There were four main mechanical outcomes that were discussed in the present systematic review reporting a higher incidence of fracture complication in ceramic abutments compared to titanium ones [19]. The inherent properties of ceramic materials, with a lower resistance to fracture and less flexural strength when compared to metals, may explain these findings [19,38]. The risk of abutment fracture may also be affected by the thickness of the material and the position and angulation of the implant with respect to the final prosthetic restoration [19]. To decrease the risk of an abutment fracture, the wall thickness of zirconia abutments must be maintained above 0.5 mm during manufacturing, while titanium has to be preferred when the thickness is less than 0.5 mm [39]. Manufacturers have also restricted the indications for zirconia abutments to limited angulation. Stock abutments provide maximum angulation of 15 to 20 degrees, while CAD-CAM custom abutments are not recommended for an angulation of over 30 degrees [40]. The abutment fracture rate of zirconia varies from 1.08% to 17.86% depending on age, gender, tooth position, abutment systems, implant systems and implant–abutment connections. Abutment fractures are more likely to happen in young male patients, most probably due to higher occlusal forces. In addition, zirconia abutments in the posterior areas seem to be more susceptible to fractures, compared to placement in the anterior areas [39].

Implant–abutment connection design also plays a huge role in the success of zirconia abutments. A one-piece internal connection showed the highest fracture rate, while external and two-piece internal connections reported a lower fracture rate [39]. Most zirconia abutment fractures occurred within 3 years after loading and several occurred during the initial fitting or subsequent tightening under a controlled torque and the primary fracture location can vary depending on the implant–abutment connection design. In one-piece internal connections, fractures mostly happen in the neck of the implant or below the implant shoulder. The region around the abutment screw head experiences the highest torque and stress concentration and is the most critical region for the stability of ceramic abutments, while the zirconia portion below the implant shoulder is the thinnest and, therefore, cannot resist high torque values. In two-piece internal connection and external connection zirconia abutments, fractures are mostly found above the implant shoulder [39].

An abutment screw fracture was mentioned only in one systematic review and, therefore, seems to be a very rare mechanical complication [42]. Irrespective of the abutment material, abutment screw loosening was found mostly in studies using external hex implants for single implant restorations. Although abutment screw loosening may not be considered a failure, repeated screw loosening can affect the success of implant therapy and patient satisfaction [42]. It has been shown that implant screws are submitted to high concentrations of stress during mastication in the posterior area and this is the most critical area for the stability of ceramic abutments. Nothdurft et al. [45] observed that screw loosening on zirconia abutments may occur due to wear induced by the unavoidable micromovements of the prosthetic component at the implant–abutment interface, which is maximized by the difference in the material properties. Ceramic abutments with titanium bases, or a two-piece Y-TZP abutment, have been proposed as an alternative to solve this problematic wear. Moreover, two-piece abutments have a reduced tendency for complications because they have a higher resistance to micromovements due to ductility of the titanium, which allows some deformation when submitted to unfavorable movements [35]. It was observed that failure in the veneer layer on zirconia abutments was considered as the most frequent problem, especially in all-ceramic restorations. Veneer failure, however, may not lead to implant or prosthetic failure, depending on the extension of the fracture, compromised esthetics or function of the prosthesis. This complication may strongly affect patient comfort and satisfaction as it may increase the treatment time, costs and complexity of maintenance. Veneer failure occurs mostly due to weak bonding between the veneering ceramic and the zirconia infrastructure [35]. When comparing zirconia and titanium abutments, apparently the incidence of a veneer failure on single crowns cemented on titanium abutments was relatively higher compared to zirconia abutments, and this may be associated with the inherent and unavoidable differences in the mechanical properties of ceramic and titanium [35,38].

The type of prostheses and the implant connection are other factors that may influence the outcome [46].

Zirconia abutments presented a better performance in the overdenture group compared to titanium abutments; however, only two original studies reported about implants that were restored with implant-supported overdentures [37].

Although according to several studies the success rates of zirconia and titanium abutments were very similar, further studies need to be performed for two-piece abutments, or hybrid zirconia–titanium abutments. In particular, further clinical research about the mechanical outcomes of this implant abutment design is needed, as some published studies were not specific on how the titanium and zirconia were bonded together as a single unit implant abutment.

As for the biological outcomes, plenty of outcomes were extracted, mostly regarding the material effect towards peri-implant soft tissue. Although zirconia and titanium have both proven to be biocompatible with the surrounding tissue, biological complications were reported in several studies. Buccal fistulas were involved in both screw and cement-retained restorations. In screw-retained restorations, this was only seen in external hexagon implants, and might have been caused by the gap between an ill-fitting abutment and an implant [34]. Increased leakage was reported for zirconia abutments compared to titanium abutments, possibly due to the lower recommended torque values used to tighten the zirconia abutment [47]. In cement-retained restorations, the fistulas were attributed to uncleaned residual cement [16,34]. The design of the abutment might also affect the amount and incidence of cement remnants in peri-implant sulcus. In fact, when the crown margins are located between 1–1.5 mm submucosally, they prevent the complete removal of cement remnants, even when using customized abutments. Resin cement is the most difficult to remove from abutments; therefore, it may be assumed that this complication is dependent on the abutment design and cementation agent, rather than on the abutment material. Hence, supragingival or epigingival margins of abutments are suggested, especially if the implant restorations are to be cemented with resin luting agents [16].

In a systematic review by Sanz-Martín et al. [1], it was reported that there were cases of suppuration without bone loss with no further explanation mentioning the cause of the complication.

The overall survival rate of the two materials shows no significant difference [1,37], but further study with longer follow-ups might be necessary to determine which of the two materials has a better survival rate.

Peri-implant mucosal recession is also considered a biological complication and was reported in several systematic reviews. This complication was predominantly reported in studies using prefabricated titanium abutments. It may be related to the fact that prefabricated abutments provide less optimal support to gingival tissues compared to customized abutments. Regarding this complication, titanium abutments have been reported in more studies because of their longer period of usage, and recession related to titanium abutments can be easily seen and recorded compared to ceramic abutments [34]. Another systematic review stated that soft tissue recession was not influenced by the selection of the abutment material, but by other important factors such as the 3D position of the implant and the presence or absence of attached mucosa [16]. This result was not in accordance with the systematic review of Cao et al. [37] that revealed that zirconia abutments could better reduce the recession in both the junctional epithelium and alveolar ridge compared to titanium abutments, although the reduction was limited [37].

The implant abutment shape design has shown effects on peri-implant soft tissue. In fact, some outcomes are more dependent on the design rather than the material of the abutment. It was reported that a comparison between concave and convex abutments showed that the soft tissue thickness was greater in the concave group [39]. The use of the concave transmucosal profiles for implant components allowed for more predictable soft tissue stability in esthetic areas. In a study by Rompen et al. [48], concave abutments of both titanium and zirconia material were used for single crown restorations, and their results showed that 87% of the sites showed facial soft tissue stability or a vertical gain, and recession in the remaining 13% was below 0.5 mm [34].

A difference was found for mucosal inflammation or bleeding on probing (BOP) when comparing abutment materials. Some systematic reviews found that mucosal inflammation was greater next to titanium abutments compared to zirconia ones [33,37,39], while the macroscopic design, surface topography or surface manipulation did not have a significant influence on this outcome [39]. Lower BOP values may be caused by less plaque retention in zirconia abutments compared to titanium ones due to the surface properties of the material [34,35,38,39]. It must be added that it is more and more common to treat patients with anticoagulant therapy [49].

This finding is also supported by an in vivo study where the bacterial colonization was compared between zirconia and titanium discs attached to a removable dental prosthesis [16,39]. Ceramic abutments are easier to polish than metal abutments, which can greatly reduce the bacterial colonization and adhesion [38]. They are also resistant to corrosion which allows for the better growth of epithelial cells and inhibits bacterial adhesion. Plaque accumulation can also be affected by several other factors besides implant abutment modifications including a cement excess, and a misfit between the prosthesis and the implant platform caused by screw loosening or ill-fitting prosthetic components [35]. Soft tissue attachment also plays a role in the degree of inflammation. In an in vitro study [42], zirconia exhibited a higher degree of fibroblast proliferation when compared to titanium. This result, however, does not translate to the differential histological outcomes in experimental studies comparing abutments made of zirconia, titanium and gold alloy. A similar soft tissue dimension was reported for titanium and zirconia abutments, while in the gold alloy abutments there was an apical shift of the epithelium barrier followed by a marginal bone loss [34,39]. It was reported that healing at zirconia and titanium abutments allowed the formation of a mucosal attachment that included an epithelial and a connective tissue portion that were about 2 mm and 1–1.5 mm high, respectively, which demonstrates that the abutments made of titanium and zirconia promoted the proper conditions for soft tissue healing [14].

Other than the different materials, the surface properties and in particular the roughness may have effects on soft tissue attachment [31]. Some histological studies on animals and humans have demonstrated that moderately rough surfaces can be beneficial for soft tissue integration, but this conclusion cannot be drawn as there is insufficient investigation on this hypothesis [31]. No significant difference was found for the periodontal probing depth between zirconia and titanium implant abutment material; however, in some studies, the cementation margin was placed subgingivally, which could be a setback in the study design. Clearly, if the restoration margin extended deeper subgingivally, the peri-implant tissues at the gingival parameters would contact the restoration material instead of the abutment material. This would influence the periodontal probing depth, accumulation of plaque and other biological parameters [17].

Titanium and zirconia abutments did not seem to influence marginal bone loss. Although zirconia abutments seem to have less marginal bone loss compared to titanium abutments, there was no significant difference between the two materials [20,38]; however, when changes in the marginal bone loss were assessed over time, a significant loss occurred for both materials except for titanium nitride. The magnitude of this loss, with a mean follow-up of 30 months, has limited clinical significance as it is smaller than the mean error of repeated radiographic measurements [20]. Surface decontamination on implant abutments seems to be an important factor affecting the peri-implant bone levels. A greater amount of bone loss was reported for steamed titanium abutments compared to plasma argon titanium abutments. It is known that the strong affinity of titanium to proteins and amino acids makes the complete cleaning of its surface rather difficult, but plasma argon cleaning has been shown to effectively decontaminate titanium surfaces in vitro. In human histological studies, it has been reported that plasma argon may promote cell adhesion and positively influence collagen fiber orientation [39].

The present systematic review of systematic reviews presents some limitations as some of the systematic reviews that were included faced their own difficulties in collecting data with less heterogeneity and bias in the studies included. The lack of differences for some clinical outcomes analyzed may also be due to the questioned reliability of periodontal parameters to assess peri-implant health. Different studies also use different methods to measure those parameters that may affect the overall results of the investigation. It has been shown that factors such as gender, implant position, or age can influence the peri-implant health parameters. Moreover, excessive probing forces may induce false-positive BOP readings and the insertion of a periodontal probe in cases with overhanging restorations may result in lower periodontal probing depth values, increasing the trauma towards the peri-implant soft tissue and leading to a false-positive BOP [39].

Another important factor to be considered for the abutment selection is its possible impact on the esthetic outcome of the implant–supported final restoration. The systematic reviews that were included had selected studies with many different esthetic indices used. Studies using spectrophotometry have shown better esthetic results with zirconia abutments compared to titanium abutments, mainly due to the color appearance of peri-implant soft tissue [20]. The blue-grayish shimmering effect of titanium abutments, especially in the case of thin peri-implant mucosal tissues, can compromise the esthetic result [34]. Hu, et al. found no difference in the discoloration of soft tissues with different abutment materials [38]. Peri-implant mucosal thickness is also of importance to achieve satisfactory results, as it has been shown that the abutment material determines minimal color changes in thicker tissue, generally more than 3 mm [17,20,41]. It is recommended to use titanium abutments when the tissue thickness is at least 3 mm, while for zirconia implant abutments, a 2 mm of soft tissue thickness would be enough [41]. Hence, the use of zirconia abutments can be esthetically appreciated only in the case of thin tissue biotype [17]. Additionally, titanium abutments with modified surfaces, such as anodized ones, seem to have more discoloration compared to zirconia [40,41]. Anodized pink titanium abutments present a “grayish” color as they are all-metallic while ceramics present a more “whitish” color [41]. The pink veneering of zirconia seems to be more promising according to spectrophotometric measurement without a significant biological or technical alteration; however, further study is needed [40]. When different abutment materials were compared to natural teeth, both titanium and zirconia abutments induced visible discoloration when assessed using a spectrophotometer in the peri-implant soft tissue [34,41]. It appears that studies using spectrophotometric analysis showed a higher sensitivity to detect peri-implant mucosal discoloration, whereas studies using subjective or objective scoring criteria reported a minimal difference in the esthetic outcomes and patient satisfaction [34]. According to Linkevicius et al. [17], the pink esthetic score (PES) index should be used for the evaluation of the final esthetic outcome because it reflects the soft tissue condition better. The PES was originally proposed by Furhauser et al. [50] and it is composed of five factors, namely, the mesial papilla, distal papilla, curvature of the facial mucosa, level of facial mucosa, and root convexity or soft tissue color and texture at the facial aspect of the implant site. For example, the PES score around zirconia abutments was significantly higher compared to titanium abutments at a 2 year follow-up [17].

There were limitations in the esthetic assessments as well because different systematic reviews included different inclusion and exclusion criteria. In those systematic reviews, different indices were used, hence, the outcome can be greatly influenced when different measurements are taken into consideration.

Additionally, some studies had taken measurements before the cementation of a prosthesis or crown, while others assessed the outcome without prosthesis.

Future research should focus on improving the study methodology in this field of implant dentistry. Although research on implant abutment materials and their modifications have been published for many years, many clinicians still have difficulty in deciding on the abutment material and design. The possible variables in the choice of the implant abutment seems to be limitless, including the material, macroscopic design, surface topography, surface treatment and even the modification of color by veneering in zirconia abutments and anodizing in titanium ones.

Other than that, implant abutment can also be influenced by how a prosthesis is placed, namely, cemented or screw retained. There is a wide variety of factors that play a role in peri-implant soft and hard tissue health.