# Pharmacokinetics of Continuous Local Antibiotic Perfusion in Fresh-Frozen Cadaver Models of Hip and Knee Arthroplasties

## TL;DR

This study explores how continuous local antibiotic perfusion can deliver high antibiotic concentrations in joint models, with results varying based on implant type and anatomy.

## Contribution

The study provides anatomical and pharmacokinetic insights into CLAP's effectiveness in different arthroplasty models using a cadaveric approach.

## Key findings

- Gentamicin concentrations peaked at 36 hours in hip models and remained high in native knee models.
- Cemented TKA blocked antibiotic diffusion from marrow to joint, while trans-acetabular perforations improved hip perfusion.
- Crystal violet dye confirmed flow pathways were implant and anatomy-dependent.

## Abstract

Background

Periprosthetic joint infections (PJI) remain a challenging complication of arthroplasty. The aim of using continuous local antibiotic perfusion (CLAP), which consists of intramedullary antibiotic perfusion (iMAP) and negative-pressure intra-articular drainage, is to maintain high local antibiotic concentrations while minimizing systemic exposure. In this study, the anatomical feasibility and pharmacokinetics of CLAP were evaluated in fresh-frozen cadavers.

Methods

A cadaver (female, 90 years old) was used to model total hip arthroplasty (THA), bipolar hemiarthroplasty (BHA), cemented total knee arthroplasty (TKA), and septic native knee. Gentamicin sulfate (1,200 μg/mL) was continuously infused at 2 mL/hour through iMAP pins. In knee models, intra-articular infusion and drainage were performed using a Salem Sump™ tube (Cardinal Health, Dublin, OH). The joint effluent was sampled at 12, 20, 36, and 48 hours and analyzed using particle-enhanced turbidimetric inhibition immunoassay. After the infusion, 20 mL of crystal violet was injected through iMAP pins to visualize the flow pathways.

Results

Gentamicin concentrations increased over time in both hip models, peaking at 36 hours (total hip arthroplasty: 1,170 μg/mL; bipolar hemiarthroplasty: 930 μg/mL). The native knee concentration remained at ≥1,000 μg/mL after 20 hours. In the cemented TKA model, concentrations remained stable (1,098-1,185 μg/mL); crystal violet did not pass from the iMAP system into the joint, indicating that the cement was blocking marrow-joint communication. Dye studies confirmed intra-articular inflow enhanced by trans-acetabular perforations in hip models. In knee models, the dye largely remained within the marrow unless a direct osseous-articular conduit was present.

Conclusions

In this cadaveric study, we demonstrated that the use of CLAP enabled high intra-articular gentamicin levels to be maintained when communication existed between the iMAP system and the joint cavity. Trans-acetabular perforations improved perfusion in the hip, whereas cemented TKA impeded diffusion, indicating that delivery routes must be adapted to the fixation status. Because this cadaveric model is based on anatomical and mechanical assumptions and lacks physiological processes such as blood flow and drug clearance, the observed pharmacokinetics should not be interpreted as directly equivalent to in vivo conditions. Although limited because only a single cadaver was used and living physiology was absent, these findings provide foundational anatomical and pharmacokinetic insights to support the optimization of CLAP in clinical practice.