Comparison of endocrowns made of lithium disilicate glass-ceramic or polymer-infiltrated ceramic networks and direct composite resin restorations: fatigue performance and stress distribution.J Mech Behav Biomed Mater. 2019 12; 100:103401.JM
This study compared the fatigue performance and the stress distribution of endodontically treated molars restored with endocrowns obtained with lithium disilicate glass-ceramic or a polymer-infiltrated ceramic network, both processed by CAD-CAM, and direct composite restorations. Forty-eight human mandibular molars were randomly assigned into 03 groups (n = 16) and restored with endocrowns (LD - lithium disilicate glass-ceramic or PICN - polymer-infiltrated ceramic network) or with direct composite restorations. Fatigue testing followed a step-stress approach (initial maximum load of 200 N and 5000 cycles, incremental step load of 200N and 10,000 cycles/step, being the specimens loaded until failure or to a maximum of 135,000 cycles at 2800 N). The fatigue failure load and number of cycles until failure were recorded and statistically analyzed. Fractographic and finite element (FEA) analyzes were conducted as well. There were no differences in fatigue failure load, number of cycles until fracture and mean survival probabilities among groups. However, indirect endocrowns had higher mechanical structural reliability, and LD restorations lasted more time before start to failing. FEA showed that the stress concentration in tooth tissues was higher for the resin composite, followed by PICN and LD in a decreasing order. Almost all fractures were restricted to the restorative material (without tooth involvement), and origins were identified at occlusal surface. The type of restoration did not influence the fatigue failure load, number of cycles until fracture and mean survival probabilities of the restorative strategies. Despite that, the mechanical structural reliability of endocrowns, especially those made of lithium disilicate, was higher and lasted more time before start to failing.