Anticonvulsant Effects of Terpenes in Animal Models: A Systematic Review with Meta-Analysis

Background/Objectives: Epilepsy affects more than 50 million people worldwide, and approximately one-third of patients exhibit drug resistance to conventional treatments. In this context, terpene compounds have emerged as promising alternatives due to their neuroprotective and anticonvulsant properties. This study aimed to analyze the preclinical evidence regarding the anticonvulsant effects of compounds belonging to the terpene class in animal models of epilepsy and seizures. Methods: A systematic review and meta-analysis (PROSPERO CRD420261449697) were conducted according to the PRISMA guidelines. Searches were performed in the PubMed, Web of Science, Embase and Scopus databases. The risk of bias was assessed using the SYRCLE Risk of Bias tool. For quantitative synthesis, only studies involving rats and mice were included. Random-effects meta-analyses were performed to assess latency to the first seizure (LFS) and latency to death (LD), with subgroup analyses according to the seizure-induction model. Leave-one-out sensitivity analyses were also conducted to assess the influence of individual studies on the pooled estimates. Results: Terpene treatment significantly increased LFS in the overall analysis (standardized mean difference–SMD = 3.08; 95% confidence interval–95% CI: 2.41–3.75; p < 0.00001), although substantial heterogeneity was observed (I2 = 81%). Significant effects were also observed in the pilocarpine (PILO; SMD = 3.45; 95% CI: 2.12–4.78; p < 0.00001) and pentylenetetrazole (PTZ; SMD = 3.00; 95% CI: 2.19–3.81; p < 0.00001) models. In contrast, the pooled effect in the picrotoxin-induced model was not statistically significant (SMD = 4.02; 95% CI: 1.19 to 6.85; Z = 2.79; p = 0.005). The isoniazid-induced model also showed a significant increase in LFS, based on a single study (SMD = 10.45; 95% CI: 4.44–16.47; p = 0.0007). Substantial heterogeneity persisted in the PILO (I2 = 82%), PTZ (I2 = 77%), and picrotoxin (I2 = 90%) subgroups. In contrast, the overall analysis of LD showed no statistically significant effect (SMD = 2.76; 95% CI: −0.24 to 5.77; p = 0.07; I2 = 86%). Subgroup analysis revealed a significant increase in LD in the PTZ model (SMD = 3.25; 95% CI: 2.06–4.45; p < 0.00001; I2 = 0%), whereas no significant effect was observed in the PILO model, which showed substantial heterogeneity (I2 = 95%) and considerable instability in sensitivity analyses. Methodological assessment identified unclear risk of bias in several domains, particularly those related to allocation procedures and randomization, while funnel plot asymmetry suggested possible small-study effects; however, the limited number of studies and substantial heterogeneity precluded firm conclusions regarding publication bias. Conclusions: Terpene compounds may exert preclinical anticonvulsant effects, with the most consistent evidence observed for increased LFS, particularly in the PILO and PTZ models. However, effects on LD were not statistically significant in the overall analysis, and model-specific findings remain preliminary. Substantial heterogeneity and methodological limitations warrant cautious interpretation. Further well-designed preclinical studies are needed to clarify the compound- and model-specific consistency, safety, and mechanisms of these effects before clinical translation can be considered.