Tendon cells isolated from patients with persistent shoulder tendinopathy show dysregulated resolution responses

Abstract

Rotator cuff tendinopathy is a common cause of pain and disability and a substantial social and economic burden affecting athletic and aging individuals.1 Patients refractory to physical therapy have increasingly been treated with arthroscopic sub‐acromial decompression (ASAD), although the benefit to patients has recently been challenged by evidence from randomized placebo‐controlled surgical trials.2 The mechanisms governing whether inflammation persists or resolves in patients with supraspinatus tendinopathy in the absence of a tendon tear are poorly understood. We previously identified increased expression of genes associated with resolving inflammation and tissue repair in tissue biopsies collected from pain‐free compared to persistently painful patients after ASAD.3 However, the biological mechanisms governing persistent pain and inflammation in patient s refractory to treatment remain unidentified. To gain insights into the biological processes underpinning resolution of shoulder tendinopathy, we investigated the bioactive lipid mediator (LM) profiles of tendon stromal cells isolated from patients with supraspinatus tendinopathy to determine whether LM profiles differed between cells isolated from symptomatic and asymptomatic patients before and after ASAD. We hypothesized that tendon stromal cells isolated from painful post‐treatment patients would show dysregulated resolution responses. Tissue biopsies were collected from symptomatic patients (n = 5) during general anesthesia prior to ASAD (painful pre‐treatment group). Samples were also collected from patients between 2 and 4 years after ASAD in whom pain had resolved completely (pain‐free post‐treatment, n = 5) or pain persisted (painful post‐treatment, n = 4). All patients were aged between 38 and 65 years and had not received cortisone treatment for 3 months prior to inclusion in the study. Patient age, gender, and BMI did not differ significantly between cohorts. In post‐treatment patients, tendon biopsies were performed under ultrasound guidance and local anesthesia using a validated biopsy technique.4 Tissues were collected under research ethics from South Central Oxford (14/SC/0222). Full informed consent according to the Declaration of Helsinki was obtained from all patients. Tendon stromal cells (CD45−CD34− fibroblasts) were isolated from biopsy samples using previously published protocols.3,5 Passage 1‐2 cells were cultured to confluence in the presence/absence of IL‐1β (10 ng mL−1) in DMEMF12 phenol red‐free medium, containing 1% heat‐inactivated human serum and 1% penicillin‐streptomycin. IL‐1β was selected as this proinflammatory cytokine is known to induce expression of NF‐κB target genes highly expressed in supraspinatus tendinopathy.3 Cells were incubated for 24 hours until harvest of the media and lysate for LM profiling, which was performed using a previously described protocol.6 RNA was isolated from collected lysates and RT‐qPCR performed as previously described,3 to determine whether 15‐PGDH mRNA expression varied between tendon stromal cells isolated from the above patient cohorts. Tendon stromal cells derived from painful pre‐treatment, painful post‐treatment, and pain‐free post‐treatment patients showed distinct clustering of the LM profiles (Figure 1A, 1B). Cells from painful post‐treatment patients showed increased cumulative levels of specialized proresolving mediators (SPMs) compared to respective cells from painful pre‐treatment and pain‐free post‐treatment patients (Figure 1C, 1D). Cumulative SPM from cells isolated from painful post‐treatment patients was increased compared to respective cells from pain‐free post‐ treatment patients under IL‐1β‐stimulated conditions (P = 0.03). Assessment of individual LM concentrations demonstrated significant increases in specific SPMs including maresin 2 and resolvin D5n3DPA in cells isolated from the tendons of post‐treatment compared to pre‐treatment patients (Figure 1 E and F). In cell incubations from painful post‐treatment patients, we also found increased levels of inflammation‐initiating eicosanoids including PGE2 and PGF2α (Figure 1G). Tendon cells isolated from pain‐free post‐treatment showed reduced 15‐PGDH mRNA expression compared to respective cells from painful pre‐treatment patients (P = 0.03, Figure 1H). This study presents a unique opportunity to understand the mechanisms underpinning persistent tendon pain and inflammation. Increased levels of SPM and inflammation‐initiating eicosanoids in cells from persistently painful patients suggest that the concentrations of SPM are not sufficient to counter‐regulate ongoing inflammatory processes in cells from these patients, reminiscent of a dysregulated resolution response characteristic of chronic inflammation. We previously identified stromal cells isolated from patients with