The key role of proinflammatory cytokines, matrix proteins, RANKL/OPG and Wnt/β-catenin in bone healing of hip arthroplasty patients

Abstract

IntroductionWe still lack understanding of why some implants fail while most remain stable after decades of use. Proinflammatory cytokines, matrix proteins and bone regulating cytokines of the RANKL/OPG (receptor activator of nuclear factor kappa B ligand/osteoprotegerin) and Wnt/β-catenin pathways are mandatory for normal bone repair but their spatial and temporal role in the healing of primary total hip arthroplasties (THA) has not been previously shown. Materials and methodsTwenty-four osteoarthritis patients with one-sided well-fixed primary THA were prospectively monitored during 18years (18Y) with repeated blood samples, clinical variables and radiographs. Eighty-one healthy donors divided in three age- and gender-matched groups and twenty osteoarthritis patients awaiting THA and serving as control of the validity of stored plasma in THA patients, were included. Plasma was analyzed for C-reactive protein (CRP), interleukin (IL)-6, IL-8, IL-1β, tumor necrosis factor (TNF)-α, osteopontin (OPN), secreted protein acidic and rich in cysteine (SPARC/osteonectin), osteocalcin (OC), bone specific alkaline phosphatase (BALP), N-terminal propeptide of collagen type I (P1NP), RANKL, OPG, the Wnt agonistic ligands (Wnt)-1 and Wnt-3a, and the Wnt antagonists sclerostin, Dickkopf (Dkk)-1, Dkk-3, Dkk-4, secreted frizzled related protein (sFRP)-1, sFRP-3 and Wnt inhibitory factor-1 (Wif-1). ResultsInflammatory mediators in arthroplasty patients (CRP, IL-6, OPN) increased significantly on day one after surgery vs preoperative value (PR) and healthy subjects and returned to baseline at 6W. TNF-α did not change relative preoperative level or healthy subjects. SPARC and OC increased in a biphasic fashion with the primary phase beginning shortly after surgery and lasting 3M (SPARC) and 2Y (OC) while the secondary phase peaked at 1Y (SPARC) and 13Y (OC), with both returning to basal level at 15Y. BALP peaked at 3M after surgery with a return to basal level at 2Y followed by a continuous increase from 5Y until 18Y. P1NP increased immediately after surgery and returned to basal level at 6W followed by a new peak at 10Y returning to basal at 13Y. IL-8 and IL-1β peaked at 5Y post-THA and returned to basal level at 10Y. RANKL/OPG and Wnt/β-catenin remained at preoperative levels until 5Y post-THA when a sustained increase in OPG level, paralleled by a sustained decrease in sclerostin, started and lasted until 18Y. Despite a strong increase by RANKL at 13Y, the OPG/RANKL-ratio remained high between 5Y and 18Y. Dkk-1 and sFRP-1 remained at basal level until 5Y followed by a peak at 7Y and a return to basal level at 15Y. Similarly, RANKL increased after 5Y, peaked at 13Y and returned to basal levels at 18Y, thus coinciding with Wnt-1. In contrast, Wnt3a, Dkk-3, Dkk-4, sFRP-3 and Wif-1 did not differ from preoperative levels or healthy subjects during the course of the follow-up. ConclusionThe primary peak of proinflammatory cytokines involved in the initiation of bone healing after trauma is in line with previous results. The primary phase of increased matrix proteins, P1NP and BALP paralleled by RANKL, OPG and Wnt/β-catenin remaining at preoperative level until 5Y, support a strong formation of mineralized matrix and to a lesser degree bone during this phase. The secondary proinflammatory peak at 5Y is likely a trigger of coupled bone remodeling and neosynthesis as it is followed by increased levels of the bone anabolic turnover marker, BALP, and mediators of the RANKL/OPG and Wnt/β-catenin pathways. A continuous increase by OPG level and the bone turnover marker, BALP, lasting from 5Y until 18Y and paralleled by a similar decrease in sclerostin level support their being key regulators of bone anabolism, whereas the transient and opposed activities of RANKL, Wnt-1, Dkk-1 and sFRP-1 serve as fine tuning tools during the coupled remodeling phase.