Tourette’s syndrome (TS) is a neuropsychiatric disorder, characterized by the presence of both motor and vocal tics. A tic is a sudden, rapid, recurrent, nonrhythmic motor movement or vocalization. Tics vary greatly between and within individuals. Most motor tics are brief, sudden and meaningless muscle movements, such as eye blinking, nose twitching or shoulder shrugging (simple motor tics). In contrast, complex motor tics appear more purposeful and involve several muscle groups. Examples include touching other people or objects, retracing steps when walking and various complex hand gestures. Similarly, vocal tics may be subdivided into simple and complex tics, ranging from meaningless sounds such as throat clearing, sniffing and barking to the sudden utterance of words, phrases and full sentences which may in a minority of cases include echolalia and coprolalia. Tic intensity can vary substantially, ranging from barely visible or audible tics to extremely forceful or loud expressions. Quite often, tics are mild, in which case they hardly attract attention from others and do not interfere with everyday life. On the other hand, powerful and frequent tics may severely interfere with everyday activities, including speech, driving and walking. In exceptional cases, tics may lead to physical injury, including joint dislocation and other tissue damage. Also, patients who display more severe or complex tics may be stigmatized as a result of the unusual, inappropriate and bizarre character of their tics. Most patients do not experience their tics as entirely beyond their control; many individuals describe premonitory urges preceding their tics, feelings that are momentarily relieved by the performance of tics and may be temporarily ignored by suppressing the tics.

The age of onset of tics is mostly between 2 and 15 years with a median of 7 (Bruun and Budman 1997). Facial tics are normally the initial symptom. Males are more commonly affected than females (Tanner and Goldman 1997). Movements generally decrease during sleep and may be suppressed for short periods while the patient is awake (Jankovic 1997). Tics typically occur in bouts during the day alternated with relatively tic-free periods within the course of a day (Peterson and Leckman 1998). Similarly, the course of tics over a period of months to years often waxes and wanes in severity (Coffey et al. 1994). Furthermore, the type of tics in an individual patient is typically variable over time, with some tics disappearing and new ones appearing in the course of time. With increasing age, however, symptoms tend to decrease in intensity and to show less variation over time regarding both severity and type of tics (Bruun and Budman 1997). TS is not rare; population studies estimate prevalence rates for TS between around 1 % (Robertson 2008).

One important feature of TS is its well-known association with a range of behavioural disorders and comorbid psychopathology, which may be more clinically relevant than the tics themselves (Mol Debes 2013). Attention-deficit/hyperactivity disorder (ADHD) is known to affect 50 % of referred patients (Olfson et al. 2011). Obsessive-compulsive symptoms constitute another common cophenomenon of the spectrum of tic disorders (Wanderer et al. 2012). Also, many children show significant problems with social functioning (Kurlan et al. 1996).

Although genetic factors are known play a pivotal role in TS, currently, our knowledge about the pathogenesis is still inconclusive. Structural and functional neuroimaging studies point to the involvement of pathways leading from cortex through basal ganglia to thalamus (Leckman 2002). The basal ganglia facilitate the ability to effectively switch between motor and mental behaviours, required for producing novel behaviour. Failures to do so may result in the repetitive production of stereotyped movements, thoughts or behaviours as in TS.

Over the past two decades a significant amount of research has been conducted on the possible role of infections and immune alterations in TS. The classically proposed model of pathogenesis of tic disorders in this research is analogous to Sydenham’s chorea (Swedo 1994). Per this model, tics and associated phenomena are thought to arise as a consequence of the immunological response to infections with Group A beta hemolytic Streptococci (GABHS). Antibodies directed against the streptococci are hypothesized to cross-react with structures of the central nervous system, subsequently leading to damage to these structures, which eventually is thought to result in tics and associated features. This supposed mechanism of autoimmunity is a classic model of molecular mimicry between host and micro-organism. Subsequent reinfections according to the presumed mechanism would lead to symptom exacerbations.

Several clinical observations have led to the hypothesis that Sydenham’s chorea might be a model for some types of TS and childhood-onset obsessive-compulsive disorder (OCD). First, it had been noted that patients with Sydenham’s chorea shared certain behavioural characteristics with patients with OCD and/or tic disorders, such as emotional lability, marked irritability, but also frank obsessive-compulsive symptoms (Swedo et al. 1993). Second, a substantial number of children with OCD were reported to show choreiform movements or tics (Steingard and Dillon-Stout 1992). In addition, in some carefully studied children with TS and/or OCD, an episodic course and/or abrupt onset of their symptoms seemed to be temporally related to signs of GABHS infections (Swedo 1994). Following these observations, case studies began to appear in the literature in the mid-1990s, in which children with OCD and/or tic disorders were described who suddenly demonstrated severe forms of tics and obsessive-compulsive symptoms and in whom a temporal relationship between symptom onset or exacerbations and GABHS or viral infections seemed apparent (e.g., Allen et al. 1995; Tucker et al. 1996). In some cases, streptococcal reinfections were associated with the reinduction of neuropsychiatric symptoms (Swedo et al. 1998).

Researchers of the National Institute of Mental Health (NIMH) subsequently proposed criteria to identify a putatively unique subgroup of children with TS and/or OCD, in whom symptom onset and/or exacerbations were abrupt, dramatic and temporally related to GABHS infections. They designated it by the term paediatric autoimmune neuropsychiatric disorders associated with streptococcal infections (PANDAS; Swedo et al. 1998). Research criteria have been outlined for the PANDAS subgroup, which require the presence of OCD and/or tic disorder, prepubertal symptom onset, sudden onset or episodic course of symptoms, temporal association between streptococcal infections and neuropsychiatric symptom exacerbations and associated neurological abnormalities (Swedo et al. 1998). What percentage of children with TS and/or OCD would meet criteria for PANDAS has not yet been investigated.

The concept of PANDAS strongly centres on Sydenham’s chorea as the putative disease model, including the suggested pathogenic role of autoantibodies that cross-react with brain antigens as a hypothesized consequence of structural homology with streptococcal antigens (Swedo 1994). Introduction of the concept of PANDAS has certainly done much to stimulate research and raise clinical awareness of the potential relevance of immune factors in tic and related disorders. Concerns regarding the validity of PANDAS have been expressed, however (Kurlan 2008; Shulman 1999; Hoekstra et al. 2002; Singer and Loiselle 2003). These include criticism about the vagueness of some of the criteria for PANDAS. For example, when should an exacerbation be considered abrupt? Or what time frame constitutes a temporal relationship between GABHS infections and symptom exacerbations? Or, what is sufficient proof of a GABHS infection? The diagnostic criteria for PANDAS are therefore not easy to apply in clinical practice as it is not straightforward to demonstrate a temporal association between GABHS infection and symptom onset and exacerbations, since streptococcal infections are fairly common in children in general, as are remissions and exacerbations in children with TS. We also lack a laboratory test for PANDAS. Case studies of patients with suspected PANDAS continue to appear in the literature (e.g., Chmelik et al. 2004; van Toorn et al. 2004; Gabbay and Coffey 2003; Storch et al. 2004; Maini et al. 2012; Das and Radhakrishnan 2012; Hachiya et al. 2013).

Some direct comparisons between PANDAS cases and non-PANDAS TS and/or OCD have also been published. In a prospective study, Kurlan et al. (2008) found that patients with PANDAS more often had exacerbations which were temporally associated (within 4 weeks) with a GABHS infection than non-PANDAS cases. Yet, the majority of the clinical exacerbations in PANDAS cases had no observable temporal relationship to GABHS infection. The authors concluded that GABHS infection is therefore not the only or even the most common antecedent event associated with exacerbations for patients fulfilling PANDAS criteria when followed over time. However, in a more recent longitudinal study, with a similar design of comparing tic exacerbation triggers between PANDAS and non-PANDAS cases, remarkably, occurrences of GABHS infections followed by a subsequent tic exacerbation were always in the non-PANDAS cases (even though also in the majority of non-PANDAS cases, exacerbations were not preceded by a GABHS infection). This latter finding challenges the validity of the clinical PANDAS criteria to discriminate between GABHS and non-GABHS related TS.

On the basis of concerns over the diagnostic criteria for PANDAS, a broader concept of childhood acute neuropsychiatric symptoms (CANS) has recently been proposed (Singer et al. 2012). The proposed CANS classification does not require association with a specific organism, limitation of symptoms to tics or OCD, a specific age range, or recurrence of symptoms. It does, however, require an acute dramatic onset, a comprehensive history and examination, and diagnostic evaluation. The usefulness of CANS as an entity remains to be investigated.

Rather than trying to identify PANDAS cases through directly implying an association of TS with infections, based on predefined PANDAS criteria, a scientifically perhaps better approach to demonstrate a relationship between TS and infections, would be to study the relevance of infections in unselected patients with TS. Interestingly, single time point studies have repeatedly found higher anti-streptococcal antibody titres (antistreptolysin O and antideoxyribonuclease B) in TS. Muller et al. (2000) found that 85 % of the subjects with TS vs 8 % of normal controls had elevated antideoxyribonuclease B levels. The same researchers subsequently showed the presence of increased titres against the streptococcal M12 and M19 proteins in patients with TS as compared with controls, while antibody titres against M1, M4 and M6 did not differ between the TS and control groups (Muller et al. 2001). Also, Cardona and Orefici (2001) reported significantly higher mean antistreptolysin titres in children with tics compared to control children, and found a positive correlation between antistreptolysin titres and severity of the tics as measured by the Yale Global Tic Severity Scale. Higher mean levels of anti-streptococcal antibody titres in patients with TS compared to controls have also been reported by others (Morer et al. 2006; Creti et al. 2004; Rizzo et al. 2006; Martino et al. 2011; Cheng et al. 2012), although there have also been negative studies, mainly in the USA, that failed to find increased single time point levels of anti-streptococcal antibodies in tic disorder subjects (Singer et al. 1998; Peterson et al. 2000; Morshed et al. 2002; Loiselle et al. 2003).

The cross-sectional design makes these findings somewhat hard to interpret, especially since antibody levels in these subjects were not assessed at the time of the first appearance of their tics, nor at a time of symptom exacerbation, but rather at an arbitrary point in time. Where could increased levels of anti-streptococcal antibodies point to? Are patients with tics more likely to encounter streptococcal infections? Or does their immune system differently respond to streptococci, which may include a more prolonged humoral response? There is indeed retrospectively collected data that indicates that patients with TS are more likely than controls to have experienced streptococcal infection prior to disease onset date (Mell et al. 2005; Leslie et al. 2008), although these findings were not confirmed in a more recent study (Schrag et al. 2009). However, we really need more large-scale prospective studies to fully understand the relevance of streptococcal infections in TS. Overall, the available longitudinal studies to date have been somewhat mixed. A first small-scale longitudinal study involving 47 patients with TS and/or OCD found that the association between symptom exacerbations and new GABHS infections among patients was no greater than that expected on the basis of chance (Luo et al. 2004). In another longitudinal study, however, positive correlations were detected between streptococcal titres and obsessive-compulsive severity rating changes in a subgroup of patients with large symptom fluctuations over time (Murphy et al. 2004). These subjects were also more likely to have elevated streptococcal titres in comparison with patients without such dramatic fluctuations. A subsequent longitudinal study in which 693 healthy children (ages 3–12 years) were enrolled who were studied monthly during 8 months revealed a strong relationship between GABHS throat culture and subsequent motor and behaviour changes, in the form of choreiform movements and hyperactivity (Murphy et al. 2007). In another longitudinal study, newly diagnosed infections were predictive of modest increases in future tic and obsessive-compulsive symptom severity, even more so when there were high levels of psychosocial stress (Lin et al. 2010). In contrast, Martino et al. (2011) who followed 144 patients with TS at 3-month intervals found that new infections did not predict clinical exacerbations.

Despite some inconsistency in these data, overall, an association between GABHS infections and TS is noticeable. The results of currently ongoing large-scale longitudinal studies on the association between infections and tic onset and tic exacerbations will have to be awaited before definitive conclusions can be drawn about the role of GABHS infections in TS.

Support for the potential of GABHS infections to induce tic-like behaviours also stems from animal work. It has been shown that exposure of rats to GABHS antigen leads to the production of antineuronal antibodies concomitant with the development of increased compulsive-like behaviour and motor disturbances (Brimberg et al. 2012). Previously it was shown that mice that were immunized and boosted with a streptococcal homogenate in Freund’s adjuvant exhibited motoric and behavioural disturbances in association with the presence of serum antibodies that were immunoreactive to several brain regions including globus pallidus and thalamus (Hoffman et al. 2004). This was not the case with mice that were boosted with Freund’s adjuvant only. These results are consistent with the hypothesis that immune response to streptococci can result in motor and behaviour alterations and suggest that anti-streptococcal antibodies cross-reactive with brain components may play a role in their pathophysiology.

Interestingly, a number of studies also seem to implicate a role for non-streptococcal infections. Case reports have suggested that both Borna virus (Dietrich et al. 2005) and mycoplasma pneumoniae (Muller et al. 2004; Matsuo et al. 2004) may be associated with tics and/or OCD. In a longitudinal study, our group reported an association between virally induced common colds and tic severity exacerbations in unselected patients with TS (Hoekstra et al. 2005). Also, a close temporal relationship between upper respiratory tract infection and the subsequent onset of tics in two adult patients was described, as was the case in three patients who developed a dystonic syndrome with presence of anti-basal ganglia antibodies after an upper respiratory infection (Edwards et al. 2004). Finally, a case–control study involving 32 patients with TS and 30 matched healthy controls demonstrated a significantly higher rate of elevated antibody titres against Chlamydia trachomatis (P = 0.017) in patients with TS as compared to controls (Krause et al. 2010). A trend toward a higher prevalence in the Tourette’s group was shown for Toxoplasma. Thus, evidence for a role of infections of infections in TS is certainly not confined to streptococcal infections.

In conclusion, relatively good evidence is available about a possible association of streptococcal infections with TS. The available evidence seems to suggest that patients with TS are more susceptible to infections. Most evidence points to more frequent streptococcal infections. Moreover animal studies have quite convincingly demonstrated that immunization with streptococcal antigens or transfer of antibodies evoked through immunization can indeed lead to tics. However, there is also some indication about the possible involvement of non-streptococcal infections, including common viral infections, in exacerbations of tic disorders.

Why would there be more frequent infections in patients with TS? Some studies have pointed to altered immune functioning as potential explanations. One investigation found that the level of IgA was decreased in patients with TS/OCD compared with control subjects, possibly explaining why the children with TS are more prone to upper respiratory tract infections (Kawikova et al. 2010); however, lowered IgA levels were not confirmed in a subsequent study, which, instead, indicated decreased serum IgG3, and possibly also IgM levels in patients with TS compared to controls (Bos-Veneman et al. 2011). In a more recent case–control comparison, patients with TS appeared to have a lower receptor expression of TLR4 after stimulation with lipopolysaccharide (meant to mimic a bacterial infection) and higher levels of soluble CD14 (Weidinger et al. 2014). These findings might represent an impaired activation of the innate immune response in TS, especially with regard to bacterial infection. The impaired response to pathogens could form an explanation for the higher susceptibility to infections. Another recent study found that the number of monocytes was significantly higher in patients with TS than in healthy controls, whereas concentrations of TNF-alpha, sIL1-ra, and sCD14 were significantly lower in patients with TS (Matz et al. 2012). The monocyte dysregulation in TS along with a possible dysbalance in innate immunity could be another important lead towards explaining why patients with TS may be predisposed to infections.

Do infections induce autoimmunity, causally related to the occurrence of tics, or is the association of TS with infections just an epiphenomenon? The presence of autoantibodies that react with parts of the brain thought to be involved in tic disorders is potentially a strong line of evidence in favour of the autoimmune hypothesis of TS. Unfortunately, as yet no standardized methodology of assessing anti-basal ganglia antibodies is available. Rather, different research centres have all used different methods, which make results difficult to compare. For example, to investigate whether a subject’s blood may contain antibodies that react with brain antigens, different substrates of brain tissue have been utilized across studies, ranging from immortalized neuronal cell lines to animal and human brains, which have, moreover, been prepared in different ways. Also, methods of assessing antibody binding to brain tissue differed between centres. Earlier studies used immunofluorescence microscopy, whereas later studies made use of enzyme-linked immunosorbent assays (ELISA).

So far, several research groups have reported the increased presence of these antineuronal antibodies in sera from patients with TS, compared to healthy controls (Kiessling et al. 1993; Murphy et al. 1997; Singer et al. 1998, 1999; Laurino et al. 1997; Morshed et al. 2002). Kiessling et al. (1993) were the first to assess antineuronal antibody status in children with recent onset of movement disorders (TS, motor and/or vocal tics, chorea, or choreiform movements), compared with a group of children referred for evaluation of ADHD, behaviour disorders, and learning disabilities who did not show signs of a movement disorder. They applied an indirect immunofluorescence technique, with unfixed frozen human caudate nucleus sections as antigenic substrate, using undiluted sera and fluorescein isothiocyanate-labeled secondary antibody directed against human IgG, and found 44 % of children with a movement disorder to be strongly positive for antineuronal antibodies versus 21 % of the control group. An additional study found positive antineuronal staining in 39 % of children with tic and/or obsessive compulsive disorder as compared to 24 % of healthy control children (Murphy et al. 1997). Subsequent studies using ELISA against either an immortalized neuronal cell line (Laurino et al. 1997; Singer et al. 1999), human basal ganglia (Singer et al. 1998), neuronal antigens (Cheng et al. 2012), or rat brain (Morshed et al. 2002; Yeh et al. 2006) in general confirmed the increased levels of serum antineuronal antibodies in patients with TS. Presence of these antibodies in sera of tic disorder patients fulfilling PANDAS criteria has also been found with high specificity and sensitivity when compared to control groups (children with neurological disease, with autoimmune disease, or with uncomplicated streptococcal infection) using both ELISA and Western blotting (Church et al. 2004). Common binding was observed to basal ganglia. An Italian study that used indirect tissue immunofluorescence identified anti-basal ganglia antibodies in almost two thirds of PANDAS cases and in less than 10 % of children with uncomplicated active streptococcal infection (Pavone et al. 2004). Investigation of oligoclonal bands of immunoglobulin G in cerebrospinal fluid in 21 patients with TS showed that 38 % of the patients exhibited pathological bands indicating a humoral immune response in the central nervous system and pointing to involvement of immunological mechanisms in TS pathology (Wenzel et al. 2011).

An interesting piece of evidence for the contribution of GABHS-induced antibodies to TS was provided by a study that tested the antibody response of tic patient sera to a representative panel of recombinant GABHS antigens (Bombaci et al. 2009). Sera from children with neither tic disorder nor overt GABHS infection were also analysed. The protein recognition patterns of these two sera groups were compared with those obtained using sera from children with GABHS-associated pharyngitis. This comparative analysis identified 25 antigens recognized by sera of the three patient groups and 21 antigens recognized by tic and pharyngitis sera, but poorly or not recognized by sera from children without tic. Interestingly, these antigens appeared to be, in quantitative terms, more immunogenic in tic than in pharyngitis patients. Additionally, a third group of antigens appeared to be preferentially and specifically recognized by tic sera. The authors concluded that this may be relevant in the context of one of the hypotheses proposing that GABHS antigen-dependent induction of autoantibodies in susceptible individuals may be involved in the occurrence of tic disorders.

There have, however, also been a number of negative studies, failing to find differences in antineuronal antibody status between patients with TS or PANDAS versus healthy controls (Brilot et al. 2011; Morris et al. 2009; Singer et al. 2005a); another study failed to find associations between clinical exacerbations of patients with TS and antineuronal antibodies (Singer et al. 2008).

It should be noted that as yet, no single neuronal antigenic structure has been identified as target for the putative antineuronal antibodies. Studies aimed at identifying neuronal antigens have indicated the identification of a candidate auto-antigen, pyruvate kinase by two independent research groups (Kansy et al. 2006; Dale et al. 2004), although binding to pyruvate kinase was not confirmed in a later study (Gause et al. 2009). Another study demonstrated that antibodies in patients with PANDAS reacted with the neuronal cell surface of the caudate-putamen and signal neuronal cells by inducing calcium-calmodulin dependent protein (CaM) kinase II activity. Depletion of serum IgG abrogated CaM kinase II cell signaling and reactivity of cerebrospinal fluid was blocked by streptococcal antigen N-acetyl-beta-d-glucosamine (GlcNAc; Kirvan et al. 2006). A recent study indicated that sera from patients with PANDAS specifically targeted dopaminergic neurons by inducing inhibitory signaling of the dopamine 2 receptor on transfected neuronal cells (Cox et al. 2013). However, binding to the dopamine receptor was not present in the sera of a series of patients with PANDAS or TS in two other studies (Dale et al. 2012; Morris-Berry et al. 2013).

Support for a pathogenic role of antineuronal autoantibodies in the disease process stems from studies in which animal models were developed to investigate whether serum or purified IgG from patients with TS can induce tic-like behaviour in rats (Hallett et al. 2000; Taylor et al. 2002). In these studies, serum or IgG was microinfused through cannulas placed in regions of the neostriatum known to induce stereotypies, after which the rats were observed for development of movements or utterances. Hallett et al. (2000) infused dilute serum from five patients with TS, with high antibody titres against human neuroblastoma, bilaterally into the ventral striatal region of the rat. Results showed a significant increase in tic-like behaviours (e.g., licks and forepaw shakes) and episodic utterances in the TS group, which was not observed when sera from healthy controls were microinfused. Taylor et al. (2002) infused serum from 12 patients with TS, who had high antibody titres against rat striatum, bilaterally into a different brain area, the ventrolateral striatal region. This led to a significant increase of oral stereotypies over a 5-day period of observation. However, these results were not confirmed in two subsequent studies aimed at replicating earlier findings about induction of stereotypic movements in rodents through passive transfer of antibodies by means of striatal microinfusions (Loiselle et al. 2004; Singer et al. 2005b). However, in a more recent study, passive transfer of GABHS-induced immunoglobulin G in mice has been shown to lead to tic-like behaviours (Yaddanapudi et al. 2010). Also, IgG from GABHS-exposed rats was demonstrated to react with D1 and D2 dopamine receptors and 5HT-2A and 5HT-2C serotonin receptors in vitro, while in vivo, IgG deposits in the striatum of infused rats colocalized with specific brain proteins such as dopamine receptors, the serotonin transporter and other neuronal proteins (Lotan et al. 2014).

In conclusion, the assessment of serum autoantibodies has appeared to be a somewhat conflicting research area in tic disorders, with a number of potentially promising leads towards characterizing antigenic structures at a molecular level, but also a fair amount of negative studies. Animal models, however, have quite convincingly pointed to the pathogenic relevance of infection-induced antineuronal antibodies. Obviously, standardization of assays in human sera should be next important steps to further our understanding of the role of antineuronal antibodies in TS.

A number of studies have indicated increased activity of the immune system in patients with TS. One marker of cellular immune activation is the increased degradation of tryptophan via the kynurenine pathway, leading to elevated plasma levels of kynurenine and subsequent metabolites (Heyes 1996; Meyer et al. 1995). Upregulation of the kynurenine pathway can be induced through increased activity of Indoleamine 2,3-dioxygenase (IDO), which is sensitive to Interferon-gamma, a major cytokine of cellular immunity, thus, the increased levels of kynurenine may possibly be reflecting immune activation. Conversion of tryptophan to kynurenine can be triggered by GABHS, as has been shown in vitro (Murr et al. 1997). Some authors also found that kynurenine increases tic-like behaviour in an animal model of TS: in mice, head-shakes which had been induced by the 5-hydroxytryptamine receptor agonist dimethoxyiodo-phenyl-aminopropane were potentiated by administration of kynurenine (McCreary and Handley 1995).

In an initial small-scale study (involving seven patients with a tic disorder versus controls) the serum kynurenine level was found to be clearly increased in all seven patients whereas serum tryptophan was normal (Rickards et al. 1996). In a subsequent larger scale study, involving 72 patients with TS and 46 matched controls, again, plasma kynurenine levels were found to be significantly elevated (Gaynor et al. 1997). We know of one other independent report of increased plasma kynurenine (Chappel et al. 1995). Interestingly, both case–control studies (Rickards et al. 1996; Gaynor et al. 1997) reported a significant positive correlation in patients with TS between levels of kynurenine and neopterin. Neopterin is a marker of cellular immunity, which is, like IDO activity, induced by cytokines. Increased levels of neopterin therefore support possible involvement of (auto)immunity. Increased levels of neopterin in patients with TS compared to controls were confirmed in two later studies (Hoekstra et al. 2007; Matz et al. 2012). Another piece of evidence for increased immune activation in patients with TS was provided by the finding of significantly elevated serum levels of vascular cell adhesion molecule-1 and E-selectin in patients with TS, compared to controls (Martino et al. 2005).