Building Antigen-Specific Immune Tolerance For Autoimmune Disease

By Erin Harris, Editor-In-Chief, Cell & Gene
Follow Me On Twitter @ErinHarris_1

Autoimmune disease has become one of the most consequential frontiers in the cell and gene therapy field, and it is gaining increasing attention across the broader advanced therapies ecosystem. Oncology remains central to the field, yet developers are increasingly pursuing approaches that may do more than suppress disease activity. The goal is to selectively retrain, retune, or reprogram the immune system.

That is why antigen-specific tolerance is an important area to watch. The aim is to intervene as precisely as possible at the source of autoimmune dysfunction, ideally avoiding the broad immunosuppression associated with many long-term treatments.

Realizing that potential will require sound science, better diagnostics, meaningful clinical endpoints, and a thoughtful approach to data and artificial intelligence. I reached out to Adam Elhofy, PhD, Chief Scientific Officer at COUR Pharmaceuticals, to better understand the company’s nanoparticle-based tolerance platform. Dr. Elhofy and I also discussed which diseases may be best suited to the approach and the clinical development questions facing the field.

From Coupled Cell Therapy to an Off-the-Shelf Platform

COUR’s platform emerged from decades of immune tolerance research, according to Dr. Elhofy. The technology traces back to the work of Stephen Miller, PhD, an immunologist and professor of microbiology and immunology at Northwestern University Feinberg School of Medicine. Dr. Miller sought to bring a coupled cell therapy approach into the clinic, including through a multiple sclerosis (MS) trial initiated in 2011.

Prospective partners consistently told the team they wanted a more scalable and straightforward therapeutic format, as well as an off-the-shelf product. The result was a particle-based platform designed to deliver the essential components needed to induce antigen-specific tolerance. “We are the first that were talking about the fact that you need a tolerogenic signal and an antigen, and you need to be delivering to the right place,” said Dr. Elhofy.

While that concept was not always widely embraced, the language surrounding autoimmune drug development has evolved quickly. Today, immune reset, immune retuning, and immune reprogramming are common terms, reflecting a growing interest in therapies that address disease-driving immune responses more selectively than broad immunosuppressive approaches.

Scientific progress, clinical activity, and patient need have converged. The question is no longer whether advanced therapies can move beyond cancer, but rather how developers will demonstrate durable and clinically meaningful benefit in chronic autoimmune disease.

Selecting Diseases Where Biology Is Clear

For COUR, the initial questions are fundamental. Is there an identified antigen? Is the disease driven by T cells? “There is anywhere from 150 to 200 identified autoimmune diseases, and 50 or so with clear T cell epitopes that have been identified,” said Dr. Elhofy. “We don’t have any shortage of diseases to go after right now.”

Not every autoimmune disease is equally ready for an antigen-specific intervention. In some diseases, the identity or relevance of the disease-driving antigen remains under debate. Dr. Elhofy pointed to Crohn’s disease as one example. Bacterial antigens have been identified by multiple laboratories, but questions remain about their role in driving disease.

Rheumatoid arthritis (RA) is another example where the understanding of relevant antigens and disease targets continues to evolve. In these settings, COUR is monitoring the science and awaiting better antigen identification tools. According to Dr. Elhofy, those tools have improved meaningfully in recent years, which could broaden the number of diseases that can ultimately be addressed.

Other diseases have more established antigen biology, including Type 1 diabetes (T1D) and primary biliary cholangitis (PBC). This distinction matters because precision approaches depend on a well-characterized biological target. When that target remains uncertain, a highly specific treatment may be premature. When the target is clear, selectively redirecting immune activity becomes much more realistic.

Measuring Benefit Beyond Disease Control

Manufacturing scale-up is a common concern across the CGT field, but for COUR, Dr. Elhofy said nanoparticle platform scale-up is not the central challenge. Defining and proving clinical benefit is more difficult. “The issue with clinical benefit is where we do have more difficulty,” said Dr. Elhofy.

COUR has an ongoing trial in myasthenia gravis (MG), where conventional measures of clinical improvement are relatively well understood. Even so, the definition of meaningful benefit continues to evolve.

Approved therapies have raised expectations, meaning it may not be enough for a new treatment to improve symptoms alone. Developers may also need to demonstrate steroid tapering, deeper responses, or benefits for additional patient populations.

PBC presents a different endpoint challenge. Pruritus and fatigue can have a significant impact on patients, yet recruiting patients based on severe pruritus can be difficult. Patient-reported outcome tools for the disease are also still developing.

Longer-term outcomes could include liver decompensation, liver transplantation, or liver failure, but trials designed around these endpoints may need to run for three to five years. The availability of approved therapies can also make placebo-controlled studies more challenging.

Dr. Elhofy believes the field needs accessible biomarkers that can provide an early signal that a therapy is working. He compared the need to something as simple and familiar as a blood pressure cuff. “If we can find something like that that will tell us long term that we’re doing something clinically beneficial, then that will help,” said Dr. Elhofy.

T1D may offer an opportunity to develop these types of measures. Continuous glucose monitoring data are widely available, while time in tight glycemic range and reduced insulin use provide measurable outcomes that are directly relevant to patients.

COUR is now focused on T1D and is exploring potential collaborations with continuous glucose monitor manufacturers. The goal is to determine whether existing real-world data can help identify meaningful patterns over time.

Earlier Diagnosis Could Change the Therapeutic Equation

Diagnosis is another critical issue. Many autoimmune diseases are diagnosed through exclusion, and by the time a diagnosis is confirmed, significant disease progression may have already occurred.

Dr. Elhofy used MS as an example. A patient may experience symptoms for several years before receiving a definitive diagnosis as providers work to rule out other possible explanations. That delay may shrink the treatment window. Earlier intervention with an immune tolerance therapy could potentially have a greater impact on the course of disease. “If you find someone that in PBC has had disease for 12 years, then you’re talking about stopping progression. You’re not talking about reversing disease,” said Dr. Elhofy. “If you caught them earlier, then you’re talking about reversing disease.”

The comparison to oncology is useful. Earlier detection can shift the goal from managing advanced disease to preventing or slowing progression.

Autoimmune medicine will need improved diagnostics, validated biomarkers, and clearer disease interception strategies. These capabilities may be essential to fully realizing the potential of antigen-specific tolerance.

AI Needs Validation, Not Blind Trust

Dr. Elhofy also offered a pragmatic perspective on AI. He sees significant potential for using AI to analyze complex and longitudinal datasets, but he stressed that AI-generated outputs must be rigorously validated. “It’s all of those things,” said Dr. Elhofy, referring to whether AI is beneficial, problematic, or something in between. “The problem is that almost all of them make so many mistakes unless they’re trained on your specific features.”

COUR is exploring bespoke AI tools built around the company’s internal data. Dr. Elhofy described an instance in which an apparently compelling data cluster could not be confirmed through several alternative methods. He also described using repeated AI-assisted checks to identify errors in a published placebo curve, including issues involving age matching, units, and the timing of treatment initiation.

AI can accelerate analysis, generate hypotheses, and help teams examine complex datasets. It cannot replace scientific judgment, appropriate controls, or rigorous validation.

That discipline will be essential as autoimmune disease development becomes increasingly data driven. The field is entering an era in which patient experience, digital measures, biomarkers, real-world data, and therapeutic innovation will need to work together rather than exist in separate silos.

If developers can make each of these elements more precise while keeping patients at the center of development, immune tolerance could become a meaningful therapeutic option for people living with chronic autoimmune disease. The opportunity is significant, but so is the challenge of proving that selectively retraining the immune system can deliver durable, clinically meaningful results.