Radiofrequency echographic multi-spectrometry and DXA for the evaluation of bone mineral density in a peritoneal dialysis setting

Background: The aim of this real-life cross-sectional explorative study was to compare radiofrequency echographic multi-spectrometry (REMS) with dual-energy X-rays absorptiometry (DXA) in the BMD assessment of patients receiving peritoneal dialysis (PD). Furthermore, we investigated the relationship between lumbar aortic calcifications (AOCs) and the DXA lumbar measurements.

Methods: Consecutive patients referring to the PD clinic of our hospital were included. Lumbar spine and femur scans were acquired with both techniques (including lumbar laterolateral DXA scans). The risk assessment of two fracture risk algorithms (FRAX^® and DeFRA^®) were compared. Cohen's k coefficients were used to assess the inter-technique agreement in the classification of patients as osteoporotic. Lumbar AOCs were estimated semi-quantitatively on laterolateral DXA scans.

Results: 41 patients were enrolled. No significant differences were documented between the BMD T-scores measured through DXA or REMS at the femur. At the lumbar spine, the DXA anteroposterior mean T-score (- 0.49 ± 1.98) was significantly higher than both the laterolateral DXA (- 1.66 ± 0.99) and the REMS (- 2.00 ± 1.94) measurements (p < 0.01 vs both). No significant differences were found between the DXA and REMS fracture risk estimates with both algorithms. The inter-technique Cohen's k coefficient (for the worst T-score, any site) was 0.421, p < 0.001. The discrepancy between the DXA laterolateral and anteroposterior lumbar T-score was positively associated with the AOCs extent and severity (r = 0.402, p < 0.01).

Conclusions: Our data showed a promising agreement, in a real-life PD setting, between DXA and REMS BMD assessment and in the consequent fracture risk estimation and confirm the AOCs interference on the diagnostic accuracy of lumbar DXA.