Abstract
BACKGROUND: Therapy with tenofovir is associated with lower bone mineral density (BMD), higher markers of bone turnover and increased fracture risk in HIV-infected adults. Bone structural parameters generated by hip structural analysis may represent a separate measure of bone strength, but have not been assessed in HIV. METHODS: Dual-energy X-ray absorptiometry (DXA) scans from 254 HIV-infected adults randomised to simplify their existing dual nucleoside analogue reverse transcriptase inhibitor therapy to coformulated tenofovir-emtricitabine or abacavir-lamivudine were analysed using DXA-derived hip structural analysis software. Hip structural parameters included femoral strength index, section modulus, cross-sectional area, and cross-sectional moment of inertia. We used one-way ANOVA to test the relationship between nucleoside analogue type at baseline and structural parameters, multivariable analysis to assess baseline covariates associated with femoral strength index, and t-tests to compare mean change in structural parameters over 96 weeks between randomised groups. RESULTS: Participants taking tenofovir at baseline had lower section modulus (-107.3 mm2, p = 0.001), lower cross-sectional area (-15.01 mm3, p = 0.001), and lower cross-sectional moment of inertia (-2,036.8 mm4, p = 0.007) than those receiving other nucleoside analogues. After adjustment for baseline risk factors, the association remained significant for section modulus (p = 0.008) and cross-sectional area (p = 0.002). Baseline covariates significantly associated with higher femoral strength index were higher spine T-score (p = 0.001), lower body fat mass (p<0.001), lower bone alkaline phosphatase (p = 0.025), and higher osteoprotegerin (p = 0.024). Hip structural parameters did not change significantly over 96 weeks and none was significantly affected by treatment simplification to tenofovir-emtricitabine or abacavir-lamivudine. CONCLUSION: In this population, tenofovir use was associated with reduced composite indices of bone strength as measured by hip structural analysis, but none of the structural parameters improved significantly over 96 weeks with tenofovir cessation. TRIAL REGISTRATION: ClinicalTrials.gov NCT00192634.