Abstract
Objectives:
The purpose of this study was to determine the longitudinal concentrations of several synovial fluid biomarkers after cartilage transplantation and potential correlations with patient reported outcomes.
Methods:
After institutional IRB approval, patients undergoing osteochondral allograft (OCA) transplantation or autochondrocyte implantation (ACI) were consented and enrolled into a prospective study involving intraoperative synovial fluid aspiration of the operative knee, as well as postoperative aspirations of the operative knee at standardized clinic follow-up timepoints: 2 weeks, 6 weeks, 6 months, and 1 year postoperatively. For this analysis, patients undergoing cartilage transplantation (OCA or ACI) with preoperative aspiration and at least one successful postoperative aspiration were included. Exclusion criteria for aspirations from the analysis included subsequent surgical procedure or articular injections on the index knee within 1 year of the transplant.
The synovial fluid samples were analyzed for the following inflammatory biomarkers with Multiplex ELISA: CCL5, MMP-1, EGF, VEGF, IL1a, FGF2, CCL2, BMP2, and Aggrecan (ACAN). The samples were digested 1:1 with 3mg/mL hyaluronidase prior to execution of the assay in accordance with the manufacturer’s protocol. A minimum of 3uL was required for analysis.
Patients prospectively completed Knee Injury and Osteoarthritis Outcome Score (KOOS), including Activities of Daily Living (ADL), Pain, Quality of Life (QoL), Sports, Symptoms, and Joint Replacement (JR) subscores, preoperatively and at 6 months and 1 year postoperatively.
All statistical analyses were performed on RStudio and STATA. Non-parametric tests were utilized to avoid assumptions of normality. Skillings-Mack tests were utilized as a non-parametric alternative for a repeated-measures ANOVA in the setting of sporadic missing values (i.e., dry aspirations). Post-hoc pairwise Wilcoxon ranksum tests with Bonferroni corrections for multiple comparisons were performed for significant tests. Univariate linear regressions were performed to determine correlations between cytokine changes and patient-reported outcomes. A minimum of 10 observations per predictor was considered sufficient for the analysis. All testing was two-sided, and significance was set at p<0.05.
Results:
A total of 63 aspirations from 19 patients were evaluated. Postoperatively, successful aspirations were able to be performed on 13 patients at 2 weeks (68%), 17 patients at 6 weeks (89%), 9 patients at 6 months (47%), and 5 patients at 1 year (26%) (Table 1). One patient underwent a subsequent procedure on the index knee (lysis of adhesions) 10 weeks after cartilage transplantation, thereby excluding their 6 month and 1-year aspirations from analysis.
Across all timepoints, there were significant differences in MMP-1 (p=0.003), ACAN (p<0.001), and FGF2 (p=0.026). Specifically, ACAN was significantly elevated at 2 (p<0.001) and 6 weeks (p<0.001), and MMP- 1 was the greatest at 6 weeks relative to transplant (p<0.001), 6 months (p=0.01), and 1 year (p=0.04). In contrast, FGF2 was significantly lower at 2 weeks compared to transplant levels (p=0.02). There were no significant differences across timepoints for CCL5 (p=0.88), VEGF (p=0.068), CCL2 (p=0.071), BMP2 (p=0.067), IL-1a (p=0.13), or EGF (p=0.01, p>0.05 for all pairwise comparisons after Bonferroni correction) (Figure 1).
A subanalysis was performed for cytokines which demonstrated at least 5 “pro-inflammatory” responses (positive fold-change) and at least 5 “anti-inflammatory” responses (negative fold-change), with a minimum of 10 observations with associated patient-reported outcomes at 6 and/or 12 months postoperatively. From the 2-week aspiration samples, CCL5 (5 increase/7 decrease), IL-1a (5/7), and CCL2 (6/7) met the inclusion criteria. Increase in IL-1a demonstrated significant linear associations with worse postoperative KOOS ADL (p=0.048), Pain (p=0.01), Quality of Life (p=0.01), and Symptoms (p=0.01) scores at 6 months (Table 2, Figure 2). Similarly, an increase in CCL-2 postoperatively was associated with worse KOOS ADL (p=0.04) and Pain (p=0.04) scores at 6 months. No significant linear associations between cytokine changes at 6 weeks and later PROMs were identified.
Conclusions:
Following cartilage transplantation of the knee, the synovial fluid profile may demonstrate early postoperative increases in MMP-1 and ACAN and a decrease in FGF-2. An early pro-inflammatory response with IL-1a within the first two postoperative weeks may be linearly associated with worse short-term patient-reported outcomes compared to an early decrease.
