Only full-thickness OA cartilage without any obvious macroscopic degenerative changes was included in this study to avoid comparison between full depth reference and partly damaged or thinner OA cartilage. As expected from macroscopic appearance, OA samples had a relatively low Mankin score. A Mankin grade assigned within the range 0–6 is accounted as “early phase of degenerative changes”
. However, in these samples cartilage oedema was present indicating that higher water content is evident at a very early stage of the disease. A strong relationship between cartilage oedema and collagen degradation has also been shown by Basser et al.
. That study suggests that the increased water content in OA cartilage is caused by loss of collagen network integrity, which in turn leads to decreased tensile stiffness and increased water imbibition, as proposed also by Maroudas et al.
. Experimental OA studies in rabbits (meniscectomy) and dogs (anterior cruciate ligament section model) have also shown increased cartilage swelling at early stages of OA
[28, 29]. The lack of relationship between water content and GAG content/dry weight in the present study confirms previous findings that swelling is not influenced by the amount of GAG
Extractable collagen seems to be related to cartilage disease as evidenced by the positive relationship between extractable collagen and water content and that more collagen is extracted from OA than from reference cartilage (Figures
2). Reduced collagen content per wet weight in OA cartilage compared to non-OA cartilage, suggests inferior cartilage properties in OA (Figure
2) and points at the importance of protecting the collagen fibrillar network. An increased extractable collagen pool, which is inversely correlated to collagen content, has previously also been identified in knee OA cartilage
[18, 23]. Taken together, this suggests a common degradative pathway of collagen in articular cartilage of different joints. Studies showing similar findings in tendon and disc diseases support that this pathway is general in connective tissue
A comparison of collagen content per dry and wet weights did not show differences between OA and reference groups (Table
1). This suggests either a loss of small amounts of collagen or a potential for chondrocytes to synthesis and deposit collagen in OA cartilage
. However, there is limited evidence that chondrocytes can recapitulate the overall collagen architecture if mature cartilage is damaged by injury or degeneration. In contrast, GAG content per dry weight was higher in OA samples than in reference samples (P = 0.012) whereas when measured by wet weight values were similar (P = 0.644). These differences draw attention to the importance of relating total molecular contents both to dry weight and to wet weight to provide maximum information on molecular content. This also suggests that cartilage may have a capacity to replace GAG; at least until advanced degradation occurs (our samples had an average Mankin score of 4). This repair capacity is known in the literature as hypertrophic repair
[33–35] and has been visualized by radiography and by MRI
[36, 37]. Increments in GAG (mainly aggrecan) content have also been seen in studies of early cartilage damage
. Hypothetically, the lower amount of GAG in the relatively older reference group may to some extent be explained by decreased physical activity, as is suggested in a contrast-MRI study
Several studies suggest age as a main risk factor for OA. Indeed, age is related to stiffer collagen, lower water content and malfunctioning chondrocytes in cartilage, all of which may predispose tissue damage
[40–42]. However, overt cartilage loss is not a major feature of aging
. It is also apparent from experimental studies that matrix changes in OA cartilage are different from those in aged cartilage
[44, 45]. Albeit based on a limited number of cartilage samples, the present study suggests that OA matrix changes were disease-related rather than age-related. An alternative explanation to a cause-relationship between age and OA could be accumulation of micro damage due to exposure of joint load over a lifetime period.
There are some limitations to this study. Inherent in the use of hip fracture patients as controls, these were older than the patients who had OA (P ≤ 0.001) (Table
1). However, the lack of correlation between age and molecular OA changes in the present study supports the use of femoral neck fracture patients, regardless of age, as a control group in these types of studies. Regarding the validity of the reference group, it has previously been shown that cartilage from patients with femoral neck fracture is very similar to that of normal controls
Cartilage was not sampled from the same position within the hip joint for all the samples. However, cartilage sampled from visibly intact full thickness regions from both groups make comparisons more valid. In support, Maroudas et al. have not found differences in GAG content in different locations of hip cartilage
The hydroxyproline assay used to quantify collagen does not distinguish between different types of collagen, which may explain the lack of correlation between immunostaining of type II collagen and percentage extractable collagen in this study. We are in the process of conducting further studies to explore the origin of the collagen in the chymotrypsin extracted cartilage.