We further conducted a meta-analysis to evaluate the efficacy of MCs in focal ischemic stroke using middle cerebral artery occlusion (MCAO) models. Four databases (WoSCC, PubMed, Embase, and Cochrane Library) were searched from 1 January 2000 to 30 June 2025, with the last search on the latter date. The search strategy combined medical subject headings with free-text terms. Detailed search strategy is provided in Supplementary Table 1. In addition, grey literature databases (Open Grey, bioRxiv, and medRxiv) were searched to identify potentially eligible preprints.

The inclusion criteria for this study were as follows: (1) Study type: Published animal experiments; (2) Study subjects: Focal ischemic stroke models induced by MCAO; (3) Intervention type: The experimental group received MCs or their derivatives, while the control group was administered conventional drugs, vehicle treatment, or a blank control; (4) Outcome measures: cerebral infarct volume, neurological function score (NFS), CBF, BBB permeability and brain water content, oxidative stress markers, inflammatory indicators [tumor necrosis factor-α (TNF-α), interleukin-1β (IL-1β)], apoptosis indicators [number of terminal deoxynucleotidyl transferase (TdT)-mediated dUTP-biotin nick end labeling (TUNEL)-positive cells], and excitotoxicity index [glutamate/N-acetylaspartate (Glu/NAA) ratio, lactate/N-acetylaspartate (Lac/NAA) ratio]. Exclusion criteria included the following: (1) Reviews, comments, duplicate publications, conference abstracts, case reports, and editorials; (2) Studies that employed other interventions in combination; (3) Studies lacking a control group; (4) Non-English published studies; (5) Studies that did not provide a clear treatment protocol; (6) Studies that failed to report sample size in detail.

Two evaluators independently conducted literature screening and data extraction processes. After removing duplicates, the preliminary screening was performed by reviewing the titles and abstracts of the articles. Then, full-text reading was carried out to further screen the articles according to the inclusion and exclusion criteria. The screening results from both evaluators were cross-checked for consistency. Finally, all eligible articles were identified, and relevant data were extracted, including title, publication year, author names, study subjects, interventions, and outcome measures. When multiple dosing times or doses were present in the literature, data corresponding to the best therapeutic effect were extracted. To comprehensively assess the intervention efficacy, we contacted corresponding authors via email for studies with data presented only in graphical form to obtain complete information necessary for our analysis. If no response was received, data were extracted with WebPlotDigitizer program. We used the following formula for data transformation: SD = SEM × n .

Two evaluators independently assessed the quality of the included studies and drew the risk of bias graph. The assessment of literature quality was conducted using the Systematic Review Center for Laboratory Animal Experimentation (SYRCLE) risk of bias tool, which comprises 10 items (Hooijmans et al., 2014). In case of disagreement during the above process, a third evaluator was invited to resolve the conflicts and achieve consensus.

The effect size of the outcome measures was described as the standardized mean difference (SMD) with 95% confidence interval (CI). A test for heterogeneity among studies was performed to select the corresponding model. If I² < 50%, the fixed-effects model was used. Otherwise, the random-effects model was applied, accompanied by subgroup analyses to explore potential sources of heterogeneity. The threshold for statistical significance was set at p < 0.05. When more than 10 studies were included for a particular outcome measure, funnel plots were constructed and the Egger test was performed to assess publication bias. All data analyses were performed with Review Manager 5.3 (RevMan, Cochrane Collaboration) and Stata 17.0 software.

In this study, 241 publications related to the application of MCs in cerebral ischemia were identified, including 144 articles and 97 reviews. The annual publication output is presented in Figure 1. The findings revealed three distinct phases of research activity. From 2000 to 2010, it was the initial stage of research, with low and fluctuating publication output. From 2011 to 2017, it entered a period of rapid growth. Since 2018, it has been a period of stable development, with publication numbers remaining high. In 2025, there was a slight decrease to 9 publications, but we only searched for early publications for that year, and subsequent publication numbers may still increase. In general, from 2000 to June 2025, the number of publications shows an overall upward trajectory with fluctuations, reflecting a sustained growth in researchers’ attention and research interest in this field.

The international cooperation network showed that 43 countries or regions were actively engaged in this research field (Figure 2A). The United States ranked first with 74 publications, followed by Italy with 32. The top 10 countries/regions in terms of number of publications are presented in Table 1, accounting for 77.6% of all publications. Notably, the United States led both productivity and the number of collaborators, highlighting its core contributions and preeminent position within this research topic.

A total of 394 institutions worldwide conducted research in related fields (Figure 2B). Table 1 lists the top 10 institutions in terms of contribution. Complutense University of Madrid topped the list in terms of the number of publications and total link strength, demonstrating the greatest influence on research concerning MCs and cerebral ischemia. It is noteworthy that there may not be a direct correlation between research output and collaboration among institutions; those exhibiting lower output may be more active in cooperation. For instance, Hospital Clinico San Carlos and Sapienza University of Rome displayed high total link strengths despite their relatively modest publication counts. In addition, the depth and breadth of cooperation between institutions required further enhancement.

A co-authorship network was constructed with authors as nodes, comprising 1,127 core authors (Figure 2C). Table 1 shows the top 10 most productive authors. The author with the greatest contribution was Jose Martinez-Orgado, who published 10 articles. The academic team centered around Jose Martinez-Orgado occupied a central position in the network, reflecting extensive collaborations and considerable influence. Many other research groups also existed within the author network. The fewer links and greater distance between these teams suggested that current cerebral ischemia research focused more on collaboration within teams rather than across different teams.

Analysis of the keyword co-occurrence network can clearly reveal the current status and hotspots in the research field. Figure 3A shows a keyword network consisting of 1,245 nodes. We compiled the top 20 keywords, as shown in Table 2. The search terms “cannabinoids,” “stroke,” and “cerebral ischemia” ranked high in frequency, reflecting that this field has attracted extensive attention and research interest of investigators. Using the log-likelihood ratio for keyword clustering and visualization analysis, we obtained 9 cluster combinations. In our clustering model (Q = 0.4168, S = 0.7763), both the cluster structure and results were robust. The 9 clusters were labeled as follows: #0 cerebral ischemia, #1 THC, #2 amyotrophic lateral sclerosis (ALS), #3 brain injury, #4 cell death, #5 artery occlusion, #6 in vivo, #7 cannabis, and #8 anandamide, as shown in Figure 3B. It is worth noting that the ALS cluster was primarily derived from the review articles included in the analysis. However, there is currently insufficient original research to directly confirm the association between the therapeutic effects of MCs on ALS and the mechanisms of cerebral ischemia. Therefore, this study focused on the core topic clusters related to cerebral ischemia.

Combined with the results of keyword co-occurrence and clustering, the research hotspots in this field were summarized into three distinct areas. The keywords “neuroprotection,” “inflammation,” “endocannabinoid system,” “cannabinoid receptor,” and “cell death” revealed the molecular mechanisms through which cannabinoids exert neuroprotective effects and mitigate brain damage after cerebral ischemia. The results related to “animal model,” “brain damage,” “in vivo,” and “artery occlusion” showed that the current hot research topics were primarily focused on pathological models and experimental studies. This study encompassed both in vivo animal models and in vitro cellular models. The in vivo models included focal cerebral ischemia models [transient middle cerebral artery occlusion (tMCAO), permanent middle cerebral artery occlusion (pMCAO), and photothrombosis model], as well as models for studying global cerebral ischemic injury (two-vessel occlusion, four-vessel occlusion, and hypoxic–ischemic brain damage model). The in vitro models comprised the oxygen–glucose deprivation (OGD) model and the oxygen–glucose deprivation/reoxygenation (OGD/R) model. In treatment, the most widely used cannabinoids were “cannabidiol” and “delta 9 tetrahydrocannabinol.”

Keyword burst analysis can reflect the prevailing topics in a research field during specific periods thereby offering a scientific basis for predicting frontier development trends. The top 15 keywords with the highest burst rates are shown in Figure 3C. In the initial stages the protective effects of the cannabinoid HU-211 in closed head injury and focal cerebral ischemia were primarily investigated with rats being the main experimental subjects. Concurrently attention began to shift towards the role of endocannabinoids (anandamide) in the nervous system. From 2004 to 2014 the overall function and mechanism of the endocannabinoid system became a hot topic. Subsequently research increasingly focused on the manifestation of CBD in ischemic encephalopathy. Over the past 5 years studies on medical cannabis drugs for post-stroke neural function recovery have gradually emerged though still remain at the preclinical stage. This suggests that future trends may prioritize translating preclinical studies of MCs into clinical applications for ischemic stroke.

A systematic search of WoSCC, PubMed, Embase and Cochrane Library databases was conducted, and 3,936 records were initially retrieved. After removing duplicates, the remaining 2,485 records were screened by title and abstract, during which 2,303 were excluded as irrelevant to the research topic. Subsequently, full-text assessments were conducted on the 182 articles. Among these, 156 were excluded for the following reasons: use of non-target cannabinoids, non-ischemic stroke models, in vitro studies, unavailable full texts, insufficient outcome reporting, absence of a control group, concomitant use of other drugs, and lack of reported sample size. Additionally, one record from the grey literature was found to be already included in the above formally published databases. Ultimately, 26 articles met the predefined criteria and were included for the meta-analysis (Lavie et al., 2001; Leker et al., 2003; Teichner et al., 2003; Hayakawa et al., 2004; Mishima et al., 2005; Hayakawa et al., 2007a; Hayakawa et al., 2007b; Hayakawa et al., 2007c; Durmaz et al., 2008; Hayakawa et al., 2008; Hayakawa et al., 2009; Ceprián et al., 2017; Khaksar and Bigdeli, 2017a; Khaksar and Bigdeli, 2017b; Khaksar and Bigdeli, 2017c; Rodríguez-Muñoz et al., 2018; Yokubaitis et al., 2021; Khaksar et al., 2022; Liu et al., 2022; Meyer et al., 2022; Lavayen et al., 2023; Xu et al., 2023; Chen et al., 2024; Villa et al., 2024a; Villa et al., 2024b; de Souza Stork et al., 2025). The screening process was performed independently by two evaluators. The study selection process is illustrated in Figure 4.

The included studies spanned 2001 to June 2025. The experimental animals primarily consisted of rats and mice, with 15 and 11 studies using these species, respectively. Rat models comprised Wistar, Sprague–Dawley (SD), and spontaneously hypertensive rats (SHR), while mouse models included ddY, C57BL/6, and CD1 strains. The majority of experiments were conducted in male animals. Among the anesthesia methods used, isoflurane was the most frequently employed in 7 studies. Other anesthetics utilized included halothane, phenobarbital, sevoflurane, and ketamine-xylazine. Anesthetic details were not reported in 3 studies. 21 studies implemented a transient ischemia model, with ischemia durations ranging from 15 min to 4 h, while 5 studies utilized a permanent ischemia model. The MCs included CBD, THC, HU-211, VCE-004.8, HU-210, Abnormal cannabidiol (AB-CBD), and Full-spectrum Cannabis sativa extract (FSC), with CBD being the most extensively studied compound. Most studies employed intraperitoneal injection as the route of administration, followed by intraventricular and intravenous injections. One study utilized oral gavage. The timing of administration was categorized into 3 approaches: pre-ischemia, post-ischemia, or combined pre- and post-ischemia. In terms of outcome measures, 22 studies reported cerebral infarct volume. NFS were assessed in 11 studies, of which 4 studies used the modified Bederson score, 3 used the Bederson score, 2 used the Zea-Longa score, and 2 used the sensorimotor deficit score. Additional outcomes included CBF in 4 studies, apoptosis (TUNEL-positive cells) in 6 studies, brain water content in 2 studies, BBB permeability in 2 studies, inflammatory cytokines (TNF-α and IL-1β) in 4 studies, and metabolic ratios (Glu/NAA and Lac/NAA ratios) in 2 studies. The basic characteristics of the included studies are shown in Table 3.

Based on the SYRCLE risk of bias tool for animal studies, we evaluated the methodological quality of the 26 included studies. 11 studies employed a randomized controlled design. Specifically, one study used computer-generated random numbers, while 10 studies only mentioned randomization without providing detail. Baseline characteristics were well-balanced across all studies, with 3 studies reporting comparable gender ratios between groups. The methods for allocation concealment and random allocation to cages were not described in any study. Regarding blinding, 3 studies reported blinding of experimenters, and 8 reported blinding of outcome assessors. No study described methods for random outcome assessment. All studies had complete data with no evidence of selective reporting. 3 studies did not report specific anesthetics used, which might introduce potential biases. A summary of the risk of bias is presented in Figure 5. The detailed results of the quality assessment of included studies are shown in Supplementary Table 2 and Supplementary Figure 1.

22 studies reported changes in cerebral infarct volume after treatment. Heterogeneity across studies was significant (p < 0.00001, I² = 60%), necessitating the use of a random-effects model for analysis. The results demonstrated a significant reduction in cerebral infarct volume in the experimental group compared to the control group (SMD = −2.12, 95% CI [−2.52, −1.73], p < 0.00001; Figure 6).

11 studies assessed the impact of MCs on ischemic stroke using NFS. No significant heterogeneity was observed across the studies (p = 0.07, I² = 42%), prompting the use of a fixed-effects model for data analysis. The results showed that the improvement in NFS was significantly greater in the experimental group compared to the control group (SMD = −1.62, 95% CI [−1.99, −1.26], p < 0.00001; Figure 7). Notably, significant improvement trends were observed on multiple scoring scales, including the Bederson score (SMD = −2.02, 95% CI [−2.78, −1.27], p < 0.00001), Zea-Longa score (SMD = −1.49, 95% CI [−2.33, −0.64], p = 0.0005), modified Bederson score (SMD = −1.46, 95% CI [−2.08, −0.85], p < 0.00001), and sensorimotor deficit score(SMD = −1.58, 95% CI [−2.31, −0.85], p < 0.0001).

A total of 4 studies reported changes in CBF. Heterogeneity was detected among the studies (p = 0.02, I² = 63%), and a random-effects model was selected. The results demonstrated that compared to the control group, the experimental group showed a significant enhancement in the recovery of CBF after ischemic injury (SMD = 2.45, 95% CI [0.07, 4.84], p = 0.04; Figure 8A).

In total, 2 studies reported on BBB permeability indicators. There was no significant heterogeneity among these studies (p = 0.81, I² = 0%), leading to the selection of a fixed-effects model. The results showed that the experimental group effectively reduced BBB permeability (SMD = −2.24, 95% CI [−3.36, −1.13], p < 0.0001; Figure 8B).

A total of 2 studies examined brain water content. No heterogeneity was observed across studies (p = 0.44, I² = 0%), thus a fixed-effects model was selected. The results demonstrated that the brain water content of the experimental group was significantly reduced after treatment (SMD = −1.97, 95% CI [−3.03, −0.91], p = 0.0003; Figure 8C).

A total of 4 studies reported changes in the number of TUNEL-positive cells. Heterogeneity was detected among the studies (p = 0.07, I² = 58%), thus a random-effects model was used for analysis. The analysis showed that the number of TUNEL-positive cells in the experimental group was significantly lower than that in the control group (SMD = −1.78, 95% CI [−2.75, −0.81], p = 0.0003; Figure 8D).

One study reported that administration of CBD in rats enhanced the activity of superoxide dismutase (SOD) (SMD = 4.02, 95% CI [1.44, 6.60], p = 0.002; Supplementary Figure 2A) and catalase (CAT) (SMD = 2.48, 95% CI [0.61, 4.35], p = 0.009; Supplementary Figure 2B) in brain tissue, while reducing malondialdehyde (MDA) levels (SMD = −4.59, 95% CI [−7.46, −1.73], p = 0.002; Supplementary Figure 2C). Another study reported that CBD treatment significantly decreased reactive oxygen species (ROS) levels in the brain (SMD = −1.67, 95% CI [−2.77, −0.56], p = 0.003; Supplementary Figure 2D). In another study, FSC treatment increased the activity of CAT in lung tissue (SMD = 5.26, 95% CI [3.24, 7.27], p < 0.00001; Supplementary Figure 2E), and reduced MDA levels (SMD = −2.01, 95% CI [−3.12, −0.89], p = 0.0004; Supplementary Figure 2F).

A total of 4 research reports examined TNF-α levels, with significant heterogeneity detected among the studies (p = 0.02, I² = 64%). Consequently, a random-effects model was selected for further analysis. The results showed that the experimental group effectively reduced the content of TNF-α (SMD = −1.56, 95% CI [−2.68, −0.44], p = 0.006, Figure 9A).

A total of 2 studies reported on IL-1β levels. There was no heterogeneity across studies (p = 0.72, I² = 0%), and a fixed-effects model was chosen for analysis. The results showed that the experimental group effectively reduced the level of IL-1β (SMD = −1.07, 95% CI [−1.96, −0.18], p = 0.02; Figure 9B).

A total of 2 studies examined the Glu/NAA ratio. There was no heterogeneity among the studies (p = 0.16, I² = 49%), so the analysis was performed using a fixed-effects model. The results showed that the experimental group significantly improved the Glu/NAA ratio (SMD = −0.89, 95% CI [−1.49, −0.30], p = 0.003; Figure 9C).

In total, 2 studies measured the Lac/NAA ratio. There was no heterogeneity among these studies (p = 0.26, I² = 20%), so a fixed-effects model was selected. The results showed that the Lac/NAA ratio was significantly decreased in the experimental group (SMD = −1.89, 95% CI [−2.59, −1.18], p < 0.00001; Figure 9D).

To explore potential sources of heterogeneity, we performed subgroup analyses of cerebral infarct volume, CBF, TUNEL and TNF-α according to animal species, anesthetic agent, model type, occlusion duration, drug class, route of administration and timing of administration.

For cerebral infarct volume, no significant subgroup heterogeneity was found across animal species (p = 0.45), model type (p = 0.20), route of administration (p = 0.11), or timing of administration (p = 0.31). Significant differences in subgroup effect sizes were detected for anesthetic agent (p = 0.03), occlusion duration (p = 0.03) and drug class (p = 0.01). In terms of drug class, various MCs showed a positive effect on reducing infarct volume. Among them, CBD exhibited the most robust therapeutic effect (SMD = −1.89, 95% CI [−2.31, −1.46]), supported by 23 comparative analyses. Among anesthetics, isoflurane demonstrated the largest effect size (SMD = −2.61, 95% CI [−3.80, −1.42]). For occlusion duration, permanent occlusion (SMD = −2.96, 95% CI [−4.36, −1.55]) exhibited the largest effect size in reducing infarct volume. Among the transient occlusion groups, a 60-min occlusion (SMD = -2.33, 95% CI [−3.13, −1.52]) yielded more consistent pooled results. Furthermore, infarct volume reductions were consistently observed in both transient and permanent MCAO models. Continuous administration (SMD = −2.56, 95% CI [−3.13, −1.98]) was more effective than pre-MCAO (SMD = −2.02, 95% CI [−2.72, −1.32]) or post-MCAO (SMD = −1.95, 95% CI [−2.57, −1.32]) administration. The results of the subgroup analysis of cerebral infarct volume are shown in Table 4.

For CBF, drug class (p = 0.002) and route of administration (p = 0.02) were the primary influencing factors. CBD significantly increased CBF (SMD = 10.29, 95% CI [4.42, 16.15]), an effect substantially greater than that observed with THC (SMD = 1.65, 95% CI [−0.02, 3.32]). Intraperitoneal injection (SMD = 4.13, 95% CI [0.77, 7.50]) was superior to intravenous administration (SMD = 0.00, 95% CI [−1.24, 1.24]). No significant heterogeneity was found for timing of administration (p = 0.07). Continuous administration showed a trend toward greater CBF improvement (SMD = 5.62, 95% CI [0.54, 10.70]). The results of the subgroup analysis of CBF are shown in Supplementary Table 3.

For TUNEL, no significant subgroup differences were found across stratified factors (p = 0.51, p = 0.35, p = 0.35). Nevertheless, two cannabinoids, HU-211 (SMD = −2.67, 95% CI [−4.62, −0.72]) and VCE-004.8 (SMD = −1.37, 95% CI [−2.45, −0.28]), showed potential protective effects. The results of the subgroup analysis of TUNEL are shown in Supplementary Table 4.

For TNF-α, no significant subgroup differences were observed for animal species (p = 0.46), anesthetic agent (p = 0.09), route of administration (p = 0.30) or timing of administration (p = 0.08). Significant heterogeneity in subgroup effect sizes was identified across drug class (p = 0.02). Among them, CBD administration demonstrated the largest effect size for reducing TNF-α levels (SMD = −1.97, 95% CI [−3.18, −0.75]). The results of the subgroup analysis of TNF-α are shown in Supplementary Table 5.

In this study, publication bias was assessed for two outcome indicators, including infarct volume and NFS (Figure 10). The analysis revealed that the funnel plots for both indicators showed asymmetric distribution characteristics. The Egger test further indicated significant publication bias for infarct volume (p < 0.001) and NFS (p < 0.001). A trim-and-fill analysis was subsequently performed to evaluate the potential impact of publication bias (Supplementary Figure 3). The results showed a minimal deviation between the adjusted and original effect sizes. This suggests that despite the presence of publication bias, the results of this study remain robust, as such bias is insufficient to materially change the overall conclusions.

MCs can enhance CBF in ischemic brain tissue by modulating cerebral vascular tone (Benyó et al., 2016; Richter et al., 2018). In a mouse model of focal cerebral ischemia, the reduction of CBF in the infarction site was reversed by treatment with 3.0 mg/kg CBD through activation of the 5-HT1A receptor (Mishima et al., 2005). Additionally, THC can facilitate redistribution of blood flow in ischemic brain regions, with acute administration increasing CBF in specific areas such as the anterior cingulate cortex, frontal cortex, and insula (Ogunbiyi et al., 2020). One study further demonstrated that both CBD and THC can significantly increase cortical blood flow when administered via intraperitoneal injection; moreover, the efficacy of CBD remained stable after repeated dosing for 14 days (Hayakawa et al., 2007c). Therefore, MCs can protect the animal brain from ischemic injury by improving perfusion.

The breakdown of BBB is a critical pathological feature following ischemic stroke (Qiu et al., 2021). The primary mechanisms involve the destruction of tight junction proteins and enhanced vesicular transport, consequently leading to leakage of peripheral neurotoxic substances and cerebral edema (Candelario-Jalil et al., 2022; Zhu et al., 2024). CBD has been shown to significantly reduce Evans Blue extravasation and the degree of cerebral edema in the ischemic hemisphere of MCAO rats (Khaksar and Bigdeli, 2017a). In the OGD model, CBD attenuated cellular permeability by activating the peroxisome proliferator-activated receptor γ (PPAR-γ) and 5-HT1A signaling pathways (Hind et al., 2016). In addition, a key mechanism underlying BBB disruption is the upregulation of matrix metalloproteinase-9 (MMP-9), which degrades extracellular matrix and tight junction proteins (Ji et al., 2023). CBD amino quinone derivatives have been found to reduce BBB leakage in stroke mice, potentially through inhibiting MMP-9 expression in brain tissue (Lavayen et al., 2023).

Following ischemic stroke, the impaired energy metabolism in brain tissue leads to overproduction of ROS and reactive nitrogen species (RNS), thereby inducing oxidative stress injury (Li et al., 2022). Experimental evidence suggested that CBD could modulate the content of ROS by targeting mitofusin-2 (MFN2) protein (Xu et al., 2023). The overactivation of inducible nitric oxide synthase (iNOS) sharply raises intracellular RNS levels (Wu et al., 2023). Administration of CBD significantly reduced post-ischemic ROS and iNOS levels, potentially mediated by regulating the activity of cyclin-dependent kinase regulatory subunit 1B (CKS1B) (Chen et al., 2024). SOD and CAT are key endogenous antioxidant enzymes that can alleviate neurotoxicity by scavenging oxygen free radicals (Afzal et al., 2023). Additionally, the overload of free radicals induces lipid peroxidation in the cell membrane, resulting in the accumulation of damaging markers like MDA (Pawluk et al., 2024). CBD enhanced SOD and CAT activity, and reduced MDA levels in ischemic lesion sites (Khaksar et al., 2022). FSC has also been demonstrated to have antioxidant effects by modulating peripheral organ oxidative stress parameters, such as MDA, CAT, and nitric oxide (NO) levels (de Souza Stork et al., 2025). Additionally, HU-210 effectively inhibited the overproduction of ROS within mitochondria, thereby enhancing the tolerance of neural tissue to ischemic injury (Cai et al., 2017).

The initial ischemic event leads to neuronal ATP depletion and ion channel dysfunction, subsequently causing cell membrane depolarization and massive influx of Ca²⁺. This process promotes the exocytosis of glutamate vesicles (Li et al., 2026). Excess glutamate activates both ionotropic and metabotropic glutamate receptors on the postsynaptic membrane, thereby inducing excitotoxic injury (Yang et al., 2024). CBD can inhibit the excitatory effects following ischemic stroke, which is associated with its reduction of Glu/NAA levels in brain tissue (Ceprián et al., 2017). Its derivative, VCE-004.8, also exhibits similar neuroprotective effects (Villa et al., 2024b). Another study demonstrated that THC suppressed glutamate release from presynaptic terminals of hippocampal neurons, thereby attenuating synapse-mediated neuronal damage (Gilbert et al., 2007). Moreover, CBD and THC have been shown to effectively reduce neurotoxicity mediated by N-methyl-d-aspartate (NMDA) and 2-amino-3-(4-butyl-3-hydroxyisoxazol-5-yl)propionic acid (AMPA) receptors (Hampson et al., 1998). HU-211 can directly inhibit the activity of NMDA receptor (Fernández-Ruiz et al., 2015). Excitotoxicity-induced disruption of intracellular calcium homeostasis is a critical pathogenic mechanism in ischemic brain injury. CBD counteracted Ca²⁺ overload-induced neurotoxicity by upregulating the expression of Na+/Ca2 + exchanger 2 (NCX2) and Na+/Ca2 + exchanger 3 (NCX3) in cortical neurons (Khaksar and Bigdeli, 2017a).

Acute CBF interruption-induced neuroinflammation is a critical pathogenesis of ischemic stroke, typically leading to poor prognosis. The inflammatory response initiates within minutes of cerebral ischemia and persists throughout all stages of the disease (Alsbrook et al., 2023). Microglia serve as key effector cells that trigger post-stroke neuroinflammation (Endres et al., 2022). Ischemic injury rapidly activates microglia, leading to the substantial release of pro-inflammatory factors such as TNF-α and IL-1β, thereby triggering an inflammatory cascade (Kumari et al., 2024). These cytokines also activate the nuclear factor-κB (NF-κB) pathway, inducing the production of more inflammatory mediators and exacerbating neurotoxicity (Mussbacher et al., 2023). The expression levels of TNF-α and IL-1β in brain tissue are closely related to the degree of nerve injury (Tirandi et al., 2023). In the MCAO models, the mRNA and protein expression levels of both TNF-α and IL-1β are significantly upregulated (DeLong et al., 2022; Xu et al., 2024). MCs such as CBD (Henshaw et al., 2021), VCE-004.8 (Navarrete et al., 2022), and FSC (Maayah et al., 2020) can reduce the levels of these two mediators, which highlights their potential in mitigating neuroinflammation.

The neuronal apoptosis in the penumbra is a pivotal mechanism underlying the progression of brain injury after ischemic stroke. The intrinsic apoptotic pathway is centrally regulated by the B-cell lymphoma 2 (Bcl-2) family, which includes pro-apoptotic members [e.g., Bcl-2-associated X protein (Bax), Bcl-xL/Bcl-2-associated death promoter (Bad)] and anti-apoptotic factors [e.g., Bcl-2, B-cell lymphoma-extra-large (Bcl-xL)] (Uzdensky, 2019). Under ischemic conditions, the balance shifts toward pro-apoptotic signaling in neurons. This process ultimately manifests in characteristic pathological features at the cellular level, such as DNA degradation, nuclear condensation, organelle destruction, and apoptotic body formation (Vitale et al., 2023). MCs have demonstrated anti-apoptotic effects. CBD not only reduced caspase-3 activity and the ratio of apoptotic bodies in hippocampal neurons but also inhibited the apoptotic pathway by modulating the Bax/Bcl-2 balance (Sun et al., 2017; Khaksar et al., 2022). HU-210 was shown to rescue neuronal apoptosis and limit expansion of the lesion in a rat MCAO model (Cai et al., 2017). HU-211 can affect both caspase-dependent and -independent apoptotic processes by reducing the activity of cathepsin B and cathepsin L (Durmaz et al., 2008; Yagami et al., 2019). Our study demonstrated that MCs reduced the number of TUNEL-positive cells after cerebral ischemia.