Movement disorders in patients with Rett syndrome: A systematic review of evidence and associated clinical considerations

Aim This systematic review identified and thematically appraised clinical evidence of movement disorders in patients with Rett syndrome (RTT). Method Using PRISMA criteria, six electronic databases were searched from inception to April 2021. A thematic analysis was then undertaken on the extracted data to identify potential themes. Results Following the thematic analysis, six themes emerged: (i) clinical features of abnormal movement behaviors; (ii) mutational profile and its impact on movement disorders; (iii) symptoms and stressors that impact on movement disorders; (iv) possible underlying neurobiological mechanisms; (v) quality of life and movement disorders; and (vi) treatment of movement disorders. Current guidelines for managing movement disorders in general were then reviewed to provide possible treatment recommendations for RTT. Conclusion Our study offers an enriched data set for clinical investigations and treatment of fine and gross motor issues in RTT. A detailed understanding of genotype–phenotype relationships of movement disorders allows for more robust genetic counseling for families but can also assist healthcare professionals in terms of monitoring disease progression in RTT. The synthesis also showed that environmental enrichment would be beneficial for improving some aspects of movement disorders. The cerebellum, basal ganglia, alongside dysregulation of the cortico‐basal ganglia‐thalamo‐cortical loop, are likely anatomical targets. A review of treatments for movement disorders also helped to provide recommendations for treating and managing movement disorders in patients with RTT.

As a complex neurodevelopmental disorder that begins in early childhood, Rett syndrome (RTT) presents with a range of symptoms, including autonomic, gastrointestinal, and neuropsychiatric disturbances. Movement disorders are a significant clinical concern, and their broad spectrum makes treatment and management in RTT challenging. Some of the abnormal movement disorders in RTT, such as hand stereotypies and gait disorders, form part of the essential diagnostic criteria for classical RTT alongside other supportive criteria such as bruxism and abnormal muscle tone. 1 Emotional, behavioral, and autonomic dysregulation (EBAD) can emerge with a wide range of symptoms and makes the treatment of RTT challenging. We have previously suggested that the behavioral component of EBAD can be exacerbated by abnormal movement behaviors. 2 Behavioral difficulties also seem to be associated with worse outcomes in patients with RTT. 3 Managing the behavioral and emotional components of EBAD by targeting specific movement behaviors in patients with RTT could potentially improve the quality of life (QoL) in this patient group by providing potentially important clinical information to understand the patients' needs better. Therefore, it is essential to systematically review movement disorders in patients with RTT because it would allow further exploration of the broad nature of movement disorders in patients with RTT.
Recent studies have explored instruments to evaluate gross motor and musculoskeletal deficits in patients with RTT 4 and provided perspectives of hand functioning in females with RTT. 5 However, while movement and motor disorders have been described in syndromic autism using systematic methods 6 and in RTT via a narrative review, 7 as far as we are aware, no study has undertaken a systematic review of movement disorders in patients with RTT with the intention of identifying themes to further understand the clinical impact of this movement impairment and whether the emerging information would be clinically relevant and useful when assisting in the management of patients. The purpose of this review was, therefore to: (i) undertake a robust systematic review of studies examining movement disorders in patients with RTT; (ii) use a thematic analysis approach to identify potential themes; and (iii) determine whether the information from these themes could be adopted by clinicians, therapists, and other healthcare professionals to improve the QoL of this patient group and disseminate these findings to the broader RTT community.

Search strategy Primary search strategy
The methodology for the systematic review followed the PRISMA criteria. 8,9 Two authors (J.S. and E.L.) independently and blindly searched the following databases: PubMed, Scopus, Cochrane, Psy-cINFO, Embase, and Web of Science in February 2021. As described, a truncation symbol (*) was used to capture as much of the literature as possible. To ensure that the primary searching captured as much literature as possible relevant to movement disorders in patients with RTT, the search strategy was focused on stereotypies. We reasoned that because nearly all patients with RTT have stereotypies 10-12 that co-occur alongside other abnormal movement disorders, the search strategy would be expected to detect most of the studies of abnormal movements in patients with RTT. To extend this primary search strategy, the first author (J.S.) searched the references from the reference list (snowballing) of studies to see whether any more eligible studies regarding movement disorders in RTT could be traced. Snowballing is a useful strategy for extending systematic reviews ensuring the best possible coverage of the literature. 13 To reduce search strategy bias, we used the principles adopted for our previous evidence synthesis. 14 First, both the first (J.S.) and second (E.L.) author independently undertook the PRISMA systematic review in a blinded manner. Second, the consensus agreement of eligible articles was based on an agreement between J.S. and E.L., and if a consensus could not be reached, the senior author (P.S.) was consulted.

Secondary searches
To supplement the primary search strategy, the first author (J.S.) also conducted an additional search of the PubMed, Scopus, Cochrane, PsycINFO, Embase, and Web of Science databases in April 2021 with the specific search terms of dystonia, Parkinsonism, bruxism, spasticity, tremor, and ataxia in RTT. This secondary search was also reviewed by the second author (E.L.) and a consensus agreement was reached on the additional articles that were included. A scoping review on the treatment of movement disorders was performed by the first author (J.S.) and reviewed by the other authors.

Search terms Primary search terms
The search of the databases used the following keywords: (Rett syndrome OR MECP2) AND (stereotypies*) Population characteristics All records within the databases that reported studies in RTT were searched.

Intervention
These included all records that mentioned or reported on movement disorders/impairments.

Eligibility criteria
The following eligibility criteria were used.
Inclusion criteria: • Complete records/articles in peer-reviewed academic/scientific journals and electronically available. • All investigations/reports performed in humans.
Exclusion criteria: • Records/articles not in English language and not available using electronic sources. • Review articles, book chapters, single case reports/studies, commentaries, conference abstracts, dissertations, letters to the editor, clinical trial protocols, and preprints.
Extraction of data and thematic analysis Data extraction and thematic analysis were performed as previously described. 14 The first author (J.S.) did the manual coding for the thematic analysis, which was independently reviewed by the second author (E.L.). A consensus on the themes was then reached, and the final themes that emerged were based on group consensus between all of the authors of the study. Microsoft Excel software 2016 was used to show the frequencies of the themes that arose.

Results
The PRISMA (Supplementary Information S1) identified 690 records, and, after duplicates were removed, 230 records were screened. Following the screening process that excluded 34 records, 196 records were assessed against the eligibility criteria. This procedure eliminated 171 records, and 25 articles were eligible. After consensus agreement between the first (J.S.) and second (E.L.) authors, a further seven articles were deemed eligible, and one study 15 was also included after snowball searching. The secondary screening of records within the databases using specific search terms identified 10 additional articles (Supplementary Information S2). In total, 43 full-text articles were included in the analysis. The data extraction from these 43 articles is shown in Table 1.

Characteristics of the eligible articles
As presented in Table 1, the 43 studies captured information across the movement disorder ecosystem in patients with RTT. About 42% of the studies used video observations to assess movement disorders. 12,[16][17][18][19][20][21][22][23][24][25][26][27][28][29][30][31][32] Others used questionnaires as the primary assessment method 11,[33][34][35][36] or as an adjunct with other methods. 18 The sample was also diverse, from an analysis of 1074 patients with RTT from the RTT Natural History Study, 10 to a study exploring a few cases. 37 The data obtained allowed a thematic analysis to be performed and are presented in the next section.
A thematic analysis of the studies When extracting data from the 43 eligible articles, six themes emerged from the thematic analysis. The themes encompassed different aspects of the movement disorder ecosystem in patients with RTT and, where appropriate, subgroups of movement disorders within the themes have been indicated. Some of the themes, such as those related to underlying neurobiological mechanisms, QoL, and treatment of movement disorders, had the lowest frequency and underscores the importance of much-needed work in these specific areas. Nonetheless, 'Clinical features of abnormal movement behaviors' emerged as the most prominent theme followed by 'Mutational profile and its impact on movement disorders'. The relevance of these themes is described below and their frequency is shown in Fig. 1.       Midline hand wringing and washing-like movements was the most common hand movement • The second most common stereotypy with hand movements was bruxism • The stereotypies of hair pulling, bruxism, and cervical retropulsion was more common in group 1 (mutation positive) • It was also indicated that mutationpositive patients had more varied stereotypies and these tended to reduce after 10 years of age 29   Thirty-two patients aged between 30 months and 28 years • Forty-one patients were seen and data were available for 32 patients • Patients also met the diagnostic criteria for RTT Video observations and MBA • The most frequent motor abnormalities noted in this cohort were stereotypies and gait abnormalities • Bruxism, abnormal eye movements, and dystonia were also observed • Drooling was also common (seen in 75% of patients) • Young patients were noted to have more hyperkinetic movements than older patients 33    • The study showed that iron accumulation in brain regions was associated with the severity of dystonia in patients with RTT • Dystonia assessment scales indicated that the severity was more pronounced in patients older than 10 years • It was suggested that the increased iron deposition in dopaminergic networks and gray matter could account for the age-related changes in the severity of dystonia 41 Saikusa et al.
One hundred females aged between 1 and 43 years (mean AE SD: 14.5 AE 11.2 years) • Diagnosis was based on genetic and clinical diagnostic criteria • Of the 100 patients, 86 had typical RTT and 14 had atypical RTT • Clinical review of information from the Japanese RTT database • Genetic testing including whole exome sequencing was performed in some patients • The study showed that walking in all age groups was associated with the ability to form meaningful words • In particular, the acquisition of words was associated with ambulatory ability after 10 years of age • The authors concluded that patients who can walk can be predicted to form meaningful words 43   • Maternal education was suggested to be a driver for increased focus to access dental services • Some patients were also said to cope with more invasive procedures such as extractions without requiring sedation 44  The clinical severity score was used to assess severity of study participants • The study findings showed that the clinical severity increases with age • Ambulation, hand function, and onset of stereotypies were also associated with disease severity; however, regardless of the initial severity, the progression of RTT becomes worse with age • It was indicated that X-chromosome inactivation does not fully account for the clinical severity seen in patients 47 Psoni et al.   age. 34 From a developmental perspective, hypotonia is noted to be present in children with RTT during the first year after birth 38 and present in about 50% of children younger than 5 years but seems to get less as the child becomes older when hypertonia and dystonic type rigidity become established. 39 This is clinically interesting because another study has shown that as the disorder progresses, features such as dyspraxia are hindered by poor muscle tone and become more prominent when complex tasks are required to be performed. 27 In another study, the clinical severity of dystonia was associated with iron accumulation in brain regions of patients with RTT. 40

Parkinsonian-like features/tremors and ataxia
Parkinsonian rigidity is prevalent in patients with RTT, accounting for about 84% in one study. 15 It tends to have an early onset (3 years of age) and increases with age. The rigidity usually starts with the ankle, and nonambulatory patients are more severely affected. In patients who were ambulatory, freezing of gait appeared to be the most characteristic movement behavior. 16 Patients who could walk were also more inclined to have the ability to form meaningful word acquisitions. 41 Interestingly, video and treadmill walking observations were also helpful in assessing ataxic gait. These features would be useful in discriminating between traditional 'Parkinsonian' signs observed from freezing behavior to those of cerebellar origin such as ataxia.

Stereotypies
Clinical severity is worse in patients with decreased hand function, 10 and increases in bradykinesia and hypertonia are thought to have a role in hand functioning. As suggested by others, 17 hand stereotypies have a heterogenous phenotype. Hand mouthing and clapping/tapping were more common than hand wringing/washing, 10 but hand wringing was the most common hand stereotypy in another study. 34 Moreover, some stereotypies such as hand 25 and bruxism 26 disappear at night. One clinical feature that was highlighted was unilateral hand movements associated with centrotemporal spikes on the electroencephalogram (EEG). These were suggested not to be epileptic but rather somatosensory or motor potential spikes originating from different brain regions. 20 Abnormal EEGs in RTT evoked by hand clapping have also been reported elsewhere. 42 These cases are useful because they show that although abnormal movements may evoke centrotemporal spikes on the EEG, they do not necessarily indicate that this is caused by epilepsy and highlights the importance of routine monitoring of movement disorders in this patient group.
Other stereotypies are also prevalent in patients with RTT, and these too can emerge differently, such as bruxism 26,28 and shifting of the weight. 34 Individuals with bruxism were also more inclined to access dental services 43 ; however, the risk of bruxism appears to decline with age. 44 Oral motility was also affected by dystonic and dyskinetic movements. 45 Other evidence suggests that while stereotypies and hand functioning remained stable over time, musculoskeletal problems became worse, and this feature continues into adulthood. 11 Regardless of the initial severity, ambulation, hand functioning, and onset of stereotypies is suggested to get worse as the disorder progresses. 46

Other features
Another study has also highlighted tongue protrusion (62%) and postural stiffness (58%) in patients, 29 while recognizing patterns in eye movements could assist in predicting the onset of stereotypic behaviors. 31 There is overlap between this theme (theme 1) and the mutational profile of patients and how this profile affects movement disorders (theme 2). Frequent comorbidities appear to be epilepsy, weight, gastrointestinal, and bowel movements in those with truncating mutations or large deletions. 36 The mutational profile of patients with RTT and its impact on movement disorders is described in theme 2.

Theme 2: Mutational profile and its impact on movement disorders
This theme provided an enriched data set that allows deeper exploration of the impact of the genotype-phenotype relationship on movement disorders in RTT.

Dystonia
When evaluating gross motor disturbances, focal dystonia is more common in patients with missense mutations than with truncating mutations. In contrast, individuals with truncating mutations tended to have a higher frequency of generalized dystonia. Additionally, the appearance of tremor can also be different. Overall, the frequency of dystonia and rigid-akinetic syndrome was also higher in patients with truncating mutations. 26

Parkinsonian-like features/tremors and ataxia
When assessing Parkinsonian rigidity scores in RTT, there was no difference between patients with truncating and missense mutations 15 ; however, patients who were less able to walk presented with worse scores on the rigidity scale. Kinetic tremors were more common with missense mutations when compared with truncating mutations where postural tremors were more abundant. 26 In the Greek sample of patients with classical RTT, the frequency of spasticity-dystonia and tremor-ataxia was higher in individuals who were mutation positive. 47

Stereotypies
One study has suggested no association between hand stereotypies and genotype, 12 while other evidence has indicated that patients with the p.R168X mutation had the poorest hand function. 23 The number of stereotypies in individuals with missense or truncating mutations were similar; however, rigidity and dystonia were more common in individuals with truncating mutations. 22 Individuals with R133C, p. R294X, p.R306C mutations, or a C-terminal deletion were also said to have better mobility and complex motor skills when 13 years or older and better complex motor skills <13 years of age. 27 Hair pulling, bruxism, and cervical retropulsion were more common in mutation-positive patients. 28 When individuals with typical and atypical RTT were compared, shifting weight on one leg, followed by handwringing and bruxism, were the most common stereotypies in individuals with typical RTT. On the contrary, bruxism and lip protrusion were the most common in patients with atypical RTT. 34 Clinical information from the International Rett Syndrome Phenotype Database (InterRett) and the Australian Rett Syndrome Database showed that patients with R255X and p.R168X mutations tended to be diagnosed at a younger age, 35 with developmental regression starting sooner. This is relevant because the study also indicated that an older age of diagnosis, especially in individuals with the p.R133C or p.R294X mutations, was linked to delayed loss of speech and hand functioning. 35 Individuals with C-terminal deletions were deemed to be less clinically severe and more likely to have learned to walk, 48 while truncating mutations were associated with worse clinical outcomes. 49

Other features
In a survey of patients from the British Isles RTT database, individuals with truncating mutations or large deletions had greater clinical severity. 36 These patients can also be clinically characterized as not being able to walk, not having actually learned to walk, and having the worst form of gross motor dysfunction. 50 This theme underscores the influence of symptoms and stressors on movement disorders. The environment is important in managing negative behavior associated with self-injury in RTT. 18 High-stress conditions can also lead to visible signs of displeasure in the faces and the vocalization of patients with RTT 19 and could have implications on how movement behaviors can be affected by these signs. Evidence further suggests that environmental and sensory influences might have an important role in reducing the incidence of secondary disabilities associated with abnormal movements in RTT. 33 In this questionnaire-based study, factors that led to a decrease in stereotypies were somnolence, pleasure, concentration, and food. These factors were said to lower the incidence of secondary skin issues and joint contractures associated with stereotypies. This premise was also supported by others who have suggested that constructive engagement through promotion of different environmental and sensory factors might have a positive influence on managing stereotypic behavior in patients with RTT. 23,24,37,51,52 However, there are data to also suggest that environmental modifications have limited impact on the behavior of hand stereotypies. 30 The neurobiological relationships between stressors and movement disorders remain unclear; however, areas within the basal ganglia and the medial prefrontal cortex are involved in regulating motor outputs based on emotional states. 53,54 Stress can also have a profound impact on dopaminergic networks, and, alongside the hypothalamic-pituitary-adrenal axis, are vital for responding to rapidly changing environmental cues. 55,56 In RTT, these pathways could be negatively impacted by stress, anxiety, or depression, resulting in a deterioration of fine and gross motor function.

Theme 4: Possible underlying neurobiological mechanisms
The neurobiological mechanisms of movement disorders was another theme that emerged from eight studies. In RTT, the underlying mechanisms appear to be diverse.

Dystonia
Increased iron accumulation in dopaminergic networks and gray matter was suggested to be correlated with the severity of dystonia in patients with RTT. 40

Parkinsonian-like features/tremors and ataxia
Earlier studies have suggested the involvement of neural mechanisms associated with the reduction of caudate heads, thalami, and presynaptic abnormalities within nigrostriatal pathways. 25 As the disorder progresses, deterioration of subcortical, cerebellar, and spinal cord networks causes ataxia and tremor and impedes the ability to walk. 57

Stereotypies
The phenotype of movement disorders might also present with different neural mechanisms. In patients with RTT, stereotypic behaviors are not observed during freezing of gait, which could suggest that the neural mechanisms underlying these processes could be independent from one another. 16

Other features
Some studies have suggested that the cortico-basal ganglia-thalamocortical (CBGTC) loop is implicated in the evolution of movement disorders in RTT. 21,34 Brainstem and cerebellar structures have also been linked when explaining the phenotype of movement disorders. 22,29 Theme 5: QoL and movement disorders This theme was relevant in the context of this evidence synthesis because it explored how psychosocial aspects of movement disorders can affect QoL. In a longitudinal 2-year follow-up of 260 patients enrolled in the RTT Natural History Study, patients with worse motor functions and an earlier onset of stereotypies had higher QoL scores on psychosocial functioning. 58 The authors reasoned that patients with more severe motor abnormalities were less likely to have negative behaviors (aggression and self-injurious behavior) that would have otherwise impacted negatively on their psychosocial QoL. This finding is relevant because it suggests that when assessing severe clinical impairments in RTT under the umbrella of movement disorders, the changes in QoL are not uniform and the outcomes can be variable.

Theme 6: Treatment of movement disorders
This theme emerged from one study. In a small case series of three patients with RTT (age range: 9-20 years), acute life-threatening episodes caused by dystonic movements could be managed in two patients using trihexyphenidyl. 59 Movement disorders can emerge with different clinical phenotypes spanning across dystonia, Parkinsonism, bruxism, spasticity, tremor, and ataxia. While there is useful information on clinical practices for managing pediatric movement disorders, 60 improved diagnostics for dystonia, 61 and guidance on the management of stereotypies, 62,63 there is little information on the treatment of movement disorders. In the following section, general treatments for movement disorders and their implications for patients with RTT will be discussed.

Discussion
The current evidence synthesis evaluated a range of clinical and neurobiological features across the spectrum of movement disorders in patients with RTT. It aimed to identify themes and suggested how the information in these themes could be extrapolated and used by clinicians and other healthcare professionals to inform the wider RTT community. The main findings identified in this study were: (i) further knowledge and learning of the clinical features of movement impairments in patients with RTT; (ii) a synthesis of information regarding genotype-phenotype relationships of movement disorders in RTT; (iii) symptoms and stressors that impact on movement disorders; (iv) deeper insight into the possible underlying neurobiological mechanisms; (v) how the QoL of patients with RTT, especially the psychosocial aspect, can be affected by movement disorders; and (vi) treatment implications for managing movement disorders in patients with RTT.
The merit of each of these themes and the associated clinical considerations are described in the next section.

Clinical features of abnormal movement behaviors
Movement disorders in RTT are wide-ranging. The impairments in gross and fine motor functioning also overlap with other organ systems, making their treatment challenging. The clinical features that emerged in theme 1 can supplement our understanding of movement disorders in RTT across the broader clinical ecosystem. Our review suggests that the developmental trajectory of movement disorders as the disorder advances is variable and not well defined ( Table 2). This is important because in RTT, other symptoms such as cardiorespiratory fatigue and EBAD are likely to lead to worsening motor function and underscores the need for ongoing clinical surveillance of movement disorders in this patient group.

Dystonia
While we have previously indicated that proactive management of dystonia would be useful for managing autonomic dysregulation and hence EBAD in patients with RTT, 64 dystonia is relatively understudied in RTT. Dystonia is often misdiagnosed in other neurodevelopmental disorders, 65,66 and an inaccurate diagnosis of spasticity or dystonia might lead to delays in optimizing surgical strategies. In patients with RTT, scoliosis caused by truncal dystonia can increase with age 34 and can exacerbate the gross motor decline if it goes unmanaged. Further data from the RTT Natural History Study show that patients who were more likely to develop scoliosis were less able to have functional use of their hands. 67 The diagnosis of dystonia in RTT can be difficult as it is often associated with other movement disorders. The nonmotor components of primary dystonia, such as sensory and neuropsychiatric abnormalities, prevalent in RTT, also need to be carefully considered 68 alongside Sandifer syndrome, which is commonly mistaken for dystonia. Given the gastrointestinal dysfunction in RTT, close attention regarding Sandifer syndrome is also needed to avoid misdiagnosis and wrong treatment. Clinical signs and symptoms of paroxysmal autonomic instability with dystonia (PAID) syndrome should also be considered. 69 PAID may present in cases where there is already an underlying autonomic dysregulation. Moving forward, the identification of clinical features and severity of dystonia in RTT could be improved by adopting the recommendations of dystonia rating scales. 70

Parkinsonian-like features/tremors and ataxia
Freezing and ataxic gait in patients could reveal distinct neurological mechanisms. In patients with RTT, Parkinsonian-like rigidity is frequent and usually appears early on in development and tends to increase with age. 15 This may further contribute to the gross motor decline.

Stereotypic movements
Specific movement behaviors such as hand stereotypies can be adversely affected by bradykinesia and hypertonia. 10 Bruxism is a common stereotypic behavior in patients with RTT and also appears to be the most common oral issue. 71 Other evidence has shown a relationship between bruxism and gastroesophageal reflux disease 72 and anxiety. 73 Because gastrointestinal issues and anxiety can worsen EBAD in patients, it would be prudent to monitor EBAD in RTT to facilitate the management of bruxism.
Nighttime bruxism is less frequent in patients 74 and, in general, the stereotypies associated with hand functioning and bruxism in patients with RTT seem to disappear at night. However, the mechanism behind this is unknown but is likely to implicate independent neural mechanisms. In RTT, autonomic dysregulation is present during wakefulness and also at night, 75,76 suggesting that the dampening down of fine motor deficits (hand stereotypies and bruxism) during sleep involves neural substrates not influenced by autonomic dysregulation, or could be more resilient to them. Further investigation would be needed to test this hypothesis.

Mutational profile and its impact on movement disorders
The second most frequently occurring theme was associated with how the mutational landscape in RTT impacts the clinical severity of movement impairments. This showed that: (i) the phenotype of movement disorders is different between typical and atypical RTT; (ii) overall, those patients with truncating mutations or large deletions had worse clinical symptoms and outcomes; and (iii) C-terminal deletions were less severe in these patients. As many specialist clinicians are involved in the care of patients with RTT, they might not be aware of the diversity of relationships between different movement disorders and the mutational profile in RTT. This could lead to delays in diagnosis or adoption of a 'wait-and-see approach' as suggested by others. 35 Our review has provided a valuable resource of information regarding mutation and clinical impact by adding to the evidence base. Despite not following a predicted clinical trajectory, when it comes to the management of movement disorders in RTT, this information would be useful for assisting clinicians in better counseling families.

Dystonia
Dystonia appears to be common in individuals with truncating mutations; however, as dystonia may show different distribution it can emerge sporadically during the lifespan. Age-related changes of dystonia in RTT are difficult to predict even if a clear mutational profile has been established.

Parkinsonian-like features/tremors and ataxia
The frequency of Parkinsonian-like rigidity does not differ between individuals with truncating and missense mutations. 15 Some other patterns also emerge, such as R294X mutations presenting with Table 2. Developmental trajectory of movement disorders in patients with RTT Stage I (period of developmental stagnation) between 6 to 18 months of age. During the initial period there is hypotonia and this lessens after rigidity becomes established. Stage II (developmental regression phase) between 12 to 48 months. Parkinsonian-like rigidity is common. It can appear early (3 years of age) and frequency increases with age. Nonambulatory patients are more severely affected. In the later parts of stage I and beginning of stage II, patients can present with bruxism. When adjusting for mutation type, bruxism declines with age. For the majority of patients, hand stereotypies appear during this stage, but hand stereotypes may also appear before regression in some patients. Hand function declines over time; however, the frequency of hand stereotypies remain stable and high across the lifespan. Stage III (pseudostationary stage) around ages 2 to 10 years. Dystonia and Parkinsonian-like rigidity persists across the lifespan of the disorder and is usually stable. The severity and trajectory of dystonia, spasticity, and Parkinsonian-like movement disorders are probably dependent on the genotype, especially in patients with truncating mutations or large deletions who have worse clinical symptoms and outcomes. Stage IV (motor regression) about 10 years of age. Patients enter into motor regression, which is primarily associated with a decline in gross motor function. RTT, Rett syndrome.

Clinical Neurosciences
Movement disorders in Rett syndrome hyperactive behavior and those with the T158M mutation having an ataxic rigid phenotype. 22

Stereotypic movements
Individuals with severe genotypes have worse oral health-related outcomes and by the age of 3 years >50% will have bruxism; however, when adjusting for mutation, the predictive risk of bruxism decreases with age. 44 The genotype-phenotype relationship of hand stereotypies is less clear. 10,12,27,77 Fine motor disturbances such as those caused by hand stereotypies can also appear in patients who are mutationnegative for methyl-CpG binding protein 2 (MECP2), suggesting that mutations in MECP2 are not entirely responsible for driving the neurobiological impairment seen in hand stereotypies. 10 It would be helpful to develop objective measures to identify patterns of stereotypies among different mutations. Machine learning could be a useful foil alongside wearable sensors to track movements in patients with RTT and help in the classification of repetitive movement patterns. In the context of this evidence synthesis, the information regarding the mutational profile and movement disorders is useful because it enriches the preexisting evidence base.

Symptoms and stressors that impact movement disorders Sensory factors and stressors influence movement disorders in RTT.
In some instances, sensory influences might help to reduce the incidence of secondary disabilities 33 ; however, in RTT, the symptoms of EBAD can worsen because of acoustic sensitivity, pain, and other coexisting symptoms and stressors. 2 The input of the sensory pathway could be altered in patients with RTT and this could make individuals more vulnerable to altered somatosensory processing especially in patients who are most at risk, i.e. less ambulatory or confined to a wheelchair. A decrease in pain sensitivity forms part of the supportive diagnostic criteria in patients 1 ; however, a decreased pain sensitivity could make this patient group more susceptible to chronic pain. This is supported by the observation that although patients with RTT do experience chronic pain, their pain expression remains unaltered, 78 implying that the pain response could be clinically reflected by a worsening of EBAD symptoms and movement behaviors. Visual cues of EBAD symptoms such as pain may also be blunted by impaired movements 79 and because previous reports have suggested that parents were unsure of whether their child had experienced pain in the last month, 80 it reinforces the notion that proactive monitoring of EBAD symptoms to reduce sensory stressors are critical in RTT. Baseline heart rate variability 81 and using electrodermal activity to monitor the impact of stressors on chronic illness in patients with RTT 82 could be options to assist clinicians in the management of EBAD and its impact on movement disorders in this patient group.

Possible underlying neurobiological mechanisms
As far as we are aware, our evidence synthesis, combined with thematic analysis, is the first to reveal the underlying neurobiological mechanisms among the movement disorder spectrum in patients with RTT. In RTT, a different neural mechanism can operate with regards to movement disorders, as was noted between freezing of gait and ataxic gait, 16 and these neural systems are probably independent of the pathways associated with autonomic dysregulation. The evidence revealed that abnormalities of the basal ganglia and dysregulation of the CBGTC loop are likely to be involved. Some of the age-related changes of movement disorders was suggested to be caused by changes in neurotransmitter density in the basal ganglion. 34 There has also been some indication of nigrostriatal pathway involvement. 25,32,83,84 Animal models further demonstrate nigrostriatal deficits. 85,86 Despite these findings, the relative contribution of the loss of nigrostriatal circuits and associated movement disorders in RTT is still unknown. Previous imaging studies in patients suggest the impact of nigrostriatal pathway activity to be mild. 87 Homovanillic acid (HVA) level is also negatively correlated with Parkinsonian-like rigidity in patients with RTT, 15 demonstrating a more specified dopaminergic involvement. Although lower levels of HVA could be associated with rigidity in patients with RTT, an extensive study of HVA levels in 1388 children with neurological disorders suggests that abnormalities in HVA levels is a frequent finding in different neurological conditions and, in some cases, probably not disease specific. 88 Nonetheless, measurement of HVA levels in RTT in a larger sample population and how these levels correlate with white matter changes 88 would be useful in furthering our understanding of the role of dopamine metabolism and movement disorders in RTT.
Studies in mice models of RTT have also suggested that motor dysfunction arises from cerebellar dysfunction 89 and repetitive movement behaviors caused by GABAergic dysfunction 90 ; however, how this relates to different patient populations and the underpinning clinical severity remains to be established. While an imaging study in females with RTT emphasizes reductions in the volume of parietal gray matter and anterior frontal lobe 91 as the most prominent anatomical abnormalities, whether these anatomical changes cause different movement impairments in patients with RTT is unknown. Recent sensitive neuroimaging investigations in patients with RTT using a susceptibility weighted imaging approach have indicated increased iron deposition in dopaminergic, and gray matter networks of the basal ganglia are associated with the severity of dystonia. 40 When viewed together, a better understanding of the neurological underpinning of movement disorders in patients with RTT has provided an enriched picture of the different neural circuits involved. This may help to assist future studies of brain imaging in RTT.

QoL and movement disorders
Understanding how abnormal movement behaviors affect QoL in RTT is not straightforward. When QoL is evaluated using the Child Health Questionnaire 50 (CHQ-PF50), patients with RTT with more severe motor impairments had better psychosocial functioning because they were less likely to engage in maladaptive behaviors such as aggression and self-injury. 58 In a recent observational study that assessed the QoL more broadly among patient with intellectual disability, including those with RTT, cerebral palsy, and Down syndrome, multivariate analysis showed that mobility was less influential on QoL. 92 This study used the validated Quality of Life Inventory-Disability (QI-Disability) tool to assess QoL. Using different tools to assess QoL in patients with complex neurodisability makes it challenging to infer clinical patterns among studies when examining how mobility and other aspects of movement disorders in RTT affect QoL.
Even when different instruments are used to assess QoL, in RTT, the genotype-phenotype relationship can also confound interpretation. In a study of 210 patients with RTT, even though the p. Arg294 mutation was the clinically milder phenotype, patients with the p.Arg294 mutation had the poorest QoL scores overall 3 and partly supports the previous observation regarding psychosocial summary. 58 This finding is important because it could be possible that in those patients with the milder clinical form, the behavioral and emotional components of EBAD could exacerbate the clinical symptoms of movement problems in RTT, which could negatively affect QoL but might not be necessarily captured using existing QoL tools. Reporting of QoL domains may also differ, 92 as do the discrimination between health-related QoL and QoL overall. While we cannot speculate on the relationship between the clinical symptoms of EBAD and genotype on QoL, we know that the QoL of patients with RTT worsens after 12 years of age. 3 Therefore, it would be important when planning future treatment and rehabilitation programs to longitudinally track the effect of EBAD on movement disorders using an appropriate QoL measure as the disorder progresses.

Treatment of movement disorders
Information regarding the treatment of movement disorders is limited in patients with RTT. From the 43 manuscripts evaluated, only one presented a small case series of two patients using trihexyphenidyl to assist in the management of life-threatening events arising from dystonic episodes. 59 By completing our systematic review followed by a thematic analysis, we can use this information to further understand the treatment of movement disorders from a wider perspective. When considering treatment of movement disorders in patients with RTT, reduction in movement is not the primary goal. Rather, the goal of treatment should be placed on reducing impairment and achieving functional movement.
As we have previously mentioned, 64 information can be extrapolated from other observations in the non-RTT population to inform our current understanding. We can now apply this knowledge to movement disorders in RTT. There is little evidence to support the use of different treatments in movement disorders. Notwithstanding this limitation, different treatment options and how they can be extrapolated to the RTT population to inform current guidelines are discussed.
As there is little research on abnormal movement treatment in RTT, we extended our search to include management of abnormal movements in other conditions in children and young people. Baclofen, gabapentin, trihexyphenidyl, levodopa, diazepam, and clonidine are common medications for managing impaired muscle tone. 98,99 Anticholinergics such as trihexyphenidyl was shown to be useful to assist in the management of dystonia in a case study of two patients with RTT. 59 There is little to no evidence on the effectiveness of trihexyphenidyl for treating dystonia in children with cerebral palsy. 100 Baclofen should be used with caution as there are no evidencebased studies on the effectiveness of using baclofen to manage dystonia in patients with RTT. The use of intrathecal baclofen in dystonia and spasticity is associated with complications that might also require further surgery. 101 Close attention should be placed on the recognition of hypoventilation and worsening of extrapyramidal symptoms in RTT.
Benzodiazepines such as clonazepam have also been used for the management of movement disorders such as dystonia despite their efficacy not being proven. However, as we have previously indicated, 14 the use of benzodiazepines should only be used in RTT when deemed strictly necessary. If benzodiazepines are prescribed, their use must be monitored carefully because of the high potential of worsening of autonomic function in this patient group. α 2 -Agonists such as clonidine have been shown to be effective in managing secondary dystonia in a cohort of patients with cerebral palsy. 102 However, a recent meta-analysis indicated that the level of evidence for clonidine treatment in improving dystonia in cerebral palsy was low to very low. 100 Some data in a small number of cases suggest that clonidine might be useful for the treatment of acute akathisia. [103][104][105] Clonidine must be used with caution given its multisystemic side effect profile. 64 The known side effects of clonidine such as sedation, hypotension, and mood changes are especially relevant in RTT. There are no published studies of clonidine for managing movement disorders in RTT.
Gabapentin has been shown to reduce the severity of dystonia and improve the QoL of children in a retrospective observational study involving 69 children with refractory dystonia. 106 Gabapentin might also be useful for treating some cases of essential tremor. [107][108][109] Again, the effectiveness of gabapentin for the management of dystonia in patients with cerebral palsy has not been established. 100,110,111 Gabapentin must be used with caution given its side effect profile, 64 and it can also induce movement disorder onset. 112,113 There is no published evidence on the efficacy of gabapentin for treating abnormal tone in RTT.
Dopaminergics have also been used to manage dystonia; very lowdose levodopa therapy is another option to treat the symptoms of motor impairments in childhood neurological disorders. 114 However, studies in RTT are needed to explore its efficacy and side effect profile, as their use can cause nausea and constipation. Gastrointestinal abnormalities such as constipation are prevalent in patients with RTT, and, therefore, dopaminergics could potentially worsen constipation in this patient group.
Botulinum toxin might be useful in treating focal dystonia. 95 Fig. 2 Schematic of brain regions implicated in movement disorders and associated neurochemical pathways. Panel A: Cross-section of the human brain showing different regions likely to be involved in movement disorders. Panel B: The array of neurochemical pathways associated with movement are complex 93 with multiple interconnecting circuits that could be dysregulated in Rett syndrome (RTT). 96 In our simplified adaptation of network schematics, 93,96 outputs from striatal medium spiny neurons project into the substantia nigra pars reticulata (SNpr) via the striatonigral pathway or into the globus pallidus (GP) via the striatopallidal pathway. These projections modulate thalamic activity and trigger or diminish motor activity through dopaminergic or GABAergic control. Impairments in dopaminergic or GABAergic circuits may give rise to the Parkinsonian-like movement features or repetitive behaviors in RTT. 15,90 Striatal integrity may also be dependent on functional methyl-CpG binding protein 2. 96 More recent evidence in mice models further suggests that motor impairments in RTT could arise from altered cerebellar architecture. 89 Evidence also indicates glutamate receptor dysregulation in the motor cortex of postmortem brain tissue in patients with RTT 97   that botulinum toxin was helpful in treating hypersalivation and may also improve other oral functions. 115 As the minimum dosing interval is about 12 weeks, long periods between injections might also be required on an individual patient basis. In summary, the lack of empirical evidence, inadequate evidence of efficacy, and the potential for multisystem side effects of pharmacological agents being used currently necessitates the need for close monitoring in patients with RTT. These treatments and the implications for using them in patients with RTT are summarized in Table 3.

Surgical and minimally invasive based options
There is some evidence that intrathecal baclofen and surgical treatments such as deep brain stimulation (DBS) of the globus pallidus pars interna are helpful in treating acquired or refractory dystonia in patients with cerebral palsy. 116,117 Another study has shown that pallidal deep brain stimulation for dystonia should be considered early on in childhood as the response becomes less effective as the dystonia becomes more established. 118 There are other aspects of DBS surgery for the treatment of dystonia that need to be considered. Psychological support preintervention and postintervention should be explored to assess the impact of DBS on social functioning and QoL in the longer term. 119 It is unclear whether such interventions would be feasible for patients with RTT. There is little information on the surgical management of dystonia in patients with RTT.

Exploratory strategies and neuroprotection
Exploratory strategies might help to reduce the progression of movement disorders in RTT. Patients with RTT with severe dystonia have age-related increased iron deposits within regions of the basal ganglia. 40 We have previously surmised that autonomic dysregulation may influence the inflammatory state in RTT, 64 and, coupled with the altered subinflammatory profile underlying Rett pathology, 120 these factors could affect redox-related physiology in patients. Together with others, 40 it would therefore be reasonable to presume that the hypoxic conditions caused by the underlying autonomic dysregulation in RTT might cause a redox imbalance that leads to increased iron mineralization. Although autonomic dysregulation and movement disorders are likely to involve different neural pathways, we suggest that buspirone could be one option to consider for the management of severe dystonia in RTT. It would be interesting to see whether reducing breathing dysregulation in RTT using buspirone could indirectly lower redox imbalances and iron accumulation and whether this reduces the clinical severity of dystonia, especially in individuals with truncating mutations who are reported to have a higher frequency of dystonia and rigidity. 22 Buspirone could also be used alongside free radical scavengers such as vitamin E as there is evidence in animal models of RTT having improved hypoxia tolerance with a vitamin E derivative. 121 Neuroprotective agents could be other options to consider. Several candidates such as levetiracetam, melatonin, memantine, omega-3, topiramate, and vitamin E have been examined for their neuroprotective potential. 122,123 Some of these have been shown to provide neuroprotection in different neurological disorders. 122 These candidate molecules could delay the progression of movement disorders in rare neurodevelopmental syndromes. Although some of these agents, such as memantine, might offer some promise in RTT, 124 there is at present no robust empirical evidence to suggest whether these molecules would offer any neuroprotective benefit for patients with RTT. Clinical trials are needed to assess the ability of these molecules to delay motor disease progression in this patient population.

Other strategies
Environmental enrichment offers an attractive nonpharmacological strategy to assist in the management of some aspects of movement disorders in RTT. The evidence synthesis showed that even minor improvements in gross and fine motor function would benefit patients. Moreover, environmental stimulation and its positive impact on the individual can also reduce the caregiver burden, as suggested by others. 37 Environmental and sensory enrichment likely leads to neuroplastic changes caused by upregulation of brainderived neurotrophic factor (BDNF) in patients with RTT. This premise is supported by a recent randomized stepped wedge trial in 12 females with RTT who showed improvements in gross motor skills and BDNF levels after 6 months of enriched environmental treatment. 125 The merits of environmental enrichment should be viewed from the perspective that access to resources among geographical regions will be varied. A Danish study explored the facilitators and obstacles of low-intensity activities in females with RTT. It highlighted themes that could facilitate health-promoting interventions in patients with RTT and ways to reduce barriers. 126 While some evidence suggests that a holistic approach to physical therapy may improve the QoL of individuals with RTT, 127 the lack of longitudinal follow-up, methodological weaknesses, and low statistical power 128,129 necessitates the need to improve the impact of nonmedical intervention among the wider RTT population. Nevertheless, environmental enrichment would be a useful adjunct alongside other interventions for managing the impact of movement abnormalities in patients with RTT.

Conclusion
Our review emphasizes six key themes. First, it provides a synthesis of clinical features among movement disorders in patients with RTT. Even though some of the clinical features identified are broad and lie across the movement disorder spectrum, under the rubric of clinical care, our information would help clinicians to better understand patterns and trends that emerge. Second, delineating genotypephenotype relationships can allow for more robust genetic counseling for families and also assist healthcare professionals in monitoring disease progression among different mutations and the potential trajectory of movement disorders. Third, environmental enrichment would be beneficial for improving some aspects of movement disorders. Fourth, the cerebellum and basal ganglia and the dysregulation of the CBGTC loop are likely anatomical targets for movement disorder problems in RTT. Some emerging and adjunct evidence also suggests that increased mineralization of iron within the basal ganglia, particularly within the substantia nigra, putamen, and globus pallidus, correlates with the severity of dystonia in patients with RTT. Fifth, movement disorders can have a variable and sometimes unexpected impact on the QoL of patients with RTT. Finally, the evidence from the systematic review and emerging from the thematic analysis allowed us to describe the developmental trajectory (Table 2) and discuss treatment implications for the management of movement disorders in patients with RTT ( Table 3).
Treatment of movement disorders in RTT would require a combined pharmacological and biopsychosocial approach. Recent evidence has demonstrated that presymptomatic training in a mouse model of RTT can help to manage symptoms and delay the onset of the functional impairments. 130 The translational impact of this work needs to be demonstrated in human studies; however, it strongly suggests that early behavioral training in an individual diagnosed with RTT could delay the onset of motor symptoms. This important finding supports the rationale for routine genetic testing of RTT in newborns. 130 Building on this work, a proactive approach of environmental enrichment together with pharmacological treatment during the very early stages of the disorder could help in the management of movement disorders, especially in high-risk individuals where the progression could be quicker. This combined strategy could help patients retain specific aspects of movement and reduce overall disease burden.

Limitations
While this evidence synthesis is valid, the clinical inferences and their applications to the broader RTT community should be treated with caution because of disparities within the subject area. Different types of movement impairments persist in RTT, and the findings should not be considered universal across the broad spectrum of movement disorders. The severity of illness, co-occurring comorbidities, and methods used to quantify movement disorders among study groups from different regions limits the generalizability of the findings. The different sample sizes of studies may also influence the findings of genotype-phenotype relationships, especially when examining patient populations with the same mutational profile. Changing definitions 131 of dystonia/spasticity over time could also be a minor confounder and this could affect diagnostic assessments and the usefulness of existing clinical rating scales. A personalized approach for the treatment and management of movement disorders in RTT is therefore warranted.
A primary search strategy was used that balanced sensitivity and specificity to answer the research questions of the systematic review. We felt that having a broader search term would have resulted in many vague search results and therefore used a more focused primary strategy. Our search strategy, while focused, could have inadvertently omitted some relevant material from the literature. However, by using a secondary search strategy, we were able to extend the reach of our systematic review.