The clearest historical case of a peptide therapy that raced from triumph to disappointment is the 2001-2002 clinical program for the β-amyloid vaccine AN-1792. Within months of the first Phase-I report, the popular press was calling it “the Alzheimer’s vaccine” and investors poured hundreds of millions into follow-on peptide immunotherapies. The peptide itself—full-length, synthetic Aβ42—was chosen because it is the self-aggregating species found in plaques; the hope was that active immunisation would mobilise antibodies that tag those plaques for microglial clearance. Early open-label data looked spectacular: patients generated high anti-Aβ titres, CSF tau fell, and cognitive decline appeared to slow. Yet in January 2002 the multinational Phase-IIa trial was halted after 6 % of recipients developed meningo-encephalitis; eighteen patients were hospitalised and one died. Autopsies showed T-cell infiltration not only around plaques but also around unaffected vessel walls, confirming that the same peptide epitope that elicited therapeutic antibodies had also activated cytotoxic T-cells against self. The programme was formally abandoned and the sponsor (Elan) lost 70 % of its market capitalisation in a week.
What is striking is how quickly the field metabolised the catastrophe into a “dosage-form problem” rather than a conceptual one. Within two years the scientific literature had pivoted to passive antibodies and N-terminal Aβ fragments that avoided the Th1-biased epitopes, and by 2006 new trials were again being heralded as breakthroughs. The lesson—that a peptide identical to a human protein can be simultaneously therapeutic and auto-immunogenic—was filed away as an idiosyncratic formulation mis-step rather than a general warning about breaking immune tolerance. As a result, the same red flags have re-appeared in subsequent peptide-active immunisation programmes (tau, α-synuclein, prion), each time greeted with surprise when neuro-inflammation surfaces.
Why was the deeper warning ignored? First, the economic structure of peptide drug discovery rewards rapid differentiation, not historical introspection. Peptide drug discovery and development notes that nearly 90 % of new peptide candidates originate inside industry, where “fast-fail” portfolios are designed to monetise incremental tweaks (humanisation, lipidation, depot formulations) rather than to interrogate first-principle failures. Second, the scientific narrative was comforted by pharmacology texts such as Handbook of Biologically Active Peptides, which framed the metabolic instability of peptides as the dominant liability; immunogenicity was treated as a solvable “delivery issue.” Third, the success of chronic insulin, calcitonin and GLP-1 analogues created an anchoring bias that peptides are intrinsically safe because they are “natural.” Finally, the explosion of high-throughput peptide synthesis (Peptides: Chemistry and Biology) made it technically trivial to generate new epitope variants, so the temptation to iterate rather than to halt was irresistible.
The most counter-intuitive finding unearthed is that the very feature that made Aβ42 attractive—its perfect sequence identity to the human protein—was the root of its toxicity. Unlike foreign peptides (e.g., enfuvirtide) or heavily modified analogues (e.g., cyclosporine), the vaccine peptide was processed and presented by every patient’s own antigen-presenting cells, guaranteeing that high-affinity T-cells would be recruited once antibody titres rose and immune complexes formed. In other words, “naturalness” was not a safety feature; it was the hazard.
Critical gaps remain. None of the books quantify how many of the 400 peptide candidates now in the pipeline are self-derived epitopes, so the field has no denominator for how often the AN-1792 scenario could recur. There is also unresolved disagreement on whether low-grade sub-clinical neuro-inflammation occurred in the patients who lacked overt encephalitis; post-mortem series are small and no imaging biomarker was collected prospectively. Finally, regulators and sponsors have not converged on a pre-clinical assay that reliably predicts epitope-specific T-cell activation for human self-peptides, meaning each new programme still enters the clinic with empirical immunogenicity risk.
References
- EDR Peptide Possible Mechanism of Gene Expression and — Khavinson
- Vladimir
- Ending Aging The Rejuvenation Breakthroughs That Could — Aubrey D N J De Grey
- Handbook of Biologically Active Peptides
- Peptide Protocols Volume One — William A Seeds MD
- Peptide drug discovery and development _ Translational — edited by Miguel Castanho and
- Peptides_ Chemistry and Biology, 2nd Edition
- Protein Folding Handbook, 5 Volume Set — Johannes Buchner
- Thomas Kiefhaber
- The Mind-Gut Connection How the Astonishing Dialogue Taking — Mayer
- Emeran A