Abstract
Aim
In cases of prosthetic joint infections the sensitivity of bacterial cultivation of tissue samples is not 100%. In fact, the reported sensitivity based on standardized criteria and rigorous tissue sampling technique probably differs between 86 to 89%. It has been claimed that sonication of explanted prostheses with subsequent culturing of sonication fluid can increase the sensitivity of the test compared to culturing of tissue samples. To what degree bacteria embedded in biofilm is dislodged during the sonication process has to our knowledge not been fully elucidated. We studied the effect of sonication as a method to dislodge biofilm embedded Staphylococcus epidermidis in vitro.
Method
46 steel plates were colonized with biofilm forming S. epidermidis ATCC 35984 in TSB with 1% glucose aerobically at 37°C for 24 hours. Plates were cleansed for non-adherent bacteria before microscopy. Biofilm embedded bacteria were stained with LIVE/DEAD ™ BacLight ™ Bacterial Viability Kit for microscopy and visualized under vital conditions using EVOS™ FL Auto 2 Imaging System (epifluorescence) and an inverse confocal laser scanning microscope LSM510 (CLSM). All steel plates were subjected to epifluorescence microscopy before and after sonication. CLSM and SEM were used to confirm the presence of biofilm embedded bacteria after sonication. Pictures from epifluorescence microscopy were processed for image analysis with help of a macro application (Fiji) and the data was expressed as biofilm coverage rate (BCR). The sonication was performed using a BactoSonic® Bandolin sonicator and the applied effect in each glass test tube (40 kHz, 800W) was measured with a Bruel og Kjær 8103 hydrophone. The amount of bacteria in the sonication fluid was quantified by counting the number of colony forming units (CFU).
Three steel plates acted as negative controls.
Results
The BCR was highly variable on the plates after sonication. The biofilm was eradicated from the majority of the plates but a considerable number of plates still had biofilm attached to the surface in a highly variable manner. The amount of bacteria in the sonication fluid correlated poorly with BCR on corresponding plates.
Conclusions
Our conclusion is that the ability of sonication to dislodge biofilm embedded S. epidermidis in vitro is not as effective as current opinion might suggest. After sonication biofilm still adhere to a significant number of plates in a highly varying manner. This prompts the need to investigate the effect of sonication on biofilm embedded bacteria formed in vivo.
