Rapid diagnosis of periprosthetic joint infection from synovial fluid in blood culture bottles by direct matrix-assisted laser desorption ionization time-of-flight mass spectrometry

Abstract

Background: The aim of this prospective study was to use direct matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS) to rapidly diagnose periprosthetic joint infections (PJIs).

Method: Synovial fluid was taken from 77 patients (80 joints, 41 hips and 39 knees) who met the International Consensus Meeting criteria for PJI, and inoculated into blood culture bottles (BCBs) and onto conventional swabs. Positive blood cultures were analyzed using either direct or routine MALDI-TOF MS. Pathogen identification and the time to identification was recorded. Differences between groups were analyzed using the Kruskal-Wallis test and Bonferroni's post-hoc test.

Results: Direct and routine MALDI-TOF MS both detected 64 positive results (80%), compared to 47 (59%) by conventional swabs (p = 0.002). Direct MALDI-TOF MS identified 85.3% of the gram-positive organisms and 92.3% of the gram-negative organisms. No fungi were identified by direct MALDI-TOF MS. In 17 BCBs that were flagged positive, identification by direct MALDI-TOF MS failed. Among the positive results in the direct MALDI-TOF MS group, Staphylococcus aureus accounted for 47%, followed by Staphylococcus epidermidis (17%), Escherichia coli (9%) and Klebsiella pneumoniae (9%). The median time to microorganism identification was significantly shorter with direct MALDI-TOF MS (12.7 h, IQR: 8.9-19.6 h) than with routine MALDI-TOF MS (39.5 h, IQR: 22.8-46.0 h) or swabs (44.4 h, IQR: 27.2-72.6 h) (p < 0.0001). In pairwise comparisons, there were significant differences in the time of microorganism identification between direct MALDI-TOF MS and routine MALDI-TOF MS (p < 0.0001) or swab culture (p < 0.0001). There was no significant difference between routine MALDI-TOF MS and swab culture (p = 0.0268).

Conclusion: Compared with current laboratory practice, direct MALDI-TOF MS shortened the time to microorganism identification and had superior results compared to conventional swabs, except for fungi. Further studies should investigate whether the earlier administration of appropriate antimicrobial agents can improve the treatment outcomes of PJIs.

Fig 1. Schematic representation of the diagnostic workflow.

Fig 2. Median time to identification of microorganisms.

The error bars represent the 25th and 75th percentiles. * Significant Bonferroni-adjusted pairwise comparisons compared to the routine MALDI-TOF MS. Δ Significant Bonferroni-adjusted pairwise comparisons compared to the swabs.