Search for a command to run...
Background Chronic hepatitis B virus (HBV) and human immunodeficiency virus (HIV) infections are among the most important global health issues. Virus-directed CAR-T and CAR-NK are promising strategies capable of targeting virally infected cells. The therapeutic potential, safety, and translational readiness of these platforms have not been fully synthesised. Objectives This study assessed preclinical and early clinical evidence of CAR-T and CAR-NK immunotherapies against HBV and HIV, including efficacy, safety and translational feasibility. Methods Databases were searched according to PRISMA 2020 guidelines. For this review, eligible studies included in vitro , in vivo , and clinical studies examining virus-directed CAR lymphocytes. Random-effects models generated pooled standardised mean differences (SMD) and risk ratios (RR). The level of specific evidence was considered by GRADE. Results Forty-three studies met the inclusion criteria (21 in vitro , 14 in vivo , and 8 clinical). Preclinical HIV CAR-T models demonstrated significant reductions in HIV p24 antigen levels (pooled SMD = −1.15, 95% CI −1.50 to −0.80). Similarly, HBV-directed engineered T-cell studies showed a marked decrease in HBsAg and HBV DNA (SMD = −1.30, 95% CI −1.70 to −0.90). CAR-NK platforms displayed comparable antiviral activity with potentially improved safety profiles. In vivo analyses also indicated consistent suppression of HIV RNA (SMD = −0.92, 95% CI −1.26 to −0.58) and moderate reductions in HBV DNA levels (SMD = −1.05, 95% CI −1.52 to −0.63). In early-phase clinical studies (phase I/II), HIV-directed CAR-T therapies produced modest decreases in circulating HIV RNA (SMD = −0.35, 95% CI −0.60 to −0.12), while HBV-targeted therapies demonstrated small but detectable antiviral responses (SMD = −0.42, 95% CI −0.78 to −0.11). Across clinical cohorts, the incidence of cytokine release syndrome (CRS) remained low, occurring in fewer than 10% of treated patients. Conclusion Virus-directed CAR-T and CAR-NK therapies show strong preclinical antiviral activity and early clinical signs of activity, showing acceptable safety. Because of heterogeneity, small sample size and limited clinical data, the quality of evidence from this population remains low to moderate. Large and well-controlled trials are necessary to optimise CAR designs, improve persistence, and investigate combinations with latency-reversing or immune-modulating drugs.