The key role of RANKL and OPG as molecular biomarkers in bone remodeling associated with different grafts regarding socket preservation: A systematic review Authors: Firas A. Jamil, Saif S. Abdulrazzaq Published: 2026 ABSTRACT ObjectivesBone remodeling usually depends on complex molecular signaling, where receptor activator of NF-κB ligand (RANKL) and osteoprotegerin (OPG) are key modulators that regulate osteoclastogenesis and the overall bone metabolism process. The aims of this systematic review were to evaluate the effects of different graft materials on RANKL and OPG expression following alveolar ridge preservation.MethodsElectronic literature searches were performed in Embase via Ovid, Medline via Ovid, CINAHL, Scopus, and Web of Science databases, in addition to manual searches regarding all related articles. The PRISMA guidelines and PICO framework were followed in this systematic review. The inclusion criteria were limited to studies published in English that focused on RANKL and/or OPG biomarker outcomes in post-extraction dental sockets preserved with various types of bone grafts. The extracted data were synthesized in a standardized MS Excel form, including general study characteristics, graft interventional details, and RANKL/OPG expression, in addition to other outcomes regarding bone healing. Risk of bias was assessed using ROBINS-I and RYRCLE tools, as well as certainty of evidence for the studies included.ResultsThe initial search yielded 307 articles. After thorough screening and reviewing stages, 12 studies were finally selected for inclusion and comprehensive analysis. The findings showed that socket preservation with different grafts led to significant differences in promoting RANKL downregulation and OPG upregulation compared with controls, where composite grafts obtained superior outcomes than other types. The lowest RANKL/OPG ratio (0.11) was reported for a biologically enhanced autograft, whereas using a synthetic graft alone had the highest value (1.88).ConclusionBone substitutes that combine bioactive agents with scaffolds may provide potential benefits for socket preservation, possibly through modulating the RANKL/OPG axis. However, their superior performance compared with other grafts and the reliability of using the RANKL/OPG ratio as a predictive biomarker remain uncertain, warranting further clinical investigation. Introduction Clinicians have employed socket preservation techniques to counter the natural bone remodeling process, where these techniques involve placing graft materials into the extraction site to maintain ridge dimensions and minimize bone resorption.1 A wide range of biomaterials, such as autografts, allografts, xenografts, and alloplasts, have been utilized in clinical settings. Their effectiveness in alveolar ridge preservation has been studied for standalone materials2, 3, 4, 5, 6, 7, 8, 9 or in combination as composite grafts.4,10, 11, 12, 13 The physical and mechanical characteristics of these grafts are well established, but we still lack a full understanding of how they affect the molecular and cellular events underlying bone remodeling, especially the regulatory mechanisms that control osteoclast and osteoblast activities. Two critical biomarkers play essential roles in the bone remodeling process. The first is receptor activator of NF-κB ligand (RANKL), which is a cytokine necessary for the activation, differentiation, and survival of osteoclasts, where increased levels of this biomarker promote bone resorption.14 The second biomarker, osteoprotegerin (OPG), acts as a decoy receptor for RANKL. By binding to RANKL, OPG prevents it from interacting with receptor activator of NF-κB (RANK) expressed on the surfaces of osteoclasts, thereby limiting the development of bone-resorbing cells and reducing bone loss.15 The RANKL/OPG ratio serves as a key indicator of bone turnover balance, where an increased ratio suggests greater resorptive activity, whereas a decreased ratio denotes a favorable osteogenic environment for bone formation.16 Numerous clinical and preclinical studies have assessed RANKL and OPG in various bone healing frameworks,17, 18, 19, 20, 21, 22 but systematic reviews of the literature have not been conducted to consolidate evidence about their specific expression in socket preservation procedures using diverse graft types. By addressing these molecular mechanisms, clinicians can select graft materials by considering their structural integrity as well as their effect on the biochemical modulation of bone healing. Therefore, the aims of this systematic review were to synthesize the available clinical evidence on how RANKL and OPG regulate bone healing in the context of socket preservation by comparing their expression profiles under different types of graft materials. In addition, possible correlations between changes in biomarker levels and clinical, radiographic, or histological outcomes were identified. To the best of the authors’ knowledge, this is the first systematic review to comprehensively evaluate the key regulatory roles of RANKL and OPG biomarkers in the remodeling process for post-extraction bony sites with respect to socket preservation technique using various grafts. Focus questions The current systematic review was designed to answer the following questions developed according to the PICO (population, intervention, comparison, and outcome) framework. How can the expression levels of RANKL and OPG affect the osteogenic potential of graft materials in socket preservation techniques? In addition, can RANKL and OPG be used as biomarkers for predicting the success of graft integration? Materials and Methods Study design This systematic review was carried out in accordance with the updated 2020 PRISMA guidelines,23 which provide best practices for reporting systematic reviews and meta-analyses. Inclusion criteria 1.Population: human or animal subjects underwent tooth extraction and socket preservation techniques with graft materials.2.Intervention: any type of bone substitute.3.Comparator: sites with no graft or comparison between different graft types.4.Outcomes: expression levels or ratios of RANKL and/or OPG biomarkers.5.Study types: clinical and preclinical studies.6.Language: articles published in English.7.Publication date: no limit. 1.Population: human or animal subjects underwent tooth extraction and socket preservation techniques with graft materials. Population: human or animal subjects underwent tooth extraction and socket preservation techniques with graft materials. 2.Intervention: any type of bone substitute. Intervention: any type of bone substitute. 3.Comparator: sites with no graft or comparison between different graft types. Comparator: sites with no graft or comparison between different graft types. 4.Outcomes: expression levels or ratios of RANKL and/or OPG biomarkers. Outcomes: expression levels or ratios of RANKL and/or OPG biomarkers. 5.Study types: clinical and preclinical studies. Study types: clinical and preclinical studies. 6.Language: articles published in English. Language: articles published in English. 7.Publication date: no limit. Publication date: no limit. Exclusion criteria 1.Reviews, unpublished theses, conference abstracts only without full data, book chapters, editorials, short communications, and case reports.2.Records without graft materials or using biologics/membranes only for socket preservation.3.Articles that reported cases with compromised socket walls.4.Studies that did not evaluate RANKL/OPG or that measured molecular markers in tissues other than bone, such as gingival crevicular fluid, saliva, or blood. Outcomes obtained from these fluids may be influenced by systemic conditions and other local inflammatory processes, and may not accurately reflect molecular events that occur in the grafted extraction socket. Therefore, direct assessment of RANKL/OPG biomarkers in bone tissue provides the most relevant evidence regarding local osteoclastic and osteoblastic activity levels. 1.Reviews, unpublished theses, conference abstracts only without full data, book chapters, editorials, short communications, and case reports. Reviews, unpublished theses, conference abstracts only without full data, book chapters, editorials, short communications, and case reports. 2.Records without graft materials or using biologics/membranes only for socket preservation. Records without graft materials or using biologics/membranes only for socket preservation. 3.Articles that reported cases with compromised socket walls. Articles that reported cases with compromised socket walls. 4.Studies that did not evaluate RANKL/OPG or that measured molecular markers in tissues other than bone, such as gingival crevicular fluid, saliva, or blood. Outcomes obtained from these fluids may be influenced by systemic conditions and other local inflammatory processes, and may not accurately reflect molecular events that occur in the grafted extraction socket. Therefore, direct assessment of RANKL/OPG biomarkers in bone tissue provides the most relevant evidence regarding local osteoclastic and osteoblastic activity levels. Studies that did not evaluate RANKL/OPG or that measured molecular markers in tissues other than bone, such as gingival crevicular fluid, saliva, or blood. Outcomes obtained from these fluids may be influenced by systemic conditions and other local inflammatory processes, and may not accurately reflect molecular events that occur in the grafted extraction socket. Therefore, direct assessment of RANKL/OPG biomarkers in bone tissue provides the most relevant evidence regarding local osteoclastic and osteoblastic activity levels. Data sources and search strategy Initial electronic searches for appropriate publications were conducted on June 3, 2025 using the following databases: Embase via Ovid, Medline via Ovid, CINAHL, Scopus, and Web of Science. On the same day, an additional manual assessment was implemented using the Google search engine (website) in various scientific journals (gray literature) to ensure that the screening process was comprehensive, and only peer-reviewed and appropriately indexed publications were included. Keywords were categorized into three main domains: condition under study (socket preservation), intervention (graft), and main outcomes involving all clinical and preclinical models (RANKL/OPG). The Boolean system utilized in the search strategy for the defined databases is presented in Supplementary File S1. The citations obtained were imported into Covidence online software (Melbourne, Australia), where all duplicates and ineligible entries were filtered out using automated tools and by manual review. Study selection process and data synthesis After the initial systematic search, two investigators independently reviewed all titles and abstracts, before detailed full-text assessment of the potentially eligible articles. Any disagreement or uncertainty between the reviewers was resolved through discussion. The independent reviewers then extracted data from the studies that met the inclusion criteria using a standardized form created in MS Excel. In cases where essential data were unclear, missing, or incomplete, the corresponding author was contacted by email to provide a clear description of the information required within a reasonable timeframe to respond. If no response was received within the specified period, the study was included in the analysis depending on the available data, and the limitation was documented accordingly. This approach was undertaken to minimize potential bias, and improve data completeness and accurate representation of the review's findings. The recorded information including the study characteristics (authors, year of publication, study type/design, sample size, subjects, and number of groups), intervention details (graft type/name, surgical data, and follow-up periods), and RANKL/OPG expression records (method of assessment, levels, ratio, and fold change over time), Furthermore, additional variables were assessed, including other biomarkers, clinical/radiographic findings, histological/histomorphometric analysis, bone formation/remodeling outcomes, inflammatory response, and complications. Risk of bias and certainty of evidence within studies The risk of bias assessment was conducted by two independent reviewers and any conflicts were resolved by discussion. The ROBINS-I version 2 (2024) tool was applied to assess the risk of bias in the non-randomized clinical trials. This tool evaluates seven bias domains, and each domain is rated as low, moderate, serious, or critical. The overall risk of bias is determined by the highest level of bias observed in any of the domains. For animal studies, the risk of bias was assessed using the SYRCLE tool, which is adapted from the Cochrane RoB tool but specifically designed for preclinical in vivo studies. This tool assesses 10 domains categorized under selection bias, performance bias, detection bias, attrition bias, reporting bias, and other biases, and each item is scored as yes (low risk), no (high risk), or unclear. The certainty of evidence was evaluated using the GRADE framework. This approach has four levels (high, moderate, low, and very low) and considers factors such as risk of bias, indirectness, inconsistency, imprecision, and publication bias. Statistical analysis The extracted data were exported into Microsoft® Excel for Mac (Microsoft 365, Version 16.99) and descriptive statistics were generated. Inter-rater agreement (percentage) and Cohen's kappa coefficient (κ) were calculated using Statistical Package for Social Science (SPSSTM) software version 24.0 (IBM Corporation, Illinois, Chicago, United States). A meta-analysis was not conducted due to the wide heterogeneity among the studies included, where they varied in terms of the study designs, participant characteristics (different animal models or human subjects), diverse types of grafting materials used, and inconsistent follow-up periods for measuring RANKL and OPG expression. In addition, the outcome data were reported in incompatible formats (e.g., mean ± standard deviation, or graphical representation only), limiting the ability to derive standardized effect sizes for quantitative synthesis. Results Study selection In total, 296 articles were initially identified by the database search. After importing these citations into Covidence, 149 duplicate entries were removed, before performing the screening process (Figure 1). The titles and abstracts of the remaining 147 records were reviewed and after resolving any conflicts between reviewers, 133 articles were excluded. The reviewers achieved a probability of agreement (Po) of 94.94% with a Cohen's kappa (κ) value of 0.781, reflecting substantial agreement according to the Landis and Koch scale (1977).Figure 1PRISMA flow diagram showing the literature search and study selection process applied in this systematic review. PRISMA flow diagram showing the literature search and study selection process applied in this systematic review. Fourteen articles were selected for full-text screening and eight were excluded for various reasons (Figure 1). During this stage, Po = 95.45% and κ = 0.904, indicating almost perfect agreement between reviewers. Finally, 12 studies were included for detailed data extraction and analysis, where six were identified from other sources (Figure 1). Study characteristics Studies were published between 2017 and 2025. There was one clinical trial24 and the others involved preclinical models.25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35 In total, 437 extraction sites were evaluated. Ten studies performed ridge preservation on anterior teeth sockets (incisors),25, 26, 27, 28, 29, 30, 31, 32,34,35 one included posterior teeth (premolars),33 and information regarding the type of teeth was unclear in one article.24 Detailed data regarding general characteristics of the articles included are presented in Table 1.Table 1General characteristics of the studies included.Table 1Study (Year)Study type/designSample typeGenderSample size (total no./subjects per group)No. of groupsTooth typeBone graft typeGraft nameAdditional biomaterial/groupCanullo et al. (2016)24Clinical trial/Prospective longitudinal case seriesHumanBoth15/151UnclearSyntheticMgHACollagen membrane/TxDjais et al. (2020)25Animal/In vivo experimental study with post-test onlyGuinea pigMale45/95IncisorCompositeMO + DFDDMNoneDjais et al. (2023)26Animal/In vivo experimental study with post-test onlyGuinea pigMale48/124IncisorCompositeChitosan (Milkfish scales) + BBXNoneIlham et al. (2025)27Animal/In vivo experimental study with post-test onlyGuinea pigMale27/93IncisorXenograftPearl oyster CaCO3 (PMNC-P)NoneKamadjaja et al. (2021)28Animal/In vivo experimental study with post-test onlyGuinea pigMale24/64IncisorXenograftHA scaffold (crab shell)Gelatin/TxKresnoadi et al. (2017)29Animal/In vivo experimental study with post-test onlyGuinea pigMale56/78IncisorCompositeMPE + DFDBBXPEG/AllKresnoadi et al. (2023a)30Animal/In vivo experimental study with post-test onlyGuinea pigMale56/144IncisorCompositeNigella sativa + BBGPEG/AllKresnoadi et al. (2023b)31Animal/In vivo experimental study with post-test onlyGuinea pigMale56/78IncisorCompositePropolis + BBGPEG/AllPrahasanti et al. (2019)32Animal/In vivo experimental study with post-test onlyWistar ratMale14/72AnteriorCell-based compositeHA scaffold + SHEDNoneRam et al. (2023)33Animal/Comparative randomized studySheepFemale30/102PremolarXenograftBovine-derived xenograft (Bio-Oss®)Collagen membrane/TxSaskianti et al. (2020)34Animal/In vivo experimental study with post-test onlyWistar ratMale10/52AnteriorCell-based compositeCAS + SHEDNoneSoekobagiono et al. (2017)35Animal/In vivo experimental study with factorial randomizationGuinea pigMale56/78IncisorCompositeMO + DFDBBXPEG/ControlNo. = number; MgHA = Magnesium enriched hydroxyapatite; Tx = treatment; MO = Moringa oleifera; DFDDM = demineralization freeze dried dentin matrix; BBX = bovine bone xenograft; PMNC-P = calcium carbonate powder of non-cultivated Pinctada maxima; HA = hydroxyapatite; MPE = mangosteen peel extract; DFDBBX = demineralized freeze-dried bovine bone xenograft; PEG = polyethylene glycol; BBG = bovine bone graft; SHED = stem cell from human exfoliated deciduous teeth; CAS = carbonate apatite scaffold. Table 1General characteristics of the studies included.Table 1Study (Year)Study type/designSample typeGenderSample size (total no./subjects per group)No. of groupsTooth typeBone graft typeGraft nameAdditional biomaterial/groupCanullo et al. (2016)24Clinical trial/Prospective longitudinal case seriesHumanBoth15/151UnclearSyntheticMgHACollagen membrane/TxDjais et al. (2020)25Animal/In vivo experimental study with post-test onlyGuinea pigMale45/95IncisorCompositeMO + DFDDMNoneDjais et al. (2023)26Animal/In vivo experimental study with post-test onlyGuinea pigMale48/124IncisorCompositeChitosan (Milkfish scales) + BBXNoneIlham et al. (2025)27Animal/In vivo experimental study with post-test onlyGuinea pigMale27/93IncisorXenograftPearl oyster CaCO3 (PMNC-P)NoneKamadjaja et al. (2021)28Animal/In vivo experimental study with post-test onlyGuinea pigMale24/64IncisorXenograftHA scaffold (crab shell)Gelatin/TxKresnoadi et al. (2017)29Animal/In vivo experimental study with post-test onlyGuinea pigMale56/78IncisorCompositeMPE + DFDBBXPEG/AllKresnoadi et al. (2023a)30Animal/In vivo experimental study with post-test onlyGuinea pigMale56/144IncisorCompositeNigella sativa + BBGPEG/AllKresnoadi et al. (2023b)31Animal/In vivo experimental study with post-test onlyGuinea pigMale56/78IncisorCompositePropolis + BBGPEG/AllPrahasanti et al. (2019)32Animal/In vivo experimental study with post-test onlyWistar ratMale14/72AnteriorCell-based compositeHA scaffold + SHEDNoneRam et al. (2023)33Animal/Comparative randomized studySheepFemale30/102PremolarXenograftBovine-derived xenograft (Bio-Oss®)Collagen membrane/TxSaskianti et al. (2020)34Animal/In vivo experimental study with post-test onlyWistar ratMale10/52AnteriorCell-based compositeCAS + SHEDNoneSoekobagiono et al. (2017)35Animal/In vivo experimental study with factorial randomizationGuinea pigMale56/78IncisorCompositeMO + DFDBBXPEG/ControlNo. = number; MgHA = Magnesium enriched hydroxyapatite; Tx = treatment; MO = Moringa oleifera; DFDDM = demineralization freeze dried dentin matrix; BBX = bovine bone xenograft; PMNC-P = calcium carbonate powder of non-cultivated Pinctada maxima; HA = hydroxyapatite; MPE = mangosteen peel extract; DFDBBX = demineralized freeze-dried bovine bone xenograft; PEG = polyethylene glycol; BBG = bovine bone graft; SHED = stem cell from human exfoliated deciduous teeth; CAS = carbonate apatite scaffold. General characteristics of the studies included. No. = number; MgHA = Magnesium enriched hydroxyapatite; Tx = treatment; MO = Moringa oleifera; DFDDM = demineralization freeze dried dentin matrix; BBX = bovine bone xenograft; PMNC-P = calcium carbonate powder of non-cultivated Pinctada maxima; HA = hydroxyapatite; MPE = mangosteen peel extract; DFDBBX = demineralized freeze-dried bovine bone xenograft; PEG = polyethylene glycol; BBG = bovine bone graft; SHED = stem cell from human exfoliated deciduous teeth; CAS = carbonate apatite scaffold. Table 2 shows the main biomarker results, where nine articles reported outcomes for both RANKL and OPG,24, 25, 26,28,30, 31, 32, 33, 34 whereas three only evaluated RANKL.27,29,35 Most of the studies utilized immunohistochemistry (IHC) as the method for marker detection,25, 26, 27, 28, 29, 30, 31, 32,34,35 and two used IHC and reverse transcription quantitative polymerase chain reaction (RT-qPCR) techniques.24,33 Different time points ranging from 3 days to 4 months were employed. RANKL/OPG ratios were obtained in five studies,24, 25, 26,32,34 with a minimum value of 0.1125 and a maximum value of 1.88.24Table 2Main biomarker outcomes for studies included in terms of the treatment groups compared with baseline or control groups at the final follow-up point.Table 2Study (Year)Bone graft typeBiomarkers (RANKL, OPG)Detection methodTime pointsRANKL level (Mean ± SD)OPG level (Mean ± SD)RANKL/OPG ratioCanullo et al. (2016)24SyntheticBothIHC + RT-qPCRBaseline (0), 4 months↑ (NS)↑ (NS)↑ (1.88)Djais et al. (2020)25CompositeBothIHC7, 14, 21 days↓ (3.00 ± 1.00)↑ (26.33 ± 0.52)↓ (0.11)Djais et al. (2023)26CompositeBothIHC3, 7, 14, 28 days↓ (7.67 ± 1.53)↑ (12.67 ± 2.52)↓ (0.61)Ilham et al. (2025)27XenograftRANKLIHC7, 14, 21 days↓ (2.67 ± 1.53)NANAKamadjaja et al. (2021)28XenograftBothIHC7, 14 days↓ Lowest (NS)↑ Highest (NS)↓ (Inferred <1)Kresnoadi et al. (2017)29CompositeRANKLIHC7, 30 days↓ (5.17 ± 1.47)NANAKresnoadi et al. (2023a)30CompositeBothIHC7, 14 days↓ Lowest (NS)↑ Highest (NS)↓ (Inferred <1)Kresnoadi et al. (2023b)31CompositeBothIHC3, 7 days↓ Lowest (NS)↑ Highest (NS)↓ (Inferred <1)Prahasanti et al. (2019)32Cell-based compositeBothIHC8 weeks↓ (4.80 ± 1.30)↑ (11.60 ± 1.14)↓ (0.41)Ram et al. (2023)33XenograftBothIHC + RT-qPCR4, 8, 16 weeksComparable (NS)Comparable (NS)No directional dominanceSaskianti et al. (2020)34Cell-based compositeBothIHC7 days↓ (4.80 ± 0.58)↑ (11.60 ± 1.08)↓ (0.41)Soekobagiono et al. (2017)35CompositeRANKLIHC7, 30 days↓ Lowest (NS)NANASD = standard deviation; IHC = immunohistochemistry; RT-qPCR = reverse transcription quantitative polymerase chain reaction; ↑ = increase; ↓ = decrease; NS = not specified; NA = not applicable. Table 2Main biomarker outcomes for studies included in terms of the treatment groups compared with baseline or control groups at the final follow-up point.Table 2Study (Year)Bone graft typeBiomarkers (RANKL, OPG)Detection methodTime pointsRANKL level (Mean ± SD)OPG level (Mean ± SD)RANKL/OPG ratioCanullo et al. (2016)24SyntheticBothIHC + RT-qPCRBaseline (0), 4 months↑ (NS)↑ (NS)↑ (1.88)Djais et al. (2020)25CompositeBothIHC7, 14, 21 days↓ (3.00 ± 1.00)↑ (26.33 ± 0.52)↓ (0.11)Djais et al. (2023)26CompositeBothIHC3, 7, 14, 28 days↓ (7.67 ± 1.53)↑ (12.67 ± 2.52)↓ (0.61)Ilham et al. (2025)27XenograftRANKLIHC7, 14, 21 days↓ (2.67 ± 1.53)NANAKamadjaja et al. (2021)28XenograftBothIHC7, 14 days↓ Lowest (NS)↑ Highest (NS)↓ (Inferred <1)Kresnoadi et al. (2017)29CompositeRANKLIHC7, 30 days↓ (5.17 ± 1.47)NANAKresnoadi et al. (2023a)30CompositeBothIHC7, 14 days↓ Lowest (NS)↑ Highest (NS)↓ (Inferred <1)Kresnoadi et al. (2023b)31CompositeBothIHC3, 7 days↓ Lowest (NS)↑ Highest (NS)↓ (Inferred <1)Prahasanti et al. (2019)32Cell-based compositeBothIHC8 weeks↓ (4.80 ± 1.30)↑ (11.60 ± 1.14)↓ (0.41)Ram et al. (2023)33XenograftBothIHC + RT-qPCR4, 8, 16 weeksComparable (NS)Comparable (NS)No directional dominanceSaskianti et al. (2020)34Cell-based compositeBothIHC7 days↓ (4.80 ± 0.58)↑ (11.60 ± 1.08)↓ (0.41)Soekobagiono et al. (2017)35CompositeRANKLIHC7, 30 days↓ Lowest (NS)NANASD = standard deviation; IHC = immunohistochemistry; RT-qPCR = reverse transcription quantitative polymerase chain reaction; ↑ = increase; ↓ = decrease; NS = not specified; NA = not applicable. Main biomarker outcomes for studies included in terms of the treatment groups compared with baseline or control groups at the final follow-up point. SD = standard deviation; IHC = immunohistochemistry; RT-qPCR = reverse transcription quantitative polymerase chain reaction; ↑ = increase; ↓ = decrease; NS = not specified; NA = not applicable. Regarding additional outcomes (Table 3), four studies evaluated other biomarkers in addition to RANKL and OPG,24,29,33,34 three studies reported clinical findings,24,33,34 one study obtained intraoral radiographs for surgical confirmation of complete tooth removal,33 three studies provided histological findings about bone remodeling process,24,29,33 and only one study described histomorphometric analysis.24 No studies reported any complications or side effects at the final time point evaluation.Table 3Additional relevant data for the studies included at the final time points.Table 3Study (Year)Bone graft typeOther markersClinical outcomes/Radiographic findingsHistology/Immunohistochemical findings of other markersHistomorphometric analysis (% ± SD)Canullo et al. (2016)24SyntheticRANKIL-6TNF-⍺No clinical symptoms or complications/NPWell-structured healing tissue, no inflammatory infiltrate, with partially osteointegrated residual MgHA granules/↑ expression levels of all other markers.Newly formed bone (16.50 ± 4.66)Osteoid tissue (9.66 ± 4.33)Residual MgHA biomaterial (38.86 ± 8.62)Connective tissue (34.98 ± 9.46)Djais et al. (2020)25CompositeNoneNR/NPNP/NANPDjais et al. (2023)26CompositeNoneNR/NPNP/NANPIlham et al. (2025)27XenograftNoneNR/NPNP/NANPKamadjaja et al. (2021)28XenograftNoneNR/NPNP/NANPKresnoadi et al. (2017)29CompositeNF-κBBMP-2NR/NPHighest osteoblast no. (12.50 ± 1.05 SD) with lowest osteoclast no. (8.50 ± 1.76 SD)/↓ NF-κB, and ↑ BMP-2.NPKresnoadi et al. (2023a)30CompositeNoneNR/NPNP/NANPKresnoadi et al. (2023b)31CompositeNoneNR/NPNP/NANPPrahasanti et al. (2019)32Cell-based compositeNoneNR/NPNP/NANPRam et al. (2023)33XenograftRANKCOL1A1TIMP3SP7MSX2Complete soft tissue healing without any complications/Intraoral PA for surgical confirmation.Greater woven bone, more structured bony bridge, with more apparent lamellar bone in test (grafted) group/Comparable expression levels of all other bone remodeling markers.NPSaskianti et al. (2020)34Cell-based compositeTGF-βVEGFRUNX2ALPOCNOPNNo adverse reactions, edema, weight loss, toxicity or death/NPNP/↑ expression levels of all other markers.NPSoekobagiono et al. (2017)35CompositeNoneNR/NPNP/NANPSD = standard deviation; RANK = receptor activator of NF-κB; IL-6 = interleukin-6; TNF-⍺ = tumor necrosis factor-alpha; NF-κB = nuclear factor kappa B; BMP-2 = bone morphogenetic protein-2; TGF-β = transforming growth factor-βeta; VEGF = vascular endothelial growth factor; ALP = alkaline phosphatase; OCN = osteocalcin; OPN = osteopontin; NP = not performed; PA = peri-apical; NR = not reported; NA = not applicable; ↑ = increase; ↓ = decrease, no. = number. Table 3Additional relevant data for the studies included at the final time points.Table 3Study (Year)Bone graft typeOther markersClinical outcomes/Radiographic findingsHistology/Immunohistochemical findings of other markersHistomorphometric analysis (% ± SD)Canullo et al. (2016)24SyntheticRANKIL-6TNF-⍺No clinical symptoms or complications/NPWell-structured healing tissue, no inflammatory infiltrate, with partially osteointegrated residual MgHA granules/↑ expression levels of all other markers.Newly formed bone (16.50 ± 4.66)Osteoid tissue (9.66 ± 4.33)Residual MgHA biomaterial (38.86 ± 8.62)Connective tissue (34.98 ± 9.46)Djais et al. (2020)25CompositeNoneNR/NPNP/NANPDjais et al. (2023)26CompositeNoneNR/NPNP/NANPIlham et al. (2025)27XenograftNoneNR/NPNP/NANPKamadjaja et al. (2021)28XenograftNoneNR/NPNP/NANPKresnoadi et al. (2017)29CompositeNF-κBBMP-2NR/NPHighest osteoblast no. (12.50 ± 1.05 SD) with lowest osteoclast no. (8.50 ± 1.76 SD)/↓ NF-κB, and ↑ BMP-2.NPKresnoadi et al. (2023a)30CompositeNoneNR/NPNP/NANPKresnoadi et al. (2023b)31CompositeNoneNR/NPNP/NANPPrahasanti et al. (2019)32Cell-based compositeNoneNR/NPNP/NANPRam et al. (2023)33XenograftRANKCOL1A1TIMP3SP7MSX2Complete soft tissue healing without any complications/Intraoral PA for surgical confirmation.Greater woven bone, more structured bony bridge, with more apparent lamellar bone in test (grafted) group/Comparable expression levels of all other bone remodeling markers.NPSaskianti et al. (2020)34Cell-based compositeTGF-βVEGFRUNX2ALPOCNOPNNo adverse reactions, edema, weight loss, toxicity or death/NPNP/↑ expression levels of all other markers.NPSoekobagiono et al. (2017)35CompositeNoneNR/NPNP/NANPSD = standard deviation; RANK = receptor activator of NF-κB; IL-6 = interleukin-6; TNF-⍺ = tumor necrosis factor-alpha; NF-κB = nuclear factor kappa B; BMP-2 = bone morphogenetic protein-2; TGF-β = transforming growth factor-βeta; VEGF = vascular endothelial growth factor; ALP = alkaline phosphatase; OCN = osteocalcin; OPN = osteopontin; NP = not performed; PA = peri-apical; NR = not reported; NA = not applicable; ↑ = increase; ↓ = decrease, no. = number. Additional relevant data for the studies included at the final time points. SD = standard deviation; RANK = receptor activator of NF-κB; IL-6 = interleukin-6; TNF-⍺ = tumor necrosis factor-alpha; NF-κB = nuclear factor kappa B; BMP-2 = bone morphogenetic protein-2; TGF-β = transforming growth factor-βeta; VEGF = vascular endothelial growth factor; ALP = alkaline phosphatase; OCN = osteocalcin; OPN = osteopontin; NP = not performed; PA = peri-apical; NR = not reported; NA = not applicable; ↑ = increase; ↓ = decrease, no. = number. Xenograft Three studies with a total of 81 extraction sites only utilized xenograft as a biomaterial for socket preservation (Table 1). Calcium carbonate (CaCO3) derived from pearl oyster shells was evaluated and the study reported progressive RANKL suppression from 8.00 ± 1.00 on day 7 to 2.67 ± 1.53 on day 21 (Table 2) in the study group, with a highly significant difference (p = 0.001, Table 4).27 Compared with the negative control group, the fold change in RANKL decreased to 4.75× on day 21. OPG was not measured. Another study used a crab shell-derived hydroxyapatite (HA) scaffold.28 On day 14, the lowest RANKL and highest OPG expression levels were in the treatment group (Table 2). No numerical values were reported, but the differences were statistically significant (p < 0.001, Table 4). Furthermore, a bovine-derived xenograft (Bio-Oss®) was studied.33 Gene expression results indicated comparable measurements between the two groups at all time points, with no significant difference (p > 0.05). However, histology confirmed enhanced woven bone formation in the grafted sockets (Table 3).Table 4Statistical findings in terms of RANKL and OPG levels considering each specific graft type within the studies included.Table 4Study (Year)Bone graft typeFold changeStatistical significance (p-value)Correlation analysis (RANKL vs OPG)InterpretationCanullo et al. (2016)24SyntheticRANKL gene and protein ↑ (NS)OPG gene ↑ ≈ 7× , but protein ↓ ≈ 0.6×Both genes = not significantRANKL protein (0.001)∗∗OPG protein (> 0.05)NS (Inferred: direct for genes as both ↑, inverse for proteins as one ↑ while the other ↓).Prolonged tissue remodeling of the synthetic graft.Djais et al. (2020)25CompositeRANKL ↓ ≈ 0.50× OPG ↑ ≈ 1.28×Both (0.000)∗∗NS (Inferred: Inverse as one ↓ while the other ↑).Enhanced bone formation potential with composite graft.Djais et al. (2023)26CompositeRANKL ↓ ≈ 1.35× OPG ↑ ≈ 2.23×RANKL (0.22)OPG (0.01)∗NS (Inferred: Inverse as one ↓ while the other ↑).Moderate bone formation pattern.Ilham et al. (2025)27XenograftRANKL ↓ ≈ 4.75×RANKL (0.001)∗∗NARobust osteoclastic suppression.Kamadjaja et al. (2021)28XenograftQualitativeRANKL ↓↓OPG ↑↑Both (0.000)∗∗NS (Inferred: Inverse as one ↓ while the other ↑).Favorable osteogenic environment.Kresnoadi et al. (2017)29CompositeRANKL ↓ ≈ 3.38×RANKL (0.002)∗∗NASuppressed bone resorption.Kresnoadi et al. (2023a)30CompositeQualitativeRANKL ↓↓OPG ↑↑Both (0.000)∗∗NS (Inferred: Inverse as one ↓ while the other ↑).Strong bone regeneration.Kresnoadi et al. (2023b)31CompositeQualitativeRANKL ↓↓OPG ↑↑Both (0.000)∗∗Strong inverse: r = – 0.665 (p < 0.05).Best response with most osteoprotective behavior.Prahasanti et al. (2019)32Cell-based compositeRANKL ↓ ≈ 2.64×OPG ↑ ≈ 1.93×RANKL (0.0005)∗∗OPG (0.0004)∗∗Strong inverse: r = – 0.912 (p = 0.0016).Promoted osteogenic potential.Ram et al. (2023)33XenograftBalancedBoth (> 0.05)NS (Inferred: direct as the levels for both were comparable).Ongoing bone remodeling of the xenograft.Saskianti et al. (2020)34Cell-based compositeRANKL ↓ ≈ 2.54×OPG ↑ ≈ 2.15×RANKL (0.0001)∗∗OPG (0.001)∗∗Strong inverse: r = – 0.0801 (p = 0.005).Improved bone regeneration.Soekobagiono et al. (2017)35CompositeQualitativeRANKL ↓↓RANKL (0.000)∗∗NAStrong osteoprotective and anti-resorptive pattern.↑ = increase; ↓ = decrease; NS = not specified; NA = not applicable; ∗ = significant; ∗∗ = highly significant. Table 4Statistical findings in terms of RANKL and OPG levels considering each specific graft type within the studies included.Table 4Study (Year)Bone graft typeFold changeStatistical significance (p-value)Correlation analysis (RANKL vs OPG)InterpretationCanullo et al. (2016)24SyntheticRANKL gene and protein ↑ (NS)OPG gene ↑ ≈ 7× , but protein ↓ ≈ 0.6×Both genes = not significantRANKL protein (0.001)∗∗OPG protein (> 0.05)NS (Inferred: direct for genes as both ↑, inverse for proteins as one ↑ while the other ↓).Prolonged tissue remodeling of the synthetic graft.Djais et al. (2020)25CompositeRANKL ↓ ≈ 0.50× OPG ↑ ≈ 1.28×Both (0.000)∗∗NS (Inferred: Inverse as one ↓ while the other ↑).Enhanced bone formation potential with composite graft.Djais et al. (2023)26CompositeRANKL ↓ ≈ 1.35× OPG ↑ ≈ 2.23×RANKL (0.22)OPG (0.01)∗NS (Inferred: Inverse as one ↓ while the other ↑).Moderate bone formation pattern.Ilham et al. (2025)27XenograftRANKL ↓ ≈ 4.75×RANKL (0.001)∗∗NARobust osteoclastic suppression.Kamadjaja et al. (2021)28XenograftQualitativeRANKL ↓↓OPG ↑↑Both (0.000)∗∗NS (Inferred: Inverse as one ↓ while the other ↑).Favorable osteogenic environment.Kresnoadi et al. (2017)29CompositeRANKL ↓ ≈ 3.38×RANKL (0.002)∗∗NASuppressed bone resorption.Kresnoadi et al. (2023a)30CompositeQualitativeRANKL ↓↓OPG ↑↑Both (0.000)∗∗NS (Inferred: Inverse as one ↓ while the other ↑).Strong bone regeneration.Kresnoadi et al. (2023b)31CompositeQualitativeRANKL ↓↓OPG ↑↑Both (0.000)∗∗Strong inverse: r = – 0.665 (p < 0.05).Best response with most osteoprotective behavior.Prahasanti et al. (2019)32Cell-based compositeRANKL ↓ ≈ 2.64×OPG ↑ ≈ 1.93×RANKL (0.0005)∗∗OPG (0.0004)∗∗Strong inverse: r = – 0.912 (p = 0.0016).Promoted osteogenic potential.Ram et al. (2023)33XenograftBalancedBoth (> 0.05)NS (Inferred: direct as the levels for both were comparable).Ongoing bone remodeling of the xenograft.Saskianti et al. (2020)34Cell-based compositeRANKL ↓ ≈ 2.54×OPG ↑ ≈ 2.15×RANKL (0.0001)∗∗OPG (0.001)∗∗Strong inverse: r = – 0.0801 (p = 0.005).Improved bone regeneration.Soekobagiono et al. (2017)35CompositeQualitativeRANKL ↓↓RANKL (0.000)∗∗NAStrong osteoprotective and anti-resorptive pattern.↑ = increase; ↓ = decrease; NS = not specified; NA = not applicable; ∗ = significant; ∗∗ = highly significant. Statistical findings in terms of RANKL and OPG levels considering each specific graft type within the studies included. ↑ = increase; ↓ = decrease; NS = not specified; NA = not applicable; ∗ = significant; ∗∗ = highly significant. Composite graft (biologically enhanced autograft/xenograft) Six studies with a total of 317 extraction sockets reported the use of composite grafts (Table 1). In a study that applied Moringa oleifera extract plus demineralization freeze-dried dentin matrix (DFDDM),25 the outcomes indicated a decrease in RANKL (3.00 ± 1.00) and increase in OPG (26.33 ± 0.52), with the most favorable effects on day 21 (Table 2). The RANKL/OPG ratio was lowest at this time point (0.11), and the differences were statistically significant (p < 0.001, Table 4). Chitosan combined with bovine bone xenograft also led to a decrease in RANKL (7.67 ± 1.53) and increase in OPG (12.67 ± 2.52), where the RANKL/OPG ratio dropped to 0.61 on day 28 compared with the placebo group.26 The differences were statistically significant for OPG (p = 0.01), but not significant for RANKL (p = 0.22, Table 4). Another study that only evaluated RANKL used mangosteen with demineralized freeze-dried bovine bone xenograft (DFDBBX).29 The combination group obtained the lowest RANKL expression at 5.17 ± 1.47 compared with 17.50 ± 1.38 for the control group (fold change = 3.38× reduction), and the difference was statistically significant (p = 0.002). Moreover, the highest osteoblast (12.50 ± 1.05) and lowest osteoclast (8.50 ± 1.76) numbers were reported in sockets treated with the composite graft, with a decrease in the NF-κB level and increase in the bone morphogenetic protein-2 level (Table 3). According to the results of two separate studies that evaluated the synergistic effect of Nigella sativa and propolis extracts with bovine bone graft,30,31 the lowest RANKL and highest OPG expression levels (Table 2) were found in the composite graft groups (exact means not provided), with significant differences (p = 0.000) for both (Table 4). The final study within this category assessed the influence of Moringa combined with DFDBBX on the RANKL expression level.35 Significant RANKL suppression was found in the composite graft (p = 0.000) on both days 7 and 30 compared with the control (Table 4). OPG expression was not evaluated. Cell-based composite graft (stem cells + scaffold) Two studies with 24 extraction sockets evaluated the use of stem cells derived from human exfoliated deciduous teeth (SHED) in grafted sites (Table 1). The first study verified the effect of combining SHED with an HA scaffold.32 After 8 weeks, the treatment group had lower RANKL levels (4.80 ± 1.30) and higher OPG levels (11.60 ± 1.14) compared with the HA only group (RANKL = 12.67 ± 2.08, OPG = 6.00 ± 1.00), and the differences were significant (p < 0.001, Table 4). Consequently, the RANKL/OPG ratio was lower than 1 (0.41, Table 2) in the combination group with a reduction of 2.64× and increase of 1.93× in terms of the fold changes for RANKL and OPG, respectively (Table 4). Similar results were reported in another study that evaluated SHED in combination with a carbonate apatite scaffold (CAS).34 On day 7, the RANKL expression level decreased whereas the OPG level increased in the CAS + SHED group (RANKL = 4.80 ± 0.58; OPG = 11.6 ± 1.08), yielding a RANKL/OPG ratio of 0.41, and it was significantly higher than that for the CAS only group (p < 0.01, Table 4). Furthermore, the expression levels of all other osteogenic markers increased (Table 3). Synthetic graft One study assessed 15 post-extracted sites in human subjects grafted with magnesium enriched HA (MgHA) as a study group (control group not present, Table 1).24 Although exact numerical values were not presented, both biomarkers increased after 4 months (Table 2). The RT-qPCR results showed that the RANKL gene level increased (fold not specified) and the OPG gene level increased by 7-fold. At the same time point, immunostaining findings indicated an increase in RANKL protein whereas the OPG protein level decreased 0.6-fold (Table 4). The RANKL/OPG ratio was reported as 1.88 ± 1.24. The statistical results were compared to the baseline period for both biomarker genes, and there was a highly significant difference for RANKL protein (p = 0.001), but no significant difference for OPG protein (p < 0.05, Table 4). Histologically, the grafted sites contained well-structured healing tissue, no inflammatory infiltrate, and incompletely osteointegrated MgHA granules. Immunohistochemical findings indicated significant increases in the protein expression levels of all other biomarkers. Histomorphometric analysis detected 16.50% newly formed bone and 38.86% residual MgHA biomaterial (Table 3). Quality assessment for the included articles According to the ROBINS-I V2 guidance, the overall rating of bias for the study by Canullo et al. (2016)24 was “moderate risk” due to some concerns with the selection of the reported results (Table 5). The risk of bias assessments obtained with the SYRCLE tool for the 11 in vivo experimental studies25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35 included in this review are summarized in Table 6. Detailed justifications/explanations for judgments are presented in Supplementary Table S1.Table 5Risk of bias assessment using ROBINS-I V2 tool for the non-randomized clinical study (Canullo et al., 2016).24Table 5DomainAssessmentJustification1. Bias due to confounding.LowThe study conduction based on consistent baseline characteristics; potential confounder (i.e., smoking) was controlled (< 10/day), and other confounders like oral hygiene or systemic diseases were clearly adjusted.2. Bias in classification of interventions.LowInterventions were clearly defined and consistently applied.3. Bias in selection of participants.LowParticipants were consecutively enrolled with clear inclusion/exclusion criteria.4. Bias due to deviations from intended interventions.LowStandard surgical procedure was described. There is no evidence of differences in co-interventions or post-operative care between groups.5. Bias due to missing data.LowNo missing data; all patients followed up to 4 months.6. Bias arising from measurement of the outcome.LowAlthough the study did not mention any type of blindness, immunohistochemistry and molecular assays (i.e., protein levels, gene expression, fold change) are inherently objective lab-based measures and unlikely to be influenced by assessor.7. Bias in selection of the reported result.ModerateThe study did not mention a pre-published protocol or trial registration. Without protocol access, it is not possible to fully rule out selective reporting of analyses.Table 6Risk of bias assessment using SYRCLE tool for the animal studies included.Table 6Study (Year)Sequence generationBaseline characteristicsAllocation concealmentRandom housingCaregiver blindingRandom outcome assessmentAssessor blindingIncomplete outcome dataSelective outcome reportingOther biasDjais et al. (2020)25NoYesNoNoUnclearUnclearUnclearYesUnclearYesDjais et al. (2023)26UnclearYesUnclearUnclearUnclearUnclearUnclearYesUnclearYesIlham et al. (2025)27UnclearYesUnclearUnclearUnclearUnclearUnclearYesUnclearYesKamadjaja et al. (2021)28NoYesNoNoUnclearUnclearUnclearYesUnclearUnclearKresnoadi et al. (2017)29NoYesNoNoUnclearUnclearUnclearYesUnclearUnclearKresnoadi et al. (2023a)30NoYesNoNoUnclearUnclearUnclearYesUnclearYesKresnoadi et al. (2023b)31NoYesNoNoUnclearUnclearUnclearYesUnclearUnclearPrahasanti et al. (2019)32UnclearYesUnclearUnclearUnclearUnclearUnclearYesUnclearYesRam et al. (2023)33YesYesUnclearUnclearUnclearUnclearUnclearYesUnclearYesSaskianti et al. (2020)34YesYesUnclearUnclearUnclearUnclearUnclearYesUnclearYesSoekobagiono et al. (2017)35UnclearYesUnclearUnclearUnclearUnclearUnclearYesUnclearYes Table 5Risk of bias assessment using ROBINS-I V2 tool for the non-randomized clinical study (Canullo et al., 2016).24Table 5DomainAssessmentJustification1. Bias due to confounding.LowThe study conduction based on consistent baseline characteristics; potential confounder (i.e., smoking) was controlled (< 10/day), and other confounders like oral hygiene or systemic diseases were clearly adjusted.2. Bias in classification of interventions.LowInterventions were clearly defined and consistently applied.3. Bias in selection of participants.LowParticipants were consecutively enrolled with clear inclusion/exclusion criteria.4. Bias due to deviations from intended interventions.LowStandard surgical procedure was described. There is no evidence of differences in co-interventions or post-operative care between groups.5. Bias due to missing data.LowNo missing data; all patients followed up to 4 months.6. Bias arising from measurement of the outcome.LowAlthough the study did not mention any type of blindness, immunohistochemistry and molecular assays (i.e., protein levels, gene expression, fold change) are inherently objective lab-based measures and unlikely to be influenced by assessor.7. Bias in selection of the reported result.ModerateThe study did not mention a pre-published protocol or trial registration. Without protocol access, it is not possible to fully rule out selective reporting of analyses. Risk of bias assessment using ROBINS-I V2 tool for the non-randomized clinical study (Canullo et al., 2016).24 Table 6Risk of bias assessment using SYRCLE tool for the animal studies included.Table 6Study (Year)Sequence generationBaseline characteristicsAllocation concealmentRandom housingCaregiver blindingRandom outcome assessmentAssessor blindingIncomplete outcome dataSelective outcome reportingOther biasDjais et al. (2020)25NoYesNoNoUnclearUnclearUnclearYesUnclearYesDjais et al. (2023)26UnclearYesUnclearUnclearUnclearUnclearUnclearYesUnclearYesIlham et al. (2025)27UnclearYesUnclearUnclearUnclearUnclearUnclearYesUnclearYesKamadjaja et al. (2021)28NoYesNoNoUnclearUnclearUnclearYesUnclearUnclearKresnoadi et al. (2017)29NoYesNoNoUnclearUnclearUnclearYesUnclearUnclearKresnoadi et al. (2023a)30NoYesNoNoUnclearUnclearUnclearYesUnclearYesKresnoadi et al. (2023b)31NoYesNoNoUnclearUnclearUnclearYesUnclearUnclearPrahasanti et al. (2019)32UnclearYesUnclearUnclearUnclearUnclearUnclearYesUnclearYesRam et al. (2023)33YesYesUnclearUnclearUnclearUnclearUnclearYesUnclearYesSaskianti et al. (2020)34YesYesUnclearUnclearUnclearUnclearUnclearYesUnclearYesSoekobagiono et al. (2017)35UnclearYesUnclearUnclearUnclearUnclearUnclearYesUnclearYes Risk of bias assessment using SYRCLE tool for the animal studies included. In terms of the quality of evidence, the final GRADE certainty level judgment for the human clinical study by Canullo et al. (2016)24 was “moderate” due to the moderate risk of bias (absence of a pre-registered protocol), and imprecision associated with a small sample size and lack of calculation power (Supplementary Table S2). Although several domains in the SYRCLE risk of bias assessment were rated as “unclear” (Table 6), this reflected insufficient reporting rather than confirmed methodological shortcomings. Therefore, these domains were not considered to represent a definitive high risk of bias, and the overall risk of bias was judged as not serious, avoiding excessive downgrading. Indirectness related to the use of animal models was acknowledged, but the experimental conditions, interventions, and outcomes were considered sufficiently relevant to the clinical question to warrant only limited concern. In addition, although protocol registration was not reported, publication bias was not strongly suspected given the consistency of reporting outcomes across studies. Consequently, the overall certainty of evidence was judged as “moderate.” Discussion Several systematic literature reviews have assessed outcomes regarding hard and soft tissue healing (i.e., alveolar site dimensions, histology, implant treatment, patient-based findings, etc.) following post-extraction socket preservation technique with different grafts either alone or in combination with other enhancement biomaterials.36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46 However, systematic evaluations have not been conducted previously of the effects of various graft types on the expression levels of RANKL and OPG markers in different studies. Although closely related, bone regeneration, osteogenesis, and bone remodeling are distinct biological processes. Bone regeneration refers to the restoration of bone defects through new tissue formation, whereas osteogenesis specifically describes the formation of new bone by osteoblasts. By contrast, bone remodeling is a continuous process involving bone resorption and formation to maintain the skeletal structure and function. These processes collectively contribute to bone healing and adaptation but should not be considered interchangeable terms. The abundant donor supply, reduced risk of human disease transmission, and lower preparation/production costs of xenografts compared with allografts and autografts have made them promising alternatives to conventional bone grafts.47 According to the current systematic review, the utilization of xenografts as socket preservation biomaterials resulted in differences in the capacity to modulate RANKL and/or OPG expression levels among graft types in animal models. In a study that used marine-derived xenograft (CaCO3) from pearl oysters (Ilham et al., 2025),27 RANKL expression was significantly reduced in the treatment group, aligning with favorable remodeling responses. These outcomes are consistent with studies that demonstrated the potential of using pearl oyster shells as alternative materials to accelerate the bone regeneration process.48, 49, 50 Kamadjaja et al. (2021)28 demonstrated the effective role of crab shell-derived HA as a graft biomaterial for significantly decreasing RANKL expression and increasing OPG expression in preserved dental sockets (Table 2, Table 4). Conversely, high RANKL and low OPG levels were reported in the graft-free group, suggesting continued osteoclastic activity. These findings are in agreement with the results obtained in two recent studies,51,52 which showed that the biocompatibility and osteoconductive properties of an HA scaffold could stimulate stem cells and osteoblasts to differentiate and proliferate, inducing new bone formation. Therefore, these xenografts (CaCO3 and HA) acted as both scaffolds and stimulants for osteogenic signaling. However, dissimilar results were obtained in another study by Ram et al. (2023)33 who used a bovine-derived xenograft (Bio-Oss®) in a ridge preservation technique, where comparable findings were obtained in both groups (i.e., expression levels of RANKL and OPG genes as well as other key remodeling biomarkers increased) over all time points (Table 2, Table 3), suggesting similar rates of osteoclastic activity. The gene expression changes in RANKL/OPG were not statistically significant (Table 4), but histological analysis indicated enhanced woven bone formation with more apparent lamellar bony bridges in grafted sockets (Table 3). Thus, Bio-Oss® positively enhanced structural healing, although the molecular signals were similar, highlighting the importance of correlating biomolecular markers with histological findings. These differences in outcomes in the latter study may be attributed to interspecies variation because sheep were enrolled as animal subjects, whereas other studies utilized guinea pigs. Interestingly, this was clearly supported by a review study that highlighted the significantly different histological structures among species (rabbits, rodents, pigs, and sheep/goats).53 These variations may affect bone remodeling and healing modulation in grafted sites. Six studies that utilized composite grafts were analyzed in this systematic review (Table 1). Djais et al. (2020)25 used a biologically enhanced autograft (Moringa + DFDDM) and observed the greatest RANKL suppression and highest OPG expression on day 21, suggesting a strong regenerative profile, which was supported by immunohistochemical data (RANKL/OPG ratio = 0.11, Table 2). Similar outcomes were obtained by Soekobagiono et al. (2017)35 who reported that a combination of Moringa plus DFDBBX led to consistent suppression of RANKL in both the early and late healing phases (days 7 and 30, respectively), implying enhanced bone formation pattern as the composite graft strongly suppressed osteoclastogenesis. This could have been due to the antioxidant, anti-inflammatory and anti-resorptive effects of Moringa oleifera leaf flavonoids and other beneficial compounds.54, 55, 56, 57 Djais et al. (2023)26 evaluated a chitosan + bovine bone composite graft. Compared with the placebo, sustained OPG elevation and RANKL reduction, with a RANKL/OPG ratio less than 1 were found in the treatment group, suggesting an appropriate bone healing process (Table 2). These outcomes were in accordance with other pre-clinical studies that showed the potential of chitosan for significantly downregulating RANKL signaling, upregulating OPG, improving the RANKL/OPG ratio and differentiation of osteoblasts, and inhibiting alveolar bone resorption.58, 59, 60, 61 The synergistic effects of combining different biological extracts such as mangosteen, Nigella sativa, and propolis with xenografts are described in this systematic review (Table 4). Kresnoadi et al. (2017)29 found a 3.38× reduction of RANKL in a mangosteen peel extract (MPE) and DFDBBX composite, with improved histological and immunohistochemical findings (Table 3). Similarly, Kresnoadi et al. (2023a)30 reported that a combination of Nigella sativa and bovine bone graft (BBG) yielded the lowest RANKL and highest OPG expression levels compared with other treatments (Table 2). Furthermore, propolis combined with BBG (Kresnoadi et al., 2023b)31 suppressed the RANKL level and increased the OPG level. The anti-inflammatory and bone formation effects of mangosteen and propolis extracts were evaluated by Lim et al. (2020).62 Their results were in agreement with the outcomes described in the current review, where both materials enhanced osteoblastic differentiation markers with mineralized matrix deposition, and significantly reduced the expression levels of pro-inflammatory cytokines. Further support was also provided by Thummuri et al. (2015)63 who investigated the effects of Nigella sativa on osteoclastogenesis. Their results clearly showed that this biological extract inhibited RANKL-induced osteoclast formation as well as preventing bone loss and preserving the bone architecture in vivo. Histologically, Kresnoadi et al. (2017)29 detected the highest osteoblast and lowest osteoclast numbers (12.50 ± 1.05 and 8.50 ± 1.76, respectively; Table 3) in sockets preserved with a composite graft (MPE + DFDBBX). These outcomes are consistent with the results reported by Azhar et al. (2017)64 who studied the effects and potency of using a combination of MPE with DFDBBX on osteoclasts, osteoblasts, and interleukin-1β (IL-1β) expression during the preservation of post-extracted sockets. MPE plus DFDBBX significantly increased the osteoblast numbers but decreased both osteoclasts and the IL-1β marker levels in the alveolar bones of animals. Osteoconductive grafts can support mesenchymal stem cells to facilitate optimal proliferation and differentiation. As described in the present systematic review, Prahasanti et al. (2019)32 and Saskianti et al. (2020)34 incorporated SHED stem cells in synthetic scaffolds (HA and carbonate HA as a type of CAS, respectively). Their outcomes indicated substantially improved bone regeneration accompanied by significant OPG elevation and RANKL suppression, confirming the immunomodulatory and osteoinductive potential of SHED cells. The RANKL/OPG ratio was 0.41 (Table 2) and the immunohistochemical results were correlated with significant fold changes (Table 4), supporting a dominant trend toward new bone formation. These important findings agree with a recent systematic review by Al-Shaibani et al. (2025)65 who evaluated the outcomes of using stem cell-seeded scaffolds (including SHED cells) for regenerating jaw and alveolar bone defects. They concluded that integrating human dental pulp stem cells or SHED cells with scaffolds consistently enhanced the bone density, volume, angiogenesis, and osteogenesis in bone defects compared with scaffold-only controls. Moreover, the porous construction of HA scaffolds closely resembles the natural structure of bone, facilitating the migration of cells, nutritional flow, and improved vascularization.66, 67, 68, 69 Canullo et al. (2016)24 conducted a clinical study to preserve human dental sockets by using a synthetic MgHA graft. The results indicated an increase in the RANKL/OPG ratio (1.88) and signs of ongoing remodeling, possibly driven by the active expression of catabolic markers (RANK, IL-6, and tumor necrosis factor-α; Table 3). Despite a 7-fold increase in the OPG gene, the protein expression levels remained low (0.6×, Table 4), indicating a lag between gene activation and protein translation. Histological analysis detected the presence of a relatively high residual graft material after 4 months (38.86%, Table 3), suggesting delayed or extended remodeling rather than complete integration. These findings are in accordance with those obtained by Crespi et al. (2011)70 who reported comparable results (32.2%) for a residual MgHA graft with the same follow-up period after ridge preservation. The gene expression profiles data are inconsistent with those presented by Crespi et al. (2009)71 who demonstrated a marked decrease in RANKL expression, comparable levels of OPG, and a decreased RANKL/OPG ratio in maxillary sinuses augmented with MgHA compared with autogenous graft sites. These results may be explained mainly by the variance in the grafted defect (i.e., maxillary sinus vs dental socket) and partly by the slightly extended follow-up period (5 months). Using a systematic approach, the present review provides a comprehensive analysis of the regulatory roles of RANK and OPG biomarkers in the context of bone remodeling regarding socket preservation techniques. However, several limitations should be considered. The applied language restriction (i.e., English only) may have led to the exclusion of relevant studies. One author did not respond regarding verification of unclear data.24 In terms of the quality of the studies included, the overall certainty of evidence was rated as “moderate,” indicating that the findings are reasonably reliable but may be influenced by future well-designed studies. In addition, most of the records (11 out of 12) were preclinical models, thereby restricting direct application to humans. Conclusions Despite the study limitations, composite grafts may have potential benefits for socket preservation. These approaches appear to influence bone remodeling processes, probably through modulating the RANKL/OPG axis. However, the current evidence is insufficient to definitively establish their superiority over other grafting approaches. Similarly, the reliability of the RANKL/OPG ratio as a predictive biomarker of bone regenerative potential and resorptive activity in post-extraction sites remains uncertain. Therefore, the results should be interpreted with caution and further high-quality clinical studies are required to clarify the clinical relevance of these molecular biomarkers. Registration and protocol The specified inclusion and exclusion criteria, and the analytical methods were carefully outlined and recorded within a registered protocol (PROSPERO ID: CRD420251046255). During the protocol refinement and literature scoping phase, a minor amendment was made to the registered review title. The original title focused solely on autogenous dentin grafts in the context of socket preservation. The revised title broadens the scope of the review to include different graft types, rather than being limited to autogenous dentin grafts, thereby providing a more comprehensive and clinically relevant synthesis of the available evidence. Importantly, this amendment did not affect this systematic review's core objective, rationale, or methodology, but instead it facilitated a wider and more representative analysis. In addition, the amendment of the title did not involve any modifications to the predefined eligibility criteria, study selection procedures, outcomes of interest, or analytical methods. Consequently, the review scope and methodological framework remained unchanged, minimizing the potential for reporting bias. Ethical approval Not applicable. Authors contributions FAJ: concept/design, data collection, data analysis/interpretation, drafting article, critical revision of article, approval of article. SSA: data collection, data analysis/interpretation, critical revision of article, approval of article. All authors have critically reviewed and approved the final draft and are responsible for the content and similarity index of the manuscript. Availability of data All original data used in this study are available within the article and supplementary materials. Any additional information can be requested from the corresponding author. Declaration of generative AI and AI-assisted technologies in the writing process During the preparation of this work, the authors used ChatGPT (OpenAI) for identifying specific background knowledge and Copilot (Microsoft) for paraphrasing certain sentences within this manuscript to enhance the scientific validity and linguistic clarity, but without any contribution to the generation of novel concepts, synthesis of data, interpretation of findings, or formulation of the final conclusions. After using these tools, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article. Source of funding This study did not receive funding from any organization. Conflict of interest All authors (Firas A. Jamil and Saif S. Abdulrazaq) state that they have no conflicts of interest to declare.