Treatment-Monitoring Capabilities of Serum and Urine Biomarkers for Meniscal Allograft Transplantation in a Preclinical Canine Model

Background: Meniscal allograft transplantation (MAT) has been developed as a treatment for meniscal deficiency. Despite promising outcomes, there are no real-time methods to evaluate graft survivorship and predict functional outcomes. Hypothesis: Assessment of serum and urine biomarkers could be used to develop biomarker panels-prognostic (1- and 3-month postsurgical time points) and diagnostic (6-month time point)-based on strong associations with clinically relevant outcome metrics obtained 6 months after surgery. Study Design: Descriptive laboratory study. Methods: Twelve adult purpose-bred research hounds were included and underwent medial meniscal release to induce meniscal deficiency. Three months after meniscal release surgery, medial menisci were replaced with fresh-frozen meniscus (n = 4), fresh meniscus (n = 4), or fresh meniscotibial osteochondral allograft (n = 4) such that a spectrum of pain and functional outcomes could be anticipated. Serum and urine from all dogs were collected preoperatively and at 1, 3, and 6 months after MAT surgery. Dogs were assessed for pain-related and functional outcomes at the same time points. To develop a prognostic panel of biomarkers, biomarker data from the 1- and 3-month post-MAT surgery time points were used to model 6-month clinical outcomes. A diagnostic panel of biomarkers was developed using data from the 6-month post-MAT surgery to model 6-month clinical outcomes. Primary outcomes for pain and function were visual analog scale (VAS) and operated limb percentage total pressure index (%TPI), respectively. Using random subject effects, linear mixed models were used to develop prognostic biomarker panels, and linear fixed-effect models were used to develop diagnostic biomarker panels, with variance explained for each panel reported (R-2) along with individual biomarker relationships. Results: Across prognostic biomarker panels, a panel including serum IL-6, IL-8, IL-10, and IL-18 was fit for the primary functional outcome, operated limb %TPI (R-2 = 0.450), whereas a panel including serum CTX-II and OPG was fit for the primary pain-related outcome, VAS (R-2 = 0.516). Across diagnostic biomarker panels, a panel including serum MMP-1 and MMP-3 and urine PINP and TIMP-1 was fit for %TPI (R-2 = 0.863). Separately, a panel including urine CTX-I, CTX-II, IL-8, MMP-2, and TIMP-1 was fit as diagnostic biomarkers for the VAS for pain (R-2 = 0.438). Conclusion: Biomarker panels of selected serum and/or urine proteins can model clinically relevant metrics for function and pain in a preclinical model of MAT.