A new method for investigating osteoarthritis using Fast Field-Cycling nuclear magnetic resonance

Abstract

Osteoarthritis in synovial joints remains a major cause of long-term disability worldwide, with symptoms produced by the progressive deterioration of the articular cartilage. The earliest cartilage changes are thought to be alteration in its main protein components, namely proteoglycan and collagen. Loss of proteoglycans bound in the collagen matrix which maintain hydration and stiffness of the structure is followed by collagen degradation and loss. The development of new treatments for early osteoarthritis is limited by the lack of accurate biomarkers to assess the loss of proteoglycan. One potential biomarker is magnetic resonance imaging (MRI). We present the results of a novel MRI methodology, Fast Field-Cycling (FFC), to assess changes in critical proteins by demonstrating clear quantifiable differences in signal from normal and osteoarthritic human cartilage for in vitro measurements. We further tested proteoglycan extracted cartilage and the key components individually. Three clear signals were identified, two of which are related predominantly to the collagen component of cartilage and the third, a unique very short-lived signal, is directly related to proteoglycan content; we have not seen this in any other tissue type. In addition, we present the first volunteer human scan from our whole-body FFC scanner where articular cartilage measurements are in keeping with those we have shown in tissue samples. This new clinical imaging modality offers the prospect of non-invasive monitoring of human cartilage in vivo and hence the assessment of potential treatments for osteoarthritis. Keywords: Fast Field-Cycling NMR; human hyaline cartilage; Osteoarthritis; T1 dispersion; quadrupolar peaks; protein interactions.

Keywords: Fast field-cycling NMR; Human hyaline cartilage; Osteoarthritis; Protein interactions; Quadrupolar peaks; T1 dispersion.

Fig. 1

Example of a basic Fast Field-Cycling Pulse sequence showing the three main periods: Polarisation, Evolution and Detection. During Polarisation the sample magnetisation is built up in a relatively strong magnetic field (B₀^P). During Evolution the magnetic field is switched to a value of interest (B₀^E) and the sample undergoes spin-lattice relaxation at that field. Finally, during the Detection period the NMR signal is read out at field B₀^D.

Fig. 2

Example of an R₁ dispersion curve from healthy femoral head cartilage. The general downward trend follows piecewise segments of lines, which in a logarithm plot correspond to power laws, with sudden changes of regimes shown by vertical dashed lines (here around 3 kHz and 2.8 MHz). Three relaxation regimes appeared, the high-field segment is difficult to see by eye but is required for a good fit, as can be seen by extending the second segment (fine dashed line). Quadrupolar Peaks are visible around 0.7, 2.1 and 2.8 MHz PLF (yellow regions).

Fig. 3

Typical NMRD profiles of the same sample of healthy human cartilage using the long and short acquisition protocols. Differences can be seen in the slopes and amplitude of the quadrupolar peaks, which are quantified in Table 1.

Fig. 4

Evolution of the Quadrupolar Peaks amplitude with the delay between the excitation pulse and the data acquisition. The amplitude obtained from the long acquisition is shown by the horizontal dashed line. The duration of the signal of interest for the characterisation of osteoarthritis seems to be on the order of hundreds of microseconds.

Fig. 5

SDS PAGE gels of the extraction buffers obtained from the contralateral (b) and osteoarthritis lesion cartilage samples (c). The molecular markers are shown on the left for reference (a).

Fig. 6

Average amplitude of the high-frequency Quadrupolar Peaks for the different cores before and after proteoglycan extraction using a GuCl buffer. A marked decrease is clearly visible with the extraction process and the Quadrupolar Peaks amplitude reached after extraction is similar to that observed from long acquisitions in the knee (note the factor 0.73 between Quadrupolar Peaks amplitude and integrated Quadrupolar Peaks amplitude).

Fig. 7

Dispersion curve of humidified collagen matrix at 50% w/w, 37 °C. Large Quadrupolar Peaks are visible (Integrated Quadrupolar Peaks amplitude of 0.17 ± 0.04 s⁻¹, offset = 1.48 ± 0.01 s⁻¹, η = 0.41 ± 0.01, ν_Q = 0.812 ± 0.005 MHz, θ = 79 ± 9°, φ = 26 ± 6°, τ₁ = 1.1 ± 0.2 μs) and the background showed a 2-component behaviour (α_high = −0.25 ± 0.02 and α_low = −0.157 ± 0.004, transition at 1.8 ± 0.3 MHz).