The usefulness of SF and/or serum-soluble biomarkers in the assessment of treatment efficacy or for monitoring disease progression in RA and OA remains controversial. The goal of the present study was to characterize and compare the concentrations of selected biological mediators (both catabolic and anabolic) in the SF obtained from RA or OA patients in comparison to SF aspirated from the knee joints of asymptomatic organ donors and to investigate whether any of these molecules might have utility as prognostic markers. Markers of catabolism included IL-1β and the IL-6 family of chemokines (IL-6, IL-8, IL-11 and LIF), which have been shown in cartilage to negatively correlate with OP-1 [12
], the anabolic marker evaluated here. In addition, COMP and osteocalcin were included as markers of matrix metabolism. Our results suggest that the concentrations of the pathophysiologically important biomarkers in SF are different between OA and RA and depend on the mechanisms that drive cellular responses in each disease. IL-11, LIF and OP-1 appear to be significant for OA processes, while IL-1, IL-6, IL-8 and OP-1 may play an important role in RA.
In line with other studies [17
], our data also indicate that RA, but not OA, is characterized by elevated IL-6 and IL-8 concentrations. The role of IL-6 in OA is unclear, and there are conflicting findings in the literature that indicate both procatabolic [19
] and anticatabolic effects of IL-6 in chondrocytes and synoviocytes [21
]. In contrast, the role of IL-6 in inflammatory processes, including RA, appears to be more consistent and involves promotion of the disease by stimulating B and T cells [24
]. Furthermore, the concentrations of IL-6 in SF were shown to positively correlate with those in the sera of patients with RA, OA, crystal deposition and other forms of inflammatory arthritis [26
]. As expected, IL-6 concentrations in biological fluids of asymptomatic organ donors were at the baseline concentrations and were significantly lower than those in patients with RA or OA. Similar to IL-6, IL-11 concentrations have been found to be significantly higher in SF than in serum, though the concentrations are highly correlated. In the present study, IL-11 concentrations were 1.4-fold higher in OA samples than those in asymptomatic organ donor samples, but not higher than those in the RA group. Trontzas et al.
] reported that SF IL-11 concentrations are higher in OA than in treated RA, but not in untreated RA. As we did not distinguish treated from untreated RA, we were not able to confirm this relationship. Although elevated concentrations of IL-11 have been found in RA, limited data are available on this cytokine [27
]. Still, our results suggest that further studies of the potential utility of IL-11 as a biomarker are warranted.
LIF, a cytokine in the IL-6 family, is downregulated by OP-1 and plays a role in bone formation and resorption. It has not been well-studied as a potential biomarker, and its association with OA has been based mainly on gene expression studies in synoviocytes [28
]. In the present study, LIF concentrations were significantly lower in both the OA and RA groups compared to the organ donors. This finding differs from that in a previous report that detected elevated SF LIF concentrations in some patients with severe RA [29
]. This discrepancy may be attributed to an inhibitory effect of other cytokines on LIF, for example, IL-4 [30
]. Nonetheless, the difference in LIF response in OA relative to the other members of the IL-6 family (IL-8 and IL-11) suggests that LIF is either involved at different stages of the diseases or has a distinct function.
The cytokine that has received the most attention among the arthritic diseases is IL-1β, yet as a biomarker, it has been studied more in either experimental OA [31
] or SF of patients with RA [32
]. In OA, it is primarily viewed as a mediator of degenerative processes in human joint tissues [33
], and substantial knowledge has been accumulated regarding its expression in cartilage and synovium, the mechanisms of its activation and interactions, its signaling, its regulation of and relationship with other active molecules, and so on. IL-1β, together with IL-6, has been shown in OA synovium to contribute to the progression of the disease by enhancing the susceptibility of chondrocytes to stimulation with proanabolic mediators [37
]. In posttraumatic OA, especially in acute phase responses, IL-1β together with tumor necrosis factor α and IL-6 are well-established regulators of cartilage degradation and resorption [1
]. As anticipated, in the current study, IL-1β and IL-6 concentrations were greater in RA patients than in OA patients or asymptomatic organ donors. However, to our surprise, the concentrations of IL-1β in SF of asymptomatic organ donors were statistically higher than those in OA, suggesting that IL-1β is involved only during the acute phase of the disease or is needed to initiate or trigger catabolic events. It is also a possibility that OA patients enrolled in our study underwent pharmacological treatment that had an inhibitory effect on IL-1β production or that only a subpopulation of patients with OA may have elevated IL-1β. The latter hypothesis is supported by a recent publication by Neu et al.
], who reported elevated concentrations of IL-1β in only a few OA samples, while in other samples IL-1β either was barely detectable or was below the detection limit. Though in previous publications IL-1β and IL-6 have been shown to be predictive of either OA or RA, our data indicate a closer association of both cytokines with RA than with OA.
As markers of matrix metabolism, we used COMP and osteocalcin. COMP is an extracellular glycoprotein and is a member of the thrombospondin family of calcium-binding proteins. COMP is associated with cartilage breakdown and has been studied as a potential diagnostic and prognostic indicator as well as a marker of disease severity or the efficacy of treatment (reviewed in [39
]). It has been reported that COMP concentrations in SF are 10 times higher than in serum and that higher COMP concentrations have been observed in patients with higher radiographic Kellgren-Lawrence grades. However, despite these expectations, here we were not able to identify an association of COMP concentrations with the type of disease or its severity, perhaps because of the limitations of our study. Samples were collected at only one time point rather than longitudinally, there was a lack of untreated controls and/or the sample size of each group was not large enough. As with COMP, we did not find differences in osteocalcin concentrations between the experimental groups, though elevation of osteocalcin has been detected previously in the destructive form of OA in comparison with nondestructive OA [40
]. In agreement with our data, the observations of Salisbury et al.
] suggested that in a predominantly older female population, the rate of normal bone turnover measured by osteocalcin in donors was not significantly different from that of OA or RA patients. Furthermore, it has been reported that OA and RA patients treated with nonsteroidal anti-inflammatory drugs showed significantly lower concentrations of SF osteocalcin than patients treated with glucocorticoids [42
]. Conflicting data on both COMP and osteocalcin indicate that only carefully designed longitudinal studies with well-controlled, large patient cohorts may shed the light on their potential as biomarkers.
Previously, we described in detail OP-1/BMP-7 in SF from organ donors or OA and RA patients [9
]. In this study, it was used primarily for correlation with other catabolic markers or because there are fewer anabolic than catabolic biomarkers, where OP-1/BMP-7 definitely belongs to the former category. As earlier, we confirmed the elevated concentrations of OP-1/BMP-7 released into the SF of RA patients in comparison to OA patients and organ donors. A higher quantity of OP-1/BMP-7 in samples characterized by higher concentrations of proinflammatory mediators may not necessarily indicate a higher synthesis of this growth factor. Our unpublished data suggest that treatment with IL-1β, for example, induce activation of pro-OP-1 and thus release of active OP-1 from the matrix. In addition, catabolic mediators lead to matrix loosening/degradation, which also may favor activation and/or release of the growth factor that has been trapped within the matrix or bound to the extracellular binding proteins or matrix components as it occurs with the transforming growth factor β latency protein [43
Biomarkers were also assessed on the basis of the activity of the disease within each patient cohort. OA was assessed on the basis of the WOMAC index score, and RA was evaluated on the basis of the synovial WBC count, ESR and CRP level. Though no significant differences between biomarkers and disease activity were found, there was a trend toward an elevation of proinflammatory mediators in the active state of OA or RA.