The CD160-HVEM axis at the crossroads of tolerance and immunity: lessons from cancer for autoimmunity and transplantation.

Background: The clinical success of immune checkpoint blockade (ICB) in oncology has validated the existence of molecular brakes that restrain T-cell immunity. While releasing these brakes unleashes antitumour responses, their physiological role is to maintain self-tolerance. This review explores the reciprocal paradigm of the CD160-HVEM axis across three disease settings: immuno-oncology, autoimmunity and transplantation. CD160-HVEM signalling can enforce pathological T-cell exhaustion in cancer, yet its failure or costimulatory co-option can precipitate autoimmunity or accelerate allograft rejection. Mechanisms: Unlike programmed cell death protein 1 (PD-1), CD160 operates within a bidirectional network involving herpesvirus entry mediator (HVEM; TNFRSF14), B- and T-lymphocyte attenuator (BTLA) and LIGHT (TNFSF14). The axis is expressed across lymphocytes, myeloid cells, epithelium, endothelium and malignant B cells. Mouse and human biology of the axis are not interchangeable: murine CD160 is a broad major histocompatibility complex (MHC) class I and CD1d receptor, whereas human natural killer (NK) triggering is dominated by HLA-C, and the transmembrane isoform CD160-TM is human-characterised. Structural diversity, from an inhibitory glycosylphosphatidylinositol (GPI)-anchored form to activating CD160-TM on NK cells, positions CD160 as a tunable rheostat rather than a linear checkpoint. Therapeutic implications: While direct anti-CD160 candidates remain in advanced preclinical development, other components of the HVEM network are undergoing human clinical testing. The most mature clinical programme is tifcemalimab, an anti-BTLA monoclonal antibody, not an anti-CD160 agent. Antagonistic anti-HVEM antibodies (Demerlé anti-HVEM18-10 in humanised mice; preclinical fully human 4CB1, mechanistically related but distinct) block HVEM engagement by both BTLA and CD160. Because CD160 is inhibitory on T cells but activating on NK cells via CD160-TM, a systemic antibody to the shared ectodomain is a dual-function drug-design constraint, not a simple checkpoint blockade. Conclusion: Precision use of this axis requires isoform, cell type, interaction geometry and disease context to be specified before a direction of intervention is chosen. Next steps are biomarker-qualified enrichment, isoform-discriminating reagents, and correlative CD160 profiling inside ongoing BTLA-HVEM studies. Pathological exhaustion that shields tumours remains a map for restoring tolerance in autoimmunity and for blocking CD28-independent rejection in transplantation.