In Vivo Antibiotic Elution and Inflammatory Response During Two-Stage Total Knee Arthroplasty Revision: A Microdialysis Pilot Study

<b>Introduction</b>: Two-stage revision with an antibiotic-loaded, temporary static cement spacer is a common treatment for periprosthetic joint infection (PJI) of the knee. However, limited data exists on in vivo antibiotic elution kinetics after spacer implantation. This pilot study uses the technique of microdialysis (MD) to collect intra-articular knee samples. The aim was to evaluate MD as an intra-articular sampling method to detect spacer-eluted antibiotics within 72 h after surgery and to determine whether they show specific elution kinetics. <b>Methods</b>: Ten patients (six male, four female; age median 71.5 years) undergoing two-stage revision for knee PJI were included. A MD catheter was inserted into the joint during explantation of the infected inlying implant and implantation of a custom-made static spacer coated with COPAL cement (0.5 g gentamicin (G) and 2 g vancomycin (V)). Over 72 h postoperatively, samples were collected and analyzed for spacer-eluted antibiotics, intravenously administered antibiotics (e.g., cefazolin and cefuroxime), metabolic markers (glucose and lactate), and Interleukin-6 (IL-6). Local and systemic levels were compared. <b>Results</b>: All catheters were positioned successfully and well tolerated for 72 h. Antibiotic concentrations in MD samples peaked within the first 24 h (G: median 9.55 µg/mL [95% CI: 0.4-17.36]; V: 37.57 µg/mL [95% CI: 3.26-81.6]) and decreased significantly over 72 h (for both <i>p</i> < 0.05, G: 4.27 µg/mL [95% CI: 2.26-7.2]; V: 9.69 µg/mL [95% CI: 3.86-24]). MD concentrations consistently exceeded blood levels (<i>p</i> < 0.05), while intravenously administered antibiotics showed higher blood concentrations. Glucose in MD samples decreased from 17.71 mg/dL to 0.89 mg/dL (<i>p</i> < 0.05). IL-6 and lactate concentrations showed no difference between MD and blood samples. <b>Conclusions</b>: Monitoring antibiotics eluted by a static spacer with intra-articular MD for 72 h is feasible. Gentamicin and vancomycin levels remained above the minimal inhibitory concentration. Differentiating infection from surgical response using metabolic and immunological markers remains challenging. Prolonged in vivo studies with MD are required to evaluate extended antibiotic release in two-stage exchanges.