Low-Dose Pamidronate Therapy for Pediatric Osteoporosis: Influence of Diagnosis on Changes in Fracture Rate and Bone Mineral Density

Special Article - Pediatric Endocrinology

J Pediatr & Child Health Care. 2017; 2(2): 1016.

Low-Dose Pamidronate Therapy for Pediatric Osteoporosis: Influence of Diagnosis on Changes in Fracture Rate and Bone Mineral Density

Crossen SS¹, Pederson J², Kumar RB³ and Bachrach LK³*

¹Department of Pediatrics, University of California at Davis, USA

²Department of Pediatrics, Feinberg School of Medicine, Northwestern University, USA

³Stanford University School of Medicine, USA

*Corresponding author: Laura K. Bachrach, Room H 314, Stanford Medical Center, 300 Pasteur Drive, Stanford, USA

Received: June 20, 2017; Accepted: September 12, 2017; Published: September 19, 2017


Controversy surrounds the optimal agent, dose and duration of bisphosphonate therapy for pediatric osteoporosis. We conducted a prospective, observational study of low-dose (4 mg/kg/year) intravenous pamidronate in 31 children with Osteogenesis Imperfecta (OI) or non-OI osteoporosis treated for a median of 39 months (range 6.5-164). Subjects in both diagnostic groups showed significant gains in spine areal Bone Mineral Density (aBMD) during the first year of therapy (29% median gain in children with OI and 15% in children with non-OI osteoporosis). Fracture frequency also declined significantly in both patient groups during the first year of treatment, including for two patients who had <10% improvement in spine aBMD over this time frame. The correlation between % change in aBMD and % change in fracture rate for our study population was weak, as demonstrated by a Spearman’s rank correlation coefficient (rho) of 0.13 (p-value 0.32, 95% confidence interval -0.32 to 1.00). Minor side effects of bisphosphonate therapy were self-limited, and no osteopetrosis, jaw osteonecrosis, or atypical femur fractures occurred during treatment for up to 13.6 years. These data suggest that low dose pamidronate is safe and effective for long-term use in pediatric osteoporosis, and that change in aBMD is an imperfect predictor of reduction in fracture risk.

Keywords: Bisphosphonate; Bone mineral density; Fracture; Pediatric osteoporosis; Osteogenesis imperfecta; Pamidronate


Bone fragility and osteoporosis (OP) are common complications of several genetic and acquired disorders of childhood. Pediatric patients with Osteogenesis Imperfecta (OI), inflammatory bowel disease, rheumatologic disorders, cerebral palsy, muscular dystrophy, cystic fibrosis, or a history of transplantation may develop low bone mass and fragility fractures [1]. Treatment of pediatric osteoporosis begins with optimizing nutrition, vitamin D stores, endocrine function, and weight-bearing physical activity [2]. When these measures are insufficient to prevent bone loss and fracture, use of pharmacologic therapies is considered [3].

Pharmacologic options for treating OP in adults include bisphosphonates to reduce bone resorption and anabolic agents to stimulate bone formation. The safety and efficacy of these medications in older patients have been established in large randomized controlled trials (RCTs), but data are limited in pediatrics [3-5]. The best studied anabolic agent, synthetic parathyroid hormone, should not be used in children due to a black box warning about the risk of osteosarcoma [6]. The anti-resorptive bisphosphonates have been used to treat primary and secondary osteoporosis in children, but the optimal agent, dose and duration of therapy remain controversial due to a lack of RCTs comparing different drugs and dosing regimens [3,7]. The response to bisphosphonates in children has been studied most extensively in patients with OI with less information on the response in children with secondary osteoporosis from chronic disease [3,4,7,8]. The most common outcome measure in these studies has been the change in areal bone mineral density (aBMD) by dual energy x-ray absorptiometry (DXA) rather than clinically relevant endpoints such as improvements in bone pain, mobility or fragility fractures [3].

Pamidronate is the bisphosphonate that has been used most extensively in pediatric patients, administered intravenously either in a high-dose protocol of 9mg/kg/year divided every 4 months [9] or low-dose at 4mg/kg/year divided every 2-3 months [10- 12]. Data comparing the relative efficacy and safety of the two regimens are limited. One retrospective study of 15 patients with non-OI osteoporosis treated for a year with low-dose or high-dose pamidronate found no dose-related differences in the changes observed for spine areal bone mineral density (aBMD) and fracture frequency [13].

Whether adverse effects vary by dose, particularly with long-term use, has also not been determined. In older adults, a “drug holiday” is often recommended after five years of bisphosphonate therapy due to concerns for over-suppression of bone turnover, osteonecrosis of the jaw, and atypical femur fractures [14]. In children, the maximal benefit from bisphosphonate therapy is achieved after 2-3 years [15], but there are risks to discontinuing drug therapy in growing patients. Fractures may occur at the junction of older “treated” bone and the more distal “untreated” bone added during growth [16]. Therefore, children with OI or ongoing risk factors for secondary osteoporosis are often maintained on a lower dose of bisphosphonates after their initial treatment period until final height is reached [3,16].

Prescribing the lowest effective dose is a priority, given concerns for over-suppression of bone turnover during many years of bisphosphonate therapy. One study of patients with OI found bone turnover markers to be suppressed below the expected range as long as two years after high dose pamidronate was discontinued [17]. Reassuringly, the same investigators found no clinical signs of oversuppression in patients treated for 10 years or more with pamidronate or zoledronic acid. No studies to date have reported long-term followup data for patients treated with low dose pamidronate.

We have previously shown that 4mg/kg/year pamidronate given for up to 30 months in 11 osteoporotic children resulted in reduced fracture rates and improved spinal aBMD [12]. This low dose was derived by extrapolation from pamidronate protocols for adults with secondary OP due to glucocorticoid therapy, transplantation, and other chronic diseases. This report summarizes the changes in spine aBMD and fracture rates and the adverse effects in 31 children treated with low dose pamidronate for up to 13.6 years for bone fragility related to OI or non-OI disorders.


Pamidronate therapy was offered on a compassionate use basis to pediatric patients with a history of low-impact long bone or vertebral fracture. Two patients without a prior fracture history were also treated due to a perceived high risk for fracture. One was a patient with Duchenne Muscular Dystrophy, chronic high dose glucocorticoid therapy, evidence of decreasing bone mineral density, and anticipated loss of ambulation. The other was a patient on chronic high-dose glucocorticoids for management of Crohn’s disease who had declining bone density; pamidronate was provided for the duration of glucocorticoid therapy. Both of these patients had baseline spine aBMD z-scores less than -2 prior to treatment. All subjects were bisphosphonate-naïve except one OI patient who had been treated at another center; this patient’s continued bone fragility at the time of entry in our study qualified her for ongoing therapy.

All patients underwent screening tests prior to the initiation of treatment, which included a complete blood count (CBC), serum calcium, phosphorus, magnesium, alkaline phosphatase, albumin, creatinine, intact parathyroid hormone, 25-hydroxy and 1,25-dihydroxy vitamin D levels, and urine calcium, phosphorus, and creatinine. The purpose of these tests was to identify any underlying bone disorder or vitamin D deficiency that would need to be addressed prior to bisphosphonate administration. Pamidronate was not administered to any child with a 25-hydroxy vitamin D concentration <50 nmol/L (20ng/ml).

Pretreatment densitometry of the lumbar spine, whole body and total hip by dual energy x-ray absorptiometry (DXA) was performed in children overage three if the examinations could be conducted without sedation. Sites with metal implants in the region of interest were excluded from analysis. Baseline lateral thoraco-lumbar spine x-rays were performed in younger children in whom a DXA scan was not feasible and in any patient suspected of having a vertebral fracture. Prior fracture history was recorded based on patient and parent reports. Clinically significant fractures were defined as any long bone or vertebral compression fracture, excluding fractures of the fingers, toes, hands, feet, ribs, and clavicles.

For children three years or older, pamidronate disodium (mixed with 0.9% normal saline) was administered intravenously every three months at a dose of 1mg/kg, up to a maximum of 30mg per dose. Children under age three received a dose of 0.75mg/kg every eight weeks due to more rapid bone turnover. Each dose was infused over four hours in an inpatient or day hospital setting at Lucile Packard Children’s Hospital. After the initial treatment period of three years, maintenance therapy was continue data dose of 1mg/kg (maximum 30mg) every six months until growth plates closed or the underlying risk factors resolved.

To monitor for acute adverse effects following the first infusion, serum calcium, magnesium, phosphorus and creatinine were measured within 3-7 days post treatment. Serum calcium, phosphorus, magnesium, creatinine and CBC were also measured prior to each subsequent infusion; 25-hydroxyvitamin D was measured annually. All follow-up infusion visits included a systematic review of interval events including fractures, symptoms or side effects and any laboratory or imaging studies. Bone densitometry by DXA was repeated at 6, 12, 24, and 36 months where feasible without sedation; spine x-rays were performed annually if a DXA could not be obtained or there was suspicion of new vertebral fracture.

Statistical analyses

Study data were managed using Research Electronic Data Capture (REDCap) [18], a secure web-based research application hosted at the Stanford Center for Clinical Informatics. Data were analyzed for OI and non-OI cohorts separately. Fracture rates were assessed as total fractures divided by total years at risk for each patient during the time frame in question. Years at risk prior to study entry was defined as 1) years lived for OI patients, 2) years since onset of highdose steroids for patients with glucocorticoid induced osteoporosis (GIO), 3) date of diagnosis for a patient with osteosarcoma, and 4) date of first fracture for other diagnoses including idiopathic juvenile osteoporosis (IJO), congenital hypomyelination syndrome, and cerebral palsy, where onset of risk was difficult to determine. Change in spine aBMD was calculated for each patient as a percent change in numeric score from baseline value, and these calculations were performed only when serial data were obtained using the same DXA equipment.

Fracture and densitometry data are reported as median and range (Table 2) or median, quartiles, and range (Figure 1) because the data were not normally distributed. P-values were calculated using Fisher’s Exact Test for categorical variables and Wilcoxon Rank Sum for quantitative variables. Spearman’s rank correlation coefficient (rho) was calculated to examine the correlation between % change in spine aBMD and % change in fracture rate during the first year of treatment in the 16 patients with complete BMD data.