Comparative Biocompatibility and Odonto-/Osteogenesis Effects of Hydraulic Calcium Silicate-Based Cements in Simulated Direct and Indirect Approaches for Regenerative Endodontic Treatments: A Systematic Review

Background: Regenerative dentistry is the operation of restoring dental, oral and maxillofacial tissues. Currently, there are no guidelines for the ideal cement/material in regenerative endodontic treatments (RET). Hydraulic calcium silicate-based cements (hCSCs) are currently the material of choice for RET. Objectives: This systematic review was conducted to gather all of the different direct and indirect approaches of using hCSCs in RET in vitro and in vivo, and to ascertain if there are any superiorities to indirect approaches. Methods and Materials: This systematic review was conducted according to the 2020 PRISMA guidelines. The study question according to the PICO format was as follows: Comparison of the biological behavior (O) of stem cells (P) exposed to hCSCs through direct and indirect methods (I) with untreated stem cells (C). An electronic search was executed in Scopus, Google Scholar, and PubMed. Results: A total of 78 studies were included. Studies were published between 2010 and 2022. Twenty-eight commercially available and eighteen modified hCSCs were used. Seven exposure methods (four direct and three indirect contacts) were assessed. ProRoot MTA and Biodentine were the most used hCSCs and had the most desirable results. hCSCs were either freshly mixed or set before application. Most studies allowed hCSCs to set in incubation for 24 h before application, which resulted in the most desirable biological outcomes. Freshly mixed hCSCs had the worst outcomes. Indirect methods had significantly better viability/proliferation and odonto-/osteogenesis outcomes. Conclusion: Biodentine and ProRoot MTA used in indirect exposure methods result in desirable biological outcomes.


Introduction
Regenerative dentistry is the operation of restoring and/or regenerating dental, oral and maxillofacial tissues and organs for therapeutic implementations [1][2][3][4]. Regenerative endodontic treatments (RET) are a large group of procedures assessed to maintain and regenerate dentine and pulpal tissues. Vital pulp therapy (VPT) sustains dental pulp vitality and maintains teeth [5]. Pulpotomy and direct pulp capping (DPC), induce the formation of regenerative dentine by human dental pulp stem cells (hDPSCs) in the treatment of exposed vital pulp [6]. Pulp capping materials develop a protective layer over the exposed vital pulp in pulpotomy, DPC, and indirect pulp capping (IPC) [7]. Ideal pulp capping materials must be biocompatible, have excellent sealing abilities, and promote migration, proliferation, and differentiation of hDPSCs [8,9]. Newly developed bioactive materials (e.g., bioactive glasses and calcium silicate-based cements) are produced/introduced every so often [10,11]. However, currently there are no guidelines for the ideal cement/material in RET.
In endodontic treatments and DPC procedures, hDPSCs and other types of alveolodental stem cells are in direct contact with hCSCs [34]. hCSCs and their toxins in direct contact with stem cells are much more harmful to the stem cells compared to indirect contact [35,36].
Consequently, many studies have tested the outcome differences of hCSCs in direct and indirect contact with stem cells, to compare their proliferative and regenerative abilities in vitro [37][38][39][40][41]. When hCSCs are clinically applied for human patients, there is no precise way to evaluate their biological outcomes, except extraction of the teeth and laboratory analysis. Therefore, a lot of the reported outcomes regarding hCSCs used in clinics do not have enough evidence to prove the toxicity/biocompatibility of hCSCs in both direct and indirect contact. However, in vitro studies, if conducted according to global standards, can be a reliable simulation of the clinical interactions between stem cells and hCSCs. Clinicians can choose their kind of hCSC and the type of contact based on studies conducted in vitro that simulate clinical environments.
To the reviewers' knowledge, there has been no comprehensive review executed on the comparison of viability/proliferation and the odonto-osteogenesis differentiation induction abilities of all of the commercially available hCSCs. Additionally, there is no review comparing the outcomes of different types of direct and indirect contacts in vitro. The main purpose of this systematic review was to gather all of the different direct and indirect approaches of using hCSCs in RET in vitro and in vivo, and to ascertain if there were any superiorities to indirect approaches when examined for biocompatibility and regeneration/differentiation abilities. Additionally, we sought to find the hCSCs with the most remarkable outcomes in each of the direct and indirect approaches in vitro, in order to help clinicians and scientists make an informed choice.

Results and Discussion
The search queries and PRISMA flow diagram (according to the PRISMA 2020 guidelines [42]) of this systematic review are displayed in Table 1 and Figure 1, respectively.

Study Selection
Database screening was performed, a total of 683 articles were initially identified and 302 of them were assessed for eligibility ( Figure 1). A total of 224 studies were excluded for the following reasons: clinical studies (n = 7) and unrelated subjects (n = 217). Hence, a total of 75 in vitro and 3 in vivo studies matched our inclusion criteria. Figure 2 showcases the distributions for all of the included studies and the range of years they were published in. All of the studies, their cells, cements, contact methods, and outcomes are detailed in Tables 2 and 3 for in vitro and in vivo studies, respectively. All of the abbreviated forms used in this review are listed in Table S1. "biodentine" OR "CEM" OR "MTA Plus" OR "MTA Fillapex" OR "endocem" [Supplementary Concept] OR "Neo MTA Plus" OR "MTA Repair HP" OR "Retro MTA" OR "Nex-cem MTA" OR "iRoot SP" [Supplementary Concept] OR "iRoot fast set" OR "well root ST" OR "AH Plus jet" OR "Portland cement" OR "accelerated Portland cement" [    a total of 75 in vitro and 3 in vivo studies matched our inclusion criteria. Figure 2 showcases the distributions for all of the included studies and the range of years they were published in. All of the studies, their cells, cements, contact methods, and outcomes are detailed in Tables 2 and 3 for in vitro and in vivo studies, respectively. All of the abbreviated forms used in this review are listed in Table S1. (See tables after the references section).

Types of Interventions
The different approaches that the authors used to place cells in contact with the materials were categorized into two major groups: direct contact and indirect contact. Furthermore, each group had different approaches, which are all displayed in Table 4 with their descriptions. Figures 4 and 5 showcase a visual description of all of the exposure methods in in vitro and in vivo studies, respectively.
Untreated stem cells were considered as a negative control group in all of the studies and all of the variables of the cements were analyzed in comparison to them. For an easier and more convenient way of comparing different outcomes, the following abbreviations were constructed: Outcomes that were significantly better and/or statistically higher than NC: significantly higher (SH).
Outcomes that showed no significant difference with NC: no significant difference (NSD). Outcomes that were significantly worse and/or statistically lower than NC: significantly lower (SL).
In addition, Figure 6 showcases a visual description of the assessment frequency of different direct and indirect exposure methods in the included studies from 2010 to 2022.

Cells/ Interventions
Methods of Assessment

Methods of Assessment
Viability and Proliferation Cellular viability and proliferation were examined in a total of 69 articles, using the following assays and methods (Supplementary Table 1

Odonto-/Osteogenesis
Alkaline phosphatase (ALP) activity was examined in a total of 25 studies using the ALP activity assay kit (colorimetric). Alizarin red staining (ARS) was assessed in a total of 22 studies. Gene expression was examined in a total of 39 studies using RT-PCR or qRT-PCR.

In Vivo Studies
Out of the three included vivo studies, none of them examined cellular migration, mineralization (ARS), or gene expressions. Only one study examined ALP activity, however, it did not compare the results of the cements with the NC group [95]. Two of the in vivo studies investigated the viability/proliferation abilities of their hCSCs and in

Methods of Assessment Viability and Proliferation
Cellular viability and proliferation were examined in a total of 69 articles, using the following assays and methods (Supplementary Table S1) Migration Cellular migration was examined in a total of 23 studies. Migration was tested using the following assays and methods: WHA, TMA, RT-PCR and Cell Tracker TM Green CMFDA.

Odonto-/Osteogenesis
Alkaline phosphatase (ALP) activity was examined in a total of 25 studies using the ALP activity assay kit (colorimetric). Alizarin red staining (ARS) was assessed in a total of 22 studies. Gene expression was examined in a total of 39 studies using RT-PCR or qRT-PCR.

In Vivo Studies
Out of the three included vivo studies, none of them examined cellular migration, mineralization (ARS), or gene expressions. Only one study examined ALP activity, however, it did not compare the results of the cements with the NC group [95]. Two of the in vivo studies investigated the viability/proliferation abilities of their hCSCs and in both of them NC showed SH results [45,95]. Only one study examined cellular attachment and reported that CEM showed cellular adhesion in both FM and 24 h set RT conditions (Table 3).

Setting Times and Conditions In Vitro
The 24 h setting in incubation (II) technique was the most used approach and had remarkable rates of SH results, while most of the cases of FM cements led to SL results compared to NC (Table 2 and Figure 3). Compared to the 24 h II technique (n = 28), the dried heat (DH) condition was used in only four studies before the application of hCSCs. However, all of the reported biocompatibility and regenerative outcomes were similar (NSD) to the NC group or significantly better than NC (SH) [55,56,75,76]. Out of the four studies that examined the DH technique, only one of them reported their exact environment and conditions-a 50 • C oven for 15 min [53]-but the remaining three studies did not specify their environments. A total of four studies used RT as their only setting condition for hCSCs and their results were a mixture of SH, NSD, and SL outcomes compared to the NC [41,43,54,116] (Figure 3).

Comparison of Different hCSCs In Vitro
To better comprehend the outcomes of different cements used in different approaches for each category of results (i.e., proliferation, odontogenesis, and osteogenesis), we designed three figures: Figure 7 (viability/proliferation), Figure 8 (odontogenesis), and Figure 9 (osteogenesis), to simplify the results. We only focused on the outcomes that showed significant differences between hCSCs. both of them NC showed SH results [45,95]. Only one study examined cellular attachment and reported that CEM showed cellular adhesion in both FM and 24 h set RT conditions (Table 3).

Setting Times and Conditions In Vitro
The 24 h setting in incubation (II) technique was the most used approach and had remarkable rates of SH results, while most of the cases of FM cements led to SL results compared to NC (Table 2 and Figure 3). Compared to the 24 h II technique (n = 28), the dried heat (DH) condition was used in only four studies before the application of hCSCs. However, all of the reported biocompatibility and regenerative outcomes were similar (NSD) to the NC group or significantly better than NC (SH) [55,56,75,76]. Out of the four studies that examined the DH technique, only one of them reported their exact environment and conditions-a 50 C oven for 15 min [53]-but the remaining three studies did not specify their environments. A total of four studies used RT as their only setting condition for hCSCs and their results were a mixture of SH, NSD, and SL outcomes compared to the NC [41,43,54,116] (Figure 3).

Comparison of Different Exposure Methods In Vitro
A detailed comparison of only the SH results of all five different exposure methods is shown in Table 5. However, in terms of NSD and SL results, the outcome differences are discussed in each of the categories below.

Viability and Proliferation
Indirect methods performed much better, with Indirect1 having the highest rate of SH results. Direct2 had the worst performance.

Cellular Attachment
Direct2 was not examined in this category. Indirect3 showed SH results in all of its experiments. Indirect2 had the weakest performance with no SH outcomes.

Summary of Outcomes of In Vitro Studies
We summarized all of the outcomes for the five different contact approaches in vitro (i.e., Direct1, Direct2, Indirect1, Indirect2, and Indirect3) into one table (Table 5). Different approaches are categorized into four groups based on their performance: (1) more than 80% of results were SH than NC; (2) 50% to 80% of results were SH than NC; (3) 33% to 50% of results were SH than NC; (4) less than 33% of results were SH than NC. Approaches that did not have even a single case of SH results were not included in Table 5.

Risk of Bias Assessment
The results of risk of bias assessments for in vitro studies and in vivo studies are displayed in Figures S1 and S2, respectively. The risk of bias was unclear for all three included in vivo studies. Out of the 75 in vitro studies, all of them had unclear risk of bias in the first three questions that represent the randomization of studies; however, all of them had low risk of bias in the remaining five questions of the questionnaire. Overall, all 75 in vitro studies had a low to unclear risk of bias.

Discussion
This systematic review was conducted to assemble all of the different direct and indirect contacts between various hCSCs and stem cells in vitro and in vivo. As mentioned in our results, there was a significant difference between the number of in vitro and in vivo studies (75 in vitro versus 3 in vivo). Amongst the five different direct and indirect approaches in vitro, indirect ones significantly outshone the direct methods in almost all different outcome categories. Indirect1 was the most used approach amongst all included studies ( Table 3). Most of the studies allowed hCSCs to set for 24 h in incubators (II). PRMTA and BD were the most frequently used hCSCs and showed significantly better biological behavior (i.e., cell viability/proliferation, attachment, migration, mineralization, odonto-/osteogenesis, and variant gene expressions) compared to other utilized cements in different exposure methods (i.e., Direct1, Direct2, Indirect1, Indirect2, and Indirect3).
In our systematic electronic search, we found four systematic reviews similar to our review. Although these reviews have analyzed similar categories of outcomes to our review, they have only focused on a very small group of hCSCs, have included only a certain type of stem cell, or had chosen only direct contact [112,114,117,118]. On the contrary, these limitations were not considered in our systematic review, enabling us to compare and discuss commercially available hCSCs more comprehensively. Additionally, the categorization of all the different direct and indirect exposure methods both in vitro and in vivo has never been conducted before.

hCSCs Differences
Calcium hydroxide deposition after hCSCs hydration is pivotal to initiate the consequential biologic reactions of hCSCs in contact with stem cells [119]. Previous studies showed that both PRMTA and BD fulfill their calcium hydroxide deposition [120]. An alkaline environment is crucial for inducing proliferation and odonto-/osteogenesis by hCSCs [121]. Different studies have reported that both PRMTA and BD induce alkaline pH in contact with cells, regardless of the evaluation periods [122,123]. Furthermore, previous findings showed that BD and PRMTA have similar cytocompatibilities [124]. BD and PRMTA both have tricalcium silicate (Ca 3 SiO 5 ) and dicalcium silicate (Ca 2 SiO 4 ) as their major components. Additionally, BD contains calcium carbonate (CaCO 3 ) (filler material) and calcium oxide (CaO) (traces), whereas PRMTA contains calcium sulfate dihydrate (CaSO 4 ·2H 2 O) (filler material) and tricalcium aluminate (Ca 3 Al 2 O 6 ) (traces) [125]. The reported data suggest that calcium sulfate dihydrate and calcium carbonate help PRMTA and BD, respectively, to be more cytocompatible for hDPSCs [126]. All of the mentioned reported outcomes corroborate our findings that BD and PRMTA have very similar abilities and both result in similar viability/proliferation and odonto-/osteogenesis outcomes.

Setting Times and Conditions
Most of our included in vitro studies allowed their cements to set in incubation (II) for at least 24 h before application. The majority of these studies saw similar outcomes (NSD) with the NC group, while some of them reported even better outcomes (SH) than NC. The 37 • C, 5% CO 2 and 95% humidity atmosphere supplied by the incubators simulates the elution of hCSC toxins in vitro and consequently prevents the damages that hCSC toxins can cause to hDPSCs, hPDLSCs and SCAP [106]. Therefore, when cements are applied immediately after mixing (freshly mixed (FM)), the biocompatibility and odonto-/osteogenesis outcomes are significantly lower (SL) and weaker than stem cells with no hCSCs (negative control (NC) group), because the hCSC toxins did not have any time to be released prior to application [48]. Additionally, some studies have reported that freshly mixed (FM) hCSCs are so toxic for stem cells that almost all of the cells were dead at the assessed evaluation periods and no cellular proliferation was observed [37]. Even when studies reported that freshly mixed hCSCs did not kill the stem cells, the odonto-/osteogenesis outcomes (e.g., DSPP gene expression, ALP gene expression, etc.) were significantly lower than NC [47]. The remarkably low number of studies using the DH or RT setting condition techniques, along with the lack of information regarding the environmental details of the DH technique, makes their current reported outcomes unreliable. Further investigations both in vitro and in vivo can examine the superiority/inferiority of the DH and RT techniques compared to II.

Direct and Indirect Approaches In Vitro
The Indirect1 approach benefits from an adequate setting time for cements (mostly in incubation (II) for 24 h), which releases the majority of toxins before making the dilution/supernatant [101]. The medium, in direct contact with the fully set cements, spends a considerable amount of time in the incubator to make sure all of the biocompatibility and regeneration-inducing molecules are released into the medium to make an hCSCenriched supernatant. The incubated supernatant is not only rich enough in hCSCsinducing molecules, but it also does not have the toxicity of cements in direct contact with stem cells, and this is why Indirect1 was so successful in not only keeping cells viable, but also inducing the proliferation and regeneration in stem cells significantly better (SH) than the NC group. Indirect2 had the most versatile SH outcomes (> 80%) across all of the different examinations and was mentioned as one of the most desirable approaches in 10 of the 14 mentioned outcome measures in Table 5. The Indirect2 method required specifically designed Transwell™ permeable inserts with extremely small pores (0.3-0.4 µm) incorporated into them. In this technique, cells were on the bottom of the plates and only in indirect contact with hCSCs through the shared medium. Most of the studies that used the Indirect2 technique allowed their cements to fully set (24 h) before placing them in the Transwell inserts. The very small pores led to a very slow release of hCSC molecules into the shared medium with the stem cells. Stem cells had enough time to respond to the hCSC chemicals without being exposed to a huge amount of toxic freshly mixed cements. Hence, this approach produced a wide range of successful outcomes throughout almost all of its examinations (Table 5).
Despite the fact that Indirect1 and Indirect2 mostly resulted in significantly remarkable results, when cements are freshly mixed, the results were SL than the NC group [47,50]. These findings help us to comprehend the fact that choosing the type of contact and the setting time/condition are equally crucial for having the most remarkable outcomes. Indirect3, similar to Indirect2, appeared in 9 of the 14 examined outcomes in Table 5 as one of the approaches with the highest rates of SH results (>80%). hCSCs in this technique were ground into powder immediately after mixing (freshly mixed (FM)) and then they were dried and at the end mixed with the medium. As our collected data show, all three different indirect approaches had remarkably better outcomes than the two different direct approaches amongst all categories of outcomes. However, it is important to mention that the number of studies that each approach was used in was outstandingly different. Indirect1 had relatively lower rates of SH (>80%) results compared to the other two indirect approaches, yet Indirect1 was used in 24 studies, which is significantly more than Indirect2 and Indirect3 (12 and 16, respectively). The exceedingly high number of studies that used Indirect1 provides its outcomes with a significant level of reliability. On the other hand, the outstanding performance of Indirect2 and Indirect3 could not be ignored. Therefore, regarding the most successful indirect contact between hCSCs and cements in vitro, Indirect1 could be a solid and safe choice, with many studies reporting similarly positive outcomes, while Indirect2 and Indirect3 have shown remarkable outcomes but have been used in significantly fewer studies.
The number of experiments in each category examining each approach (direct or indirect) was significantly different. In some cases, studies with different approaches had similar results, and because they were assessed a similar number of times, their outcomes were perfectly comparable and all of them had the same level of reliability (e.g., all three different indirect approaches had high rates of SH results in ALP activity and were assessed a similar number of times). On the other hand, some of these immense differences resulted in unreliability when an approach was used in very limited studies. Some of the outcomes that were reported in a very limited number of studies were as follows: (1) Indirect3 was assessed for cellular attachment in only 2 studies, whereas, on the other hand, Indirect1 was assessed 21 times; (2) for cellular migration, Indirect3 and Direct1 both had the highest rates of SH results. However, Indirect3 was used 12 times and Direct1 only 5; (3) when assessed for mineralization, Indirect2, Indirect3 and Direct2 all showed remarkably high rates of SH results, but Direct2 was used only 2 times, and Indirect2 and Indirect3 were used 7 and 11 times, respectively.

Study Limitations and Suggestions
(A) The very limited number of in vivo studies (n = 3) makes their findings noncomparable with the findings of the in vitro studies (n = 75). Hence, the main focus of this review was directed at in vitro studies.
(B) Since our main goal by designing this review was to compare the abilities of different direct and indirect approaches to each other, we only investigated gene expressions that were examined in at least one approach of each of the direct and indirect groups. The expression of BMP1, BMP2, OC, CAP and CEMP1 genes were only examined in Indirect1. Since we did not have any study with a direct approach examining the expression of the mentioned genes, we were unable to include them.
(C) Given the fact that indirect approaches outshone direct ones in most of the categories of outcomes, we suggest that scientists and manufacturers design and use more indirect approaches in vivo. By doing this, we can discover whether the superiority of indirect approaches in vitro also applies for in vivo cases.
(D) There was a significant difference in the number of experiments that each approach was used for, and since the whole purpose of this review was to compare different approaches' abilities, if the number of experiments was equal, this comparison would have had much more value. We suggest designing a broad and comprehensive in vitro study focusing on all five different approaches and all the different outcomes. By doing this, not only would we have an equal number of experiments in each group, we would also be able to trust the compared results even more due to them being published by one group of authors instead of us comparing the results of different studies conducted by different groups of authors.
(E) Indirect pulp capping (IPC) is one of the few indirect approaches in VPT that is used commonly in clinics. However, IPC requires at least 0.5 mm of residual healthy dentine left on top of the pulp. Therefore, in cases of non-existing dentine, direct pulp capping (DPC) is performed. The results of this review suggest that indirect approaches lead to much better outcomes in almost all categories of results in vitro. Given the superiority of IPC to DPC in in vitro studies, we suggest that more scientists lean onto using and inventing different hydrogels and other biomaterials to simulate IPC when there is not enough healthy dentine left on top of the pulp to perform a conventional IPC. The biomaterials used to simulate IPC must have selective penetration and permeability, just like healthy dentine. This way, hDPSCs would not be in direct contact with hCSCs, and the integrity of the pulp would also be preserved.
(F) In this review, we did not investigate the outcome differences of different cellular assays, cellular culture conditions, and different resource variations in our included studies, for the following reasons: A) for a proper comparison amongst different cellular assays (e.g., MTT, XTT, LDA, SEM, MTS, CCK-8, etc.), we had to compare studies that had complete similarity in all other elements of their study (e.g., similar type of stem cells, similar type of hCSCs, similar setting time, similar setting condition, etc.), and this level of similarity was not available in our included studies; B) Even if we found enough similar studies to compare their outcomes of these cellular assays and cellular culture conditions, classification of these culture conditions and cellular assays would be impossible. For viability/proliferation evaluations, our included studies assessed 19 different assays (MTT, XTT, LDA, SEM, MTS, CCK-8, etc.), and numerous assays were assessed for other evaluations as well, such as cellular attachment (10 assays), cellular migration (4 assays), etc.

Conclusions
When they are assessed for viability/proliferation and the odonto-/osteogenesis of stem cells in vitro, BD and PRMTA have similarities and both have significantly better outcomes than TCLC, PC and many other commercially available hCSCs in both direct and indirect approaches. Allowing hCSCs to set for at least 24 h in incubation (II) before application results into the most desirable outcomes. Indirect contact between hCSCs and stem cells is significantly less cytotoxic for stem cells and induces remarkably higher rates of odonto-/osteogenesis compared to direct contact in vitro. Moreover, Indirect1 is the most tested contact between hCSCs and stem cells for both viability/proliferation and odonto-/osteogenesis outcomes.

Materials and Methods
This study has been prepared and organized according to the preferred reporting items for systematic reviews and meta-analyses (PRISMA) 2020 guidelines [42]. This systematic review has been registered at PROSPERO (Registration ID: CRD42023387828). The study question according to the PICO format was as follows: Comparison of biological behavior (O) of stem cells (P) exposed to hCSCs through direct and indirect methods (I) with untreated stem cells (C). Additionally, the stem cells' behavior with direct exposure was also compared to the hCSCs with indirect exposure.

Types of Studies
No limitation was considered for the type of the included studies, and all of the in vitro and in vivo articles evaluating the behavior of stem cells that were exposed to cements were included.

Population
All of the studies that used stem cells to analyze the biological features of at least one type of commercially available hCSC through direct or indirect methods were included. We did not apply any restrictions on the type of stem cells (human or animal).

Intervention
Studies that analyzed any type and form of direct and/or indirect contact between hCSCs and stem cells in vitro and in vivo were included.

Control
Studies that considered untreated stem cells as a negative control (NC) group were included.

Types of Outcome Measures
Studies that analyzed the following outcomes were included: (1) Setting time and setting environment of each cement; (2) The types of tests assessed for each type of outcome; (3) Biocompatibility; cellular migration, cellular attachment, cellular viability/proliferation; (4) Odonto-/osteogenesis; ALP activity, mineralization activity (tested via ARS) and odonto-/osteogenesis-related gene expressions.

Information Sources and Search Strategy
An electronic search was executed in Scopus, Google Scholar, and Medline via PubMed to identify eligible studies only in the English language. The search included articles up to 6 December 2022. The search queries mentioned in Table 1 were considered for electronic search.

Study Selection and Data Collection
Two reviewers (AY and SM) independently screened the titles and abstracts of articles and excluded articles based on the exclusion criteria mentioned above. Selected articles were then fully read to see if they passed our inclusion criteria. In the case of any disagreement, a third reviewer (HN) was consulted. The data and outcomes from selected studies were then extracted and tabulated. The same reviewers performed the data extraction, and any conflicts were solved by a third expert (HN).
In reporting the results of gene expression, the following point was considered: For the genes that were expressed in the early phases (e.g., ALP, OCN, OPN, DSPP, and DMP1), only results related to the early phase of the differentiation were mentioned and the comparison related to the late phases of the differentiation for these genes was not mentioned. Similarly, for the late-phase genes (e.g., BMP, OC, and Runx2), merely the comparisons related to the expression in the late phases were noted.

Synthesis Methods
Based on the extracted data, different direct and indirect techniques used for VPT and endodontic treatments were widely diversified. Hence, it was not possible to perform a meta-analysis. A descriptive analysis of the data extracted from clinical studies, along with narrative and graphical synthesis, was performed.

Risk of Bias Assessment
The risk of bias of the included studies was assessed individually by two reviewers (AY and SM), using the CRIS guidelines (checklist for reporting in vitro studies) for in vitro studies and Cochrane's risk of bias assessment tool for in vivo studies. The CRIS checklist consists of 10 questions, 2 of which were not considered for this review due to not agreeing with the risk of bias analysis of in vitro studies.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/jfb14090446/s1, Figure S1: Risk of bias assessment for in vitro studies; Figure S2: Risk of bias assessment for in vivo studies; Table S1: Complete list of all of the abbreviations in alphabetic order; Table S2: Commercially available hCSCs used in in vitro and in vivo studies; Table S3: Modified hCSCs used in in vitro and in vivo studies.