The Biologic Response to Polyetheretherketone (PEEK) Wear Particles in Total Joint Replacement: A Systematic Review

Polyetheretherketone (PEEK) and its composites are polymers resistant to fatigue strain, radiologically transparent, and have mechanical properties suitable for a range of orthopaedic applications. In bulk form, PEEK composites are generally accepted as biocompatible. In particulate form, however, the biologic response relevant to joint replacement devices remains unclear. The biologic response to wear particles affects the longevity of total joint arthroplasties. Particles in the phagocytozable size range of 0.1 µm to 10 µm are considered the most biologically reactive, particularly particles with a mean size of < 1 µm. This systematic review aimed to identify the current evidence for the biologic response to PEEK-based wear debris from total joint arthroplasties.

(1) What are the quantitative characteristics of PEEK-based wear particles produced by total joint arthroplasties? (2) Do PEEK wear particles cause an adverse biologic response when compared with UHMWPE or a similar negative control biomaterial? (3) Is the biologic response affected by particle characteristics?

Embase and Ovid Medline databases were searched for studies that quantified PEEK-based particle characteristics and/or investigated the biologic response to PEEK-based particles relevant to total joint arthroplasties. The keyword search included brands of PEEK (eg, MITCH, MOTIS) or variations of PEEK types and nomenclature (eg, PAEK, CFR-PEEK) in combination with types of joint (eg, hip, knee) and synonyms for wear debris or immunologic response (eg, particles, cytotoxicity). Peer-reviewed studies, published in English, investigating total joint arthroplasty devices and cytotoxic effects of PEEK particulates were included. Studies investigating devices without articulating bearings (eg, spinal instrumentation devices) and bulk material or contact cytotoxicity were excluded. Of 129 studies, 15 were selected for analysis and interpretation. No studies were found that isolated and characterized PEEK wear particles from retrieved periprosthetic human tissue samples.

In the four studies that quantified PEEK-based particles produced using hip, knee, and spinal joint replacement simulators, the mean particle size was 0.23 µm to 2.0 µm. The absolute range reported was approximately 0.01 µm to 50 µm. Rod-like carbon particulates and granular-shaped PEEK particles were identified in human tissue by histologic analysis. Ten studies, including six animal models (rat, mouse, and rabbit), three cell line experiments, and two human tissue retreival studies, investigated the biologic response to PEEK-based particles. Qualitative histologic assessments showed immunologic cell infiltration to be similar for PEEK particles when compared with UHMWPE particles in all six of the animal studies identified. However, increased inflammatory cytokine release (such as tumor necrosis factor-α) was identified in only one in vitro study, but without substantial suppression in macrophage viability. Only one study tested the effects of particle size on cytotoxicity and found the largest unfilled PEEK particles (approximately 13 µm) to have a toxic effect; UHMWPE particles in the same size range showed a similar cytotoxic effect.

Wear particles produced by PEEK-based bearings were, in almost all cases, in the phagocytozable size range (0.1–10 µm). The studies that evaluated the biologic response to PEEK-based particles generally found cytotoxicity to be within acceptable limits relative to the UHMWPE control, but inconsistent when inflammatory cytokine release was considered.

To translate new and advanced materials into clinical use more quickly, the clinical relevance and validity of preclinical tests need to be improved. To achieve this for PEEK-based devices, human tissue retrieval studies including subsequent particle isolation and characterization analyses are required. In vitro cell studies using isolated wear particles from tissue or validated joint replacement simulators, instead of manufactured particles, are also required.