What Is Immunogenicity and Why Does It Matter?

Immunogenicity is the ability of a substance to provoke an immune response in the body. In medicine, the term cuts two ways: vaccine developers want high immunogenicity so the body builds lasting protection, while makers of therapeutic drugs generally want low immunogenicity so the body does not attack the treatment itself. Understanding where immunogenicity helps and where it hurts is central to modern drug development, gene therapy, and the growing field of personalized cancer vaccines.

When the Immune System Fights the Medicine

Most conversations about immunogenicity in drug development focus on an unwanted problem: the body treating a therapeutic protein as foreign. Biologic drugs, which include monoclonal antibodies, enzymes, and hormones made from living cells, are large, complex molecules. Unlike a simple chemical pill, their size and structure can trigger the immune system to produce what are called anti-drug antibodies (ADAs). These antibodies latch onto the drug and can neutralize it, speed up its removal from the bloodstream, or, in rare cases, cause dangerous allergic or autoimmune reactions.

The consequences range from mild to severe. On the mild end, ADAs may quietly reduce how long the drug stays in circulation, meaning a patient needs higher or more frequent doses. On the severe end, ADAs can completely neutralize the drug’s effect, leaving the patient without a working treatment. Determining whether any particular antibody response actually matters clinically depends on whether those antibodies measurably change the drug’s behavior in the body, its effectiveness, or its safety profile.¹

How Anti-Drug Antibodies Clear Drugs From the Body

When ADAs bind to a therapeutic protein, they form clusters called immune complexes. The body’s cleanup systems recognize these clusters and remove them from circulation, much the way they would clear debris from an infection. Research has shown that this clearance depends on the size and structure of the complexes: larger clusters get swept away faster. In preclinical studies, total drug concentration dropped by roughly 20 to 60 percent within minutes when the drug was bound up in immune complexes, and drugs with intact immune-signaling regions on their structure were cleared even faster than modified versions.² Clinical data from patients who developed ADAs confirmed this pattern, with larger immune complexes disappearing from the blood more quickly than smaller ones.³

This matters practically because standard blood tests that measure total drug levels can be misleading. If a patient’s drug is bound up in immune complexes, the test may show adequate drug concentrations, but much of that drug is effectively neutralized and unable to do its job. Specialized assays that distinguish free, active drug from antibody-bound drug give a more accurate picture of what is really happening.

What Makes a Biologic Drug More or Less Immunogenic

Several features of the drug molecule itself influence how likely it is to trigger an immune response. Two of the most studied are protein aggregation and glycosylation.

Protein aggregation happens when individual drug molecules clump together during manufacturing, shipping, or storage. These clumps are a well-recognized risk factor for immunogenicity. Aggregated proteins can shift the type of immune response they provoke: in mouse studies, subvisible aggregates (too small to see but still present) of two different proteins triggered a different branch of the immune system compared to the same proteins in their normal, non-clumped form. The aggregates pushed the immune system toward a response associated with stronger, more inflammatory antibody production.⁴ Controlling aggregation during manufacturing is therefore a major focus in the pharmaceutical industry, and regulatory agencies require companies to screen for it.⁵

Glycosylation refers to sugar molecules attached to the surface of the drug protein. These sugar patterns vary depending on what type of cell is used to produce the drug. When a biologic is manufactured in non-human cell lines, the resulting sugar patterns can differ from what the human immune system considers “self,” potentially triggering hypersensitivity reactions or antibody formation. The high variability in glycosylation across manufacturing batches adds another layer of complexity, since it is difficult to keep these sugar patterns perfectly consistent.⁶

The Patient Side of the Equation

The drug’s molecular features are only part of the story. The patient’s own biology plays a major role in whether ADAs develop.

Genetics is the strongest patient-level predictor. A person’s immune system presents fragments of foreign proteins to immune cells using molecules encoded by a group of genes called the human leukocyte antigen (HLA) system. Different people carry different versions of these genes, and some versions are better at presenting drug-derived fragments, making an immune response more likely. In inflammatory bowel disease patients treated with biologic drugs, carrying a specific HLA variant called HLA-DQA1*05 was repeatedly associated with higher rates of antibody formation, lower drug levels in the blood, and treatment failure.⁷ In cancer patients treated with the checkpoint inhibitor atezolizumab, a different allele, HLA-DRB1*01:01, was the strongest genetic predictor of ADA development, nearly doubling the odds.⁸

Other patient factors include disease state, immune suppression status, and prior exposure to similar biologics. Patients on immunosuppressive medications alongside their biologic drug tend to develop ADAs at lower rates, which is one reason many treatment regimens combine biologics with drugs like methotrexate.

Does the Route of Injection Matter?

A longstanding assumption held that injecting a drug under the skin (subcutaneously) was more likely to provoke an immune response than delivering it directly into a vein (intravenously). The skin is dense with immune cells, so the logic seemed intuitive. But a meta-analysis covering 17 therapeutic proteins and 48 treatment groups found no statistically significant difference in ADA rates between the two routes.⁹ The only variable that significantly predicted ADA rates in that analysis was which protein was being given, not how it was injected. Preclinical data in non-human primates reached the same conclusion: the route of administration is not a systematic risk factor, and immunogenicity depends more on the drug itself and the patient than on where the needle goes.¹⁰

A Cautionary Tale From Erythropoietin

One of the most dramatic examples of unwanted immunogenicity involved erythropoietin (EPO), a hormone that stimulates red blood cell production. Recombinant versions of EPO have been used for decades to treat anemia in patients with chronic kidney disease. Before 1998, antibody-mediated reactions to EPO were extremely rare. Then, from 1999 to 2002, clinicians documented a sharp increase in cases of pure red cell aplasia, a condition where the body stops making red blood cells almost entirely.¹¹

The problem was traced to anti-EPO antibodies. In affected patients, these antibodies did not just neutralize the injected drug; they also blocked the body’s own natural erythropoietin. Testing confirmed that serum from patients with PRCA blocked red blood cell colony formation in laboratory bone marrow samples, and that this inhibition was reversed by adding EPO back.¹² The episode served as a wake-up call for the entire industry, demonstrating that immunogenicity against a drug could, in rare cases, knock out a patient’s own critical biology. Changes to EPO formulations and manufacturing eventually brought the incidence back down, but the incident permanently raised the bar for immunogenicity monitoring.

When Immunogenicity Is the Goal

Vaccines are the mirror image of the ADA problem. The entire point of a vaccine is to provoke a strong, targeted immune response. Here, maximizing immunogenicity is the design objective.

Many vaccines struggle to be immunogenic enough on their own, which is why adjuvants are added. Adjuvants are substances that amplify the immune response to a vaccine’s target. They work by activating the innate immune system, the body’s first line of defense, through receptors that sense patterns associated with infection or tissue damage. This innate activation in turn ramps up the adaptive immune response, the part that produces antibodies and memory cells for long-term protection.¹³ Advances in understanding how these innate sensing receptors work have opened the door to more precisely designed adjuvants, including ones that activate specific immune pathways tailored to a given pathogen.¹⁴

Personalized Cancer Vaccines

One of the most exciting frontiers for designed immunogenicity is personalized cancer vaccines. These vaccines are built around neoantigens, mutated proteins unique to a patient’s tumor. The idea is to train the patient’s immune system to recognize and attack cells carrying those mutations. Early results are encouraging. In a first-in-human trial, a personalized mRNA neoantigen vaccine called EVM16 triggered strong immune responses against the intended targets in eight of nine patients, with the strength of the response increasing at higher doses.¹⁵ Preclinical work has also shown that an AI-guided platform can identify promising neoantigens and package them into multi-target mRNA vaccines that retain antigen-specific immunogenicity and show antitumor activity in mice.¹⁶

Predicting and Measuring the Immune Response

Given the stakes, both drug developers and clinicians need reliable ways to predict and detect immunogenicity.

On the prediction side, one of the key tools used during drug development is the MAPPs assay, which maps out which fragments of a drug protein are likely to be presented to the immune system by HLA molecules. This helps developers identify risky regions in a protein before it ever reaches patients, and it can guide the selection of drug candidates with lower immunogenicity risk.¹⁷

On the detection side, once a drug is in patients, blood tests are used to monitor for ADAs. The older standard was a conventional ELISA assay, but these tests can be blinded by the presence of the drug itself in the blood sample: if drug levels are high, the drug molecules compete with the test reagent and mask the antibodies. Newer electrochemiluminescence (ECL) assays are far more tolerant of drug interference. In a head-to-head comparison for one biologic drug, vedolizumab, the ECL assay tolerated drug concentrations roughly 100 times higher than the ELISA before losing sensitivity.¹⁸ Validated ECL-based assays for other biologics, such as infliximab, have demonstrated high sensitivity and precision for clinical patient testing.¹⁹ The same platform has also been adapted to detect ADAs against smaller therapeutic peptides, broadening its usefulness across the biologic drug landscape.²⁰

Engineering Drugs to Fly Under the Immune Radar

If you know which parts of a protein drug trigger immune recognition, you can try to redesign those parts. This approach, broadly called deimmunization, has become a sophisticated field in its own right.

The most familiar example is the humanization of antibody drugs. Early therapeutic antibodies were derived entirely from mice, and the human immune system recognized them readily as foreign. Replacing the mouse structural regions with human sequences while keeping the drug’s targeting region intact produced a far less immunogenic product.²¹ Even fully humanized antibodies retain some residual immunogenicity, though, because the drug’s unique targeting regions can still contain sequences the immune system flags. Targeted amino acid modifications in those regions can further reduce the risk while preserving the drug’s function.

Computational tools have pushed this further. Algorithms now search for the minimal set of conservative mutations that will eliminate predicted immune-triggering sequences from a protein without wrecking its structure or activity.²² More recently, a platform integrated into the Rosetta protein design suite allows engineers to explore large sets of possible redesigns simultaneously, scoring them for both low immunogenicity and physical stability.²³ The result is a growing library of methods for designing biologics that keep their therapeutic punch while reducing the chance the patient’s immune system will interfere.

Biosimilar Drugs and the Immunogenicity Question

Biosimilars, the biologic drug equivalents of generics, face a unique immunogenicity challenge. Because biologics are made by living cells, no two manufacturing processes produce exactly identical molecules. A biosimilar must demonstrate that any differences from the original drug do not translate into clinically meaningful differences in immune response. Head-to-head immunogenicity comparison between a biosimilar and its reference product has therefore been a required part of regulatory approval.²⁴

That regulatory landscape is shifting, though. After two decades of accumulating clinical data, agencies are moving toward streamlined biosimilar development that relies more heavily on analytical and pharmacokinetic similarity, without always requiring large comparative efficacy trials. The reasoning is that well-designed pharmacokinetic studies, especially in healthy volunteers, can provide sensitive immunogenicity assessments while avoiding the confounding factors that come with studying sick patients on multiple medications.²⁵ For patients, this could mean faster access to more affordable versions of expensive biologics.

Gene Therapy and Pre-Existing Immunity

Gene therapies that use viral vectors to deliver corrective genes face a different immunogenicity problem than traditional biologics. Adeno-associated virus (AAV) vectors, among the most widely used delivery vehicles, are derived from viruses that commonly infect humans without causing disease. The catch is that many people have already been exposed to wild-type AAV, which means they carry pre-existing antibodies against the viral shell. Those antibodies can prevent the therapeutic vector from ever reaching its target cells.²⁶

The immune challenges do not stop at entry. Once the vector delivers its cargo and cells start producing the new protein, the immune system can mount responses against both the viral capsid and the new protein itself. Capsid-specific killer T cells can destroy the transduced cells, erasing the therapeutic effect. Patients receiving high systemic doses of AAV vectors for conditions like spinal muscular atrophy and Duchenne muscular dystrophy have experienced immune-mediated side effects ranging from liver inflammation to more severe complications. Screening patients for pre-existing AAV antibodies before treatment, and developing strategies to blunt or redirect the immune response, are active areas of research in the field.

The Gut Microbiome and Vaccine Responses

An emerging and somewhat surprising factor in immunogenicity is the community of bacteria living in the gut. A systematic review found that the composition of the intestinal microbiome can influence responses to both oral and injected vaccines, affecting the strength of both antibody and cellular immune responses. Certain bacterial groups, particularly Bifidobacterium, were associated with better vaccine responses, while others, especially Gammaproteobacteria, appeared to dampen them.²⁷

This connection has been observed specifically with COVID-19 vaccines as well. In a study of people receiving the BNT162b2 mRNA vaccine, baseline gut microbial composition appeared to influence the strength of the immune response, including in people living with HIV.²⁸ The practical implications are still taking shape, but the finding raises the possibility that interventions targeting the gut, whether through diet, probiotics, or other means, could someday be used to enhance vaccine effectiveness, especially in populations that tend to respond poorly. It also reinforces how immunogenicity is never purely about the molecule being injected; the whole biological context of the person receiving it matters.²⁹