Mylène Jansen

328 Chapter 16 Participant biological samples A maximum of 2 ml of SF was aspirated by needle from the index knee at baseline visit (whilst participant under anesthesia and prior to the distraction frame being fitted), subsequently at midpoint of distraction (3–4 weeks, under local anesthesia) and at endpoint of distraction (at 6–7 weeks, immediately after the distraction frame was removed under anesthesia; Figure 1A). Within 2 h, all samples were centrifuged for 20 min at 3000G. Supernatants were stored in 200 μl aliquots in cryovials at -80°C in monitored freezers. Comparator ranges Normal : These were calculated in previously collected SF from patients undergoing amputation for treatment of lower limb tumor, at Royal National Orthopaedic Hospital (Stanmore), London, UK, or transplant donation, at Charing Cross Hospital, London, UK (REC 09/ H0710/60), who had macroscopically normal knee articular cartilage at the time of surgery and no evidence of arthritis or tumor invasion into the joint. 23 OA : These were calculated from measurements in SF from research tissue bank samples of patients with a confirmed diagnosis of OA undergoing partial or total joint replacement surgery at the Nuffield Orthopaedic Centre, Oxford, UK (REC 09/H0606/11 + 5). SF had been processed and stored as above. Reagents General laboratory reagents were the best available grade from either Sigma–Aldrich (Dorset, UK) or BDH (Dorset, UK) unless otherwise stated. MesoScale Discovery (MSD) plates and MSD SULFO-TAG labeled Streptavidin (#R32AD-5) were fromMSD (Rockville, MD, USA). Enzyme-linked immunosorbent assays (ELISAs) were from commercial providers (Table 1). Assays Assays were conducted for 10 predefined candidate molecules listed in Table I. All assays were carried out as per manufacturers’ instructions unless stated otherwise. Each assay had either previously undergone validation by us 23 or else underwent structured performance assessment and optimization for SF for this project, and all also passed quality performance requirements during sample reads (Table I). ELISA plates were read using Berthold Mithras LB940 reader andMSD plates byMSDQuickPlex SQ120 reader (analyzed withMSDDiscoveryWorkbench software v4.0.12). For TSG-6, each plate well (MSD, Rockville, USA, L15XA) was custom- coated with 30 μl 10 μg/ml TSG-6 capture antibody (Merck, MABT108) in phosphate- buffered saline (PBS) overnight at 4°C. Methods were then as described 23 . Mean concentrations of analytes were calculated from duplicate assay reads for each participant for each time point. Inter- and intra-assay coefficients of variation (CVs) were calculated for all assays. The lower limit of quantitation (LLOQ) was calculated for all assays. Where a measurement was below LLOQ, 50% of this value was used. 22,23 Lower and upper limits were also calculated for all assays for normal ranges using the geometric mean ± 2 standard deviations (SDs).