Correlations of low-field NMR and variable-field NMR parameters with osteoarthritis in human articular cartilage under load

Abstract

NMR experiments carried out at magnetic fields below 1 T provide new relaxation parameters unavailable with conventional clinical scanners. Contrast of T generally becomes larger towards low fields, as slow molecular reorientation processes dominate relaxation at the corresponding Larmor frequencies. This advantage has to be considered in the context of lower sensitivity and frequently reduced spatial resolution. The layered structure of cartilage is one example where a particularly strong variation of T across the tissue occurs, being affected by degenerative diseases such as osteoarthritis (OA). Furthermore, the presence of H- N cross-relaxation, leading to so-called quadrupolar dips in the H relaxation time dispersion, provide insight into the concentration and mobility of proteoglycans and collagen in cartilage, both being affected by OA. In this study, low-field imaging and variable-field NMR relaxometry were combined for the first time for tissue samples, employing unidirectional load to probe the mechanical properties. 20 human knee cartilage samples were placed in a compression cell, and studied by determining relaxation profiles without and with applied pressure (0.6 MPa) at 50 μm in-plane resolution, and comparing with volume-averaged T dispersion. Samples were subsequently stored in formalin, prepared for histology and graded according to the Mankin score system. Quadrupolar dips and thickness change under load showed the strongest correlation with Mankin grade. Average T and change of maximum T under load, as well as its position, correlate with thickness and thickness change. Furthermore, T (ω) above 25 mT was found to correlate with thickness change. While volume-averaged T is not a suitable indicator for OA, its change due to mechanical load and its extreme values are suggested as biomarkers available in low-field MRI systems. The shape of the dispersion T (ω) represents a promising access to understanding and quantifying molecular dynamics in tissue, pointing toward future in vivo tissue studies.

Multicomponent analysis of T1 relaxation in bovine articular cartilage at low magnetic fields

“…The FFC technique is not uniquely suitable for such a type of research; one can envisage T 1 distribution measurements performed on a low‐field MRI scanner equipped with a permanent magnet. The advantage of the FFC relaxometry over the steady‐field relaxometry is that it introduces through varying B 0 the second dimension that can be exploited in a complementary way or in its own right to monitor the OA‐induced changes in cartilage . Thus, attempts were made to analyze quasi‐continuous T 1 dispersion profiles before and after trypsin digestion in terms of elementary Bloembergen‐Purcell‐Pound–type relaxation processes with individual correlation times.…”

“…The MRI techniques developed for early diagnosis of OA exploit spatially resolved measurements of T 1 , T 2 , and T 1ρ as potential biomarkers, thanks to their correlation with the composition and structure of AC . In addition to the MRI mapping of T 1 and T 2 , fast field‐cycling (FFC) NMR relaxometry has recently been used in the attempt to correlate the biological state of AC cartilage with T 1 dispersion . The FFC experiments were carried out in the magnetic fields B 0 = 0.2 mT to 470 mT, which span quadrupolar relaxation enhancement (QRE) peaks.…”

“…The FFC experiments were carried out in the magnetic fields B 0 = 0.2 mT to 470 mT, which span quadrupolar relaxation enhancement (QRE) peaks. A statistically significant difference was found between osteoarthritic and healthy cartilage for both the dispersion amplitude and the area of the QRE peaks, as well as for the maximum values of T 1 and T 2 when the samples were subject to external load . The power‐law exponent α in at B 0 < 25 mT was found to correlate weakly with Mankin grade but significantly with other biomarkers .…”

“…The power‐law exponent α in at B 0 < 25 mT was found to correlate weakly with Mankin grade but significantly with other biomarkers . A number of parameters obtained at low magnetic field strengths were found to correlate with disease status only under external load …”

“…Such a decomposition is generally possible mathematically, and is expected to be robust, regardless of the underlying correlation functions and detailed molecular dynamics models. The formalism of the multi‐Lorentzian decomposition is introduced here instead of the power‐law approximation to the T 1 dispersion in the attempt to draw a correlation between g ( τ c ) and the changes caused by enzymatic treatment of cartilage.…”