A bone histological approach to distinguish periprosthetic joint infection from aseptic osteolysis in revision joint replacement.

Abstract

Aims: Accurate differentiation between periprosthetic joint infection (PJI) and aseptic osteolysis (AO) remains a major clinical challenge in revision joint arthroplasty. We examined bone biopsies for alterations in bone matrix integrity and the osteocyte lacunocanalicular network in order to differentiate between these pathologies.

Methods: Bone biopsies were taken from areas of osteolysis in patients who underwent revision arthroplasty for either PJI, diagnosed according to 2018 International Consensus Meeting criteria, or AO. Controls were obtained during primary arthroplasty. Biopsies were assessed histologically by Masson's trichrome, silver and RGB trichrome staining, and for osteocyte lacunocanalicular characteristics. The diagnostic potential of histological measures was investigated using univariate area under the receiver operating characteristic curve (AUROC) analysis.

Results: RGB trichrome staining was most effective for revealing both bone matrix and lacunocanalicular features. Osteocyte lacunar area and circularity were increased in PJI relative to both AO and Control bone. PJI bone exhibited reduced canalicular density compared to AO (p < 0.001), and shorter, narrower canaliculi than AO and Controls. Predictive univariate modelling revealed that all histological parameters measured except lacunar area were potentially diagnostic for PJI. Degraded bone matrix differentially predicted PJI (AUROC = 0.875, 88% sensitivity, 80% specificity) and AO (AUROC = 0.733, 63% sensitivity, 76% specificity), as did lacunar circularity (PJI: AUROC = 0.903, 75% sensitivity, 96% specificity; AO: AUROC = 0.797, 69% sensitivity, 84% specificity). Canalicular area fraction, length, and width all predicted PJI (AUROCs = 0.915, 0.982, 0.883), with canalicular length yielding 94% sensitivity and 100% specificity. Canalicular density differentially predicted PJI (AUROC = 0.770, 69% sensitivity, 80% specificity) and AO (AUROC = 0.757, 88% sensitivity, 60% specificity).

Conclusion: Histological assessment of bone matrix degradation and the osteocyte lacunocanalicular network reveals new potential diagnostic markers for PJI to support clinical decision-making and provides novel measures for distinguishing PJI from AO, which may be particularly useful in cases of low-grade and chronic infections. Based on these findings, larger confirmatory studies are warranted.