The inclusion and exclusion criteria were framed based on the PICOS model. In vivo studies and clinical trials, and in vitro studies that investigated both biocompatibility and bioactivity were included in the assessment. Studies applying Ti or its alloys coated with bioactive glasses or the composite coatings on the Ti implant containing bioactive glasses were included in this review. The outcome indicators discussed in this review contain biocompatibility, bioactivity, and antibacterial properties, which are shown in Figure 1. Biocompatibility refers to 1) cytotoxicity and cell activity, 2) hemocompatibility, 3) anti-inflammatory properties, and 4) bioactivity. Bioactivity comprises 1) osteointegration, 2) osteogenesis, 3) angiogenesis, and 4) soft tissue adhesion.

Articles published in the last 10 years were included. There are no restrictions on the research type. Articles not published in English and whose full texts were unavailable were excluded.

Four databases, namely, PubMed, Embase, Scopus, and Web of Science, were searched in this study. The search strategy is shown in Table 1.

Duplicates were removed using Endnote X9.3.2. The first screening was performed by filtering the title and abstract and inclusion of studies was determined after reading the full texts. The screening was done following the inclusion and exclusion criteria and was performed by two independent authors. Any conflict was resolved by a third author.

Data were collected using Microsoft Excel. The extracted data included substance material and samples’ shapes and sizes, glass models and composites, experimental subjects, manufacturing methods, and special process parameters. Other included information can be seen in detail in the following contents.

The data were extracted independently by two researchers, and any problems were solved through discussion and a third author's help.

The process of literature screening is shown in Figure 2. After screening, 49 articles were included in this systematic review. The characteristics of the included studies are shown in Table 2.

Ti-6Al-4V, which has been widely applied in orthopedic prostheses and dental implants, is the most widely applied alloy in the included studies. Pure Ti has also been investigated. Bioactive glasses 45S5, S35P5 (Massera et al., 2012), and 58S (Sepulveda et al., 2002) are the most commonly studied basic bioactive glass.

Substrate pretreatment plays an important role in bio-interaction (Pattanaik et al., 2012), which can improve corrosion resistance (Kim et al., 1996) and osteointegration (Buser et al., 2012; Fischer and Stenberg, 2012). Surface roughness can also be increased as well as the adhesion between substrate and coating. Sandblasting is a simple, low-cost method (Zhou et al., 2016). Aluminum oxide and silicon carbon can be injected onto the substrate using high-speed compressed air, which improves the surface roughness. Sandblasting combined with acid etching forms a microporous structure and removes the residual abrasive particles (Pattanaik et al., 2012). Chemical pretreatments are also performed. Alkali and heat-treated implants, which apply hydroxide at high temperatures, form a titanate layer on the Ti surface. This improves the connection between the bone tissue and the implant and enables higher implant stability (Nishiguchi et al., 2003). The porous nanostructures can also increase the bond strength between coatings and substrates (Nesabi et al., 2021). Surface topology is an important surface structure that can effectively regulate the behavior of cells. Numerous studies have shown that rough or micro-/nano-sized topological structures can effectively improve cell behavior, thereby enhancing the integration ability between implants and bone interfaces (Geng et al., 2021a). Micro-arc oxidation can generate a uniform, rough, and porous oxide layer, which contributes to a tighter connection between substrate and coatings (Ma et al., 2016). Among the different coating technologies, polishing and sandblasting the titanium substrate to increase its surface roughness, followed by cleaning the substrate with distilled water, acetone, or ethanol, are the more common pretreatment procedures.

PLD is performed under confined conditions. The schematic diagram is shown in Figure 3A. The pulsed molecular laser source can be used to sinter bioactive glass. In a physical coating preparation process, the precise ratio of the coating components can be guaranteed, and the prepared coatings are more uniformly attached. To ensure uniform adhesion and prevent laser single-point substrate surface corrosion, the target is usually coated in a rotary manner (Ma et al., 2016; Wang et al., 2018; Wang et al., 2020).

Table 6 shows the process parameters of the included studies. In the included studies, PLD was usually performed under an environmental temperature of 200°C–800°C. The substrate temperature mainly regulates crystal composition and physiochemical and biological properties by changing the alignments of the target material on the surface and the bonding level of the coating (Figure 4) (Serra et al., 2004; Wang et al., 2020). When the substrate temperature reaches 200°C, the BG coating gains the best mechanical properties and surface appearance and can best bond with the substrates (Serra et al., 2004; Zhao et al., 2008). However, another study has proven that coatings formed under 700°C have the best biological properties (Dhinasekaran et al., 2021). To reduce the overlap between the laser and vapor that it generates, the angle of incidence of the laser projection on the substrate is maintained at 45° (Shaikh et al., 2019).

Coatings formed by PLD have a dense surface, a better chemical composition ratio of BG (Ledda et al., 2016), and a microsphere (Palangadan et al., 2014). The nanostructure on the surface can help in implant osteointegration (Rau et al., 2020). PLD mixed with micro-arc oxidation enables the porous morphology of coatings, which results in a better surface appearance and structure (Ma et al., 2016).

Magnetron sputtering is usually performed under low-pressure conditions, where the deposition chamber is filled with gas, which usually comprises argon (Ar) (Figure 3B). Ar atoms are ionized to produce Ar⁺ and new electrons, with the electronic field accelerating the electrons that bombard the targets and sputter the ions from the target atoms. After deposition, the substrate is heated to optimize the coatings (Shi et al., 2008).

The distance between the substrate and target may influence the deposition rate. A closer distance means a higher deposition rate. However, the increased momentum of the charged particles leads to an increase in the substrate temperature (Maximov et al., 2021). The acceleration voltage, heat treatment time, vacuum pressure, and filling gas also have an impact on the process (Wolke et al., 2005; Popa et al., 2017).

Magnetron sputtering makes the thickness of the coating uniform and adjustable and results in higher adhesion and purity of the coating (Shi et al., 2008), which makes it suitable for covering large areas of the substrate, and the technology is easily scalable to the industrial level under alternating current conditions (Popa et al., 2015).

Dip coating technology is usually combined with sol–gel, which is a process of preparing bioactive glass (Figure 5A). The glass precursor is obtained from a solution of metal alkoxides and nitrates in ethanol, which is subjected to sufficient hydrolysis and condensation reactions by stirring (Catauro et al., 2016), and the substrates are then immersed in gel. In other cases, the prepared glass is mixed into the solution and infiltrates the substrates. After dipping, drying and calcination are performed to remove excess material and stabilize the coating.

Various conditions such as the extraction speed, dipping times, viscosity of sol, and process of drying influence the coating formation (Brown, 1989). The extraction speed is directly related to the coating thickness. The slower the speed, the thinner the coating and the more the original shape of the substrate can be maintained.

When compared with others, dip coating can be performed under lower temperatures, and the process is cheaper and easier. The coatings are more uniform and have higher purity, which stabilizes the substrate's shape (Fu et al., 2011; Catauro et al., 2016; Fu et al., 2017a). Other components, such as nanoparticles, mesoporous agents, and antimicrobial agents added to the solution, may lead to complications in the coating structure and composition (Tian et al., 2016; Rivadeneira and Gorustovich, 2017).

Electrophoretic deposition (EPD) forms a coating by applying a direct current or alternating current electric field between two electrodes, causing charged particles to be dispersed in suspension and move in the direction of the substrate electrode (Figure 5B). The particles are deposited in an orderly manner on the substrate (Wang et al., 2014; Sultana et al., 2021). Heat treatment at 800°C–900°C makes the coatings denser and more stable (Ananth et al., 2013; Khanmohammadi et al., 2020).

Table 7 shows the parameters in the process of EPD in the included studies. The factors affecting the properties of coatings prepared via EPD can be concluded as solutions’ stability, conductivity (Zhitomirsky et al., 2009), and BG powder concentration. The distance between the electrodes, applied electric field voltage, deposition time, temperature, and pH all influence the deposited coatings. Changes in the BG concentration in the solution may increase the electrophoretic mobility, and adjustments made in the deposition parameters can make the coatings denser and more uniform (Jugowiec et al., 2017).

Hydrothermal deposition is a multiphase reaction of dissolution and recrystallization of materials under high temperature and pressure with an aqueous solution as the reaction system in a closed autoclave, leading to the formation of precipitates (Pore et al., 2021). This method can be used for the preparation of powders and as a coating technique for materials.

The process essentially involves attaching thermocouples and pressure sensors to the reactor assembly, setting the parameters, and heating the reactor. Subsequently, pressure builds up and the coating is deposited on the sample in a supercritical environment (Ali et al., 2018).

The method is simple and inexpensive and can be used to synthesize coatings with uniform thickness, orientation, and shape directly in an aqueous solution (Valanezhad et al., 2015). The coating structure can be controlled by changing the synthesis parameters. However, with this method, it is not easy to control the crystal structure formed, and due to the environmental requirements of high temperature and high pressure, it is more dependent on the experimental equipment.

Cytotoxicity is the killing of cells by chemicals without involving the cellular mechanisms of apoptosis or necrosis. After implantology, ions that reach the cytotoxic concentration leach out of the glass coating and interact with the cells (Al-Noaman et al., 2012). As an important indicator to assess the safety of biological materials, cytotoxicity is assessed by in vitro studies that simulate the survival environment of cells. A lower cell survival rate indicates higher cytotoxicity, indicating clinical risk (Wataha et al., 1994). Cell viability above 70% is usually considered non-cytotoxic (Wei and Ding, 2017). The release of large amounts of alkaline ions can have adverse effects on living cells. The increase in pH at the implantation site is accompanied by the dissolution of bioactive glass, which increases by-products and ultimately leads to toxic effects on the surrounding tissues (Jones, 2013). The adverse effect causes tissue reactions such as inflammation, necrosis, induction of immunity, and carcinogenesis (Costa, 1991).

Cytotoxicity is influenced by the substance and doped content in the glass coating. BG doped with silver ions, cobalt oxide, and titanium dioxide has shown that the doping of cobalt oxide causes higher cytotoxicity than that of silver and titanium dioxide (Lung et al., 2021). This is mainly because cobalt ions induce oxidative stress and activate intracellular nicotinamide adenine dinucleotide phosphate (NADPH) oxidase to produce ROS, which causes oxidative damage to cells (Chattopadhyay et al., 2015) and affects cell morphology and viability (Fleury et al., 2006). A high level of silver in the coating increases the level of released nitrate and improves its cytotoxicity (Catauro et al., 2015).

Cell activity refers to the ability of cells to maintain or resume normal physiological activities, such as cell adhesion, proliferation, migration, differentiation, and metabolism (Patel et al., 2019), which is affected by environmental factors, such as cell culture parameters, attached drugs, and growth factors. Cell adhesion refers to the cellular ability to contact and bind to adjacent cells or the extracellular matrix (ECM) (Humphries et al., 2009). Cell spreading is the behavior of cells on the surface of a biomaterial, which is influenced by certain protein molecules (Cuvelier et al., 2007). Cell proliferation is the process of cell division by DNA replication and other reactions under the action of cycle regulators, resulting in an increase in cell numbers, which is the basis for normal tissue development and maintenance (Xynos et al., 2001). Cell migration is the movement of cells after receiving a certain signal or concentration gradient of a substance and is essential for proper immune response and wound repair (Trepat et al., 2012). The morphological and functional changes that occur because of the selective expression of cellular genes are defined as cell differentiation (Ponzetti and Rucci, 2021).

BG 45S5 is generally regarded as the gold standard for bioactive glass, and its ionic lysate induces adhesion and proliferation of cells (Abushahba et al., 2020). Calcium silicate-based materials release calcium and silicate ions, which induce osteoblast proliferation by gene activation (Catauro et al., 2016). Uniform coatings give the best cell metabolic, whereas inhomogeneous coatings, where some cells are in direct contact with the Ti matrix, are less biocompatible (Catauro et al., 2016). However, studies have shown that, when compared with HA coatings, BG-coated samples lead to the rupture and contraction of cells (Dhinasekaran et al., 2021). The surface roughness and profile of the coatings influence cell adhesion and proliferation (Gaweda et al., 2018). Fluorescence staining was performed on cells to observe the effect of BG coatings on the distribution of cytoskeleton, a typical sub-apical localization of the cytoskeleton around the cell membrane can be found. This phenomenon confirms the adhesion and proliferation ability of the cells and demonstrates that BG coatings contribute to the differentiation of the Caco-2 cell line [Figure 6A(a,b)] (Ledda et al., 2015). Doping zinc oxide into BG 45S5 stimulates osteoblasts proliferation and, thus, improves the combination between the implants and bone tissue (Ishikawa et al., 2002; Oki et al., 2004). BG-coated implants doped with Ag₂O increase cell viability. A large number of cells can be observed on the surface with Ag₂O in 0.008 % mol/mol, which may have resulted from the release of nitrate ions (Catauro et al., 2015). On the surface of composite coatings comprising chitosan and BG, more cell adhesion, a higher rate of proliferation, and an extended and expanded cytoskeleton can be observed (Patel et al., 2019; Zarghami et al., 2021). On the surface of yttria-stabilized zirconia (YSZ)-BG composite coatings, a large coverage of osteoblasts can be observed, with many filamentous adhesions between the cells and visible nodule formations, which is an early feature of cell differentiation [Figure 6A(c,d)]. However, increasing the relative content of YSZ in the coating decreases the cell activity, as more yttria Y³⁺ ions are released (Ananth et al., 2013).

In in vivo experiments, after the formation of HA on the surface, five stages of biological events occurred: 1) growth factors adsorption, 2) bone progenitor cells adhesion, 3) proliferation, 4) differentiation, and 5) production of the extracellular matrix, which enhances bone healing (Popa et al., 2015). The surface roughness contributes to a larger cell attachment area. It has been observed that cell density and bone healing in implants with rough surfaces are significantly better than in those with smooth surfaces (Klyui et al., 2021). Dissolved ions such as Ca, Mg, and Si can activate the expression of bone-related genes by regulating bone-related cell growth and metabolism (Taguchi et al., 2019; Zhang et al., 2021). The wettability and hydrophilicity of the surface also promote cell proliferation (Soares et al., 2018).

Hemocompatibility is the ability of blood to tolerate a material without causing significant adverse blood reactions when the material is in contact with blood (Nalezinkova, 2020). The main adverse blood reactions involve thrombosis. The absorbance of blood proteins on the surface of the materials triggers a series of cascade reactions, resulting in thrombosis, and the coagulation cascade spreads rapidly, leading to death in severe cases (Manivasagam et al., 2021). Hemolysis, which reflects hemocompatibility, is caused by adverse reactions to any toxic substance that comes in contact with blood (Dhinasekaran et al., 2021). The percentage of hemolysis (hemolysis%) is calculated on the basis of the following formula: Hemolysis % = free   hemoglobin   concentration total   hemoglobin   concentration × 100 % .

If the hemolysis% of a sample is <2%, it is non-hemolytic; a hemolysis% between 2% and 5% indicates that it is slightly hemolytic; if the hemolysis% >5%, it is considered hemolytic. Materials that are blood compatible are considered to have a hemolysis% less than 5%.

Studies have demonstrated that the hemolysis% of BG powders in all concentrations is lower than that of HA at the same concentrations, while more red blood cells (RBCs) are ruptured with BG coating. This may be due to the release of sodium ions from the BG coating, causing RBCs to rupture and, thus, exhibit hematotoxicity, which is corrected by washing after coating (Durgalakshmi et al., 2020; Dhinasekaran et al., 2021). Bargavi et al. (2022) found that BG coatings doped with various concentrations of alumina (Al) exhibited non-hemolytic properties and improved hemocompatibility when compared to pure BG coatings, where BG coatings doped with 10% Al had the best hemocompatibility. The hemocompatibility of BG coatings doped with zirconia (Zr) has also been investigated. With an increase in Zr concentration, the hemolysis% of the coating slightly decreased; while BG coatings doped with 5% and 10% Zr showed non-hemolysis, BG coatings doped with 15% Zr showed slight hemolysis (hemolysis % <2.5%) (Bargavi et al., 2020). Generally, BG-coated implants show great hemocompatibility.

Inflammation is an immune response of the body to resist harmful irritation, which helps maintain tissue homeostasis during injury or infection (Medzhitov, 2010; Wu et al., 2019; Chang and Xiong, 2020); however, excessive inflammatory responses form fibrous capsules that prevent implants' osteointegration. Therefore, superior anti-inflammatory property is critical for implant success.

There is no significant difference in the expression of anti-inflammatory factors in human amniotic mesenchymal stromal cells (hAMSCs) on RKKP glass-ceramic coating when compared with the control group, which indicates that the coating did not affect the expression of hAMSCs' anti-inflammatory factors (Ledda et al., 2016). Wu et al. (2014) found that bioactive Sr₂MgSi₂O₇ (SMS) ceramic coatings exhibited superior anti-inflammatory effects compared to HAp coatings, and their mechanism of inhibiting the inflammatory response may be due to the 1) inhibition of the Wnt5A/Ca²⁺ pathway, which enhances the inflammatory response by decreasing the Ca²⁺ concentration or 2) inhibition of inflammatory cytokine expressions by the Toll-like receptor (TLR) pathway, which induces an immune response by the release of Mg²⁺ and Sr²⁺ (Figures 7A a–c).

For in vivo experiments, the inflammatory response of the host to the implant is a normal bodily reaction, often manifested as a local inflammatory response and vascular congestion, which disappears after some time (Taguchi et al., 2019). The gingival index determines the inflammatory status by observing the gingival condition, while the periodontal pocket is a manifestation of the pathological inflammatory response (Löe, 1967; Donos, 2018). In clinical trials, in follow-up survey statistics, the gingival index and depth of periodontal pockets were smaller in BG-coated groups, which showed a higher success rate of implantology (Mistry et al., 2016).

Antibacterial properties are an essential characteristic in grafts implanted clinically, which refers to the grafts’ ability to reduce microbial growth on their surfaces. The formation of a biofilm is the first stage of bacterial growth, which also inhibits the proliferation of osteoblasts.

Better antibacterial properties are reflected in larger bacterial inhibition zones, lower reduction rates, fewer colony-forming units, and lower minimum inhibitory concentrations. A study has shown that as the content of silver ions in the coating increased, the number of bacteria on the sample surface decreased significantly and the bacterial inhibition area on the surface became larger [Figures 7B(a–f)] (Fu et al., 2017b). The cobalt- and Ti-doped glass coatings have better antibacterial properties than traditional 58S glass (Lung et al., 2021). As more zirconium oxide is added to samples, they exhibit a higher ability to inhibit bacterial growth (Bargavi et al., 2020), whereas borate-based glass exhibits worse antibacterial properties (Rau et al., 2020). Glass composited with drugs can help eradicate bacteria (Zarghami et al., 2021). By observing bacteria in saliva on the implant surface, Costa et al. (2020) found that fewer pathogenic bacteria were observed on the surface.

The antibacterial property is mainly increased by doping silver and other metallic ions in the manufacturing of chitosan composted coatings and in combination with antibiotics like tetracycline and vancomycin. The inhibition of bacteria by silver ions is mainly due to direct contact, which results in the deformation of cell membranes (Lung et al., 2021) and ROS in bacteria (Fu et al., 2017b). The electrostatic effects leading to changes in cell membrane permeability, thereby influencing cell signal transduction and production of ROS, are the main reasons for antibacterial properties caused by metallic compounds (Lung et al., 2021). Due to the slow-release behavior, BG containing chitosan and vancomycin has higher bactericidal effects (Zarghami et al., 2020). Chitosan increases cell membrane permeability, leading to the release of intracellular substances and thus causing cell death (Ordikhani et al., 2016). The composite coating of bioactive glass as a drug carrier can also significantly improve antimicrobial properties. BG has a rougher surface (Matter et al., 2021) and creates a more alkaline biological environment (Echezarreta-López and Landin, 2013; Brauer, 2015) through dissolution and ion release, where bacteria grow poorly.

In in vivo experiments, the plaque index and gingival recession were assessed to evaluate the degree of oral hygiene (Silness and Löe, 1964; Kassab and Cohen, 2003). A clinical trial proved that patients with BG-coated implants have reduced plaque index and gingival recession and, therefore, reduced occurrence of oral disease (Mistry et al., 2016).