Comparison of three in vitro implant leakage testing methods

Aims: To assess the accuracy and sensitivity in detecting implants leakage with a gas-enhanced permeation test (GEPT) and to compare with a molecular- and a bacterial-based leakage tests. Materials and methods: Three implants systems were tested (n = 20 per group): Nobel Biocare (NB), Astra Tech (AT) and Biomet 3i (B3i). Implants were mounted in PVC disks and were first tested for gas pressure change and infiltrated saline volume over 40 min. The same implants were then subjected to a molecular leakage evaluation using fluorescent Dextran for 28 days. After cleaning and sterilization, bacterial permeation (E. faecalis) was evaluated by selective media turbidity for another 28 days. Slopes in the pressure change and the perfused saline rate were used as a measure of leakage in the GEPT model and the times of positive events, that is, color change, after molecular and bacterial tests were recorded. Data were analyzed using Kolmogorov-Smirnov/Shapiro-Wilk, Kruskal-Wallis H and Spearman's Rho tests (P < 0.05). Results: The gas and saline (ml) leakage values accounted for 0.85 ± 0.71 and 0.56 ± 0.50 ml (AT), 0.23 ± 0.030 and 0.12 ± 0.20 ml (NB) and 0.01 ± 0.01 and 0 ± 0 ml (B3i), respectively, and were significantly different from each other (P < 0.001). Slope in the pressure change over time showed a significant positive correlation with the collected saline solution (r = 0.91; P < 0.001). Molecular and bacterial leakage was positive at the same implants, which also showed increased leakage values in the GEPT setup. The development of positive events in the timeline of the bacterial leakage evaluation corresponded well to the GEPT leakage model. Conclusion: The GEPT proved to be a reliable method to quantify leakage. The B3i showed the best sealing among the tested systems.