A structured qualitative systematic review following the PRISMA (Preferred Reporting Item for Systematic Review and Meta-Analyses) recommendations and the PRISMA checklist [4] [5] .

We have tried to answer the following question: does bacterial adhesion on lithium disilicate ceramics differ from that on other dental ceramics, and what factors influence bacterial adhesion?

The population, intervention, comparison, outcomes, and study designs for this systematic review are defined in Table 1.

Studies (in situ or in vivo biofilm) that evaluated bacterial adhesion to dental ceramics, (including lithium disilicate ceramics). Articles published between 2009 and 2022.

Exclusion criteria: literature reviews, Systematic reviews; articles published before 2009. Studies not dealing with lithium disilicate ceramics were excluded. Studies not meeting the objectives of our work based on abstract reading and critical reading of the full text were also excluded.

Individual search strategies were developed for each of the following electronic databases: SCOPUS, PubMed/Medline, Google Scholar, and Science direct. These databases were last consulted on: 24/03/2023.

Key words: Based on the PICO sections (Table 1), we first identified the main concepts of our thesis topic: Concept 1: bacterial adhesion, concept 2: dental ceramics. In a second step, we found the keywords via these concepts: biofilm, bacterial adhesion, dental ceramics, lithium disilicate.

The first search was performed by a single author (R.S) who was responsible for eliminating several articles according to the inclusion criteria already specified, then the selection of studies and quality assessment were carried out independently by two readers (S.M; R.S) so as to reduce the risk of excluding relevant studies, minimize the risk of error of judgment and subjectivity, and ensure reproducibility of results. In the event of a difference of opinion, the articles concerned were discussed between the two readers in order to reach a consensus.

The quality and risk for bias of the included studies were assessed by the quasi-experimental studies appraisal tool by the Joanna Briggs Institute that had been adapted for another systematic review of in vitro studies [6] . This step enabled the final selection of potentially eligible articles. This risk-of-bias assessment stage enabled the final selection of potentially eligible articles.

Figure 1 shows an overview of the study selection procedure. A total of 701 studies were identified in the initial survey. After reading the titles and abstracts, we

were left with a total of 117 articles. We excluded 54 articles that did not comply with the inclusion criteria. We then excluded 49 articles after full-text evaluation of 63 articles, and finally retained those relevant to our systematic review (14 articles).

The results of the quality evaluation of the studies are summarized in Table 2. The risk of bias was evaluated with the use of an adapted quasi-experimental studies appraisal tool by the Joanna Briggs Institute. Of the 14 studies included in this systematic review, most had a low risk of bias, with the exception of questions 5 and 8. Most answers to question 5 (Were there multiple measurements of the outcome both pre and post the intervention/exposure?) were uncertain, we can explain this by the fact that we chose studies without intervention on the ceramics used. Regarding question 8 (Were outcomes measured in a reliable way?), most answers were uncertain, because most studies did not provide information on the number of assessors, assessor training, intra-assessor reliability and inter-assessor reliability within the study, this criteria presents a high risk of bias. Five studies [7] [8] [9] [10] [11] are no control groups.

After reading the full text of the articles, we tried to extract the results of most

interest to our study, the results of the articles are summarized in Table 3 and Table 4, those tables includes 11 in vitro studies and 3 in vivo studies, the table provides several information about the studies namely: Name of authors, purpose of the study, ceramics used, control group, tests used, types of bacteria, results and conclusions.

Material composition and surface properties can influence initial bacterial adhesion and compromise dental and periodontal health [12] .

According to the study conducted by Diane T. Vo et al. [13] , lithium disilicate ceramics (Press and CAD) showed no significant differences in bacterial adhesion, but these ceramics were significantly lower than the samples (ZirPress/Ceram), which were significantly lower than the samples (Ceram Glaze), the authors linked these differences to manufacturing factors, Indeed, the glaze application has the roughest surface finish due to the absence of post-fabrication polishing, and the ZirPress samples have been modified by a technique involving the addition of a fluorapatite glazing ceramic, which may explain the non-uniform surface characteristics and the introduction of an additional surface. The manufacturing technique therefore influences the surface properties of lithium disilicate as well as bacterial adhesion.

However, a study conducted by Contreras LPC et al. concluded that ceramic manufacturing technique does not influence surface properties such as surface free energy [14] .

Ra: Surface roughness; B.A: Bacterial adhesion.

Stephanie Francoi poole et al. [15] , showed that bacterial growth and adhesion were similar in the lithium disilicate and zirconia-reinforced lithium disilicate glass-ceramic groups, however, the monolithic zirconia group revealed a greater susceptibility to bacterial adhesion, This can be explained by the composition of zirconia, which differs from lithium disilicate-based glass-ceramics: zirconia is a glass-free ceramic, a polycrystalline ceramic with a higher percentage of crystalline content and larger crystal size than lithium disilicate-based glass-ceramics. These hypotheses are the same in other published studies [16] [17] .

Ra: Surface roughness; B.A: Bacterial adhesion.

Maciej Dobrzynski et al. have reported that initial bacterial adhesion can be influenced by material composition, and that lithium disilicate glass-ceramics have good anti-adhesive properties [18] . Nevertheless, L. Viitaniemi et al. showed that lithium disilicate exhibited lower bacterial adhesion than the other materials examined in their study.

Shlomo et al. [19] concluded from their results that lithium disilicate may be less susceptible to bacterial adhesion than zirconia.

In contrast, 5 studies [20] [21] [22] [23] [24] have shown that lithium disilicate ceramics exhibit higher bacterial adhesion than zirconia. The glass content has been suggested as a factor responsible for differences in bacterial adhesion, the results of one study [25] reinforcing this hypothesis since it showed that feldspathic ceramics, containing more glass than lithium disilicate exhibited higher bacterial adhesion than lithium disilicate and in another study [26] the control material (glass plate) showed higher values for bacterial adhesion than all ceramic materials.

However, no conclusions can be drawn since in previous studies [15] [18] , zirconia showed higher bacterial adhesion even though it has no glassy phases. In this respect, further studies are needed to determine whether there is a correlation between bacterial adhesion and the glass content of the ceramic.

Comparison of the studies was difficult for several reasons: firstly, the techniques used to assess bacterial adhesion and the surface roughness of the materials studied are different; secondly, we only included three in vivo studies, as in vitro measurements are not always capable of simulating the complex conditions present in the oral environment.

A rougher surface and complicated topography present a greater affinity for bacterial adhesion than smoother surfaces, and therefore greater difficulty in completely removing biofilm by mechanical brushing [27] . Surface roughness seems to affect only the number of bacteria in the biofilm, not the species; a rough surface increases the surface area available for colonization compared with a smooth surface [28] . In addition, the crevices created by roughness provide shelter for bacteria, giving them time to secure their attachment to the film.

In our study, 9 studies investigated the association between surface roughness and bacterial adhesion, with six studies [12] [13] [15] [24] [29] showing that surface roughness favors bacterial adhesion. In a literature review by Bollen [30] , a maximum surface roughness of Ra = 0.2 μm was suggested as a threshold value for bacterial retention; below this value, no further reduction was observed, while above this value, biofilm accumulation increases with roughness. A systematic review by Wim Teughels et al. concluded that an increase in surface roughness above the Ra threshold of 0.2 μm facilitates biofilm formation on restorative materials 29.

Other studies reported that surfaces with a roughness threshold above 0.2 μm showed no difference in biofilm formation [31] [32] . Meier et al. found that a five-fold increase in roughness did not result in a greater number of adherent bacteria [33] , in our work, three studies [19] [23] [26] among the included studies showed that there is no correlation between surface roughness and bacterial adhesion.

It has been noted that the ceramic finishing procedure (polishing and/or glazing) can lead to changes in free energy, with the same material exhibiting changes in surface free energy depending on the finishing and polishing procedures applied [14] [34] . Indeed, the microbiota present in the oral environment has a high free energy and adheres preferentially to high free energy substrates.

Several studies [15] [18] [25] [29] have investigated the influence of polishing techniques on bacterial adhesion. AMO Dal Piva et al. [12] , showed in their study that glazed surfaces have greater surface roughness and tend to accumulate more biofilm. This agrees with the results of another study by Scotti R et al. [35] . Another study showed that the presence of glaze on the surface does not prevent the formation of dental biofilm.

Maciej Dobrzynski et al. [18] , showed in their study that differences in the adhesion of micro-organisms to the surface of polished and unpolished ceramics were statistically significant, with unpolished surfaces being more susceptible to adhesion by micro-organisms. This is consistent with the results of several studies [36] [37] . Stephanie et al. [15] also reported that grinding with diamond burs resulted in greater roughness on ceramic surfaces, and the surface roughness of ceramic materials appeared to favor susceptibility to P. intermedia adhesion.

Consequently, grinding with diamond burs is not recommended, to minimize the risk of bacterial adhesion [25] .

In one of the studies selected for our systematic review, Patrcia et al. [29] concluded that surface treatment influences bacterial adhesion, in fact, the unpolished surface showed the highest bacterial adhesion among the groups studied, they also concluded that mechanical polishing achieved a lower surface roughness than glazing as well as minimal morphological changes to the ceramic.

Within our systematic review, the three in vivo studies [24] [25] [29] used a polymicrobial biofilm to assess microbial adhesion to ceramics, after the formation of the acquired exogenous film (AEP), pioneer colonizing microorganisms adhere to this film, creating a basic or immature biofilm, and vary according to the environment and materials to which they adhere. After the pioneer colonizers, the secondary colonizers follow, composed of various species depending on the bacterial composition of the environment; these species can be used to predict the bacterial composition of the mature biofilm. The mature stage of oral biofilm generally occurs after 3 to 5 days [24] .

Pioneer colonizers of oral biofilms have been identified as Streptococcus sanguinis, S. oralis, S. gordonii, S. mitis, S. mutans, S. sobrinus, Actinomyces naeslundii and Capnocytophaga ochracea [15] .

The in vivo studies selected in our work used Streptococcus mutans [12] [13] [18] [23] [25] [29] [38] , Streptococcus sanguinis [12] [19] [22] [26] , Streptococcus gordonii [21] [26] , Streptococcus oralis [26] , Candida albicans [12] [18] [29] and Prevotella intermedia [15] as pioneering colonizers.

Streptococcus sanguinis is considered the initial colonizer, while S.mutans is considered the colonizer associated with the development of carious lesions, also facilitating the adhesion of more complex pathogenic bacteria. C. albicans is considered a colonizer associated with caries, periodontal disease and candidiasis [19] .

Streptococcus mutans is the most widespread bacteria in the oral environment. It is the most adhesive of all bacteria, irrespective of ceramic type and finishing method [12] [18] .

With regard to P.intermedia, Stephanie et al. [15] concluded in their study that grinding with diamond burs results in high roughness on ceramic surfaces, and the roughness of ceramic material surfaces seems to favor the susceptibility to adhesion of P. intermedia, Gram-negative bacteria predominantly have a higher surface energy, between 35 and 65 mN/m, while most Gram-positive bacteria have lower values, between 0 and 25 mN/m [39] . The closer the surface free energy of the material and microorganism, the greater the probability of adhesion [40] .

Some studies show that certain bacterial species interact preferentially with certain materials, depending on differences in the physico-chemical properties of the surface [18] [41] . Moreover, bacterial adhesion to a given substrate depends on both the hydrophobicity of the surface and the hydrophobicity of the bacteria. Thus, hydrophobic bacteria such as S. mutans, S. oralis and S. sanguinis adhere more readily to hydrophobic surfaces, while hydrophilic bacteria such as S. mitis adhere more readily to hydrophilic surfaces [42] .

A number of approaches are being studied to interrupt biofilm formation or prevent its spread. Some strategies are based on materials engineering, where anti-adhesive surfaces are created or antibacterial additives are incorporated into substrates. Other research focuses on ways of acting directly on bacteria, either by inhibiting the quorum sensing system, preventing extracellular matrix formation and, among other things, inhibiting secondary messenger signaling pathways.

With regard to the material’s topography, it was seen thath characteristics concerning the size, shape and distribution of roughness patterns affect both attachment and biofilm formation of different bacterial strains on various substrates. Bacterial adhesion decreases as the size of the topographic pattern get smaller, and in this sense, topographies on a micron-scale mainly affect bacterial attachment, while topographies on a nanoscale can have bactericidal effects [43] .

With regard to the incorporation of antimicrobial agents into biomaterials, particularly in dentistry, various nanoparticles and agents have been incorporated to prevent biofilm formation without altering physicochemical and mechanical properties [44] , such as silver-vanadate (AgVO3), which is an antimicrobial that has the advantage of being stable and not forming agglomerations. It inhibited the growth of Candida albicans, Streptococcus mutans, Staphylococcus aureus and Pseudomonas aeruginosa.

Although there are many antibacterial agents applied to reduce biofilm formation, the current context still shows species resistant to antimicrobial therapies, so more studies are needed to control or reduce pathogenic biofilms by seeking new products and techniques. In this respect, the efforts of materials engineering are essential [43] .