We searched Web of Science, PubMed, Embase, Cochrane Library, Sinomed, and China National Knowledge Infrastructure (CNKI) for articles on the cardioprotective effects of saponins in myocardial ischemia-reperfusion injury. The search terms were “myocardial ischemia-reperfusion injury,” “myocardial I/R injury,” “myocardial ischemia-reperfusion injury,” “myocardial ischemia-reperfusion injury,” “saponins,” “ginsenosides,” “Holothurin,” and “Quillaja Saponins,”. The publication time was limited to before October 2022. Two reviewers (Fan Yang and Xin Su) independently assessed articles for eligibility, and disagreements were resolved by discussion with the corresponding author.

Inclusion criteria: (a) Type of study: In vivo animal studies. (b) Interventions and controls: Saponin extracts derived from traditional Chinese herbs compared with placebo treatment. (c) Outcomes: At least one of the following outcome measures should be reported: Infarct size [infarct size expressed as a percentage of infarct area over the area at risk (AAR)], left ventricle ejection fraction percentage (LVEF), left ventricle fractional shortening percentage (LVFS).

Exclusion criteria: (a) Duplicate reports. (b) Studies with incomplete data and unclear outcome indicators. (c) Studies where the outcome indicators could not be transformed into mergeable data. (d) Studies using incorrect statistical methods that could not be corrected.

Two researchers (Jiahao Sun and Jiarong Fan) independently extracted baseline information (i.e., sample size, country, author, and year), animal characteristics (e.g., age, weight, strain, or species), detailed treatments of included studies Strategy (such as the type of saponin, source plant, dose, method and time of administration) and methods for measuring myocardial infarction size after myocardial I/R injury.

The quality of included studies was assessed and scored by two researchers (Yang Fan and Su Xin) according to the CAMARADES(Collaborative Approach to Meta Analysis and Review of Animal Data from Experimental Stroke) list (18). For each article, we scored 1 point for each of the following terms: statement of a potential conflict of interest, sample size calculation, random allocation to groups, a peer-reviewed publication, compliance with animal welfare regulations, and blinded outcomes assessment. We define 1–2 points as high risk, 3–4 points as medium risk, and 5–6 points as low risk. Disagreements were resolved through discussions with the corresponding author.

This meta-analysis used continuous variables as mean and standard deviation. We calculated pooled effect sizes using random effects models. Effect sizes were expressed as weighted mean differences (WMD) with 95% confidence intervals (CI). To assess heterogeneity, we used the Q statistic and the I² statistic for quantifying. Publication bias was evaluated initially by funnel plots and further detected by Egger's test. If there were significant heterogeneity among studies (p < 0.10), the analysis would be performed by deleting each study, post hoc subgroup analysis (i.e., duration of reperfusion, study type, species, and schedule of preconditioning), and trim-and-fill method to achieve sensitivity analysis. Univariate regression (i.e., study type, species, states, sample size, route of administration and duration of reperfusion, and temporal regimen of preconditioning) was proposed to explore potential heterogeneity sources. For all results except heterogeneity, P < 0.05 was considered statistically significant. We used STATA version 15.1 (STATA Corporation, CollegeStation, TX, USA) for statistical analyses and graphs.

We obtained 300 related articles (PubMed: 64, EMBASE: 47, Web of Science: 55, Cochrane Library: 1, CNKI: 91, Sinomed: 42) by searching in databases, and 182 articles were retained after removing duplicate articles, among which 154 were excluded after the title and abstract screening, and 28 were eligible for full-text review. Five articles did not report infarct size as mentioned in the inclusion criteria, three lacked a placebo group, and three used saponin conversion products as an intervention. Finally, 17 studies of 386 animals (254 in the saponin treatment group and 132 in the control group) met the predetermined inclusion criteria and were eventually included in our meta-analysis (Figure 1) (4, 5, 10, 13, 17, 19–30). Table 1 shows comprehensive information for each study. All studies used mice (C57BL/6) and rats (Sprague-Dawley or Wistar) to investigate the cardioprotective effects of saponins. All studies measured infarct size by Evans blue/TTC double staining. Saponins were administered orally or intravenously, or intraperitoneally. In most studies, the administration began a few days before myocardial ischemia, and some studies were administered within half an hour before ischemia or during reperfusion. All 17 studies were conducted in China and were published between 2013 and 2022.

The quality of the included studies was evaluated in Table 2. Most studies (82.4%) were scored 3 to 4, indicating that the data were generally reliable and the risk of bias was acceptable. Three studies were judged as having a high risk of bias.

A summary analysis of 24 arms with 290 animals (intervention group = 192, control group = 98) showed the consistent effect of saponins on myocardial infarction size (Figure 2). Random effect model analysis showed that compared with placebo treatment, saponins significantly reduced infarct size (WMD: −3.60, 95% CI: −4.45 to−2.74, P < 0.01). There is a large amount of heterogeneity between studies (I²: 84.7%, P < 0.001). The funnel plot and Egger test (P < 0.001) showed the existence of publication bias (Figure 3). After introducing possibly unpublished studies through the trim and fill method, the results were steady (WMD: −2.81, 95% Cl: −3.69 to−1.93, P < 0.001), and the systematic deletion of each study did not affect the summary WMD and corresponding P values (Figure 4), which means that the beneficial effect of saponins in myocardial protection is stable. Subgroup analysis to explore the sources of heterogeneity between studies showed that sample size was a possible source of heterogeneity (Figure 5). Univariate regression failed to reveal any significant correlation between research covariables and effect sizes (Table 3).

Necrosis is thought to be caused by chemical or physical damage. A decrease in ATP production can cause intracellular calcium overload during myocardial ischemia. During reperfusion, the restoration of energy supply under the premise of calcium overload-induced the oscillatory release and uptake of calcium by the sarcoplasmic reticulum, resulting in uncontrolled excessive muscle fiber contraction and damage to cellular structures. The study of Wang et al. showed that Calenduloside E(CE) could repress L-type calcium current induced by L-type calcium channels(LTCC) agonist and regulate calcium homeostasis (37).

After reperfusion, excessive reactive oxygen species (ROS) are produced, which may lead to the oxidation of nucleic acids, lipids, and proteins, further leading to changes in membrane damage, protein function, metabolic disorders, gene mutation, and resulting in oxidative stress (38–41). Studies have shown that Sirt1, Sirt3, and Nrf2 have the function of regulating oxidative stress. He et al. found that Saponins from Rhizoma Panacis Majoris (SPRM) protect against MIRI by activating Sirt1 and Nrf2-related signaling pathways and attenuating oxidative stress (10). Tubeimoside I (TBM) reduces the production of ROS by inhibiting NOX2. On the other hand, TBM increased antioxidant factors such as Nrf2 and NQO1 (26). Zhu et al. found that a new flavonoid glycoside (APG) increased SOD activity, decreased H2O2 and MDA, inhibited mitochondrial oxidative stress, and finally attenuated MIRI via activating PKCε signaling (19).

After MIRI, neutrophil adhesion and infiltration occur in the coronary artery (5, 42–44). The expression of adhesion molecules increases on the surface of microvascular endothelial cells and leukocytes (9, 45, 46), which promotes adhesion, aggregation, and chemotaxis of neutrophils and resists microcirculation, accelerating the process of MIRI (47). In recent years, the role of NLRP3 inflammatory agents in MIRI has become a research hotspot. NLRP3 can combine with Caspase-1 and Asc to form NLRP3 inflammasome, which requires the activation of NF-κB. Yu et al. found that the pretreatment of Panax notoginseng saponins (PQS) significantly down-regulated the expression of NLRP3, ASC, and caspase-1. This effect may be achieved by inhibiting the TLR4/MyD88/NF- κ B signal pathway (24).

Na⁺-K⁺-ATPase is an essential system of cellular energy metabolism. Na⁺-K⁺-ATPase activity decreased significantly after MIRI (48). This situation increases intracellular Na⁺, and more Ca²⁺ enters the cell through Na⁺-Ca²⁺ exchange, resulting in calcium overload (49–51). Total saponins of A. Elata (aralosides, AS) can significantly increase the activity of Na⁺-K⁺-ATPase and Ca²⁺-Mg²⁺-ATPase (17). The study of Wang et al. showed that Calenduloside E(CE) could promote the interaction between LTCC and bcl2-related athanogene 3 (BAG3), reduce MIR-induced calcium overload and play a protective role in the heart (22).

Autophagy is a phenomenon in which cells use lysosomes to digest their organelles or some proteins, and it is the key to maintaining normal metabolism and renewal of some organelles. During myocardial ischemia, autophagy can provide critical nutrients for cell survival and inhibit cell apoptosis and necrosis by digesting non-functional proteins. O^2•− is a potential mechanism of autophagy. Huang et al. found that Saponin astragaloside IV(ASIV) could increase SOD2 levels and reduce the accumulation of O^2•− and autophagosome, alleviating I/R-induced cell death and protecting cardiomyocytes (13).

In short, the pathogenesis of MIRI begins with calcium overload and oxidative stress caused by sudden changes in the environment (54). Calcium overload and oxidative stress interact with each other (55, 56), resulting in the destruction of cell structure and function (57, 58), leading to necrosis or apoptosis (1). The necrotic cells cause a different inflammatory response, stimulating other cells to start pyroptosis. Saponins can regulate calcium balance by improving membrane ion channel activity, reducing calcium overload, and preventing calcium overload-induced cell death and oxidative stress. At the same time, saponins reduce the production of reactive oxygen species, promote the removal of reactive oxygen species, reduce autophagy (59–61), apoptosis and inflammatory responses caused by oxidative stress, and reduce pyroptosis caused by inflammatory responses (6, 62–64). Therefore, traditional Chinese medicine saponins can act on multiple targets in the process of MIRI and protect MIRI through the interaction of many pathways.