Circulating surgery-induced gene signatures are associated with recovery from total knee arthroplasty for knee osteoarthritis

Abstract

Purpose: Osteoarthritis (OA) is a major cause of pain and disability. Local and peripheral inflammation are known to contribute to the disease activity, joint stiffness and pain. Local inflammation (defined as synovitis) correlates with changes in pain severity, and the degree of synovitis is also associated with abnormal pain processing and sensitization in patients with knee OA. Peripheral immune signatures have been associated with joint inflammation, disease activity and pain, and dysregulated single-cell immune signatures have been correlated with unsatisfactory recovery in patients after joint replacement surgery. However, the disease- and surgery-specific signatures dysregulated in OA patients that develop stiffness after total knee arthroplasty (TKA) are not well understood. As a consequence, we are limited in our ability of identifying patients at risk of developing poor surgical outcomes early after TKA. In this study we integrated clinical outcomes with transcriptomics analyses in blood samples from OA patients undergoing TKA, aiming to identify surgery-induced gene signatures that are associated with recovery from surgery. Methods: Patients: we enrolled 179 patients with idiopathic end-stage OA scheduled for TKA, with IRB approval and patient consent. Seventeen patients were excluded from analysis after enrollment, leaving 162 patients for analysis of clinical data and biological samples. We collected demographics (including age, sex, BMI, and ethnicity) at baseline, and range-of-motion (ROM) at baseline and 6 weeks after surgery. For sample selection for RNA-seq analyses, patients were divided onto 2 groups before sequencing: (1) cases: patients who were found to have postoperative stiffness at 6 weeks, and (2) age-, sex-, race- and BMI-matched controls: patients without post-TKA stiffness at 6 weeks. Stiffness was defined as ≤95° ROM measured by goniometer at 6 weeks (±2 weeks). Sample collection: We collected PAXgene RNA tubes from all patients at the time of surgery (DOS) and at 24 hours after surgery (POD1). We also collected heparinized venous whole blood for peripheral blood mononuclear cell (PBMC) isolation from 6 consecutive patients at DOS and POD1. PBMC isolation: PBMCs were isolated by centrifugation using Ficoll-Paque PLUS (GE Healthcare) density gradient according to manufacturer’s instructions. RNA from isolated PBMCs was extracted using the RNeasy mini kit. PAXgene RNA isolation and globin depletion: RNA was isolated from frozen PAXgene RNA tubes using the PAXgene Blood RNA System, following the manufacturer’s instruction. After isolation, samples were globin-depleted using the GLOBINclear Kit. NanoString gene expression analyses: A total of 100 ng of total RNA obtained from PBMCs isolated at DOS and POD1 from 6 patients was used for NanoString analyses using the human Immunology Panel and analyzed using the accompanying nSolver 4.0 software. RNA-seq analyses: A total of 100ng of globin-depleted RNA isolated from PAXgene tubes collected at DOS and POD1 from cases (n=9) and age-, sex,- race- and BMI-matched controls (n=9) were used for RNA-seq, performed using an Illumina HiSeq 4000 at the Genomics Resources Core Facility of Weill Cornell Medicine. After sequencing the reads were processed using a dedicated RNAseq pipeline developed by bioinformaticians at the David Z. Rosensweig Genomics Research Center at Hospital for Special Surgery. A q value < 0.05 was considered significant. Results: NanoString analyses of RNA isolated from PBMCs purified from whole blood collected at DOS and at POD1 identified significant changes in gene expression following total knee arthroplasty in our patient population, including increased expression of IL1B mRNA and enrichment in NF-kB signaling pathways at 24h after surgery. We also uncovered a relative enrichment in macrophage and neutrophil signatures accompanied by depletion of T-cell signatures following surgery. RNAseq analyses in RNA from whole blood collected at DOS and POD1 from controls and cases using PAXgene Blood RNA tubes confirmed the surgery-specific signature, with enrichment of Toll-like receptor and IL1 signaling pathways at 24 h after surgery, consistent with the NanoString analyses in PBMCs. Comparison of the responses to surgery in cases and controls also uncovered differences in gene signatures in patients that developed stiffness after surgery, including changes in the MAPK pathway. Conclusions: Together, our results show that peripheral gene signatures can be used to evaluate pathways involved in the responses to surgery, and that patients that develop stiffness after TKA may have dysregulated responses. We believe that these dysregulated surgery-induced signatures can be used to predict patients at risk of developing complications following surgery, and to develop targeted preventative therapies to avoid the development of knee stiffness following TKA.