Introduction

Rett syndrome (RTT), although considered rare, is one of the most common causes of intellectual disability in females with an incidence of 1 in 8,500 by the age of 15 years [1]. Most individuals with RTT express a mutation in the MECP2 gene, which either activates or represses neural transcription when it binds to methylated cytosines in DNA [2,3]. However the severity of the disorder varies depending on the type of MECP2 mutation [4–6] and the pattern of X-chromosome inactivation [7]. Clinical outcomes for this syndrome are complex, with varying degrees of autonomic dysfunction [8], motor impairments influencing mobility and oromotor control [5,9–11], epilepsy [12,13] and poor growth [14,15]. A high proportion also develop skeletal abnormalities such as scoliosis [16,17], low bone density and mass and a high frequency of fractures [18]. As almost 60% of those with RTT will live beyond 37 years, physicians must begin to address these chronic health issues early in life [19].

Studies assessing bone mineral content (BMC, g) and areal bone mineral density (aBMD, g/cm²) in RTT have shown that these parameters were lower compared to the gender matched control groups [18,20–22]. Bone mineral content and aBMD z-scores for age and height tended to become more negative with age [18,21–23], however individuals as young as three and four years of age had low BMC and aBMD values in the lumbar spine [18,23], total body and femoral neck [18].

Fractures are also a common occurrence in RTT. The fracture incidence in an Australian RTT population was shown to be 43.3 per 1000 person years, which equates to a rate nearly four times that of the general population [24] within whom fractures are often associated with sport and playing activities [25]. Fractures can occur spontaneously, with trivial trauma or a fall, and occur predominantly in the long bones of the upper and lower limbs [24,26]. Fractures have also been closely linked with mobility levels and ability to bear weight, and in a Danish study there were significantly more non-ambulant patients who had fractured compared to the ambulant non-fractured healthy control group [26]. To date the prevalence and incidence of vertebral fractures in RTT remains unknown. As vertebral fractures, often considered a manifestation of osteoporosis, may be asymptomatic, they may escape medical attention [27].

Our Australian study, which assessed the BMC and aBMD in the lumbar spine, femoral neck and total body using dual energy x-ray absorptiometry, found that subjects who were unable to walk or wheelchair dependent had lower mean height calculated BMC and aBMD z-score values compared to those who were ambulant [18]. The z-scores were predicted for height and sex, as these were more appropriate to use than comparisons with normal values by age, given the generally small stature of those with RTT. In a US study (n = 49), one third of individuals who were non-ambulant had decreased bone mass in the lumbar spine [23]. Small bone size and reduced lean tissue mass have also been observed and correlated with low bone density and mass values [18,22]. Lean tissue mass has been shown to be a strong predictor of bone density in young women in the general population, particularly during adolescence [28]. Reduced lean tissue mass, often seen in those who are immobilised, decreases the physiological adaptations that occur within bone as response to muscle forces. This negatively affects bone development (modeling), leading to narrower bones with thin cortices and reduced trabecular bone mass [29,30].

Despite the high frequency of fractures and low bone mass and density identified in RTT, information about best clinical practice for the diagnosis, treatment and prevention of osteoporosis in this disorder is lacking. Due to its rarity, many clinicians may have little or no experience with RTT. This study therefore developed clinical guidelines for the management of bone health in RTT.

Methods

This project was coordinated from the Telethon Kids Institute in Western Australia (WA). Ethics approval was obtained from Curtin University in Western Australia.

Guideline development followed the format described by the National Health and Medical Research Council, Australia (1999) and consensus was achieved using a modified Delphi technique as used in previous studies [31]. Potential expert panel members were invited to participate via email which provided information about the study and described the development process. As approved by the Curtin University ethics committee, the expert panel members provided their consent to participate in the study via email, which was then stored on a password-protected server at the Telethon Kid's Institute.

Results

Initial Guidelines Draft and Redrafting Using a Modified Delphi Technique

The draft guidelines included 41 statements, 11 questions (representing all RettNet topics) and a reference list. The comments made by expert panel members were either supportive of the responses made to statements or used to construct statements for round 2. Of the 38 participants who responded to the first round, 33 (86.7%) responded to the second round which included nine new statements. Agreement was achieved for all statements in the second round, allowing these statements to form part of the final guidelines. The final document contained 39 separate statements, all of which reached consensus and are listed in Tables 1–5.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Assessment of Bone Health.

https://doi.org/10.1371/journal.pone.0146824.t001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Bone Mineral Density Assessment.

https://doi.org/10.1371/journal.pone.0146824.t002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Bone Mineral Density Assessment Technique.

https://doi.org/10.1371/journal.pone.0146824.t003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Non-Pharmacological Intervention.

https://doi.org/10.1371/journal.pone.0146824.t004

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Pharmacological Intervention.

https://doi.org/10.1371/journal.pone.0146824.t005

Discussion

Reduced bone mass and density have been identified in RTT throughout childhood and into adulthood [18,21–23,38,39]. In addition bone area is reduced [18,22] and muscle mass is low [18,31]. One impetus for developing these guidelines was initially in response to concerns expressed by parents on Rettnet regarding the increased likelihood of bone fractures in RTT, as well as their questions regarding the assessment and improvement of bone density in their child. Furthermore, research indicating low bone mass and density [18] and increased fracture risk [24] in RTT also highlighted the need for these guidelines.

This project identified relevant literature relating to RTT or developmental disabilities and aspects of bone growth during childhood, adolescence and adulthood. A series of statements and questions were developed from the literature and from the listserv Rettnet. An international expert panel and a parent reference group subsequently reviewed the statements. This comprehensive document, which aims to guide clinicians on best practices for maintaining bone health and reducing fractures in RTT, was compiled using a modified Delphi technique. The panel provided expert advice for the scope of the guidelines with limited “drop out” between the first and second round review. The option “not my area of expertise” made it possible for the clinicians to only express opinion in their respective field of experience, enabling inclusion of multiple health professions in the expert panel. Using similar methods as our previously developed guidelines for management of scoliosis [17] and assessment and management of nutrition and growth [31] in RTT, the communication via email and the online review of the guidelines occurred in a timely and efficient manner.

Limitations of the study were the lack of supporting peer-reviewed literature specific to RTT for aspects of the guidelines, particularly statements relating to intervention strategies. Twenty-two of 63 statements (which includes the different parts of some statements) were formulated based on clinical experiences of the expert panel rather than peer-reviewed literature. These statements particularly related to the types of biochemical testing used to assess bone health, technique, frequency and skeletal site targeted for bone density measurements, vitamin D measurements and use of bisphosphonates to treat low bone density. There is also limited evidence on the effects of physical activity on bone mass density in RTT. Where literature specific to RTT was absent, the guidelines not only relied on the opinion of the expert panel but literature on bone health in the general population and in those for other disabilities.

Biochemical assessments were described taking into consideration the likelihood of nutritional deficits observed in RTT [11]. Radiological investigations were also outlined with techniques and analyses tailored to accommodate the decreased stature [15,40] and movement problems [5] seen in RTT. For example, the guidelines include recommendations to reduce unnecessary movements during bone density scanning procedures as individuals with RTT often display stereotyped hand movements, dystonic sustained muscle contractions or abnormal postures [41]. Lastly both pharmacological and non-pharmacological approaches to improving bone density included focusing on realistic expectations of increasing weight bearing activities, calcium and vitamin D intake. These statements were made with the proviso that the tested efficacy of this approach on preventing fractures in at–risk children with RTT is unknown although there is evidence in cerebral palsy that could be relevant. A randomised clinical trial in 26 non-ambulant children with cerebral palsy (CP) found that increasing normal standing time by 60%, with a total time between 180–675 minutes/week over the course of one year, improved vertebral volumetric trabecular bone mineral density compared to the control group with CP whose standing time was unchanged [42]. However, no change was observed in the volumetric trabecular bone mineral density of the proximal tibia [42]. A more recent study, which analysed factors associated with fracture in 536 children with CP, found that those who did not use a standing device were at a nearly four times greater risk of fracture [43]. Panel members recommended the combination of increased time standing and vitamin D and calcium supplementation.

Poor growth, continuing into adulthood, has been observed in RTT from an early age [15,40,44]. Although first observed as a deceleration of head growth, height and weight are also reduced in RTT [44], consistent with DXA findings of narrow bones for chronological age [18,22]. The cause of this growth failure is complex and often correlated with clinical severity and genotype [45]. However functional abnormalities of the digestive tract, oromotor problems and other factors that affect feeding may also lead to nutritional deficits and poor growth [31,46]. Assessing dietary intake, in particular consumption of foods high in calcium, and measuring serum levels of bone markers such as calcium and 25-hydroxyvitamin D (25(OH)D) [47], are essential steps when investigating bone health. A recent study found that dietary protein, calcium and phosphorus intakes expressed as a proportion of Dietary Reference Intakes for age and sex, showed significant positive associations with TB aBMD z-scores in RTT [48].

Typically 25(OH)D is used to assess vitamin D status, however there is some conjecture as to the optimal level to support skeletal health [49]. The clinical practice guidelines for the evaluation, treatment and prevention of vitamin D deficiency developed by the Endocrine Society in the United States (US) and our guidelines, recommend that circulation 25(OH)D levels should reach a minimum threshold of 30ng/ml (75nmol/litre) [49]. However, the committee at the Institute of Medicine in the US, suggest a lower threshold at 20ng/ml (50nmol/litre) as they felt there was little or no benefit to bone health above this value [50]. The current position statement for Australia and New Zealand and others, conclude that further research is needed to evaluate non-skeletal benefits of 25(OH)D levels >50 nmol/L /litre (30ngmL) [51,52]. As the precise level recommended in RTT is unknown, given the altered bone in this condition, it may be beneficial to aim for a higher minimal threshold, although our panel recommended adherence to local protocols. The guidelines also include the recommended dietary reference intakes for calcium based on Australian standards [53]. These serve as a guide only and may differ from a clinician’s local protocol. We therefore suggest clinicians follow their local recommended calcium reference intakes. Where dietary intake is lacking or bone marker serum levels are low, supplementation is suggested.

Growth failure also needs to be considered when assessing bone mass and density using DXA analysis as the two-dimensional measurement is influenced by size and underestimates a person’s BMC and aBMD when subjects are small for age [18,54]. The guidelines therefore recommends calculation of height formulated z-scores and volumetric bone mineral density (bone mineral apparent density–BMAD) where possible [55]. Overall DXA assessment was agreed upon by the expert panel, quantitative ultrasound (QUS) may also be considered given that it is less influenced by small stature as it measures the propagation of mechanical vibrations through tissue, rather than a two dimensional measurement. Quantitative ultrasound not only provides information on the density of peripheral bones such as the calcaneous, phalanx, tibia and fibula but their biomechanical properties such as elasticity and compressive strength. Furthermore, unlike DXA, QUS is non-invasive due to the absence of radiation exposure. Although QUS is not as widely accepted as DXA given that only peripheral bone density assessments can be made, its clinical use is increasing due in part to its cost effectiveness, portability and ease of use [56]. In addition, QUS has been shown to predict future fractures similarly to that of DXA [57]. With increasing normative data for peripheral bones [56,58] QUS may be considered when assessing bone health in RTT.

A further risk factor for reduced bone density is poor mobility. In the Australian study of 97 females with RTT, we found that those with reduced mobility had lower bone mass and density measurements in the lumbar spine, femoral neck and total body [18]. Decreased lean tissue mass, also identified in RTT, would reduce the physiological load placed on bone, thereby impairing bone development and decreasing bone strength [18,30]. The recommendation to increase mobility, particularly through time spent walking or weight bearing, reached a high level of agreement by panel members. A daily training program in four independently mobile individuals with RTT aged between 8.5–11 years of age, found that both physical fitness and functional capabilities improved [59]. Due to the variability in physical capabilities in RTT, referral to a physiotherapist for development of a physical activity plan may be helpful. With reduced mobility comes the need for parents and carergivers to increasingly handle the children during transfers. It is therefore important to ensure that carergivers are trained in transfer methods and use of equipment and that the home and school environments are safe, so as to reduce the risk of accident and fracture.

Delayed puberty is another risk factor for low bone density that needs to be considered in RTT [33]. Oestrogen plays an important part in bone formation during puberty by stimulating endocortical bone formation whilst inhibiting periosteal apposition [60]. Delays in puberty may therefore negatively affect bone development. Our investigation of pubertal trajectory in 213 females with RTT in Australia found that the median age of menarche (14 years) was slightly delayed compared to the normal population (12–13 years) [33] and a US study found that 19% of girls with RTT experience delayed menarche [34]. Additionally, being underweight or having the presence of the C-terminal or p.Arg168* MECP2 mutations correlated with delayed adrenarche, thelarche and menarche [33]. The increased height velocity and weight gain, which typically occur during puberty, have also been shown to be absent in RTT and height has been shown to be below the normative mean as early as 17 months of age in RTT [40]. Therefore the clinical assessment of bone health should include measurement of height and weight and assessment of pubertal development using Tanner staging [61].

Furthermore, the prescription of medications to regulate the menstrual cycle should be reconsidered, in particular the use of depot medroxyprogesterone acetate (DMPA) as its use has been found to increase the risk of fracture in individuals with developmental disabilities [62]. An additional class of medications that are highlighted as a risk factor for low bone density is the anticonvulsants [62], in particular valproate, which we have shown increases fracture risk three-fold in RTT, compared to the risk associated with no or any other prescribed antiepileptic drug [63].

The ISCD 2013 position paper on “Fracture Prediction and the Definition of Osteoporosis in Children and Adolescents” has recently updated its definition of osteoporosis in the young. It is now recognized that vertebral fractures can occur in at-risk children with BMD Z-scores better than the traditional -2.0 standard deviation threshold. As such, the new position paper states that BMD Z-score criteria are no longer required to define osteoporosis in children with vertebral fractures. This underscores the importance of determining whether vertebral fractures are present in high-risk populations such as RTT. According to the current ISCD criteria in children without vertebral fractures, a clinically significant fracture history plus a low BMD ≤-2.0 standard deviation are both required to diagnose osteoporosis. Clinically significant fractures include either two or more long bone fractures by age 10 years or three or more long bone fractures at any age up to 19 years [36]. It should be noted that there is marked variability in the age and gender-matched BMD Z-scores that are generated by different normative reference databases, which may place an individual above or below the absolute score of -2.0 depending on the normative data that is used [64]. This further highlights the need to investigate for the presence of both vertebral and non-vertebral fractures and monitor the occurrence of future fractures when considering osteoporosis in children. Overall, the diagnosis of osteoporosis is now based on a more functional measure of bone health—fracture history (including vertebral fractures), rather than BMD alone [65].

If vertebral fractures are present with or without low BMD z-scores or clinical significant fractures combined with low z-scores persist, the use of bisphosphonates should be considered [36]. Albeit there is only one single case study supporting its use in RTT (Lotan et al., 2013), bisphosphonates have become the standard of care in children with moderate to severe osteogenesis imperfecta where low BMC and aBMD are accompanied by recurrent long bone or vertebral compression fractures (Glorieux, 2007). However, it is still possible that the low bone mass observed in RTT may be due to fewer bone forming cells with decreased bone modeling as opposed to increased bone remodeling (Motil et al., 2014). Since bisphosphonates increase bone density by inhibiting osteoclasts reducing resorption of existing bone (Glorieux, 2007), they may have a less positive effect on bone mass in RTT than in the general population. With growth, there is continued bone accretion, which together with reduced bone resorption, increases cortical bone thickness and trabecular bone number [37]. During the first 12 months of bisphosphonate therapy in children with secondary osteoporosis, spine BMC and aBMD tend to increase significantly more than total body values [66]. Due to the paucity of evidence in RTT, re-evaluation after one year of treatment is recommended.

Fracture represents a substantial burden on those with RTT and their caregivers [67] and can occur unrecognized, given the decreased pain sensitivity and communication issues observed in RTT [68]. One goal of developing this document is to reduce the frequency of fractures and improve bone health of those with this disorder. The authors, along with an international clinical panel, have successfully developed guidelines using consensus methods, for the assessment, treatment and management of low bone density in RTT, taking into consideration the nutritional deficits and limited mobility in this disorder as well as other risk factors. Given the rarity of this disorder, these guidelines may be used by clinicians with little or no experience managing patients with RTT. Wherever possible, statements were supported by peer-reviewed literature. However in the absence of evidence, the guidelines relied heavily on expert panel opinion. The lack of supporting literature opens the door for further research to enhance our understanding of how to improve bone density in RTT and the identification of successful treatment protocols when bone mass and density are reduced.

Acknowledgments

The authors wish to thank our panel of parents from the Australian Consumer Reference Group for their useful insights and the International Rett Syndrome Foundation, Rettsyndrome.org for organising the listserv RettNet. We would also like to thank the UK for funding this stud, the NIH and NHMRC for funding the Australian Rett Syndrome research program, the NIH for funding the Natural History Study in the US and the International Rett Syndrome Foundation for funding InterRett research program. The authors also acknowledge the valuable contributions of the expert panel who participated in the Delphi technique. The expert panel comprised: Chhavi Agarwal, MD, MS, MRCP, FAAP, Pediatric Endocrinology of New York, New York, USA; Arthur Beisang, MD, Gillette Children's Hospital, St. Paul, Minnesota, USA; Bruria Ben-Zeev, MD, Pediatric Neurology Unit, Edmond & Lily Safra Children’s Hospital, Chaim Sheba Medical Center, Tel Hashomer, Israel; Anne-Marie Bisgaard, MD, PhD, Center for Rett Syndrome, Department of Clinical Genetics, Rigshospitalet, Copenhagen, Denmark; Torkel Brismar, MD, PhD, Department of Clinical Science, Intervention and Technology, Karolinska Institutet, Solna, Stockholm, Sweden; Angus Clarke, DM, FRCP, School of Medicine, Medical genetics, Cardiff University, Wales, UK; Giangennaro Coppola, MD, Clinic of Child and Adolescent Neuropsychiatry, Department of Medicine and Surgery, University of Salerno, Salerno, Italy; Matthew Drake, MD, PhD, Endocrinology, Diabetes, Metabolism and Nutrition Department, Mayo Clinic, Minnesota, USA; Carolyn Ellaway, MBBS, PhD, FRACP, Western Sydney Genetics Program, The Children’s Hospital at Westmead, Disciplines of Paediatrics and Child Health and Genetic Medicine, University of Sydney, Sydney, New South Wales, Australia; Michael Freilinger, MD, Department of Pediatrics and Adolescent Medicine, Medical University of Vienna, Vienna, Austria; Suzanne Geerts, RD, MS, Civitan International Research Center, University of Alabama at Birmingham, Birmingham, Alabama, USA; Richard Haas, MD, Department of Neurosciences, University of California San Diego, La Jolla, California, USA; Gabriele Haeusler, MD, Department of Pediatrics and Adolescent Medicine, Medical University of Vienna, Vienna, Italy; Ingrid Holm, MD, MPH, Division of genetics and Genomics, Boston Children's Hospital, Boston, USA; Peter Humphreys, MD, FRCP, Division of Neurology, Children’s Hospital of Eastern Ontario, Ottawa, Ontario, Canada; Mary Jones, MD, MPH, Katie's Clinic for Rett syndrome and related disorders, UCSF Benioff Children's Hospital, Oakland, California, USA; Walter Kaufmann, MD, Centre for Genetic Disorders of Cognition & Behavior, Kennedy Krieger Institute, School of Medicine, Johns Hopkins University, Baltimore; Salman Kirmani, BS, MBBS, MD, Mayo Clinic, Rochester, New York, USA; Jane Lane, RN BSN, Civitan International Research Center, University of Alabama at Birmingham, Birmingham, Alabama, USA; Gunilla Larsson, PhD, Swedish Rett Centre, Froson, Sweden; Agnes Linglart, MD, French Institute of Health and Medical Research, Pediatric Endocrinology and Diabetes Service, Bicêtre Hospital, Paris, France; Meir Lotan, PhD, Department of Physiotherapy, Ariel University, Ariel, Israel; Kathleen Motil, MD, PhD, Department of Pediatrics, USDA/ARS Children's Nutrition Research Center, Baylor College of Medicine, Section of Gastroenterology, Hepatology and Nutrition, Texas Children’s Hospital, Houston, Texas, USA; Craig Munns, MBBS, PhD, FRACP, Institute of Endocrinology and Diabetes, The Children’s Hospital at Westmead, Sydney, New South Wales, Australia; Yoshiko Nomura, MD, Segawa Neurological Clinic For Children, Tokyo, Japan; Alan Percy, MD, Department of Pediatrics and Neurology, University of Alabama at Birmingham, Birmingham, Alabama, USA; Mercedes Pineda, MD, PhD, Fundació Hospital Sant Joan de Déu, Centre for Biomedical Research on Rare Diseases, Instituto de Salud Carlos III,Barcelona, Spain; Carolyn Schanen, MD, PhD, Nemours Biomedical Research, Alfred I. duPont Hosptial for Children, Wilmington, DE, USA; Jay Shapiro, MD, Bone and Osteogenesis Imperfecta Department, Kennedy Krieger Institute, Baltimore, MD, USA; Kevin Sheridan, MD, Gillette Children's Hospital, St. Paul, Minnesota, USA; Steven Skinner, MD, Greenwood Genetic Center, Greenwood, South Carolina, USA; Bronwyn Stuckey, BA, MBBS, FRACP, Keogh Institute for Medical Research, Perth, WA, Australia; Birgit Syhler, BSc, Center for Rett Syndrome, Department of Clinical Genetics, Rigshospitalet, Copenhagen, Denmark; Daniel Tarquinio, DO, MS, Children’s Healthcare of Atlanta, Emory University, Atlanta, Georgia, USA; Sue Thompson, BSc, AdvAPD, Genetic Metabolic Disorders Service, The Children’s Hospital at Westmead, Sydney, New South Wales, Australia; Leanne Ward, MD, FRCPC, Research Chair in Pediatric Bone Health, Department of Pediatrics, Faculty of Medicine, University of Ottawa, Division of Endocrinology and Metabolism, Children’s Hospital of Eastern Ontario, Ottawa, Canada; Batia Weiss, MD, Division of Pediatric Gastroenterology and Nutrition, Edmond & Lily Safra Children's Hospital, Chaim Sheba Medical Center, Tel Hashomer, Israel; Ingegerd Witt Engerstrom, MD, PhD, Swedish National Rett Center, Frösön, Sweden.