Peptide Immunogenicity: Why Antidrug Antibodies Matter

Antidrug antibodies (ADAs) are host-produced immunoglobulins that specifically recognize a therapeutic biologic—peptide, protein, or conjugate—administered to the host. ADA formation is the endpoint of the standard adaptive immune response: antigen-presenting cells (macrophages, dendritic cells, B cells) internalize the peptide, process it into shorter fragments, present those fragments on major histocompatibility complex (MHC) class II molecules, and activate CD4+ T-helper cells carrying T-cell receptors complementary to the peptide-MHC complex. The activated T-helper cells provide cytokine signals (IL-4, IL-21) that license B cells to undergo class switching, affinity maturation, and differentiation into antibody-secreting plasma cells [1].

The threshold for ADA formation depends on whether the peptide provides both B-cell epitopes (short conformational or linear sequences recognized by B-cell receptors) and T-cell epitopes (linear fragments presented on MHC class II). Very short peptides (fewer than about 9 amino acids) often fail to provide T-cell epitopes and are therefore poorly immunogenic unless conjugated to carrier proteins. Peptides in the 20-50 residue range can act as haptens: insufficient to independently activate T cells, but immunogenic when covalently linked to albumin or immunoglobulins in vivo.

Host tolerance mechanisms normally prevent antibody formation against endogenous peptides and proteins. Central tolerance eliminates T and B cells with receptors strongly reactive to self during thymic and bone-marrow development; peripheral tolerance anergizes or deletes self-reactive lymphocytes that escape central tolerance. Therapeutic peptides structurally identical to endogenous sequences generally escape ADA formation—unless aggregation, formulation excipients, or contaminants create neoepitopes that tolerance mechanisms never learned to ignore.

A 2025 systematic review published on PubMed Central aggregated safety data from pivotal clinical trials of therapeutic peptides approved between 2010 and 2024. The review analyzed trials covering GLP-1 receptor agonists (semaglutide, liraglutide, tirzepatide, dulaglutide), parathyroid hormone analogs (teriparatide, abaloparatide), calcitonin analogs, oxytocin analogs (carbetocin), and vasopressin analogs [2].

Across the included trials, the review reported ADA-positive rates varying from under 1% (semaglutide in the SUSTAIN program) to over 50% (teriparatide in pivotal osteoporosis trials). The substantial variation reflects differences in peptide sequence similarity to endogenous counterparts, formulation and delivery route, dose interval, and sensitivity of the assay methods used. The high teriparatide ADA rate is largely clinically silent because the antibodies are non-neutralizing and do not measurably alter bone mineral density response [3].

The review separated ADA incidence from clinically meaningful consequences: treatment-emergent ADAs that correlate with loss of efficacy, altered pharmacokinetics (accelerated clearance), or infusion reactions. Using this framing, clinically consequential ADA formation was reported in under 5% of subjects across most peptide classes, with serious adverse events directly attributable to ADA-mediated mechanisms occurring in well under 1% of exposed subjects.

The 2025 review reported that across aggregated peptide therapeutic trials, serious adverse event (SAE) rates attributable to the investigational peptide were consistently below 3%, with treatment-related deaths well under 0.5%. SAE definitions follow ICH E2A guidance: events that are fatal, life-threatening, require hospitalization or prolong existing hospitalization, cause persistent disability, or constitute a congenital anomaly [4].

The under-3% figure should be read with several caveats. First, clinical trial populations are selected for absence of contraindications and typically exclude subjects with active malignancy, severe renal or hepatic impairment, or significant cardiovascular disease. Real-world populations include these comorbidities and generally experience higher adverse event rates. Second, SAE attribution in open-label or placebo-controlled trials is investigator-judged and subject to reporting bias. Third, rare SAEs (thyroid C-cell tumors with GLP-1 agonists, anti-drug-antibody-mediated cross-reactive endogenous hormone suppression) require post-marketing surveillance to detect.

The clinical trial safety record cannot be extrapolated to research-grade or compounded peptides of uncertain provenance. The <3% SAE rate applies specifically to GMP-manufactured, regulator-reviewed products administered under medical supervision with defined dose regimens. Research-grade peptide exposure outside these conditions has higher documented adverse event rates, as discussed in our companion article on pancreatitis risk.

Individual susceptibility to ADA formation is substantially determined by human leukocyte antigen (HLA) class II alleles, which encode the MHC class II molecules that present peptide fragments to T-helper cells. Each HLA class II molecule has a characteristic binding groove that accommodates peptide fragments with specific anchor residues; different HLA alleles thus present different subsets of any given peptide sequence. A patient carrying HLA alleles whose grooves bind fragments of the therapeutic peptide strongly is immunologically predisposed to form T-helper responses and, downstream, ADAs [5].

Pharmacogenomic studies of factor VIII replacement in hemophilia A, interferon-beta in multiple sclerosis, and more recently of therapeutic peptides have identified specific HLA alleles associated with higher ADA rates. For example, HLA-DRB1*15:01 has been associated with ADA formation against interferon-beta. Similar associations are emerging for therapeutic peptides, though peptide-specific HLA associations are still being catalogued due to the relatively small clinical trial populations available for each peptide.

Pre-treatment HLA screening is not standard clinical practice for most peptide therapeutics, both because of cost and because the association strength is usually modest (odds ratios 1.5-3x rather than deterministic). However, in-silico immunogenicity prediction tools (NetMHCIIpan, EpiMatrix) now allow peptide designers to scan candidate sequences against a population-representative HLA allele panel and identify subsequences likely to be strong T-cell epitopes in the target population. Iterative design can then replace problematic residues with immunosilent alternatives that preserve pharmacological activity.

Delivery route substantially modulates ADA formation for identical peptide sequences. Subcutaneous administration—the default for most chronic peptide therapies—is generally associated with the highest ADA rates because subcutaneous tissue contains resident dendritic cells (Langerhans cells, interstitial DCs) that efficiently sample, process, and present injected antigens to draining lymph node T cells [6].

Intravenous administration distributes peptide throughout the circulation before immune tissue uptake, reducing the local antigen concentration available to any single dendritic cell and generally producing lower ADA rates at comparable total dose. However, IV administration of protein aggregates or endotoxin-contaminated material can produce acute infusion reactions (cytokine release) independent of ADA formation. Oral administration (as in oral semaglutide) is associated with low ADA formation both because bioavailability is limited and because gut-associated lymphoid tissue (GALT) has active tolerance-promoting mechanisms (regulatory T cell induction, IgA class switching) that dampen systemic IgG responses.

Inhaled and intranasal routes produce mixed results: local mucosal immune responses can be prominent, and aerosolization can induce peptide aggregation. Implantable and microneedle delivery systems are being evaluated for sustained peptide delivery with potentially lower immunogenicity due to the reduced peak local concentration and co-delivery of tolerizing signals. Formulation-level interventions—PEGylation, albumin fusion, or lipid modification to extend half-life—generally reduce dose frequency and, modestly, reduce ADA rates compared to native peptides dosed more frequently.

Not all ADAs are clinically consequential. ADA assays typically distinguish binding antibodies (which recognize the peptide but do not interfere with its pharmacological activity) from neutralizing antibodies (NAbs, which block the peptide from binding its target receptor or engaging its mechanism of action). Binding ADAs are much more common than NAbs and are often clinically silent, though they can accelerate peptide clearance through immune-complex formation and Fc-receptor-mediated uptake by hepatic and splenic macrophages [7].

Neutralizing antibodies are the high-consequence subset. A NAb that blocks semaglutide from binding the GLP-1 receptor functionally replicates treatment discontinuation; dose escalation cannot overcome sufficiently high NAb titers. More problematically, NAbs against therapeutic peptides that share sequence with endogenous peptides can cross-react with the endogenous ligand, potentially causing deficiency of the natural hormone. This mechanism has been observed for erythropoietin (pure red cell aplasia from anti-EPO NAbs) and is a theoretical concern for GLP-1 analogs and for recombinant calcitonin.

The clinical interpretation of ADA results therefore requires parallel NAb testing whenever binding ADA is detected. Regulatory guidance (FDA 2019 guidance on immunogenicity testing) specifies a tiered approach: screening assay for binding antibodies, confirmatory assay to rule out false positives, titer characterization, and neutralization assay for confirmed positive samples.

ADA detection relies on bridging immunoassays, where the peptide antigen is coated on a solid phase (ELISA) or biotinylated and ruthenium-labeled (ECL immunoassay). Patient serum is incubated with the antigen; ADAs form bridges between labeled and capture antigens, producing signal proportional to ADA concentration. ELISA and electrochemiluminescence (ECL, Meso Scale Discovery platform) are the two dominant formats, with ECL generally preferred for higher sensitivity and better tolerance of drug interference [8].

Drug interference is a central analytical challenge: patients on active therapy have circulating peptide that competes with the assay antigen for ADA binding, producing false negatives. Acid dissociation protocols (brief exposure to low pH to release ADAs from circulating peptide before neutralizing pH for the bridging assay) and anti-idiotype calibrators are standard mitigations. Assay validation must demonstrate drug tolerance at concentrations relevant to the clinical dosing window.

Neutralizing antibody assays use cell-based (receptor-reporter systems) or competitive ligand-binding formats. Cell-based NAb assays are more physiologically relevant but more technically demanding; ligand-binding formats are simpler but may miss NAbs that block functional signaling through epitopes distant from the ligand-binding site. Reporting conventions include titer (dilution at which signal returns to baseline), cut-point-adjusted positivity, and where possible, absolute NAb concentration derived from calibration curves.

Peptide engineering can reduce immunogenicity at design time. Key strategies include: (1) Deimmunization by substituting residues in predicted T-cell epitopes with immunosilent alternatives, guided by in-silico MHC class II binding predictors; (2) PEGylation or lipidation, which shield the peptide from antigen-presenting cells while extending half-life; (3) Incorporation of D-amino acids or N-methylated residues, which resist proteolytic processing and therefore reduce presentation on MHC class II; (4) Humanization for peptides derived from non-human species, replacing non-human residues with human counterparts while preserving the pharmacophore [9].

• Deimmunization: substitute predicted T-cell epitope residues
• PEGylation/lipidation: shield from antigen-presenting cells
• D-amino acids and N-methylation: reduce MHC presentation
• Humanization: replace non-human residues
• Aggregation control: formulation to prevent multimerization
• Endotoxin control: purification to <0.5 EU/mg for clinical material

Formulation-level controls are equally important. Peptide aggregation is a major driver of immunogenicity because aggregated peptides display repetitive epitopes that efficiently cross-link B-cell receptors. Buffer selection, pH, ionic strength, surfactant addition (polysorbates), and lyoprotectant choice (trehalose, sucrose, mannitol) all influence aggregation. Stability studies during development include sub-visible particle counting (micro-flow imaging, light obscuration) to track aggregation under stress conditions.

Explore detailed chemical profiles and research guides for compounds discussed in this article:

• Semaglutide: Complete Research Guide → /learn/semaglutide
• Tirzepatide: Complete Research Guide → /learn/tirzepatide
• Liraglutide: Complete Research Guide → /learn/liraglutide

• Read more: Peptide Allergic Reactions → https://www.chemverify.com/learn/peptide-allergic-reactions-researchers-guide
• Read more: Endotoxin Testing for Peptides → https://www.chemverify.com/learn/endotoxin-testing-for-peptides-essential-safety-protocols-for-research
• Read more: Peptide Pancreatitis Risk → https://www.chemverify.com/learn/peptide-pancreatitis-cases-unregulated-sources