The Right Key, the Wrong Lock: TIGIT Checkpoint Blockade and the Road to Precision Immunotherapy

Abstract

T-cell immunoreceptor with immunoglobulin and immunoreceptor tyrosine-based inhibitory motif (ITIM) domains (TIGIT) emerged as one of the most promising next-generation immune checkpoint targets following the success of programmed cell death protein 1 (PD-1), programmed death-ligand 1 (PD-L1), and cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) blockade. TIGIT suppresses antitumor immunity through interaction with cluster of differentiation 155 (CD155), inhibition of CD226-mediated co-stimulation, and promotion of immunosuppressive regulatory T-cell (Treg) activity within the tumor microenvironment (TME). Strong preclinical evidence demonstrated that TIGIT blockade, particularly in combination with PD-1/PD-L1 inhibition, restored T-cell and natural killer (NK) cell function and produced durable antitumor responses in multiple tumor models, leading to rapid clinical development. Despite this compelling biological rationale, most late-stage clinical programs failed to reproduce early success. Although the phase II CITYSCAPE trial showed encouraging activity in PD-L1-high non-small cell lung cancer (NSCLC), subsequent phase III trials, including SKYSCRAPER-01, SKYSCRAPER-02, SKYSCRAPER-03, SKYSCRAPER-14, AdvanTIG-302, KEYVIBE, and STAR-221, failed to improve survival outcomes or meet primary endpoints. The notable exception was SKYSCRAPER-08 in esophageal squamous cell carcinoma, suggesting that TIGIT blockade may be effective only in selected biological contexts. This review critically examines the molecular biology of the TIGIT–CD155–CD226 axis, its role in immune regulation and tumor immune evasion, and the preclinical and clinical evidence supporting TIGIT-targeted therapy. Particular emphasis is placed on understanding the causes of clinical failure, including CD226 loss during T-cell exhaustion, checkpoint network redundancy, Fc-engineering uncertainty, immunosuppressive TMEs, inadequate biomarker-guided patient selection, and tumor-type-specific dependence on the TIGIT pathway. We also present original bioinformatics analyses demonstrating that broader checkpoint network signatures outperform TIGIT expression alone for patient stratification. Finally, we evaluate emerging solutions including biomarker-guided precision immunotherapy, Fc-optimized antibodies, bispecific checkpoint inhibitors, TIGIT-engineered chimeric antigen receptor T-cell (CAR-T) cells, radiotherapy combinations, and multi-checkpoint blockade. Collectively, current evidence suggests that the future of TIGIT-directed therapy lies not in universal checkpoint inhibition but in biologically informed, precision-guided immunotherapy strategies.