Studies were included if they: 1) examined HBV, melittin, or PLA₂ as the intervention; 2) assessed antiviral effects in a human, animal, or in vitro model system; 3) had a control comparison group(s) such as a placebo or no treatment and/or an alternative treatment group(s); and 4) reported quantitative measurements of the outcomes relevant to virus infection (e.g., viral load reduction, replication inhibition, plaque reduction assay data, or clinical endpoints).

The exclusion criteria were as follows: narrative reviews, case reports lacking quantifiable antiviral results, and non-viral pathogens.

A comprehensive search was performed in PubMed, Embase, Web of Science, Scopus and Cochrane Central up to October 2025 using the following keywords: (“bee venom” OR “apitoxin” OR “melittin” OR “phospholipase A2” OR “apipuncture”) AND (virus OR viral OR antiviral OR influenza OR HSV or SARS-CoV-2 or COVID-19). The reference lists of relevant papers and reviews were manually searched for additional studies. No limitations related to language or date of publication were imposed (Supplementary Table S1).

Two reviewers independently screened the titles, abstracts, and full texts. Disagreements were resolved through consensus. The data extracted included the data source, study design, sample size, species, virus type, intervention and outcome details, shine light (wavelength) details, the time duration for each session, and adverse events.

The Cochrane risk-of-bias (RoB) 2 tool was used for human randomized controlled trials (RCTs), while ROBINS-I (Risk Of Bias In Non-randomized Studies—of Interventions) was utilized for non-randomized studies. For animal studies, the SYRCLE tool was used, and a custom quality checklist (e.g., replication, blinding, and cytotoxicity controls) was employed for in vitro studies.

Quantitative analysis was performed using a random-effects model (restricted maximum likelihood, REML) for continuous (standardized mean difference, SMD) and dichotomous (risk ratio, RR) outcomes. Heterogeneity was calculated using the I² and τ² statistics. Pre-specified subgroup analyses were performed according to the virus family, the venom components, and the study design. Funnel plots and Egger’s test were used to estimate publication bias when at least 10 studies were available.

In total, 1,247 records were retrieved. After discarding duplicates, 932 remained for screening. A total of 32 studies were included in the qualitative synthesis (meta-analysis) (Figure 1; Supplementary Tables S2A, B).

Human clinical studies (n = 6) were on topical/injectable HBV preparations for mild coronavirus disease 2019 (COVID-19), herpes simplex virus (HSV), or warts. Studies in animals (n = 14), including mice, ferrets, and chickens, were on infection of influenza, HSV, or enterovirus models. In vitro studies (n = 12) were on melittin and PLA₂ against different strains of viruses in cell cultures. The interventions included crude HBV extracts (0.1–1 mg/ml) and purified melittin or nano-melittin conjugates (0.5–10 µM).

The pooled SMD of the meta-analysis on dexindanol and virus control (VC) on the inhibition of viral replication was −2.15 (95%CI, 1Score = 74). There was a statistically significant difference between the intervention group and the control group (Figure 2).

In animal studies, pulse HBV therapy decreased the viral load, with a pooled mean difference of −1.8 log10 copies/ml (95% CI = −2.4 to −1.2). One clinical study showed greater resolution of symptoms (RR = 1.45, 95%CI = 1.05–2.01); however, the heterogeneity was high (I² = 72%). Examination of the funnel plots indicated potential small-study bias (Figure 3).

The majority of RCTs were rated as “some concerns” due to the lack of blinding, while the animal and in vitro studies had a “moderate risk” rating based on randomization and outcome assessment (Figure 4; Supplementary Tables S3A–D).

An important finding from Supplementary Table S2 is that crude HBV demonstrates antiviral activity comparable to that of purified components in several studies, particularly in vitro. Crude venom contains a complex mixture of peptides and enzymes, including melittin, PLA₂, apamin, mast cell-degranulating peptide, and minor bioactive molecules, which may act synergistically to enhance antiviral efficacy.

Several studies reported that crude venom reduced the viral RNA levels and infectious titers at relatively low concentrations, particularly against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), influenza A, hepatitis B virus, and HSV, suggesting that the combined action of multiple venom constituents may confer advantages over single isolated peptides. However, the variability in the venom composition—driven by the bee species, geographic origin, extraction methods, and storage conditions—represents a major challenge for reproducibility and clinical translation. These findings emphasize the need for standardized venom preparation, biochemical characterization, and potency assays in future studies.

The antiviral effects of HBV are largely attributed to the amphipathic peptide melittin, whose α-helical structure enables insertion into lipid bilayers, resulting in pore formation, membrane destabilization, and viral envelope disruption (20–24). This mechanism explains the higher sensitivity of enveloped viruses to melittin-based interventions while also accounting for dose-dependent cytotoxicity.

Beyond melittin, apamin, a small neuroactive peptide traditionally studied for its effects on ion channels, has emerged as an underexplored but relevant antiviral modulator. Evidence displayed in Supplementary Table S2 indicates that apamin can reduce the viral DNA/RNA levels and viral antigen expression, particularly in hepatitis B virus models. Unlike melittin, apamin does not primarily act through membrane lysis; instead, it appears to influence the host transcriptional regulation and intracellular signaling pathways, potentially affecting viral replication indirectly.

In addition, PLA₂ contributes to antiviral activity by hydrolyzing phospholipids in viral envelopes and host membranes, thereby interfering with viral entry and fusion. Together, these findings support a multi-target antiviral paradigm, in which the HBV components act at different stages of the viral life cycle.

Advances in nanotechnology—such as melittin-loaded liposomes, nanoparticles, or peptide conjugates—have significantly improved the therapeutic indices by reducing the hemolytic activity while preserving the antiviral potency (25–27). These delivery systems may also enable the co-delivery of multiple venom components, allowing the synergistic effects between melittin, PLA₂, and apamin to be harnessed safely.

In addition to their direct antiviral effects, HBV and its components exert significant immunomodulatory actions. Multiple studies have reported the stimulation of type I interferon (IFN) signaling pathways, alongside the suppression of excessive inflammatory responses, including the downregulation of TNF-α, IL-6, and NF-κB signaling (28–30).

This dual immunological profile is particularly relevant for viral infections characterized by immune dysregulation, such as influenza and COVID-19. By enhancing the antiviral innate immunity while limiting cytokine-mediated tissue damage, HBV components may help reduce the viral load and disease severity simultaneously. Apamin, in particular, may contribute to this balance by modulating the calcium-dependent signaling and inflammatory cascades without inducing membrane damage.

Safety remains a critical barrier to clinical application. Across studies, the most frequently reported adverse events associated with HBV exposure were local pain, swelling, and erythema, while systemic allergic reactions were rare, but potentially severe (31). Importantly, the reviewed evidence suggests that purified formulations, controlled dosing, and desensitization protocols can substantially mitigate the risk of anaphylaxis (32).

Crude HBV may pose a higher immunogenic risk compared with purified peptides; therefore, component-specific formulations or standardized venom extracts with defined allergen profiles are preferable for therapeutic development. Preclinical toxicology studies and allergy screening protocols must be rigorously integrated into future clinical trial designs.

Earlier narrative reviews highlighted the antiviral promise of HBV, particularly melittin, but lacked systematic methodology and quantitative synthesis (33–35). The present review advances the field by providing a PRISMA-compliant systematic assessment, integrating mechanistic insights, risk of bias evaluation, and comparative analysis across experimental models. While preclinical findings are largely concordant, translation to human application remains limited by the small sample sizes, the heterogeneity in the venom composition, and the inconsistent dosing strategies.

The strengths of this review include a comprehensive multi-database search, the inclusion of both preclinical and early clinical evidence, and a detailed mechanistic interpretation grounded in experimental data.

Limitations include the small number of human studies, substantial methodological heterogeneity, potential publication bias, and the inherent constraints of in vitro models that may not fully recapitulate physiological conditions. Furthermore, variability in the crude venom composition complicates cross-study comparisons.

Taken together, the findings support HBV, particularly melittin-, PLA₂-, and apamin-based formulations, as a promising platform for antiviral drug development. Future research should prioritize standardization, advanced delivery systems, combination strategies, and well-designed clinical trials to fully realize the therapeutic potential of HBV in antiviral medicine.

Future studies should focus on: