Assessing cortical bone porosity with MRI in an animal model of chronic kidney disease

Abstract

Chronic kidney disease (CKD) is characterized by secondary hyperparathyroidism and an increased risk of hip fractures predominantly related to cortical porosity. Unfortunately, bone mineral density measurements and high-resolution peripheral computed tomography (HR-pQCT) imaging have shortcomings that limit their utility in these patients. Ultrashort echo time magnetic resonance imaging (UTE-MRI) has the potential to overcome these limitations by providing an alternative assessment of cortical porosity. The goal of the current study was to determine if UTE-MRI could detect changes in porosity in an established rat model of CKD. Cy/+ rats (n = 11), an established animal model of CKD-MBD, and their normal littermates (n = 12) were imaged using microcomputed tomography (microCT) and UTE-MRI at 30 and 35 weeks of age (which approximates late-stage kidney disease in humans). Images were obtained at the distal tibia and the proximal femur. Cortical porosity was assessed using the percent porosity (Pore%) calculated from microCT imaging and the porosity index (PI) calculated from UTE-MRI. Correlations between Pore% and PI were also calculated. Cy/+ rats had higher Pore% than normal rats at both skeletal sites at 35 weeks (tibia = 7.13 % +/- 5.59 % vs. 0.51 % +/- 0.09 %, femur = 19.99 % +/- 7.72 % vs. 2.72 % +/- 0.32 %). They also had greater PI at the distal tibia at 30 weeks of age (0.47 +/- 0.06 vs. 0.40 +/- 0.08). However, Pore% and PI were only correlated in the proximal femur at 35 weeks of age (ρ = 0.929, Spearman). These microCT results are consistent with prior studies in this animal model utilizing microCT imaging. The UTE-MRI results were inconsistent, resulting in variable correlations with microCT imaging, which may be related to suboptimal bound and pore water discrimination at higher magnetic field strengths. Nevertheless, UTE-MRI may still provide an additional clinical tool to assess fracture risk without using ionizing radiation in CKD patients.

Keywords: Chronic kidney disease; MRI; Porosity index; UTE; microCT.

Fig. 1.

Experimental Timeline. Animals underwent in vivo microCT imaging of the distal tibiae at 30 weeks. The distal tibiae and proximal femora were imaged at both 30 and 35 weeks of age with in vivo UTE-MRI. After euthanasia at 35 weeks, the distal tibiae and proximal femora were imaged using ex vivo microCT imaging prior to mechanical testing of the femoral neck.

Fig. 2.

UTE-MRI Analysis. Representative images from Cy/+ rats and their normal littermates are depicted. VOIs were obtained from both echoes at the distal tibia and the proximal femur. A = anterior, P = posterior, M = medial, L = lateral.

Fig. 3.

microCT Analysis. Representative images from Cy/+ rats and their normal littermates are depicted. VOIs were obtained at the distal tibia and the proximal femur. A = anterior, P = posterior, M = medial, L = lateral.

Fig. 4.

microCT Results. (a) Cy/+ rats did not have a greater distal tibia Pore% than the normal rats at 30 weeks (in vivo) (p = 0.009, ANOVA; p = 0.751, Tukey’s test), but they did have a greater Pore% at 35 weeks (ex vivo) (*p = 0.009, ANOVA; p < 0.001, Tukey’s test). Pore% in Cy/+ rats increased over the course of the study (⁺p = 0.009, ANOVA; p = 0.013, Tukey’s test). (b) Cy/+ rats also had a greater ex vivo proximal femur Pore% than the normal rats at 35 weeks (*p < 0.001, Welch’s t-test).

Fig. 5.

UTE-MRI Results. (a) Cy/+ rats had a greater distal tibia PI than the normal rats at 30 weeks (*p = 0.009, ANOVA; p = 0.019, Tukey’s test) but not at 35 weeks (p = 0.467, Tukey’s test). The distal tibia PI in Cy/+ rats did not increase over the course of the study (p = 0.575, Tukey’s test). (b) Cy/+ rats did not have a greater proximal femur PI than the normal rats at 30 weeks or at 35 weeks, and the proximal femur PI in Cy/+ rats did not increase over the course of the study (p = 0.083, ANOVA).

Fig. 6.

Femoral Neck Mechanical Properties. Cy/+ rats had a lower femoral neck ultimate load than normal rats at 35 weeks (*p = 0.001, Welch’s t-test).

Fig. 7.

Mechanics and Porosity Correlations. (a) There was no correlation between the distal tibia PI and femoral neck ultimate load in CKD animals at 35 weeks (p = 0.505) (b) or between the proximal femur PI and femoral neck ultimate load (p = 0.505). (c) However, there was a correlation between the distal tibia Pore% and femoral neck ultimate load (*p < 0.001) (d) and between the proximal femur Pore% and femoral neck ultimate load (*p = 0.020).

Fig. 8.

microCT and UTE-MRI Correlations. (a) There was no correlation between the distal tibia PI and Pore% in CKD animals at 30 weeks (p = 0.630) (b) or at 35 weeks (p = 0.337). (c) However, there was a correlation in the proximal femur at 35 weeks (*p = 0.003).