Abstract
Objectives:
Post-traumatic osteoarthritis (PTOA) is a significant adverse outcome of injuries such as anterior cruciate ligament (ACL) tears, which results in pain and functional limitation. Approximately 35% of patients develop symptomatic PTOA within 10 years of an ACL injury, and there are currently no treatments that can prevent or reverse its onset. Early detection of PTOA is critical for timely intervention, yet current methods are limited. MRI, including advanced quantitative techniques like T1ρ and T2 mapping, offers a promising alternative. These methods can assess the biochemical composition of cartilage, with T1ρ reflecting proteoglycan content and T2 mapping correlating with collagen structure. Our prior research has demonstrated that these MRI sequences can detect early osteoarthritis-related changes in cartilage within a year following ACL reconstruction. However, it remains unclear whether these early imaging biomarkers can predict the long-term development of PTOA.
The purpose of this study is to determine the association between early cartilage degeneration, as measured by T1r and T2 relaxation times, with long-term whole-joint knee health, as measured by semi-quantitative Whole-Organ MRI scoring (WORMS) at approximately 10 years after ACL reconstruction.
Methods:
A total of 13 patients who previously participated in two longitudinal cohort studies of PTOA after ACL reconstruction agreed to participate in this study. All procedures were IRB-approved, and all patients provided informed consent to participate. Patients had previously participated in imaging-based studies after ACL reconstruction that included quantitative knee MRI at 1 year after surgery. At the time of subsequent imaging, all patients were 10 years post-operative from ACL reconstruction surgery. The mean age was 32.4±5.4 years at the time of injury and body mass index (BMI) was 24.2±2.6 kg/m2. All patients underwent ACL reconstruction with soft tissue graft (8 autografts). All patients completed a standardized post-operative rehabilitation protocol.
The MR images one year after surgery were acquired using a 3T MRI scanner with an eight-channel knee coil. The imaging protocol included the sagittal T1ρ/T2 quantification sequence [TR/ TE, 8 ms/3 ms; FOV, 14 cm; matrix, 256x128; slice thickness, 4 mm; views per segment, 64; spin-lock frequency, 500 Hz; T1ρ time of spin-lock: 0, 10, 40, 80 ms; T2 preparation TE: 0, 13.7, 27.3, 54.7 ms[MOU1] [MOU2] [MOU3] [DL4] for 8 cases, 0, 12.87, 25.69, 51.39 ms for 5 cases[MOU5] [DL6] ]. The MR images 10-year after surgery were acquired using a 3T MRI scanner with a 16-channel knee coil, including the bilateral 3D proton-density fat-saturated fast-spin-echo (CUBE) [TR/TE, 1,200/27 ms; echo train length, 32; matrix, 512 × 512 pixels; field of view (FOV), 16 cm; slice thickness, 0.6 mm (interpolated into 0.3 mm)] for evaluate cartilage and meniscal morphology.
The postprocessing for one-year T1ρ and T2 analysis was performed using in-house developed software with Matlab. We calculated the mean values in each cartilage region; medial femur (MF), medial tibia (MT), lateral femur (LF), lateral tibia (LT). Furthermore, MT was divided into three sub-compartments concerning the meniscus, anterior MT (aMT), central MT (cMT), and posterior MT (pMT) (Figure 1). The WORMS scores were evaluated by two radiologists on the CUBE images 10 years after surgery. In the WORMS, we used WORMS total as the total score out of all scores, MM_sum as the total medial meniscus (MM) score, LM_sum as the total lateral meniscus (LM) score, and CART and BME as cartilage and bone marrow edema score of each region.
As primary analyses, the relationships between the T1ρ and T2 values of each region and MT sub-compartment one year postoperatively and the WORMS total score 10 years postoperatively were analyzed using Spearman test. As secondary analyses, we compared the one-year T1ρ and T2 values and the 10-year WORMS between the ipsilateral and contralateral sides using paired t-tests. The MT and each sub-compartment were investigated the correlations with WORMS scores using Spearman test. Statistical significance was set at p < .05.
Results:
The one-year T1ρ value in MT was significantly correlated with the 10-year WORMS total (r = 0.609, p = 0.029), and there was a moderate correlation that was not statistically significant between the 10y_WORMS total and the one-year T2 value in MT (r = 0.523, p = 0.067). The other region did not correlate with the 10-year WORMS total. For the MT sub-compartment analyses, the one-year T1ρ in pMT also had a significant correlation with the 10-year WORMS total (r=0.606 and p=0.03) (Table 1). The MT and each sub-compartment had some correlations with each WORMS score, such as BME in LT, CART in LT, CART in MF, MM_sum, and LM_sum (Table 2).
Comparing one-year T1ρ and T2 values with the contralateral knee, the T1ρ in MT, cMT, and pMT were greater in the ipsilateral knee (p = 0.009, 0.017, and 0.004) and the T1ρ and T2 in the other region did not differ in sides (Figure 2). Comparing the 10-year WORMS total with the contralateral knee, there was no significant difference between in the ipsilateral and in the contralateral knees (p = 0.515). However, the ipsilateral knee had a greater WORMS cartilage score in the lateral compartment (p = 0.039) and in the LM_sum (p = 0.049).
Conclusions:
The elevated cartilage T1ρ values in the medial tibia one year postoperatively were found and correlated with the degree of the entire knee joint destruction using WORMS at 10 years postoperatively. This is the initial report of an association between quantitative cartilage assessment early postoperatively and knee joint destruction long-term postoperatively. This study could reveal that the early elevated T1ρ was truly associated with future OA progression and might suggest the qMRI could be used to show if an intervention is working for PTOA.
