Changes in serum markers failed to predict persistent infection after two-stage exchange arthroplasty

Background

The proper timing of reimplantation is importation to treatment success in the two-stage exchange revision. The 2018 International Consensus Meeting suggested that a variation trend toward normalization in serum markers was useful for determining the proper timing of reimplantation. However, the opposite results were found by previous studies, and the normalization of serum markers was reported to fail to predict infection control. We investigated whether value changes and percent changes in four common serum markers (erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), interleukin-6 (IL-6), and fibrinogen) can predict persistent infection.

Methods

A retrospective review of 141 patients treated with the two-stage revision from 2014 to 2018 was conducted. The variation trend in serum indicators was evaluated by the percent changes (using values of serum markers pre-reimplantation divided by values pre-resection) and value changes (using values of serum markers pre-resection minus values pre-reimplantation). Treatment success was defined according to the Delphi-based consensus criteria with a minimum follow-up of 1 year, and the receiver operator characteristic (ROC) was used to examine the usefulness of changes in serum markers.

Results

Twenty-two patients (15.60%) were persistently infected. No significant difference was found in either the value change or percent change in serum markers between reinfection and non-reinfection patients. When predicting persistent infection, the area under the curves (AUC) demonstrated that both percent changes and value changes in serum markers were poor indicators. The AUC of value changes was 0.533 for the CRP, 0.504 for the IL-6, 0.508 for the ESR, and 0.586 for fibrinogen when predicted persistent PJI. In addition, the AUC indicated that percent changes in the CRP (0.464), the IL-6 (0.534), the ESR (0.527), and fibrinogen (0.586) were all poor markers.

Conclusions

We have shown that both value changes and percent changes in serum markers were not sufficiently rigorous to aid in persistent infection diagnosis. The proper timing of reimplantation must, therefore, take into account various clinical tests rather than the downward trend of serum markers only.