Liraglutide, a glucagon-like peptide-1 receptor agonist, inhibits bone loss in an animal model of osteoporosis with or without diabetes

Introduction Liraglutide (Lrg), a novel anti-diabetic drug that mimics the endogenous glucagon-like peptide-1 to potentiate insulin secretion, is observed to be capable of partially reversing osteopenia. The aim of the present study is to further investigate the efficacy and potential anti-osteoporosis mechanisms of Lrg for improving bone pathology, bone- related parameters under imageology, and serum bone metabolism indexes in an animal model of osteoporosis with or without diabetes. Methods Eight databases were searched from their inception dates to April 27, 2024. The risk of bias and data on outcome measures were analyzed by the CAMARADES 10-item checklist and Rev-Man 5.3 software separately. Results Seventeen eligible studies were ultimately included in this review. The number of criteria met in each study varied from 4/10 to 8/10 with an average of 5.47. The aspects of blinded induction of the model, blinding assessment of outcome and sample size calculation need to be strengthened with emphasis. The pre-clinical evidence reveals that Lrg is capable of partially improving bone related parameters under imageology, bone pathology, and bone maximum load, increasing serum osteocalcin, N-terminal propeptide of type I procollagen, and reducing serum c-terminal cross-linked telopeptide of type I collagen (P<0.05). Lrg reverses osteopenia likely by activating osteoblast proliferation through promoting the Wnt signal pathway, p-AMPK/PGC1α signal pathway, and inhibiting the activation of osteoclasts by inhibiting the OPG/RANKL/RANK signal pathway through anti-inflammatory, antioxidant and anti-autophagic pathways. Furthermore, the present study recommends that more reasonable usage methods of streptozotocin, including dosage and injection methods, as well as other types of osteoporosis models, be attempted in future studies. Discussion Based on the results, this finding may help to improve the priority of Lrg in the treatment of diabetes patients with osteoporosis.


Introduction
The World Health Organization (WHO) defined osteoporosis as a progressive systemic skeletal disease characterized by low bone mass and microarchitectural deterioration of bone tissue, leading to increased bone fragility and susceptibility to fracture (1,2).Aside from established risk factors including age, cigarette smoking, low physical activity, the use of drugs such as glucocorticoids, and low calcium and vitamin D levels (3,4), diabetes has recently gained increased attention as a potential risk factor for osteoporosis and fragility fractures (5,6).The likely reasons are related to insulin deficiency (7) and the impact of high glucose on calcium and phosphorus metabolism (8).Given that diabetes is a systemic disease associated with a range of chronic and severe complications, the disability and mortality rates are high once fractures occur in patients (6).Although conventional anti-osteoporosis drugs such as calcium tablets, vitamin D, bisphosphonates, denosumab, and teriparatide have been used to treat osteoporosis (2), they do not address the sustained effects of insulin deficiency and high glucose toxicity on bone metabolism.Therefore, besides conventional treatments, it is advantageous to explore drugs that offer both hypoglycemic and anti-osteoporosis effects.
New therapies for diabetes such as glucagon-like peptide-1 receptor agonists (GLP1Ras) have been shown to exert multiple effects on various organs and tissues, including the cardiovascular system (9)(10)(11)(12), arteries (13)(14)(15), lipid metabolism (16), and bone metabolism (17,18).Liraglutide (Lrg), a representative GLP1Ras, is a novel anti-diabetic and widely used drug that mimics the endogenous GLP-1 to potentiate insulin secretion (19).Studies have demonstrated that osteoblastic cells express functional receptors for GLP-1 (20), and continuous subcutaneous infusion of GLP-1 or Lrg in diabetesrelated osteoporosis models normalized their impaired trabecular architecture and promoted bone formation (8,21).These findings highlight the potential use of Lrg in combating diabetes-related bone loss.However, the evidence provided by a single literature source is limited, and the mechanism of Lrg for osteoporosis-or diabetesrelated osteoporosis has not been systematically summarized.Thus, the present study aims to investigate the pre-clinical evidence and possible mechanisms of Lrg in animal models of osteoporosis.

Methods
The Preferred Reporting Items for Systematic Review and Meta Analyses (PRISMA) checklist was used to structure this study (22).

Data sources and search strategies
A literature search was conducted to identify all published animal experimental studies of Lrg for osteoporosis in PubMed, EMBASE, Cochrane library, Web of Science database, WanFang, Chinese Science and Technology Journal Database, Chinese Biomedical Database, and China National Knowledge Infrastructure from their inception dates to April 27, 2024."Liraglutide OR Victoza" AND "Osteoporosis OR Bone Loss OR Osteopenia OR Bone Metabolism" were used as the search terms in PubMed and were modified to suit other databases.A complete record of search strings in PubMed is provided as an example in Appendix 1.Additionally, the reference lists of potential articles were searched for relevant studies.

Eligibility criteria
The studies were screened by two independent authors (ZW and WD) and included if they met the following criteria: (1) studies assessing the efficacy of Lrg for osteoporosis or bone loss in animal models were included, (2) the treatment group used Lrg as monotherapy with unrestricted medicament type, dosage, duration, and route of administration, compared with a blank control or placebo in the control group, and (4) bone pathology and/or bone mineral density [including lumbar spine bone mineral density (L-BMD) and femur bone mineral density (F-BMD)] and/or bone histomorphometric parameters under micro-CT [trabecular number (Tb.N) and trabecular thickness (Tb.Th)] and/or bone maximum load and/or bone turnover markers [C-terminal crosslinked telopeptide of type I collagen (CTX), N-terminal propeptide of type I procollagen (PINP), and osteocalcin (OC)] and/or indicators of adverse reactions were selected as the primary outcome measures.Indicators reflecting the mechanisms of anti-osteoporosis action of Lrg were selected as secondary outcome measures.Studies were excluded if they (1) were not controlled experiments or in vivo animal experiments, (2) included combination medication in the treatment group, (3) lacked primary outcome indicators or had incomplete data, (4) had inconsistencies between graphic and textual data, and (5) were duplicate publications.

Data extraction
Two reviewers (ZW and YY) independently and systematically performed data extraction, focusing on study design characteristics, animal information, modeling methods, anesthetic details, interventions, and outcomes.Only data pertaining to the highest dose and the final time point were included when the experiments featured multiple Lrg dose groups or various measurement times.Graphical data were measured using Photoshop when results were only available in graphic from and no response was received from the corresponding authors.

Risk of bias in individual studies
Two independent authors (WD and JX) utilized the CAMARADES 10-item quality checklist (23) with minor modifications to assess study quality.The modifications included F -anesthetics without significant bone toxicity or protective activity and G-appropriate animal model with complications or risk factors (including aged, diabetes, hyperlipemia, or hypertensive).The authors first independently selected studies, extracted data, and scored the studies and then discussed disagreements with the corresponding author (QZ) until a consensus was reached.

Statistical analysis
We performed all of the analyses available using RevMan 5.3 software.For continuous data, standardized mean differences (SMDs) and 95% confidence intervals (95% CIs) were calculated to estimate the combined overall effect sizes.Heterogeneity was assessed using the Cochrane Q-statistic test (P < 0.05 was considered statistically significant) and the I 2 statistic test (I 2 < 50% was considered homogeneous).Data were aggregated using a random-effects model if there was high heterogeneity (I 2 > 50%); otherwise, a fixed-effects model was adopted.Potential publication bias was assessed by a visual inspection of the funnel plot and asymmetry test to ensure the reliability of results.Sensitivity analysis and subgroup analyses were performed if necessary.

Study quality
The number of criteria met in each study ranged from 4/10 to 8/ 10, with an average of 5.47.The review authors' judgments on each risk of bias item for each included study are presented in Table 2. Summary of the process for identifying candidate studies.

Bone pathology
Bone pathology was used as the primary outcome measure in three studies (26,27,37).Wang et al. (26) reported that diabetes osteoporosis rats treated with Lrg showed a marked improvement of osteoblasts on the surface of the femoral head, flattening of osteocytes, empty bone lacunae, and pyknosis of bone nuclei in the subchondral region.Two studies (27, 37) found that Lrg could increase trabecular bone and reduce trabecular bone spacing in diabetes osteoporosis rats compared with the control group.

Subgroup analysis
Does the combination of diabetes make a difference in the effect of Lrg on bone resorption?We conducted a subgroup analysis on the primary outcome measure BMD, considering whether diabetes was present.The results indicated that, although the difference was not statistically significant, the effect value of Lrg in the osteoporosis with diabetes group was better than that in the osteoporosis without diabetes group (SMD 2.05 ± 0.52 vs. SMD 1.85 ± 0.50, P = 0.59; Figure 6).Although not pronounced, Lrg appears to potentially increase the efficacy by mitigating the harmful effects of high blood sugar on osteoporosis while combating the disease itself.Further animal research is required to verify this potential advantage in the future.

Summary of evidence
This is the first animal systematic review to include 17 studies with acceptable quality that estimate the efficacy and mechanisms of Lrg in models of osteoporosis.The findings indicate that Lrg possesses anti-osteoporosis potential while also lowering blood glucose levels.

Limitations
Several limitations exist within the current studies: (1) the potential for negative studies to exaggerate the efficacy of Lrg in osteoporosis due to reporting biases, (2) selection bias is likely due to the exclusive search of Chinese and English language databases, (3) methodological deficiencies are evident in the lack of blinded induction of models, blinded assessment of outcomes, and adequate sample size calculations, (4) the impact of obesity factors on osteoporosis remains controversial (40)-thus, caution is advised in interpreting results from studies treating osteoporosis caused by hyperlipidemia with liraglutide, (5) no study reported on disinfection during invasive procedures such as intraperitoneal injections, subcutaneous injections, and blood glucose measurements, which are crucial in maintaining integrity in animal studies, especially diabetic models, (6) the majority of studies did not document the incidence of rats being dropped due to complications during the modeling process, and (7) few studies addressed bone pathology directly.

Implication
In terms of methodology, although most studies met the scoring points of the CAMARADES 10-item quality checklist (23), the absence of crucial standards such as blinded induction of models, blinded assessment of outcomes, and rigorous sample size calculations could undermine the reliability of the findings.Adherence to the ARRIVE guidelines (41) is recommended to address these issues.When describing sample size calculation, the rational for the number of animals used should be clearly stated along with details of any calculations performed (42,43).Measures taken to minimize the effects of subjective bias when allocating animals to treatments (e.g., randomization procedures) and when assessing results (e.g., details on who was blinded and at what stage) should also be documented.Wang et al. (44) provided a robust example of how to describe sample size, random grouping postmodeling, and blinded evaluation of outcomes.Experimental animals with comorbidities such as advanced age, obesity, hypertension, hyperglycemia, or other risk factors may more closely mirror the physiology of patients with osteoporosis, potentially increasing the clinical relevance of research findings (45).However, it is necessary to adjust modeling approaches, such as drug dosage and mode of administration to optimize the success rate and safety of complex models in animals.In the included studies, six utilized an ovariectomized osteoporosis model with diabetes (8,21,(26)(27)(28)(29).Based on bilateral oophorectomy, three studies (8,21,26) established a diabetes model using an intraperitoneal injection of STZ at doses greater than or equal to 60 mg/kg; three studies (27-29) used STZ doses between 30 and 35 mg/kg combined with a high-fat and high-sugar diet.The inappropriate use of high doses of STZ has been associated with increased animal suffering and mortality (46).Previous studies (46)(47)(48) have shown that doses of 60 mg/kg body weight and above can be harmful or lethal to rats.Therefore, it is inappropriate to inject large doses of STZ in conjunction with major surgical models such as bilateral ovariectomy without concurrently reporting mortality, side effects, and corresponding treatments (8,21,26).The multiple Forest plot: effects of liraglutide on femur bone mineral density in the subgroup of whether diabetes is combined.Forest plot: effects of liraglutide for increasing the level of osteoprotegerin (OPG) both in serum and bone compared with the control group.3 (49-58).The authors recommend that low-dose STZ or low-dose STZ plus high-fat feeding may be more suitable for composite models.Moreover, it is worth noting that almost all included diabetes models used intraperitoneal instead of intravenous injections of STZ for modeling.This is significant as an accidental delivery of STZ into the sub-peritoneal or bowel space may decrease the success rate and increase the mortality (46,59).Osteoporosis models with increased bone resorption as the dominant mechanism, including ovariectomized osteoporosis, diabetic osteoporosis, and glucocorticoid models, were used in the present studies.It is suggested that future composite models can be based on other osteoporosis models rather than solely on the ovariectomized osteoporosis model.The possible mechanisms of Lrg-mediated bone protection from the current findings are summarized as follows: (1) In bone tissue, OPG competitively binds to RANKL, blocking its blinding to RANK on the surface of osteoclasts, thus inhibiting osteoclast maturation (60).Studies indicate that the OPG/RANKL/RANK signaling pathway is increased to counteract bone resorption after Lrg treatment.STAT3 has been identified as a potential target activated by Lrg to upregulate OPG/RANKL (P < 0.05) ( 26); (2) The levels of OPG, RANKL, and RANK are regulated by a variety of cytokines and hormones that either promote or inhibit osteoclast formation, Forest plot: effects of liraglutide for reducing the level of receptor activator of nuclear factor-k B ligand (RANKL) both in serum and bone compared with the control group.

Conclusion
The pre-clinical evidence reveals that Lrg is capable of partially reversing osteopenia in animal models likely by activating osteoblast proliferation through promoting the Wnt signal pathway and p-AMPK/PGC1a signal pathway and inhibiting the activation of osteoclasts by inhibiting the OPG/RANKL/RANK signal pathway through anti-inflammatory, anti-oxidant, and anti-autophagic pathways.This finding may help to improve the priority of Lrg in the treatment of diabetes patients with osteoporosis.

FIGURE 3 Forest
FIGURE 3Forest plot: effects of liraglutide for increasing lumbar spine bone mineral density (L-BMD) compared with the control group.
FIGURE 4 (A) Forest plot: effects of liraglutide for increasing trabeculae linear density (Tb.N) compared with the control group.(B) Forest plot: effects of liraglutide for increasing trabeculae thickness (Tb.Th) compared with the control group.(C) Forest plot: effects of liraglutide for increasing object surface/volume ratio (BV/TV) compared with the control group.

FIGURE 5 Forest
FIGURE 5Forest plot: effects of liraglutide for increasing the level of osteocalcin (OC) compared with the control group.

TABLE 1
Characteristics of the included studies.

TABLE 2
Risk of bias of the included studies.

TABLE 3
Discrepancies between blood glucose levels, type of diabetes, and mortality with varying doses of STZ.