Purpose or Case Report: Increased fracture risk is a complication that occurs in the context of primary bone diseases such as osteogenesis imperfecta (OI). Despite being considered as the reference-standard, the use of dual-energy X-ray absorptiometry (DXA) to evaluate fragility fractures in OI has not been validated by prior systemic review. Identifying patients at greatest risk for bone fragility fractures and determining skeletal health markers that can monitor bone mass concerning response to bone-active treatments are important issues for clinicians. In this systematic review we assessed the clinical utility of DXA for evaluating osteoporotic bone in OI pediatric patients according to the U.S. Preventive Services Task Force guidelines.
Methods & Materials: We retrieved articles that assessed abnormal bone quality in patients of mean age ≤18 years (MEDLINE and EMBASE, 1946-2015). Evidence was accrued for concurrent and predictive validity, and responsiveness of DXA according to the questions: (1) Is DXA a good predictor and/or indicator of bone fragility fracture risk in patients with OI? (2) Is DXA responsive to an intervention (bone-active treatment, e.g. bisphosphonate, calcitriol, exercise) in OI patients? Two reviewers independently evaluated articles’ quality of reporting and methodological quality using Standards for Reporting of Diagnostic Accuracy and Quality Assessment of Diagnostic Accuracy Studies tools.
Results: Out of 162 retrieved references we included 18 studies (778 patients; 357 reported boys and 315 girls; age range, 0–18 years). The mean quality of reporting score was “moderate” (77.8% of articles); the mean methodological quality score was “high” (81.4%). For the proposed questions, conflicting evidence exists in the literature to support DXA’s ability to diagnose fracture in patients with OI (Grade C) and insufficient evidence is available to make a recommendation regarding the value of DXA for prediction of future fractures (Grade I). Nevertheless, fair recommendation exists for the use of lumbar spine bone mineral density to assess the effect of bisphosphonates on bone health (Grade B). Gaps in the literature included lack of information concerning effects of different statistical adjustment applications during measurement differences between OI subtypes.
Conclusions: DXA-based measurements can respond to bisphosphonate interventions (fair evidence), however insufficient evidence exists to support other clinimetric properties of this technique.