Quantification of cartilage and bone degeneration in ACL-injured knees using advanced MRI and MRSI

Daniel Kuo, Jonathan Cheng, Radu Bolbos, Alekos Theologis, Benjamin Ma, Xiaojuan Li

Anterior cruciate ligament (ACL) injuries are one of the most common ligament injuries of the knee. Although ACL reconstruction enables patients to return to active lifestyles, recent long-term studies have demonstrated that 10 to 15 years after surgery, 50% to 70% of these patients have documented radiological changes of osteoarthritis (OA). Because the average age of patients with ACL injuries is approximately 22 years, this results in significant morbidity for a very young and active population. Unfortunately, the mechanical and biochemical mechanisms responsible for cartilage degeneration in ACL-injured knees remain elusive. This presents a profound and largely unsolved diagnostic and treatment challenge. Our long-term goal is to develop advanced imaging techniques to noninvasively diagnose and monitor joint degeneration, with the expectation of providing a better understanding of the disease pathophysiology and ultimately improving patient management. The goals of this study are to: (1) explore the clinical significance of biochemical changes in the knee directly after ACL injury, as indicated primarily by bone marrow edema-like lesions (BMELs); and (2) detect early cartilage and bone degeneration after ACL reconstruction using quantitative magnetic resonance imaging (MRI) and magnetic resonance spectroscopic imaging (MRSI).

Patients with acute knee ACL injuries are scanned two to three months after initial injuries and prior to any surgery, and two weeks, six months, one year and two years after ACL reconstruction. The imaging protocol consists of T2-weighted fat-suppressed fast spin echo (FSE) images, high resolution spoiled gradient recalled (SPGR) images, T1ρ and T2 quantification using in-house developed sequences, and 3-D magnetic resonance spectroscopic imaging (3-D MRSI) in bone marrow. BMEL are segmented semi-automatically from T2-weighted FSE images and the volume of BMEL and the signal intensity increase in BMEL relative to normal bone marrow are quantified. Cartilage is segmented semi-automatically in SPGR images. From this, 3-D contours for cartilage overlying BMEL and surrounding cartilage are defined. The follow-up images are registered to the baseline images acquired prior to surgeries. T1ρ values in the cartilage overlying the baseline BMEL are calculated. T1ρ values are also quantified in all compartments of the joint at all time points, and in deep and superficial layers respectively.

At baseline, BMEL were found to be most common and largest in the lateral tibia (LT) and lateral femoral condyle (LFC) (n = 14). The lesions were observed to resolve over time as seen in a reduction in average volume and relative signal. At one- and two-year follow-up, 42% and 86% of BMEL were resolved, respectively (n = 7). Significant elevated water and unsaturated lipids and decreased saturated lipids were found in BMEL compared with normal bone marrow. Volume of BME correlated significantly with volume of elevated water based on 3-D MRSI but not with volume of elevated unsaturated lipids. At baseline, the average T1ρ values in cartilage overlying were significantly higher than that in surrounding cartilage in LT. Furthermore, elevated T1ρ values for injured cartilage have been confirmed using correlation with clinical evaluation during arthroscopy in two patients. At one-year follow-up, despite of the resolution of BMEL, T1ρ in the cartilage overlying the original BMEL were still significantly higher than those in the surrounding cartilage, suggesting potential persistent damage of the cartilage. In medial sides, T1ρ values were higher in medial femoral condyle (MFC) and medial tibia (MT) at one-year follow-up than baseline. The T1ρ values in middle part of MT 08/25/2008 were also significantly higher than T1ρ values in healthy controls, suggesting potential early degeneration in these regions. Our results suggest that advanced quantitative MRI (T1ρ) and 3-D MRSI may detect early biochemical changes in the joint that may predispose OA in ACL-injured knees.


Related Content

MQIR Research Directions