Microindentation for in vivo measurement of bone tissue mechanical properties in humans

Abstract

Bone tissue mechanical properties are deemed a key component of bone strength, but their assessment requires invasive procedures. Here we validate a new instrument, a reference point indentation (RPI) instrument, for measuring these tissue properties in vivo. The RPI instrument performs bone microindentation testing (BMT) by inserting a probe assembly through the skin covering the tibia and, after displacing periosteum, applying 20 indentation cycles at 2 Hz each with a maximum force of 11 N. We assessed 27 women with osteoporosis-related fractures and 8 controls of comparable ages. Measured total indentation distance (46.0 +/- 14 versus 31.7 +/- 3.3 microm, p = .008) and indentation distance increase (18.1 +/- 5.6 versus 12.3 +/- 2.9 microm, p = .008) were significantly greater in fracture patients than in controls. Areas under the receiver operating characteristic (ROC) curve for the two measurements were 93.1% (95% confidence interval [CI] 83.1-100) and 90.3% (95% CI 73.2-100), respectively. Interobserver coefficient of variation ranged from 8.7% to 15.5%, and the procedure was well tolerated. In a separate study of cadaveric human bone samples (n = 5), crack growth toughness and indentation distance increase correlated (r = -0.9036, p = .018), and scanning electron microscope images of cracks induced by indentation and by experimental fractures were similar. We conclude that BMT, by inducing microscopic fractures, directly measures bone mechanical properties at the tissue level. The technique is feasible for use in clinics with good reproducibility. It discriminates precisely between patients with and without fragility fracture and may provide clinicians and researchers with a direct in vivo measurement of bone tissue resistance to fracture.

Fig. 1

Indentation procedure for measuring material properties of bone in vivo and SEM imaging of an indent on a human bone sample. (A) Illustration of the method for obtaining indentation measurements, including insertion of the test probe assembly, displacing the periosteum with the reference probe, first-cycle indentation, and last-cycle indentation, which determines the IDI with respect to the first cycle. (B) SEM image of an indentation (encircled by dashed line) being compared to a dime (the smallest U.S. coin). (C) This magnified SEM image of the indentation shows microcracks created during the repetitive loading cycles at a constant force.

Fig. 2

Parameters are calculated from force-versus-distance data obtained by the RPI instrument. The parameters include indentation distance increase (IDI), total indentation distance (total ID), and creep indentation distance (creep ID) measured in the first cycle. (A) The IDI is defined as the increase in the indentation distance in the last cycle relative to the indentation distance in the first cycle (see Fig. 1A). The creep ID is determined by the increase in distance while the force is held constant at the maximum value for a duration of one-third of the first indentation cycle. The total ID is defined as the total distance the test probe is inserted into the bone from touchdown to the end of the twentieth cycle. (B–D) Results from clinical trials of each parameter with fracture (n = 27) and control (n = 8) patients. Note that fracture patients usually had higher indentation distances. The subscript H on the graphs indicates that the parameters were measured with the Hospital del Mar protocol. This is important because the values of these parameters depend on the measurement protocol.

Fig. 3

Data results including statistics and a receiver operating characteristic (ROC) curve. (A) Age-adjusted statistical results for IDI (µm), creep ID (µm), total ID (µm), femoral neck bone mineral density (FN BMD, g/cm²), and total-hip bone mineral density (TH BMD, g/cm²). (B) The ROC curve displays the clinical results from Hospital Del Mar, Barcelona. The area under the curve (AUC) is a scalar quantity to gauge the performance of the curve. An AUC of 100% would represent a perfect model; however, an area going along the line of discrimination (dashed diagonal) would be a completely random model.

Fig. 4

SEM images of cadaveric human bone samples that were fractured and exhibit crack bridging, which resists crack extension. The crack growth toughness of samples was compared with the indentation distance increase (IDI). (A–C) The samples in panels A and C were fractured in fluid,(27) and microcracks were observed, whereas the sample in panel B displays a microcrack created by the RPI instrument during repetitive indentations. It resembles the microcracks in both A and C. (D) Comparison between IDI and crack growth toughness(22) (slope of R curve) obtained for samples from five donors. High IDI and low crack growth toughness are associated with bones that are prone to fracture. The graph shows this trend by relating high IDI to low crack growth toughness and vice versa. The linear fit has a Pearson correlation of −0.904, with p = .018 (one-tailed) and p = .035 (two-tailed). We believe that the one-tailed test is justified because we anticipated the direction of the trend: High IDI corresponds to low crack growth toughness. Because of the limited number of samples and subjects, this correlation should be regarded as preliminary until a more complete investigation is done.