Bulk fracture and loss of restoration have historically been reported to be the main reasons for clinical failures of inlays and onlays41, 47-48, 91. Lithium disilicate ceramic
however, with its improved mechanical durability compared to other glass ceramics, may offer increased clinical longevity for partial coverage restorations53, 55, 59.
The use of bonded ceramics may be advantageous in clinical scenarios where traditional mechanical retention form is limited, and may allow for more conservative tooth preparations compared to mechanically retained restoration preparations. In these cases, the need for conventional means of retention such as ideal taper and axial height of the preparation, is reduced due to the bonding potential of ceramic to dentin and enamel with composite resin cement. Alternately, the success of these partial coverage ceramic restorations is, to a high degree, dependent on sufficient support and an optimal adhesive bond to the underlying tooth structure. The resin cement seals the intaglio surface of the ceramic from crack initiation, maximizing its potential strength properties. This is a major advantage in the use of bonded ceramic onlays, in that they may be used in cases with minimal or no retention to conserve healthy tooth structure. Classic crown preparation
approaches may, in these cases, involve extensive preparation and/or endodontic treatment, in combination with a post or pin(s) and core placement in order to achieve retention. Rather, the minimal preparation for the bonded ceramic partial coverage restoration is less traumatic for the tooth, and pulp vitality may be preserved.
The clinical performance of adhesively bonded all ceramic restorations has mostly been studied in the short- and medium-term. There are few studies with extended
observation periods of 10 years or more. Frankenberger et al. found the failure rate for IPS Empress (leucite-reinforced feldspathic porcelain) inlays and onlays of 12% at 12 years, mostly due to bulk fracture47. Van Dijken and Hasselrot conducted a 15-year evaluation of extensive dentin-enamel bonded IPS Empress partial and complete coverage restorations. In the clinical trial, the authors included onlays that preserved very thin portions of buccal or lingual cusps. Overall, the cumulative failure rate was 24% after 15 years. Most of the failures were due to restoration loss or fracture, with only 1 tooth in the entire study experiencing a cusp fracture48.
Long term, randomized controlled clinical trials are lacking that evaluate lithium disilicate’s use for onlay restorations. In an effort to evaluate the clinical performance of IPS e.max, there have been two studies that sought to indirectly gain insight into
restoration longevity by analyzing dental laboratory data from dentists making remake requests. One study showed 0.99% of e.max onlays were requested to be remade due to fracture over 7.5 years67. The second study reported 0.70% of IPS e.max CAD onlay restorations be remade after 3.5 years due to fracture77. Although these studies do not replace the need for long-term randomized controlled clinical trials, they do show that in
the medium-term, IPS e.max onlays do not experience a high rate of catastrophic failure. However, these studies may tend to under report negative outcomes since the only restorations counted are the ones that were remade by the original laboratory and the patient by the same dentist. Furthermore, patients, who had received the IPS e.max onlay(s), who had moved to other regions where not included. It can be reasonably said that if there had been large numbers of these restorations failing that this would have surfaced through these indirect studies.
Ideal preparation design standards for dentin-enamel-bonded crowns or partial coverage restorations are lacking. A common clinical dilemma regarding inlay and onlay restorations relates to cavity design. Specifically, there is confusion amongst dentists regarding occlusal isthmus width, remaining cusp thickness and when cuspal coverage is indicated. The traditional principals of cavity preparation with respect to cuspal coverage are based on cast metal or amalgam restorations that do not adhere to the dental tissue5-7, 92.
Several in vitro studies have shown that the use of adhesive techniques may provide cuspal reinforcement and enhance fracture resistance of nontraditional onlay preparation designs93,96,113. The present study also demonstrated the high fracture resistance of
adhesively bonded onlay restorations. Additionally, unlike the previous mentioned studies, this experiment sought to more thoroughly mimic clinical conditions by exposing samples to simultaneous thermocycling and mechanical loading, creating an environment where ceramic is typically subject to fatigue failure94, 100-101. The samples were exposed to 1.2million cycles, an equivalent to five clinical years100. Additionally, the 3D motion of the
teeth during chewing was simulated with both vertical and lateral movements of the stainless-steel antagonist. In all test groups, an effort was made to replicate the typical cusp-to-fossa contacting relationship. During each mechanical loading cycle the stainless- steel stop made contact with the functional cusp, then moved laterally 2.0mm to contact the central fossa. The use of lateral movements on restorations has a deteriorating effect, especially in wet environments. Therefore, it is recommended that any laboratory
simulation intended to establish the longevity of an all-ceramic restoration include lateral movements to more closely resemble clinical oral conditions111-112.
In the present study, there was no association found between the fracture resistance of premolars restored with CAD/CAM lithium disilicate onlays and the preserved buccal- lingual wall width of the remaining cusps. Therefore, the null hypothesis that surrounding wall thickness does not affect the fracture resistance of the tooth-restoration complex was accepted. Notably the nonfunctional cusps, which were exposed to a degree of lateral forces when the stainless-steel stop rotated to the central fossa, did not experience any fractures. There was a significant difference in the failure rates between the groups that preserved the functional cusp and the groups that preserved the nonfunctional cusp. Therefore, the null hypothesis that functional or nonfunctional cusp preservation does not affect the fatigue resistance of a tooth-restoration complex was rejected.
Group 1 displayed the highest failure rate of 75.0%. It was noted after visual examination of the test specimens that the antagonist happened to make contact on the margin between the preserved functional cusp and the restoration (Figure 2.5). Magne et al. studied premolar cuspal flexure as a function of restorative material and occlusal
contact location. It was found that antagonist contact with the restoration margin
demonstrated the most amount of cuspal deformation98. In the present study, antagonist forces directly contacting the restoration margin on the functional cusp in Group 1 may have caused cuspal deflection, stressed the adhesive bond of the ceramic onlay to the tooth and led to a high rate of onlay debonding. Group 6 preparation design was identical to Group 1, with added mesial and distal boxes for retention. Remarkably, this group demonstrated a 0.0% failure rate. Although the antagonist also made contact with the restoration margin in this group, the added retentive boxes may have minimized potential flex or movement of the onlay away from the cusps, thus preventing the onlays from debonding. In addition, the presence of retentive boxes likely limited the ability of shear forces to overcome the adhesive interface. Therefore, a clinician may want to consider added retentive features to an onlay preparation design if the opposing occlusion will be on the restoration margin. In Group 2, a 2.0mm buccal-lingual width of the functional cusp was preserved in each sample. This design allowed for the antagonist to make majority contact with the onlay restoration, and not the onlay margin or remaining cusp (Figure 2.6). This group demonstrated a 0.0% failure rate.
Figure 2.5. An occlusal view of a test specimen from group 1 after simultaneous thermocycling and mechanical loading. The wear pattern of the antagonist demonstrates
initial mechanical loading primarily on the tooth and restoration margin.
Figure 2.6. An occlusal view of a test specimen from group 2 after simultaneous thermocycling and mechanical loading. The wear pattern of the antagonist demonstrates
initial mechanical loading primarily on the ceramic restoration. There was a statistically significant difference in the cumulative failure rates
functional cusps are generally well protected and subject to more compressive forces while nonfunctional cusps tend to receive more tensile stresses90, it is important to note that none of the failures in this study were due to cusp fracture. The mechanism of failure in the functional cusp group may be more related to the margin location in relation to antagonist occlusion, as previously discussed. Although thin cusps were preserved in groups 2, 3 and 4, the antagonist occlusion was not directly on the restoration margin and they experienced only one debonded restoration and no cusp or restoration fractures.
The tooth structure saving preparation designs used in this study may minimize the risk of pulpal complications in vital teeth. In addition, other advantages may include
preserving a tooth’s natural hues and shades, intracoronal geometry and position in the arch. The use of adhesively retained lithium disilicate ceramics, that demonstrate improved biomechanical properties, for partial coverage restorations may be a viable alternative to full coverage restorations, and may challenge the once widely accepted principals related to preparation resistance and retention form.