Quantitative MRI of articular cartilage and its clinical applications

Abstract

Cartilage is one of the most essential tissues for healthy joint function and is compromised in degenerative and traumatic joint diseases. There have been tremendous advances during the past decade using quantitative MRI techniques as a noninvasive tool for evaluating cartilage, with a focus on assessing cartilage degeneration during osteoarthritis (OA). In this review, after a brief overview of cartilage composition and degeneration, we discuss techniques that grade and quantify morphologic changes as well as the techniques that quantify changes in the extracellular matrix. The basic principles, in vivo applications, advantages, and challenges for each technique are discussed. Recent studies using the OA Initiative (OAI) data are also summarized. Quantitative MRI provides noninvasive measures of cartilage degeneration at the earliest stages of joint degeneration, which is essential for efforts toward prevention and early intervention in OA.

Keywords: cartilage; osteoarthritis; quantitative MRI.

Fig. 1

Articular cartilage and degeneration during osteoarthritis (OA). Changes at early stages of OA include hydration, loss of proteoglycan, thinning and disruption of collagen. Changes in late stages include further loss of proteoglycan and collagen, dehydration, extensive fibrillation and cartilage thinning, and eventually denudation of the subchondral bone.

Fig. 2

Representative MR images with different stages of cartilage lesions and the corresponding WORMS scores.

Fig. 3

Cartilage thickness maps of the knee cartilage overlaying over the femur (left) and tibia (right) respectively.

Fig. 4

An example of intra-subject longitudinal cartilage thickness matching using the technique described by Carballido-Gamio et al (11). DESS: dual echo steady state.

Fig. 5

dGEMRIC images of the hips of a patient with unilateral hip dysplasia. The symptomatic hip has a much lower dGEMRIC index. Numbers in the images refer to the average dGEMRIC index for the region shown in color. (Figure reprint from reference (29), with permission).

Fig. 6

Variation of T₂ quantification was observed using different sequences in phantoms (left) and in vivo knees (right) (30). SE: spin-echo; FSE-4: fast spin-echo (FSE) with echo train length (ETL) as 4; FSE-8: FSE with ETL as 8; FSE-16: FSE with ETL as 16; MESE: multi-echo spin-echo; SGPR: spoiled gradient echo acquisition.

Fig. 7

T₂ maps of a sedentary subject (A), a light exerciser (B), and a moderate/strenuous exerciser (C) from the OAI cohort with risk factors for knee OA. In subjects with risk factors for knee OA from OAI cohort, light exercisers had lower T₂ values compared with both sedentary and moderate/strenuous exercisers (48), suggesting that there could be a “U” shape of the effect of exercise to cartilage biochemistry.

Fig. 8

T_1ρ and T₂ maps of a healthy control (a), a subject with mild OA (b) and a subject with severe OA (c). Significant elevation of T_1ρ and T₂ values were observed in subjects with OA. T_1ρ and T₂ elevation had different spatial distribution and may provide complementary information associated with the cartilage degeneration.

Fig. 9

T_1ρ maps of the lateral side (A and B) and medial side (C and D) of an ACL-injured knee at baseline (A and C) and one-year follow up (B and D) (65). T_1ρ values in lateral-posterior tibial cartilage (the region overlying bone bruises, white arrows) were elevated significantly in ACL-injured knees at baseline and remained high at one-year follow despite resolution of bone bruise in lateral tibia. T_1ρ values in the contacting area of medial femoral condyle and medial tibia (blue arrow) were significantly elevated in ACL-injured knees at one-year follow up.

Fig. 10

gagCEST images of a human patella in vivo with irradiation at δ = −1.0 ppm (left), δ = +1.0 ppm (middle), and the difference image (right), displaying a clear demarcation of a cartilage lesion and GAG loss on the medial facet (indicated by arrows) (Figure reprint from reference (73), with permission).

Fig. 11

High-spatial-resolution morphologic MR images and MR imaging parameter maps (ADC, FA, and T2) in healthy volunteer (top row: 31-year-old man, right knee) and a subject with OA (bottom row: 60-year-old woman, Kellgren-Lawrence grade of 3) (Figure reprint from reference (79), with permission).

Fig. 12

In vivo sodium imaging at 7T and 3T respectively. 3T: 15000 projections, RF 80°/0.5 ms, TR 80 ms, 2mm isotropic resolution, time of acquisition (TA) 20 min; 7T: 10000 projections, RF 90°/0.5 ms, TR 100 ms, 2mm isotropic resolution, TA 17 min. IR RECT: 3D radial with the rectangular inversion; IR WURST: 3D radial with the adiabatic Wide-band Uniform Rate and Smooth Truncation (WURST) inversion pulse (“IR WURST” experiment). Images Courtesy of Dr. Ravinder Regatte.

Fig. 13

Representative T_1ρ (left) and T₂ (right) maps 3–6 months and 1-year after microfracture (96). RT: repaired tissue; NC: normal cartilage.