RANK, RANKL, and OPG in Dentigerous Cyst, Odontogenic Keratocyst, and Ameloblastoma: A Meta-Analysis

The of this study to assess and compare RANK, RANKL, and OPG immunoexpression in dentigerous cyst, odontogenic keratocyst, and ameloblastoma. The protocol was registered in PROSPERO (CRD42018105543). Seven databases (Embase, Lilacs, LIVIVO, PubMed, Scopus, SciELO, and Web of Science) were the primary search sources and two databases (Open Grey and Open Thesis) partially captured the “grey literature”. Only cross sectional studies were included. The JBI Checklist assessed the risk of bias. A meta-analysis with random effects model estimated the values from the OPG and RANKL ratio reported by the individual studies and respective 95% confidence intervals. The heterogeneity among studies was assessed with I2 statistics. Only nine studies met the inclusion criteria and were considered in the analyses. The studies were published from 2008 to 2018. Two studies presented low risk of bias, while seven studies presented moderate risk. The meta-analysis showed the highest OPG>RANKL ratio for dentigerous cyst (ES=43.3%; 95% CI=14.3-74.8) and odontogenic keratocyst (ES=36.8%; 95% CI=18.8-56.7). In contrast, the highest OPG<RANKL ratio was found for ameloblastoma (ES=73.4%; 95% CI=55.4-88.4) and it was higher in the stromal region compared to the odontogenic epithelial region. The results may explain the aggressive potential of ameloblastoma from the higher OPG<RANKL ratio in this tumor, while it was lower for dentigerous cyst and odontogenic keratocyst. aimed to assess and compare, through a systematic review of the literature, the immunoexpression of RANK, RANKL, and OPG on dentigerous cyst, odontogenic keratocyst, and ameloblastoma, in order to verify whether the odontogenic keratocyst is more similar to the neoplastic or the cystic profile. The authors tested the following hypothesis: the odontogenic keratocyst will be more similar to the cystic profile.

The aim of this study was to assess and compare RANK, RANKL, and OPG immunoexpression in dentigerous cyst, odontogenic keratocyst, and ameloblastoma. The protocol was registered in PROSPERO (CRD42018105543). Seven databases (Embase, Lilacs, LIVIVO, PubMed, Scopus, SciELO, and Web of Science) were the primary search sources and two databases (Open Grey and Open Thesis) partially captured the "grey literature". Only cross sectional studies were included. The JBI Checklist assessed the risk of bias. A meta-analysis with random effects model estimated the values from the OPG and RANKL ratio reported by the individual studies and respective 95% confidence intervals. The heterogeneity among studies was assessed with I2 statistics. Only nine studies met the inclusion criteria and were considered in the analyses. The studies were published from 2008 to 2018. Two studies presented low risk of bias, while seven studies presented moderate risk. The meta-analysis showed the highest OPG>RANKL ratio for dentigerous cyst (ES=43.3%; 95% CI=14.3-74.8) and odontogenic keratocyst (ES=36.8%; 95% CI=18. 8-56.7). In contrast, the highest OPG<RANKL ratio was found for ameloblastoma (ES=73.4%; 95% CI=55.4-88.4) and it was higher in the stromal region compared to the odontogenic epithelial region. The results may explain the aggressive potential of ameloblastoma from the higher OPG<RANKL ratio in this tumor, while it was lower for dentigerous cyst and odontogenic keratocyst.
inhibits the osteoclastogenic process, blocking the RANKL and RANK binding (11).
A tumor may invade bone tissue and affect the balance between resorption and apposition (12). Several studies (4,(12)(13)(14) have been focused on investigating the correlation of the immunologic expression of RANK, RANKL, and OPG to the development of odontogenic cysts and tumors. However, the scientific literature is still controversial regarding the expression of such proteins in the development of these lesions. Therefore, this study aimed to assess and compare, through a systematic review of the literature, the immunoexpression of RANK, RANKL, and OPG on dentigerous cyst, odontogenic keratocyst, and ameloblastoma, in order to verify whether the odontogenic keratocyst is more similar to the neoplastic or the cystic profile. The authors tested the following hypothesis: the odontogenic keratocyst will be more similar to the cystic profile.

Introduction
Odontogenic cysts and tumors constitute one of the most important lesion groups of the oral and maxillofacial complex (1). Odontogenic tumors are rare and manifest from various clinical features, as well as different histopathological presentations of the epithelium and odontogenic ectomesenchyme (2). On the other hand, cysts originate from Malassez epithelial rests remaining from the dental blade or the reduced enamel epithelium (3). One of the most significant biological events in the pathogenesis of these lesions is osteoclastic cell activation, which results in bone resorption (4).
The balance between bone formation and resorption is required for bone homeostasis, so the tissue may fully function (5,6). Hence, the unbalance of this system has been associated with several bone neoplasias (7). Among the regulating factors of bone resorption, the system that stands out the most is composed by the receptor activator triad of nuclear factor kappa B (RANK), nuclear factor kappa B ligand (RANKL), and osteoprotegerin (OPG) (8). The RANK works as a signaling receptor of RANKL, while the latter is expressed in the osteoblastic cells of the periodontal ligament (9), binding to RANK and activating osteoclasts to promote bone resorption (10). On the other hand, OPG Expression of RANK, RANKL and OPG (15) and the Cochrane guidelines (16). The systematic review protocol was registered in the PROSPERO database under no. CRD42018105543.

Study Design and Eligibility Criteria
This study was a systematic review that aimed to answer the following guiding question: "Is the profile of the RANKL/ OPG ratio in the odontogenic keratocyst, assessed by the immunoexpression of RANK, RANKL, and OPG, more similar to the neoplastic or the cystic profile?".
The studies included assessed the expression of the osteoclastogenic factors RANK, RANKL, and OPG in ameloblastoma, dentigerous cyst, and non-syndromic odontogenic keratocyst by immunohistochemistry, without restrictions of year, language, or publication status (In Press).
The following were excluded: 1) Studies involving inflammatory odontogenic cysts; 2) Studies not related to the topic; 3) Review studies, case reports, letters to the editor or editorials, congress abstracts, personal opinions, and books and/or book chapters; 4) Studies with high risk of bias.

Sources of Information and Search
The Embase, Latin-American and Caribbean Health Sciences Literature (LILACS), LIVIVO, PubMed (including MedLine), SciELO, Scopus, and Web of Science databases were the primary study sources. OpenThesis and OpenGrey partially captured the "grey literature". A manual search was also performed through a systematized analysis of the references of the eligible articles. All steps aimed to minimize selection and publication biases.
The MeSH (Medical Subject Headings), DeCS (Health Sciences Descriptors), and Emtree (Embase Subject Headings) resources were used to select the search descriptors. The Boolean operators "AND" and "OR" enhanced the research strategy through several combinations (

Study Selection
The studies were selected in three different phases. In the first phase, as a calibration exercise, the reviewers discussed the eligibility criteria and applied them to a sample of 20% of the studies retrieved, so to determine inter-examiner agreement. After achieving a proper level of agreement (Kappa ≥ 0.81), two eligibility reviewers [IFPL and FRM] performed a methodical analysis of the titles of the studies, independently. The reviewers were not blind to the names of authors and journals. Titles not related to the topic were eliminated in this phase. In phase 2, two reviewers [IFPL and FRM] also analyzed the abstracts systematically. Studies not related to the topic, review studies, case reports, letters to the editor or editorials, congress abstracts, personal opinions, and books and/or book chapters were excluded. The studies related to the topic, but without abstracts available were fully analyzed in the third phase.
In this phase, the full texts of preliminary eligible studies were analyzed to verify whether they fulfilled the eligibility criteria. When there was no agreement in the assessment, a third reviewer [LRP] was consulted to make a final decision. The studies rejected were registered separately, explaining the reasons for exclusion.

Process of Data Collection and Extraction
After the selection, the studies were analyzed and two reviewers [IFPL and FRM] extracted their data for the following information: Identification of the study (author, year of publication, and study location); sample characteristics (number of cases); cysts and tumors assessed; ethical criteria involved; specimen fixation; diagnostic method used; immunoexpression of RANK, RANKL, and OPG; and correlation of RANKL and OPG.
To ensure the consistency among reviewers, both reviewers [IFPL and FRM] performed a calibration exercise, in which information were extracted jointly from an eligible study. The reviewers solved any disagreement through discussions, and when both reviewers disagreed, they consulted a third one [LRP] for a final decision.

Risk of Individual Bias of the Studies
The risk of bias of the studies selected was assessed by the Joanna Briggs Institute Critical Appraisal Tools for use in JBI Systematic Reviews Checklist for Analytical Cross Sectional Studies (17). Two authors [IFPL and FRM] assessed each domain independently and systematically regarding their potential risk of bias, as recommended by the PRISMA-P (15). The reviewers solved any disagreement through discussions, and when both reviewers disagreed, they consulted a third one [LRP] for a final decision.
The risk of bias was ranked as High when the study reached up to 49% of "yes" score, Moderate when the study reached from 50% to 69% of "yes" score, and Low when the study reached over 70% of "yes" score.

Study Selection
The first selection phase resulted in 245 studies distributed in nine electronic databases. After removing duplicates, 120 studies remained for the analysis of titles and abstracts. Then, after reading the titles, 15 studies continued to the analysis of abstracts. After analyzing the abstracts, only 11 studies were considered eligible for the full text reading. The references of the 11 studies were carefully assessed to check for studies retrieved through the main search strategy, but none was found. From the 11 studies included in this phase, two were removed for the following reasons: 1) use of Polymerase Chain Reaction (PCR) for the analysis; and 2) diagnostic study. Therefore, nine studies continued to the qualitative analysis of results. Figure 1 reproduces the process of search, identification, inclusion, and exclusion of articles. Table 2 shows a summary of the main characteristics of the studies. Most part of the studies (five) (13,14,(21)(22)(23) were performed in Brazil, while the other four studies were conducted in Mexico (24), Turkey (4), Greece (25), and Malaysia (26). The analysis of the nine studies resulted in a total sample of 285 specimens. From the nine studies analyzed, only three (14,21,22) mentioned the ethical criteria involved. As for specimen fixation, five studies (13,14,21,23,25) used 10% formaldehyde and four studies (4,22,24,26) did not mention the means of fixation. All studies (4,13,14,(21)(22)(23)(24)(25)(26) used immunohistochemistry as the diagnostic method.

Discussion
Odontogenic cysts and tumors are lesions of the oral cavity with the potential to affect the balance of the bone resorption and apposition system (12). However, the scientific literature is still controversial regarding the expression of RANK, RANKL, and OPG in the development of such lesions.
The dentigerous cyst is one of the most common odontogenic cysts of the oral and maxillofacial complex.
the immunoexpression of OPG. The three lesions assessed were stained for RANK in both the odontogenic epithelium and stroma. However, the immunoexpression of RANKL and OPG ranged from 0% to 100% in the odontogenic epithelium and the stroma ( Table 4). The OPG>RANKL ratio ranged from 0% to 70% in the odontogenic epithelium and from 0% to 77.8% in the stroma (Fig. 2), while the OPG<RANKL ratio ranged from 0% to 50.0% in the odontogenic epithelium and from 0% to 75% in the stroma (Fig. 3). The OPG=RANKL ratio ranged from 20.0% to 60% in the odontogenic epithelium and from 0% to 90.0% in the stroma (Fig. 4).

Synthesis of Results and Meta-Analysis
Considering the OPG>RANKL ratio (Fig. 2), the highest estimate was perceived for dentigerous cyst (ES=43.3%; 95% CI=14.3-74.8) analyzed in the epithelium, although bigger, this ratio was statistically similar to that calculated for odontogenic keratocyst (ES=36.8%; 95% CI=18.8-56.7). In contrast, the highest OPG<RANKL ratio was found in      Although it is asymptomatic and rarely aggressive, in some cases it may cause significant bone expansion and teeth dislocation (27). On the other hand, the ameloblastoma and the odontogenic keratocyst present a more invasive biological behavior with a potential destructive growth (28). Hence, several studies (4,13,14,21) have been focused on investigating the correlation of the immunologic expression of the osteoclastogenic factors RANK, RANKL, and OPG in the development of such pathologies. The present study proposed to verify whether the profile of the odontogenic keratocyst is more similar to a cystic or neoplastic lesion, which excluded inflammatory cysts from the analysis. It is worth noting that this study extracted only the results from non-syndromic odontogenic keratocysts, considering that other variables may affect the expression of RANK, RANKL, and OPG in the syndromic odontogenic keratocyst. Immunohistochemistry is a diagnostic method that uses antibodies as specific reagents for detecting antigens in tissue and cell cut-offs, and it is often used for diagnosing neoplasias (29). This type of technique depends on the means of specimen fixation (30), for which 10% formaldehyde is mostly recommended (31). All the eligible studies in the present systematic review performed the immunohistochemistry technique for analyzing the expression of RANK, RANKL, and OPG, agreeing the recommendations of the scientific literature. However, only five studies (13,14,21,23,25) reported specimen fixation in 10% formaldehyde (Table 2).
DNA amplification reactions, as in PCR, are one of the most important molecular biology techniques today (32). Despite this, it is a high-cost method, which involves extraction of genetic material and need experienced professionals to avoid errors inherent to the sensitivity of the technique, such as failure in amplification (33). As a consequence, several laboratories find it difficult to apply it in their routine (34). Based on this, we only included studies that used immunohistochemistry to evaluate the immunoexpression of RANK, RANKL and OPG, because, despite being a technique with less precision than PCR, it is widely used in laboratories.
Most studies (13,14,(21)(22)(23)25) used polyclonal antibodies to perform the immunohistochemical reaction. Polyclonal antibodies have low specificity, as they bind to different epitopes, increasing the chances of promoting a cross reaction and, consequently, generating unspecific markings (35). The widespread use of polyclonal antibodies in studies the present systematic review may have occurred due to their low cost, when compared to monoclonal antibodies. Therefore, we believe -and strongly suggest -that monoclonal antibodies should be chosen for immunohistochemical reactions, whenever possible.
The RANK is a member of the tumor necrosis factor (TNF) receptor family (36) that works as a signaling receptor of RANKL, while the latter is expressed in the osteoblastic cells of the periodontal ligament (9), binding to RANK and activating osteoclasts to promote bone resorption (10). Thus, the RANKL, which is a protein that also belongs to the TNF family, works by regulating osteoclastic activity (37). Confirming such affirmation, six studies (4,13,21,23,24,26) of the present systematic review assessed the immunoexpression of RANK in dentigerous cyst, odontogenic keratocyst, and ameloblastoma, and all of them presented positive expression of RANK in both the odontogenic epithelium and stroma, reinforcing the action of this protein in the osteolytic process (Table 4).
The immunoexpression of RANKL was assessed in all the studies included in this systematic review. From the nine studies included, eight (4,13,14,(21)(22)(23)(24)(25) presented positive expression of RANKL in dentigerous cyst, odontogenic keratocyst, and ameloblastoma in both the odontogenic epithelium and stroma. This result corroborates the findings of the study by Qian and Zuang (38), which assessed 24 ameloblastoma specimens and verified that RANKL was the key factor for the development of osteoclastogenesis. However, in the present systematic review, only one study (26) did not present positive expression of RANKL for ameloblastoma, which is incompatible with the biological behavior of this pathology (Table 4). Ameloblastomas are known to be aggressive and locally invasive/destructive tumors. In some cases, bone expansion is so significant that radical resection is the only form of treatment (38). The referred authors of the study (26) suggest that the expression of RANKL may be absent because either the sample studied was a group of indolent tumors or the undergoing dynamic process was associated with bone remodeling induced by the tumor.
The OPG is a receptor of the TNF family and it is secreted by a number of cell types, including osteoblasts (39). This protein works by inhibiting the osteoclastogenic process, thus blocking the RANKL and RANK binding (11). Hence, studies (40,41) have been developed aiming to investigate its potential as a therapeutic agent for bone diseases. All the studies included in the present systematic review assessed the immunoexpression of OPG, and immunoreactivity was not positive in only two of them (4,26), whereas both verified OPG in the ameloblastoma (Table 4).
Studies (14,25) have shown that assessing the ratio of immunoexpression of RANKL and OPG is particularly important, considering it may indicate the biological activity and the osteolytic potential of the odontogenic cyst or tumor. From the nine studies included in this systematic review, five (13,14,21,22,25) assessed such ratio. Thus, the highest OPG>RANKL ratio was found for odontogenic keratocyst and dentigerous cyst. These findings agree with the biological behavior of the odontogenic keratocyst and the dentigerous cyst, considering these are lesions with lower bone resorption ability than ameloblastoma, which is a very aggressive tumor (42). It is known that RANKL binds to RANK in the surface of osteoclast precursors, recruiting and activating the tumor necrosis factor receptor-associated factor-6 (TRAF-6). Thus, the TRAF-6 stimulates the activation of the nuclear factor-kB (NF-kB) through the interaction of p62 and the atypical protein kinase C (aPKC), which triggers the transcription of osteoclastogenic genes (43). The NF-kB signaling is essential for osteoclastogenesis (44). Thus, the OPG inhibits the osteoclastogenic events by not allowing the binding of RANKL to RANK and preventing the entire activation cascade of NF-kB from being activated (45). Confirming such findings, authors (12) verified a similar ratio with higher prevalence of OPG than RANKL in the calcifying odontogenic cyst, reinforcing the proposal of the cystic nature of the odontogenic keratocyst. Such ratio may suggest that OPG may be involved in other different biological processes of bone remodeling (22).
However, the pathogenesis of the odontogenic keratocyst is still uncertain (46). The Sonic Hedgehog (Shh) activation has been indicated as one of the main mechanisms involved in the progression of this pathology (47). In normal cells, the Patched (PTCH) transmembrane receptor hinders Shh activation (46). However, in neoplastic cells, the Shh protein binds to the PTCH1 receptor, activating the Smoothened (Smo) transmembrane protein and inducing cell proliferation from the expression of Glioma cytoplasmic proteins (Gli-1 and Gli-2) (48). Consequently, the NF-kB is activated, which is a signaling pathway that participates directly in the osteolytic process (44). It is known that PTCH1 works as an inhibitor of Smo in the absence of the Shh ligand. A study (49) featuring the treatment of cerebral ischemia in rats using polydatin verified an overexpression of PTCH1 and a reduction of NF-kB, which may explain indirectly the OPG>RANKL ratio found in the odontogenic keratocyst, considering that the lower NF-kB activation implies in lower expression of RANKL and, consequently, lower osteolytic potential.
In contrast, the OPG<RANKL ratio was higher for ameloblastoma, considering the stromal region is significantly larger. The results agree with the ameloblastoma activity, considering the stromal region participates actively in the invasion and proliferation of tumor cells (50). A study (51) performed with the immunohistochemical marker for detecting myofibroblasts and anti-α-actin smooth muscle antibody (α-SMA) verified that from the 15 cases of solid ameloblastomas examined, only one did not express α-SMA, indicating a high myofibroblast activity in the development of ameloblastomas. Such aggressiveness from the stroma may be explained by the potential changes in the components of the mitogen-activated protein kinase (MAPK), especially in the BRAF gene, which may be activated by the fibroblast growth factor (FGF) (52). The expression of FGF in ameloblastomas, especially FGF1, FGF2, and FGF3, are important for the positive MAPK regulation (53). The MAPK activation allows the phosphorylation of Raf, MEK, and ERK proteins, inducing cell proliferation (54).
Moreover, a recent study (55) suggested that the I. F. P. Lima et al.
transforming growth factor-β (TGF-β) and interleukin-1α (IL-1α) have the ability to induce the expression of RANKL in stromal fibroblasts of ameloblastomas. The positive regulation of RANKL is associated with a negative regulation of OPG, which causes the system to work in favor of osteoclastogenesis (43). The higher expression of RANKL allows its binding to RANK and consequently the activation of NF-kB activation, which is an essential transcription factor for the osteolytic potential of tumor lesions (44). Reinforcing these findings, authors (56) verified a higher immunoexpression of RANKL than OPG in osteosarcoma, which is a tumor similar to ameloblastoma regarding its aggressive and osteolytic behavior. Similarly, Sambandam et al. (57) studied a squamous cell carcinoma sample and verified a positive regulation of RANKL, suggesting that this protein works in osteoclastic differentiation and bone destruction. The present study is not free of limitations since immunohistochemistry is a semi-quantitative technique that, in some cases, may not be objective. Besides, there is no way to control the intensity of staining among the studies evaluated. Because of this, our systematic review did not measure the levels of staining intensity, but only the simple presence of staining. However, this study has strengths that should be considered. This is an original systematic review and meta-analysis, which used an extensive search strategy including the "grey literature", without restrictions of language or publication status. Also, the recent classification of odontogenic cysts and tumors proposed by the WHO (58) establishes that the dentigerous cyst and the odontogenic keratocyst are considered odontogenic cysts, while ameloblastoma is considered a benign odontogenic tumor. Thus, this systematic review corroborates the classification proposed by the WHO. This is particularly important in clinical practice since the odontogenic keratocyst is more similar to the cystic profile, requiring greater care to perform systematic curettage technique after removal of the lesion, in order to reduce the chances of recurrence.
The results of the present systematic review suggest that the OPG<RANKL ratio was higher for ameloblastoma, which may explain its aggressive potential, and it was lower for dentigerous cyst and odontogenic keratocyst, reinforcing the WHO classification of odontogenic keratocyst as an odontogenic cyst.