Magnetic resonance imaging, X-ray and dual X-ray absorptiometry techniques for assessment and monitoring of knee osteoarthritis

Aims

Research projects completed and ongoing by members of our research team include: (i) validation of quantitative cartilage measurements using a 1 T extremity magnetic resonance imaging (MRI) scanner; (ii) radiographic evaluation of soft-tissue structures – a comparison between 1 T and 1.5 T MRI scanners; (iii) short-term and long-term reproducibility of computer-determined medial minimum joint space width (mJSW) of the knee using a fixed-flexion X-ray protocol; (iv) determination of 'normal' values of minimum medial joint space width of the knee, bone density using dual X-ray absorptiometry around the knee joint and cartilage volume and thickness in men and women per decade between the ages of 20 and 69 years; (v) determination of values of medial mJSW of the knee, bone density around the knee joint and cartilage volume and thickness in men and women with varying degrees of knee osteoarthritis (OA); (vi) development of a longitudinal database of patients with OA of the knee, documenting pain, physical function, quality of life, history, medication use, family history, physical examination, blood tests, plain film radiographs and MRI of the knee using a 1 T extremity MRI scanner; (vii) development of a more fully automated technique for cartilage segmentation of three-dimensional (3D) MRI data; and (viii) quantification of 3D trabecular bone structure using high-resolution MRI.

Methods

Healthy volunteers and patients with knee OA referred for assessment to a rheumatologist are invited to participate in the various studies. All patients are asked to complete preset questionnaires. Radiographs of the knee are performed postero-anteriorly in a standardized fixed-flexion position by trained technicians. Radiographs are subsequently digitized and analyzed for mJSW using an automated computer algorithm. A 1 T extremity MRI scanner is used for imaging each individual's knee. Sequences include a saggital T₁-weighted, 3D spoiled gradient echo with fat saturation, a saggital T₁-weighted fast spin echo (FSE), a coronal T₂-weighted FSE and an axial FSE inversion recovery sequence.

Results

Quantitative cartilage measurements of normal and OA subjects' knees were obtained using a validated segmentation technique on both 1 T and 1.5 T MRI images. The 1 T extremity MRI scanner has been found to be reliable and of comparable precision in comparison with a 1.5 T whole-body scanner.

The use of a fixed-flexion position and a computer algorithm to determine medial mJSW of the knee in patients having normal or osteoarthritic knees has been found to be reliable in both short-term and long-term studies. Forty-seven pairs of knee radiographs were assessed in the short-term study while 14 pairs in a long-term study were assessed. Reliability for both studies was measured by means of intraclass correlation coefficient and found to be 0.95–0.99 for both healthy subjects and those with OA.

A cross-sectional study of 45 patients referred for assessment of knee pain is being completed. Data analyses will include comparison of pain scales, history and physical examination with MRI findings in the knee.

Conclusions

Quantitative cartilage measurements using a 1 T office-based extremity MRI scanner have been found to be reliable and of comparable precision with those obtained using a 1.5 T whole-body MRI scanner. The use of a fixed-flexion radiographic technique for assessing mJSW has been found to be reliable in a long-term reproducibility study at our center. Early diagnosis and prompt assessments of patients with OA can be accessible using an office-based extremity MRI scanner and could prove to be a valuable clinical tool with the development of disease-modifying agents for the treatment of OA.