Despite the increasing interest in knee cartilage volume as an outcome measure in studies of osteoarthritis (OA), it is unclear what components of knee cartilage will be most useful as markers of structural change in the tibiofemoral joint. This study was undertaken to longitudinally compare changes in femoral and tibial cartilage volume in patients with OA.
One hundred seventeen patients with knee OA (58.1% women; mean ± SD age 63.7 ± 10.2 years) were examined. Femoral and tibial cartilage volumes (medial and lateral tibiofemoral joints) were determined from T1-weighted fat-saturated magnetic resonance images of the knee from coronal views.
The study population was followed up for a mean ± SD of 1.9 ± 0.2 years. In the medial tibiofemoral joint, the mean ± SD loss of cartilage was 0.15 ± 0.30 ml/year for femoral cartilage and 0.10 ± 0.25 ml/year for tibial cartilage. In the lateral tibiofemoral joint, the average loss was 0.15 ± 0.22 and 0.12 ± 0.16 ml/year for femoral and tibial cartilage, respectively. There was a significant correlation between the degree of loss of tibial cartilage and the degree of loss of femoral cartilage, in both tibiofemoral joints (r = 0.81, P < 0.001 at the medial tibiofemoral joint; r = 0.71, P < 0.001 at the lateral tibiofemoral joint).
Longitudinal changes in tibial cartilage and those in femoral cartilage are strongly related to one another. This suggests that in tibiofemoral disease, measuring tibial cartilage alone may be adequate, given the facts that measurements of the total femoral cartilage are less reproducible and there are difficulties inherent in identifying the most appropriate component of femoral cartilage to measure.
There has been increasing interest in the use of magnetic resonance imaging (MRI) in the measurement of knee cartilage volume as a possible outcome measure in arthritis (1–3). MRI has been shown to provide a valid quantification of cartilage volume as compared with findings of anatomic dissection. The method is reproducible, with coefficients of variation (CVs) of ∼2% (2–4). This technique has been used to explore factors that influence knee cartilage in healthy adults and children (3, 4).
One problem is that with most techniques currently in use to measure knee cartilage volume, manual manipulation has to be performed to varying degrees, and the techniques are consequently quite time-consuming and their use is limited to relatively few institutions (1–3). One potential approach is to attempt to limit the components of knee cartilage being measured and still retain cartilage measures that are a valid gauge of the state of joint cartilage. We have previously shown a strong correlation between femoral cartilage volume and tibial cartilage volume measured in both the medial and lateral tibiofemoral compartments of the knee in healthy subjects and those with osteoarthritis (OA) (5). Since measurements of the total femoral cartilage are less reproducible and there are difficulties in identifying the most appropriate component of femoral cartilage to measure, we speculated that for tibiofemoral disease, measuring tibial cartilage alone may provide useful information. In the present study, we performed a longitudinal investigation comparing the changes in femoral cartilage volume and tibial cartilage volume in patients with OA of the knee.
PATIENTS AND METHODS
Patients were recruited by using a combined strategy including advertising through local newspapers and the Victoria branch of the Arthritis Foundation of Australia, as well as collaboration with general practitioners, specialist rheumatologists, and orthopedic surgeons. Details about the patients have been reported previously (6). The study was approved by the Ethics Committee of the Alfred and Caulfield Hospitals in Melbourne, Australia. All patients provided informed consent. Patients ages ≥40 years who fulfilled the American College of Rheumatology clinical and radiographic criteria for knee OA (pain and radiographic evidence of osteophytes) (7) were examined. Patients were excluded if any other form of arthritis was present, if there was a contraindication to MRI, or if a total knee replacement was planned.
Each patient had an MRI performed on the symptomatic knee at baseline and ∼2 years later. If symptomatic OA was present in both knees, the knee with less severe OA as determined radiographically was studied. The patients' knees were imaged in the sagittal plane on a 1.5T whole-body MR unit (Sigma Advantage GE Medical Systems; Milwaukee, WI) with use of a commercial receive extremity coil as previously described (3, 4). The method to transform the images to the coronal plane has also been described previously (5).
Since the femoral cartilage is a continuous structure and forms part of 3 joints (the patellofemoral and medial and lateral tibiofemoral joints), the coronal view was used to measure both the femoral and tibial cartilage. This sequence allows best visualization of the femoral cartilage component of the medial and lateral tibiofemoral joint and is satisfactory for visualization of the tibial cartilage (5). Similar tibial cartilage volumes are obtained from the original sagittal sequence and the reformatted coronal data (5).
Articular cartilage volumes were determined by means of 3-dimensional image processing on an independent workstation, using the Osiris software package (University of Geneva, Geneva, Switzerland). The medial and lateral tibial cartilage plates cover the respective tibial plateaus and are easily identified as separate structures from other structures in the knee, in an analogous way to that described for tibial cartilage plates in sagittal scans (5, 6). The volumes of the medial and lateral tibial cartilage plates were isolated by manually drawing disarticulation contours around the cartilage boundaries on a section-by-section basis in an analogous method to that described for sagittal images (6). The medial and lateral femoral cartilage plates do not have a clear anatomic boundary. For these we developed a series of rules. The posterior boundary was defined as the first section in which articular cartilage was clearly identified as a discrete structure. The anterior boundary was the first image in which tibial cartilage also appeared. The outer boundary (either medial or lateral) was defined as the outer boundary of the respective cartilage plate. The inner boundary of the medial and lateral tibial cartilage was defined as the anatomic edge of the respective cartilage plate that can be clearly identified. The inner boundary of the medial and lateral femoral cartilage is not a clear anatomic structure, so a rule was used. This was defined as the peak of the angle subtended by the medial aspects of the medial and lateral femoral condyles.
The intraobserver reproducibility for repeat measures of cartilage volume from single acquisitions as measured by CV was as follows: medial tibial cartilage volume 2.3%, lateral tibial cartilage volume 2.4%, medial femoral cartilage volume 2.6%, lateral femoral cartilage volume 2.8%. The areas of the medial and lateral tibial plateaus were directly measured by manually drawing contours on the reformatted axial data (6). The CVs as a measure of intraobserver reproducibility for tibial plateau measures were as follows: medial 2.7% and lateral 2.8%.
To assess the relationship between change in femoral cartilage volume and change in tibial cartilage volume at the medial and lateral tibiofemoral joints, Spearman's correlation coefficient was used. Two-tailed P values less than 0.05 were considered significant. All analyses were performed using the SPSS statistical package (version 10.0.5; SPSS, Chicago, IL).
A total of 123 patients were included in the original OA study (6). Information was incomplete or missing for 6 patients. There were no significant differences between patients who were lost to followup and those who completed the study, in terms of age, sex, body mass index, or distribution of radiographic OA. The mean ± SD age was 63.7 ± 10.2 years, and 58.1% of the patients were female (Table 1). Fifty-three percent of the patients had grade 1 or 2 OA as measured by the Kellgren/Lawrence scale (8) (Table 1).
Table 1. Characteristics of the study population (n = 117)
Age, mean ± SD years
63.7 ± 10.2
Time between scans, mean ± SD years
1.9 ± 0.2
Kellgren/Lawrence grade, no. (%)
There was strong correlation between femoral cartilage volume and tibial cartilage volume, measured in both the medial and lateral tibiofemoral joint compartments at baseline and followup (Tables 2 and 3). In the medial compartment, the mean ± SD baseline volume of the femoral cartilage was 1.94 ± 0.85 ml and that of the tibial cartilage was 1.54 ± 0.46 ml (r = 0.91, P < 0.001). In the lateral compartment, the baseline volume of the femoral cartilage was 1.76 ± 0.62 ml and that of the tibial cartilage was 1.63 ± 0.55 ml (r = 0.92, P < 0.001). The correlation (r) between tibial and femoral cartilage volumes at followup was 0.81 (P < 0.001) in both compartments.
Table 2. Mean ± SD femoral and tibial cartilage volumes (ml) in the medial and lateral tibiofemoral joints
Difference in volume over study period
Difference in volume per year
Medial tibiofemoral joint
1.94 ± 0.85
1.66 ± 0.85
0.31 ± 0.51
0.15 ± 0.30
1.54 ± 0.46
1.36 ± 0.49
0.18 ± 0.46
0.10 ± 0.25
Lateral tibiofemoral joint
1.76 ± 0.62
1.47 ± 0.63
0.29 ± 0.41
0.15 ± 0.22
1.63 ± 0.55
1.42 ± 0.54
0.24 ± 0.31
0.12 ± 0.16
Table 3. Correlation between femoral and tibial cartilage volumes in the medial and lateral tibiofemoral joints
Correlation between tibial and femoral cartilage
Medial tibiofemoral joint
Lateral tibiofemoral joint
Changes in volumes
At followup of the study population after a mean ± SD of 1.9 ± 0.2 years, there was a significant correlation between the degree of loss of tibial cartilage and the degree of loss of femoral cartilage, in both the medial and lateral tibiofemoral joints (Tables 2 and 3). In the medial tibiofemoral joint, the mean ± SD loss of femoral cartilage was 0.15 ± 0.30 ml/year and the loss of tibial cartilage was 0.10 ± 0.25 ml/year (r = 0.81, P < 0.001). In the lateral tibiofemoral joint, the loss of femoral cartilage and the loss of tibial cartilage were 0.15 ± 0.22 ml/year and 0.12 ± 0.16 ml/year, respectively (r = 0.71, P < 0.001) (Tables 2 and 3 and Figure 1).
In this study, we have shown that there is a strong correlation between change in femoral cartilage volume and change in tibial cartilage volume, in the medial and lateral tibiofemoral joints of patients with radiographically evident OA. No previous studies have longitudinally compared change in cartilage volume at the tibia and the femur in the same joint compartments. We previously showed that similar information about structure of the lateral and medial tibiofemoral joint can be obtained by measuring either the femoral or the tibial cartilage (5). In both the medial and lateral tibiofemoral joints there was a strong correlation between findings in the femoral cartilage as measured from the coronal sequence and findings in the tibial cartilage, in normal subjects and those with OA (5), and a strong inverse relationship between the radiologic grade of joint space narrowing in both tibiofemoral joints and femoral and tibial cartilage volume has been demonstrated (9). We have extended these observations in the current study by demonstrating a strong correlation between the amount of tibial cartilage lost and the amount of femoral cartilage lost in the medial and lateral tibiofemoral joints.
A potential problem in our study is that the original MR images were acquired in the sagittal plane, and from this sequence it is difficult to reproducibly identify the femoral cartilage that would correspond to the tibiofemoral joint as seen on standing knee radiographs. For this reason, we reformatted the data into the coronal plane. Reconstruction of MRI-acquired images into different planes has been widely used clinically (10–12). We found that similar tibial cartilage volumes are obtained from the original sagittal sequence and the reformatted coronal data. The average over- and underestimate of lateral tibial cartilage volume from the reformatted coronal scans was 3.5% and that of medial tibial cartilage volume was 3.8%, compared with the originally acquired sagittal sequences (5). Although sagitally acquired images yield very good definition of the tibial cartilage, to avoid any differences introduced by reformatting we compared the tibial and femoral cartilage volumes after calculating both from the reformatted coronal MRI data. The rate of cartilage change using this method was similar to the rate we have recently reported from sagittal MRI (6).
Whether features of cartilage can be used as interim markers for OA has yet to be determined, although preliminary data indicate that this will be the case (6, 13, 14). The results of this study suggest that measuring either the tibial cartilage or the femoral cartilage alone may be adequate as a marker of OA at the tibiofemoral joint. This has the potential advantage of saving time in analyzing MRI data. Measurements of the total femoral cartilage are less reproducible (1–3), and there are difficulties inherent in identifying the most appropriate component of femoral cartilage to measure when examining the tibiofemoral joint. This is because the femoral cartilage articulates with the patellar and medial and lateral tibial cartilages and is part of the 3 joints that comprise the knee (the patellofemoral and medial and lateral tibiofemoral joints). There is no anatomic or easily reproducible delineation defining these articulations. Measuring the tibial cartilage has a potential advantage compared with femoral cartilage measurement since the former is a clearly defined anatomic structure and is thus less susceptible to error in its identification across different study sites and by different operators. However, in the case of studies that examine cartilage defects, which may not necessarily be symmetric across the joint, measurement of both femoral and tibial cartilage may be necessary.
MRI cartilage measurement is a rapidly evolving field. Measurement of cartilage volume is difficult and time-consuming because manual processing is still needed to obtain valid and reproducible results. However, determination of cartilage volume has the potential to be a useful tool in understanding factors that affect articular cartilage in health and disease, as well as an outcome measure in clinical trials. Our data suggest that it is possible to simplify the measures used, which may make cartilage volume measurement more feasible, especially in large-scale epidemiologic studies.
We would like to extend special thanks to the people who participated in and made this study possible.