gms | German Medical Science

Objectives: Periprosthetic joint infections (PJI) occur in 1–3% of primary arthroplasties, with high failure rates after mandatory revision surgery. We theorise that the occurrence of PJI significantly and lastingly affects the skeletal joint components of the knee. These changes mediated by cytokines and cell-cell interactions may lead to increased risk of aseptic loosening and recurrent PJI after revision arthroplasty. This study investigates the potential impact of PJI on bone metabolism and whether two-stage revision arthroplasty can restore the physiological bone environment.

Methods: 30 specimens of human distal femoral bone were collected at primary total knee arthroplasty (control), at septic implant removal, and at prosthesis reimplantation. µCT analysis was used to assess trabecular thickness, distance, and bone mineral density. Osteoclast and osteoblast cell numbers were determined after TRAP and Alcian blue staining. Expression of Osteocalcin, Osteopontin, TNF-stimulated gene 6 protein (TSG-6), and IL-17a was analysed using PCR.

Results and conclusion: In PJI, bone volume decreased to 0.17 at explantation and 0.22 at reimplantation compared to control (BV/TV: 0.31, p<0.001). Trabecular thickness (164.2±18.8 µm) and separation (1833.9±178.5 µm) were altered in patients with PJI compared to control (Tb.Th: 250.4±19.6 µm, p<0.001; Tb.Sp: 1367.4±101.0 µm, p<0.001). Both parameters improved at reimplantation, however, trabecular thickness remained lower (Tb.Th: 221.1±19.6 µm, p<0.001; Tb.Sp: 1417.6±122.7 µm, p<0.001). Femoral bone mineral density was reduced in PJI at prosthesis removal (-8.02%, p<0.001) and reimplantation (-3.18%, p=0.047). Osteoblast per bone surface cell count at prosthesis explantation (Ob./B.S.: 0.149; p<0.001) and reimplantation (Ob./B.S.: 0.231; p<0.001) was elevated compared to control (Ob./B.S.: 0.020). Osteoclasts per bone surface was increased at explantation (Oc./B.S.: 0.663 vs 0.012, p<0.001) and reimplantation (Oc/B.S.: 0.298, p=0.001).

Expression of osteoblast function markers osteocalcin (0.105, p=0.002) and osteopontin (femur: 0.414, p=0.007) was reduced at prosthesis explantation. While no difference was observed for osteopontin, expression of osteocalcin remained lower at reimplantation (0.252, p=0.044). Expression of TSG-6, a protein promoting bone remodelling, was reduced at prothesis explantation (0.385, p=0.008) and reimplantation (0.523, p=0.044). Expression of osteoclastogenesis-stimulating cytokine IL-17a was increased at explantation (1.844, p=0.022) and reimplantation (1.844, p=0.004).

We observed significant changes in expression of markers linked to bone metabolism in patients with PJI suggesting that inflammation also affects the bone tissue on a cellular and microstructural level. Altered gene expression at time of reimplantation suggests that, despite a two-stage prosthesis exchange, bone metabolism has not returned to baseline activity and may play a crucial role in high prosthesis failure rates after revision arthroplasty.