Crown pull-off test (crown retention test) to evaluate the bonding effectiveness of luting agents

The purpose of this review was to assess the influencing factors which affect laboratory tests that evaluate the effectiveness of luting agents on the retention of crowns in prepared dentin and - based on the results of the review - to propose a reasonable experimental setup.

The database MEDLINE was systematically searched for laboratory methods that evaluated the effectiveness of luting agents by pulling off crowns from prepared extracted teeth.

Eighteen studies were included into the systematic review. The studies varied largely with regard to tooth type (molars, premolars), number of specimens (9-25), stump height (3-6mm), convergence angle (4.8-33 degrees ), standardization and measurement of preparation surface, seating force (25-200N), artificial ageing, crosshead speed for tensile force and statistical analysis. The coefficient of variation of the test results varied from 3% to 100%. The most important influencing factors for the crown dislodgement were stump height and convergence angle as well as the luting agent. Panavia and RelyX Unicem generally produced the highest values followed by glass ionomer and zinc phosphate cements. When pooling and normalizing the data, the mean difference between glass ionomer and resin-based materials as well as between glass ionomer and zinc phosphate cements was statistically significant (Wilcoxon, p<0.05). Seating force, roughness, type of cutting bur and use of a desensitizing agent had all a negligible effect on the test results. Artificial ageing like thermocycling had no influence with glass ionomer cements whereas for resin-based cements thermocycling and prolonged water storage generated similar a failures stress than thermocycling alone. The comparison with clinical results did not reveal conclusive evidence that the results of the laboratory methods completely reflect the results of prospective clinical trials in conjunction with single crowns and fixed dental prostheses. A reasonable experimental setup includes: at least 20 specimens per group, stump height 3mm, convergence angle 20 degrees , thermocycling of specimens (5000x), avoidance of shearing forces during dislodgement and failure probability statistics (Weibull).

The results of pull-off tests with crowns that are cemented with luting agents varied largely and reflect only partially the results from clinical trials. The most influencing factors (standardized tooth preparation, avoidance of shearing stress during dislodgement of crown) had to be controlled to get meaningful results.