Children and adolescents with rheumatic diseases, such as systemic lupus erythematosis (SLE) and juvenile idiopathic arthritis (JIA) are recognized to be at risk for impaired bone health. In those who develop overt osteoporosis, multiple risk factors are implicated, including increased concentrations of inflammatory cytokines, diminished nutrition, physical activity and muscle bulk [1, 2]. Use of glucocorticoids, in particular, is also known to compromise bone mass, quality and strength [3]. As a consequence, in children with rheumatic conditions, a 10 to 30 % prevalence rate of vertebral fractures has been reported [4–7], with a 6 % incidence of vertebral fractures within 12 months of glucocorticoid initiation [8].

Bone mineral density (BMD) as measured by dual-energy x-ray absorptiometry (DXA), is one of the most common tools used to assess bone health. Low BMD is associated with an increased risk for fractures in otherwise healthy children [9]. Children with rheumatic diseases including SLE and JIA have been shown to have reduced BMD compared to healthy age-matched controls [1, 10–12]. Decreased lumbar spine BMD Z score in the first 6 months of initiation of glucocorticoid treatment has been demonstrated to be an important predictor for increased vertebral fracture risk [13]. DXA has, however, limitations in the assessment of bone health in pediatric cohorts given that the results are influenced by bone size [14]. In addition bone strength is determined not only by bone density but also by the underlying bone quality and micro-architecture.

Histomorphometric analysis of trans-iliac bone samples allows characterization of bone microarchitecture including bone quantity, structure and metabolism. Histomorphometric studies in adults with rheumatoid arthritis have shown evidence of reduced trabecular bone volume with decreased trabecular thickness and bone turnover [15, 16]. Moreover, in adult cohorts, vertebral fractures have been demonstrated to relate to trabecular bone micro-architecture independent of the bone density [17].

The mineral content of bone has similarly been demonstrated to be an important determinant of bone stiffness and strength [18]. The degree and heterogeneity of bone mineralization can be measured from trans-iliac bone samples using quantitative backscattered electron imaging (qBSE). Deviations in bone mineralization have been demonstrated in patients with bone fragility, such as in children with osteogenesis imperfecta [19] and adults with postmenopausal osteoporosis [20], but have not been assessed to date in patients with rheumatic diseases.

Studies of bone histomorphometry and mineralization in pediatric patients treated with glucocorticoids have contributed to the understanding of underlying pathology and clinical factors with the potential to predict severity of bone disease [21, 22]. Given these features have not to date been described in pediatric rheumatalogical patients, the aims of this pilot study were therefore to characterize bone histomorphometry and mineralization in a small cohort of children with rheumatic diseases and to assess associations between clinical and radiological indices to bone micro-architecture.

All patients with a rheumatic disease who had been referred to the Bone Health Clinic at the Hospital for Sick Children, Toronto between 2003 and 2011 and who had undergone a trans-iliac bone biopsy were included in this retrospective descriptive analysis. Patients had been referred to the clinic for assessment of bone health because of the finding of at least one vertebral compression fracture. All patients were on calcium and vitamin D supplements and had undergone baseline investigations including biochemical assessment of parameters of calcium homeostasis, as well as spinal radiographs and DXA. The bone biopsy was undertaken as part of the clinical evaluation for symptomatic osteoporosis. Exclusion criteria included children with impaired renal function or other chronic medical conditions (other than their rheumatic disease) that could affect bone health. This study was approved by the Hospital for Sick Children’s Research Ethics Committee.

This study demonstrated that children with rheumatic diseases on chronic glucocorticoids who presented with vertebral fractures, have evidence of significantly decreased bone structural and formation histomorphometric indices compared to healthy age matched peers. The bone mineralization findings revealed greater heterogeneity of mineralization as well as a significant shift towards hypermineralization within the trabecular bone.

Patients with rheumatic conditions are at increased risk for bone fragility and fractures. [5–7]. Higher glucocorticoid dose have been shown to correlate with the presence of vertebral fractures [7, 8]. Similarly in our study, those children with greater cumulative glucocorticoid dose had a more severe grade of vertebral height loss. The results from this study add to this literature by demonstrating that the underlying histomorphometric deficits in trabecular bone in these children also relate to glucocorticoid dose. In our cohort, structural and formation parameters were lower than the age-specific reference range. Adults with rheumatoid arthritis have been shown to have evidence of decreased trabecular bone volume and turnover [15, 35]. Reduced trabecular thickness and osteoblast surface have similarly been found in other cohorts of children treated with glucocorticoids such as those with Duchene muscular dystrophy and steroid dependent nephrotic syndrome [21, 22]. Unlike, however, the histomorphometry results from the pediatric patients with nephrotic syndrome, our subjects did not have evidence of osteomalacia, as seen by the reduced osteoid thickness and volume. The primary effect of the glucocorticoid treatment in our subjects would appear to be through the reduction in bone formation decreasing osteoblastogenesis, increasing osteoblast apoptosis and inhibiting osteoblast production of bone matrix [36–38].

As well as the relationship between glucocorticoid dose and trabecular bone deficit, we were able to demonstrate that those children with more pronounced systemic effects from their steroid treatment (specifically reduced height and increased BMI Z score) have lower trabecular thickness and osteoblast surface. This association between increased BMI and greater bone fragility has also been demonstrated in a prospective study of children with rheumatic conditions which assessed determinants of vertebral fractures [13]. Combined, these findings would support that clinicians should have heightened vigilance around bone health screening in those rheumatological patients with systemic glucocorticoid effects.

In addition to structural and formation deficits, children with rheumatic diseases on glucocorticoids have significant differences in trabecular bone mineralization compared to healthy controls. Trabecular bone consists of individual osteons and bone packets that have different mineral contents depending on when they were produced during the (re) modeling cycle. During secondary mineralization the mineral content of these bone packets increases with time as free water in the bone matrix is replaced by mineral [18]. Bone mineralization as measured by qBSE quantifies the degree and heterogeneity of mineralization across the cross-sectional area of the trans-iliac bone biopsy, and is determined by bone turnover rate and by the time course of mineral accumulation within the bone matrix [18]. In our subjects, the shift in the bone mineralization towards a more hypermineralized profile would be consistent with decreased bone turnover leading to a greater percentage of mineralized bone packets. The inverse association between mineralization peak and histomorphometric bone formation parameters as well as serum bone turnover marker alkaline phosphatase, would support this finding. The increased heterogeneity of the matrix mineralization is also likely a result of a significant decrease in bone turnover. In experimental models, decreased bone turnover has been shown to lead to increased heterogeneity in bone mineralization [39]. Increased degree of mineralization through decreased bone remodeling is probably a significant contributor to the underlying bone fragility in our subjects. In other conditions, bone hypermineralization has been demonstrated to contribute to bone brittleness and tendency to fracture [40, 41].

BMD as measured by DXA is a commonly used screening tool to assess bone health in children on chronic glucocorticoids. A meta-analysis demonstrated a difference of −0.18 [−0.25; −0.1] g/m² lower spine BMD in children with JIA on glucocorticoids compared to healthy controls [42]. Similarly we demonstrated an association between average daily glucocorticoid dose and BMAD Z score. While low bone mass in pediatric patients has been defined by a BMD Z score of less than or equal to −2 [43], only 60 % of our cohort, all of whom had vertebral compression fractures, had a lumbar spine BMAD Z score that met this criteria. We also did not find a correlation of either lumbar spine areal BMD or BMAD to the histomorphometric or mineralization indices. While the density of bone is measured in BMD values, our findings would suggest that DXA BMD values do not always reflect some of the other underlying pathological mechanisms that glucocorticoids have on bone. In addition clinical variables such as short stature, delayed skeletal maturation, ethnicity and lean body mass can confound BMD results [43, 44]. The lack of association of bone histomorphometry to BMD highlights the limitations of relying on BMD values alone to evaluate bone health in children with chronic diseases. Given its invasive nature, trans-iliac bone biopsies are not suitable as a bone health screening tool, but should be limited to those patients who appear to have severe bone fragility out of keeping of their medical history or bone health risk factors. Instead, yearly lateral thoracolumbar spine X-rays to assess for vertebral height loss should be used in conjunction with BMD values to screen children with rheumatic conditions on glucocorticoids for evidence of bone fragility.

The treatment of bone fragility in children with rheumatic conditions is challenging, in part because of limited therapeutic options in the pediatric age range. Given higher rates of vitamin D deficiency in children with rheumatic conditions [45, 46], ensuring adequate vitamin D and calcium intake to maintain bone mineralization is an important first line therapy, although may not be sufficient to help prevent fragility fractures [13]. The use of biologic agents in JIA, has been demonstrated to help facilitate new bone formation after discontinuation of glucocorticoids [47]. These effects are however, attenuated in those patients who remain on glucocorticoids [48]. Bisphosphonates have been shown to improve BMD in children with secondary osteoporosis [49, 50]. Our findings of low baseline osteoblast activity in children on chronic glucocorticoids, emphasizes the need for cautious use of anti-resorptive agents such as bisphosphonates, which can lead to further suppression of bone turnover [21]. Consideration of commencing bisphosphonate therapy should be reserved for those patients presenting with significant fragility fractures or bone pain.

There are limitations to our study that deserve consideration, in particular the small number of subjects included in this analysis. Given the small subject numbers further data from a larger number of subjects is needed to confirm the findings from this pilot study. In addition the subjects in our study presented with a significant degree of bone fragility and vertebral compression fractures. It is uncertain to what degree similar histomorphological and mineralization deficits would be seen in a cohort of asymptomatic children on equivalent glucocorticoid doses. While we demonstrated a linear relationship between glucocorticoid dose and trabecular structural deficits, we cannot comment as to whether there is a threshold glucocorticoid dose below which there would not be a significant detrimental effect on bone quality. Only static and not dynamic histomorphometry was able to be performed, as the subjects had not received tetracycline prior to the biopsy being taken.

In conclusion, our results from this small pilot study demonstrate that children with rheumatic diseases on chronic glucocorticoid treatment who present with vertebral compression fractures have evidence of decreased trabecular bone volume and formation with increased and more heterogeneous mineralization. The association between glucocorticoid dose to the bone micro-architectural deficits, highlights the importance of decreasing the steroid dose in chronic disease states where possible. Treatment with bisphosphonates is sometimes used in children with glucocorticoid induced osteoporosis, however can lead to further suppression of bone formation [21]. Our findings support the need for further research in anabolic treatment options for children, in order to address the underlying bone pathology.