Inhibition of sclerostin activity promotes bone regeneration in experimental periodontal disease: A systematic review and meta‑analysis of animal models

Introduction

Periodontal bone regeneration is a major clinical challenge. Current surgical strategies, such as autologous bone grafting combined with guided tissue regeneration, achieve long-term success rates of ~58% in restoring functional attachment (1-4). These approaches are further constrained by donor site morbidity, unpredictable graft resorption and diminished efficacy in patients with systemic conditions such as diabetes mellitus (5-8). Collectively, these limitations underscore the need for biological therapies that directly modulate bone metabolism.

Sclerostin (Scl; encoded by the SOST gene) is a 22-kDa secreted glycoprotein that has emerged as a key regulator of skeletal homeostasis, particularly in bone remodelling and craniofacial biology. It serves as a key mediator of the interaction between osteoblasts and osteoclasts, maintaining the balance between bone formation and resorption (8). Mechanistically, Scl exerts its effects through a number of interrelated pathways (9-12). By antagonizing the canonical Wnt/β-catenin signaling cascade, Scl binds to low-density lipoprotein receptor-related proteins (LRPs) including LRP5/6 on osteoblast precursors, thereby suppressing osteoblast proliferation, differentiation and matrix mineralization. This inhibition decreases the capacity for new bone formation (13-15). Scl indirectly promotes osteoclastogenesis through bone morphogenetic protein (BMP)/Wnt signaling interactions. It increases the receptor activator of NF-κB ligand (RANKL) to osteoprotegerin (OPG) ratio by stimulating RANKL secretion from osteocytes while decreasing OPG expression, favoring osteoclast differentiation and activity. Thus, Scl not only limits bone formation but also actively enhances bone resorption (16-19).

Beyond local effects in bone, Scl may also act systemically. Evidence suggests (16,17) that it can influence distant organs such as the kidney or vascular system, linking bone metabolism to broader physiological processes. In periodontal disease, these regulatory networks become dysregulated. The alveolar bone undergoes rapid turnover and is subject to continuous mechanical loading from mastication and occlusal forces (16-18). When combined with chronic biofilm-induced inflammation, these stressors amplify disruption of the Scl-Wnt-RANKL axis, accelerating alveolar bone resorption and undermining periodontal tissue stability (19).

Given its dual role in suppressing osteoblast activity and stimulating osteoclastogenesis, therapeutic inhibition of Scl has attracted extensive research (14-19), particularly following the clinical success of romosozumab in osteoporosis. However, whether these benefits extend to biofilm-driven periodontitis, remains uncertain. The regulatory pathways by which Scl influences alveolar bone remodeling, including inhibition of canonical Wnt/β-catenin signaling, modulation of the RANKL/OPG pathway, interactions with BMP signaling and amplification through inflammatory cytokines are illustrated in Fig. S1.

Research into therapeutic inhibition of Scl has progressed since the approval of romosozumab for osteoporosis and preclinical studies have confirmed its anabolic effects in long-bone defect models (20-22). Whether these effects extend to polymicrobial biofilm-induced periodontitis remains unclear, given the unique challenges of the periodontal environment, including alveolar bone turnover, estimated to be 200-300% faster compared with that of tibial bone (23-27), constant bidirectional loading from mastication (28,29) and chronic biofilm-driven inflammation (30,31).

To the best of our knowledge, the present meta-analysis is the first to synthesize preclinical evidence on Scl inhibition in experimental periodontitis. Existing studies (14-19) demonstrate limited insight into how treatment effects vary by dose, anatomical site or systemic conditions. By integrating available data, the present analysis aimed to outline the dose-response association between SOST suppression and alveolar regeneration, as well as potential differences across defect types such as interproximal and furcation lesions and interactions with systemic bone-modifying factors. Together, these findings may provide a more comprehensive understanding of the therapeutic potential of Scl inhibition in periodontal regeneration.

Materials and methods

Protocol and registration

The present protocol was registered on PROSPERO (https://www.crd.york.ac.uk/PROSPERO; registration ID: CRD42023388345). The present review aimed to evaluate whether inhibition of Scl activity promotes alveolar bone repair and regeneration in animal models of periodontal disease, based on the recommendations from the Cochrane Handbook for Systematic Reviews (32) and PRISMA statement for reporting systematic reviews and meta-analysis (33).

Inclusion criteria

A population, intervention, comparison, outcomes and study design framework was applied to define the criteria. Inclusion criteria were as follows: i) Animal models of experimental periodontal disease; ii) evaluated the inhibition or neutralization of Scl activity; iii) used animal models with periodontal disease but without Scl inhibition as the control group; iv) encompassed primary outcomes associated with decreased alveolar bone loss (ABL) and secondary outcomes including changes in bone volume fraction (BVF), bone mineral density (BMD), bone-specific serum markers and the number of SOST-positive cells; and v) in vivo randomized controlled trials.

Exclusion criteria

Studies that did not replicate periodontal pathology, in vitro experiments and studies lacking a control group or using non-periodontitis animals as controls were excluded.

Search strategy

Studies were identified through a systematic s earch of the following electronic databases: PubMed (https://pubmed.ncbi.nlm.nih.gov/), the Cochrane Library (https://www.cochranelibrary.com/), Embase (https://www.embase.com/) and Web of Science (all databases, https://www.webofscience.com/). Studies published from inception to January 2025 were included, with no language restrictions. The search combined medical subject headings with free-text terms applied to titles, keywords and abstracts. In addition, a manual search was performed by screening the reference lists of articles and related reviews for additional relevant studies. The complete search strategy is provided in Table SI.

Study selection

All retrieved records were imported into EndNote software (version X9; Clarivate Plc) to remove duplicates. A total of two reviewers then independently screened the remaining articles against the inclusion and exclusion criteria. First, titles and abstracts were reviewed to exclude irrelevant studies. Full text of potentially eligible studies was assessed to confirm their relevance. Any discrepancies were resolved by discussion and consensus or consultation with a third reviewer.

Data extraction

Study publication date, author name, sample size, animal species, the sex and diet of the animals, construction of the periodontal disease model, the adopted interventions, route and dose of administration, the duration of administration, outcomes and types of detection (e.g., histological, radiographic or molecular analyses) were collected.

If data were insufficient or presented only graphically, attempts were made to contact the authors and obtain the numerical values. If the information could not be retrieved, WebPlotDigitizer (V5), a digital ruler software, was used to measure graphical data (automeris.io/WebPlotDigitizer/) (34). If studies reported outcomes at multiple time points, only data from the most commonly reported time point across studies (e.g., 3 or 6 weeks) were acquired for meta-analysis. If the experiments included a number of groups, only periodontitis animal models were taken as the experimental and control group for meta-analysis. The data were extracted independently by two authors, with any disagreements being resolved through discussion and consensus.

Quality assessment

Risk of bias was independently evaluated by two reviewers using the Systematic Review Centre for Laboratory Animal Experimentation (SYRCLE) risk of bias tool (35), which consists of ten domains adapted from the Cochrane Collaboration tool (36). According to the criteria recommended by Hooijmans et al (35), each domain was rated as low risk (score=1) if sufficient methodological details were reported, high risk (score=-1) if clear methodological shortcomings were present or unclear risk (score=0) when information was insufficient to permit judgement. The assessed domains included sequence generation, baseline characteristics, allocation concealment, random housing, blinding of caregivers/experimenters, random outcome assessment, blinding of outcome assessors, incomplete outcome data, selective outcome reporting and other potential sources of bias. Discrepancies between reviewers were resolved by discussion or consultation with a third reviewer.

Data synthesis and statistical analysis

As only morphometric results, such as the distance between the cementoenamel junction and alveolar bone crest (CEJ-ABC), tissue mineral density (TMD) and BVF, yielded sufficient data, these continuous variables were extracted for meta-analysis. Analysis was conducted using REVMAN (version 5.3; The Cochrane Collaboration) software to determine pooled effects. To identify and measure heterogeneity in the results, χ2 tests and I2 statistics were implemented. χ2 with a significance level of α=0.1 was considered to indicate heterogeneity.

As all outcome measures were continuous variables, the standardized mean difference (SMD) with 95% CI was used to describe the efficacy of the intervention effect. SMD was used instead of weighted mean difference because the studies included reported heterogeneous outcomes (BMD, BVF and ABL) in different units and scales. The use of SMD allowed for standardization and ensured comparability across studies. In accordance with the Cochrane Handbook for Systematic Reviews of Interventions, a random-effects model was applied for all meta-analyses, irrespective of the I2 value, since heterogeneity is expected across experimental studies.

Subgroup analyses were conducted to explore potential sources of heterogeneity, such as type of intervention (direct vs. indirect inhibition), species, sex, dosage, route of administration, model construction and whether the ligature was removed (most models were ligature-induced). To minimize potential bias from individual studies, sensitivity analysis was conducted by sequentially excluding each trial to evaluate its influence on the pooled results.

Level of evidence

A total of two investigators independently assessed the certainty (level) of evidence for each outcome using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) system (GRADEPRO.org), based primarily on five domains, including type of study design, risk of bias, inconsistency, indirectness, imprecision and other comprehensive consideration of sample size and confounding factors. Any disagreements were resolved through discussion or by a third researcher.

Results

Search outcomes

The literature search yielded a total of 99 studies (Fig. S2). After removing the duplicates and screening the remaining publications, 39 articles were eligible for full-text evaluation. Of these, 31 were excluded. A total of eight studies was retrieved for systematic review and assessed by meta-analysis (37-44)

Study characteristics

The present review included articles published from 2013-2023. Although there were no restrictions on the type of animal or publication language, all periodontal disease models identified were rat or mouse studies published in English. The number of animals ranged from 14-40, with a total of 119 males and 64 females [sex not reported in one study (41)]. Models included Sprague-Dawley rats (37,38,42), Wistar albino rats (44), inbred F344 rats (39), OPG-deficient mice (40) and periostin-knockout mice (41). Diet primarily comprised standard chow and tap water, except for the high-sugar water provided to all groups by Chen et al (37) and the 10% sucrose given to the experimental group by Yang et al (43).

For interventions targeting Scl activity, Scl antibody (Scl-Ab) was used in five studies (37,38,41,42,44), SOST gene knockout was performed in two studies (41,43) and specific pharmacological agents were tested in three studies, namely caffeic acid phenethyl ester (CAPE) (44), low-dose doxycycline (LDD) (44), infliximab (39) and WP9QY (40).

In addition to ligature-induced periodontitis, five studies also applied additional conditions unfavourable for bone regeneration, such as ovariectomy (37,38), periostin knockout (41), OPG deficiency (40) and streptozotocin-induced type I diabetes (37-41,43). A total of one study used silk ligatures saturated with Porphyromonas gingivalis (43).

Systemic subcutaneous injection of Scl-Ab was employed in three studies (37,40,42), while intraperitoneal injection was used in two studies (38,41). The dosage and duration included 25 mg/kg twice weekly for 2-8 weeks in four studies (37,41,42,44) and 5 mg/kg weekly for 22 weeks in one study (38). Local Scl-Ab application (125 µg) was tested in one study (42) but proved less effective compared with systemic administration. In addition, Yiğit et al (44) applied CAPE and LDD at 10 mg/kg for 14 days, delivered by intraperitoneal injection and oral gavage, respectively. CAPE was more effective compared with LDD at decreasing Scl expression. In the other pharmacological studies, infliximab was administered intraperitoneally at 5 mg/kg once or twice before euthanasia (39) and WP9QY was given subcutaneously at 10 mg/kg three times daily for 5 days (40).

The primary outcome assessment methods included micro-computed tomography (CT) and histological examination to evaluate ABL, BVF and BMD. Immunohistochemistry was used to assess the number of SOST-positive cells and osteoclasts. In addition, serum analysis, ELISA and reverse transcription-quantitative PCR were performed to measure serum biomarkers, such as procollagen type I N-terminal propeptide (P1NP), osteocalcin (OCN), tartrate-resistant acid phosphatase 5b (TRAP5b), RANKL, C-terminal telopeptide of type I collagen (CTx-1) and ALP. The characteristics of the included studies are summarized in Table I.

Table I Characteristics of included studies.

Table I

Characteristics of included studies.

	First author, year
Characteristic	Taut et al, 2013	Ren et al, 2015	Chen et al, 2015	Hadaya et al, 2019	Yang et al, 2016	Yiğit et al, 2022	Kim et al, 2016	Ozaki et al, 2017
Sample size	16	16	24	40	20	20	16	14
Animal	SD rat	Periostin knockout mouse	SD rat	SD rat	C57BL/6 mouse	Wistar albino rat	Inbred F344 rat	OPG-/- C57BL/6 mouse
Sex	Male	N/A	Female	Female	Male	Male	Male	Male
Diet	Regular rodent chow diet	N/A	Food and high-sugar drinking water	Standard diet NIH-31 modified	Standard solid mouse chow, 10% sucrose for EX group and sterile water for CON group	Standard rat food and tap water ad libitum	Standard rat chow and water ad libitum	Sterilized water and diet ad libitum
Construction of periodontal disease model	Ligature	Periostin knockout	Ligature; OVX	Ligature; OVX	Ligature of silk saturated with P.g	Ligature	Ligature; streptozoto- cin	OPG deficiency
Intervention, EX/CON	Scl-Ab/PBS	Scl-Ab, SOST and periostin double knockout/ PBS	Scl-Ab/ empty vehicle	Scl-Ab/empty vehicle	SOST gene knockout/none	CAPE; LDD/ PBS	Infliximab/ PBS	WP9QY/PBS
Route of administration	s.c.	i.p.	s.c.	i.p.	N/A	CAPE, i.p.; LDD, oral gavage	i.p.	s.c.
Dose	25 mg/kg	25 mg/kg	25 mg/kg	5 mg/kg Scl-ab	N/A	CAPE, 10 µmol/ kg/day; LDD, 10 mg/kg/day	5 mg/kg	10 mg/kg
Duration	Twice/week for 3 or 6 weeks	Twice/week for 8 weeks	Twice/week for 6 weeks	Weekly for 12 weeks before OVX and 10 weeks after OVX	N/A (constructive knockout); 2 months	Daily for 14 days	Once for the 3 day group (on day 0); twice for the 20 day group (on days 7 and 14)	3 times/day for 5 days
Secondary outcomes	BVF, TMD. OCN, P1NP, TRAP5b	Bone deposition rate, mor- phological changes of osteocytes, BV/TV	TRAP5b, CTx-1, BVF, BMD	P1NP, TRAP5b	OPG, RANKL, JNK, p38MAPK, ERK1/2	PNMLS, BMP2 and Scl-positive expression	IL-1β, Scl, cathepsin K-, RANKL- and Scl-positive cells	BV/TV; number of TRAP-5b-, sterix- orScl- positive cells; serum ALP
Detection technique	Bone-specific serum biomar- kers; micro-CT; histology immunohisto- chemistry; fluorescent calcian labeling	Electron microscopy; micro-CT; histology/ immunohis- tochemistry; FITC staining and Imaris analysis	Bone- specific serum biomarkers; micro-CT; histology	Serum analysis; micro-CT; histology	Immunohisto- chemistry; micro- CT	Histomorpho- metry, histopa- thology and immunohisto- chemistry	Histology, histomorpho- logy and immunohisto chemistry; RT-PCR	Micro-CT; histomorpho- logy and immunohisto- chemistry; ELISA (serum marker)
(Refs.)	(42)	(41)	(37)	(38)	(43)	(44)	(39)	(40)

[i] Primary outcome was alveolar bone loss in all studies. SD, Sprague-Dawley; OPG, osteoproteprin; BV, bone volume; TV, total volume; TMD, tissue mineral density; PNMLS, polymorphonuclear leukocytes; RANKL, receptor activator of NF-κB ligand; i.p., intraperitoneal; s.c., subcutaneous injections; OVX, ovariectomy; TRAP5b, tartrate-resistant acid phosphatase; OCN, osteocalcin; P1NP, procollagen type I N-terminal propeptides; CAPE, caffeic acid phenethyl ester; LDD, low-dose doxycycline; SOST, sclerostin; Scl-Ab, sclerostin antibody; EX, experimental; CON, control; BMP2, bone morphogenetic protein 2; BVF, bone volume fraction; BMD, bone mineral density; CTx-1, C-terminal telopeptide of type I collagen; ALP, alkaline phosphatase; RT, reverse transcription; FITC, fluorescein isothiocyanate; N/A, not applicable; NIH-31, standard laboratory rodent diet; P.g, Porphyromonas gingivalis; CT, computed tomography.

Quantitative synthesis and data analysis

The outcome variables with sufficient data, including changes in the CEJ-ABC distance, ABL, BMD, BVF, number of Scl-positive cells and bone-specific serum biomarkers such as OCN, P1NP and TRAP5b, were pooled for further analysis (Table SII).

Primary outcomes. ABL

ABL was evaluated by measuring the linear distance or volume between the CEJ and ABC before and after treatment. This parameter directly reflects periodontal tissue changes as assessed by histological or micro-CT examination. All studies (37-44) reported quantitative changes in ABL, therefore this was used as the primary outcome variable of the present meta-analysis. Extracted data were expressed as linear distance, surface area or volume.

Pooled analysis revealed a significant decrease in the CEJ-ABC distance in the Scl inhibition group compared with the control group (Fig. 1; SMD: -2.68; 95% CI: -3.76 to -1.59; I2=85%; P<0.0001). In a study by Yiğit et al (44), two drugs (LDD and CAPE) with Scl-inhibiting potential were tested and results were analysed separately. While numerous ABL data were derived from micro-CT measurements, two studies employed histological assessment (39,44), resulting in different measurement units (linear distance, area or volume).

Figure 1 Forest plot of alveolar bone loss. Scl, sclerostin; df, degrees of freedom; LDD, low-dose doxycycline; CAPE, caffeic acid phenethyl ester; IV, inverse variance model.

Subgroup analysis was performed according to species, sex, dose, route of administration, type of Scl inhibition and periodontitis model. With the exception of dose and route of administration, no significant subgroup differences were observed (Fig. 2). Low-dose (5 mg/kg; Z=7.70) and subcutaneous (Z=5.26) treatments yielded improved outcomes compared with high-dose (25 mg/kg; Z=3.00; P=0.006; Fig. 2B) and intraperitoneal (Z=4.06; P=0.04) treatments (Fig. 2E).

Figure 2 Subgroup analysis of alveolar bone loss. (A) Direct versus indirect inhibition. (B) Dose. (C) Species. (D) Sex. (E) Route of administration. (F) Induction method (ligature and gene-knockout). ABL, alveolar bone loss; Scl, sclerostin; df, degrees of freedom; LDD, low-dose doxycycline; CAPE, caffeic acid phenethyl ester; SMD, standardized mean difference; IV, inverse variance model.

A number of studies used a therapeutic design initiated after ligature removal (37,41-43). Taut et al (42) also incorporated a preventive arm, but no quantitative data were reported and this subgroup could not be included in the present meta-analysis. By contrast, Hadaya et al (38), in which ligatures were not removed and chronic inflammation persisted during treatment, exhibited a larger effect size compared with ligature-removed models (Fig. S3). These findings suggested that sustained inflammation may modify the response to Scl inhibition and decrease its net regenerative benefit.

Sensitivity analysis was conducted using a leave-one-out approach. After sequentially removing each included study, the pooled effect size and its 95% CI did not notably change (Fig. S4). This indicated that no single study had a decisive influence on the overall results, supporting the validity of the present meta-analysis.

Secondary outcomes. BVF

BVF is a volumetric parameter calculated as the ratio of BV to tissue volume based on micro-CT scanning. A total of three studies (37,40,42) reported the BVF and the newly formed bone area was significantly larger in the experimental groups compared with the control groups (Fig. 3A; SMD: 2.70; 95% CI: 0.69-4.71; I2=82%; P=0.009). In a study by Taut et al (42), both systemic and local administration were tested, but only systemic data were extracted for meta-analysis to ensure consistency.

Figure 3 Forest plots of BVF and TMD. (A) Bone volume fraction. (B) Tissue mineral density. Scl, sclerostin; df, degrees of freedom; SMD, standardized mean difference; IV, inverse variance model; BVF, bone volume fraction; TMD, tissue mineral density.

BMD/TMD. A total of two studies (37,42) reported BMD/TMD and revealed that compared with control treatment, systemic Scl-Ab treatment significantly increased BMD (Fig. 3B; SMD: 2.51; 95% CI: 1.58-3.44; I2=0%; P<0.00001).

Other bone-specific serum biomarkers. Bone-associated serum markers (P1NP, OCN, OPG, RANKL, TRAP5b, CTx-1 and ALP) were evaluated in four studies (37,38,42,43). Except for OCN in two studies (37,42), no significant pooled results were obtained (Fig. 4A). Hadaya et al (38) reported that P1NP levels were notably higher in the Scl-Ab group at 6 and 12 weeks post-ovariectomy but decreased below baseline at the end of treatment (38). Taut et al (42) reported that OCN levels were notably elevated in the Scl-Ab group at 3 and 6 weeks. P1NP was also notably higher at 3 but not at 6 weeks (42). Chen et al (37) observed marked increases in OCN and OPG in the experimental group after 6 weeks, accompanied by decreases in TRAP5b and CTx-1(37). Ozaki et al (40) reported that WP9QY did not reduce ALP levels in OPG-deficient mice but increased osteoblast differentiation in the M1 interradicular septum.

Figure 4 Forest plots of bone-specific biomarkers and SOST-positive cells. (A) Bone-specific serum biomarkers (P1NP, OCN and TRAP5b). (B) Histological test for TRAP5b. (C) SOST-positive cells. Scl, sclerostin; df, degrees of freedom; LDD, low-dose doxycycline; CAPE, caffeic acid phenethyl ester; SMD, standardized mean difference; TRAP5b, tartrate-resistant acid phosphatase 5b; SOST, sclerostin; IV, Inverse Variance Model; Pp1NP, procollagen type I N-terminal propeptide; OCN, osteocalcin; TRAP5b, tartrate-resistant acid phosphatase 5b. Supplementary figure legends

For serum TRAP5b, no notable differences were observed between the groups. The detection methods varied: Two studies (37,42) assessed TRAP5b expression through Scl-associated assays, while two others (38,40) used histological staining to quantify the number of osteoclasts/unit area (Fig. 4C). Although Taut et al (42) concluded that Scl inhibition had no effect on TRAP5b expression, the remaining studies indicated that Scl-Ab therapy decreased TRAP5b expression, which was consistent with increases in the expression of other osteogenic markers.

Number or ratio of SOST-positive cells. A total of three studies (39,40,44) using specific drugs assessed SOST-positive cells, consistently showing that the number of Scl-positive cells increased significantly in untreated periodontitis models but decreased markedly after treatment (Fig. 4B; SMD: -6.06, 95% CI: -10.29 to -1.84; I2=93%; P=0.005).

Risk of bias. The risk of bias across ten domains was assessed using SYRCLE (Table SIII). Overall, the methodological quality of the included studies was moderate, with a number of domains judged as unclear or high risk. Adequate sequence generation was reported in 1/8 studies, while baseline characteristics were comparable in 7/8 (37-41,43,44). Allocation concealment was not described in any report and therefore was rated as high risk across all the studies. Random housing and outcome assessments were not reported in any study, resulting in universally unclear ratings. A total of three studies (37,38,44) described the blinding of experimenters and the blinding of outcome assessors. By contrast, incomplete outcome data were adequately addressed in all studies and selective outcome reporting was judged as low risk in 7/8 (37-41,43,44). Other sources of bias were consistently rated as low risk. Collectively, these findings indicated that although the internal validity of the included studies was limited by insufficient reporting of randomization, housing and blinding, the risk of attrition and reporting bias was generally low.

Level of evidence. Level of evidence for SOST-positive cells, ABL and related subgroups was determined to be moderate. However, the levels of evidence for TMD, BVF, OCN, P1NP and TRAP5b were lower (Table SIV). The decrease in the level of evidence was primarily due to inconsistency and imprecision or less overlap in CI, small P-values for the heterogeneity test, large I2 value and failure to perform sensitivity analyses. None of the included studies reported osteonecrosis or other adverse events associated with sclerostin inhibition

Discussion

The present analysis highlighted the potential of Scl inhibition in mitigating ABL and promoting osteogenesis in experimental periodontitis models (37,38,41-43). The mechanisms underlying these effects are complex but may involve key pathways including Wnt/β-catenin signaling, which regulates osteoblast differentiation and bone formation (13-15,41,43). In addition, Scl inhibition may act via the RANKL/OPG pathway, thereby balancing osteoclast activity and limiting excessive bone resorption (16,37,40,42). The decrease in the number of SOST-positive cells further suggested that Scl inhibition effectively targets its intended biological pathway at the histological level (39,41,44).

Scl inhibition not only attenuated alveolar bone loss (ABL) but also promoted regeneration through three primary mechanisms. First, activation of the canonical Wnt pathway restores β-catenin translocation to the nucleus, rescuing osteogenic differentiation in periodontal ligament stem cells (13-15,41,43,45). Second, modulation of the RANKL/OPG pathway limits osteoclastogenesis by suppressing macrophage colony-stimulating factor secretion from platelet-derived growth factor receptor-β (PDGFRβ)-positive osteoprogenitors (16,37,40,42,46,47). Third, disruption of the TNF-α/NF-κB/SOST feedback loop reduces inflammation-induced osteocyte apoptosis and RANKL production, thereby supporting bone preservation and regeneration (39,48,49). Together, these multi-pathway interactions highlight the key role of Scl in periodontitis, namely bone microenvironment crosstalk.

Low-dose regimens (5 mg/kg) (38) showed greater efficacy compared with higher doses (25 mg/kg), particularly in maintaining bone formation and decreasing ABL (37,41,42,44). This paradox suggests that treatment duration and timing may be more influential compared with dose alone. In periodontal models, longer treatment (22 vs. 4-6 weeks) promoted cumulative Wnt activation, underscoring the importance of optimizing both dose and duration for clinical translation (38). Administration route also influenced outcomes, with subcutaneous delivery (37,40,42) outperforming intraperitoneal injection, potentially due to decreased hepatic first-pass metabolism (38,41,45). However, the current evidence is limited and the superiority of lower-dose regimens should be interpreted cautiously. More data are needed to confirm the optimal dose-time association and future studies should systematically examine dosing schedules and treatment durations to validate these observations.

The present study primarily reflects therapeutic models, as preventive evidence was scarce. A number of studies applied therapeutic interventions following ligature removal (37,41-43). Taut et al (42) tested both preventive and therapeutic arms, but only the therapeutic arm reported quantitative data, hindering pooled analysis of preventive effects. Hadaya et al (38) retained ligatures during treatment to model persistent inflammation. Compared with inflammation-resolved therapeutic models, this design produced a larger effect size for ABL but less pronounced regenerative benefits. These findings suggested that persistent inflammation limits the anabolic effects of Scl inhibition, even when short-term improvements in bone turnover markers are observed. The coexistence of preventive and therapeutic designs may therefore obscure biological interpretation and should be considered when extrapolating to clinical contexts.

To assess the robustness of the present results, a leave-one-out sensitivity analysis was performed, sequentially excluding each study and recalculating the pooled effect size. Results remained consistent across all iterations, indicating that no single study disproportionately influenced the overall findings. This supported the validity of the present conclusions despite heterogeneity and small sample sizes.

Notably, transient increases in bone formation markers such as P1NP have been reported, consistent with findings from osteoporosis models (38,42,46). None of the included studies reported osteonecrosis or similar adverse events (38). This suggested a favorable safety profile compared with antiresorptive therapies such as bisphosphonates, which are associated with medication-related osteonecrosis of the jaw. The absence of major adverse outcomes supports the translational potential of Scl inhibition.

A number of limitations should be acknowledged. First, heterogeneity across studies was moderate-high, stemming from differences in animal species, periodontitis induction method, intervention protocol (dose, duration, route of administration, preventive vs. therapeutic design and presence or absence of residual inflammation). These variations complicate direct comparisons and may explain the variability in effect sizes. However, sensitivity analysis demonstrated that the overall conclusions remained robust despite this heterogeneity. Second, the overall risk of bias was high or unclear across a number of domains. A number of studies did not adequately describe randomization procedures and allocation concealment was rarely reported, raising concerns regarding potential selection bias. While a number of studies reported blinding of investigators or outcome assessors, these practices were inconsistent, increasing the risk of performance and detection bias. In addition, small sample size and incomplete methodological reporting further decreased the certainty of evidence and may limit reproducibility. These shortcomings highlight the need for stricter adherence to established guidelines such as the ARRIVE guidelines (50) or SYRCLE (35), which emphasize transparent reporting of randomization, allocation concealment and blinding. Strengthening these aspects in future animal studies may minimize bias and improve reliability. Third, the use of different outcome assessment methods created additional inconsistency. Techniques such as micro-CT, histology, immunohistochemistry and serum biomarkers vary in sensitivity, quantification accuracy and the biological processes they capture. For example, micro-CT provides precise three-dimensional measurements of alveolar bone, whereas histology and immunohistochemistry yield localized and often semi-quantitative information. Serum biomarkers, by contrast, reflect systemic rather than site-specific changes in bone turnover. Furthermore, different biomarkers (such as OCN, P1NP and TRAP5b) and detection approaches (immunohistochemistry vs. serum assays) were used across studies. These discrepancies complicate comparisons, contribute to variability in effect sizes and may explain why some results had low statistical significance. Such inconsistencies may also affect the reliability of the results. Consequently, the overall certainty of evidence was downgraded to moderate to low according to GRADE. Future studies should use standardized and validated detection protocols or coordinate primary outcome measures and biomarker panels, to enhance comparability, reproducibility and reliability. Finally, the predominance of ligature-induced models (7/8 studies) limits the generalizability of findings to bacterially driven periodontitis (37-44). These limitations underscore the need for more rigorous and standardized protocols in future. Clear reporting of randomization, allocation concealment and blinding, along with coordination of biomarker detection methods and outcome assessments, are key in reducing heterogeneity and improving evidence quality. In parallel, systematic investigation of preventive vs. therapeutic settings, optimized dose-time regimens, alternative delivery routes and diverse disease models are necessary to delineate the translational potential of Scl inhibition in periodontal regeneration (37-44).

In conclusion, the present analysis suggested Scl inhibition notably decreased ABL and enhanced the expression of bone formation markers in preclinical periodontitis models. The improved efficacy of low-dose subcutaneous regimens (5 mg/kg) suggested that sustained Wnt activation rather than transient peak-dose stimulation may underlie the superior therapeutic response. These findings warrant clinical trials evaluating Scl-Ab as an adjunct to conventional periodontal therapy (for example, scaling and root planing), prioritizing dose optimization and standardized imaging protocols.

Supplementary Material

Mechanism of sclerostin-mediated regulation of bone remodeling, integrating Wnt/β-catenin, RANKL/OPG, BMP and inflammatory pathways. LRP, lipoprotein receptor-related protein; RANKL, receptor activator of nuclear factor κ-B ligand; OPG, osteoprotegerin; Runx2, RUNX family transcription factor 2; SOST, sclerostin.

Flow chart of study selection according to PRISMA guidelines. WOS, Web of Science.

Subgroup analysis comparing studies with ligature removal vs. retention. Scl, sclerostin; df, degrees of freedom; LDD, low-dose doxycycline; CAPE, caffeic acid phenethyl ester; SMD, standardized mean difference; IV, inverse variance model.

Leave-one-out sensitivity analysis showing the pooled effect size following sequential exclusion of each study. LDD, low-dose doxycycline; CAPE, caffeic acid phenethyl ester.

Search strategy.

Details of outcomes variables.

SYRCLE (The Systematic Review Centre for Laboratory Animal Experimentation) results.

Level of evidence.

Acknowledgements

Not applicable.

Funding

Funding: No funding was received.

Availability of data and materials

The data generated in the present study may be requested from the corresponding author.

Authors' contributions

MY contributed to data analysis using RevMan and EndNote software, constructed figures, analyzed data, and wrote and revised the manuscript. YM conceived the study, constructed figures, analyzed data, and wrote and revised the manuscript. XQ designed the study methodology, contributed to project administration and resources, and provided supervision. ZQ contributed to the study conception, critical revision of the manuscript for important intellectual content, and overall supervision of the project. JS contributed to study validation, data interpretation and resource coordination. MY and YM confirm the authenticity of all the raw data. All authors have read and approved the final manuscript.

Ethics approval and consent to participate

Not applicable.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References