James F. Leckman, Bradley S. Peterson, David L. Pauls

During the course of the past two decades, Tourette's syndrome (TS) and related conditions have emerged as model disorders for researchers interested in the interaction of genetic, epigenetic, and neurobiological factors which shape clinical outcomes from health to chronic disability over the life span (50,61). Figure 1 depicts these interactions and provides a framework for much of the research described in this chapter.

PHENOMENOLOGY AND NATURAL HISTORY

TS is a chronic neuropsychiatric disorder of childhood onset that is characterized by tics that wax and wane in severity and an array of behavioral problems including attention deficit hyperactivity disorder (ADHD) and some forms of obsessive compulsive disorder (OCD) (19,78,88). Many of the features of the syndrome - including the age of onset, course, anatomic distribution of tics, the sensory phenomena, the sex ratio, the comorbidity with OCD and ADHD, and the marked diminution of tics during sleep -provide promising clues concerning the underlying neurobiology of this syndrome (59).

Tic symptoms. The range of symptoms is enormous and includes motor and phonic tics. Tics are defined as sudden repetitive movements, gestures or utterances that typically occur in bouts and resemble fragmentary aspects of normal behavior. Motor tics vary from abrupt movements such as rapid forceful eye blinking, sudden head jerks, and shoulder shrugs, to more complex and purposive or dystonic behaviors such as gestures of the hands or face or a slow sustained head tilt. Phonic or vocal tics range from simple sniffing or throat clearing to fragments of words and phrases.

Usually of brief duration, individual tics rarely last more than a second. Several tics can occur at the same moment or in an orchestrated sequence. An example would be a head-shake, shoulder shrug, and a inarticulate shout all erupting at the same moment. Clinical rating instruments presently in use focus on the number, frequency, intensity, and complexity of tics as well as the degree of interference that they can cause (63). Other useful instruments focus on how noticeable the tics are to others, whether they elicit comments or curiosity, and the extent to which they make the patient's appearance odd or bizarre (93). Videotaped tic counts can also provide valuable information for medication trials and acute challenge studies (15).

Premonitory urges. Less well appreciated are the sensorimotor phenomena that frequently accompany tics and obsessive-compulsive behavior. These experiences include premonitory feelings or urges that are relieved with the performance of the tic, and a need to perform tics or compulsions until they are felt to be "just right" (65,66). Figure 2 presents the anatomical distribution of premonitory urges. The premonitory feelings associated with tics are usually described as being a physical sensation - something like a faint, but insistent, urge to itch. Whereas, the premonitory awareness that something is not "just right" in the environment is typical described as be more a mental than a physical experience.

Natural history. As described by Georges Gilles de la Tourette and as is reported in virtually every series, the modal age of onset of motor tics is 6 to 7 years of age with a range from 2 to 16 or 17 (93). The symptoms usually begin as transient bouts of motor tics involving the eyes, face, or head. Typically, the phonic symptoms begin somewhat later at 10 to 11 years of age. This is also the point at which the patient begins to have some awareness of the premonitory sensory phenomena. Tics increase during periods of stress and fatigue. Indeed any period of heightened emotionality, positive or negative, may prompt an increase. Typically, an individual's repertoire tics changes over time and this may be influenced by extrinsic events. For example, bouts of sniffing and throat clearing may emerge following a head cold and persist for long after other symptoms of the cold have passed. Sleep is associated with a marked reduction of tics (33).

Bouts of tics, and bouts of bouts of tics can last minutes to hours and can disrupt intentional behavior (81). These bouts of tics also occur in bouts. Over longer time scales of weeks to months a waxing and waning of tic symptoms is characteristic. Similar neurobiological mechanisms may underlie this fractal quality, and non-linear dynamical modeling suggests the presence of a deterministic process (81). Although a more or less stable and distinctive repertoire of tics emerges over time for a given individual, the range and variability of these symptoms are remarkable. Tic symptoms gradually worsen from the time of onset through early puberty with gradual improvement thereafter. Figure 3 presents the course of tic severity in a birth cohort of TS patients over the first two decades of life (67). In its most extreme form, TS is a lifelong condition that is chronically disabling in adulthood due to nearly constant volleys of motor and phonic tics that interfere with work tasks and verbal communication. Although rare, self-abusive tics such as hitting the face can also occur as can virulent coprolalia and copropraxia. The factors that govern the persistence of tics into adulthood or that prompt major exacerbations in adults are not well understood.

The factors that influence the natural history of TS have yet to be fully established but are likely to include gender specific hormonal factors, adverse prenatal and perinatal events, post-infectious autoimmune reactions, post-natal emotional stress and drug exposures, thermal stress, as well as comorbid medical and psychiatric conditions (see section on Epigenetic and Environmental Factors below).

Comorbidity. Tic symptoms are usually not the child's first or only difficulty. Attentional deficits, impulsivity, and motoric hyperactivity are commonplace, affecting 40 to 50% of all referred TS cases (19). When present, these symptoms are finely interwoven with the tic symptoms and can often have disastrous effects on peer acceptance, school performance, and parental supports (26,104). The etiological relationship between TS and the various forms of ADHD remains controversial. As discussed below, it is likely that these disinhibition syndromes likely share an overlapping set of susceptibility genes and environmental risk factors.

For more than 60% of TS patients, persistent obsessive compulsive symptoms appear a few years after the onset of the tic symptoms (19,79,88). Some portion of these individuals go on to develop full blown OCD and may experience enduring OC symptoms even though their TS symptoms have otherwise diminished. As noted below (see section on Genetic Factors), the OCD associated with TS may be the result of the same underlying genetic vulnerability as the tic symptoms. Tic-related OCD typically includes symptoms of aggressive, sexual and religious obsessions and related compulsions as well as obsessions of symmetry and exactness and associated arranging and counting compulsions (53). Although similar in many respects to other forms of OCD, tic-related OCD may be distinctive in terms of its earlier age of onset; the prominence of such symptoms as ritualized touching, tapping, and rubbing; and its relative refractoriness to serotonin reuptake inhibitors and responsiveness to augmentation with typical and atypical neuroleptics (57).

Many TS patients suffer from other comorbid psychiatric disorders including major depression and various anxiety disorders (18,19). The presence of a bipolar diathesis may herald a particularly problematic course (8, 47).

Epidemiology. Once thought to be a rare condition, the prevalence of TS is now estimated to be between 1 and 6 cases per 1,000 boys, and milder variants of the syndrome are likely to occur in 2 to 10 percent of the population (4,20,23, 70). Consistent with the course of illness described above, the frequency of TS among adults is reduced by at least an order of magnitude (68).

GENETIC FACTORS

The etiology of TS remains unknown. Twin studies indicate that genetic factors are likely to play an important role in the transmission and expression of TS and related phenotypes (45, 87,113). Specifically, monozygotic (MZ) twin pairs have been found to be highly concordant for TS (53%), and if other tic disorders are included, the overall concordance for the MZ pairs is substantially higher (75-90%). The concordance of same-sex dizygotic twin pairs is much lower (8% for TS or 23% if other tic disorders are included). These concordance figures are consistent with a complex genetic etiology with variable phenotypic expression.

In addition to other tic disorders, twin and family studies also provide evidence that some forms of OCD may be etiologically related to TS (78,79). The family genetic data have been less helpful in sorting out the relationship between TS and ADHD. In the largest series, ADHD among the first degree relatives was most strongly associated with the TS proband having ADHD as well suggesting that the vulnerability to ADHD symptoms is familial in those families (80). In addition, there is a greater than chance association of TS and ADHD among those first degree relatives who have TS. However, this association is not observed in the case of other first degree relatives who have lesser variants of TS.

Using mathematical models of genetic transmission and either a broad or narrow definition of the affected phenotype, the distribution of affected individuals within some well characterized families follows a pattern consistent with an autosomal dominant form of transmission with a polygenetic background (27,79). This finding, coupled with recent advances in human genetics, has led directly to the initiation of genetic linkage studies in an effort to identify the chromosomal location of the putative TS vulnerability genes. Thus far classical linkage studies within high-density families have not been successful in detecting genes of major effect (5). This has led to a re-examination of the patterns of segregation within families, and more recent analyses have suggested either an intermediate mode of inheritance or the presence of a polygenic background (43,112). As with other complex behavioral traits, permissive strategies such as studies of affected sibling pairs and other designs based on identity by descent within large population isolates, probably offer the greatest promise for identifying the chromosomal location of the TS susceptibility alleles.

Advances in molecular neurobiology have fostered efforts to examine the possible role of specificcandidate genes in TS. Numerous candidate genes, mostly involved in central monaminergic pathways and histocompatibility, are being characterized and there is limited evidence to support the role of some of these loci as genes of minor effect in TS (21,22,40). Most of these results, however, have not been confirmed by other investigators (9,30,76).

The search for the TS vulnerability genes has also naturally led to the study of epigenetic and environmental factors that influence early brain development as reviewed below.

EPIGENETIC AND ENVIRONMENTAL FACTORS

Studies of monozygotic twins indicate that epigenetic or environmental factors play an important role in mediating the extent to which the genetic vulnerability to TS and related disorders is expressed. Apart from gender (males are 3 to 4 times more likely than females to develop TS) and adverse perinatal conditions (within symptomatically discordant monozygotic twin pairs the more severely affected twin almost always has a lower birth weight than the less severely affected twin), other epigenetic factors that mediate expression have not been unequivocally identified (45,60,62). Candidiates include exposure to high levels of gonadal androgens and stress hormones during early CNS development (54,95) as well as conditions that effect the delivery of oxygen and nutrients to the developing fetus (77,114). Other candidates include exposure to chronic intermittent psycho-social stress in the postnatal period (12, 96); exposure to thermal stress (68); exposure to anabolic steroids (64); exposure to cocaine or other CNS stimulants (24, 35,73); and recurrent streptococcal infections (48,107).

Considerable interest in possible autoimmune forms of TS has emerged with the recognition that late neuropsychiatric manifestations of rheumatic fever can include tics, OCD, and motoric hyperactivity as well as the classical choreiform movements (105,109). The presence of increased titers of antibodies directed against a specific B cell antigen (75,108) as well as the presence of antineuronal antibodies that cross react with basal ganglia structures (98) have served to strengthen this line of investigation.

The successful identification of risk and/or protective factors may lead directly to early interventions that will limit, if not prevent, clinically significant forms of TS and OCD. This line of investigation is also complementary to efforts to define genetic susceptibility factors. For example, genetically determined host susceptibility factors are likely to contribute to an individual's risk to develop postinfectious autoimmune forms of TS. Similarly, the detection of epigenetic risk factors will sharpen the focus on specific neuroanatomic regions and systems that provide the neuroanatomic substrates for these disorders.

NEUROANATOMIC SUBSTRATES

There is a substantial body of data that implicate the basal ganglia and related cortical and thalamic structures in the pathobiology of TS and related disorders. These data include a modest collection of neuropathological and neurosurgical studies as well as a rapidly expanding literature based on in vivo neuroimaging techniques (for reviews see refs. 60,97).

Functionally, the basal ganglia are composed of fiber tracks that contribute to the multiple parallel cortico-striato-thalamocortical circuits (CSTC) that concurrently subserve a wide variety of sensorimotor, motor, oculomotor, cognitive and "limbic" processes (1,38). Leckman and co-workers have hypothesized that TS and etiologically related forms of OCD are associated with a failure to inhibit subsets of these CSTC mini-circuits (56,61). For example, it is anticipated that tics involving the face would be associated with a failure of inhibition of those mini-circuits that include the ventromedian areas of the caudate and putamen that receive somatotopic projections from the orofacial regions of the primary motor and premotor cortex (Figure 4). Using similar logic, we and others (56) have hypothesized that obsessions with aggressive and sexual themes would be associated with a failure to inhibit portions of the limbic mini-circuits, while counting obsessions and obsessive need for symmetry and exactness would be the result of a failure to inhibit some number of prefrontal mini-circuits.

Evidence to support these hypotheses includes data from both postmortem and neuroimaging studies. For example, functional magnetic resonance imaging (MRI) studies of tic suppression indicate that deficits in reducing basal ganglia activation or difficulties in activating key prefrontal corticostriatal inhibitory pathways may be associated with increased tic severity (85, Figure 5).

Figures 6A and 6B depict key aspects of this model in which TS symptoms are seen to be the product of several partially independent mechanisms that contribute to an imbalance between the direct and indirect pathways and the eventual disinhibition of specific thalamocortical pathways (61). As discussed below, the proposed mechanisms include a dopaminergic hyperinnervation of the striatum (69,117) increased density of dopamine D₂ receptors in the caudate (115,116), and a failure of activation within the subthalamic nucleus (2,3).

In addition, structural MRI studies have revealed abnormalities in the size and structural lateralization of basal ganglia structures in both children and adults with TS (84,102). These structural alterations are likely to have functional consequences as well as provide clues concerning the location and timing of events in the developing CNS that directly influence the emergence and course of TS and related symptomatology such as ADHD and the premonitory urges. The mechanisms responsible for these changes have not been elucidated but may involve perinatal hypoxia events and/or injury secondary to postinfectious autoimmune reactions.

The issue of right-left laterality has also focused attention on the potential role of the corpus callosum and the presence or absence of symmetrical or asymmetrical activation of the CSTC circuits. However, the available data on the corpus callosum are limited to structural studies and these data are controversial (6,82). Nevertheless lateralized systems within the CSTC circuits may well be a crucial determinant of tic persistence and severity.

It is intriguing to note that many of these same areas have been implicated in other PET studies of OCD and ADHD. OCD appears to involve metabolic changes in orbital frontal cortices and the caudate (7, 92), while ADHD may involve superior sensorimotor areas centered in the premotor cortices (118,119). Other aspects of the circuitry of the basal ganglia may provide important clues concerning the anatomic distribution of motor tics and the "choice" of obsessive themes frequently encountered in forms of OCD related to TS. Specifically, the unidirectional input from the amygdala to widespread areas of the nucleus accumbens and ventral portions of the caudate and putamen appears to overlap those areas most affected in TS related OCD (46).

In addition, reciprocal connections between midbrain sites (periaqueductal grey, substantia nigra, and the ventral tegmental area), portions of the hypothalamus and structures in the basal ganglia and amygdala are likely to play a critical role in the genesis and maintenance of the symptoms of TS. These connections may also contribute to the stress sensitivity (including sensitivity to thermal stress observed in a limited number of subjects) of the disorder and the more frequent expression of TS in males than females. Many of these structures contain receptors for corticosteroids and gonadal steroids and are likely to be responsive to alterations in their hormonal milieu.

While the neurophysiologic defect that underlies TS and etiologically related conditions remains unknown, a complete understanding of these disorders will likely illuminate mechanisms that regulate the activity of the multiple parallel CSTC circuits that subserve much of our normal cognitive, behavioral, and emotive repertoire. Advances in this area will also lead to the identification of specific neuroanatomical sites that may be crucially involved in the genesis of TS and OCD symptoms that may in turn be of value in isolating candidate genes that are uniquely expressed in these regions.

NEUROCHEMICAL MECHANISM AND NEUROPHARMACOLOGICAL EFFECTS

Among the most remarkable developments in the neurobiology of the basal ganglia and related structures over the past decade are the extensive immunohistochemical studies which have demonstrated the presence of a wide spectrum of differently distributed classic neurotransmitters, neuromodulators, and neuropeptides. The functional status of a number of these systems has been evaluated in Tourette's syndrome (see refs. 60,96 for reviews). Thus far, the mesencephalic monoaminergic (dopaminergic, noradrenergic, and serotonergic) projections that modulate the activity of the CSTC circuits have received the greatest amount of attention in TS and in related conditions such as OCD and ADHD. Despite their central role in the functioning of the CSTC circuits, the data concerning the roles of various excitory and inhibitory amino acid neurotransmitter systems remains inconclusive. Other evidence has focused attention on the tonically active, presumably cholinergic, interneurons and their key organizing role within the straitum (39).

Dopaminergic systems. Based largely on parallels between the tics, vocalizations and obsessive-compulsive behaviors seen in some patients with encephalitis lethargica, Devinsky was amomng the first to suggest that TS is the result of altered dopaminergic function in the midbrain (25). Altered central dopaminergic projections also play a crucial role in the model of pathogenesis depicted above (Figure 6B). Support for this model includes clinical trials in which haloperidol and other neuroleptics which preferentially block dopaminergic D₂ receptors have been found to be effective in the partial suppression of tics in a majority of TS patients (89,94). Second, tic suppression has also been reported following administration of agents that reduce dopamine synthesis (110). Third, increased tics have been reported following withdrawal of neuroleptics and following exposure to agents that increase central dopaminergic activity such as l-dopa and CNS stimulants including cocaine (35,49,73). Finally, investigators have also reported that TS patients have lower mean levels of homovanillic acid (HVA), a major metabolite of brain dopamine in CSF (97), but this has not been a consistent finding in larger well controlled studies (52).

Efforts to integrate these data have led to two major hypotheses concerning dopaminergic activity in TS: dopaminergic hyperinnervation of the striatum and supersensitive striatal D₂ receptors. Dopaminergic hyperinnervation would contribute to the imbalance between the direct and indirect pathways and is consistent with the pharmacological effects dopamine synthesis inhibitors and of D₂ blocking agents described above. Direct support for this includes preliminary evidence of increased levels of the dopamine transporter sites in the striatum of some TS subjects based both on in vivo neuroimaging techniques and studies of postmortem brain tissue (69,100,117).

The second hypothesis of striatal D₂ supersensitivity is also consistent with the effects of neuroleptics in reducing tics as well as the emergence of tardive tic syndromes following neuroleptic withdrawal. Preliminary PET studies of brain D₂ dopamine receptors also indicate that supersensitivity of D₂ receptors may be an important mechanism for some, but not all, individuals with TS (115,116). Additional imaging studies are needed to address fully the potential abnormalities of receptor number, affinity and distribution across the growing family of dopamine receptors over the course of brain development.

The role of dopaminergic systems in tic-related OCD has not been as carefully studied as it has in TS, despite both preclinical and clinical evidence implicating dopamine in some OC behaviors. The preclinical evidence consists of animal models for compulsive behavior that involve the induction of stereotypies by agents that increase dopaminergic activity, such as amphetamine, bromocriptine, the DA precursor L-DOPA, and the D2/D3 agonist quinpirole (see ref. 37 for a review). Complex repetitive behaviors that resemble naturally occurring OC symptoms have also been observed in subjects who abuse CNS stimulants such as cocaine (24). These so-called "punding" behaviors include repetitive cleaning, washing, grooming, and hoarding. In addition, there have been multiple reports of the occurrence of OC symptoms, alone or in association with tics, in post-encephalitic parkinsonian patients.

Direct evidence for a role of dopamine in mediating tic-related OCD includes the observations of McDougle and co-workers that many OCD patients with a personal or family history of a tic disorder are less responsive to serotonergic reuptake blocking agents than other OCD patients (71 and that neuroleptic augmentation of the reuptake blocker leads to further improvement in the OCD symptoms (72).

In sum, dopaminergic systems have been repeatedly implicated in the pathophysiology of TS and related disorders. The next decade should bring notable advances both in terms of the availability of highly selective pharmacological agents that may have therapeutic value as well as novel ligands that will support more informative neuroimaging studies.

Excitatory amino acid systems. As presented in figure 6A, glutamate is an excitatory neurotransmitter of both corticostriatal and thalamocortical neurons within the CSTC circuitry. Glutamatergic neurons in the subthalamic nucleus also play an important role in modulating the output of the basal ganglia by activating regions of the globus pallidus and the pars reticulata of the substantia nigra. Indeed, in our model of TS neuropathology a failure of glutamatergic subthalamic neurons to activate normally is the final common pathway leading to TS (Figure 6B). Direct support for this hypothesis is limited to the finding of reduced levels of glutamate in the globus pallidus in a small number of postmortem specimens (2,3). In our model this failure of subthalamic neurons to activate normally could be due either to pathology intrinsic to the nucleus or upstream effects leading to a pathological inhibition of the subthalamus. In any case, this is a key element in the proposed model and further studies are needed to confirm or reject this view.

Inhibitory amino acid systems. Although inhibitory amino acid neurotransmitters, particularly gamma aminobutyric acid (GABA), are the principal neurotransmitters of the straitopallidal and pallidothalamic projections that form two major portions of CSTC circuits, neurochemical studies of GABA levels in various brain regions, CSF, and plasma have not shown significant differences between TS and controls (2,3, 99). Similarly, although the long acting benzodiazepine clonazepam is frequently used in the treatment of refractory TS, the scientific data supporting its efficacy are limited to small-scale trials (36). Additional work is needed to document the efficacy of clonazepam or related compounds and to clarify the role of inhibitory amino acids in the pathobiology of TS and related conditions. Given the abundance of GABAegric neurons in the basal ganglia, refinements in magnetic resonance spectroscopy (MRS) should permit in vivo measurements of this neurotransmitter.

Endogenous opioid peptides. Endogenous opioid peptides (EOP) are localized in structures of the extrapyramidal system and are known to act as neuromodulators within GABAergic neurons. Two of the three families of EOPs, dynorphin and met-enkephalin, are highly concentrated and similarly distributed in the basal ganglia and substantia nigra. Dynorphin, along with substance P, colocalizes within GABAergic striatopallidal neurons within the direct pathway; while met-enkephalin colocalizes with GABAergic neurons in the indirect pathway. Significant levels of opiate receptor binding have been detected in both primate and human neostriatum and substantia nigra, and EOPs are likely to be importantly involved in the gating of motor functions and to interact with central dopaminergic pathways.

EOPs have been directly implicated in the pathophysiology of TS. Haber and co-workers who reported decreased levels of dynorphin A(1-17) immunoreactivity in striatal fibers projecting to the globus pallidus in postmortem material from a small number of TS subjects (41,42). These observations, coupled with the neuroanatomic distribution of dynorphin, its broad range of motor and behavioral effects, and its modulatory interactions with striatal dopaminergic systems, acting at both pre- and postsynaptic sites, suggest that dynorphin may play a role in the pathobiology of TS. The results of pharmacological challenge studies and small scale clinical trials of selective kappa receptor antagonists present a more complex picture (14). Further postmortem, in vivo neuroimaging, and psychopharmacological studies are warranted to clarify the meaning of these preliminary findings.

Cholinergic systems. Throughout the cortex and striatum local circuit cholinergic neurons are likely to play an important role in modulating the CSTC circuits. Tonically active neurons (TANs) located in the intermediate zones between the striasomal and matrisomal compartments of the striatum are likely to be cholinergic and to play an important in the coordination of goal directed behaviors (39). TANs are also known to be responsive to dopaminergic activation in conditioning paradigms.

The most compelling recent evidence of cholinergic involvement in the pathobiology of TS concerns the augmentation of the tic reducing effects of D₂ dopamine receptor blocking agents by nicotine gum and patches (90). The remaining neurochemical studies (including reports of normal levels of cortical and CSF acetylcholinesterase activity) and psychopharmacological data (involving clinical trials of choline, lecithin, and deanol and challenge studies with physostigmine) have been conflicting (101,103,111).

Noradrenergic system. Central noradrenergic pathways arising from the locus coeruleus and adjacent sites project widely and may modulate CSTC circuits involved in TS. This modulation could occur at the level of the cortex or be mediated by indirect effects on dopaminergic neurons in the ventral tegmental area (VTA). VTA neurons innervate mesolimbic areas (nucleus accumbens, olfactory tubercles, amygdala) as well as mesocortical sites such as the prefrontal cortex, receives noradrenergic afferents from the brain stem.

Evidence of noradrenergic involvement in the pathophysiology of TS is based on the well documented heightened stress responsivity seen in TS patients (13, 16), elevation of CSF and urinary norepinephrine levels (16,52), and the beneficial effects of alpha2 adrenergic agonists including clonidine and guanfacine in some, but not all, TS patients (17,34,55). Additional support has been based on the rebound exacerbations of tics in patients abruptly withdrawn from clonidine and from the finding of blunted growth hormone response to clonidine challenge in children and adults with TS (50,57,73).

Serotonergic systems. Ascending serotonergic projections from the dorsal raphe to the basal ganglia and cerebral cortex have been repeatedly invoked as playing a role in the pathophysiology of both TS and OCD. The most compelling evidence relates to OCD and is based largely on the well-established efficacy of potent serotonin reuptake inhibitors such as clomipramine, fluoxetine, and fluvoxamine in the treatment of OCD. However, these agents appear to have limited effectiveness in suppressing tics (91).

Studies of postmortem brain, CSF, and plasma studies of serotonin present a mixed picture. For example, preliminary postmortem studies in TS have shown that serotonin and the related compounds tryptophan (TRP) and 5-hydroxy-indoleacetic acid (5-HIAA) may be globally decreased in the basal ganglia and other areas receiving projections from the dorsal raphe (2,3). These postmortem findings are consistent with some previous observations of significantly lower levels of CSF 5-HIAA , plasma TRP, whole blood serotonin, and 24-hour urinary serotonin in TS patients compared to normal controls, but not others (see ref. 97 for a review). However, given the range of confounding factors that may influence these measures (the effects of chronic illness, medication, normal compensatory responses, among others), caution is warranted in interpreting these data.

Gender specific endocrine factors. Men are more commonly affected with TS than are women. This observation is true of both clinical cohorts and epidemiological samples (4,93). This observation has led us and others to hypothesize that androgenic steroids acting at key developmental periods (the prenatal period when the brain is being formed, adrenarche when adrenal androgens first appear at age 5 to 7 years, and puberty) may be involved in determining the natural history of TS and related disorders (83). This may be a direct effect of androgenic steroids or it may be due in part to the action of estradiol (formed in key brain regions by the aromatization of testosterone).

Normal surges in testosterone and other androgenic steroids during critical periods in male fetal development are known to be involved in the production of long-term functional augmentation of subsequent hormonal challenges (as in adrenarche and during puberty) and in the formation of structural CNS dimorphisms. In recent years several sexually dimorphic brain regions have been described, including portions of the amygdala (and related limbic areas) and the hypothalamus (including the medial preoptic area that mediates the body's response to thermal stress). These regions contain high levels of androgen and estrogen receptors and are known to influence activity in the basal ganglia both directly and indirectly. It is also of note that some of the neurochemical and neuropeptidergic systems implicated in TS and related disorders, such as dopamine, serotonin, and the opioids, are involved with these regions and appear to be regulated by sex-specific factors. Similarly, a number of auoimmune conditions also show altered sex ratios suggesting a role for gonadal steroids or other sex specific factors in these conditions.

Limited support for a role for androgens comes from case reports of tic exacerbation following anabolic steroid abuse (64) and the beneficial effects of antiandrogens in some patients with TS (86).

CONCLUDING REMARKS

Tic disorders are probably best seen as conditions in which there is a failure to inhibit bits of normal behavior. In TS fragments of movements, gestures, and sounds intrude and can interfere with goal directed behaviors. OCD and ADHD may be part of a larger spectrum of related cortical-subcortical disinhibitory syndromes. The relevant neurobiological substrates are likely to reside in those CSTC circuits that normally are engaged to learn, express, or to suppress behavioral routines. This area of study has benefited enormously from the characterization of these circuits in other animal species.

Systematic and sustained investigation of possible pathobiological mechanisms in TS over the past two decades has led to explicit models pathogenesis and to promising new treatment strategies. Advances in genetics and the neurosciences are continuing at a rapid pace and there is hope that the dynamic complexity of the neuroimaging, neurochemical, neuroendocrine, neuroimmunological, and neuropharmacological data in TS can be understood with greater clarity as the relevant pathobiological mechanisms are elucidated. However, as the hope for identification of a single gene mutation recedes, the diverse and multifactorial origins of tic disorders are being more fully appreciated. Susceptibility genes acting in concert with environmental events and epignetic factors are seen to shape and constrain the development of CSTC circuits. Building on the advances of the past decade and the synergistic potential among the various areas of active research, substantial progress can be anticipated in the identification of risk and protective factors that mediate the expression of TS. Advances in this area may lead to preventative interventions as well as novel approaches to treatment.

The need for suitable animal models for this disease is clear. Until such models are in place, the pace of research will continue to be slow. On the other hand, given the multifactorial nature of the disorder it may only be possible to develop partial models focused on particular mechanisms. Some of the more promising leads that could be developed into animal models include the effects on the structure and function of CSTC circuits of transient perinatal hypoxia and post-infectious autoimmunity to certain strains of beta hemolytic streptococci.

The continued development of in vivo neuroimaging techniques, from ligand-based studies to functional and spectroscopic MRI techniques, offers some of the most appealing studies for the next generation of patient-based TS research. For example, recent success in acquiring event-related fMRI data may permit a more accurate understanding of distributed processes that must occur in the moments before and following bouts of tics. Similarly, the application of neuroimaging techniques to document autoimmune-based inflammatory changes in volume or to monitor the effects of treatment may be a useful guide for clinical decision-making.

ACKNOWLEDGEMENTS

This work was supported in part by NIH grants MH44843, MH49351, MH00508, NS16648, HD03008, RR00125, & RR06022 (General Clinical Research Centers), MH30929 (Mental Health Clinical Research Center), and the Tourette Syndrome Association. The authors would also like to acknowledge the many important scientific and practical contributions made by Dr. Donald J. Cohen who initiated this program of research twenty years ago. Other investigators who helped to shape this research program include Drs. George M. Anderson, Mark A. Riddle, Phillip B. Chappell, Wayne K. Goodman, Paul J. Lombroso, Lawrence D. Scahill, and Heping Zhang.

published 2000