# Gentamicin fails to eradicate Staphylococcus aureus biofilm in vitro, even in combination with rifampin

## TL;DR

High doses of gentamicin, with or without rifampin, cannot fully eliminate Staphylococcus aureus biofilm in laboratory tests, suggesting these antibiotics may not cure device-related infections.

## Contribution

Demonstrates that high-dose gentamicin and rifampin combinations fail to eradicate mature S. aureus biofilm in vitro, challenging current treatment assumptions.

## Key findings

- Gentamicin, even at high concentrations, did not eradicate mature S. aureus biofilm in any tested regimen.
- Adding rifampin did not improve biofilm eradication and sometimes increased bacterial counts.
- Persistent biofilm viability may be due to antibiotic tolerance rather than resistance.

## Abstract

Introduction: Biofilm formation is one of the key elements making orthopaedic device-related infections (ODRIs) difficult to eradicate. Aminoglycosides such as gentamicin are frequently applied via local carriers, and systemic rifampin is added for its anti-biofilm activity. However, robust in vitro evidence of their ability to eradicate mature biofilm is limited. This study assessed whether gentamicin, alone or in combination with rifampin, can eradicate established Staphylococcus aureus biofilm in vitro. Methods: A clinical methicillin-susceptible S. aureus isolate was grown as a 5 d old biofilm on a peg lid microtiter plate. Three exposure regimens were tested: (i) continuous exposure to gentamicin (15–2000 mg L−1) for 28 d, (ii) intermittent 2 h exposures twice daily (at 15, 250 and 2000 mg L−1) for 28 d to reflect systemic twice-daily dosing and (iii) a 14 d burst release starting at 2000 mg L−1 with stepwise decline to model release from local carriers. Rifampin was either absent or added at 3.3 mg L−1, approximating peri-implant concentrations from preclinical pharmacokinetic studies. Biofilm viability was quantified as colony-forming units (CFUs) from sonicated pegs, and selected surviving isolates underwent susceptibility testing. Results: Across all regimens, concentration- and time-dependent decreases in CFU counts were observed, but no regimen resulted in bacterial counts falling below the lower limit of detection (LLOD). The addition of rifampin did not result in the sustained enhancement of biofilm killing, and, in some regimens, resulted in higher CFU counts. Isolates recovered from culture-positive pegs remained largely susceptible to gentamicin, whereas rifampin resistance arose sporadically. Conclusion: High-dose gentamicin exposures failed to eradicate 5 d old S. aureus biofilm in vitro, whatever the administration regimen. Rifampin co-administration did not alter the final outcome of biofilm persistence, despite its well-recognised anti-biofilm activity. These findings challenge the reliance on aminoglycoside-loaded carriers as curative strategies for ODRIs and suggest that persistent viability may reflect antibiotic tolerance that may not be overcome by antibiotics alone.