Bone Mass and Microarchitecture in CKD Patients with Fracture

Patients with predialysis chronic kidney disease (CKD) have increased risk for fracture, but the structural mechanisms underlying this increased skeletal fragility are unknown. We measured areal bone mineral density (aBMD) by dual-energy x-ray absorptiometry at the spine, hip, and radius, and we measured volumetric BMD (vBMD), geometry, and microarchitecture by high-resolution peripheral quantitative computed tomography (HR-pQCT) at the radius and tibia in patients with CKD: 32 with fracture and 59 without fracture. Patients with fracture had lower aBMD at the spine, total hip, femoral neck, and the ultradistal radius, the last having the strongest association with fracture. By HR-pQCT of the radius, patients with fracture had lower cortical area and thickness, total and trabecular vBMD, and trabecular number and greater trabecular separation and network heterogeneity. At the tibia, patients with fracture had significantly lower cortical area, thickness, and total and cortical density. Total vBMD at both radius and tibia most strongly associated with fracture. By receiver operator characteristic curve analysis, patients with longer duration of CKD had area under the curve of >0.75 for aBMD at both hip sites and the ultradistal radius, vBMD and geometry at the radius and tibia, and microarchitecture at the tibia. In summary, patients with predialysis CKD and fractures have lower aBMD by dual-energy x-ray absorptiometry and lower vBMD, thinner cortices, and trabecular loss by HR-pQCT. These density and structural differences may underlie the increased susceptibility to fracture among patients with CKD.Fracture rates in patients with ESRD are elevated,¹ as high as individuals who have normal kidney function and are older by 10 to 20 years.² Recently, there has been increasing recognition that patients with predialysis chronic kidney disease (CKD) also experience an increased fracture burden.^2–5 In 2006, we reported that participants who were older than 50 years in the Third National Health and Nutrition Examination Survey (NHANES III) and had an estimated GFR (eGFR) between 15 and 59 ml/min (stages 3 and 4 CKD) had a two-fold higher risk for hip fracture than individuals without CKD.⁶ Subsequent studies confirmed our findings and also demonstrated that fracture risk increases as kidney function declines.^3–5 In one study, hip fracture risk was as high in patients with stage 4 CKD as in patients with ESRD.⁴ Given the rapid expansion of the population of individuals who are older than 65 years worldwide and the high prevalence of CKD in the elderly,⁷ it is highly important to improve our understanding of the structural and biologic mechanisms that contribute to increased fracture rates in patients with CKD so that we can develop strategies to identify those who are at risk for fracture.In patients with ESRD, relationships between areal bone mineral density (aBMD) measured by dual-energy x-ray absorptiometry (DXA) and prevalent fractures are inconsistent; some studies detected no difference in aBMD,^8–11 whereas others, including a meta-analysis, found lower aBMD in those with prevalent spine and nonspine fractures^8,11,12; therefore, recent Kidney Disease: Improving Global Outcomes (KDIGO) guidelines for the diagnosis and management of CKD-mineral and bone disorder (CKD-MBD) do not recommend routine measurement of aBMD by DXA in patients with ESRD or in those with late stages 3 through 5 CKD, because more severe CKD is commonly associated with renal osteodystrophy.¹³ In contrast, however, KDIGO guidelines recommend DXA to assess fracture risk in patients with stage 1 through early stage 3 CKD, as long as biochemical testing does not suggest CKD-MBD.Fractures occur when bone strength is not sufficient to withstand an applied force. Bone strength is determined both by the density of bone present and by the quality of that bone. Whereas bone density is measured by DXA, other factors that contribute to bone quality, such as cortical and trabecular microarchitecture, are not. Chronic parathyroid hormone excess, a common biochemical feature of CKD, is generally associated with trabecularization of endocortical bone, cortical thinning, and increased cortical porosity.^14–16 Whereas effects of hyperparathyroidism on trabecular bone are less consistent, trabecular structure and connectivity are generally maintained.¹⁴ The resolution of DXA is too low to distinguish between cortical and trabecular bone, both of which influence resistance to fracture; this may account for its inconsistent utility in discriminating fracture status in patients with ESRD.Bone biopsy studies have demonstrated that histomorphometric abnormalities of renal osteodystrophy may begin early in the development of CKD¹⁷; therefore, DXA may also have limited utility in patients with predialysis kidney disease. Recently, Bacchetta et al.¹⁸ used high-resolution peripheral quantitative computed tomography (HR-pQCT; voxel size 82 μm) to demonstrate that both cortical and trabecular microarchitecture are abnormal in patients with early CKD compared with healthy control subjects. In addition, they reported thinner cortices and abnormal trabecular microarchitecture in a small number of patients with CKD and with fractures; however, they did not measure aBMD by DXA. In this study, we measured aBMD by DXA and volumetric BMD (vBMD) and bone microarchitecture by HR-pQCT in patients with predialysis CKD (stages 2 through 5), with and without prevalent fracture. On the basis of results in postmenopausal women with normal kidney function reported by Boutroy et al.¹⁹ and by our group,²⁰ we hypothesized that HR-pQCT would detect abnormalities of vBMD and bone geometry and microstructure in patients with CKD and a history of fracture, whereas measurement of aBMD by DXA would not.