Comparison of Three Methods to Quantify Repair Cartilage Collagen Orientation

Abstract

Objective: The aim of this study was to determine if the noninvasive or minimally invasive and nondestructive imaging techniques of quantitative T2-mapping or multiphoton microscopy (MPM) respectively, could detect differences in cartilage collagen orientation similar to polarized light microscopy (PLM). It was hypothesized that MRI, MPM, and PLM would all detect quantitative differences between repair and normal cartilage tissue.

Methods: Osteochondral defects in the medial femoral condyle were created and repaired in 5 mature goats. Postmortem, MRI with T2-mapping and histology were performed. T2 maps were generated and a mean T2 value was calculated for each region of interest. Histologic slides were assessed using MPM with measurements of autocorrelation ellipticity, and by PLM with application of a validated scoring method. Collagen orientation using each of the 3 modalities (T2-mapping, MPM, and PLM) was measured in the center of the repair tissue and compared to remote, normal cartilage.

Results: MRI, MPM, and PLM were able to detect a significant difference between repair and normal cartilage (n = 5). The average T2 value was longer for repair tissue (41.43 ± 9.81 ms) compared with normal cartilage (27.12 ± 14.22 ms; P = 0.04); MPM autocorrelation ellipticity was higher in fibrous tissue (3.75 ± 1.17) compared with normal cartilage (2.24 ± 0.51; P = 0.01); the average PLM score for repair tissue was lower (1.6 ± 1.02) than the score for remote normal cartilage (4.4 ± 0.42; P = 0.002). The strongest correlation among the methods was between MRI and PLM (r = -0.76; P = 0.01), followed by MPM and PLM (r = -0.58; P = 0.08), with the weakest correlation shown between MRI and MPM (r = 0.35; P = 0.31).

Conclusion: All 3 imaging methods quantitatively measured differences in collagen orientation between repair and normal cartilage, but at very different levels of resolution. PLM is destructive to tissue and requires euthanasia, but because MPM can be used arthroscopically, both T2-mapping and MPM can be performed in vivo, offering nondestructive means to assess collagen orientation that could be used to obtain longitudinal data in cartilage repair studies.

Keywords: articular cartilage; cartilage repair; outcome measures; tissue.

Figure 1.

Sagittal quantitative T2-mapping of the stifle joint performed on a clinically relevant field strength (3 T) MRI unit. Regions of interest were assessed centrally within the defect (single arrow) and remote to the defect (double arrow).

Figure 2.

Axial plane histologic sections (hematoxylin and eosin) of cartilage repair region from 3 animals (A, B, C). In the repair region, the tissue is histologically characterized as fibrocartilagenous (single arrows) as opposed to surrounding normal articular cartilage (double arrows). All images are the same magnification and size; differences in cartilage thickness are due to variability between animals. Bar = 1 mm.

Figure 3.

Sagittal magnetic resonance images from the same 3 animals (A, B, C) as depicted in Figure 2, with T2-mapping to include X as location of 2 mm² region of interest (yellow) over the region of repair and Y as location of 2 mm² region of interest (yellow) over region of cartilage remote to repair. Green circles represent areas of cartilage most affected by magic angle (approximately 55° to Bo). In all animals, a significant difference was detected in T2 relaxation time between repair and normal cartilage collagen orientation. MRI T2 repair values: A = 37.6, B = 27.8, C = 49.7. MRI T2 normal values: A = 33.6, B = 23.8, C = 25.3.

Figure 4.

Axial plane multiphoton microscopy images obtained from the same 3 animals (A, B, C) and regions of repair and normal cartilage as presented in Figures 2 and 3. In all panels, superficial is toward the top of the image. A significant difference was observed between repair and normal cartilage collagen structure. Note that the resolution of multiphoton microscopy is at the cellular level. Autocorrelation ellipticity repair values: A = 4.51, B = 1.92, C = 3.01. Autocorrelation ellipticity normal values: A = 2.46, B = 2.94, C = 2.74.

Figure 5.

Axial plane polarized light microscopy (PLM) images of the same 3 animals (A, B, C) and regions of repair and normal cartilage as presented in Figures 2-4. In all panels, superficial is toward the top of the image. PLM was able to detect a significant difference between repair and normal cartilage collagen structure. PLM repair scores: A= 0.5, B = 2.5, C = 2.5. PLM normal scores: A = 4, B = 4.5, C = 5.