Rett syndrome (RTT) is a profoundly disabling neurodevelopmental disorder predominantly affecting females (1–3) and is primarily caused by mutations in the gene encoding X-linked methyl-CpG-binding protein 2 (MeCP2) (4).

Individuals with RTT undergo a period of apparently normal development during the first 6 months of life followed by developmental regression between 12 and 30 months of age characterized by partial or complete loss of spoken language and hand function skills, impaired or absent gait, and the development of repetitive hand stereotypies, which all represent the fundamental diagnostic criteria for RTT (5). Other common features include epilepsy, breathing disruptions while awake, gastrointestinal difficulties, autonomic abnormalities, scoliosis, and limited nonverbal communication skills (e.g., intense eye communication) (5–8). Behavior disturbances occur in nearly all individuals with RTT across their lifespan, with internalizing behaviors such as anxiety or mood swings being much more common overall than externalizing behaviors such as aggression, hyperactivity or self-injury, with anxiety-like behavior considered a significant parental concern (9, 10). Recommendations for regular annual assessments in the primary care setting include those related to behavior such as anxiety and depression (11).

A number of caregiver- and clinician-assessed measures have been developed to assess disease severity, functional ability, and primarily neurological and sociobehavioral abnormalities in clinical studies. The Rett Syndrome Behaviour Questionnaire (RSBQ) assesses the severity of neurobehavioral problems from the perspective of the caregiver and is one of the most widely used measures due to the specificity of its psychometric profile to the core features of RTT and its acceptance by the United States Food and Drug Administration (FDA) for use in RTT studies.

In this review, we will provide a summary of the role for caregivers and healthcare providers in the management of RTT as well as describe the characteristics of the RSBQ and the association between its domains and clinical features of RTT, its reliability (sensitivity and specificity), and performance in a phase 2 study (12) and the recent phase 3, placebo-controlled LAVENDER study (Clinicaltrials.gov: NCT04181723) (13, 14) in girls and women with RTT aged 5–20 years.

Primary care providers and other healthcare professionals caring for individuals with RTT are required to manage the evolving medical comorbidities of RTT effectively throughout an individual's lifespan, which can last beyond 50 years of age (15–17); however, many have limited first-hand experience in managing the disorder due to its rare occurrence (11).

Recent consensus guidelines for primary care providers recommend regular medical assessments (follow-up visits at least every 6 months) to screen for issues that can appear quickly, progress rapidly, and require intervention (11). In terms of behavioral assessments, it is recommended that primary care providers regularly screen for symptoms of anxiety and depression, such as withdrawal, screaming, and irritability. Respiratory assessments include screening for awake-disordered breathing (hyperventilating, breath holding, color change) and air swallowing (Table 1). Neurological assessments should include screening for abnormal movements (stereotypies and dystonia) and their level of impact on daily activities, fine motor skills (hand use), and gross motor skills (sitting, standing, and walking) (Table 1).

Caregivers are tasked with managing daily activities and communicating with the individual with RTT on a daily basis. Together with neurologists and pediatric neurologists, caregivers are primarily responsible for relaying feedback on clinical assessments in trials and determining improvement in the individual with RTT; this emphasizes the need to address the top concerns of caregivers, who often cite communication as being one of the most important (18–20).

Understanding the natural history of these core symptoms in RTT is critical to directing diagnosis, guiding prognosis, and testing the efficacy of new interventions in clinical trials. As an established efficacy measure in RTT clinical trials, it is important that healthcare providers and caregivers are able to interpret the psychometrics of the RSBQ.

The RSBQ is a validated assessment tool that has been accepted by the FDA as a primary outcome measure for clinical studies in RTT and has been used on numerous occasions to assess symptoms in clinical trials in RTT (12, 13, 21, 22). The RSBQ is a caregiver-completed scale assessing a wide range of neurological and behavioral symptoms in RTT that has been used across a range of ages (2–47 years) in clinical studies (23–25). The RSBQ consists of 45 items of which 38 items are grouped into 8 domains/subscales that reflect the core features of RTT: General Mood; Breathing Problems; Hand Behaviors; Repetitive Face Movements; Body Rocking and Expressionless Face; Nighttime Behaviors; Fear/Anxiety; and Walking/Standing) (Table 2). When the RSBQ was first developed, each item was grouped into the appropriate subscale based on a factor analysis (26); the 7 items that did not belong under any of the subscales were classed as “uncategorized” but contribute to the overall total score. Each item is rated on a Likert scale as 0 (behavior “not true”), 1 (behavior “somewhat or sometimes true”) or 2 (behavior “often true”), with the total score ranging from 0 to 90 (higher scores indicate increased severity) (26).

The percentage occurrence of potentially characteristic behaviors in RTT include hand stereotypies, which are almost universal (99%), teeth grinding (58%), sleeping difficulties and nighttime laughing (64%), anxiety or inappropriate fear (73%), low mood/changeable mood (77%), breath holding (63%), and hyperventilation (77%) (25), which reflect many of the caregiver concerns and are all captured in the domains of the RSBQ.

The heterogeneity of the RTT phenotype presents a challenge both in clinical trial assessments and routine clinical practice as often only a limited set of difficulties may predominate at any given time; however, the total scale scores for the RSBQ are influenced by multiple items, some of which are neurologic and others behavioral in nature, and it is this broad scope that allows changes in an individual feature to be captured whether it be related to physical functioning or behavior.

The RSBQ was initially developed as a diagnostic tool to differentiate females with RTT from those with severe intellectual disability after it was shown to be highly discriminatory between individuals with RTT (n = 143) and intellectual disability (n = 85) (26). The RSBQ transitioned to a clinical trial outcome assessment after it was shown to adequately describe behavioral characteristics in RTT in a UK and Australian cohort (23). In a subsequent study in 74 girls with RTT, the Fear/Anxiety subscale was deemed reliable and valid for use in clinical and research settings (24). The RSBQ was first used in a placebo-controlled clinical trial to investigate the efficacy of mecasermin (recombinant human insulin-like growth factor 1) as a treatment for RTT though, together with most of the study outcomes, failed to show a statistically significant difference between drug and placebo (22).

The RSBQ adds clinical context by indicating severity and frequency of symptoms, is relatively simple to complete, and is correlated with functioning (23–25). More research is needed to improve understanding of the correlation of RSBQ domains with the overall score, the floor or ceiling effects due to the limited range between the absence and presence of a characteristic, and a minimal clinically important difference. While the RSBQ addresses most of the features of RTT, there are important clinical domains that are common in RTT including gastrointestinal and nutritional problems (27, 28), and scoliosis (29), that are not covered by the RSBQ.

The number of items and range of scores in a subscale is important in terms of the validity and reliability of a psychometric profile. The domains/subscales of the RSBQ and their clinical relevance in the diagnosis and assessment of disease severity are summarized below.

Studies have investigated the reliability of the RSBQ (test–retest reliability, intra-rater reliability), validated its discriminatory diagnostic potential in a RTT population versus a population with intellectual disability, and compared behavior profiles in UK and Australian populations (23, 26). Moderate to high internal consistency was reported for the total score and the 8 subscales, with good inter-rater and test–retest reliability scores, and significantly higher scores in a RTT population versus those with intellectual disability, thus validating its use as a diagnostic tool (26). The Cohen's d effect sizes were all large (40), the smallest being 0.81 for the Walking/Standing domain and the largest for Hand Behaviors (2.24) suggesting that Hand Behaviors is the domain with the greatest specificity to RTT. Hand behaviors and breathing problems were virtually only present in individuals with RTT, while mood fluctuations, fear/anxiety behaviors, inconsolable crying and screaming at night, and repetitive mouth and tongue movements and grimacing were more frequent or severe in RTT compared to those with intellectual disability (26). Physical ability was significantly associated with the domains for Hand Behaviors and Walking/Standing (26); thus, these domains provide an additional assessment of physical functioning. Variability in each domain ranged from 4.6% for Walking/Standing and up to 11.9% for General Mood (26), which may be affected by the number of items in each domain. There was also good correlation in terms of the behavior profile identified using the RSBQ scores between Australian and UK populations with RTT (23). In addition, a recent examination of meaningful changes in RSBQ symptoms offered insights into the essential domains of life experiences of children with Rett syndrome and their caregivers (41).

Cianfaglione et al. (2015) analyzed potential differences in the RSBQ based on mutation and clinical severity in a UK national sample of 91 girls and women with RTT and showed that the Hand Behaviors domain contributed the most to the total score (25). Means for each domain were generally near half of the maximum scores possible and ranges were broad, suggesting an absence of scoring to the floor/ceiling and variability in the manifestation of the behavioral phenotype across individuals. There was a high degree of positive intercorrelation between all RSBQ domains other than the Walking/Standing domain. Barnes et al. investigated the reliability of the Fear/Anxiety domain in 74 girls with RTT and showed that internal consistency was comparable with that reported by Mount et al. (26), that age did not affect scores, and that severity of general anxiety was inversely correlated with clinical severity (RTT-Clinical Severity Scale) (24).

Despite evidence of a positive discriminatory benefit with the RSBQ in RTT, in a recent assessment of the RSBQ, it was reported that half of the items tested exhibited floor or ceiling effects (42). However, this study was itself criticized for inadequate representation of the population, the use of a clinically heterogeneous sample, and missing psychometric evaluations (43). In their rebuttal, the authors recommended that efforts should be directed to improve metrics for specific functions (walking, hand use, and nonverbal communication) and disabilities (seizures, sleep dysfunction, and gastrointestinal comorbidities) that caregivers routinely highlight as primary concerns in order to improve caregivers' interpretation of clinically meaningful improvements in RTT (44).

The efficacy and safety of trofinetide, a synthetic analog of a tripeptide (glycine-proline-glutamate) derived from the N-terminus of insulin-like growth factor 1 (45) and the first drug to be approved for the treatment of RTT, was investigated in a placebo-controlled, phase 2 study (12), and the phase 3, placebo-controlled, LAVENDER study (13, 14).

In the phase 2 study, treatment with trofinetide at 200 mg/kg twice daily for 6 weeks showed statistically significant improvements relative to placebo on the RSBQ (Figure 1A) and was supported by improvements on overall functioning as measured by the CGI-I (Figure 1B). All of the domains except Walking/Standing were directionally in favor of the 200 mg/kg treatment group, with notable improvement in mood dysfunction and disruptive behavior (General Mood domain, p = 0.007), breathing problems (Breathing Problems domain, p = 0.095), and repetitive movements (Repetitive Face Movement domain, p = 0.047) (Figure 1C). The RSBQ total score had medium Cohen's d effect sizes (−0.487) that were similar to the CGI-I (−0.645) (12).

In the LAVENDER study, which included females with RTT aged 5–20 years, after 12 weeks' treatment, twice-daily trofinetide demonstrated statistically significant differences from placebo for both coprimary endpoints (RSBQ and CGI-I) (14). Mean (standard error, SE) change from baseline to week 12 in the RSBQ total score was −5.1 (0.99) and −1.7 (0.98) in the trofinetide and placebo group, respectively. The least squares (LS) mean treatment difference was −3.1 (95% CI, −5.7 to −0.6; p = 0.0175, Cohen's d effect size = 0.37; Figure 2A). The change in RSBQ was aligned with the CGI-I using RTT-specific anchors (38), suggesting that improvement in behavioral components is predictive of overall clinical status (Figure 2B). Cohen's d effect sizes for the coprimary endpoints fell in the 0.4–0.5 medium effect size range (0.37 for the RSBQ, 0.47 for the CGI-I) (40). The changes from baseline for all RSBQ domains were all directionally in favor of trofinetide (Figure 2C), with notable improvements in the domains for Body Rocking and Expressionless Face (nominal p = 0.0132; Cohen's d effect size = 0.39) and Fear/Anxiety (nominal p = 0.0003; Cohen's d effect size = 0.58). The Cohen's d effect sizes reported in the phase 2 study and in LAVENDER were of medium size and comparable to those reported for FDA-approved treatments and suggestive of clinically meaningful changes.

The RSBQ is a valuable tool for the evaluation of RTT and, as an efficacy outcome measure in clinical trials, is easy to understand and specific to the core features of RTT. The heterogeneity of the domains that constitute the RSBQ total score complements the heterogeneity of findings in RTT and allows the detection of a drug effect whether improvements are observed in one or all of the 8 domains. Given that caregivers are required to complete the assessments in the RSBQ in clinical trials in RTT, and that primary care providers are increasingly expected to engage in routine healthcare assessments for individuals with RTT, it is important that both groups are familiar with the RSBQ and its metrics. Future trials should consider additional assessment tools to augment the domains of the RSBQ such as scales measuring verbal and nonverbal communication and the recently validated Gastrointestinal Health Questionnaire for Rett Syndrome (46).