- Split View
-
Views
-
Cite
Cite
M. Sharif, R. Granell, J. Johansen, S. Clarke, C. Elson, J. R. Kirwan, Serum cartilage oligomeric matrix protein and other biomarker profiles in tibiofemoral and patellofemoral osteoarthritis of the knee, Rheumatology, Volume 45, Issue 5, May 2006, Pages 522–526, https://doi.org/10.1093/rheumatology/kei216
- Share Icon Share
Abstract
Objectives. There is some evidence that tibiofemoral osteoarthritis (TFJ OA) and patellofemoral osteoarthritis (PFJ OA) may have different risk factors. To investigate the possibility that these conditions are separate disease entities, we compared biomarker profiles of patients with each disease.
Methods. Serum samples were taken from 222 patients who had knee pain and X-ray signs of knee OA. Eighty-two had only medial TFJ OA and 38 only PFJ OA in one or both knees. The remaining patients had either mixed disease or equivocal radiographic evidence of OA. The following biomarkers were measured in serum samples from baseline and follow-up visits: cartilage oligomeric matrix protein (COMP), glycosaminoglycan, keratan sulphate epitope 5D4, YKL-40, osteocalcin, C-telopeptide of type I collagen, hyaluronan and C-reactive protein.
Results. The two subsets of OA (TFJ and PFJ) had similar radiographic disease severity and there were no significant differences in the presence and patterns of pain scores (visual analogue scale and Western Ontario and McMaster Universities Osteoarthritis Index). No difference was found for the biomarkers between the two groups, with one exception. Both baseline and area under the curve per month COMP concentrations were significantly higher in the TFJ than the PFJ group (P<0.01).
Conclusions. The reduced serum COMP in PFJ disease compared with TFJ OA could be due to small articular cartilage volume in the latter or to a qualitative difference in cartilage metabolism.
Medial tibiofemoral osteoarthritis (TFJ OA) and patellofemoral osteoarthritis (PFJ OA) frequently occur in the same joint, although patients can be identified with either TFJ or PFJ, especially at early stages of the disease. The disease processes may differ between the two conditions. Thus no association exists between changes in patellar and tibial cartilage volume over 2 yr in the same knee [1]. Also, there is discordance between the anatomical changes of OA (X-ray changes) and pain and disability [2], determinants of osteoarthritic activity and outcome such as general health status and psychological factors [3], and muscle weakness [4, 5]. This discordance is probably due to differences between the tibiofemoral and patellofemoral joints themselves. For example, climbing stairs and steps will be a greater problem than flat walking for patients with PFJ disease and vice versa for those affected by TFJ disease. Obesity and meniscectomy are stronger risk factors for TFJ disease, while Heberden's nodes and family history are more closely associated with PFJ involvement [6]. In addition, squatting (both occupational and non-occupational) is a major risk factor for TFJ OA but not for PFJ OA [6, 7]. If TFJ and PFJ are different diseases then the patterns of serum biomarkers of joint tissue metabolism in the two diseases might also be divergent. In order to test this hypothesis we compared the clinical, radiographic and biomarker profiles of patients with each subset of OA.
Methods
Patients
Two hundred and twenty-two patients with persistent (>3 months) pain in one or both knees and radiographic evidence of TFJ or PFJ OA in at least one knee as considered by the screening clinician were recruited (see [8], where only the TFJ patients are reported). Approval for the study was given by the institutional review board (Research Ethics Committee) of the United Bristol Healthcare NHS Trust, Bristol, UK. Age, gender, body mass index (BMI), pain [visual analogue scale (VAS), 0–100 mm] and the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) [9] were recorded at baseline to describe the patient cohort.
Radiographs and intra-articular classification of patients
Plain X-rays were taken standing for the bearing anteroposterior view and lying lateral views in 30° flexion using a standardized protocol [10]. Radiographs obtained at study entry and over the 5 yr of follow-up were interpreted independently by two experienced readers during several sessions, in random order, with the identity labels concealed. Joint space width (JSW) was measured to the nearest 0.5 mm using a transparent ruler and taken as the average of the two readers. The coefficient of variation was 11.3% for a random sample of 20. The readers also made a visual assessment of PFJ joint space narrowing as absent (0) or present (1), and the mean value was used. Osteophytes were graded as none (0), mild (1), moderate (2) or severe (3) for the medial and patellofemoral compartments, and the mean used in the analysis. The criteria of Kellgren and Lawrence (K&L) [11], usually applied to the joint as a whole, were used to grade the medial TFJ and the PFJ, taking the lowest ‘compartmental K&L’ grade of the two readers. Patients were divided into those with TFJ OA (one or both medial TFJ K&L scores >2 but both PFJ scores <3), and those with PFJ OA (both medial TFJ K&L scores <3 but one or both PFJ scores >2).
Blood samples
Serum was collected at baseline and up to nine occasions during the following 5 yr as described previously [8].
Serum biomarkers of cartilage, bone and synovial tissue metabolism
Cartilage oligomeric matrix protein (COMP), a marker of cartilage damage/loss, was measured using the COMP™ enzyme-linked immunosorbent assay (ELISA) kit (AnaMar Medical, Sweden) as described recently [8]. Concentrations of glycosaminoglycan (GAG) and keratan sulphate epitope 5D4 (KS), which are putative markers of cartilage loss, were measured by the dimethylmethylene blue assay [12] and an ELISA inhibition assay [13], respectively. YKL-40, a possible marker of cartilage destruction and/or synovial inflammation, was determined by a two-site, sandwich-type ELISA (Quidel, Mountain View, CA, USA) [14]. Osteocalcin (OC), a proposed marker of bone formation and the C-telopeptide of type I collagen (CTX-I), a marker of bone resorption, were measured using immunoassays (N-MID Osteocalcin and β-CrossLaps/serum, respectively, Roche Diagnostics, Mannheim, Germany). C-reactive protein (CRP), a systemic marker of inflammation and known to be elevated in OA, was measured using a sensitive capture ELISA [15]. Serum levels of hyaluronate (HA), a putative marker of synovial inflammation, was measured by an ELISA inhibition assay using a slight modification of the method used by Goldberg [16, 17].
Statistical analysis
To ensure a clear-cut comparison between the two knee compartments (tibiofemoral and patellofemoral), patients were only included if the compartmental K&L score was >2 in that compartment in one or both knees and <3 in the other compartment in both knees. The primary analysis of serum biomarkers used baseline measurements, but the availability of sequential samples over 5 yr allowed a secondary analysis using the area under the curve (AUC) (eliminating patients who had total knee replacement [8]). Clinical symptoms and the pattern of biomarkers in the two groups were compared using the two-tailed t-test, allowing for the effects of age, sex and BMI (after natural log transformation where data were not normally distributed). Original mean values were calculated on the two groups and the corresponding adjusted differences of the means were compared using two-tailed t-tests. The distribution of radiographic abnormalities was compared using the χ2 test.
Results
Separation of radiographic groups
Of the 222 subjects, 82 had only medial TFJ OA, 38 had only patellofemoral OA, 29 had both compartmental K&L scores <3, 34 had OA in both compartments, 37 had normal X-rays and two of the X-rays were lost. The characteristics of the TFJ only and PFJ only patients are summarized in Table 1. There was no difference in the age distribution, but there were more females with PFJ OA and this group also had a higher BMI. The medial tibial JSW was normal in the PFJ group while in the TFJ group it was significantly reduced. PFJ joint space narrowing (JSN) was more severe in the PFJ group (Table 2) and medial osteophytes were more severe in the TFJ group. Thus the intended separation into two different radiographic groups was achieved.
Variable . | TFJ group (n = 82) . | PFJ group (n = 38) . | Difference between TFJ and PFJ groups . | P . |
---|---|---|---|---|
Clinical | ||||
Age (yr) | 64.4 (62.4, 66.4) | 65.4 (62.7, 68.1) | −1.0 (−4.6, 2.5) | 0.555 |
Gender (M:F) | 46:35 | 6:32 | <0.001 | |
Body mass index (kg/m) | 34.2 (32.0, 36.4) | 40.3 (36.5, 44.1) | −6.0 (−10.2, −1.9) | <0.01 |
Pain score (VAS) | 52.0 (5.6, 5.8) | 51.2 (4.1, 6.0) | 0.08 (−1.1, 1.2) | 0.892 |
WOMAC | ||||
Pain | 7.8 (6.9, 8.6) | 8.0 (6.7, 9.3) | −0.25 (−1.7, 1.2) | 0.742 |
Stiffness | 3.6 (1.9, 5.3) | 3.9 (3.3, 4.5) | −0.24 (−0.9, 0.5) | 0.501 |
Function | 25.1 (22.4, 27.8) | 26.9 (22.6, 31.2) | −1.74 (−6.8, 3.3) | 0.499 |
Radiographic | ||||
Medial JSW (mm)a | 2.3 (1.9, 2.7) | 4.5 (4.2, 4.8) | −2.2 (−2.8, −1.6) | <0.001 |
Biomarkersb | ||||
Cartilage | ||||
COMP (U/l) | 12.99 (12.4, 13.6) | 10.78 (10.1, 11.4) | 1.93 (0.77, 3.08) | 0.001 |
GAG (μg/ml) | 41.34 (39.1, 43.6) | 38.24 (35.5, 41.0) | 0.42 (−3.47, 4.31) | 0.832 |
KS (ng/ml) | 1158 (1108, 1207) | 1090 (1016,1165) | 38.25 (−59.1, 135.6) | 0.439 |
YKL-40 (ng/ml)c | 4.49 (4.4, 4.6) | 4.52 (4.3, 4.8) | 0.09 (−1.15, 0.32) | 0.463 |
Bone | ||||
OC (ng/ml) | 18.48 (17.0, 20.0) | 18.05 (16.1, 20.0) | 1.89 (−0.79, 4.59) | 0.167 |
CTX-I (pg/ml) | 0.26 (0.2, 0.3) | 0.26 (0.2, 0.3) | 0.026 (−0.03, 0.08) | 0.339 |
Synovium/inflammation | ||||
HA (ng/ml)c | 5.21 (5.1, 5.3) | 5.09 (4.9, 5.3) | 0.137 (−0.11, 0.38) | 0.276 |
CRP (μg/L)c | 7.82 (7.6, 8.1) | 7.64 (7.2, 8.1) | 0.129 (−0.41, 0.67) | 0.637 |
Variable . | TFJ group (n = 82) . | PFJ group (n = 38) . | Difference between TFJ and PFJ groups . | P . |
---|---|---|---|---|
Clinical | ||||
Age (yr) | 64.4 (62.4, 66.4) | 65.4 (62.7, 68.1) | −1.0 (−4.6, 2.5) | 0.555 |
Gender (M:F) | 46:35 | 6:32 | <0.001 | |
Body mass index (kg/m) | 34.2 (32.0, 36.4) | 40.3 (36.5, 44.1) | −6.0 (−10.2, −1.9) | <0.01 |
Pain score (VAS) | 52.0 (5.6, 5.8) | 51.2 (4.1, 6.0) | 0.08 (−1.1, 1.2) | 0.892 |
WOMAC | ||||
Pain | 7.8 (6.9, 8.6) | 8.0 (6.7, 9.3) | −0.25 (−1.7, 1.2) | 0.742 |
Stiffness | 3.6 (1.9, 5.3) | 3.9 (3.3, 4.5) | −0.24 (−0.9, 0.5) | 0.501 |
Function | 25.1 (22.4, 27.8) | 26.9 (22.6, 31.2) | −1.74 (−6.8, 3.3) | 0.499 |
Radiographic | ||||
Medial JSW (mm)a | 2.3 (1.9, 2.7) | 4.5 (4.2, 4.8) | −2.2 (−2.8, −1.6) | <0.001 |
Biomarkersb | ||||
Cartilage | ||||
COMP (U/l) | 12.99 (12.4, 13.6) | 10.78 (10.1, 11.4) | 1.93 (0.77, 3.08) | 0.001 |
GAG (μg/ml) | 41.34 (39.1, 43.6) | 38.24 (35.5, 41.0) | 0.42 (−3.47, 4.31) | 0.832 |
KS (ng/ml) | 1158 (1108, 1207) | 1090 (1016,1165) | 38.25 (−59.1, 135.6) | 0.439 |
YKL-40 (ng/ml)c | 4.49 (4.4, 4.6) | 4.52 (4.3, 4.8) | 0.09 (−1.15, 0.32) | 0.463 |
Bone | ||||
OC (ng/ml) | 18.48 (17.0, 20.0) | 18.05 (16.1, 20.0) | 1.89 (−0.79, 4.59) | 0.167 |
CTX-I (pg/ml) | 0.26 (0.2, 0.3) | 0.26 (0.2, 0.3) | 0.026 (−0.03, 0.08) | 0.339 |
Synovium/inflammation | ||||
HA (ng/ml)c | 5.21 (5.1, 5.3) | 5.09 (4.9, 5.3) | 0.137 (−0.11, 0.38) | 0.276 |
CRP (μg/L)c | 7.82 (7.6, 8.1) | 7.64 (7.2, 8.1) | 0.129 (−0.41, 0.67) | 0.637 |
Values are means (95% CI) for TFJ and PFJ OA cohorts and difference of means (95% CI). Medial JSW data are for the worst knee at baseline. Column 4 shows the difference of means (95% CI) after allowing for gender, age and sex. The P values correspond to the adjusted values.
aIn the few patients where the worst knee was not specified the average of both knees was taken.
bn = 73–77 for the TFJ group and 32–26 for the PFJ group depending on biomarker.
cSerum HA, CRP and YKL-40 were found to be skewed and therefore the values were log transformed (Ln).
Variable . | TFJ group (n = 82) . | PFJ group (n = 38) . | Difference between TFJ and PFJ groups . | P . |
---|---|---|---|---|
Clinical | ||||
Age (yr) | 64.4 (62.4, 66.4) | 65.4 (62.7, 68.1) | −1.0 (−4.6, 2.5) | 0.555 |
Gender (M:F) | 46:35 | 6:32 | <0.001 | |
Body mass index (kg/m) | 34.2 (32.0, 36.4) | 40.3 (36.5, 44.1) | −6.0 (−10.2, −1.9) | <0.01 |
Pain score (VAS) | 52.0 (5.6, 5.8) | 51.2 (4.1, 6.0) | 0.08 (−1.1, 1.2) | 0.892 |
WOMAC | ||||
Pain | 7.8 (6.9, 8.6) | 8.0 (6.7, 9.3) | −0.25 (−1.7, 1.2) | 0.742 |
Stiffness | 3.6 (1.9, 5.3) | 3.9 (3.3, 4.5) | −0.24 (−0.9, 0.5) | 0.501 |
Function | 25.1 (22.4, 27.8) | 26.9 (22.6, 31.2) | −1.74 (−6.8, 3.3) | 0.499 |
Radiographic | ||||
Medial JSW (mm)a | 2.3 (1.9, 2.7) | 4.5 (4.2, 4.8) | −2.2 (−2.8, −1.6) | <0.001 |
Biomarkersb | ||||
Cartilage | ||||
COMP (U/l) | 12.99 (12.4, 13.6) | 10.78 (10.1, 11.4) | 1.93 (0.77, 3.08) | 0.001 |
GAG (μg/ml) | 41.34 (39.1, 43.6) | 38.24 (35.5, 41.0) | 0.42 (−3.47, 4.31) | 0.832 |
KS (ng/ml) | 1158 (1108, 1207) | 1090 (1016,1165) | 38.25 (−59.1, 135.6) | 0.439 |
YKL-40 (ng/ml)c | 4.49 (4.4, 4.6) | 4.52 (4.3, 4.8) | 0.09 (−1.15, 0.32) | 0.463 |
Bone | ||||
OC (ng/ml) | 18.48 (17.0, 20.0) | 18.05 (16.1, 20.0) | 1.89 (−0.79, 4.59) | 0.167 |
CTX-I (pg/ml) | 0.26 (0.2, 0.3) | 0.26 (0.2, 0.3) | 0.026 (−0.03, 0.08) | 0.339 |
Synovium/inflammation | ||||
HA (ng/ml)c | 5.21 (5.1, 5.3) | 5.09 (4.9, 5.3) | 0.137 (−0.11, 0.38) | 0.276 |
CRP (μg/L)c | 7.82 (7.6, 8.1) | 7.64 (7.2, 8.1) | 0.129 (−0.41, 0.67) | 0.637 |
Variable . | TFJ group (n = 82) . | PFJ group (n = 38) . | Difference between TFJ and PFJ groups . | P . |
---|---|---|---|---|
Clinical | ||||
Age (yr) | 64.4 (62.4, 66.4) | 65.4 (62.7, 68.1) | −1.0 (−4.6, 2.5) | 0.555 |
Gender (M:F) | 46:35 | 6:32 | <0.001 | |
Body mass index (kg/m) | 34.2 (32.0, 36.4) | 40.3 (36.5, 44.1) | −6.0 (−10.2, −1.9) | <0.01 |
Pain score (VAS) | 52.0 (5.6, 5.8) | 51.2 (4.1, 6.0) | 0.08 (−1.1, 1.2) | 0.892 |
WOMAC | ||||
Pain | 7.8 (6.9, 8.6) | 8.0 (6.7, 9.3) | −0.25 (−1.7, 1.2) | 0.742 |
Stiffness | 3.6 (1.9, 5.3) | 3.9 (3.3, 4.5) | −0.24 (−0.9, 0.5) | 0.501 |
Function | 25.1 (22.4, 27.8) | 26.9 (22.6, 31.2) | −1.74 (−6.8, 3.3) | 0.499 |
Radiographic | ||||
Medial JSW (mm)a | 2.3 (1.9, 2.7) | 4.5 (4.2, 4.8) | −2.2 (−2.8, −1.6) | <0.001 |
Biomarkersb | ||||
Cartilage | ||||
COMP (U/l) | 12.99 (12.4, 13.6) | 10.78 (10.1, 11.4) | 1.93 (0.77, 3.08) | 0.001 |
GAG (μg/ml) | 41.34 (39.1, 43.6) | 38.24 (35.5, 41.0) | 0.42 (−3.47, 4.31) | 0.832 |
KS (ng/ml) | 1158 (1108, 1207) | 1090 (1016,1165) | 38.25 (−59.1, 135.6) | 0.439 |
YKL-40 (ng/ml)c | 4.49 (4.4, 4.6) | 4.52 (4.3, 4.8) | 0.09 (−1.15, 0.32) | 0.463 |
Bone | ||||
OC (ng/ml) | 18.48 (17.0, 20.0) | 18.05 (16.1, 20.0) | 1.89 (−0.79, 4.59) | 0.167 |
CTX-I (pg/ml) | 0.26 (0.2, 0.3) | 0.26 (0.2, 0.3) | 0.026 (−0.03, 0.08) | 0.339 |
Synovium/inflammation | ||||
HA (ng/ml)c | 5.21 (5.1, 5.3) | 5.09 (4.9, 5.3) | 0.137 (−0.11, 0.38) | 0.276 |
CRP (μg/L)c | 7.82 (7.6, 8.1) | 7.64 (7.2, 8.1) | 0.129 (−0.41, 0.67) | 0.637 |
Values are means (95% CI) for TFJ and PFJ OA cohorts and difference of means (95% CI). Medial JSW data are for the worst knee at baseline. Column 4 shows the difference of means (95% CI) after allowing for gender, age and sex. The P values correspond to the adjusted values.
aIn the few patients where the worst knee was not specified the average of both knees was taken.
bn = 73–77 for the TFJ group and 32–26 for the PFJ group depending on biomarker.
cSerum HA, CRP and YKL-40 were found to be skewed and therefore the values were log transformed (Ln).
. | . | Number of knees . | . | . | |
---|---|---|---|---|---|
Radiographic measurement . | . | Tibiofemoral group . | Patellofemoral group . | P (difference in distributions medial vs patellofemoral) . | |
Medial K&L | Score 0 | 9 | 27 | <0.001 | |
Score 1 | 1 | 5 | |||
Score 1.5 | 1 | 1 | |||
Score 2 | 3 | 5 | |||
Score 3 | 66 | 0 | |||
Score 4 | 2 | 0 | |||
Patellofemoral K&L | Score 0 | 51 | 7 | <0.001 | |
Score 1 | 18 | 2 | |||
Score 1.5 | 0 | 2 | |||
Score 2 | 13 | 2 | |||
Score 3 | 0 | 23 | |||
Score 4 | 0 | 2 | |||
Patellofemoral JSN | Score 0 | 76 | 8 | <0.001 | |
Score 0.5 | 6 | 7 | |||
Score 1 | 0 | 23 | |||
Medial osteophytes | Score 0 | 26 | 23 | 0.049 | |
Score 0.5 | 18 | 6 | |||
Score 1 | 19 | 6 | |||
Score 1.5 | 11 | 1 | |||
Score 2 | 7 | 1 | |||
Score 3 | 1 | 1 | |||
Patellofemoral osteophytes | Score 0 | 40 | 13 | 0.227 | |
Score 0.5 | 13 | 11 | |||
Score 1 | 15 | 10 | |||
Score 1.5 | 7 | 3 | |||
Score 2 | 7 | 1 | |||
Score 3 | 0 | 0 |
. | . | Number of knees . | . | . | |
---|---|---|---|---|---|
Radiographic measurement . | . | Tibiofemoral group . | Patellofemoral group . | P (difference in distributions medial vs patellofemoral) . | |
Medial K&L | Score 0 | 9 | 27 | <0.001 | |
Score 1 | 1 | 5 | |||
Score 1.5 | 1 | 1 | |||
Score 2 | 3 | 5 | |||
Score 3 | 66 | 0 | |||
Score 4 | 2 | 0 | |||
Patellofemoral K&L | Score 0 | 51 | 7 | <0.001 | |
Score 1 | 18 | 2 | |||
Score 1.5 | 0 | 2 | |||
Score 2 | 13 | 2 | |||
Score 3 | 0 | 23 | |||
Score 4 | 0 | 2 | |||
Patellofemoral JSN | Score 0 | 76 | 8 | <0.001 | |
Score 0.5 | 6 | 7 | |||
Score 1 | 0 | 23 | |||
Medial osteophytes | Score 0 | 26 | 23 | 0.049 | |
Score 0.5 | 18 | 6 | |||
Score 1 | 19 | 6 | |||
Score 1.5 | 11 | 1 | |||
Score 2 | 7 | 1 | |||
Score 3 | 1 | 1 | |||
Patellofemoral osteophytes | Score 0 | 40 | 13 | 0.227 | |
Score 0.5 | 13 | 11 | |||
Score 1 | 15 | 10 | |||
Score 1.5 | 7 | 3 | |||
Score 2 | 7 | 1 | |||
Score 3 | 0 | 0 |
Radiographic data are for the worst knee at baseline. When the worst knee was missing the average of both knees was taken. P values for the nominal radiographic data (JSN, osteophytes and K&L) correspond to the χ2 test for equality of distributions.
. | . | Number of knees . | . | . | |
---|---|---|---|---|---|
Radiographic measurement . | . | Tibiofemoral group . | Patellofemoral group . | P (difference in distributions medial vs patellofemoral) . | |
Medial K&L | Score 0 | 9 | 27 | <0.001 | |
Score 1 | 1 | 5 | |||
Score 1.5 | 1 | 1 | |||
Score 2 | 3 | 5 | |||
Score 3 | 66 | 0 | |||
Score 4 | 2 | 0 | |||
Patellofemoral K&L | Score 0 | 51 | 7 | <0.001 | |
Score 1 | 18 | 2 | |||
Score 1.5 | 0 | 2 | |||
Score 2 | 13 | 2 | |||
Score 3 | 0 | 23 | |||
Score 4 | 0 | 2 | |||
Patellofemoral JSN | Score 0 | 76 | 8 | <0.001 | |
Score 0.5 | 6 | 7 | |||
Score 1 | 0 | 23 | |||
Medial osteophytes | Score 0 | 26 | 23 | 0.049 | |
Score 0.5 | 18 | 6 | |||
Score 1 | 19 | 6 | |||
Score 1.5 | 11 | 1 | |||
Score 2 | 7 | 1 | |||
Score 3 | 1 | 1 | |||
Patellofemoral osteophytes | Score 0 | 40 | 13 | 0.227 | |
Score 0.5 | 13 | 11 | |||
Score 1 | 15 | 10 | |||
Score 1.5 | 7 | 3 | |||
Score 2 | 7 | 1 | |||
Score 3 | 0 | 0 |
. | . | Number of knees . | . | . | |
---|---|---|---|---|---|
Radiographic measurement . | . | Tibiofemoral group . | Patellofemoral group . | P (difference in distributions medial vs patellofemoral) . | |
Medial K&L | Score 0 | 9 | 27 | <0.001 | |
Score 1 | 1 | 5 | |||
Score 1.5 | 1 | 1 | |||
Score 2 | 3 | 5 | |||
Score 3 | 66 | 0 | |||
Score 4 | 2 | 0 | |||
Patellofemoral K&L | Score 0 | 51 | 7 | <0.001 | |
Score 1 | 18 | 2 | |||
Score 1.5 | 0 | 2 | |||
Score 2 | 13 | 2 | |||
Score 3 | 0 | 23 | |||
Score 4 | 0 | 2 | |||
Patellofemoral JSN | Score 0 | 76 | 8 | <0.001 | |
Score 0.5 | 6 | 7 | |||
Score 1 | 0 | 23 | |||
Medial osteophytes | Score 0 | 26 | 23 | 0.049 | |
Score 0.5 | 18 | 6 | |||
Score 1 | 19 | 6 | |||
Score 1.5 | 11 | 1 | |||
Score 2 | 7 | 1 | |||
Score 3 | 1 | 1 | |||
Patellofemoral osteophytes | Score 0 | 40 | 13 | 0.227 | |
Score 0.5 | 13 | 11 | |||
Score 1 | 15 | 10 | |||
Score 1.5 | 7 | 3 | |||
Score 2 | 7 | 1 | |||
Score 3 | 0 | 0 |
Radiographic data are for the worst knee at baseline. When the worst knee was missing the average of both knees was taken. P values for the nominal radiographic data (JSN, osteophytes and K&L) correspond to the χ2 test for equality of distributions.
Severity of disease in each group
The WOMAC scores were the same in both groups (Table 1). The overall radiographic severity (Table 2) shows that there were no significant differences comparing the TFJ group medial K&L score and osteophyte grades with the PFJ group K&L score and osteophyte grades, suggesting that the severity in the two compartments was similar.
Serum biomarkers
The serum concentrations of the biomarkers representing turnover of cartilage, bone and synovial tissue are presented in Table 1. In both groups all biomarkers, except COMP, had similar mean values. Gender, age and BMI were found to be general confounders in the data and therefore all between-group comparisons in Table 1 have been adjusted for these variables. The only biomarker showing a significant difference between TFJ and PFJ OA groups, after allowing for gender, age and BMI, was COMP. The serum COMP concentrations adjusted for age and BMI in men and women in the two subsets of OA are shown in Fig. 1. The serum COMP levels in both male and female patients with TFJ OA was significantly higher compared with the PFJ group (P<0.05).
Longitudinal measurements of some of the biomarkers were also available for up to 5 yr; the data were therefore analysed and expressed as AUC per month for the whole of the study period. These data also show that serum GAG, KS, YKL-40, HA and CRP concentrations were similar in the two subsets of OA, but serum COMP was significantly higher in TFJ OA compared with PFJ OA. The mean [95% confidence interval (CI)] AUC/month serum COMP was 12.61 (12.00, 13.22) in the tibiofemoral group and 10.75 (10.03, 11.47) U/l in the patellofemoral group (P = 0.001). In order to determine whether there are any differences in the rate of loss of cartilage in the two subsets of OA, we compared the changes in the medial and patellofemoral joint space over time. The results show that there was no significant difference in the rate of cartilage loss (loss of JSW) over 5 yr in the two groups [0.05 (–0.01, 0.11) mm/yr in the TFJ vs 0.08 (–0.03, 0.18) mm/yr in the PFJ, P = 0.636]. However, as expected, a much greater proportion of the PFJ knees exhibited JSN in the patellofemoral compartment compared with the TFJ knees (71 vs 14%, P<0.001). Moreover, the medial compartment JSW in the TFJ group was significantly reduced compared with the PFJ group [2.44 mm (2.16, 2.72) vs 4.61 mm (4.31, 4.91), P<0.001] and the JSN in the patellofemoral compartment was more severe in the PFJ group than the TFJ group (P<0.001). Therefore, this knee-based analysis is consistent with the patient-based data presented in Table 1.
Discussion
These results show that serum COMP concentrations are higher in patients with TFJ OA than in patients with PFJ OA of similar severity. The two cohorts of patients described in this study were selected by a single process from the same population of patients with knee pain, which serves to delineate any differences in clinical, imaging (X-ray) and biochemical measures of disease activity in TFJ OA and PFJ OA. No significant differences in VAS or WOMAC (pain, stiffness and function) were found between the two subsets of OA. Moreover, the osteophyte scores and K&L grades were similar in respective compartments in the two groups, suggesting that the two subsets of OA had similar radiographic severity. Thus, the 82 patients with TFJ OA and the 38 with PFJ OA identified from the main cohort represent suitable subsets of patients for investigating biochemical differences in joint tissue metabolism.
The concentrations of OC and CTX-I were similar in the two subsets of OA, suggesting that there is little difference in bone metabolism between TFJ OA and PFJ OA. Similarly, the lack of any significant differences between serum levels of HA and CRP between the two groups suggests little difference in synovial activity at the time point measured. By contrast, the serum concentrations of COMP at entry and AUC/month were significantly higher in tibiofemoral disease compared with PFJ OA. These differences remain significant after the major confounding variables (age, gender and BMI) were allowed for in the analysis, suggesting that the increased COMP concentrations in the TFJ group are likely to be real and not an artefact.
Several independent studies have now demonstrated that serum COMP is a marker of cartilage loss [18, 19]. Although COMP is present in some other joint tissues, the concentrations are very low compared with cartilage and menisci [20]. Thus the data presented here may indicate that there is increased degradation and loss of cartilage and/or menisci in the TFJ group compared with the PFJ group. This finding is consistent with the demonstration that there is poor correlation between loss of cartilage in the patellofemoral compartment compared with the tibiofemoral compartment, while there is good positive correlation between loss of cartilage in the medial and lateral tibiofemoral compartments [1]. The increase in serum COMP in the TFJ subset could result merely from a larger volume of affected cartilage being present in the TFJ group as compared with the PFJ group [21]. If so, it would be expected that the other markers of cartilage metabolism used in this study (GAG, KS and YKL-40) would show some differences between TFJ and PFJ. But the concentrations of GAG, KS and YKL-40 were not significantly different in the two subsets of patients, suggesting that the differences in cartilage metabolism observed between the two groups of OA are not simply due to cartilage volume. Moreover, the rate of radiographic cartilage loss was similar in the two subsets and therefore an alternative explanation may be that there are some qualitative differences in cartilage metabolism between the two groups. The value of the cartilage markers, especially GAG, KS and YKL-40, as measures of cartilage damage/loss is questionable [22] and in several of our previous studies we have not found any clear association between serum levels of GAG, KS and YKL-40 and radiographic measures of cartilage loss [17]. In addition, other studies have found KS and YKL-40 to be better markers of systemic inflammation than cartilage turnover [23, 24]. Thus, until further measurements of more reliable markers of cartilage damage and/or loss are available, we cannot establish whether there are definite quantitative and qualitative differences in cartilage metabolism in the two subsets of OA.
The authors have declared no conflicts of interest.
References
Cicuttini F, Wluka A, Wang Y, Stuckey S. The determinants of change in patella cartilage volume in osteoarthritic knees.
Claessens AA, Schouten JS, van den Ouweland FA, Valkenburg HA. Do clinical findings associate with radiographic osteoarthritis of the knee?
Davis MA, Ettinger WH, Neuhaus JM, Mallon KP. Knee osteoarthritis and physical functioning: evidence from the NHANES I Epidemiologic Followup Study.
McAlindon TE, Cooper C, Kirwan JR, Dieppe PA. Determinants of disability in osteoarthritis of the knee.
Baker KR, Xu L, Zhang Y et al. Quadriceps weakness and its relationship to tibiofemoral and patellofemoral knee osteoarthritis in Chinese: the Beijing osteoarthritis study.
Cooper C, McAlindon T, Snow S et al. Mechanical and constitutional risk factors for symptomatic knee osteoarthritis: differences between medial tibiofemoral and patellofemoral disease.
Cicuttini FM, Spector T, Baker J. Risk factors for osteoarthritis in the tibiofemoral and patellofemoral joints of the knee.
Sharif M, Kirwan JR, Elson CJ, Granell R, Clarke S. Suggestion of nonlinear or phasic progression of knee osteoarthritis based on measurements of serum cartilage oligomeric matrix protein levels over five years.
Bellamy N, Buchanan WW, Goldsmith CH, Campbell J, Stitt LW. Validation study of WOMAC: a health status instrument for measuring clinically important patient relevant outcomes to antirheumatic drug therapy in patients with osteoarthritis of the hip or knee.
Cooper C, Cushnaghan J, Kirwan JR et al. Radiographic assessment of the knee joint in osteoarthritis.
Kellgren JH, Lawrence JS. Radiological assessment of osteo-arthrosis.
Farndale RW, Buttle DJ, Barrett AJ. Improved quantitation and discrimination of sulphated glycosaminoglycans by use of dimethylmethylene blue.
Creamer P, Sharif M, George E et al. Intra-articular hyaluronic acid in osteoarthritis of the knee: an investigation into mechanisms of action.
Harvey S, Weisman M, O’Dell J et al. Chondrex: new marker of joint disease.
Sharif M, Shepstone L, Elson CJ, Dieppe PA, Kirwan JR. Increased serum C reactive protein may reflect events that precede radiographic progression in osteoarthritis of the knee.
Goldberg RL. Enzyme-linked immunosorbent assay for hyaluronate using cartilage proteoglycan and an antibody to keratan sulfate.
Sharif M, George E, Shepstone L et al. Serum hyaluronic acid level as a predictor of disease progression in osteoarthritis of the knee.
Saxne T, Heinegard D. Cartilage oligomeric matrix protein: a novel marker of cartilage turnover detectable in synovial fluid and blood.
Clark AG, Jordan JM, Vilim V et al. Serum cartilage oligomeric matrix protein reflects osteoarthritis presence and severity: the Johnston County Osteoarthritis Project.
Neidhart M, Hauser N, Paulsson M, DiCesare PE, Michel BA, Hauselmann HJ. Small fragments of cartilage oligomeric matrix protein in synovial fluid and serum as markers for cartilage degradation.
Ding C, Garnero P, Cicuttini F, Scott F, Cooley H, Jones G. Knee cartilage defects: association with early radiographic osteoarthritis, decreased cartilage volume, increased joint surface area and type II collagen breakdown.
Fawthrop F, Yaqub R, Belcher C, Bayliss M, Ledingham J, Doherty M. Chondroitin and keratan sulphate epitopes, glycosaminoglycans, and hyaluronan in progressive versus non-progressive osteoarthritis.
Spector TD, Woodward L, Hall GM et al. Keratan sulphate in rheumatoid arthritis, osteoarthritis, and inflammatory diseases.
Author notes
Department of Anatomy, University of Bristol and 1Academic Rheumatology Unit, Bristol Royal Infirmary, Bristol, UK, 2Department of Rheumatology, University of Copenhagen, Copenhagen, Denmark and 3Department of Pathology and Microbiology, University of Bristol, UK.
- diagnostic radiologic examination
- area under curve
- biological markers
- cartilage
- collagen type i
- epitopes
- glycosaminoglycans
- keratan sulfate
- knee region
- osteoarthritis
- knee osteoarthritis
- c-reactive protein
- osteocalcin
- knee pain
- visual analogue pain scale
- serum specimen
- ontario
- pain score
- cartilage metabolism
- severity of illness
Comments