Ultrasound attenuation in normal and spontaneously degenerated articular cartilage

Abstract

High-frequency ultrasound (US) measurements may provide means for the quantification of articular cartilage quality. Bovine patellar cartilage samples (n = 32) at various degenerative stages were studied using US attenuation measurements in the 5- to 9-MHz frequency range. The results were compared with the histologic, biochemical and mechanical parameters obtained for the same samples, to identify which structural or functional factors could be related to the attenuation and its variations. Attenuation, as calculated in the frequency or time domain, correlated significantly with the histologic tissue integrity (i.e., Mankin score, Spearman r = −0.576 or −0.571, p < 0.01), but the slope of attenuation vs. frequency was not related to Mankin score. Ultrasound speed was, however, the most sensitive indicator of Mankin score (r = −0.755, p < 0.01). Cartilage quality index (CQI), a combination of structural and functional parameters, correlated significantly with the attenuation or speed (r = −0.655 or −0.872, p < 0.01). Our results suggest that US attenuation and speed may be suited for the diagnostics of cartilage degeneration. (E-mail: [email protected])

Introduction

Physically, ultrasound (US) attenuation depends on the absorption and scattering properties of the tissue and, thereby, is specific for the tissue composition and structure. For biologic tissues that are typically inhomogeneous and anisotropic, the attenuation mechanisms are more complicated compared with those in most nonbiologic materials. Attenuation in biologic tissues is highly dependent on the frequency, and increases linearly or nonlinearly as a function of frequency. The frequency-dependent attenuation has been reported to be 0.063f dB MHz⁻¹ mm⁻¹ (at 0.8 to 7.0 MHz), 0.10f dB MHz⁻¹ mm⁻¹ (at 0.3 to 4.5 MHz) and 4.1f dB MHz⁻¹ mm⁻¹ (at 1 MHz) for fat, kidney and lung tissues, respectively (f is the frequency of interest) (Wells 1969). High attenuation in lung tissue, compared to other soft tissues, is related to scattering effect induced by air-filled alveoli, whereas other soft tissues are acoustically more homogeneous, consisting of less acoustical discontinuities. US attenuation is also related to tissue temperature and varies between living and nonliving tissues (Wells 1969).

US attenuation in normal articular cartilage (AC), as measured in vitro at 100 MHz, has been reported to vary between 92 to 147 dB mm⁻¹ for superficial tissue of normal bovine AC (Agemura et al. 1990). At 10 to 40 MHz, integrated attenuation varied from joint to joint between 2.8 to 6.5 dB mm⁻¹ for normal bovine AC in vitro (Senzig et al. 1992). Importantly, it has been suggested that US attenuation in AC is related to tissue degeneration (Joiner et al., 2001, Nieminen et al., 2002) and, thereby, attenuation measurements could serve as a quantitative technique in diagnostics of osteoarthrosis (OA). Previously, however, attenuation has been studied only in normal or chemically degraded cartilage (Agemura et al., 1990, Joiner et al., 2001, Nieminen et al., 2002, Senzig et al., 1992). Earlier studies have reported slightly conflicting results on the role of proteoglycan (PG) content as a determinant of attenuation. Contrary to Joiner et al. (2001) and Nieminen et al., 2002, Agemura et al., 1990 suggested that attenuation decreased after chemical degradation of PGs. Inconsistent results may arise from the different frequency in use and different methods used for depleting PGs from the tissue. Enzymatic degradation of superficial cartilage collagen has been shown to have little effect on the attenuation (Nieminen et al. 2002), even though collagen controls significantly the sound reflection at the cartilage surface (Chérin et al., 1998, Töyräs et al., 1999). The changes in organization of collagen fibril network after PG depletion may, however, affect US attenuation (Joiner et al. 2001). Tissue changes after experimental enzymatic treatments are typically restricted to the superficial tissue and are rather specific (Rieppo et al. 2003) as compared with in vivo experimental models, such as surgical anterior cruciate ligament transection model (Setton et al. 1994) or intra-articular injection of papain (Murray 1984), iodoacetate or collagenase (van Osch et al. 1994). Spontaneously degenerated cartilage is typically softer, shows collagen network disruption, PG depletion and increase of water content as compared with intact tissue (Armstrong and Mow, 1982, Buckwalter and Mankin, 1997, Maroudas and Venn, 1977).

In the present study, we studied bovine cartilage samples at variable stages of spontaneous degeneration. Because the factors affecting the attenuation in spontaneously degenerated cartilage are still not determined, we assessed the inter-relationships between the US attenuation and various histologic, biochemical and mechanical parameters measured in bovine cartilage samples. Cartilage samples were graded into three groups using histologic Mankin score, and the dependence of US attenuation parameters and speed on cartilage degeneration were studied. With these analyses, we aimed to expand knowledge on AC attenuation mechanisms and to reveal the diagnostic ability of attenuation measurements to detect typical osteoarthrotic alterations (i.e., changes in the cartilage stiffness and contents of PG, collagen or water). Additionally, a novel quality parameter, cartilage quality index (CQI), was established to evaluate tissue integrity even more objectively than by using the present histologic grading methods.