Curr Osteoporos Rep (2014) 12:313–318 DOI 10.1007/s11914-014-0224-1
PEDIATRICS (M LEONARD AND L WARD, SECTION EDITORS)
Fractures in Children with Cerebral Palsy M. Zulf Mughal
Published online: 26 June 2014 # Springer Science+Business Media New York 2014
Abstract Children with moderate to severe cerebral palsy are at increased risk of sustaining fracture following minimal trauma. Such fractures predominantly occur in lower limb bones and are associated with low bone mineral density. Risk factors for fracture in this group include nonambulatory status, anticonvulsant use, presence of a joint contractures, immobilization after surgery, and poor nutrition. Aims of this review are to describe the prevalence and pathogenesis of fractures in nonambulant children with cerebral palsy. Interventions and treatments that improve low bone mineral density and which may help to reduce the fracture risk in this population are also discussed. Keywords Cerebral palsy . Fractures . Bone mineral density . Physical therapy . Bisphosphonates
Introduction Cerebral palsy (CP) is a term used to describe a group of nonprogressive disorders of movement and posture, resulting from an insult to the developing brain. It is one of the commonest chronic disabling conditions of childhood, with prevalence of 2.0–3.5 per 1000 live births [1]. The most common motor abnormality is spasticity, which may be categorized into diplegia, hemiplegia, and quadriplegia. Other forms of CP include dystonia choreo-athetosis, ataxic or a mixture of these disorders. The Gross Motor Function Classification System (GMFCS) [2] is a widely used five level clinical standardized system to classify the gross motor function of children with CP, with emphasis on function in sitting and walking. Children with CP often have other disabilities, which may affect their quality of life and life expectancy. M. Z. Mughal (*) Royal Manchester Children’s Hospital, Central Manchester University Hospitals NHS Foundation Trust, Oxford Road, Manchester M13 9WL, UK e-mail: [email protected]
These include intellectual impairment, behavioral problems, hearing and visual problems, feeding difficulties, poor growth, recurrent respiratory infections, and epilepsy. Secondary musculoskeletal problems include joint contractures, kyphoscoliosis and hip subluxation.
Fractures in Children with Cerebral Palsy Children and adolescents with CP are prone to low trauma fractures, which occur during normal activities such as dressing and transferring [3•, 4]. Such fractures are more common in nonambulant children who are at the severe end of the spectrum of CP, defined as level IV or V according to the GMFCS [5•]. Fractures not only cause pain but they further limit the mobility of children with CP leading to muscle wasting through disuse, hospitalization, and missed schooling. The cause of injury may not be clear in over 50 % of cases, and delay in diagnosis is not uncommon [6]. In the literature, the term ‘spontaneous fracture’ has been used to describe such fractures as they apparently occur without any known external cause [7]. Lack of a clear history of the injury causing the fracture, difficulties in communication, and delay in presentation sometimes lead to suspicion of child abuse [4, 7]. There have been a few large epidemiologic studies of fragility fractures in children with CP. In a study of 763 such children, Leet and colleagues reported a fracture prevalence of 12 % [8]. In another large study of 1637 patients with CP, Presedo and colleagues reported a fracture prevalence of 6 % [6]. More recently Maruyama et al studied fracture prevalence in 525 children attending 38 schools for physically disabled children in the Hokuriku-Koshinetsu District of Japan [9]. In this population, they reported 1 year and lifetime fracture prevalence of 3.6 % and 9.7 %, respectively. In children with moderate to severe CP (GMFCS III to V) Stevenson and colleagues reported a fracture incidence of 4 % per year [10]. Brunner and Doderlein [11] surveyed 37 patients with CP who had sustained 54 fractures with minimal trauma and found that the majority (74 %) were in the femoral shaft and
314
the supracondylar region. In a population-based study of 763 CP patients (1.3–18 years), Leet et al found that over 70 % of fractures occurred in lower limb bones [8]. In the study by Presedo et al over 80 % of fractures occurred in the lower limbs [6]. These investigators also found that over 10 % of CP children developed complications after a fracture, which included further fractures, malunion, nonunion, and infections, including pneumonia [6]. In summary, nonambulant children with CP are prone to fragility fractures. Such fractures predominantly occur in lower limb bones and are associated with a high complication rate.
Risk Factors for Fractures in Children with Cerebral Palsy Development of bone is dependent on an inherited genetic ‘template’ that is modified by mechanical loading, and nutritional and endocrine environments. Postnatal bone development is modulated by the mechanical forces to which it is subjected to, which primarily arise from muscle contractions [12]. The skeleton of a healthy growing child continuously adapts to increasing mechanical loading from bigger and stronger muscles by increasing bone mass and altering bone geometry [13]. By contrast, in a disabled nonambulant child with CP, muscle weakness and habitual inability to participate in normal load-bearing activities results in reduced periosteal bone expansion. This leads to development of slender long bones, which have increased propensity to fracture. Therefore, it is not surprising that radiographs of the bones of CP children often show signs of impaired bone development, such as slender or ‘gracile’ long bones. They may also show thinning of cortices, prominent trabecular pattern due to osteopenia, or the “washed-out appearance” (Fig. 1).
Curr Osteoporos Rep (2014) 12:313–318
Measurement of bone mineral density (BMD) and bone size parameters using dual energy x-ray absorptiometry (DXA), peripheral quantitative computed tomography (pQCT), and quantitative computed tomography (QCT) in children with CP have provided insight into pathogenesis of skeletal fragility in this setting. In a prospective longitudinal study of 69 subjects with moderate to severe spastic CP (ages 2–17.7 years; GMFC levels III to V), Henderson et al [14] noted that DXA measured lateral distal femoral areal BMD (aBMD) Z-scores decreased in spite of the increases in aBMD. These results suggest that bone mass accrual in children with CP was less than that expected in healthy children. Using pQCT, Binkley and colleagues showed that subjects with CP had reduced distal tibial cortical bone area, periosteal and endosteal circumferences, thickness, bone mineral content, and polar strength-strain index, compared with controls [15]. Wren and colleagues found that midtibial cross-sectional area and cortical bone area, measured using QCT, was significantly lower in CP children with GMFCS levels III and IV in comparison with those with GMFCS levels I and II [16]. Taken together, results of these studies suggest that increased fragility of a long bone, such as the femur, in subjects with CP arises because of slower rate of bone mass accrual, smaller periosteal diameters, and thinner cortices. As detailed below, low areal bone mineral density (aBMD) measured by DXA in the distal lateral femur is associated with fracture risk in children with CP. Besides nonambulatory status, anticonvulsant use, undernutrition requiring gastrostomy feeding and low vitamin D status are associated with low BMD in children with CP [3•, 5•]. While, these factors might adversely affect bone development and mineralization, Henderson has suggested that anticonvulsant use and gastrostomy feeding are more common in nonambulant CP children, and, therefore, these risk factors may merely be markers of CP ‘severity’ [17]. Previous fracture was also a predictor of subsequent fractures, presumably due to bone loss associated with immobilization [5•, 10]. Immobilization of the hip in a spica following surgery was also associated with an increase in risk of fracture [18, 19]. Pre-existent contracture, stiff or dislocated joints, which restrict limb movements, also increase fracture risk [11]. In summary, pre-existing factors that limit limb movements, prior fracture, and postfracture or postoperative immobilization are associated with increased risk of further fracture in children with CP.
Assessment of Bone Mineral Density in Children with Cerebral Palsy Fig. 1 A supracondylar femoral fracture in a nonambulant a 13-year old boy with quadriplegic CP. The fracture came to light when he appeared to be distressed and his right knee was noted to be swollen; there was no history of trauma. Note thinning of cortices and osteopenia or the “washed-out appearance” of bones
A number of studies have reported low bone mineral density (BMD), defined as areal BMD Z-scores less than or equal to – 2.0 SD, in children with CP. A systematic review by Mergler
Curr Osteoporos Rep (2014) 12:313–318
and colleagues found that impaired ambulation, feeding difficulties, previous fractures, anticonvulsant use, and lower body fat mass were associated with low BMD Z-scores [5•]. In severely disabled children with CP, whole body, hip, or spine assessment of BMD by DXA can be challenging because of joint contractures, hip dysplasia, scoliosis, or use of metal fixation devices. As mentioned previously, femur is the most common site of fracture in nonambulant children with CP [6, 8, 11]. Harcke and colleagues have described the technique for measuring aBMD at the distal femoral region, the most common fracture site, in children with CP [20, 21]. The lateral distal femur (LDF) scans are divided into 3 rectangular subregions, representing metaphyseal bone (LDF Region 1), the transition zone from the metaphysis to the diaphysis (LDF Region 2), and diaphyseal bone (LDF Region 3). Zemel et al have published LDF reference data based on over 800 children and adolescents [22••]. At this site, the prevalence of aBMD Z-score less than – 2, was 77 % [23]. Henderson and colleagues studied the relation between LDF aBMD Z-scores and fracture history in a cross-sectional study of 619 children aged 6 to 18 years with muscular dystrophy (n=112) or moderate to severe cerebral palsy (n=507), cared for at eight canters in the USA [24]. There was a strong correlation between fracture history and LDF aBMD Z-scores; 35 %–42 % of those with aBMD Z-scores less than –5 had fractured compared with 13 %–15 % of those with aBMD Z-scores greater than –1. Each 1.0 decrease in LDF aBMD Z-scores increased the fracture risk by 6 %–15 %. In a cross-sectional study of 59 children with motor disabilities (CP, myelomeningocele, muscular dystrophy, and syndromes causing motor disability; GMFCS levels II–V), which involved systematic radiographic screening, KilpinenLoisa et al found compression vertebral fractures in 25 % of subjects [25]. In this study, 17 % of subjects had sustained peripheral fractures after minor trauma. Low spinal bone mineral apparent density z-score and hypercalciuria were significant predictors for fractures, with five- and sevenfold risk for fractures, respectively. In summary, nonambulant children with CP have low BMD, which is associated with increased risk of fracture. Robust normative data are available for estimation of BMD at the LDF site. In children with chronic immobilization, the 2013 Pediatric Official Positions of the International Society for Clinical Densitometry recommends assessment of BMD at the LDF site, which is the most common fracture site in children with CP [26••].
The Role of Physical Therapy Interventions in Improving Bone Mineral Density in Children with Cerebral Palsy A number of small trials of physical therapy in children and adolescents with CP have been undertaken with BMD as an outcome measure.
315
Chad et al [27] undertook an 8-month program of weight physical activity intervention, consisting of upper, lower, and trunk exercises, in nine children with CP, and a same number of controls. They observed a 5.6 % increase in femoral neck volumetric BMD in children in the intervention group, compared with a –6.3 % change in the control group. Caulton et al undertook a randomized controlled trial (RCT) of the effects of an increased duration of static standing program in 26 nonambulant CP children, aged between 4.3–10.8 years [28]. CP children in the intervention group stood for 50 % longer than their usual duration in their standing frames while those in the control group continued with their usual duration of standing. After 9 months, the mean vertebral volumetric trabecular BMD (vTBMD), estimated by QCT, in the intervention group showed a 6 % increase (P=0.01) compared with the control group. However, no change was observed in the vTBMD of the proximal tibia in response to longer period of static standing program. Ward et al undertook a pilot RCT of low magnitude, high frequency mechanical loading therapy in a heterogeneous group of disabled but ambulant subjects [29]. Twenty subjects aged 4–19 years stood on vibrating platforms (90 Hz, 0.3 gravity) or placebo platforms for 10 min/day, 5 days/week for a 6-month duration. The proximal tibial vTBMD of children who stood on active devices increased by 6.3 % over baseline value. In contrast there was significant drop (11.9 % decrease) in vTBMD measured in subjects who stood on a placebo device (net benefit of treatment 17.7 %, P=0.003). Another RCT using the same mode of vibration therapy, 31 children with CP aged 6–12 years (GMFS I–IV) stood on a vibrating platform (at home for 10 minutes/day for 6 months, and on the floor without the platform for another 6 months [30]. There was there was a significantly greater increase in QCT measured tibial cortical bone area and moments of inertia during the vibration period (P ≤ 0.03). In contrast, a 6-month pilot RCT of 9 minutes of side to side alternating whole body vibration therapy (peak vibration frequency of 18 Hz and acceleration of 2.6 g) administered daily, during school days in 5- to 13-year-old CP children (GMFCS III–IV) did not increase DXA measured aBMD of the lumbar spine [31]. The LDF aBMD of metaphyses and diaphysis regions decreased in the treatment group and increased in the control group, a finding that was opposite to what was predicted. Whole body vibration therapy did result in a significant improvement in the mobility function, as assessed by the speed in the 10-minute walk test. Damcott and colleagues studied changes in LDF aBMD in five children with CP (age 4–9 years; GMFCS III to IV) who stood on dynamic sanders (30 min/day, 5 days/week, for 15 months) in which novel footplates were incorporated to provide reciprocal loading that mimics the forces applied to the lower limbs during the natural walking gait [32]. Four control subjects stood on static standers. Dynamic standing resulted in a significant increase in LDF aBMD, whereas in
316
passive standers the baseline aBMD appeared to be preserved. A 12-week, home based RCT of virtual cycling training in ambulatory children with spastic CP (age 6–12 years; GMFCS I–II) resulted in a significant increased distal femoral aBMD, but not lumbar spine aBMD [33]. In summary, results of these small preliminary trails suggest that dynamic physical therapy programs may have the potential of improving BMD at fracture prone sites in children with CP. Adequately powered RCTs, ideally with fracture prevention as an outcome measure, are required to confirm these preliminary findings.
Curr Osteoporos Rep (2014) 12:313–318
which in the opinion of the paper’s authors, had acted as “stress risers” [a mechanical defect(s) that concentrates stress in that area]. In summary, cyclical intravenous pamidronate therapy in nonambulant children with CP results in increase in BMD and reduction in fracture rate. In their recent systematic review, Fehlings et al [41••] recommended that the lack of long-term efficacy data and potential risk of side effects needs to be carefully weighed when considering the use of bisphosphonates in this population. Since cyclical treatment regimens appear to be creating stress risers, trials of alternative treatment regimens may be required in this group of patients.
Bisphosphonate Therapy in Children with CP Bisphosphonates are a group of drugs that inhibit osteoclastic bone resorption resulting is an increase in bone mass, improvement in bone strength and reduction in the risk of fragility fractures. In children with moderate to severe osteogenesis imperfecta, cyclical intravenous pamidronate disodium pentahydrate (pamidronate) therapy has been associated with improvements in vertebral BMD, improvement in bone pain, and reduction in fracture rate [34]. Systemically administered bisphosphonates, have also been used to treat children with secondary osteoporosis. A number of cross-sectional studies have shown an improvement in vertebral and femoral aBMD in nonambulant children with CP treated with cyclical intravenous pamidronate [35–37]. Henderson and colleagues [36] conducted a placebo-controlled RCT of intravenous pamidronate in 6 pairs of nonambulatory CP children, for 12 months. The aBMD of the lateral distal femoral metaphyseal region (LDF region 1) increased by 89 % in the intervention group, whereas the control group had a mean increase of 9 %. During the 18-month trial period, there were three fractures in placebo treated subjects and none in pamidronate treated subjects. The observed increase in aBMD in region 1 of the LDF had virtually returned back to baseline 2 years after the treatment was discontinued [38]. Despite this, a retrospective study by the same group of investigators [39], showed that 1 year of cyclical pamidronate therapy in 25 nonambulatory children with CP (GMFCS IV and V) with a prior history of fractures, resulted in a significant (P=0.02) decrease in fracture rate from 30.6 % per year to 13 % per year, over a mean post-treatment follow-up period of approximately 4 years. These data suggest that the protective effect of intravenous pamidronate treatment persists long after the bone density values have returned to baseline. A small subset of this cohort experienced fractures immediately upon stopping of pamidronate treatment. Recently, Harcke and colleagues undertook a review of radiographs of CP children who and suffered fractures following cyclical intravenous pamidronate therapy [40] Approximately 60 % percent of fractures were located adjacent to the margin of a ‘pamidronate bands’,
Calcium and Vitamin D Supplementation in Children with Cerebral Palsy Low body stores of vitamin D (as reflected by serum 25hydroxyvitamin D concentrations) are prevalent are prevalent in nonambulant CP children and they might contribute to low BMD in this population [42]. In a small study of institutionalized severely disabled Black South African children and adults with CP, Bischof et al [43] found vitamin D status of subjects to be an important factor in the etiology of fractures. These investigators also reported an association between the number of fractures and use of anticonvulsants; older anticonvulsants, such as phenobarbitone, are known to increase catabolism of vitamin D. Three months of vitamin D administration (calciferol 5000 iu/day) resulted in a significant decrease in serum alkaline phosphatase activity, and a significant increase in serum calcium and phosphate levels. In a pilot study, Jekovec-Vrhovsek et al [44] treated 13 children with the severe form of CP in full-term care, who had been treated with anticonvulsants, with 1,25-dihydroxy-cholecalciferol vitamin D (0.25 mcg daily) and with calcium (500 mg daily) for 9 months. Seven control children acted as controls. During the study period lumbar spine aBMD in the treated group increased (P
PEDIATRICS (M LEONARD AND L WARD, SECTION EDITORS)
Fractures in Children with Cerebral Palsy M. Zulf Mughal
Published online: 26 June 2014 # Springer Science+Business Media New York 2014
Abstract Children with moderate to severe cerebral palsy are at increased risk of sustaining fracture following minimal trauma. Such fractures predominantly occur in lower limb bones and are associated with low bone mineral density. Risk factors for fracture in this group include nonambulatory status, anticonvulsant use, presence of a joint contractures, immobilization after surgery, and poor nutrition. Aims of this review are to describe the prevalence and pathogenesis of fractures in nonambulant children with cerebral palsy. Interventions and treatments that improve low bone mineral density and which may help to reduce the fracture risk in this population are also discussed. Keywords Cerebral palsy . Fractures . Bone mineral density . Physical therapy . Bisphosphonates
Introduction Cerebral palsy (CP) is a term used to describe a group of nonprogressive disorders of movement and posture, resulting from an insult to the developing brain. It is one of the commonest chronic disabling conditions of childhood, with prevalence of 2.0–3.5 per 1000 live births [1]. The most common motor abnormality is spasticity, which may be categorized into diplegia, hemiplegia, and quadriplegia. Other forms of CP include dystonia choreo-athetosis, ataxic or a mixture of these disorders. The Gross Motor Function Classification System (GMFCS) [2] is a widely used five level clinical standardized system to classify the gross motor function of children with CP, with emphasis on function in sitting and walking. Children with CP often have other disabilities, which may affect their quality of life and life expectancy. M. Z. Mughal (*) Royal Manchester Children’s Hospital, Central Manchester University Hospitals NHS Foundation Trust, Oxford Road, Manchester M13 9WL, UK e-mail: [email protected]
These include intellectual impairment, behavioral problems, hearing and visual problems, feeding difficulties, poor growth, recurrent respiratory infections, and epilepsy. Secondary musculoskeletal problems include joint contractures, kyphoscoliosis and hip subluxation.
Fractures in Children with Cerebral Palsy Children and adolescents with CP are prone to low trauma fractures, which occur during normal activities such as dressing and transferring [3•, 4]. Such fractures are more common in nonambulant children who are at the severe end of the spectrum of CP, defined as level IV or V according to the GMFCS [5•]. Fractures not only cause pain but they further limit the mobility of children with CP leading to muscle wasting through disuse, hospitalization, and missed schooling. The cause of injury may not be clear in over 50 % of cases, and delay in diagnosis is not uncommon [6]. In the literature, the term ‘spontaneous fracture’ has been used to describe such fractures as they apparently occur without any known external cause [7]. Lack of a clear history of the injury causing the fracture, difficulties in communication, and delay in presentation sometimes lead to suspicion of child abuse [4, 7]. There have been a few large epidemiologic studies of fragility fractures in children with CP. In a study of 763 such children, Leet and colleagues reported a fracture prevalence of 12 % [8]. In another large study of 1637 patients with CP, Presedo and colleagues reported a fracture prevalence of 6 % [6]. More recently Maruyama et al studied fracture prevalence in 525 children attending 38 schools for physically disabled children in the Hokuriku-Koshinetsu District of Japan [9]. In this population, they reported 1 year and lifetime fracture prevalence of 3.6 % and 9.7 %, respectively. In children with moderate to severe CP (GMFCS III to V) Stevenson and colleagues reported a fracture incidence of 4 % per year [10]. Brunner and Doderlein [11] surveyed 37 patients with CP who had sustained 54 fractures with minimal trauma and found that the majority (74 %) were in the femoral shaft and
314
the supracondylar region. In a population-based study of 763 CP patients (1.3–18 years), Leet et al found that over 70 % of fractures occurred in lower limb bones [8]. In the study by Presedo et al over 80 % of fractures occurred in the lower limbs [6]. These investigators also found that over 10 % of CP children developed complications after a fracture, which included further fractures, malunion, nonunion, and infections, including pneumonia [6]. In summary, nonambulant children with CP are prone to fragility fractures. Such fractures predominantly occur in lower limb bones and are associated with a high complication rate.
Risk Factors for Fractures in Children with Cerebral Palsy Development of bone is dependent on an inherited genetic ‘template’ that is modified by mechanical loading, and nutritional and endocrine environments. Postnatal bone development is modulated by the mechanical forces to which it is subjected to, which primarily arise from muscle contractions [12]. The skeleton of a healthy growing child continuously adapts to increasing mechanical loading from bigger and stronger muscles by increasing bone mass and altering bone geometry [13]. By contrast, in a disabled nonambulant child with CP, muscle weakness and habitual inability to participate in normal load-bearing activities results in reduced periosteal bone expansion. This leads to development of slender long bones, which have increased propensity to fracture. Therefore, it is not surprising that radiographs of the bones of CP children often show signs of impaired bone development, such as slender or ‘gracile’ long bones. They may also show thinning of cortices, prominent trabecular pattern due to osteopenia, or the “washed-out appearance” (Fig. 1).
Curr Osteoporos Rep (2014) 12:313–318
Measurement of bone mineral density (BMD) and bone size parameters using dual energy x-ray absorptiometry (DXA), peripheral quantitative computed tomography (pQCT), and quantitative computed tomography (QCT) in children with CP have provided insight into pathogenesis of skeletal fragility in this setting. In a prospective longitudinal study of 69 subjects with moderate to severe spastic CP (ages 2–17.7 years; GMFC levels III to V), Henderson et al [14] noted that DXA measured lateral distal femoral areal BMD (aBMD) Z-scores decreased in spite of the increases in aBMD. These results suggest that bone mass accrual in children with CP was less than that expected in healthy children. Using pQCT, Binkley and colleagues showed that subjects with CP had reduced distal tibial cortical bone area, periosteal and endosteal circumferences, thickness, bone mineral content, and polar strength-strain index, compared with controls [15]. Wren and colleagues found that midtibial cross-sectional area and cortical bone area, measured using QCT, was significantly lower in CP children with GMFCS levels III and IV in comparison with those with GMFCS levels I and II [16]. Taken together, results of these studies suggest that increased fragility of a long bone, such as the femur, in subjects with CP arises because of slower rate of bone mass accrual, smaller periosteal diameters, and thinner cortices. As detailed below, low areal bone mineral density (aBMD) measured by DXA in the distal lateral femur is associated with fracture risk in children with CP. Besides nonambulatory status, anticonvulsant use, undernutrition requiring gastrostomy feeding and low vitamin D status are associated with low BMD in children with CP [3•, 5•]. While, these factors might adversely affect bone development and mineralization, Henderson has suggested that anticonvulsant use and gastrostomy feeding are more common in nonambulant CP children, and, therefore, these risk factors may merely be markers of CP ‘severity’ [17]. Previous fracture was also a predictor of subsequent fractures, presumably due to bone loss associated with immobilization [5•, 10]. Immobilization of the hip in a spica following surgery was also associated with an increase in risk of fracture [18, 19]. Pre-existent contracture, stiff or dislocated joints, which restrict limb movements, also increase fracture risk [11]. In summary, pre-existing factors that limit limb movements, prior fracture, and postfracture or postoperative immobilization are associated with increased risk of further fracture in children with CP.
Assessment of Bone Mineral Density in Children with Cerebral Palsy Fig. 1 A supracondylar femoral fracture in a nonambulant a 13-year old boy with quadriplegic CP. The fracture came to light when he appeared to be distressed and his right knee was noted to be swollen; there was no history of trauma. Note thinning of cortices and osteopenia or the “washed-out appearance” of bones
A number of studies have reported low bone mineral density (BMD), defined as areal BMD Z-scores less than or equal to – 2.0 SD, in children with CP. A systematic review by Mergler
Curr Osteoporos Rep (2014) 12:313–318
and colleagues found that impaired ambulation, feeding difficulties, previous fractures, anticonvulsant use, and lower body fat mass were associated with low BMD Z-scores [5•]. In severely disabled children with CP, whole body, hip, or spine assessment of BMD by DXA can be challenging because of joint contractures, hip dysplasia, scoliosis, or use of metal fixation devices. As mentioned previously, femur is the most common site of fracture in nonambulant children with CP [6, 8, 11]. Harcke and colleagues have described the technique for measuring aBMD at the distal femoral region, the most common fracture site, in children with CP [20, 21]. The lateral distal femur (LDF) scans are divided into 3 rectangular subregions, representing metaphyseal bone (LDF Region 1), the transition zone from the metaphysis to the diaphysis (LDF Region 2), and diaphyseal bone (LDF Region 3). Zemel et al have published LDF reference data based on over 800 children and adolescents [22••]. At this site, the prevalence of aBMD Z-score less than – 2, was 77 % [23]. Henderson and colleagues studied the relation between LDF aBMD Z-scores and fracture history in a cross-sectional study of 619 children aged 6 to 18 years with muscular dystrophy (n=112) or moderate to severe cerebral palsy (n=507), cared for at eight canters in the USA [24]. There was a strong correlation between fracture history and LDF aBMD Z-scores; 35 %–42 % of those with aBMD Z-scores less than –5 had fractured compared with 13 %–15 % of those with aBMD Z-scores greater than –1. Each 1.0 decrease in LDF aBMD Z-scores increased the fracture risk by 6 %–15 %. In a cross-sectional study of 59 children with motor disabilities (CP, myelomeningocele, muscular dystrophy, and syndromes causing motor disability; GMFCS levels II–V), which involved systematic radiographic screening, KilpinenLoisa et al found compression vertebral fractures in 25 % of subjects [25]. In this study, 17 % of subjects had sustained peripheral fractures after minor trauma. Low spinal bone mineral apparent density z-score and hypercalciuria were significant predictors for fractures, with five- and sevenfold risk for fractures, respectively. In summary, nonambulant children with CP have low BMD, which is associated with increased risk of fracture. Robust normative data are available for estimation of BMD at the LDF site. In children with chronic immobilization, the 2013 Pediatric Official Positions of the International Society for Clinical Densitometry recommends assessment of BMD at the LDF site, which is the most common fracture site in children with CP [26••].
The Role of Physical Therapy Interventions in Improving Bone Mineral Density in Children with Cerebral Palsy A number of small trials of physical therapy in children and adolescents with CP have been undertaken with BMD as an outcome measure.
315
Chad et al [27] undertook an 8-month program of weight physical activity intervention, consisting of upper, lower, and trunk exercises, in nine children with CP, and a same number of controls. They observed a 5.6 % increase in femoral neck volumetric BMD in children in the intervention group, compared with a –6.3 % change in the control group. Caulton et al undertook a randomized controlled trial (RCT) of the effects of an increased duration of static standing program in 26 nonambulant CP children, aged between 4.3–10.8 years [28]. CP children in the intervention group stood for 50 % longer than their usual duration in their standing frames while those in the control group continued with their usual duration of standing. After 9 months, the mean vertebral volumetric trabecular BMD (vTBMD), estimated by QCT, in the intervention group showed a 6 % increase (P=0.01) compared with the control group. However, no change was observed in the vTBMD of the proximal tibia in response to longer period of static standing program. Ward et al undertook a pilot RCT of low magnitude, high frequency mechanical loading therapy in a heterogeneous group of disabled but ambulant subjects [29]. Twenty subjects aged 4–19 years stood on vibrating platforms (90 Hz, 0.3 gravity) or placebo platforms for 10 min/day, 5 days/week for a 6-month duration. The proximal tibial vTBMD of children who stood on active devices increased by 6.3 % over baseline value. In contrast there was significant drop (11.9 % decrease) in vTBMD measured in subjects who stood on a placebo device (net benefit of treatment 17.7 %, P=0.003). Another RCT using the same mode of vibration therapy, 31 children with CP aged 6–12 years (GMFS I–IV) stood on a vibrating platform (at home for 10 minutes/day for 6 months, and on the floor without the platform for another 6 months [30]. There was there was a significantly greater increase in QCT measured tibial cortical bone area and moments of inertia during the vibration period (P ≤ 0.03). In contrast, a 6-month pilot RCT of 9 minutes of side to side alternating whole body vibration therapy (peak vibration frequency of 18 Hz and acceleration of 2.6 g) administered daily, during school days in 5- to 13-year-old CP children (GMFCS III–IV) did not increase DXA measured aBMD of the lumbar spine [31]. The LDF aBMD of metaphyses and diaphysis regions decreased in the treatment group and increased in the control group, a finding that was opposite to what was predicted. Whole body vibration therapy did result in a significant improvement in the mobility function, as assessed by the speed in the 10-minute walk test. Damcott and colleagues studied changes in LDF aBMD in five children with CP (age 4–9 years; GMFCS III to IV) who stood on dynamic sanders (30 min/day, 5 days/week, for 15 months) in which novel footplates were incorporated to provide reciprocal loading that mimics the forces applied to the lower limbs during the natural walking gait [32]. Four control subjects stood on static standers. Dynamic standing resulted in a significant increase in LDF aBMD, whereas in
316
passive standers the baseline aBMD appeared to be preserved. A 12-week, home based RCT of virtual cycling training in ambulatory children with spastic CP (age 6–12 years; GMFCS I–II) resulted in a significant increased distal femoral aBMD, but not lumbar spine aBMD [33]. In summary, results of these small preliminary trails suggest that dynamic physical therapy programs may have the potential of improving BMD at fracture prone sites in children with CP. Adequately powered RCTs, ideally with fracture prevention as an outcome measure, are required to confirm these preliminary findings.
Curr Osteoporos Rep (2014) 12:313–318
which in the opinion of the paper’s authors, had acted as “stress risers” [a mechanical defect(s) that concentrates stress in that area]. In summary, cyclical intravenous pamidronate therapy in nonambulant children with CP results in increase in BMD and reduction in fracture rate. In their recent systematic review, Fehlings et al [41••] recommended that the lack of long-term efficacy data and potential risk of side effects needs to be carefully weighed when considering the use of bisphosphonates in this population. Since cyclical treatment regimens appear to be creating stress risers, trials of alternative treatment regimens may be required in this group of patients.
Bisphosphonate Therapy in Children with CP Bisphosphonates are a group of drugs that inhibit osteoclastic bone resorption resulting is an increase in bone mass, improvement in bone strength and reduction in the risk of fragility fractures. In children with moderate to severe osteogenesis imperfecta, cyclical intravenous pamidronate disodium pentahydrate (pamidronate) therapy has been associated with improvements in vertebral BMD, improvement in bone pain, and reduction in fracture rate [34]. Systemically administered bisphosphonates, have also been used to treat children with secondary osteoporosis. A number of cross-sectional studies have shown an improvement in vertebral and femoral aBMD in nonambulant children with CP treated with cyclical intravenous pamidronate [35–37]. Henderson and colleagues [36] conducted a placebo-controlled RCT of intravenous pamidronate in 6 pairs of nonambulatory CP children, for 12 months. The aBMD of the lateral distal femoral metaphyseal region (LDF region 1) increased by 89 % in the intervention group, whereas the control group had a mean increase of 9 %. During the 18-month trial period, there were three fractures in placebo treated subjects and none in pamidronate treated subjects. The observed increase in aBMD in region 1 of the LDF had virtually returned back to baseline 2 years after the treatment was discontinued [38]. Despite this, a retrospective study by the same group of investigators [39], showed that 1 year of cyclical pamidronate therapy in 25 nonambulatory children with CP (GMFCS IV and V) with a prior history of fractures, resulted in a significant (P=0.02) decrease in fracture rate from 30.6 % per year to 13 % per year, over a mean post-treatment follow-up period of approximately 4 years. These data suggest that the protective effect of intravenous pamidronate treatment persists long after the bone density values have returned to baseline. A small subset of this cohort experienced fractures immediately upon stopping of pamidronate treatment. Recently, Harcke and colleagues undertook a review of radiographs of CP children who and suffered fractures following cyclical intravenous pamidronate therapy [40] Approximately 60 % percent of fractures were located adjacent to the margin of a ‘pamidronate bands’,
Calcium and Vitamin D Supplementation in Children with Cerebral Palsy Low body stores of vitamin D (as reflected by serum 25hydroxyvitamin D concentrations) are prevalent are prevalent in nonambulant CP children and they might contribute to low BMD in this population [42]. In a small study of institutionalized severely disabled Black South African children and adults with CP, Bischof et al [43] found vitamin D status of subjects to be an important factor in the etiology of fractures. These investigators also reported an association between the number of fractures and use of anticonvulsants; older anticonvulsants, such as phenobarbitone, are known to increase catabolism of vitamin D. Three months of vitamin D administration (calciferol 5000 iu/day) resulted in a significant decrease in serum alkaline phosphatase activity, and a significant increase in serum calcium and phosphate levels. In a pilot study, Jekovec-Vrhovsek et al [44] treated 13 children with the severe form of CP in full-term care, who had been treated with anticonvulsants, with 1,25-dihydroxy-cholecalciferol vitamin D (0.25 mcg daily) and with calcium (500 mg daily) for 9 months. Seven control children acted as controls. During the study period lumbar spine aBMD in the treated group increased (P
