Mylène Jansen

336 Chapter 16 We also compared the change in analyte levels over the 6 weeks of distraction in those making the MCID of 10 points or more by KOOS 4 (responders) with those who did not improve by this amount (non-responders). The clinical response to joint distraction was most pronounced at 6 months, with 11/15 (73%) of individuals with available data reaching a MCID on KOOS 4 . Responders at 6 months had a greater increase in TGF β -1 and FGF-2 during the distraction period than non-responders (Figure 4B, Supplementary Table 4 and Supplementary Figure 2). Similar analyte changes were also seen in responders and non-responders at 3 months, when TIMP-1 levels were also different between the 2 groups (ES 497 ng/ml, p= 0.02, Supplementary Figure 3). Discussion Easily detectable, substantial changes in levels of 8 putative mechanosensitive molecules of the inflammatory response (activin A, LTBP2, TGF β -1, FGF-2, TIMP-1, IL-6, MCP-1, and IL- 8) were seen in SF over the period of KJD. There were also associations between several of these molecules over time. These changes would not appear to be due to SF volume change because whilst some analytes increase, others stay the same or even decrease. Of the regulated molecules, whilst IL-6 and MCP-1 (also known as CCL-2) have been associated with degeneration or pain in the osteoarthritic joint 34,35 , FGF-2 and TGF β -1 are more typically associated with repair. 27,36 It is perhaps not surprising that a mechanically-induced inflammatory response should include both catabolic and reparative processes. But that an intervention which apparently leads to net articular cartilage repair involves the activation of traditionally inflammatory pathways would go against current convention. Overall there was substantial variation between individuals for certain molecules in the extent and sometimes direction of this response. This supports the notion that an individual’s biological response to the intervention could vary and be related to their clinical response. On the other hand, some molecules like TGF β -1 and activin A showed very consistent directional changes following KJD. Our proof-of-concept study appears to suggest an association between this measurable biological response to joint distraction and subsequent clinical outcome. The clinical response to joint distraction was most pronounced at 6 months. Several of the associations between change in analytes and KOOS 4 at 6 months were also apparent at 3 and 12 months, and when individuals were stratified, either by their molecular response or their clinical response. This supports that elements of this biological response to distraction appeared to be associated with a clinically meaningful response: for example, FGF-2 and TGF β -1, typically associated with cartilage anabolism/anticatabolism, were raised in responders. 37,38 One molecule, activin A, strikingly fell in all individuals to what we estimate are normal levels in human SF. Activin A is produced by osteoarthritic and injured articular cartilage and