R-2 ' measured in trabecular bone in vitro: Relationship to trabecular separation

Measurement of key parameters of the microstructure of trabecular bone is c ritical to the study of osteoporosis and bone strength. Density based metho ds cannot provide this information, and give only the total amount of bone present, and not its arrangement. Magnetic resonance imaging has shown the potential to provide information related to the microarchitecture of the tr abecular bone matrix. Twelve samples (8 x 8 x 8 mm(3) bone cubes) were cut from sheep vertebrae such that the trabeculae ran either parallel or perpen dicular to each face. Detailed measurements of the structure of these bone cubes were made by histomorphometry, and compared to R-2' and R-2* measured with a spin and gradient-echo sequence, Partially Refocused Interleaved Mu ltiple Echo, at 1.5 Tesla. The precision of the R-2' measurement (% coeffic ient of variation) was 8.7 +/- 5.1, and 7.7 +/- 4.3 for R-2* Uncorrected va lues of R-2' and R-2* were significantly correlated to density measured by quantitative computed tomography (r = 0.87, p = 0.0005, and r = 0.90, p = 0 .0002, respectively), and trabecular bone area measured by histomorphometry (r = 0.80, p = 0.002, and r = 0.83, p = 0.0008, respectively). Density cor rection was effected by imaging the same slice of bone in two orientations (90 degrees and 0 degrees) to the main magnetic field. For both R-2' and R- 2* there was a significant difference between measurements in the 90 degree s and 0 degrees orientations (p < 0.01). The difference between the two val ues was used, and termed R-2net' or R-2net*. The net parameters were indepe ndent of bone mass. R-2net' and R-2net* were significantly correlated to tr abecular separation (p < 0.05) with r = -0.58 and -0.62, respectively. Thes e results demonstrate the ability of magnetic resonance imaging to characte rize a key measure of the trabecular microstucture. An increase in trabecul ar separation has important biomechanical consequences in osteoporosis. Thi s result also strengthens the hypothesis that the sensitivity of R-2' to os teoporosis-related bone changes is due to magnetic susceptibility effects i n which rapid transitions between bone and marrow create local magnetic fie ld inhomogeneities that result in an increase in R-2' values. (C) 1999 Else vier Science Inc.