MRI breakthrough for cartilage diagnosis wins Lanier Award
By Sally Chapralis
Development of the diagnostic dGEMRIC measure or index permits ‘visualization’ of the charged GAG distribution in cartilage with the resolution of proton MRI, and can be done in vivo or in vitro
Martha L. Gray, PhD; Deborah Burstein, PhD; Young-Jo Kim, MD, PhD; and Alice Maroudas, PhD, have received the 2007 Elizabeth Winston Lanier Award for their manuscript on “Magnetic Resonance Imaging (MRI) of Cartilage Glycosaminoglycan: Basic Principles, Imaging Technique, and Clinical Application.”
With an increasing number of therapeutic strategies to prevent, correct, or slow the progression of osteoarthritis, diagnostic measures that can serve as proxies for the present or predicted state of cartilage are critical in deciding which approach is most appropriate in any given situation.
Their report introduces a new MRI technique—delayed Gadolinium Enhanced MRI of Cartilage (dGEMRIC)—to “noninvasively image the glycosaminoglycan (GAG) concentration of articular cartilage…based on the concept of fixed [electrical] charge” resulting from GAG.
The team’s development of the diagnostic dGEMRIC measure or index “permits ‘visualization’ of the charged GAG distribution in cartilage with the resolution of proton MRI, and can be done in vivo or in vitro.” Clinical applications of dGEMRIC illustrate the technique’s ability to recognize cartilage degeneration in its early stages and allow clinicians to intervene with appropriate preventive and disease-reversing therapies.
Research, theory focus on electrical charges
Diseased cartilage lacks GAG, and the associated electrical charge that influences the mechanical properties of cartilage. Because the charges are essential elements of a GAG molecule, they are “fixed” (in contrast to mobile ions) to the cartilage extracellular matrix (ECM) and the density of charge associated with the GAG is referred to as the fixed charge density (FCD).
The fact that GAG concentration, and thus FCD, is lower in diseased cartilage than it is in normal cartilage led to the team’s research focus on FCD.
The team used the Donnan theory as its theoretical base. It was, as the authors write, “originally developed to describe the partitioning of ions and water across a semi-permeable membrane, where charged species restricted to one side of the membrane influenced the equilibrium concentration of electrolytes.”
In pioneering work, team researcher Alice Maroudas “demonstrated that the semipermeable membrane model was analogous to having charges fixed within the cartilage matrix and so the fixed charge influenced the composition of interstitial electrolytes in an analogous way.” This evidence supported the relationship between FCD and ion partitioning in cartilage as described by the Donnan theory.
A radiotracer method utilizing charged radioactive sodium was first used to measure GAG and FCD in cartilage. This research “provided among the first quantititative measurements of the distribution of GAG (or equivalently FCD) with depth from articular surface, with topological position, and with diseases.”
However, while the radiotracer and other methods were enabling, they had limitations. As the researchers point out, these analyses “were rarely applied to situations where one could monitor changes in response to an intervention” and “they were not feasible clinically.” The problem, as the team notes, was “related to the methods for measuring the concentration of an interstitial ionic probe.”
Donnan, MRI and FCD
Because the Donnan-based approach, using the measurement of mobile ion concentration to compute tissue [GAG], had been independently validated in non-MR studies, the team concluded that the biophysical principles on which this method is based are sound and generalizable to any mobile ion that could be measured by MR.
The researchers projected that, in principal, MR-based methods could provide a map or image reflecting GAG concentration, given the extent to which the concentration of mobile ion probe (sodium [Na+] and gadolinium diethylenetriamine-pentaacetic acid [Gd(DTPA)2]) can be measured by MRI and the equilibrium conditions of Donnan achieved. In vitro, where equilibrium conditions are known, the map would reflect absolute GAG concentration; where equilibrating conditions are unknown (in vivo), the map would reflect relative GAG concentrations.
In either case, three criteria would be critical: First, that the “image contrast is dominated by the mobile ion concentration (to provide the desired specificity)”; second, that the “mobile ion being measured must not bind to the matrix,” and, finally, that “the mobile ion being measured must be in (quasi-) equilibrium with the bathing solution or synovial fluid/blood.” These criteria would ensure that the map reflected GAG rather than binding or transport of the mobile ion.
dGEMRIC measures FCD and GAG
The principles were first applied to sodium MRI, and correspondence between the MRI-derived results and biochemistry and histology were demonstrated. Given the relatively low availability and resolution of sodium MRI, the next step was to validate that dGEMRIC (i.e. measuring [Gd(DTPA)2-] in cartilage) could be used to measure both FCD and GAG. As an in vitro method, dGEMRIC was validated against several gold standard methods.
“Specifically, [GAG] measured by dGEMRIC corresponded very closely to that measured by sodium MR and by the biochemical dimethylmethylene blue (DMMB) assay for both human and bovine cartilage. In addition, there is abundant qualitative evidence that dGEMRIC images correspond closely to the corresponding histological images,” noted the team.
However, because no gold standard measures exist by which to compare dGEMRIC in vivo, the research team used two approaches. First, they “demonstrated that in vivo images taken before total joint arthroplasty were similar to the corresponding in vitro images and histology of tissue harvested during surgery.”
The second validation strategy verified the essential role of charge in the measurement. Researchers compared two images—one made using Gd(DTPA)2- and the other using similarly-sized non-ionic contrast agent. They found a relatively uniform distribution of the non-ionic contrast agent in the joint cartilage, but a non-uniform distribution of the Gd(DTPA)2-. This difference in distribution, they write “was a consequence of the Gd(DTPA)2- being charged (and thus being sensitive to local fixed charge density.”
Overall, the researchers note, their “data support the overall conclusion that dGEMRIC can be a sensitive and specific measure of cartilage GAG.” But, “until other factors involved in quantifying GAG from a measurement of T1 are more fully understood, in vivo studies of dGEMRIC report the T1 measurement directly as a “dGEMRIC Index”.”
dGEMRIC presents opportunities
The team’s development of this new methodology to nondestructively measure FCD on a spatially localized basis offers a number of opportunities.
Researchers now have the ability to “track the distribution of GAG across cartilage over time in culture with high resolution, thereby enabling long-term in vitro studies of the evolution of degradation, development, or repair and of factors that might be involved.” dGEMRIC has also led to a number of in vivo cross-sectional studies that have provided “valuable clinical insights into human joint physiology and disease.”
The team notes that the dGEMRIC Index has opened other doors. It appears to “be sensitive to cartilage-modifying injuries,” for example, when evaluating either ACL or PCL injuries. It can also aid in the evaluation of pre-radiographic osteoarthritis. dGEMRIC data offers additional evidence that joint degeneration in dysplastic hips “can occur early in young adulthood and the onset is related to the severity of dysplasia.” This information can be clinically applied to improve future therapeutic strategies.
A pilot study to see if dGEMRIC could provide additional information about the timing and use of pelvic osteotomies proved relevant. “In this pilot study,” the writers explain, “nine of 52 osteotomies failed, and the dGEMRIC index was found to be the best predictor of failure.”
Finally, the team addressed functionality and mechanical properties. “The possibility that dGEMRIC might provide information in line with the mechanical information obtained at arthroscopy was demonstrated in a clinical study where the dGEMRIC index was lower in arthroscopically diseased compartments than arthroscopically reference compartments.”
In discussing their research into using dGEMRIC to measure cartilage GAG as a possible imaging biomarker, the writers say that “although much work remains to determine if such GAG measurements (perhaps combined with other measures) are predictive of ultimate clinical outcome, methods like dGEMRIC enable us to gain new insights into “normal” and therapeutically induced changes in cartilage GAG in vivo, information that was previously not available.”
The researchers received consulting and/or research funds from Procter and Gamble, Pharmacia and Pfizer, as well as grant funding from the National Institutes of Health, Orthopaedic Research and Educational Foundation, the Arthritis Foundation, and the Medical Research Council of Great Britain.