Purpose To investigate the effect of the location of a partial ferrule on 2 walls and the influence of ferrule height variations on remaining walls, fracture resistance, and failure mode of maxillary incisors endodontically treated and restored with fiber posts. Materials and Methods Sixty intact human maxillary central incisors were divided into 6 groups (n = 10): no ferrule (NF), 2‐mm ferrule (CF2), absence of lingual wall and 1 proximal wall with remaining walls’ height, 1 mm (IF1), 2 mm (IF2), 3 mm (IF3), and 4 mm (IF4). They were restored with a fiberglass post and composite resin core. Complete metal crowns were then cemented on all specimens. Completed specimens were subjected to thermal cycling (6000 cycles, 5°C/55°C) followed by the immediate testing of fracture resistance. After failure, the specimens were sectioned buccolingually and evaluated to identify the mode of failure. The data were analyzed with an ANOVA and the Student‐Newman‐Keuls multiple comparison tests (α = 0.05). Results An incomplete ferrule (IF2) with the lingual wall missing and 1 proximal wall had a fracture resistance of 494 ± 137 N, which was significantly less that of a complete ferrule (CF2) (707 ± 162 N, p = 0.002). An increase of 3 to 4 mm in height of the remaining walls had negligible influence on the fracture resistance: 514 ± 117 N (IF3), 557 ± 177 N (IF4). Partial decementation was observed in all specimens of IF3 and IF4, in 9 of IF1 and IF2, in 6 in NF, and in 3 in CF2. Root fractures occurred in 7 specimens, 4 in CF2, 1 in IF1, 1 in IF2, and 1 in IF4. Cracks under the cemento‐enamel junction occurred in 4 specimens (NF and IF1), in 5 (IF2 and IF3), and in 6 (CF2 and IF4) and were considered catastrophic failures. Conclusions Specimens with a complete ferrule of 2 mm height were more resistant to fracture than specimens with 2 mm height and absence of the lingual wall and 1 proximal wall. A 3 or 4 mm wall height increase was associated with an insignificant increase in fracture resistance and cannot compensate for the absence of the lingual wall and 1 proximal wall.

SUMMARY Purpose To investigate the effect of a cervical cavity extending 1 mm apical to the cemento–enamel junction (CEJ) on fracture resistance and failure mode of maxillary central incisors that have been treated endodontically, present with complete and incomplete ferrules, and are restored with and without a fiber post. Methods and Materials 50 intact human maxillary central incisors were divided into five groups (n=10): CG (control group) 6-mm fer-rule height, no cervical cavity, and without post; (CO) 6-mm ferrule height without post, with a cervical cavity (access to root canal and cervical cavity restored with composite resin), cervical cavity; and post with ferrule heights of 1 mm (CP1), 2 mm (CP2), and 6 mm (CP6) restored with fiberglass post and composite resin core. After complete metal crowns were cemented on all specimens, they were subjected to thermal cycling (6000 cycles, 5°C/55°C), followed by immediate testing of fracture resistance. After failure, the specimens were sectioned buccolingually to evaluate and identify the mode of failure. The data were analyzed with an analysis of variance (ANOVA) and the Student–Newman–Keuls multiple comparison tests (α =0.05). Results A 1-mm ferrule height (CP1) fracture resistance was significantly lower (531±125 N) compared to the 6-mm ferrule height (CP6) (769±175 N) (p<0.05). With respect to the groups with similar residual dentin, with and without a cervical cavity, CG (667±119 N) and CO (668±119 N), the presence of a post (CP6) increased the resistance to fracture, although no statistically significant difference was demonstrated. Partial decementation was observed in all specimens of CG and CP6, in nine of CP1 and CP2, and in three in CO. Root fractures occurred in 23 specimens. The root surface was exposed 2 mm below the CEJ to simulate bone level. Propagation of subosseous cracks occurred in four specimens in CG and CP2, in seven specimens in CP6, in two specimens in CP1, and in six specimens in CO. All were considered catastrophic failures. Conclusions Within the limitations of this study it is suggested that, when restoring an endodontically treated maxillary central incisor that has a cervical lesion and needs to be restored with a complete crown, a fiber post is cemented to improve fracture resistance.