Determinants and Trends of Bone Mineral Density in Transfusion-Dependent Thalassaemia: A Retrospective Single-Centre Cohort Analysis  — ASN Events

Determinants and Trends of Bone Mineral Density in Transfusion-Dependent Thalassaemia: A Retrospective Single-Centre Cohort Analysis  (#102)

Evan EZ Zhang 1 2 , Alma AC Corker 1 2 , Leanne LC Crnek 1 , Annmarie AB Bosco 1 2 , Amanda AB Beech 2 3 , Michael MB Bennett 2 4 , Giselle GKG Kidson-Gerber 1 2
  1. Department of Haematology, Prince of Wales Hospital, Randwick, NSW, Australia
  2. University of New South Wales, Sydney, NSW, Australia
  3. Obstetric Medicine and Endocrinology, Royal Hospital for Women, Randwick, NSW, Australia
  4. Department of Endocrinology, Prince of Wales Hospital, Randwick, NSW, Australia

Background/aim 

Bone disease is a major complication of transfusion-dependent thalassaemia (TDT), but longitudinal site-specific bone density trajectories and their clinical and biochemical determinants remain incompletely characterised. We aimed to define these trajectories and identify key correlates of skeletal deterioration. 

Methods 

Retrospective single-centre cohort study of 74 patients with TDT. 71 had analysable DXA data, contributing 424 scans (median 5 per patient; IQR 3.0–8.5). Linear mixed models assessed change in lumbar spine, femoral neck, and total-hip BMD and Z-scores over time. Prespecified covariates included age, sex, BMI, hypogonadism, diabetes, hypothyroidism, hypoparathyroidism, splenectomy, bisphosphonate exposure, and fragility fracture. Biochemical variables were linked to each scan using prior 12-month median or nearest available values. 

Results 

The cohort was predominantly beta thalassaemia (93.2%). Median follow-up was 15.1 years (IQR 7.3–21.4); median scans per patient were 5 (IQR 3.0–8.5). Comorbidity burden was substantial: splenectomy 56.8%, hypogonadism 29.7%, diabetes 21.6%, hypothyroidism 20.3%, hypoparathyroidism 10.8%, and fragility fracture 16.2%. At latest DXA, mean age was 39.0 ± 11.4 years; mean lumbar, femoral-neck, and total-hip BMD were 0.976, 0.845, and 0.824 g/cm², respectively. At latest DXA (mean age 39.0 ± 11.4 years), Z-score ≤ −2 was present in 36.6% at lumbar spine, 31.0% at total hip, and 12.7% at femoral neck. In mixed models, femoral-neck BMD declined significantly (−0.0053 g/cm²/year, p<0.001) while lumbar spine BMD was stable. Lumbar Z-score improved modestly (+0.016/year, p<0.001), whereas femoral-neck and total-hip Z-scores declined (both approximately −0.020/year, p<0.001). Hypogonadism was the most consistent adverse clinical determinant, independently associated with lower femoral-neck BMD, total-hip BMD, and total-hip Z-score. ALP was the most reproducible biochemical correlate, with higher values associated with lower femoral-neck BMD and lower Z-scores at all three sites after FDR correction. Higher PTH was associated cross-sectionally with lower BMD at all sites. Ferritin showed no independent association with BMD in adjusted models. 

Conclusions  

In TDT, longitudinal skeletal deterioration is most evident at the hip, particularly the femoral neck. Hypogonadism, elevated ALP, and PTH dysregulation may be more informative correlates of adverse bone health than ferritin, supporting targeted endocrine surveillance and hip-focused DXA monitoring.