Abstract
In last years the use of zirconia in dentistry has become very popular. Unfortunately, the clinical indications for a dental use of zirconia are not completely clear yet, neither are their limitations.
The objective of this review was to evaluate the basic science knowledge on zirconia and to discuss some aspects of the clinical behavior of zirconia-based restorations. In particular, one of the goals was highlighting the possible correlation between in vitro and in vivo studies. The definition of concepts like success, survival and failure was still debated and the correlation between in vitro results and predictability of clinical behavior was investigated.
1
Introduction
Since its introduction into the dental market, zirconia (polycrystalline zirconium dioxide) has been widely used to produce fixed partial dentures (FPDs) and implant abutments. Zirconia properties are highly suitable for a dental use: high mechanical properties, natural-tooth appearance, unsolubility in water environment, no cytotoxicity, reduction of bacterial adhesion, radiopacity and low corrosion potential .
Zirconia can mainly be employed according to two different technical solutions. The first, as a ‘white metal’ for manufacturing copings (for single crowns) and frameworks (for multi-unit fixed prostheses); such supporting structures need finally to be veneered with porcelain, in order to achieve the final occlusal/anatomic shape and to exploit the high esthetic potential of this material . This allows to highly improve the esthetic properties of zirconia restorations, although a problem arises: the need of matching mechanical properties and behavior of the two different bilayered materials. As a matter of fact, the most frequent complications of zirconia–ceramic restorations is chipping of the veneering material itself; for this reason, more recently the possibility of using zirconia as a “monolithic” material, shaped in the final anatomic and esthetic tooth morphology, has been advocated.
In its specific form of “yttria-stabilized zirconia polycrystal (Y-TZP)”, zirconia is a high-strength ceramic. The much higher mechanical performances of this material (flexure strength, fracture toughness) compared to most of the other metal-free materials, make framework bulk fractures quite unlikely . On the contrary, a major concern is the chipping of the esthetic ceramic veneer, showing a high incidence, as demonstrated by the majority of clinical trials and systematic reviews . The problem is specific to the bilayer nature of these restorations, as discussed later, and is multifactorial.
Various in vitro studies were performed in order to test the zirconia mechanical properties as a dental material ; at the same time, a large number of in vivo studies (clinical trials) were carried out aimed at the clinical performance of the zirconia-based restorations over time, focused on single crown copings and, prevalently, on FDP frameworks and on implant abutments .
The problem is, both extrapolating the clinical predictability of a certain kind of restoration from in vitro data and correlating its success and failures reported in clinical trials with the material properties emerging from in vitro data do not lead to correct interpretations of scientific results. This is mainly due to the lack of homogeneity in the goals of the in vitro and in vivo study protocols, together with an objective difficulty of controlling too many variables in the tested samples. The designs of in vitro studies usually take into account only single variables ( e.g. , the thickness of porcelain layer and/or of zirconia coping, etc.), whilst in vivo trials are usually conditioned by a high number of variables, not easy to control and often confounding the results of the analyses.
After the widespread diffusion of zirconia as a dental material, two main problems have been evidenced in the clinical practice: chipping of porcelain veneering in single crowns and FDPs and fracture of zirconia abutments .
In particular, chipping/delamination of the veneering ceramic has been described as the most frequently occurring problem of bilayered zirconia restorations .
This paper was aimed at identifying a possible correlation between the most relevant properties of zirconia, as shown under laboratory conditions, and the most significant results of clinical trials, pointing out concepts like clinical success, survival and failure of such restorations.
2
In vitro data
First of all, it has to be noticed that the design of an in vitro research protocol is very sensitive to the technical variables; e.g. , the value of zirconia fracture toughness is highly dependent upon the shape/dimension of the notch that is used to initiate the fracture experimentally . This makes the results of such studies quite hard to compare.
To date, many factors have been reported to be related to the prosthetic complications in zirconia restorations : pressing and structural defects of the frameworks, grinding damages, improper cooling rates, not compatible coefficients of Thermal Expansion, incorrect surface treatment procedures ( e.g. aggressive sandblasting), wrong framework design and thickness, type of finishing margins, incorrect luting procedures, material aging.
2
In vitro data
First of all, it has to be noticed that the design of an in vitro research protocol is very sensitive to the technical variables; e.g. , the value of zirconia fracture toughness is highly dependent upon the shape/dimension of the notch that is used to initiate the fracture experimentally . This makes the results of such studies quite hard to compare.
To date, many factors have been reported to be related to the prosthetic complications in zirconia restorations : pressing and structural defects of the frameworks, grinding damages, improper cooling rates, not compatible coefficients of Thermal Expansion, incorrect surface treatment procedures ( e.g. aggressive sandblasting), wrong framework design and thickness, type of finishing margins, incorrect luting procedures, material aging.
3
Pressing and structural defects
Zirconia mechanical properties are affected by grain size and pressing modalities: higher temperatures and longer sintering times induce the formation of larger grain sizes . Above a critical size, zirconia is less stable and more vulnerable to spontaneous t–m transformation than with smaller grains (<1 μm) . Moreover, below a certain grain size (approximately 0.2 μm), the transformation is not possible, leading to a reduction in fracture toughness. The fabrication process of zirconia frameworks may introduce defects into the material itself . The possible presence of structural defects, such as micro voids and flaws within the material, can concentrate stress resulting in a starting site of internal fracture under loading .
4
Grinding damages
Grinding procedures are often performed in three different phases of the realization of zirconia restorations: the first during machining procedures , the second when reshaping and finishing the morphology and surface of the zirconia copings/frameworks before proceeding with ceramic veneering by dental technicians, , the third when, after the final cementation, occlusal adjustment are needed in the dental office.
It should be considered that such procedures are often due to a poor CAD programming and/or to an inaccurate occlusal design, so the main operative recommendation is to perform a CAD CAM programming in strict compliance with the final design of the restoration. However, it is undeniable that adjustment procedures are not infrequent in the daily practice, sometimes exposing the zirconia core of the restoration to the oral cavity fluids, thinning the porcelain veneering layer and making its chipping likely under clinical service . Some data from in vitro studies, both based upon crystallographic and fractographic techniques, suggest that grinding should be always avoided, in order to maintain strength and thoughness of zirconia restorations unaltered and in order to avoid creation of cracks extended deeper into its bulk . Polishing procedures and surface finish degree seem to have an effect on the onset of residual stress and, consequently, on aging of zirconia. It was suggested that rough polishing can result in the development of compressive surface stress, which would be beneficial for aging resistance, while smooth polishing produces transformation around scratches, due to tensile residual stresses .
Other studies showed that fine polishing after grinding can remove the compressive layer of m-phase from the surface of the zirconia, reducing the severity and the amount of surface defects flaws to a degree where the internal strength of the material becomes the main factor determining its mechanical performance , and generating an increase of flexural strength .
Nevertheless, to date the surface roughness cannot be either considered a predictable parameter or a single indicator for the in vivo stability .
5
Relation between cooling rates, CTE and thickness
The influence of microstructure on the mechanical properties of porcelains, compared with those of glass–ceramics and glass-infiltrated alumina, has been detailed .
Flexural strength of veneering materials generally ranges between 60 and 120 MPa (compared with >450 MPa of core materials). The value of this property, as determined experimentally, can noticeably vary with test conditions .
Several studies have shown that the residual thermal stress due to the lack of compatibility between the CTE (coefficient of thermal expansion) of the core material and that of the veneering ceramic (thermal mismatch) can be one of the main causes of the ceramic chipping . In order to reduce the residual thermal stresses, using a veneering ceramic with a CTE as close as possible to that of the zirconia framework has always been recommended as a paramount prerequisite .
Moreover, slow heating as well as slow cooling rate during the veneering process have been shown to reduce the incidence of chipping in these restorations and do not reduce the mechanical resistance of a zirconia framework itself, while affecting the flexural strength of the veneering ceramic, as well as its bonding strength to the core material . A recent in vitro study evaluated the effects of number of firings on the flexural strength of bi-layered PFZ specimens, showing that three to five veneering layers and firings determine higher flexural resistance when compared to the effect of a single firing and recommending a 3-layers veneering procedure to increase the final flexural resistance ; if a 5-layer procedure is necessary to improve esthetics, it does not decrease this resistance.
Zhang et al. recently pointed out that fast cooling can lead to thermal fracture of different bi-layered ceramic structures and that cooling rate (in terms of heat transfer coefficient) plays a critical role in crack initiation and propagation .
6
Sandblasting and luting procedure
The issue of defining proper luting procedures and materials for zirconia restorations is one of the hottest topic in restorative dentistry. It has to be considered that one of the main problems of metal-free restorations is the difficulty to remove crowns or bridges in case of complications ( e.g. need of endodontic treatment) without creating a damage to their structural integrity through the mechanical stress and indentations related to the procedure.
Different internal surface treatments have been proposed to increase the bond strength between zirconia and cement , although the minimum amount of bond strength that is needed to maintain a crown retention under occlusal loading has not been evidenced yet . Although conventional luting cements like zinc phosphate and glass-ionomer are still extensively used for this purpose, in the last years sandblasting the internal surface of zirconia and cementing the restoration with self-adhesive resin cements has become an increasingly popular procedure. Sandblasting and possible primer application can considerably increase the bond strength of the cement to the crown surface. Some studies have suggested that airborne particle abrasion can reduce the strength and longterm survival of zirconia restorations and that the duration and particle size of air-particle abrasion affects the roughness and phase transformation of Y-TZP: in fact, longer treatment times with larger particles may result in degradation of this material ; however, to date it is clear that for zirconia restorations the sandblasting procedure can be an useful procedure promoting a better performance of luting, enhancing adhesion of luting cement to the framework and providing a strengthening effect for Y-TZP, at the expenses of reduced stability when small particles are used (50 μs aluminum oxide for 5 s at 0.35 MPa pressure at a distance of 2.0 cm) . It has been shown that sandblasting performed after grinding procedures can remove some larger grinding-induced cracks and weakly adherent surface grains and, simultaneously, can produce surface compressive stresses that strengthen the material and affect only the superficial layer of few μs , improving the aging resistance of yttria-stabilized zirconia .
In conclusion, a sandblasting procedure with 30 μm particles can improve the fatigue behavior Y-TZP ceramic materials and can therefore be recommended for adhesive cementation procedures . Similarly, grit blasting with CoJet™ Sand or Rocatec results in an increase of surface roughness, with a removal of maximum 3 μm of material, and coating the surface with submicron silica and alumina particles is also accepted .
Other treatments of internal zirconia surface were investigated, such as grinding alone , plasma spray and glass drops , selective etching , chemical etching .
The choice of the luting material is another important topic. Cements containing phosphate monomer showed an increase of the bond strength to zirconia surface . On the other hand, Inokoshi et al. pointed out that, regarding luting techniques, the combination of mechanical and chemical pre-treatment seems to be particularly crucial to obtain durable bonding to zirconia restorations. The cement choice was not demonstrated to be a determining factor after aging conditions, as long as composite cement was used .
Another main clinical issue related to luting zirconia restorations regards how to remove eventually the cement excess, in particular around fixtures, in order to prevent periodontal or peri-implant complications .
However, it has been already noticed that, up to now, how much bond strength is needed to achieve a clinical long-term stable cementation of zirconia crowns to natural or titanium/zirconia abutments is not clear yet.
7
Framework design
The framework design is another important factor, quite technique-sensitive, that can significantly influence the mechanical performance of all the bi-layered restorations, like the metal–ceramic and zirconia–ceramic ones . An improperly modelled coping substructure, in fact, can lead to a sensible increase of the failure rates, strongly conditioning the fracture modes of final restorations .
The relevance of the framework design for porcelain veneer restoration has been addressed in various studies. According to Rosentritt et al. , the chipping of the veneering ceramic is strongly related to the shape of the zirconia substructure.
Bonfante et al. compared the fracture resistance of glass infiltrated alumina and porcelain fused to metal premolar crowns, designed with a traditional coping design (with a thin uniform thickness) versus a modified framework, reporting that the anatomic modification of framework design determined an increase in the mechanical strength of both the PFM restoration and the all-ceramic restorations . In a recent study the same authors investigated the effect of zirconia framework design on bilayered molar crowns after thermo-mechanical simulation. For the modified framework design, where the lingual side of the core was built up to lower the overall bulk of the cusps, an increase in laboratory fatigue resistance was observed; so, the results confirmed those of the previous study. Silva et al. also reported differences in the fracture load of different zirconia molar crowns related to the framework design. Higher strength was recorded for a modified framework design – core thickness increased by 1 mm in the proximal area, and a 2.5 mm height cervical margin on the lingual site – compared to the traditional coping design (0.5 mm uniform thickness).
Kokubo et al. evaluated four different zirconia framework designs for molar crowns: a conventional zirconia coping with flat occlusal surface, a zirconia coping with 0.6 mm circumferential shoulder collar, a zirconia coping with two inclined cusp planes reproducing cuspal anatomy and a zirconia coping designed for supporting occlusal force with an increase of wall thickness in correspondence of the occlusal plate. The authors concluded that the zirconia copings with a modified shape, designing a cuspal configuration to obtain an even thickness of veneering porcelain layer, showed the highest fracture load.
Kim et al. , in another in vitro study based on Vickers’ indentation test, examined the pattern of failure for brittle bi-layered all ceramic materials, concluding that cracks originating from the veneer top surface are likely to be arrested at the core interface. From this observation the authors speculated that if the ceramic structure is adequately supported, reducing the thickness of the veneering porcelain, the damage could be fairly limited.
More recently, three different coping designs have been tested under in vitro conditions . The frame designs were tested evaluating an abutment contour with flat design, another group with a anatomically guided frame design (similar to that chosen for porcelain-fused-to-metal crowns (PFM)) , and a last group with a PFM design completed by “ribbings” in order to keep the thickness of the overlying porcelain veneering more uniform and to increase the surface for porcelain bonding. Porcelain veneering was made by pressure layering technique and crowns were cemented using a self-adhesive resin cement. The PFM-like framework with ribbings resisted significantly higher loads than flat and PFM-like frame designs and showed a higher percentage of repairable failures. The results of this recent study confirm the importance of a properly designed zirconia framework design, aimed at obtaining a veneering ceramic layer thickness as uniform as possible, never exceeding 2 mm, so reducing the onset of tensile stress and chippings.
8
Type of finishing margins
The effect of different finish line designs on the fatigue, fracture resistance and failure modality of veneered zirconia restorations was evaluated in a study : a complete narrow chamfer, a narrow chamfer with a lingual ledge, and a complete ledge are the three finishing margins that were proposed and analyzed for zirconia crowns. The study showed that the finish line design did not have any statistically significant influence either the fracture resistance or on the failure type of zirconia crowns. In another study it was showed that after experimentally induced veneering fracture the framework remained intact, independently of the finish line (25).
Some evidence is growing that crowns fabricated with digital procedures exhibit better marginal fit compared to conventional techniques .
As to the overall fidelity, many authors reported that marginal discrepancy of single zirconia crowns is comprised in a clinical acceptable range of 0–78 μm whilst FDPs exhibit worse openings, between 9 and 149 μm . However several aspects are not clarified yet.
The studies on marginal discrepancy usually report about mean value of gaps that may be in a clinical acceptable range , but when all recorded values at the margins and the standard deviations are evaluated it is clear that if only the mean value is considered the conclusions can be misinterpreted. Also when external margins fidelity (marginal fit) is compared to internal fit, usually the latter shows higher discrepancies .
Another open issue is how much marginal gap can be accepted clinically. McLean and later Beschnidt and Strub, pointed out 120 μm as a limit for clinically acceptable marginal discrepancies, Holmes placed the ideal marginal gap no more than 50 μm, whilst Audenino gaps between 50 and 300 μm for metal-free restorations. Cagidiaco et al. reported a marginal gap of 20 μs with metal bevel margins and 120 μs with butt joint porcelain margins, but Jahangiri specified that with clinical devices, such as the explorer, only an assessment marginal accuracy over 124 μm can be achieved.
In any case, it must be pointed out that only insufficient data is available about this topic and how much discrepancy at the margins can be accepted clinically is still unclear .
Recently, Navarra et al. showed that after one year of simulated chewing activity, zirconia copings with different marginal designs (feather edge, deep chamfer or slight chamfer) did not show any signs of t–m transformation, neither where the load was applied, nor at the margins.
In conclusion, although being the chamfer the most advocated finishing line of porcelain fused to zirconia crowns, the choice of the marginal design of the preparation should be carefully evaluated, based on the condition of the specific restored tooth.
9
Software efficiency and reliability
In the digital era, efficiency and reliability of the softwares implemented in CAD/CAM systems are among the main factors of success. Many studies have reported the effect of different surface treatments to be potentially detrimental on the flexure strength of zirconia, using polished zirconia as control . However, in the daily practice zirconia can be subjected to different types of surface damage as a result of the CAD/CAM milling procedure and of other routine laboratory procedures.
Softwares can be managed in order to design frameworks, copings, abutments and pontics through a virtual wax-up aimed at creating the final design of milled zirconia pieces. Unfortunately, sometimes the software cannot achieve a proper design and that can introduce internal stress at the interface between zirconia coping and porcelain layer. The milling stresses were considered to be strongly correlated to the incidence of chipping . The results of some other studies reported that an anatomically designed framework could be recommended to minimize the risk of chipping .
Unfortunately, most of the softwares cannot produce different coping designs by default, thus time-consuming procedures are usually needed to obtain this anatomical frame designs. In order to get appropriate anatomical frame designs, wax-up made by the technician and then scanning procedures are still a common procedure.
In conclusion, individual skill in the use of software is a factor that can strongly influence the clinical success of zirconia restorations and needs to be improved in quality, reliability and efficacy.
10
Aging
Aging of zirconia is a synonym for ‘low temperature degradation’ (LTD), a phenomenon due to a spontaneous slow transformation of the crystals from the tetragonal phase to the less stable monoclinic phase in absence of any mechanical stress . Aging is directly related to various factors, like the stabilizer type and content, residual stress and grain size , surface defects, temperature, vapor and processing techniques . In particular, mechanical stress and wetness together (the so-called “hydrothermal stress”) may accelerate the aging of zirconia structures.
Lughi and Sergo , in an engineering-oriented review, suggested that t-zirconia should be prepared and used in a way to make it non-transformable. This would make the material completely unaffected by LTD although with a lower strength. However, the same authors also pointed out that already existing zirconia made with a grain size of 0.5 μs at a temperature range between 70 and 130 °C, in water environment, can have a lifetime projection at 37 °C of 6–12 years, with a predicted degradation to monocline zirconia at the same T ° in 10 years of 21–36%. All these values are not far from a clinical acceptability. It should be taken into consideration that at present there are no clinical reports demonstrating that aging affects the success of zirconia restorations . Moreover, very frequently framework/coping design is made exposing zirconia into the oral environment in order to give proper support to porcelain layers and/or reduce the thickness of the restoration in case of limited prosthetic space: in such cases, up to now, short term clinical observations never showed visible degradation in the oral environment under clinical service.
11
In vivo data
Before discussing some literature reporting a literature review about the available clinical trials and systematic reviews, some aspects strictly related to patient’s view and perception of the received therapies are to be considered. Some authors of systematic reviews defined ‘survival’ as the percentage of restorations remained in situ with or without modifications but still under clinical acceptability, dividing ‘complications’ into biological and mechanical events . However, when zirconia implant-supported restorations are taken into account, reported biological and technical complications are relatively frequent. This, in turn, means that substantial amount of chair time and economical resources have to be accepted by the patient and the dental service following the incorporation of zirconia implant- supported FDPs.
Anusavice defined ‘Success’ the fulfillment of treatment planning goals and expectations of the patient, while ‘failure’ should represent the inability of a restoration to perform as expected under typical clinical and patient conditions . Also Hickel et al. drew special attention to the patients’ dissatisfaction due to failures/replacements of the restorations .
Heintze and Rousson in a systematic review reported that survival of zirconia can be classified in three grades, according to the different dentist’s intervention: grade 1 = polishing, grade 2 = repair, grade 3 = replacement .
Although chipping of ceramic surfaces may seem to be a relatively insignificant event in the field of prosthetic reconstruction, it represents a situation that may lead to psychologic trauma for patients, especially when it affects the anterior region of the dental arch, and to an unexpected expenditure of time and money for dental practitioners. It can represent an unexplainable phenomenon, showing unacceptable level of uncertainty regarding the quality control capability of procedures for fabricating relatively costly prostheses .
In this perspective, chippings of grade 3 in the classification by Heintze and Rousson elicit some considerations: (1) Fracture surface extends into a functional area and repair is not feasible. (2) Recontouring will result in a significant unacceptable alteration of the anatomic form from the original anatomy. (3) Recontouring will significantly increase the risk of pulp trauma by the generation of heat. (4) Repair with a resin composite will result in esthetic changes and unreliable long-term results that are unacceptable by the patient.
In summary, in order to avoid misunderstanding between definition of success and survival, “success” can be defined as the percentage of restorations that remained in situ without any modification, “survival” should be defined as the percentage of restorations that remained in situ with modifications but still under clinical acceptability, whilst “failure” can be defined as the percentage of restorations that needed to be replaced.
From this point of view, some data from clinical trials and systematic reviews have to be considered. Several systematic reviews of literature evaluated clinical performances of zirconia luted restorations . The wider number of published papers was on zirconia bridges and clearly showed the importance of connectors with proper shape and size to avoid fracture of the framework .
Chipping of porcelain was also reported in all studies evaluating both zirconia bridges and single crowns . In the latter a 93% of survival rate was reported whilst the survival rate was between 78% and 100% for FDPs after 5 years of clinical service. In these studies, usually natural teeth were used as abutments and only one paper reported about zirconia restorations luted on titanium abutments . The incidence of framework fracture was directly related to the design of FDPs, where inlay-retained FDPs showed the highest failure rate, about 10% after only 12 months . The most common complication observed in zirconia restorations was chipping of porcelain veneering, with or without exposing the underlying zirconia framework. Higher frequency of chipping was reported in posterior teeth, compared to the anterior .
12
Monolithic zirconia crowns
Fabrication of monolithic crowns is now considered as a possible alternative to porcelain-fused-to-zirconia restorations in order to avoid ceramic chipping, but only a few data is available to date about these prostheses. The performance of crowns fabricated from an experimental monolithic shrinkage-free ZrSiO 4 –ceramic was described in a randomized controlled trial and the 12-month survival rate of 95.1% matched that of gold crowns. This material, when used as a monolithic, not requiring an esthetic veneer, allows lower cost of production. However, the authors did not address the quality of the esthetics that could be achieved with such a technique.
Preliminary observations of monolithic vs porcelain fused to zirconia single crowns in the posterior area are available . A follow-up of these crowns is being performed at the periodical recalls by using a replica technique (rubber impressions) , in order to control the integrity of occlusal surfaces under natural loading and the surface of exposed zirconia areas of porcelain fused to zirconia crowns. The occlusal surface of monolithic crowns after 1 year of clinical service showed a superficial roughness ( Fig. 1 ) that seems higher and deeper than that of zirconia after grinding ( Fig. 2 ). As to the porcelain fused to zirconia crowns, the exposed surface of the coping did not show evident superficial signs of degradation after 1 year of clinical service ( Fig. 3 ).
13
Correlation between in vitro and in vivo of zirconia restorations
Unfortunately, the wide majority of clinical trials available in the literature do not report in detail the specific laboratories procedures by which restorations were made. For this reason, it is hardly possible to correlate results of in vitro investigations with in vivo performances, such as success, survival and failure rates. As seen before, many laboratory parameters have a direct influence on the clinical performance and on the possible complications, e.g. chipping of porcelain layers. While these parameters are usually reported and discussed in in vitro tests, no one of the clinical trials available in the literature reported detailed technical specifications: how zirconia was milled and processed, the used veneering technique by which porcelain was layered and fired, the thickness of porcelain and zirconia and how such thickness was kept under control during laboratory steps, the temperature programs of firing and cooling, the shape of copings and/or frameworks.
Recently, three different CAD/CAM shapes of zirconia copings for single-unit porcelain-fused-to-zirconia crowns were tested, both under in vitro and in vivo conditions , in order to evaluate their fracture resistance. After three years of clinical service, it was noticed that the group with flat design showed fractures of porcelain in 6 cases of 30 whilst the two other tested groups reported no chipping at all. These promising results were not in complete agreement with those obtained under in vitro conditions for one of the three tested groups; this could be explained considering the difficulty to reproduce the real clinical environment in the laboratory and to standardize all the different variables. These two studies, planned and performed in parallel, showed that it is possible to correlate only partially, under certain limitations, the results of in vitro and in vivo investigations.
14
Zirconia abutments
In the last decade the use of zirconia as a material for implant abutments has been increasingly widespread, for its white color, differently from metal abutments, for its high level of biocompatibility and for its superior mechanical properties, compared to the dental ceramics.
The increased worlwide acceptance of zirconia abutments is mainly due to esthetic implications; they elicit a good chromatic response when translucent, metal free restorations are luted onto them and, at the same time, in case of soft tissue recession, no titanium is exposed. Moreover, zirconia abutments allow to avoid the grayish appearance exhibited by the titanium abutments through the soft tissues; however, it has been shown that when the thickness of mucosa covering titanium or zirconia abutments exceeds 2 mm the difference in color and light reflection of soft tissues is no longer noticeable for the human eye .
To date, evidence is growing that zirconia provides good biological responses, as showed in the past for titanium abutments .
As regards the clinical service, the performance comparison between metal and zirconia abutments is not easy at the moment, due to a lack of medium- to long-term data for the latter .
From in vitro tests, the main significant data show that the type of connection influences the mechanical strength of zirconia abutments, in that a superior structural resistance can be achieved by means of internal connection via a secondary metallic component; moreover, they show that the use of a secondary metallic component might have a beneficial influence on the stability of zirconia abutments, and that the use of implant-prosthetic zirconia abutments in the molar area is not recommended . Although the strength of both tested systems seems to be adequate to resist physiologic chewing forces in the premolar area , zirconia abutments show significantly lower fracture strength than titanium abutments .
In agreement with these results, Foong et al. pointed out that one-piece zirconia abutments exhibited a significantly lower fracture resistance than titanium. The mode of failure is specific to the abutment material and design, with the zirconia abutment fracturing before the retentive abutment screw.
Some papers on clinical performances of zirconia abutments are available in the literature .
In three articles concerning a clinical trial at 1-, 3- and 5-year, survival rates estimated from annual failure rates appeared to be similar for ceramic and metal abutments . However, the information about zirconia abutments of this clinical trial reported at three different recalls were limited to one type of external exagon abutment, in a small number of samples (18 patients with 18 zirconia and 12 titanium abutments), with a high patient drop-out (30% at 3–5 years) and a high variability in the location of abutments (zirconia group: 2 canines, 11 premolars and 5 molars; titanium group: 2 canines and 10 peremolars).
Glauser et al. in a prospective clinical study, reported that zirconia abutments offer sufficient stability to support implant-supported single-tooth reconstructions in anterior and premolar regions, with a very favorable soft and hard tissue response. However, also in this study the drop-out was pretty high, at 4-year recall (up 30%).
Similarly, very positive and promising results were reported by Zembic et al. at 11-year on zirconia implant abutments in anterior and posterior regions, showing 96.3% of cumulative success rate, but with a drop-out around 24% and with a scattered distribution in the mouth (at the baseline, 25 incisors, 14 canines, 15 premolars in both jaws).
Recently, the first prospective clinical trial on zirconia and titanium abutments with internal exagon has been carried out . At a three-year recall, titanium abutments success rate was 100%, whilst that of zirconia abutments was 88%. Fracture of internal connections reported in all zirconia failed abutments was possibly due to imprecision of milling or micromotion, addressing the risk of a possible internal deformation of the implant shape, as evidenced by metallic discoloration on retrieved abutments.
Untouched zirconia abutment connections were also observed microscopically, in order to evaluate the thickness and consistency of shape of the exagon itself ( Fig. 4 ). The connection thickness appears to be sufficient for titanium abutments, but is seems weak for zirconia abutements based on the flexural strength available data. More specifically, the thickness in the thinner area measured at the exagon was, about 230 μm; assuming 1000 Mpa as flexural resistance of dental zirconia after sintering, this determines a 11.5 N fracture load limit (appx) (1300 MPa to 15 N appx). The mean of the 6 recorded thickness was 337 μm, (appx 30 N for 1000 MPa and 40 N for 1300 MPa). In the thinnest measured area of the six parts (to be considered a locus minor resistentiae ) the thickness was 177 μm for a fracture load limit in the range of 9.5–11 N. It looks clear that such a thin thickness of zirconia at the exagonal connection with the fixture can not resist the torque stress to 25 N/cm as well as stand the occlusal stresses ( Fig. 4 ).
