The dominance of peptides for GLP-1 therapeutics isn't a failure of antibodies but a success for picking the right tool. Peptides have a natural advantage when the therapeutic strategy involves engineering a natural ligand, making them a better starting point for certain targets like GPCRs.
In the real world, the selection of a therapeutic modality like an antibody or peptide is often driven by a company's existing expertise and technology platform rather than a purely agnostic approach to finding the single best tool for a clinical problem. Organizations default to the tools in their toolbox.
While GLP-1 has been a known target for a long time, the recent explosion in peptide therapeutics was primarily enabled by solving the historical challenge of poor half-life and exposure. Achieving one- or two-week half-lives through techniques like fatty acid acylation was the critical technological unlock for the field.
The debate isn't about peptides replacing antibodies but about combining them. The future lies in hybrid therapeutics, such as grafting peptides into antibody CDRs or creating fusions that use a peptide for optimal target binding and an antibody scaffold for effector functions, half-life extension, and stability.
Despite the clear potential of hybrid peptide-antibody drugs, their development is slow. This is attributed to human nature in science: researchers tend to stick with familiar, comfortable modalities and the tools available in their specific lab or company, creating a barrier to cross-disciplinary innovation.
The primary hurdle for the entire biologics field is enhancing the therapeutic index (efficacy vs. toxicity). Because most conditions like cancer and autoimmune disorders are 'diseases of self,' therapeutics often have on-target, off-tumor effects. This fundamental problem drives the need for innovations like masking and conditional activation.
A key barrier to complex peptide-antibody drugs is manufacturing (CMC). Current methods require separate synthesis and conjugation steps. A fully genetically encoded system—where the entire hybrid molecule is produced in a single cell line—would dramatically lower the barrier to entry and simplify manufacturing, unlocking new drug designs.
As biologics evolve into complex multi-specific and hybrid formats, the number of design parameters (valency, linkers, geometry) becomes too vast for experimental testing. AI and computational design are becoming essential not to replace scientists, but to judiciously sample the enormous design space and guide engineering efforts.