Antibody–drug conjugates (ADCs) are a rapidly advancing class of targeted cancer therapeutics that couple the antigen specificity of monoclonal antibodies (mAbs) with the potent cytotoxicity of small-molecule drugs. In their core design, a tumor-targeting antibody is covalently linked to a cytotoxic payload via a chemical linker, enabling the selective delivery of highly potent agents to malignant cells while sparing normal tissues, thereby improving the therapeutic index. Humanized and fully human immunoglobulin G1(IgG1) antibodies are the most common ADC backbones due to their stability in systemic circulation, robust Fcγ receptor engagement for immune effector functions, and reduced immunogenicity. Antibody selection requires balancing tumor specificity, internalization rate, and binding affinity to avoid barriers to tissue penetration, such as the binding-site barrier effect, while emerging designs exploit tumor-specific antigen variants or unique post-translational modifications to further enhance selectivity. Advances in antibody engineering, linker chemistry, and payload innovation have reinforced the clinical success of ADCs, with more than a dozen agents FDA approved for hematologic malignancies and solid tumors and over 200 in active clinical trials. This review critically examines established and emerging conjugation strategies, including lysine- and cysteine-based chemistries, enzymatic tagging, glycan remodeling, non-canonical amino acid incorporation, and affinity peptide-mediated methods, and discusses how conjugation site, drug-to-antibody ratio (DAR) control, and linker stability influence pharmacokinetics, efficacy, and safety. Innovations in site-specific conjugation have improved ADC homogeneity, stability, and clinical predictability, though challenges in large-scale manufacturing and regulatory harmonization remain. Furthermore, novel ADC architectures such as bispecific ADCs, conditionally active (probody) ADCs, immune-stimulating ADCs, protein-degrader ADCs, and dual-payload designs are being developed to address tumor heterogeneity, drug resistance, and off-target toxicity. By integrating mechanistic insights, preclinical and clinical data, and recent technological advances, this work highlights current progress and future directions for next-generation ADCs aimed at achieving superior efficacy, safety, and patient outcomes, especially in treating refractory cancers.

Conventional immunotherapy, including immune checkpoint blockade and chimeric antigen receptor (CAR)-T cells, has revolutionized cancer therapy over the past decade. Yet, the efficacy of these therapies is limited by tumor resistance, antigen escape mechanisms, poor persistence, and T-cell exhaustion, particularly in the treatment of solid tumors. The emergence of unconventional immunotherapies offers novel opportunities by leveraging diverse immune cell subsets and synthetic biologics. This review explores various immunotherapy platforms, including gamma delta T cells, invariant natural killer T cells, mucosal-associated invariant T cells, engineered regulatory T cells, and universal CAR platforms. Additionally, it expands on biologics, including bispecific and multispecific antibodies, cytokine fusions, agonists, and oncolytic viruses, showcasing their potential for modular engineering and off-the-shelf applicability. Distinct features of unconventional platforms include independence from the major histocompatibility complex (MHC), tissue-homing capabilities, stress ligand sensing, and the ability to bridge adaptive and innate immunity. Their compatibility with engineering approaches highlights their potential as scalable, efficient, and cost-effective therapies. To overcome translational challenges such as functional heterogeneity, immune exhaustion, tumor microenvironment-mediated suppression, and limited persistence, novel strategies will be discussed, including metabolic and epigenetic reprogramming, immune cloaking, gene editing, and the utilization of artificial intelligence for patient stratification. Ultimately, unconventional immunotherapies extend the therapeutic horizon of cancer immunotherapy by breaking barriers in solid tumor treatment and increasing accessibility. Continued investments in research for mechanistic insights and scalable manufacturing are key to unlocking their full clinical potential.

Chimeric antigen receptor (CAR) T-cell therapy has transformed the treatment of haematological malignancies, but conventional viral transduction results in semi-random genomic integration, contributing to heterogeneous CAR expression and potential insertional effects. Site-specific genome engineering offers an alternative by directing CAR insertion to defined genomic loci, allowing greater control over transgene expression and cellular function. This narrative review evaluates advances in site-specific CAR T-cell engineering, focusing on integration mechanisms, donor platforms, genomic loci, safety, and translational readiness. Homology-directed repair (HDR) and homology-independent targeted integration (HITI) are critically compared alongside viral and non-viral donor systems. Candidate loci are evaluated according to their functional consequences and level of evidence, with TRAC representing the most extensively characterised target, while PDCD1, CD7, CD247, and other loci offer distinct functional opportunities but remain supported by varying levels of preclinical and translational evidence. The review also examines genomic risks associated with targeted editing, including structural rearrangements and chromosome-scale abnormalities, and considers emerging applications in allogeneic, solid-tumour, and in vivo CAR T-cell engineering. Overall, site-specific integration provides a framework for linking CAR placement to therapeutic function, but no single locus or integration strategy is universally optimal. Clinical evidence remains limited relative to the expanding preclinical landscape, and genomic safety, scalable manufacturing, and clinical validation remain major priorities for translation.

Targeted protein degradation has transformed drug discovery by replacing occupancy-driven inhibition with event-driven removal of disease-causing proteins, opening previously “undruggable” targets, including transcription factors, scaffolding proteins, and intrinsically disordered regions, to therapeutic intervention. Within this landscape, peptide-based molecular glues have emerged as a distinct and underexplored modality, combining the extended, adaptable interaction surfaces of peptides with the mechanistic simplicity of induced-proximity pharmacology. Unlike bifunctional degraders such as PROTACs, which require the concurrent optimization of two linked ligands, peptide molecular glues act through a single engineered interface, stabilizing cooperative ternary complexes between a target protein and an E3 ubiquitin ligase to drive ubiquitination and proteasomal degradation. This review highlights the current understanding of peptide molecular glue biology across five interconnected dimensions: the mechanistic basis of E3 ligase recruitment and cooperative ternary complex formation; rational design strategies spanning sequence selection, structural stabilization, and computational and AI-guided optimization; the structural and biophysical determinants, such as conformational flexibility, interface complementarity, and thermodynamic and kinetic stability, that distinguish productive from non-productive complexes; and emerging therapeutic applications in oncology, hematological malignancy, neurodegeneration, and inflammatory disease, several of which have already reached clinical evaluation. We further examine the comparative advantages of peptide-based architectures, including reduced molecular weight, access to challenging interfaces, and expansion of the addressable E3 ligase repertoire, against persistent translational barriers in proteolytic stability, membrane permeability, immunogenicity, and manufacturability. Taken together, the evidence positions peptide-based molecular glues not as incremental variants of existing degrader platforms, but as a mechanistically distinct therapeutic class capable of substantially widening the druggable proteome.