Chapter 42, Not Your Father’s Lobotomy: Psychiatric Surgery Revisited (continued)

Subcaudate Tractotomy

Another stereotactic procedure geared towards interrupting fibers from the orbitofrontal cortex to the thalamus is subcaudate tractotomy (innominatomy). Developed by Geoffrey Knight in 1965 in London, the operation was designed to relieve depressive, anxiety, and obsession symptoms while minimizing postoperative epilepsy, as well as cognitive and personality deficits. , The lesion is created by placement of multiple 1 х 7 mm rods of yttrium-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 and known as the substantia innominata has been traditionally 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. Initially, placing two rows of four rods each made the lesion. Later studies, having refined the technique, have created the lesion by RF thermocoagulation. Such a lesion would be expected to interrupt reciprocal connections involving OFC and subgenual ACC to the striatum and thalamus. Also, amygdalofugal fibers to OFC and subgenual ACC would be affected.

Depression has been the most common diagnosis for patients undergoing this technique. A review of the literature has revealed a response rate of 55 to 68 %. OCD patients undergoing this procedure have been shown to have similar response rates in the range of 50 %.16 The most comprehensive review of subcaudate tractotomy included 1300 patients. In this study, 40 to 60% of patients lead normal to near-normal lives. Only 1 % of subcaudate tractotomy patients committed suicide compared with a 15 % rate of suicide among patients with similar major affective disorders (AD).16

As with other stereotactic interventions, the most common side effects were transient headache, confusion, and lethargy. One death was reported as a direct result of the surgery. In a 1975 study, 6.7 % of 208 patients had mild long-term personality changes, whereas a study in 1994 reported no such changes in 1300 patients.16 As experience with cingulotomy and capsulotomy became more prevalent and political difficulties with psychiatric surgery persisted, subcaudate tractotomy fell by the wayside.

Limbic Leucotomy

Whereas the three aforementioned procedures each target a single anatomic substrate, limbic leucotomy is designed to interrupt fibers at two separate areas, one involving a frontothalamic loop and the other involving an area of the Papez circuit. Termed limbic leucotomy, the procedure was developed in England by Desmond Kelly and Alan Richardson in the early 1970s. The operation itself consists of three 6-mm thermocoagulative or cryogenic lesions in the lower medial quadrant of each frontal lobe (to interrupt frontothalamic connections) and two 6 mm lesions in each cingulum. Essentially it is a combined subcaudate tractotomy and cingulotomy. This intervention would impact on multiple CSTC loop systems as defined above.

In 2002, the most recent experience with limbic leucotomy using modern diagnostic and system assessment procedures found a 50% response rate in OCD and depression patients. Historically, Kelly and his collaborators reported no deaths or seizures in a prospective group of 66 patients. Again, transient headaches, lethargy, apathy, and incontinence were the most common side effects. One patient had severe memory impairment owing to improper lesion placement, whereas 12 % had persistent lethargy in an average follow up of 16 months. More recent experience with limbic leucotomy corroborates the incidence of these adverse effects.21


Our understanding of the neurocircuitry of psychiatric disorders is rapidly evolving. What is most strongly supported is the role of CSTC loops in the pathophysiology of psychiatric symptoms. In the mid-1980s, Alexander et al. , suggested there were at least five functionally segregated basal ganglia thalamotocortical circuits. These circuits were hypothesized to be anatomically segregated and to subserve different physiological functions. One, a “motor” circuit, centers on the sensorimotor portions of the caudate/putamen, globus pallidus, substantia nigra, thalamus, and premotor areas. Other loops include the oculomotor, dorsolateral prefrontal, anterior cingulate, and lateral orbitofrontal cortex.31 Each of these circuits is segregated through the aforementioned basal ganglionic structures and has specific cortical projections. (Fig. 42.9)

Within each circuit are “direct” and “indirect” pathways. (Fig. 42.10) The “direct” pathway has two excitatory and two inhibitory pathways, making it a net positive feedback loop. The “indirect” pathway with its three inhibitory and one excitatory connection can be conceived of as a net negative feedback loop.14 In the case of the latter three circuits, the cortical areas involved are the dorsolateral prefrontal, the cingulate, and the orbitofrontal cortices as opposed to the pre-motor and motor cortices involved with the “motor” loop. It is based on this framework that modern neurosurgical intervention in Parkinson’s disease (PD) has been developed. A similar scheme can be used to develop a strategy for the surgical intervention of psychiatric disorders. Again, one major limitation facing the development of psychiatric surgery is that a convincing animal model to test hypotheses for OCD and other psychiatric disease does not exist. Although the situation is somewhat better for depression, there remain few system studies of the circuitry that is hypothesized to underlie the development of depression.

For the purposes of this chapter, two of the most elucidated psychiatric diseases that have been the targets of neurosurgical interventions will be discussed. They in turn can be used as a framework to begin to study the use of DBS as a surgical strategy for the treatment of psychiatric disease in general.

Obsessive-Compulsive Disorder

Cortico-striato-thalamocortical interactions are strongly implicated in the pathogenesis of psychiatric disease in humans and specifically in the mediation of OCD symptoms and its response to treatment. Evidence for this is derived from several sources, which are discussed below.

Whereas movement disorders and psychiatric disease might seem dissimilar on the surface, common neural substrates are implicated in their symptomatology. From the earliest observations of OCD, the central role of neuronal areas subserving motor function in its pathogenesis has been speculated Tourette’s disorder, a disease characterized by motor tics as well as OCD-like symptoms, demonstrates the possibility of a neural substrate capable of producing motor as well as psychiatric disease states. Studies demonstrating the strong clinical and genetic association between Tourette syndrome and OCD have suggested the central role of the basal ganglia in the genesis of OCD symptoms. , A similar basal ganglia circuit to the one implicated in PD has been proposed to explain the production of both motor and obsessional symptoms in Tourette’s syndrome. Further analysis of the clinical spectrum of PD has revealed many striking similarities between the “motor” disease of PD and the psychiatric diseases of OCD and AD.

Neuropharmacological hypotheses of OCD pathophysiology have also implicated CSTC loops in the pathogenesis of OCD. There is strong evidence that serotonergic systems modulate OCD symptoms. Potent inhibitors of serotonin transporter function (serotonin reuptake inhibitors [SRI]) are unique among antidepressants in producing at least some clinical benefit in most patients with OCD. Interestingly, both the serotonin transporter and some serotonin receptor subtypes, such as 5-HT2A and 5-HT2C, implicated in OCD are highly expressed in the ventral striatum where they could influence the functioning of the CSTC and functionally-related circuits, which are of most interest in OCD. In theory, other neurotransmitter systems within CSTC loops may play a role in susceptibility, course, or response to OCD treatment. For example, dopaminergic mechanisms have been implicated by controlled studies demonstrating that neuroleptics, ineffective as monotherapy in OCD, are beneficial when added to ongoing SRI treatment. Other CSTC neurotransmitter systems that are candidates for involvement in OCD on the basis of the anatomical localization or functional roles in cortico-striato-pallidothalamic circuits include glutamate, aminobutyric acid (GABA), Substance P, cholinergic, and endogenous opiod mechanisms.

Based on these observations and those of several other authors, one can begin to construct a neuronal architecture for the basis of OCD and develop a rationale for surgical intervention. (Fig. 42.11) When constructing such a model, it is important to recognize how rapidly our knowledge of psychiatric pathoneurophysiology is changing. What is most evident is that it is unlikely that a single center or anatomic or physiological defect is responsible for the pathogenesis of psychiatric symptoms (e.g., loss of dopamine in PD). It is more likely that a dysregulation between several neural circuits is involved in the genesis of the disorder. From this follows a slightly different strategy for psychiatric neurosurgery than what has transpired in the past with movement disorder surgery. Historically, in movement disorder surgery, surgical intervention focused on specific anatomic areas thought to either be hyper- or hypoactive. Surgical procedures were designed to “rectify” this over- or underactivity. Just as this picture has grown more complex in movement disorders through animal models and functional imaging, in psychiatric conditions, it is becoming evident that there are several surgical targets that could potentially modulate entire neural systems and thus bring about an amelioration of symptoms. This systems approach is vital in assessing the somewhat conflicting data in the literature and making the choice of a surgical target. The multicircuit model hypothesizes that the primary pathogenic mechanism lies in a dysregulation of the basal ganglia/limbic striatal circuits that modulate neuronal activity in and between portions of the orbitofrontal and anterior cingulate cortices as well as the medial, dorsomedial and anterior thalamic nuclei. ,14,

There are three components to this neuronal model of OCD. The first involves a reciprocal positive-feedback loop involving the orbital and prefrontal cortex and the dorsomedial (DM) thalamic nucleus, by way of the anterior limb of the internal capsule. The corticothalamic projection is excitatory and mediated primarily by glutamate and aspartate. Although the reciprocal thalamocortical projection’s neurotransmitter remains to be identified, multiple studies suggest it to be excitatory and, most probably, glutamatergic.45,

The second component of the OCD model involves the orbitofrontal/prefrontal cortex, the ventral caudate, the dorsomedial pallidum, and the intralaminar, anterior and DM thalamic nuclei. Alexander et al. outlined this relationship in their OFC - CSTC loop. While projections from the ventral striatum to the dorsomedial pallidum involve multiple neurotransmitters including GABA and substance P, the output of this pathway by way of the dorsomedial pallidum to the thalamus is almost exclusively inhibitory, mediated by GABA. This component is thought to serve as a modulator for the excitatory positive-feedback orbitofrontal thalamic loop described earlier. Another vital aspect of this second component of the OCD model involves serotonergic projections from the dorsal raphe nuclei of the midbrain to the ventral striatum. These are speculated to be inhibitory in nature.

The third constituent of this model involves the limbic system and the circuit of Papez. Many manifestations of OCD have features in common with anxiety disorder. The impact of the patient’s various obsessions and compulsions on his or her emotional state is the hallmark of the disease. In 1937, Papez concluded that participation from the cerebral cortex is essential for the subjective emotional experience and that emotional expression is dependent on the integrative action of the hypothalamus. Papez devised a circuit (Fig. 42.12) based on his observations on neuroanatomic connections to integrate these two structures. The pathway begins from the hippocampal formation to the mammillary body via the fornix. The projection, via the mammillothalamic tract, continues on to the anterior thalamic nuclei. From here, there are widespread connections to the cingulate gyrus. In the aforementioned OCD model, there are numerous connections to the Papez circuit via the DM nuclei and the OFC. There are also heavy projections from the anterior cingulate cortex (ACC) to the nucleus accumbens region of the striatum. These connections could subserve the anxiety/emotional component of OCD.

Recent functional imaging studies have consistently found evidence that corroborate this model of OCD pathogenesis. It is important to assess functional imaging data of psychiatric disease carefully as there are overlaps of activation phenomenon between normal and psychiatric patients. Furthermore, it is often difficult to distinguish changes that are markers for symptom amelioration versus activity subserving “normal” function. Nevertheless, functional imaging data does further implicate the role of CSTC loops in OCD. This appears in both the neutral and provoked state. The neutral state is considered the baseline obsessional and compulsive behavior of the OCD patient. The provoked state occurs when the OCD patient is presented with a stimulus known to exacerbate his or her OCD profile. After treatment with appropriate medications, including selective serotonin reuptake inhibitors (SSRI), and behavioral therapies, these areas of abnormally increased metabolism were shown to decrease by positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) studies.55,14, Such areas of activation and responses to treatment might prove useful in assessing future neurosurgical treatments for OCD.

Affective Disorder

CSTC interaction has been implicated in the pathoneurophysiology of ADs, including major depression (MD) and bipolar disorder. (Fig. 42.13)Other limbic elements are important as well, including the amygdala, hippocampus, and the hypothalamic-pituitary axis (HPA). Symptoms of AD have cognitive, motor, affective, and neuroendrocrinological components, each with its own particular neural circuit. , 14

The neural circuitry of AD seems to consist of a dorsal, a ventral, and a modulatory component. The dorsal component, involved with the motor and cognitive aspects of AD, consists of the prefrontal, dorsal anterior cingulate, and premotor cortices. This compartment accesses the dorsal striatum and projects to the thalamus via its projections from the dorsomedial portion of the pallidum, closing the loop. Next, a ventral component, involved with affective aspects of depression, consists of the subgenual anterior cingulate (Brodmann’s area 25), orbitofrontal, and insular cortices. Interaction with the ventral striatum through the medial/rostral pallidum and subsequently the thalamus, closes this component.

Finally, there is a modulatory component consisting of the pregenual ACC, the amygdala and the HPA (hypothalamic-pituitary axis), which is thought to regulate relative ventral-dorsal component activity and subserve the neuroendocrine aspects of AD symptoms. The Amygdala-HPA axis and the pregenual ACC modulate the aforementioned ventral-dorsal compartment relationship via the amygdala’s tendency to drive activity to the ventral compartment and the pregenual ACC’s inhibitory projections to both compartments. The hippocampus, in turn, modulates activity in HPA.

With regard to the endocrine and humoral aspects of depression, connections between the corticomedial amygdala and the hypothalamus via the stria terminalis regulate the release of cortisol and epinephrine in relation to emotional stimuli. Basolateral amygdala connections with the basal ganglia, indirectly via connections such as the ventral amygdalofugal pathway and direct cortical connections, influence skeletomotor motivation and behaviors in response to emotional stimuli.

These three components can be condensed into a model analogous to the one proposed for OCD. In this model, faulty transmission, such as altered levels of excitability and temporal patterning of activity in parallel, but interconnected, CSTC systems can be associated with depression symptoms: a ventral affective and limbic-thalamocortical loop consisting of the cingulate, OFC, and the anterior or dorsomedial thalamic nuclei; a dorsal cognitive thalamocortical loop consisting of the prefrontal, premotor and cingulate cortices and the anterior or dorsomedial thalamic nuclei, and a modulating circuit consisting of the amygdala and the HPA.

Much of the work implicating the basal ganglia and other structures in the pathogenesis of AD is derived from imaging studies using PET and fMRI. The neuroimaging of AD is hampered by the heterogeneity of the disease; however, some consistent features are evident. Abnormalities in metabolism have been demonstrated in the OFC, cingulate cortex, basal ganglia, and amygdala. , , Increased metabolism has been found in orbitofrontal, anterior insular, and subgenual cingulate cortices in transient states of sadness and major depression episodes. Conversely, decreased metabolism has been demonstrated in the dorsolateral prefrontal cortex in MD and transient sadness. Both of these findings have been reversed with successful treatment. Finally, increased metabolism in the amygdala, also reversed by treatment, has been associated with depressed patients. Metabolism in the subgenual cingulate and hippocampus has been variable.14 Based on functional imaging data, there seems to be a relative increase in activity in the ventral compartment and hypoactivity in the dorsal compartment. Reciprocal inhibitory connections between the dorsal and ventral compartments, combined with amygdala hyperactivity and abnormal hippocampal activity could generate this overall relationship leading to AD symptoms. With regard to the implication of these interactions for DBS for AD, “Thus, successful treatment of MD (via any of a number of modalities) may rely on some combination of deactivation of the ventral compartment, inhibition of the amygdala, stimulation (or protection) of the hippocampus...” 14 It is important to remember that, unlike the model for PD, these models of psychiatric disease, because of their inherent nature, have little support from animal models. Whereas some models for depression exist, and a number of models of OCD have been proposed, all have significant limitations. There is recently more interest in developing models of different component symptoms of syndromes and of associated and predisposing factors. Future work in this area may yet yield some appropriate animal models.

Neural circuit models of psychiatric disease are based on anatomic connections and the aforementioned functional imaging studies. One major concern, especially for depression, but also true for OCD, is that there are significant disagreements among researchers over models of affective disorders. Another aspect of the above models involves not only the physiological and anatomic connections themselves, but also the “weights” of the connections. There are bodies of evidence implicating involvement of the above brain circuits in symptoms of OCD and depression, and in emotional processing more generally, including “dispositional mood” (personality traits such as neuroticism and extraversion) that seem to have major influence on risk on psychopathology. , Thus, although these models may seem too simple, they serve as a springboard for future functional imaging and physiological mapping studies and form the basis from which neurosurgical and pharmacological therapies can be developed in a similar fashion as they were for PD.

Previous Page | Page 193 Next Page