Weissman JD, Epstein CM, Davey KR. Magnetic Cortico-cortical connectivity of the human mid- brain stimulation and brain size: relevance to ani- dorsolateral frontal cortex and its modulation by mal studies. Electroencephalogr Clin Neurophysiol repetitive transcranial magnetic stimulation. Eur J 1992;85:215. Neurosci 2001;14:1405. [75] Nobler MS, Oquendo MA, Kegeles LS, et al. [88] George MS, Stallings LE, Speer AM, et al. Decreased regional brain metabolism after ECT. Prefrontal repetitive transcranial magnetic stimu- Am J Psychiatry 2001;158:305. lation (rTMS) changes relative perfusion locally [76] Teneback CC, Nahas Z, Speer AM, et al. Two and remotely. Hum Psychopharmacol 1999;14:161. weeks of daily left prefrontal rTMS changes [89] Li X, Teneback CC, Nahas Z, et al. Lamotrigine prefrontal cortex and paralimbic activity in de- inhibits the functional magnetic resonance imaging pression. J Neuropsychiatry Clin Neurosci 1999; response to transcranial magnetic stimulation in 11:426. healthy adults [abstract]. Biol Psychiatry 2002; [77] Nahas Z, Teneback CT, Kozel AF, et al. Brain 51:8S. effects of transcranial magnetic stimulation de- [90] Kimbrell TA, Ketter TA, George MS, et al. livered over prefrontal cortex in depressed adults: Regional cerebral glucose utilization in patients the role of stimulation frequency and distance from with a range of severities of unipolar depression. coil to cortex. J Neuropsychiatry Clin Neurosci Biol Psychiatry 2002;51:237. 2001;13:459. [91] George MS, Ketter TA, Gill D, et al. Brain regions [78] Shajahan PM, Glabus MF, Gooding PA, Shah PJ, involved in recognizing facial emotion or iden- Ebmeier KP. Reduced cortical excitability in tity: an O15 PET study. J Neuropsychiatry Clin depression–Impaired post-exercise motor facilita- Neurosci 1993;5:384. tion with transcranial magnetic stimulation. Br J [92] George MS, Ketter TA, Post RM. SPECT and Psychiatry 1999;174:449–54. PET imaging in mood disorders. J Clin Psychiatry [79] Speer AM, Kimbrell TA, Wasserman EM, et al. 1993;54:6. Opposite effects of high and low frequency rTMS [93] George MS, Huggins T, McDermut W, Parekh PI, on regional brain activity in depressed patients. Rubinow D, Post RM. Abnormal facial emotion Biol Psychiatry 2000;48:1133. recognition in depression: serial testing in an ultra- [80] Mitchel P. 15 Hz and 1 Hz TMS have different rapid-cycling patient. Behav Modif 1998;22:192. acute effects on cerebral blood ﬂow in depressed [94] George MS. An introduction to the emerging patients [abstract]. Int J Neuropsychopharmacol neuroanatomy of depression. Psychiatr Ann 1994; 2002;5:S7. 24:635. [81] Shastri A, George MS, Bohning DE. Performance [95] George MS, Ketter TA, Parekh PI, et al. Regional of a system for interleaving transcranial magnetic brain activity when selecting a response despite stimulation with steady state magnetic resonance interference: An H215O PET study of the Stroop imaging. Electroencephalogr Clin Neurophysiol and an emotional Stroop. Hum Brain Mapping Suppl 1999;51:55. 1994;1:194. [82] Bohning DE, Shastri A, McConnell K, et al. A [96] George MS, Ketter TA, Parekh PI, Horwitz B, Combined TMS/fMRI study of intensity-depen- Herscovitch P, Post RM. Brain activity during dent TMS over motor cortex. Biol Psychiatry transient sadness and happiness in healthy women. 1999;45:385. Am J Psychiatry 1995;152:341. [83] Bohning DE, Shastri A, Wassermann EM, et al. [97] George MS, Ketter TA, Post RM. What functional BOLD-fMRI response to single-pulse transcranial imaging studies have revealed about the brain basis magnetic stimulation (TMS). J Magn Reson of mood and emotion. In: Panksepp J, editor. Imaging 2000;11:569. Advances in biological psychiatry. Greenwich [84] Baudewig J, Siebner HR, Bestmann S, et al. (CT): JAI Press; 1996. p. 63. Functional MRI of cortical activations induced [98] George MS, Ketter TA, Parekh PI, et al. Blunted by transcranial magnetic stimulation (TMS). Neu- left cingulate activation in mood disorder subjects roreport 2001;12:3543. during a response interference task (the Stroop). [85] Nahas Z, Lomarev M, Roberts DR, et al. J Neuropsychiatry Clin Neurosci 1997;9:55. Unilateral left prefrontal transcranial magnetic [99] Ketter TA, Andreason PJ, George MS, et al. stimulation (TMS) produces intensity-dependent Anterior paralimbic mediation of procaine- bilateral effects as measured by interleaved BOLD induced emotional and psychosensory experiences. fMRI. Biol Psychiatry 2001;50:712. Arch Gen Psychiatry 1996;53:59. M.S. George et al / Neurosurg Clin N Am 14 (2003) 283–301 299

[100] Beer B. Uber das Auftretten einer objectiven [115] Grunhaus L, Dannon PN, Schreiber S, et al. Lichtempfindung in magnetischen Felde. Klinische Repetitive transcranial magnetic stimulation is Wochenzeitschrift 1902;15:108. as effective as electroconvulsive therapy in the [101] Kolbinger HM, Hoflich G, Hufnagel A, Moller treatment of nondelusional major depressive H-J, Kasper S. Transcranial magnetic stimulation disorder: an open study. Biol Psychiatry 2000; (TMS) in the treatment of major depression—a 47:314. pilot study. Hum Psychopharmacol 1995;10:305. [116] Janicak PG, Dowd SM, Martis B, et al. Repetitive [102] Grisaru N, Yarovslavsky U, Abarbanel J, Lam- transcranial magnetic stimulation versus electro- berg T, Belmaker RH. Transcranial magnetic convulsive therapy for major depression: prelimi- stimulation in depression and schizophrenia. Eur nary results of a randomized trial. Biol Psychiatry Neuropsychopharmacol 1994;4:287. 2002;51:659. [103] Nobler MS, Sackeim HA, Prohovnik I, et al. [117] Pridmore S. Substitution of rapid transcranial Regional cerebral blood flow in mood disorders, magnetic stimulation treatments for electroconvul- III. Treatment and clinical response. Arch Gen sive therapy treatments in a course of electrocon- Psychiatry 1994;51:884. vulsive therapy. Depress Anxiety 2000;12:118. [104] George MS, Wassermann EM, Williams WA, et al. [118] Dannon PN, Dolberg OT, Schreiber S, Grunhaus Daily repetitive transcranial magnetic stimulation L. Three and six-month outcome following courses (rTMS) improves mood in depression. Neuro- of either ECT or rTMS in a population of severely report 1995;6:1853. depressed individuals—preliminary report. Biol [105] George MS, Wassermann EM, Williams WE, et al. Psychiatry 2002;51:687. Mood improvements following daily left prefron- [119] Klein E, Kreinin I, Chistyakov A, et al. Thera- tal repetitive transcranial magnetic stimulation in peutic efficacy of right prefrontal slow repetitive patients with depression: a placebo-controlled transcranial magnetic stimulation in major de- crossover trial. Am J Psychiatry 1997;154:1752. pression: a double-blind controlled study. Arch [106] Pascual-Leone A, Rubio B, Pallardo F, Catala Gen Psychiatry 1999;56:315. MD. Beneficial effect of rapid-rate transcranial [120] Herwig U, Padberg F, Unger J, Spitzer M, magnetic stimulation of the left dorsolateral pre- Schonfeldt-Lecuona C. Transcranial magnetic frontal cortex in drug-resistant depression. Lancet stimulation in therapy studies: examination of the 1996;348:233. reliability of ‘‘standard’’ coil positioning by neuro- [107] Loo C, Mitchell P, Sachdev P, McDarmont B, navigation. Biol Psychiatry 2001;50:58. Parker G, Gandevia S. A double-blind controlled [121] McConnell KA, Nahas Z, Shastri A, et al. The investigation of transcranial magnetic stimulation transcranial magnetic stimulation motor threshold for the treatment of resistant major depression. depends on the distance from coil to underlying Am J Psychiatry 1999;156:946. cortex: a replication in healthy adults comparing [108] McNamara B, Ray JL, Arthurs OJ, Boniface S. two methods of assessing the distance to cortex. Transcranial magnetic stimulation for depression Biol Psychiatry 2001;49:454. and other psychiatric disorders. Psychol Med 2001; [122] Kozel FA, Nahas Z, DeBrux C, et al. How the 31:1141. distance from coil to cortex relates to age, motor [109] Burt T, Lisanby SH, Sackeim HA. Neuropsychiat- threshold and possibly the antidepressant response ric applications of transcranial magnetic stimula- to repetitive transcranial magnetic stimulation. tion. Int J Neuropsychopharmacol 2002;5:73–103. J Neuropsychiatry Clin Neurosci 2000;12:376. [110] Holtzheimer PE, Russo J, Avery D. A meta- [123] Mosimann UP, Marre SC, Werlen S, et al. analysis of repetitive transcranial magnetic Antidepressant effects of repetitive transcranial stimulation in the treatment of depression. magnetic stimulation in the elderly: correlation Psychopharmacol Bull 2001;35:149. between effect size and coil-cortex distance. Arch [111] Kozel FA, George MS. Meta-analysis of left pre- Gen Psychiatry 2002;59:560. frontal repetitive transcranial magnetic stimula- [124] Nahas Z, Oliver NC, Johnson M, et al. Feasi- tion (rTMS) to treat depression. J Psychiatr Pract bility and efficacy of left prefrontal rTMS as a 2002;8:270–5. maintenance antidepressant [abstract]. Biol Psy- [112] Martin JLR, Barbanoj MJ, Schlaepfer TE, et al. chiatry 2000;48:57. Transcranial magnetic stimulation for treating de- [125] Belmaker RH, Grisaru N. Does TMS have bipolar pression. The Cochrane Library. Oxford: Update efficacy in both depression and mania [abstract]? Software; 2002. Biol Psychiatry 1998;48:157. [113] Thase ME. The need for clinically relevant re- [126] Grisaru N, Chudakov B, Yaroslavsky Y, Belmaker search on treatment-resistant depression. J Clin RH. TMS in mania: a controlled study. Am J Psychiatry 2001;62:221. Psychiatry 1998;155:1608. [114] Thase ME. Studying new antidepressants: if there [127] Nahas Z, Bohning DE, Molloy M, Outz JA, Risch were a light at the end of the tunnel, could we see SC, George MS. Safety and feasibility of repetitive it? J Clin Psychiatry 2002;63:24. transcranial magnetic stimulation in the treatment 300 M.S. George et al / Neurosurg Clin N Am 14 (2003) 283–301

of anxious depression in pregnancy. J Clin stimulation in patients with major depression and Psychiatry 1999;60:50. schizophrenia. Depress Anxiety 1998;7:65. [128] Pascual-Leone A, Valls-Sole J, Brasil-Neto JP, [141] Puri BK, Davey NJ, Ellaway PH, Lewis SW. An Cohen LG, Hallett M. Akinesia in Parkinson’s investigation of motor function in schizophrenia disease. I. Shortening of simple reaction times with using transcranial magnetic stimulation of the focal, single-pulse transcranial magnetic stimula- motor cortex. Br J Psychiatry 1996;169:690. tion. Neurology 1994;44:884. [142] Nahas Z, McConnell K, Collins S, et al. Could left [129] Ghabra MB, Hallett M, Wassermann EM. Simul- prefrontal rTMS modify negative symptoms and taneous repetitive transcranial magnetic stimu- attention in schizophrenia [abstract]? Biol Psychi- lation does not speed fine movement in PD. atry 1999;45:37. Neurology 1999;52:768. [143] Hoffman R, Boutros N, Berman R, Krystal J, [130] Boylan LS, Pullman SL, Lisanby SH, Spicknall Charney D. Transcranial magnetic stimulation and KE, Sackeim HA. Repetitive transcranial magnetic hallucinated ‘‘voices’’ [abstract]. Biol Psychiatry stimulation to SMA worsens complex movements 1998;43:93s. in Parkinson’s disease. Clin Neurophysiol 2001; [144] Hoffman RE, Boutros NN, Hu S, Berman RM, 112:259. Krystal JH, Charney DS. Transcranial magnetic [131] Mally J, Stone TW. Therapeutic and ‘‘dose- stimulation and auditory hallucinations in schizo- dependent’’ effect of repetitive microelectroshock phrenia. Lancet 2000;355:1073. induced by transcranial magnetic stimulation in [145] Greenberg BD, George MS, Dearing J, et al. Effect Parkinson’s disease. J Neurosci Res 1999;57:935. of prefrontal repetitive transcranial magnetic [132] Mally J, Stone TW. Improvement in Parkinsonian stimulation (rTMS) in obsessive compulsive disor- symptoms after repetitive transcranial magnetic der: a preliminary study. Am J Psychiatry 1997; stimulation. J Neurol Sci 1991;162:179. 154:867. [133] Siebner HR, Rossmeier C, Mentschel C, Peine- [146] Cora-Locatelli G, Greenberg BD, Harmon A, mann A, Conrad B. Short-term motor improve- et al. Cortical excitability and augmentation strat- ment after sub-threshold 5-Hz repetitive egies in OCD [abstract]. Biol Psychiatry 1998; transcranial magnetic stimulation of the primary 43:77s. motor hand area in Parkinson’s disease. J Neurol [147] Greenberg BD, Ziemann U, Cora-Locatelli G, Sci 2000;178:91. et al. Altered cortical excitability in obsessive- [134] Shimamoto H, Morimitsu H, Sugita S, Nakahara compulsive disorder. Neurology 2000;54:142. K, Shigemori M. Therapeutic effect of repetitive [148] Alonso P, Pujol J, Cardoner N, et al. Right pre- transcranial magnetic stimulation in Parkinson’s frontal TMS in OCD: a double-blind, placebo- disease. Rinsho Shinkeigaku 1999;39:1264. controlled study. Am J Psychiatry 2001;158:1143. [135] Shimamoto H, Takasaki K, Shigemori M, Imai- [149] Sachdev PS, McBride R, Loo CK, Mitchell PB, zumi T, Ayabe M, Shoji H. Therapeutic effect Malhi GS, Croker VM. Right versus left prefrontal and mechanism of repetitive transcranial magnetic transcranial magnetic stimulation for obsessive- stimulation in Parkinson’s disease. J Neurol 2001; compulsive disorder: a preliminary investigation. 248(Suppl):III48. J Clin Psychiatry 2001;62:981. [136] Siebner HR, Tormos JM, Ceballos-Baumann AO, [150] McCann UD, Kimbrell TA, Morgan CM, et al. Auer C, Catala MD, Conrad B, et al. Low fre- Repetitive transcranial magnetic stimulation for quency repetitive transcranial magnetic stimula- posttraumatic stress disorder [letter]. Arch Gen tion of motor cortex in patients with writer’s Psychiatry 1998;55:276. cramp. Neurology 1999;52:529. [151] Grisaru N, Amir M, Cohen H, Kaplan Z. Effect [137] Chae JH, Nahas Z, Wassermann EM, et al. A pilot of transcranial magnetic stimulation in posttrau- study using rTMS to probe the functional neuro- matic stress disorder: a preliminary study. Biol anatomy of tics in Tourette syndrome. Neuropsy- Psychiatry 1998;44:52. chiatry Neuropsychol Behav Neurol [in press] [152] Lisanby SH, Luber B, Sackeim HA, Finck AD, [138] Reutens DC, Puce A, Berkovic SF. Cortical hy- Schroeder C. Deliberate seizure induction with perexcitability in progressive myoclonus epilepsy: repetitive transcranial magnetic stimulation in non- a study with transcranial magnetic stimulation. human primates [letter]. Arch Gen Psychiatry Neurology 1993;43:186. 2001;58:199. [139] Wedegaertner F, Garvey M, Cohen LG, Hallett [153] Lisanby SH, Luber B, Barriolhet L, Neufeld E, M, Wassermann EM. Low frequency repetitive Schlaepfer T, Sackeim HA. Magnetic seizure ther- transcranial magnetic stimulation can reduce apy (MST)—acute cognitive effects of MST action myoclonus [abstract]. Neurology 1997;48: compared with ECT [abstract]. J ECT 2001;17:77. A119. [154] Sackeim HA, Decina P, Kanzler M, Kerr B, Malitz [140] Feinsod M, Sreinin B, Chistyakov A, Klein E. S. Effects of electrode placement on the efficacy Preliminary evidence for a beneficial effect of of titrated, low-dose ECT. Am J Psychiatry 1987; low-frequency, repetitive transcranial magnetic 144:1449. M.S. George et al / Neurosurg Clin N Am 14 (2003) 283–301 301

[155] Sackeim HA, Prudic J, Devanand DP, et al. Effects [159] Marangell LB, Rush AJ, George MS, et al. Vagus of stimulus intensity and electrode placement on nerve stimulation (VNS) for major depressive epi- the efﬁcacy and cognitive effects of electroconvul- sodes: longer-term outcome. Biol Psychiatry 2002; sive therapy. N Engl J Med 1993;328:839. 51:280. [156] George MS, Wassermann EM. Rapid-rate trans- [160] Elger G, Hoppe C, Falkai P, Rush AJ, Elger CE. cranial magnetic stimulation (rTMS) and ECT. Vagus nerve stimulation is associated with mood Conv Ther 1994;10:251. improvements in epilepsy patients. Epilepsy Res [157] Rush AJ, George MS, Sackeim HA, et al. Vagus 2000;42:203. nerve stimulation (VNS) for treatment-resistant [161] Harden CL, Pulver MC, Ravdin LD, Nikolov B, depressions: A multicenter study. Biol Psychiatry Halper JP, Labar DR. A pilot study of mood in 2000;47:276. epilepsy patients treated with vagus nerve stimula- [158] Sackeim HA, Rush AJ, George MS, et al. Vagus tion. Epilepsy Behav 2000;1:93. nerve stimulation (VNS) for treatment-resistant [162] Limousin P, Krack P, Pollak P, et al. Electrical depression: efﬁcacy, side effects and predictors of stimulation of the subthalamic nucleus in advanced outcome. Neuropsychopharmacology 2001;25:713. Parkinson’s disease. N Engl J Med 1998;339:1105. Neurosurg Clin N Am 14 (2003) 303–319

From psychosurgery to neuromodulation and palliation: history’s lessons for the ethical conduct and regulation of neuropsychiatric research Joseph J. Fins, MD Division of Medical Ethics, Departments of Medicine and Public Health, Weill Medical College of Cornell University, New York–Weill Cornell Medical Center, 525 East 68th Street, F-173, New York, NY, USA

We are certain that these experiments ment of Parkinson’s disease and contemplating shall stir up keen discussions in the medical, their use for severe psychiatric illnesses, such as psychiatric, psychological, philosophical, obsessive-compulsive disorder (OCD) [4,5] and social and other fields. We expect that, but the modulation of consciousness in traumatic hope at the same time that this discussion brain injury [6–8]. shall promote the progress of science and Any student of medical history would have to above all the benefit of mental patients [1]. ask if these developments are ethically appropri- Egas Moniz, 1936 ate and whether the promise of neuromodulation will be able to transcend the potential peril ...We, the doctors, are so fallible, ever associated with the manipulation of motor, psy- beset with the common and fatal facility of chiatric, or cognitive function. Can today’s inves- reaching conclusions from superficial observations, and constantly misled by the tigators avoid the moral blindness of their ease with which our minds fall into the rut predecessors? Will society tolerate this new foray of one or two experiences [2]. into somatic therapy and seek to regulate it as William Osler, 1903 legitimate science, or will a lingering memory of psychosurgery be so overwhelming as to make this impossible? Optimism tempered by history Lay journalists covering the most exciting As we contemplate the emerging era of neuro- developments associated with deep brain stimula- modulation and imagine the utility of deep brain tion often ask these questions [9]. A recent stimulation for disease entities in neurology and editorial in The Economist, for example, asserted psychiatry, our enthusiasm is immediately tem- that these new developments in neurobiology pose pered by history. Just a generation ago, other a ‘‘greater threat to human dignity’’ than the confident investigators were heralding invasive debate over cloning [10]. Editorialists like William somatic therapies like prefrontal lobotomy to Safire have raised questions about the world of treat psychiatric illness. That era of psychosurgery ‘‘neuroethics’’ [11], and The Washington Post ended with widespread condemnation, congres- Magazine featured a story on the lobotomist sional calls for a ban [3], and a vow that history Walter Freeman in response to the excitement should never repeat itself. Now, just 30 years later, over deep brain stimulators [12]. In the lay press, neurologists, neurosurgeons, and psychiatrists are the link between psychosurgery and present implanting deep brain stimulators for the treat- efforts in neuromodulation is clear, and the message has been cautionary [13]. Historical analogies are important, but they E-mail address: jjfi[email protected] only tell part of the story. From a scientific

1042-3680/03/$ - see front matter Ó 2003, Elsevier Inc. All rights reserved. doi:10.1016/S1042-3680(02)00118-3 304 J.J. Fins / Neurosurg Clin N Am 14 (2003) 303–319 standpoint, there are critically important differ- By reviewing the history of the psychosurgery ences between psychosurgery and neuromodula- movement, I will demonstrate that this work tion. Although both are somatic therapies, deep continued in an unregulated and ethically dispro- brain stimulation seeks to modulate brain func- portionate fashion despite well-articulated and tion through the insertion of electrodes. In con- early criticism. I will maintain that this was pos- trast, psychosurgery destroyed vital brain tissue sible because there was no mechanism to regulate through lesioning. These lesions were also perma- the conduct of these practices until the advent nent, whereas the presently known effects of of modern bioethics in the late 1960s and early neuromodulation are reversible. In addition, the 1970s as a scholarly discipline able to influence permanent cognitive impairment seen with psy- clinical practice and public policy. chosurgical ablation does not occur with deep Earlier critiques, although often cogent and brain stimulation. well informed, were unable to influence clinical There are also important differences between or research practice because they were made in Freeman’s prefrontal lobotomy and psychosur- a practice environment that had yet to overcome gical procedures performed today. Psychosurgi- the prerogative of the physician to direct care cal operations of the past were crude and their without additional oversight. As we shall see, indi- rationale was based on a mix of anecdotal ex- viduals who criticized their colleagues’ practice perience and supposition [14,15]. Current psycho- of psychosurgery in the literature or at national surgical procedures utilize modern neuroimaging meetings were powerless to ensure that their techniques and are more precise and procedu- concerns were heard at the psychosurgeon’s home rally diverse. They have a more benign side institution. There was no mechanism to collect effect profile and demonstrate evidence of efficacy these arguments and regulate these activities. for some refractory neuropsychiatric conditions It would take broader societal changes, such as [16]. the decline of physician paternalism and the emer- Modern neuromodulation’s knowledge base, gence of civil and patient rights [20] reflected in although still a nascent area of investigation, far the establishment of bioethics think-tanks like the exceeds what was known to psychosurgeons at Hastings Center and the regulatory efforts midcentury [17]. Today, clinical investigators are of the National Commission for the Protection supported by an extensive platform of structural of Human Subjects of Biomedical and Behavioral and functional neuroimaging and detailed ana- Research, before broader regulations governing tomic and electrophysiologic studies that allow clinical research were put into place. for more precise hypotheses concerning neural Sadly, much of this history has already been systems underlying disease states. forgotten or deemed irrelevant. In an oddly analo- These improvements have led to greater pro- gous way, the debate over lobotomy and psycho- cedural precision and better diagnostic capacity surgery has gone the way of a Tommy Dorsey to identify mechanisms of action, efficacy, and long-playing record or a Led Zeppelin eight-track lack of therapeutic response. In the treatment tape. As we shall see, many of the regulatory of Parkinson’s disease, these advances have co- recommendations made by the National Com- alesced so that neuromodulation is now a mature mission, such as the proposed National Psycho- and established therapy for refractory disease surgery Advisory Board or recommendations on [18,19]. Deep brain stimulation procedures for research involving the institutionalized mentally movement disorders are covered in the United infirmed, never went into law. This lacuna in the States by Medicare and have become the standard law necessitated the Clinton era 1998 National of care for refractory disease. Bioethics Advisory Commission’s ‘‘Report on Although these scientific differences are signif- Research Involving Persons With Mental Disor- icant, a more salient distinction is the respective ders That May Affect Decisionmaking Capacity’’ historical and cultural contexts of psychosurgery [21]. Ironically, this is yet another set of recom- and neuromodulation. Historically, psychosur- mendations that have yet to be ensconced in gery developed before the rise of modern bio- Federal regulations. ethics. Neuromodulation, today, is emerging both The emergence of neuromodulation as credible against the historical backdrop of psychosurgery science with real and potential clinical applica- and within an ethical and regulatory context that tions, however, makes it essential that we revisit should be far more attentive to human subject these earlier ethical and clinical debates so that the protections. hard-won wisdom and scholarship of earlier eras J.J. Fins / Neurosurg Clin N Am 14 (2003) 303–319 305 can help to navigate what will be demonstrated as Psychiatric Association in San Francisco, the recurrent ethical challenges. Many familiar ques- New York psychiatrist urged professional humil- tions about consent, coercion, and personhood ity. He cautioned investigators to have some humil- are now resurfacing in the debate over applying ity in their speculations and an awareness of the neuromodulation techniques, neé psychosurgery, investment that they have in their own theoretic to OCD, the first of many psychiatric entities to speculations. To mitigate against these excesses, which this modality may be applied. he urged investigators to assume as ‘‘an inflexi- It is my hope that a reconsideration of this rich ble duty’’ to gain knowledge of the mechanism history in light of these promising scientific de- of any proposed therapy [30]. Specifically, he velopments will deepen the insight of a new observed: generation of psychiatrists, neurologists, and neu- In any scientific treatment which is not rosurgeons who are destined to face challenges fully understood there is the serious danger that their forbears could only have imagined. of not being able to predict possible damage. It should be an inflexible duty to When psychosurgery was therapy become thoroughly familiar with the drug or procedure to be employed. This implies It is one of those strange paradoxes in history a thorough understanding of the phy- that 2 years after the articulation of the Nurem- siological and psychobiological functions berg Code and the development of the sterotac- which may have a bearing on the proposed treatment. No one has the right to invent tic technique, Egas Moniz won the Nobel Prize theories to suit his desires without having for the development of the prefrontal leukotomy given full consideration to what has been [22,23]. He began his work in 1935 [24], just 3 established or found plausible [30]. years after the initiation of the Tuskegee Syphilis Study [25]. Both the Tuskegee Syphilis Study and In the same address, Diethelm implicitly com- psychosurgery would be criticized decades later mented on issues that would later inform the lit- as an abrogation of patient rights invoking the erature on informed consent. He warned of the same ethical principles articulated in the Nurem- vulnerability of lay people to the latest scientific berg Code. In 1949, however, the Nobel Prize cele- trend and the powerlessness of patients, ‘‘who are brated Moniz’s work in medicine and physiology. forced to follow’’: Although some have maintained that the ...it is important in medicine to recog- Nobel Prize was really meant for Moniz’s more nize fully the responsibility with regard to enduring discovery of cerebral angiography [26], those who follow voluntarily that is the the citation on his medal read simply, ‘‘For his physicians; to those who follow blindly, discovery of the therapeutic value of prefrontal that is lay people; and to those who are leucotomy in certain psychoses’’ [27]. The un- forced to follow, that is the patients. The plea- derstatement in this inscription fails to capture the sure of being an inventor and pathfinder is alluring but leads to all the dangers of ad- lack of effective therapy for severe mental illness venture [30]. and the desperation of patients and the frustration of their families and caregivers. When Moniz was In these comments, Diethelm anticipates the given the prize, his work was hailed as being at the Nuremberg Code with its emphasis on the scien- vanguard of therapeutics for severe mental illness tific basis of human experimentation and the [28]. In the era before the introduction of anti- centrality of informed consent and voluntariness psychotics, prefrontal lobotomy offered a potential [31]. therapy for the severely and persistently mentally A young psychiatrist attending the same meet- ill who otherwise would require institutionaliza- ing tells us that, as yet, there was no mention of tion. It was viewed as a therapy of last resort and Moniz’s work at the conference and that Die- positively portrayed in the American media, at thelm did not mention lobotomy specifically in his least in the early years after its introduction [29]. address [32]. His comments were more generic and All was not sanguine in professional circles, directed toward the growing place of somatic ther- however. As early as 1938, Oskar Diethelm, a con- apy in psychiatry, a subdiscipline of the field of temporary of Freeman and Watts in the United which psychosurgery would become a part. States, poignantly warned about the emerging Diethelm’s warnings would be applicable to somatic therapies in psychiatry [30]. In an ad- the leading proponents of psychosurgery as illus- dress at the Annual Meeting of the American trated by the hubris of both Freeman and Moniz. 306 J.J. Fins / Neurosurg Clin N Am 14 (2003) 303–319

Moniz himself was smug and overly optimistic opposed to a ‘‘...mutilating operation that about his work and methods. In an early sum- destroys brain tissue.’’ mary article originally published in the American Ultimately, Grinker questioned Freeman’s Journal of Psychiatry in 1937, he summarily dis- methodology because ‘‘...once one cuts, there is missed any discussion of the risks and benefits of no return.’’ Given the irreversibility of the pro- his prefrontal leukotomy. He simply concluded by cedure, he voiced concern about how patients asserting: were selected for the procedure and the vagueness Following this exposition I do not wish of entry criteria that required the patient to have to make any comment since the facts speak failed ‘‘conservative measures.’’ He was particu- for themselves. These were hospital patients larly aware of the vulnerability of most of the who were well studied and well followed. patients who came from state facilities and whose The recoveries have been maintained. I preoperative assessment was suspect: cannot believe that the recoveries can be explained upon simple coincidence. Prefrontal ...We are dealing with a large number of leucotomy is a simple operation, always patients who have been chosen from state safe, which may prove to be an effective hospital populations. There have been some surgical treatment in certain cases of men- private cases, it is true, but when one thinks tal disorder [33]. of what conservative treatment means, of what actual, thoroughgoing psychologic Freeman, too, was overly confident. Just study and treatment means, it is difficult months before the American Psychiatric Associ- to imagine that the patients who come from ation meeting, he demonstrated the traits that state hospital populations are getting the Diethelm believed were so dangerous in a clinical benefit of that type of treatment [35]. investigator. Lawrence M. Weinberger recalls that He expressed his ‘‘surprise’’ that patients were as a young physician, he met Freeman in the lobotomized after only a few months of mental spring of 1938. Weinberger was a second-year illness before a spontaneous remission might oc- fellow in neurologic surgery at the University of cur and in young patients in their 30s. He also Pennsylvania and had traveled to the Delaware questioned the application of the procedure to State Hospital to meet Freeman and to learn how a wide range of diagnoses and was concerned to perform the novel procedure. Freeman’s meth- about the postoperative defects and the means to od of patient selection and his ‘‘purely observa- assess their impact on intelligence. tional’’ technique of determining who might be a Addressing the risks of the procedure, he suitable candidate for surgery mortified Wein- suggested that more attention needed to be paid berger. He asked Freeman about his selection to cognitive impairment and the late effects of criteria and ‘‘...was answered with a prolonged scarring, which might lead to the development silent stare and finally one word: ÔExperience!Õ’’ of seizure disorders. He also noted the vague [34]. Weinberger reported being ‘‘squelched’’ and way in which cognitive defects were described saying no more [34]. noting, ‘‘There must be developed a technique of Concerns were not limited to junior trainees. measuring very carefully both the existing Roy Grinker publicly questioned Walter Freeman function and the defect in the individuals before at a panel discussion before the Section on and after the operation’’ [35]. Grinker also Nervous and Mental Diseases at the annual observed that a technologic solution to psychic session of the American Medical Association held problems was not always in order, although the in Cleveland in 1941 [35]. At that time, Grinker alternatives were often more time-consuming and was Chief of the Division of Neuropsychiatry at taxing: Michael Reese Hospital in Chicago. He would later become a training analyst and Editor-in- It is obvious that if the anxiety and Chief of the Archives of General Psychiatry [36]. suffering, the things which one wants to Grinker sought to moderate attitudes toward relieve, are based on conflict, the rational basis of treatment is a psychological ther- lobotomy, about which he acknowledged ‘‘...a apy. This can not always be done. It is great deal of preconception and emotional bias.’’ sometimes very tedious and very costly, but He acknowledged that there was controversy if it were more possible we would not between those who believed that the psychoses hesitate to ask for more psychologic work had an organic basis and might be amenable to in state hospitals rather than more operat- ‘‘physical therapy in psychiatry’’ and those ing rooms [35]. J.J. Fins / Neurosurg Clin N Am 14 (2003) 303–319 307

Building on a therapeutic approach that was [40]. During this meeting, although lobotomy was broader than the merely surgical, Grinker also being used as a therapy in the community, the outlined the professional responsibilities of the editor observed that, ‘‘although we seem well surgeon and his obligation to ensure that ade- warranted to continue the procedure, it is not even quate assessment occurred before any procedure clear that psychosurgery, as performed now, is was performed. Without equivocation, he asserted more beneficial than harmful’’ [40]. Experts were that: unable to reach a consensus except to conclude that ‘‘...an enormous amount of further research I do not believe that the surgeon can is an imperative need in nearly every aspect of the assume the responsibility of being the oper- field’’ [40]. ator at the behest of his neurologic and Among these needs was greater clarity about psychiatric friends and say ‘‘I did it because whom the procedure was being performed on and they told me to.’’ He must assume the responsibility of insisting that before such how patients fared. Eleven years after Weinberger operations are done all conservative personally questioned Freeman about his method means—and by ‘‘conservative means’’ I of selecting patients and evaluating outcomes, mean thorough going psychological treat- methodologic issues remained a concern. One ment-have been employed [35]. thoughtful commentator on that era has maintained that the patient’s diagnosis was less Grinker’s criticisms are all the more credible relevant than the severity of the patient’s symp- because he did not categorically condemn the toms in determining whether or not he or she procedure. Instead, he saw it not as a therapy but would be a surgical candidate [41]. It was against ‘‘still an experiment.’’ Amid proponents and crit- this imprecision that experts agreed that there was ics of psychosurgery, he confessed that his point a need for a comprehensive means of engaging of view, ‘‘...perhaps might disappoint some, as I in outcome measures to discern whether the pro- am not iconoclastic about the operation.’’ He cedure was efficacious and on which subjects. believed that the operation had a ‘‘usefulness’’ but Efforts to introduce quantitative analysis were that ‘‘the delimitations of its usefulness have not complicated, however, because markers of im- been clarified.’’ Until those limitations became provement, such as hospital discharge, could be better understood, he cautioned that ‘‘...I do not a result of the patient’s family situation [42]. think this is the time to disseminate it widely to Indeed, if any consensus was achieved at the First the profession or to the public...’’ Research Conference on Psychosurgery, it was Unfortunately, Grinker’s pointed-and bal- that a majority of assembled experts urged the anced-criticism of the lobotomy in 1941 had little adoption of a ‘‘...universally accepted rating impact. Although he urged restraint and viewed scale, with patients individualized by diagno- lobotomy as experimental, the procedure was sis or other category...[to] provide an effective widely disseminated as therapy over the next 10 method for both case-selection and evaluation of years. It is estimated that by 1951, 19,000 to operative results’’ [40]. 20,000 Americans were lobotomized, with many In the second edition of his comprehensive of those subjected to the procedure being return- textbook on psychiatric treatment, Diethelm ing veterans [37]. In spite of growing professional noted that the operative procedures under con- uncertainty about the effectiveness of the pro- sideration had ‘‘undergone considerable modifi- cedure and its safety, public sentiment was cation’’ [43]. He viewed these modifications as positive enough to compel the Veterans Admin- evidence of what he described as ‘‘therapeutic istration and state hospitals to introduce the insecurity which should exert a strong critical hesi- procedure after the war [38]. It is estimated that tancy on the part of the clinician who is consid- nearly 3000 returning veterans underwent psycho- ering surgical procedure’’ [43]. surgery [39]. Despite growing controversy about the safety Although Grinker’s early criticism did little to and efficacy of psychosurgery at subsequent na- stem the tide of lobotomy, his well-considered tional conferences [40,44] and concerns about the concerns about methodology, patient selection, side effects of the procedure [45], these concerns outcome data, and the morbidity associated with did not decrease the incidence or popularity of the procedure would be revisited by the First the procedure. Although articulated at confer- Research Conference on Psychosurgery convened ences or in journals, these criticisms had little im- by the National Institute of Mental Health in 1949 pact on the incidence of psychosurgery in the 308 J.J. Fins / Neurosurg Clin N Am 14 (2003) 303–319

United States. Between 1936 and 1946, some 6000 and not on the important issue of efficacy. The patients were operated on in the United States question was not whether lobotomy altered the alone [46]. By the late 1950s, 40,000 to 50,000 status of the patient’s symptoms but whether procedures had been performed in the United the cognitive side effects were justified [49]. Nolan States, with some 4000 estimated by Freeman to Lewis, Director of the New York Psychiatric have been done or directed by him in 30 hospitals Institute, asked whether ‘‘the quieting of a patient’’ in 15 states nationwide [47]. Rosemary Kennedy, was indeed a ‘‘cure’’ and worried about the num- the sister of President Kennedy, was among those ber of ‘‘zombies’’ and ‘‘mental invalids’’ produced operated on. as a consequence of the procedure [50]. The popularity of the procedure needs to be The decline of psychosurgery was not promp- contextualized against the dismal condition of ted by ethical concerns but rather by the advent of state mental hospitals and the absolute lack of modern psychopharmacology and effective ther- effective treatment for the persistently and chron- apy for psychoses with the major tranquilizers, ically mentally ill. One representative report from such as chlorpromazine in 1954. Indeed, two com- a state hospital during that era describes lobot- mentators have opined that even with the grow- omy as a ‘‘fruitful method in the treatment of ing backlash against psychosurgery in the early chronic mental illness,’’ with 37.4% of patients 1950s, the lack of an effective alternative would being able to be discharged from hospital after the have been sufficient to keep the procedure in procedure [48]. The authors of this study, which ‘‘common use’’ [51]. mostly involved patients with schizophrenia, maintained that: Psychosurgery and the body politic ...While there are many limitations and Not even chlorpromazine was enough of an failures with this treatment, they are over- advance to remove psychosurgery entirely from shadowed by the generally satisfactory and therapeutic consideration in the 1960s and early not infrequently brilliant results. It is still difficult to assess the extent to which 1970s. Some physicians, such as the Harvard technical difficulties or inherent resistance neurosurgeon H. Thomas Ballantine, Jr, main- of the disease process contribute to opera- tained that psychosurgical procedures like cin- tive failures. However, it is truly gratifying gulotomy retained a role in conjunction with to observe a patient who was previously standard psychiatric care for refractory patients. a tremendous problem in management- He articulated guidelines to regulate the judicious secluded, untidy, aggressive, destructive, use of the procedure for the relief of the patient’s and combative to the point of requiring suffering and improvement of functioning in half a dozen attendants to carry out the society. Procedures were to be reserved for pa- basic necessities of personal care-become tients who failed all other methods of treatment. a quiet cooperative individual, taking pride in her personal appearance, assisting with Decisions to operate were to be made in conjunc- various ward tasks, and finally returning to tion with a psychiatrist, who would also make her own home. Possibly only those who psychiatric follow-up available, and patients and have served in a state hospital can appre- family were to be informed of potential risks and ciate fully these problems. The comments benefits. Most critically, he condemned the use we have heard occasionally to the effect that of psychosurgery for political or social purposes, postlobotomy patients lack judgment, and articulating instead a solely patient-centered cannot, for example, plan a meal or play rationale for the procedure [52]. a good game of bridge seem to lack It was the social uses of psychosurgery for satisfactory perspective when one regards what was called behavior control that, which the level from which these individuals were plucked...Lobotomy is not a cure-all but it Ballantine and others condemned, that caused can well be regarded as an encouraging a furor during that era. As distinct from its earlier therapeutic weapon for a very malignant iteration as a means to address a patient’s de- disease [48]. pression or schizophrenia, this more modern dimension of psychosurgery sought to modify Although studies like this one failed to take behavior. Amid the social turmoil of that era, into account a potential placebo effect or the sociobiologists began to suggest that psychosur- possibility of spontaneous remission, criticism of gery might have a role in addressing problems like lobotomy hinged on the question of side effects violence or civil unrest. It is perhaps hard to J.J. Fins / Neurosurg Clin N Am 14 (2003) 303–319 309 imagine today that there could be serious con- Delgado analogized cerebral pacemaking to cerns about such futuristic attempts at social the growing role of cardiac pacemakers as a means control, but they were quite real. One news report to suggest the utility of brain pacemaking in the in the Medical News column in the Journal of the future. He acknowledged concerns about the American Medical Association sought to reassure ethical implications of his work and urged ‘‘in- the wary reader that ‘‘Logistically, psychosurgery telligent collaboration of the best minds’’ to for social control is highly unlikely, simply be- address the field’s ‘‘fundamental medical, social, cause there are not enough neurosurgeons’’ [53]. and even philosophical implications.’’ Nonethe- This second period in the history of psycho- less, he urged continued scientific progress, noting surgery could be said to have begun through the that ‘‘We are certainly facing ethical, philosophic, work of Jose M.R. Delgado. Coupling psycho- and practical problems not exempt from risks, surgery with burgeoning efforts in computer tech- but we should also expect important medical nology and solid state electronics, Delgado application of the new methods to epilepsy, advanced the idea of ‘‘psychocivilizing society’’ intractable pain, involuntary movements, and using an implantable brain implant that could be mental disorders’’ [62]. His position might be operated by remote control [54]. Delgado came to best summed up by his observation that ‘‘Fears international attention in 1965 when he returned have been expressed that this new technology to his native Spain for a now famous publicity brings with it the threat of possible unwanted and stunt in which he demonstrated the potential of ethical remote control of the cerebral activities his work by stopping a charging bull in Cordoba’s of man by other men, but this danger is quite bullring using a ‘‘stimociever’’ he had developed improbable and is outweighed by the expected [55]. clinical and scientific usefulness of the method’’ Delgado’s work raised concerns about the [63]. possibilities of mind control, and his legacy Leading proponents of psychosurgery for the remains a lingering question for the current era control of violence were Frank R. Ervin, a psy- of neuromodulation. Recent reports in the lay chiatrist at the Neuropsychiatric Institute at the press describing a remotely controlled ‘‘cyborg’’ University of California at Los Angeles, and rat with a brain implant [56] alluded to Delgado’s Vernon H. Mark, a Harvard neurosurgeon, who work [57,58], thus resurfacing the question of together coauthored Violence and the Brain [64]. mind control, which had explosive political Much of their work hinged on seeking to dem- consequences when first introduced. onstrate the relation between organic brain dis- A physician and physiologist working in the orders, such as temporal lobe epilepsy (TLE), and Department of Psychiatry at Yale University, violent or aggressive behavior. In one early case, Delgado studied aggression in primates and then they were able to demonstrate a left tempo- manipulated their response through the use of ral horn lesion by means of a pneumoencepha- implantable electrodes, arguing that ‘‘a better logram in a young woman with TLE in whom understanding of the neurophysiological mecha- violent outbursts were inducible using implantable nisms responsible for aggressive and destructive electrodes and telemetric equipment supplied reactions may provide man with greater capacity ‘‘through the courtesy and assistance of Dr to educate and direct his own behavior’’ [59]. Delgado’’ [65]. Delgado, some of whose work was funded by the Echoing Delgado’s notion of ‘‘greater personal United States Public Health Service, the Office of freedom,’’ Mark maintained that ‘‘I believe the Naval Research, and the Department of the Air correction of that organic condition gives the pa- Force [60,61], argued that society was on the cusp tient more rather than less, control over his own of a new era in which the human mind could behavior. It enhances, and does not diminish, his influence its own evolution through the use of dignity. It adds to, and does not detract from, his technology. Using notions of self-dominion, he en- human qualities’’ [66]. Although many of Mark’s visioned an escape from the blind chance of nor- aspirations for psychosurgery seem overly optimal evolution to one where man and technology mistic, he did foreshadow developments in psy- would alter human history, ultimately leading to chiatry, moving that field from being dominated a‘‘...future man with greater personal freedom by psychoanalysis and ‘‘political psychiatrists’’ and originality, a member of a psychocivilized toward those having an interest in the organic society, happier, less destructive, and better basis of disease [67]. balanced than present man’’ [54]. He wrote about the ‘‘absurd split’’ and the: 310 J.J. Fins / Neurosurg Clin N Am 14 (2003) 303–319

...historical dichotomy between Ôpurely and psychosurgery is an ethical, political and organic’ and Ôpurely social’ abnormalities. spiritual crime. It should be made illegal’’ [75]. Specifically, physicians tend to categorize Breggin’s accusations seem more suited to the a few abnormal behaviors, such as paralysis, clinical work of Freeman than to the positions blindness and dementia as neurological taken by Mark, who was opposed to any social problems. At the same time, certain other application of his work. In a talk delivered at abnormalities, such as depression and ag- the Hastings Center’s Institute of Society, Ethics gression, have found a hard niche within the and Life Science’s Working Group on Behavior domain of the psychiatrists, sociologists and Control, he sought to refute the charge of racism criminologists. Many of them view these lodged against the neurosurgical treatment of behaviors as nothing but the reflections of particular environments. They tend to be- violent epileptics [76]. He asked, ‘‘Does the theory lieve that brain function or dysfunction is that some violence is caused by brain disease lead not an important determinant of abnormal us to expect it is a characteristic mainly of black behavior [68]. people? Certainly not. On the other hand the theory that personal violence is caused exclusively As much as his work in psychosurgery, these by social conditions might very well lead us to views seemed to have engendered a backlash from look at the black ghettos. The environmental cues the more purely socially oriented practitioners to personal violence may very well cluster around who saw psychiatry in political or sociologic terms racially differentiated areas’’ [76]. Citing the [69]. This schism made Mark the target of what he prevalence of domestic violence across all demo- described as an ‘‘anti-psychiatry campaign’’ [70]. graphics, both rich and poor as well as black and Reacting to a brief letter he wrote with colleagues white, he asserted that ‘‘From this perspective, to the Journal of the American Medical Association claret is the predominant color [of violence], not in 1967 on the potential relation between brain black or white’’ [76]. He would consistently main- disease and urban riots [71], Mark was accused of tain that violence was colorblind in both domes- racism by proponents of social psychiatry [72]. tic and international contexts [77]. This assault on psychosurgery was led by Peter Indeed, he was one of the first to advocate an Roger Breggin, a Washington social psychiatrist. ‘‘integrated approach’’ to psychosurgical care. At A self-described political conservative and civil Boston City Hospital, where he was Director of libertarian [73], Breggin sought to characterize Neurosurgery, he sought to overcome the di- psychotherapy and on a continuum of a totalitar- chotomous practices of his neurologic and psy- ian-libertarian axis in which a ‘‘high degree of chiatric colleagues and to develop a ‘‘holistic’’ autonomy and personal freedom characterizes therapy in which ‘‘...treatment of a patient should more libertarian therapies’’ [74]. In this context, involve not only his brain but his family, living public psychiatric hospitals and psychosurgery conditions, job and role in society. It is very were seen as vectors of social control and described important, therefore, to imbed a neurological as ‘‘custodial concentration camps’’ and ‘‘powerful diagnosis of problems of violence into a larger totalitarian technologies,’’ respectively. integrated approach to human behavior.’’ Along In Congressional hearings before the Sub- these lines of comprehensive care, he also advo- committee on Health of the Senate Committee cated a ‘‘committee of some sort’’ that would over- on Labor and Public Welfare, Breggin charged see consent while not accepting patients who do that psychosurgeons were unethical because of not want therapy. how they obtained consent for the procedures In retrospect, Mark’s attempt to address they performed. He was deeply suspicious of how organic illness with due attention to the patient’s psychosurgeons sought to regulate their activities voluntary consent comes across as moderate during through review committees that relied on ‘‘pro- immoderate times prone to hyperbole, overreac- fessional ethics and medical control’’ to maintain tion, and suspicion. There was a prevailing sense physician control of the situation. Ultimately, of alarm that psychosurgery was being broadly Breggin said, ‘‘It creates for themselves an elitist applied in law enforcement. Contemporary ac- power over the human mind and spirit. If America counts of that era observed that one of the ever falls to totalitarianism, the dictator will be ‘‘appeals’’ of psychosurgery was viewed as the a behavioral scientist and the secret police will be willingness of law enforcement agencies to embrace armed with lobotomy and psychosurgery. And by this technology for the purpose of addressing the way, lobotomy is still with us...Lobotomy seemingly intractable problems. One international J.J. Fins / Neurosurg Clin N Am 14 (2003) 303–319 311 perspective from a New Zealander behavioral ment for racist repression’’ [80]. The National Com- scientist observed that: mission report would later confirm this finding. In a review of 600 psychosurgery procedures Still another cultural aspect which adds performed in 1974, only six Hispanics and one to the appeal of psychosurgery is the black patient were identified. The Commission readiness of the American government to noted that ‘‘The fact that so few patients from seek solutions to its domestic problems minority groups have undergone psychosur- which hide the causes of the problems. Psychosurgery can be one such solution if gery...is due not to discrimination on the part used to ‘‘cure’’ the nation’s social problems. of surgeons but to the economic realities and pub- The Justice Department and several state lic policy’’ [81]. departments of correction (notably Califor- Furthermore, although minority communities nia’s) have shown great interest both in were indeed fearful of these procedures, not all establishing that black rioters and aggres- were in opposition. Two black neurosurgeons sive inmates are suffering from brain speaking at the 1976 National Minority Confer- dysfunction and in curing them through ence on Human Experimentation in Reston, psychosurgery. Three leading proponents of Virginia, urged limited use of psychosurgery with psychosurgery have advanced the view that appropriate oversight and review boards [82]. Dr many of those involved in ghetto rebellions acted violently because of brain dysfunc- Jesse Barber, Chief of Neurosurgery at Howard tion. The accumulated results of racism and University, observed, ‘‘I personally feel guilty poverty were discounted as causal factors about not developing a program (of psychosur- since not everyone rioted [78]. gery) at Howard University.’’ He added that ‘‘When it was considered we were reluctant to Although this citation would suggest wide- face the opposition and destroy our image in the spread use of psychosurgery for social control, the black community.’’ data suggest that it was otherwise. A 1974 study Although there was little evidence for the use conducted by the Behavioral Control Research of psychosurgery for law enforcement purposes, Group of the Hastings Institute surveyed the the issue of behavior control dominated the delib- Commissioners of Corrections for all 50 states to erations of advocates of all stripes whether they ascertain the prevalence of behavior control in the were physicians, philosophers, attorneys, or nas- nation’s prisons [79]. Forty-seven states and the cent bioethicists. In the political climate of that District of Columbia responded. If behavior modi- era, there was, it seemed, a fine line between scien- fication was done at all, the least coercive treat- tific fact and science fiction [83]. In contrast to ments, such as group therapy or token economy the media’s favorable—and distorted—depiction systems, were employed. None of the respon- of psychosurgery in the early years of lobot- dents used psychosurgery as a treatment proce- omy [84], the popular culture of the 1970s height- dure, although one added, ‘‘not at this time.’’ ened fears of abuse. Works like Crichton’s The Although these data would suggest that things Terminal Man depicted the ‘‘treatment’’ of a vio- were better than feared, the author of the report lent criminal with implantation of electrodes. The cautioned the reader to be wary of the results. setting for the procedure was a fictitious neuro- Perhaps reflecting the distrust prevalent during psychiatric institute located in Los Angeles, the end of the Nixon administration and the which, not so coincidentally, borrowed the name Watergate scandal, she notes that ‘‘In dealing with of Ervin’s own institution [85]. prisons, however, the publicly announced pro- In this broader cultural context, fears of psy- grams may be only a small percentage of what is chosurgery transcended the internecine battles actually occurring’’ [79]. of psychiatric subdisciplines, as exemplified by This prevalence data concerning psychosur- Breggin’s less than restrained ideologic attacks. gery for nonmedical purposes was corroborated More balanced criticism came from individuals by a study conducted by the American Psychiatric like Edward Mearns, a professor of law at Association Task Force on Psychosurgery and by Case Western Reserve University, who observed the work of the National Commission. As the that: American Psychiatric Association report put it, ‘‘Both reports concluded that there is no reliable There is something promising about the evidence that psychosurgery has been used for notion that the effort to cure sick individ- political purposes, social control or as an instru- uals may result in considerable social 312 J.J. Fins / Neurosurg Clin N Am 14 (2003) 303–319

benefit as well. But there is something court went on to cite the entirety of the Nurem- disturbing about the conscious effort to berg Code to argue that reasoned and voluntary use medicine and medical men institutions consent in prison was so greatly impaired as to to cure a ‘‘sick’’ society. For it is one thing make lawful consent in that setting impossible to use medicine as an instrument of healing [92,93]. Indeed, one of his attorneys, Professor and quite another to use it as an instrument Robert A. Burt, pointed out that the prisoner for social control [86]. himself illustrated the coercive nature of incarcer- The philosopher Robert Neville, who directed ation when his confinement was declared uncon- the Hastings Center Working Group on Psycho- stitutional and he had the opportunity to revisit surgery, echoed this sentiment. Neville asserted his decision after initially consenting to psycho- that ‘‘The ethics of good medicine isn’t easily surgery. When these new developments allowed transferred to good social control, so there is a him to imagine being set free, he suspended his danger of having the appearance of therapy, consent to reconsider surgery given his new cir- when the real purpose is punishment’’ [87]. Hal cumstances [94]. Edgar of Columbia Law School pointed out that although procedures on the brain are justifiable if Regulatory bioethics: the National they result in a cure of a physical illness like Commission’s report on psychosurgery Parkinson’s disease without undue side effects, it becomes more complex when the focus of the Elliot S. Valenstein, commenting on this intervention is mental illness which is ‘‘...a period, observed that ‘‘The coalition of civil- concept which is heavily influenced by social rights, anti-psychiatry, and minority groups op- norms, and a label which is often imposed in posed to psychosurgery proved to be much more a particular case because people engage in effective politically than the earlier opposition ÔstrangeÕ behavior of one sort or another’’ [88]. to psychosurgery from within the medical profes- Echoing the sentiments of the times about sion had been’’ [95]. Indeed, public sentiment, questioning authority, he asked, ‘‘...how does catalyzed in part by an organized and growing one know whether the sick are being cured or scholarly bioethics movement, led the National whether medicine is being used as yet another tool Commission for the Protection of Human Sub- in society’s ever present effort to secure comfort jects of Biomedical and Behavioral Research to through conformity’’ [88]. issue a report on psychosurgery in 1977 [81]. As Much of the resistance to psychosurgery—and such, the report was a political compromise in lieu the broader issue of behavior control—was closely of a moratorium or an outright ban as proposed, related to emerging concerns about the use of the most prominently, by Senator J. Glenn Beall, a procedure on prisoners, who might be coerced or Republican from Maryland [96]. unable to give appropriate consent to experimen- The National Commission was created by the tal procedures [89]. Some of this was related to National Research Act of 1974 [97] in the wake of a fear of somehow legitimizing a deeply flawed revelations about the Tuskegee Syphilis Study and penal system through an affiliation with medicine other research ethics abuses. The National Com- [90], but much of it was a central concern about mission was constituted and staffed by a number the ability of inmates to give voluntary and of prominent bioethicists, and their work was informed consent. prolific. They issued a number of reports in ad- This issue came to national prominence in the dition to the one on the ethics of psychosurgery, 1973 case of Kaimowitz v. Department of Mental although they were specifically mandated to Health, in which a three-judge panel in Michigan address this issue by Congress. Concerns about sought to determine whether a confined prisoner psychosurgery were prominent in testimony lead- could voluntarily consent to an experimental ing up to the passage of the act. Willard Gaylin, procedure that might temper his aggressive and President of the Hastings Center, warned of the criminal behavior [91]. The court opined that dangers of psychosurgery during 1973 Senate there was no scientific basis to suggest that hearings [98]. psychosurgery would be therapeutic in the ab- It is somewhat ironic, given the public origins sence of a discernible disorder like epilepsy and of the report, that it failed to reach the conclu- that the risk-benefit ratio was disproportionate sion that most of the public wanted. Instead of given the current state of knowledge and the adopting the more politically correct position, known risks of the procedure. More critically, the to invoke a more modern phrase, the National J.J. Fins / Neurosurg Clin N Am 14 (2003) 303–319 313

Commission found sufficient evidence of efficacy them...The inconsistency results from char- of the modern procedures like cingulotomy and acterizing psychosurgery at either point in not the prefrontal leukotomies of Freeman’s time as solely therapy or solely experimen- day to endorse continued experimental work in tation. It is clearly both [86]. psychosurgery as long as strict regulatory guide- The ambiguity over psychosurgery’s status as lines and limitations were in place. As Chairman experimental or therapy is also implicit in the of the National Commission, J. Kenneth Ryan of report. Psychosurgery is generally viewed as in- Harvard Medical School told a Science corre- vestigational, although the National Commission spondent: does acknowledge that some patients had been helped by these interventions. We looked at the data and saw they did not support our prejudices. I, for one, did For the purposes of the report, psychosurgery not expect to come out in favor of psy- was defined as surgery whose ‘‘primary object of chosurgery. But we saw that some very sick the performance...is to control, change, or affect people had been helped by it, and that it did any behavioral or emotional disturbance...’’ and not destroy their intelligence or rob them of included both classic psychosurgeries like ablation their feelings. Their marriages were intact. as well as its more modern iteration of electric They were able to work. The operation stimulation: ‘‘Psychosurgery includes the implan- shouldn’t be banned [96]. tation of electrodes, destruction or direct stimu- The National Commission’s view of psycho- lation of the brain by any means...’’ [81]. It is surgery was tempered, however, by the endorse- interesting to note that brain surgery for the ment of stringent regulatory guidelines and treatment of movement disorders like Parkinson’s procedural safeguards that would prohibit psy- disease or for epilepsy and pain management was chosurgery for anything other than a patient- excluded from this definition. centered application: Given the investigational nature of psychosurgery, the National Commission recommended The Commission affirms that the use of strict oversight by specially constituted institu- psychosurgery for any purpose other than to tional review boards (IRBs) with a subcommittee provide treatment to individual patients would of Department of Health, Education, and Welfare be inappropriate and should be prohibited. (DHEW)-sanctioned experts or consultants in Accordingly, the Commission is recommending safeguards that should prevent neurology, neurosurgery, psychology, and psychi- the performance of psychosurgery for pur- atry to review the technical aspects of the surgery. poses of social or institutional control or IRB review was recommended until the ‘‘safety other such misuse [81]. and efficacy of any psychosurgical procedure have been demonstrated.’’ The IRB was charged with Furthermore, although the National Commis- assessing the competence of the surgeon, the sion decided not to recommend a ban of psycho- appropriateness of the procedure for a selected surgery, it was careful to note that it did not patient, and the adequacy of the patient’s in- recognize psychosurgery as ‘‘accepted practice.’’ formed consent. In addition, harkening back to This addressed a lingering question since the historical concerns about the evaluation of pro- 1940s, posed most eloquently by Professor Mearns cedural efficacy, the National Commission recom- of Case Western Reserve Law School in 1975: mended that the IRB ensure adequate pre- and ...Moreover, psychosurgeons do not postoperative evaluations. submit research protocols to review com- Addressing concerns about regulatory over- mittees for another reason. They simply do sight, informed consent, and the risks of coercion not perceive their operations to be experi- for the voluntarily institutionalized psychiatric ments. They view them as therapy...They inpatient, the National Commission recommended see the patient as ill with a condition that the establishment of a National Psychosurgery requires specific treatment geared to his Advisory Board that would determine whether the particular illness. This need for treatment suggests speed or at least the avoidance of ‘‘specific psychosurgical procedure has demon- cumbersome review procedures... Some- strated benefit for the treatment of the psychiatric how, the inconsistency of characterizing symptom or disorder of the patient.’’ If the pro- psychosurgery as therapy at the time it is cedure was part of a research study, the advisory performed and as experimentation at the board would determine whether enrollment was time of publication seems to escape in compliance with the National Commission’s 314 J.J. Fins / Neurosurg Clin N Am 14 (2003) 303–319 recommendations on research involving the determined to ban psychosurgery for prisoners, institutionalized mentally infirmed [81,99]. children, the involuntarily confined mentally ill, The report set even more stringent guidelines and those who were decisionally or legally incom- for prisoners and those who were involuntarily petent and unable to provide consent. Further- committed, had a legal guardian, or were believed more, he endorsed the establishment of a joint to be unable to give informed consent because of committee on psychosurgery, composed of leading impaired decision-making capacity. Attempting to professional organizations, ‘‘to establish mech- balance ‘‘access to potentially beneficial therapy’’ anisms for the voluntary regulation and reporting against the coercion or a breach of voluntariness, of psychosurgical procedures.’’ This was a position the National Commission outlined a complex reg- he later reversed. ulatory approach requiring, among other stric- Despite Califano’s narrowing of the National tures, that the proposed National Psychosurgery Commission’s broader recommendations regard- Advisory Board determine whether the specified ing whom psychosurgery might be performed on, psychosurgicalprocedurehadademonstrablebene- he and the report were still criticized liberally. One fit for the patient’s condition and if the operation of the more strident columns was by William were to be performed as an element of a research Raspberry of The Washington Post. In an Op-Ed protocol, whether these efforts would be in com- written after Califano’s determination, he said pliance with the recommendations on research that he wished the Commission had called involving the institutionalized mentally infirmed psychosurgery ‘‘...by its right name: making holes [81]. in people’s heads and slicing their brains in a hit- To engage in some rudimentary health services or-miss attempt to make them behave them- research, the National Commission also suggested selves’’ [101]. He criticized Califano’s about-face the establishment of a national data collection ontheNationalPsychosurgeryAdvisoryBoardand registry to assess the safety and efficacy of these urged the ‘‘immediate cessation of butchery-in- procedures. This fundamental issue remained an the name of medicine.’’ He concluded by suggest- open question given the small sample of cases re- ing that someone take Califano to see ‘‘One Flew viewed by National Commission consultants. In Over the Cuckoo’s Nest’’ [101]. Despite this re- addition to this data, the National Commission sponse, progressive commentators like George recommended that this registry note the indica- Annas viewed the report, as submitted to the Sec- tions for procedures and the demographics of retary, as balanced. He observed that ‘‘Certain pro- those who underwent psychosurgery. visions are unlikely to please either avid promoters In addition, with one abstention, the National of psychosurgery or those favoring a complete Commission actually encouraged (their word) the ban; nonetheless, it is a reasonable response to Secretary of the DHEW ‘‘...to conduct and a highly complex problem, and its basic ap- support studies to evaluate the safety of specific proach is likely to gain general acceptance’’ psychosurgical procedures and the efficacy of such [102]. procedures in relieving specific psychiatric symp- Professor Annas was right to appreciate the toms and disorders, provided that the psychosur- complexity of these issues and the report’s gery is performed in accordance with these balanced approach to them. He was, however, recommendations’’ [81]. overly optimistic that the National Commission’s Finally, the National Commission recommen- recommendations would eventually be accepted. ded to Congress the imposition of strict sanctions Three years after the report was issued, Califano’s and the threat of Federal defunding if these successor, Health and Human Services (HHS) recommendations were violated and the exclusion Secretary Patricia Harris, still had failed to make of Federal agencies from psychosurgery funding a determination over proposed regulations [103]. ‘‘unless such agencies or components are primar- Today, we still have failed to reach a national ily concerned with health care or the conduct of consensus on research on subjects whose medical, biomedical and behavioral research’’ [81]. psychiatric, or neurologic conditions make it Not all the National Commission’s recommen- impossible for them to provide voluntary consent. dations were accepted by DHEW Secretary The 1998 National Bioethics Advisory Com- Joseph Califano in his determination [100]. He mission report on ‘‘Research Involving Persons decided to promulgate regulations that would With Mental Disorders That May Affect Deci- limit psychosurgery to those individuals who could sionmaking Capacity’’ attempted to address as- provide informed and voluntary consent. Thus, he pects of this question, but it too remains in J.J. Fins / Neurosurg Clin N Am 14 (2003) 303–319 315 bureaucratic limbo [21,104,105]. The proposed context of OCD. OCD has already been the object Psychosurgery National Advisory Board would of sophisticated psychosurgical interventions, have been empowered to adjudicate these morally most notably the cingulotomy as currently per- complex decisions in the context of psychosurgery formed by Cosgrove’s group [110] and the closed or neuromodulation. Because this body never radioablation techniques employing the gamma came into being, we have been left without both knife as studied by Greenberg et al [111]. a regulatory framework and a national locus for If we begin with definitional issues, we will a debate about the ethical implications of neuro- immediately realize that the distinction between modulation. somatic and psychiatric interventions is no longer Recent history suggests that there remains as neatly sequestered as in the report. Increas- a need for such a dialogue as well as a national ingly, the line between mind and brain—and body empowered to review the science behind neurology and psychiatry—has become blurred, proposed clinical trials to ensure that our ethical and it becomes increasingly problematic to treat sensibilities are not offended. In the United States, one brain condition differently than another. For there is no such body to provide this oversight or example, although the National Commission dis- catalyze this discussion. The most evolved forums tinguished treatment of Parkinson’s disease from for such deliberations are local bodies like the psychosurgery, we now know that treatment of Cingulotomy Committee at the Massachusetts the movement disorders of Parkinson’s disease General Hospital, which provides oversight for with deep brain stimulation can affect the therapeutic interventions for patients who are able patient’s emotions. As was described in the New to provide consent themselves [106]. Professional England Journal of Medicine, French investigators groups are also evolving to articulate ethical have demonstrated that deep brain stimulation principles for the conduct of neuromodulation has the unrecognized potential to alter a mood research as in the case of the Obsessive-Compul- state and induce a reversible but acute depression sive Disorder Working Group [107]. [112]. What is the boundary between modulating In France, the neurosurgeon Alim Benabid a movement disorder and potentially altering voluntarily sought the approval of the French emotions? Is one intervention a neurologic pro- National Bioethics Commission for approval of cedure and the other psychosurgery, or are both a clinical trial of deep brain stimulation in re- neuromodulation? fractory OCD, the first of many psychiatric This overlap phenomenon is especially confus- maladies for which neuromodulation is likely ing if the mechanisms of Parkinson’s disease are to be investigated [108,109]. Benabid’s actions compared with those of severe OCD. They share suggests a continued need for extramural over- mechanistic similarities and are characterized by sight of these still contentious interventions along hypersynchronous activity. The pathophysiology the lines of the proposed National Psychosurgery and neural circuits of both diseases share cortico- Advisory Body. basal gangliothalamic interactions [18] and both could be construed as tremors. In the case of Parkinson’s disease, it is a motor tremor. In OCD, Summary: fairness, palliation, it is a limbic-thought tremor. and psychosurgery Although this analogy is not airtight, it is Moving forward, we need to create regulatory reminiscent of Vernon Mark’s prescient anticipa- mechanisms that will balance human subject tion of a time when the mind-brain dichotomy protections for individuals with intractable neu- would be less tenable as an ethical demarca- rologic and psychiatric disorders against scien- tion, justifying treatment for one disorder but tific progress and access to potentially beneficial proscribing another. This categorization should interventions. This fiduciary obligation of practi- be disconcerting to those who embrace equity in tioners, clinical investigators, and public policy the treatment of mental and physical disease as makers can be facilitated by building on the yet a value, especially when the distinction is pre- uncompleted efforts of the 1977 National Com- dicated on a Cartesian dualism unsubstantiated mission Report on Psychosurgery. This report by modern neuroscience. This regard for fair- should serve as a foundation on which additional ness becomes even more troubling if we recall analysis could be built. Surgeon General David Satcher’s call for parity In the spirit of this evolution, I would like in the treatment of physical and mental illness to anticipate some of this analytic work in the [113]. 316 J.J. Fins / Neurosurg Clin N Am 14 (2003) 303–319

A second analytic issue, in my view, will be to entirely unexpected because of the downstream move beyond viewing psychosurgery or neuro- effects of the stimulation target, the frontal lobe’s modulation as either research or therapy and cingulum gyrus, which is a site often ablated in the consider it instead as a form of palliative care. The classic prefrontal leukotomy of Moniz and ethos of pain and symptom management articu- Freeman. These affective changes can range from lated in the palliative care community is appli- euthymia to varying degrees of blunting or hypo- cable to this work. These interventions are not mania. Either extreme should be avoided, espe- directed toward cure but rather at the masking of cially the blunting end of the spectrum, which is intractable symptoms [114] and the alleviation of reminiscent of Nolan Lewis’s concerns about suffering, a strategy consistent with palliative care the ‘‘the quieting of a patient’’ after the primitive [115]. lobotomy [50]. I believe the National Commission may have These affective changes may unsettle some anticipated the relationship between psychosur- observers. If the past is a prologue, we might gery and neuromodulation and palliation in its expect new charges of mind control and a call for willingness to endorse surrogate consent. Justifi- the prohibition of this new technology. If history’s cation for surrogate authorization may have been lessons are heard, however, we will appreciate that linked to an implicit view of psychosurgery as a more fruitful response is articulating an ethic of palliation when it was being applied to intractable responsibility that palliates the suffering of those psychiatric suffering. Viewing the relief of suffer- with intractable psychiatric illness. As Willard ing as an ethical mandate, the National Commis- Gaylin told us a quarter of a century ago, ‘‘To sion may have been willing to allow surrogates to be afraid of our technology is to be afraid of our- provide authorization even when the intervention selves. It is only essential that we protect ourselves was yet unproven to be therapeutic. Whether here, as everywhere, from arrogance and insen- notions of palliation entered into the National sitivity. The answer is not to prohibit technology Commission’s balancing and specification of ethi- but to insist that it always be subservient to the cal principles is a question that merits additional transcending values of human worth and human study. Although the recommendation regard- dignity’’ [116]. ing surrogate consent was reversed by Secretary Califano, the National Commission’s thinking might suggest an alternative analytic framework Acknowledgments than that offered by the more restrictive National The author thanks Nicholas D. Schiff, Benja- Bioethics Advisory Commission report [106]. min D. Greenberg, Franklin G. Miller, G. Rees The analogy to palliative care may also be Cosgrove, and Amy B. Ehrlich for their helpful helpful in fostering a collaborative and patient- insights, Diane Richardson of the Oskar Diethelm centered model of care. In this regard, I offer an Library of the Payne Whitney Clinic and Chris analogy to patient-controlled analgesia (PCA). In McKee of the Hastings Center Library for their PCA, the patient is given the ability to determine archival assistance, and Ali R. Rezai and Alim the frequency of dosing of pain medication to Benabid for the opportunity to present earlier address his/her perception of pain and level of versions of this paper to the neuromodulation suffering within preset and safe limits. This pre- meetings held in Aix-Les-Bain, France, in January cedent for joint patient and physician control has and June 2002. important implications for deep brain stimulators given the history of ‘‘behavior control.’’ The still References present fear of mind control by the modern equivalent of Delgado’s stimoceiver makes it [1] Moniz E. How I came to perform prefrontal critical that patients have an appropriate oppor- leucotomy. In: Proceedings of the First Inter- tunity to adjust the frequencies of their stimula- national Congress of Psychosurgery. Lisbon; 1949. tors within safe and monitored parameters. This p. 15–21. [Cited in: Tierney AJ. Egas Moniz and the origins of psychosurgery: a review commem- will allow for a sharing of power and help ame- orating the 50th anniversary of Moniz’s Nobel liorate fears of co-optation of personhood. Prize. J Hist Neurosci 2000;9(1):22–36.] This is more than a hypothetic issue, because [2] Osler W. Teacher and student. Aequanimitas and investigators have begun to recognize that sub- other addresses. London: HK Lewis; 1904. p. 21–43. jects with OCD have affective changes depending [3] Medical News. Congress weighs 2 measures. on the setting of their stimulator. This is not JAMA 1973;225(9):1044–6. J.J. Fins / Neurosurg Clin N Am 14 (2003) 303–319 317

[4] Roth RM, Flashman LA, Saykin AJ, Roberts [22] Raju TNK. The Nobel chronicles. 1949: Walter DW. Deep brain stimulation in neuropsychiatric Rudolf Hess (1881–1973) and Antonio Egas Moniz disorders. Curr Psychiatr Rep 2001;3:366–72. (1874–1955). Lancet 1999;353(9160):1281. [5] Rapoport JL, Inoff-Germain G. Medical and sur- [23] Damasio AR. Egas Moniz, pioneer of angiogra- gical treatment of obsessive-compulsive disorder. phy and leucotomy. Mt Sinai J Med 1975;42(6): Neurol Clin 1997;15(2):421–8. 502–13. [6] Schiff ND, Pulver M. Does vestibular stimulation [24] Moniz E. Prefrontal leucotomy in the treatment activate thalamocortical mechanisms that reinte- of mental disorders, 1937. Am J Psychiatry 1994; grate impaired cortical regions? Proc R Soc 151(Suppl):236–9. London B Biol Sci 1999;266(1417):421–3. [25] Jones JH. Bad blood, the Tuskegee Syphilis ex- [7] Schiff ND, Plum F, Rezai AR. Developing periment. New York: The Free Press; 1993. prosthetics to treat cognitive disabilities resulting [26] Ligon BL. The mystery of angiography and the from acquired brain injuries. Neurol Res 2002; ‘‘unawarded’’ Nobel Prize: Egas Moniz and 24(2):116–24. Hans Christian Jacobaeus. Neurosurgery 1998; [8] Fins JJ. A proposed ethical framework for inter- 43(3):602–11. ventional cognitive neuroscience: a consideration [27] Jasper HH. A historical perspective: the rise and of deep brain stimulation in impaired conscious- fall of prefrontal lobotomy. Adv Neurol 1995;66: ness. Neurol Res 2000;22:273–8. 97–114. [9] Carmichael M. Healthy shocks to the head. [28] Tierney AJ. Egas Moniz and the origins of Newsweek June 24, 2000. p. 56–8. psychosurgery: a review commemorating the 50th [10] The future of mind control. The Economist. May anniversary of Moniz’s Nobel Prize. J Hist Neuro- 25, 2002. sci 2000;9(1):22–36. [11] Safire W. The but-what-if factor. The New York [29] Diefenbach GJ, Diefenbach D, Baumeister A, Times May 16, 2002. p. A-25. West M. Portrayal of lobotomy in the popular [12] El-Hai J. The lobotomist. The Washington Post press: 1935–1960. J Hist Neurosci 1999;8(1):60–9. Magazine February 4, 2001. p. 16–31. [30] Diethelm O. An historical review of somatic [13] Herbert W. Psychosurgery redux. US News & therapy in psychiatry. Am J Psychiatry 1939;95(4): World Report November 3, 1997. p. 63–4. 1165–79. [14] Valenstein ES. Great and desperate cures: the rise [31] Trials of war criminals before the Nuremberg and decline of psychosurgery and other radical military tribunals under Control Council Law treatments for mental illness. New York: Basic No. 10, vol. 2. Washington (DC): US Government Books; 1986. Printing Office; 1949. p. 181–2. [15] Diering SL, Bell WO. Functional neurosurgery [32] Romano J. Reminiscences: 1938 and since. Am J for psychiatric disorders: a historical perspective. Psychiatry 1990;147(6):785–92. Stereotact Funct Neurosurg 1991;57:175–94. [33] Moniz E. Prefrontal leucotomy in the treatment [16] Marino JR, Cosgrove GR. Neurosurgical treat- of mental disorders. 1937. Am J Psychiatry 1994; ment of psychiatric illness. Psychiatr Clin North 151(6 Suppl):236–9. Am 1997;20(4):933–43. [34] Weinberger LW. Lobotomy: a personal memoir. [17] Koppell BH, Rezai AR. The continuing evolution Pharos 1998;Spring:17–8. of psychiatric neurosurgery. CNS Spectrums 2000; [35] Panel Discussion at Cleveland Session. Neurosur- 5(10):20–31. gical treatment of certain abnormal mental states. [18] Montgomery EB Jr. Deep brain stimulation JAMA 1941;117:517–27. reduces symptoms of Parkinson disease. Cleve [36] American men of medicine. 3rd edition. Institute for Clin J Med 1999;66(1):9–11. Research in Biography, Farmingdale, NY, 1961. [19] Just H, Ostergaard K. Health-related quality of [37] Pressman JD. Last resort, psychosurgery and the life in patients with advanced Parkinson’s disease limits of medicine. New York: Cambridge Univer- treated with deep brain stimulation of the sub- sity Press; 1998. thalamic nuclei. Mov Disord 2002;17(3):539–45. [38] Overholser W, editor. Proceedings of the Second [20] Fins JJ. Truth telling and reciprocity in the doctor- Research Conference on Psychosurgery. Evalua- patient relationship: a North American perspec- tion of change in patients after psychosurgery. tive. In: Bruera E, Portenoy RK, editors. Topics in Federal Security Agency, Public Health Service, palliative care, vol. 5. New York: Oxford Univer- National Institutes of Health: Public Health Ser- sity Press; 2001. p. 81–94. vice Publication No. 156. Washington (DC); Gov- [21] National Bioethics Advisory Commission. Re- ernment Printing Office; 1952. search involving persons with mental disorders [39] Frankel G. Psychosurgery’s effects still linger. The that may affect decisionmaking capacity: report Washington Post 1980;April 6:A24–A25. and recommendations of the National Bioethics [40] Bigelow N, editor. Proceedings of the First Advisory Commission, vol. 1. Rockville (MD): Research Conference on Psychosurgery. Cri- National Bioethics Advisory Commission; 1998. teria for the selection of psychotic patients for 318 J.J. Fins / Neurosurg Clin N Am 14 (2003) 303–319

psychosurgery. Federal Security Agency, Public [59] Delgado JMR. Social rank and radio-stimulated Health Service, National Institutes of Health: aggressiveness in monkeys. J Nerv Ment Dis 1967; Public Health Service Publication No. 16. Govern- 144(5):383–90. ment Printing Office: Washington (DC); 1951. [60] Delgado JMR. Aggression and defense under [41] Swayze VW. II. Frontal leukotomy and related cerebral radio control. UCLA Forum Med Sci 1967; psychosurgical procedures in the era before anti- 7:171–93. psychotics (1935–1954): a historical overview. Am [61] Delgado JMR. Implantation of multilead elec- J Psychiatry 1995;152(4):505–15. trode assemblies. Technical documentary report, [42] Sands SL, Malamud W. A rating scale analysis of ADL-TDR. Holloman Air Force Base (NM): the clinical effects of lobotomy. Am J Psychiatry United States Aeromedical Research Laboratory; 1949;105:760–6. 1969. p. 1–19. [43] Diethelm O. Treatment in psychiatry. 2nd edition. [62] Delgado JMR. Radio-controlled behavior. NY Springfield: Charles C. Thomas; 1950. State J Med 1969;69(3):413–7. [44] Proceedings of the Second Research Conference [63] Delgado JMR. Radio stimulation of the brain in on Psychosurgery. Public Health Service Publi- primates and man. Fourth Becton, Dickinson and cation No. 156. Washington (DC): Government Company Oscar Schwidetzky Memorial Lecture. Printing Office; 1952. Anesth Analg 1969;48(4):529–42. [45] Hoffman JL. Clinical observations concerning [64] Mark VH, Ervin FR. Violence and the brain. schizophrenic patients treated by prefrontal leu- New York: Harper and Row; 1970. kotomy. N Engl J Med 1949;241:233–6. [65] Ervin FR, Mark VH, Sweet WH. Focal cerebral [46] Flor-Henry P. Psychiatric surgery 1935–1973: disease, temporal love epilepsy and violent behav- evolution and current perspectives. Can Psychiatr ior. Trans Am Neurol Assoc 1969;94:253–6. Assoc J 1975;20(2):157–67. [66] Mark VH. Brain surgery in aggressive epileptics. [47] Frankel GDC. Neurosurgeon pioneered ‘‘opera- Hastings Cent Rep 1973;3(1):1–5. tion icepick’’ technique. The Washington Post [67] Medical News. Neurosurgeon cites work with epi- April 7, 1980. p. A1–2. leptic patients. JAMA 1974;225(8):917–20. [48] Oltman JE, Brody BS, Friedman S, Green WF. [68] Mark VH. A psychosurgeon’s case for psychosur- Frontal lobotomy: clinical experience with 107 gery. Psychol Today 1974;8(2):28. cases in a state hospital. Am J Psychiatry 1949; [69] Breggin PR. Psychiatry and psychotherapy as a 105:742–51. political process. Am J Psychiatry 1975;29(3): [49] Diering SL, Bell WO. Functional neurosurgery 369–82. for psychiatric disorders: a historical perspective. [70] Mark VH. Psychosurgery versus anti-psychiatry. Sterotact Funct Neurosurg 1991;57:175–94. Boston University Law Review 1974;54(2):217–30. [50] Lewis N. Symposium on lobotomy. Am J Psychi- [71] Mark VH, Sweet WH, Ervin P. Disease in riots atry 1944;101:523. and urban violence. JAMA 1967;201:895. [51] Feldman RP, Goodrich JT. Psychosurgery: [72] Breggin PR. Psychosurgery. JAMA 1973;226(9): a historical overview. Neurosurgery 2001;48(3): 1121. 647–57. [73] Medical News. Does behavior-altering surgery [52] Medical News. Cingulotomy helps some patients, imperil freedom? JAMA 1973;225(8):916–8. surgeon says. JAMA 1973;225(9):1036–7. [74] Breggin PR. Therapy as applied utopian politics. [53] Medical News. Debate over benefits and ethics of Mental Health Soc 1974;1(3,4):129–46. psychosurgery involve public. JAMA 1973;225(8): [75] Medical News. Political overtones could obscure 913–14. clinical issues in psychosurgery. JAMA 1973; [54] Delgado JM, Anshen RN, editors. Physical con- 225(9):1035–44. trol of the mind: toward a psychocivilized society. [76] Mark VH. Brain surgery in aggressive epileptics. New York: Harper and Row; 1969. The Hastings Center Report 1973;3(1):1–5. [55] Osmundsen JA. ‘‘Matador’’ with a radio stops [77] Mark VH. Brain injury and criminality. J Clin wired bull: modified behavior in animal subjects Psychiatry 1978;39(5):486–7. of brain study. New York Times. May 17, 1965. [78] Older J. Psychosurgery: ethical issues and a pro- [56] Talwar SK, Xu S, Hawley ES, Weiss SA, Moxon posal for control. Am J Orthopsychiatry 1974; KA, Chapin JK. Rat navigation guided by mind 44(5):661–74. control. Nature 2002;417(6884):37–8. [79] Blatte H. State prisons and the use of behavior [57] Boyce N. Enter the cyborgs: promise and peril control. Hastings Cent Rep 1974;4(4):11. in a marriage of brains and silicon. US News & [80] Donnelly J. The incidence of psychosurgery in World Report May 13, 2002. p. 56–8. the United States 1971–1973. Am J Psychiatry [58] Onion A. Rat robots, scientists develop remote- 1978;135(12):1476–80. controlled rats. Available at: ABCnews.com http: [81] The National Commission for the Protection of www.abcnews.go.com/sections/scitech/DailyNews/ Human Subjects of Biomedical and Behavioral rats020501.html. Research. Use of psychosurgery in practice and J.J. Fins / Neurosurg Clin N Am 14 (2003) 303–319 319

research: report and recommendations of National Commission for the Protection of Human Subjects Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. Federal of Biomedical and Behavioral Research. Federal Register 1978;43(221):53242–4. Register 1977;42(99):26318–32. [101] Raspberry W. Turning patients into zombies. The [82] Cohn V. Psychosurgery makes gain, 2 black Washington Post December 11, 1978. p. A25. doctors urge limited use of operation. The [102] Annas GJ. Psychosurgery: procedural safeguards. Washington Post January 8, 1976. p. A3. Hastings Cent Rep 1977;7(2):11–3. [83] Martindale D. Psychosurgery: furor over the [103] Frankel G. Cabinet official weighs psychosurgery cuckoo’s nest. New Physician 1977;26(2):22–5. funding. The Washington Post May 10, 1980. [84] Diefenbach GJ, Diefenbach D, Baumeister A, p. A-7. West M. Portrayal of lobotomy in the popular [104] Fins JJ, Miller FG. Enrolling decisionally in- press: 1935–1960. J Hist Neurosc 1999;8(1):60–9. capacitated subjects in neuropsychiatric research. [85] Crichton M. The terminal man. New York: Bal- CNS Spectrums 2000;5(10):32–42. lantine Books; 1998 [originally published in 1972]. [105] Miller FG, Fins JJ. Protecting vulnerable research [86] Mearns EA. Law and the physical control of subjects without unduly constraining neuropsy- the mind: experimentation in psychosurgery. chiatric research. Arch Gen Psychiatry 1999;56: Case Western Reserve Law Review 1975; 701–2. 25(3):565–603. [106] Diering SL, Bell WO. Criteria for the selection of [87] Medical News. Debate over benefits and ethics of psychiatric patients [reprint]. Functional neuro- psychosurgery involve public. JAMA 1973;225(8): surgery for psychiatric disorders: a historical 913–4. perspective. Stereotact Funct Neurosurg 1991; [88] Gaylin WM, Michels R, Edgar HSH, Callahan D. 57:175–94. The law and the biological revolution. Columbia [107] Nuttin B, Gybels J, Cosyns P, et al. Deep brain Journal of Law and Social Problems 1973;10(1): stimulation for psychiatric disorders. Neurosur- 47–76. gery 2002;51(2):519. [89] Gaylin W, Blatte H. Behavior modification in [108] Goodman S. France wires up to treat obsessive prisons. The American Criminal Law Review disorder. Nature 2002;417(13):677. 1975;13(1):11–35. [109] Fins JJ. The ethical limits of neuroscience. The [90] Steinfels P. A clockwork orange—or just. Hastings Lancet Neurology 2002;1:213. Cent Rep 1974;4(2):10–2. [110] Dougherty DD, Baer L, Cosgrove GR, et al. [91] New World. Future of psychosurgery in doubt. Prospective long-term follow-up of 44 patients Nature 1973;224(5412):128–9. who received cingulotomy for treatment-refractory [92] Annas GJ, Glantz LH. Psychosurgery: the law’s obsessive-compulsive disorder. Am J Psychiatry response. Leg Med Annu 1975;329–48. 2002;259(2):269–75. [93] Solvenko R. Commentary: on psychosurgery. [111] Greenberg BD, Murphy DL, Rasmussen SA. Hastings Cent Rep 1975;5(5):19–22. Neuroanatomically based approaches to obses- [94] Burt RA. Reflections on the Detroit psychosurgery sive-compulsive disorder. Neurosurgery and trans- case. Why we should keep prisoners from the cranial magnetic stimulation. Psychiatr Clin North doctors. Hastings Cent Rep 1975;5(1):25–34. Am 2000;23(3):671–86. [95] Valenstein ES. Great and desperate cures. New [112] Bejjani B-P, Damier P, Arnulf I, et al. Transient York: Basic Books; 1986. p. 288. acute depression induced by high-frequency deep- [96] Culliton BJ. Psychosurgery: National Commission brain stimulation. N Engl J Med 1999;340(19): issues surprisingly favorable report. Science 1976; 1476–80. 194(4262):299–301. [113] Satcher DS. Executive summary: a report of the [97] The National Research Act. Pub L, 93–348. surgeon general on mental health. Public Health [98] Jonsen AR. The birth of bioethics. New York: Rep 2000;115(1):89–101. Oxford University Press; 1998. p. 96. [114] Fins JJ. Case study commentary: palliation in the [99] National Commission for the Protection of Hu- age of chronic disease. Hastings Cent Rep 1992; man Subjects of Biomedical and Behavioral Re- 22(1):41–2. search. Research involving those institutionalized [115] Fins JJ. Principles in palliative care: an overview. as mentally infirm: report and recommendations. J Respir Care 2000;45(11):1320–30. Washington (DC): US Department of Health, [116] Gaylin WM. The problem of psychosurgery. In: Education, and Welfare; 1978. Gaylin WM, Meister JS, Neville RC, editors. [100] Determination of the secretary regarding the re- Operating on the mind: the psychosurgery conflict. commendation on psychosurgery of the National New York: Basic Books; 1975. p. 22. Color Plates Plate 1. Magnetoencephalographic data from three patients in this series. Power spectra (four top panels) and coherence plots (three bottom panels) are displayed. The top left panel shows the power spectra from controls (blue) and thalamocortical dysrhythmia (TCD) patients, including neuropsychiatric as well as parkinsonian, neurogenic pain, and tinnitus patients. A peak shift into the theta domain and power increase in the theta and beta bands are demonstrated for TCD patients in comparison to controls. The top right panel presents the power spectrum of Patient 2, with a postoperative (blue) curve demonstrating the reappearance of the alpha peak and the power decrease in the theta and beta bands. The postoperative power spectra of Patient 7 (middle left) and Patient 1 (middle right) are mainly characterized by a reduction in the theta and beta power, allowing reappearance (Patient 7) or better visualization (Patient 1) of the alpha peak. Coherence was analyzed applying a cross-correlation analysis of the variation along time of the spectral power for frequencies between 0 and 40 Hz. The bottom left panel shows such coherence for controls, the middle panel shows coherence for Patient 2 before surgery, and the right panel shows coherence for the same patient after surgery. (See also Fig. 3 in article by Jeanmonod et al.) Plate 1. Plate 2. Magnetoencephalographic source localization for Patient 2. Projection of 4- to 10-Hz activity onto the patient’s MRI examination before (eight top images) surgery. The thalamocortical dysrhythmia of this patient is localized in the right-sided paralimbic domain comprising the temporopolar, anterior parahippocampal, orbitofrontal, and basal medial prefrontal areas. This low-frequency focus disappears after surgery. (See also Fig. 4 in article by Jeanmonod et al.) Plate 2. Plate 3. Schematic diagram of the thalamocortical circuits that support the thalamocortical dysrhythmia mechanisms. Three thalamocortical modules are shown, each with its specific (yellow) and nonspecific (green) thalamic relay cell projecting to the cortex and reticular nucleus, one (blue) pyramidal cell with its corticothalamic and corticoreticular output, and two reticular cells (red) with their projections back to thalamic relay cells. Three thalamocorticothalamic loops are thus displayed, with their feed-forward thalamocortical activation and their respective recurrent feedback is proposed to generate increased cortical activation through temporal coincidence. Either thalamic cell disfacilitation or overinhibition (central module) hyperpolarizes thalamic cell membranes. This allows the deinactivation of calcium T- channels and the generation of low-threshold calcium spike bursts, thus resulting in low-frequency thalamocortical substrate for low-frequency coherent discharge of an increasing number of thalamocortical modules. At the cortical level, low-frequency activation of corticocortical inhibitory interneurones (red), by reducing lateral inhibitory drive, results in high-frequency coherent activation of the neighboring (right) cortical module (edge effect). (See also Fig. 5 in article by Jeanmonod et al.) Plate 3. Neurosurg Clin N Am 14 (2003) 321–325 Index Note: Page numbers of article titles are in bold face type.

A Capsulotomy, for psychiatric disorders N-Acetyl aspartate, in magnetic resonance depression, 219, 232–233 spectroscopy, in obsessive-compulsive history of, 191–192 disorder, 214–215 obsessive-compulsive disorder, 204–208, 219, 232–233 Amygdala, connections to, in depression, 216–218 Caudate nucleus, connections to in depression, 216–219 Annas, George, on psychiatric neurosurgery in obsessive-compulsive disorder, 214, regulations, 314 218–219 Anticonvulsant action, of vagus nerve Cerebrospinal ﬂuid analysis, in depression, in stimulation, 276–277, 279–280 vagus nerve stimulation, 280 Antidepressants, for depression, 201. See also Chlorpromazine, for psychiatric disorders, history Transcranial magnetic stimulation. of, 190 Anxiety disorders, transcranial magnetic Cingulate cortex, anterior, connections to stimulation for, 293–294 in depression, 216–219 B in obsessive-compulsive disorder, 214, 218–219 Ballantine, H. Thomas, Jr., on psychiatric neurosurgery, 308 Cingulotomy, for psychiatric disorders, 191–192, 225–235 Barber, Jesse, on psychiatric neurosurgery, 311 anatomic considerations in, 225–227 Behavioral therapy, for obsessive-compulsive complications of, 230 disorder, 200–201 depression, 204–206, 208–209, 218, 225–228 history of, 225 Benabid, Alim, deep brain stimulation studies obsessive-compulsive disorder, 204–206, of, 315 208–209, 216, 225–231 Bipolar mood disorder, thalamocortical SPECT in, 238–249 dysrhythmia in. See Thalamocortical patient selection for, 227–228, 231 dysrhythmia. physiologic considerations in, 225–227 Body dysmorphic disorder, neuroimaging and results of, 228–231, 241–243 neurocircuitry in, 214 technique for, 228, 239 versus alternative surgical methods, 231–233 Broca, Paul, localization philosophy of, 181 Corticostriatothalamocortical circuitry Burckhardt, Gottlieb, frontal lobotomy in depression, 216–219 procedure of, 181–182 in obsessive-compulsive disorder, 214, 218–219 C Califano, Joseph, psychiatric neurosurgery D regulations approved by, 314 Delgado, Jose M.R., on psychiatric neurosurgery, 309 Capsule, internal, anterior limbs of, electrical stimulation of, in obsessive-compulsive Deniker, Pierre, psychoactive drug therapy disorder, 267–274 discovery by, 190

Depression, 199–212 Ethical issues, in neuropsychiatric surgery, clinical features of, 201 303–319 epidemiology of, 201 historical aspects of, 303–308 medical treatment of, 201 regulatory aspects of, 312–315 neurocircuitry models for, 216–219 sociopolitical aspects of, 308–312 neuroimaging in, 216–219 neurosurgery for F capsulotomy, 219, 232–233 Fiamberti, Amarro, transorbital lobotomy cingulotomy, 204–206, 208–209, 225–228 procedure of, 189 current awareness of, 201–202 eﬀectiveness of, 207–209 Freeman, Walter, psychiatric neurosurgical history of, 202–203 procedures of, 182–183, 185–190, 303–306 limbic leukotomy, 205–207, 209, 232 Frontal lobotomy, history of, 182–190, 203, modern procedures for, 203–205 303–308 neuroimaging and neurocircuitry related to, 218–219 Fulton, John F., neurophysiologic studies of, safety of, 205–207 182, 185–187 subcaudate tractotomy, 204–205, 207, Functional neuroimaging, in obsessive- 218, 231 compulsive disorder, 215 thalamocortical dysrhythmia in. See Thalamocortical dysrhythmia. transcranial magnetic stimulation for, G 288–293 Gall, Franz Joseph, on phrenology, 181 maintenance, 292 Gamma knife capsulotomy, for psychiatric meta-analysis of, 289–292 disorders, 204–208 with intentional seizure generation, 294–295 Goltz, Friedrich, temporal lobotomy procedure vagus nerve stimulation for, 275–282 of, 181 Diethelm, Oskar, on ethics of neuropsychiatric Grant, Francis, lobotomies by, 186 surgery, 305–307 Grinker, Roy, on ethics of neuropsychiatric surgery, 306–307 E H Electrical stimulation of internal capsule anterior limbs, for Hallucinations, thalamocortical dysrhythmia in. obsessive-compulsive disorder, 267–274 See Thalamocortical dysrhythmia. of vagus nerve, for depression, 275–282 Hippocampus, connections to, in depression, Electroconvulsive therapy, for psychiatric 216–218 disorders, 184, 201 Hypothalamic-pituitary-adrenal axis, connections versus transcranial magnetic stimulation, to, in depression, 216 287–289, 291–292, 294 Electroencephalography, in depression, in vagus I nerve stimulation, 280 Impulse control disorder, thalamocortical Eloquent cortex, localization of, transcranial dysrhythmia in. See Thalamocortical magnetic stimulation for, 287 dysrhythmia. Epilepsy Institutional review boards, for psychiatric transcranial magnetic stimulation for, neurosurgery research, 313 286–287, 293 vagus nerve stimulation for, 276–277 J Ervin, Frank R., on psychiatric neurosurgery, Jacobsen, Carlyle, neurophysiologic studies of, 309 182 Index / Neurosurg Clin N Am 14 (2003) 321–325 323

K Magnetoencephalography, in thalamocortical Kelly, Desmond, limbic leukotomy procedure dysrhythmia, 253–256, 260 of, 193 Mania, transcranial magnetic stimulation for, Knight subcaudate tractotomy, 193 292 Mark, Vernon H., on psychiatric neurosurgery, L 309–310, 315 Laborit, Henri, psychoactive drug therapy Mearns, Edward, on psychiatric neurosurgery, discovery by, 190 311–313 Leksell, Lars, capsulotomy procedure of, 191 Mixter, W. Jason, lobotomies by, 187 Leukotomy Moniz, Egas, psychiatric neurosurgical limbic, for psychiatric disorders, 193 procedures of, 182–183, 185–186, 190, 305–306 depression, 205–207, 209, 232 obsessive-compulsive disorders, 205–207, Movement disorders, transcranial magnetic 209, 218–219, 232 stimulation for, 293 versus cingulotomy, 232 orbitomedial, for obsessive-compulsive N disorder, 218 National Commission for the Protection of Lima, Almeida, early frontal lobotomies by, 182 Human Subjects of Biomedical and Behavioral Research, psychiatric neurosurgery and, Limbic leukotomy, for psychiatric disorders, 193 312–314, 316 depression, 205–207, 209, 232 obsessive-compulsive disorders, 205–207, 209, National Psychosurgery Advisory Board, 218–219, 232 313–315 Limbic system, in psychiatric disorders Neurocircuitry models, for psychiatric disorders, dysfunction of, 226 213–223 pathway disruption in. See Cingulotomy; depression, 216–219 Limbic leukotomy. obsessive-compulsive disorder, 214–216, 218–219 Lobotomy frontal, history of, 182–190, 203, 303–308 Neuroimaging. See also speciﬁc techniques. prefrontal, history of, 182–190 for psychiatric disorders, 213–223 transorbital, 189 depression, 216–219 obsessive-compulsive disorder, 214–216, Lyerly, James, psychiatric neurosurgical 218–219 procedures of, 186–187 Neville, Robert, on psychiatric neurosurgery, 312 M Magnetic resonance imaging O blood oxygenation level-dependent Obsessive-compulsive disorder, 199–212 in depression, in vagus nerve stimulation, behavioral therapy for, 200–201 279 clinical features of, 200 in transcranial magnetic stimulation stud- diagnosis of, SPECT in, 238, 241–242 ies, 288 epidemiology of, 200 in stereotactic cingulotomy, 228, 238 medical treatment of, 200 morphometric neurocircuitry models for, 214–216, 218–219 in depression, 217 neuroimaging in, 214–216, 218–219 in obsessive-compulsive disorder, 214 neurosurgery for capsule anterior limb electrical stimulation, Magnetic resonance spectroscopy, in obsessive- 267–274 compulsive disorder, 214–215 capsulotomy, 204–208, 219, 232–233 Magnetic stimulation, transcranial. See cingulotomy, 204–206, 208–209, 216, Transcranial magnetic stimulation. 225–231, 238–249 324 Index / Neurosurg Clin N Am 14 (2003) 321–325

Obsessive (continued) for depression. See Depression, neurosurgery current awareness of, 201–202 for. eﬀectiveness of, 207–209 for obsessive-compulsive disorder. See history of, 202–203 Obsessive-compulsive disorder, limbic leukotomy, 205–207, 209, 218–219, neurosurgery for. 232 for thalamocortical dysrhythmia, 251–265 modern procedures for, 203–205 historical perspective of, 181–197, 202–203 neuroimaging and neurocircuitry related to, capsulotomy, 191–192 218–219 cingulotomy, 191–192 safety of, 205–207 early years, 181–190 subcaudate tractotomy, 204–205, 207, 218, ethical aspects of, 303–308 231–232 future techniques derived from, 193–196 pathophysiology of, 226 limbic leukotomy, 193 positron emission tomography in, 226, stereotactic methods, 190–193 246–248 subcaudate tractotomy, 193 SPECT in, 237–250 neuroimaging and neurocircuitry models for, diagnostic, 238, 241–242 213–223 postoperative, 238–241, 245–249 palliative, 316 thalamocortical dysrhythmia in. See research on Thalamocortical dysrhythmia. ethical considerations in, 303–319 transcranial magnetic stimulation for, regulation of, 312–315 293–294 single-photon emission computed tomography in, 237–250 OCD. See Obsessive-compulsive disorder. sociopolitical aspects of, 308–312 Orbitofrontal-subcortical connections transcranial magnetic stimulation, 283–301 in depression, 216–219 in obsessive-compulsive disorder, 214, 218–219 interruption of, in subcaudate tractotomy, R 193, 204–205, 207 Raspberry, William, on psychiatric neurosurgery regulations, 314 Orbitomedial leukotomy, for obsessive- compulsive disorder, neuroimaging in, 218 Regulation, of neuropsychiatric surgery research, 312–315 P Richardson, Alan, limbic leukotomy procedure of, 193 Palliative care, for psychiatric disorders, neurosurgery in, 316 Ryan, J. Kenneth, on psychiatric neurosurgery, 313 Pallidotomy, anterior medial, for thalamocortical dysrhythmia, 252–263 magnetoencephalography with, 253–256, 260 S patient selection for, 252 Schizophrenia results of, 255, 259–263 lobotomy for, 308 strategy for, 252–253, 259 transcranial magnetic stimulation for, 293 Parkinson’s disease Seizures. See also Epilepsy. deep brain stimulation for, 315 induction of, in transcranial magnetic transcranial magnetic stimulation for, 293 stimulation, 294–295 Phrenology, 182 Selective serotonin reuptake inhibitors, for Positron emission tomography, in obsessive- obsessive-compulsive disorder, 200 compulsive disorder, 226, 246–248 Shock therapies, for psychiatric disorders, 184 Prefrontal lobotomy, history of, 182–190 Single-photon emission computed tomography Psychiatric neurosurgery. See also speciﬁc in depression, in vagus nerve stimulation, 279 procedures, e.g., Cingulotomy. in obsessive-compulsive disorder, 237–250 Index / Neurosurg Clin N Am 14 (2003) 321–325 325

diagnostic, 238, 241–242 versus cingulotomy, 231–232 postoperative, 238–241, 245–249 Transcranial magnetic stimulation, 283–301 SPECT. See Single-photon emission computed animal models of, 287 tomography. drug use with, 293 electrical requirements of, 285 Stereotactic surgery for anxiety disorders, 293–294 cingulotomy. See Cingulotomy. for depression, 288–293 history of, 190–193 maintenance, 292 Subcaudate tractotomy, for psychiatric disorders, meta-analysis of, 289–292 193 with intentional seizure generation, depression, 204–205, 207, 218 294–295 obsessive-compulsive disorder, 204–205, 207, for eloquent cortex localization, 287 218, 231–232 for epilepsy, 286–287, 293 versus cingulotomy, 231–232 for mania, 292 for movement disorders, 293 T for schizophrenia, 293 functional imaging with, 287–288 Talairach capsulotomy procedure, 191 high-frequency, 285–286 Thalamocortical dysrhythmia, 251–265 mechanisms of action of, 284–286 pathophysiology of, 251, 259, 262 paired-pulse, 286–287 surgery for, 252–263 repetitive, 284 magnetoencephalography with, 253–256, research uses of, 286–288 260 safety of, 286 patient selection for, 252 versus electroconvulsive therapy, 287–289, results of, 255, 259–263 291–292, 294 strategy for, 252–253, 259 Transorbital lobotomy, 189 Thalamotomy Trichotillomania, neuroimaging and central lateral, for thalamocortical neurocircuitry in, 214 dysrhythmia, 252–263 magnetoencephalography with, 253–256, 260 V patient selection for, 252 Vagus nerve, electrical stimulation of results of, 255, 259–263 clinical applications of, 280 strategy for, 252–253, 259 for depression, 275–282 dorsomedial, for psychiatric disorders, 191 anatomic considerations in, 275 Thalamus, connections to clinical trials of, 276 in depression, 216–219 long-term results of, 278–279 in obsessive-compulsive disorder, 214, 218–219 mechanism of action of, 277, 279–280 open-label study of, 277–278 Thermocapsulotomy, for psychiatric disorders, placebo-controlled study of, 278–279 204–208 practical considerations in, 280–281 Tics, transcranial magnetic stimulation for, 293 preclinical investigations of, 275–276 Tourette syndrome Valentine, Elliot S., on psychiatric neurosurgery, neuroimaging and neurocircuitry in, 214 312 transcranial magnetic stimulation for, 293 W Tractotomy, subcaudate, for psychiatric Watts, James, psychiatric neurosurgical disorders, 193 procedures of, 185–187, 189 depression, 204–205, 207, 218 obsessive-compulsive disorder, 204–205, 207, Weinberger, Lawrence M., on ethics of 218, 231–232 neuropsychiatric surgery, 306