Traditional chemotherapy, which relies on chemical agents, often comes with significant side effects. This treatment works by administering toxic substances that inhibit cell division, targeting rapidly dividing cells. However, this approach not only destroys cancer cells but also damages fast-growing healthy cells.

Recently, immunotherapy, which enhances the function of a patient’s immune cells to selectively attack cancer cells, has been gaining attention. But how does it specifically target only cancer cells?

Our immune cells naturally develop self-tolerance, a mechanism that prevents immune responses against normal cells. By disrupting the self-tolerance of immune cells surrounding tumors, immunotherapy directs the body’s defense system to attack only cancer cells.

These immunotherapies have shown effectiveness against various cancer types and serve as a foundation for combination therapy, which integrates multiple treatment approaches for enhanced efficacy.

How Are Personalized Cancer Vaccines Developed?

Personalized cancer vaccines target mutated antigens* present in cancer cells to trigger an immune response. Therefore, identifying the right antigen to target is the key to their development. The process involves several crucial steps.

*Antigen: A foreign substance that triggers a response from the immune system. Antigens can take various forms, including bacteria, viruses, and cancer cells.

First, cancerous tissue from the patient is obtained through a biopsy or surgery. The DNA from the tumor is then analyzed using sequencing technology to decode its genetic information. This process helps identify neoantigens, which are newly formed proteins resulting from cancer cell mutations. Among these, researchers look for mutated sequences that can activate the immune system.

Based on the identified neoantigens, messenger RNA (mRNA) is designed. To ensure safe and effective delivery, lipid nanoparticles (LNPs) or other lipid-based carriers are used to manufacture personalized cancer vaccines, which are then administered to the patient. Simply put, this process identifies highly immunogenic mutated sequences within the patient’s tumor, enabling a precise and targeted immune attack that selectively destroys cancer cells.

Once the vaccine is administered, the patient’s cells read the genetic information and produce proteins corresponding to the selected neoantigens. These proteins are then presented to the immune system via the Major Histocompatibility Complex (MHC) / Human Leukocyte Antigen (HLA). By exposing the neoantigens to immune cells, particularly T cells, the vaccine triggers an immune response, leading to the targeted destruction of cancer cells.

Thus, since personalized cancer vaccines can only be developed after securing a patient’s tumor tissue, reducing the manufacturing time is a critical challenge in maximizing the vaccine’s effectiveness.

Accelerating Vaccine Development with AI

The key to shortening vaccine production time lies in AI-driven neoantigen prediction models. The currently developing neoantigen prediction model utilizes AI analysis to identify neoantigens that can be used in therapeutic applications.

Even within the same type of cancer, genetic mutations vary significantly, and there are over 40,000 known variants of patient-specific Major Histocompatibility Complex (MHC). This makes it extremely challenging to determine whether each individual neoantigen can effectively trigger an immune response. However, by leveraging computer algorithms and machine learning, AI can rapidly identify neoantigens most likely to induce an immune response.

To better understand this process, let’s explore the mechanism of immune cell activation using neoantigens.

Neoantigens are processed into peptide (protein chain) forms containing tumor-specific mutations. These peptides are then presented within the Major Histocompatibility Complex (MHC) of Antigen-Presenting Cells (APCs), which capture antigens and present them to T cells. For effective capturing, the specific sequence of the neoantigen peptide must bind to the patient’s unique MHC. Once this binding occurs, the peptide is recognized by T cell receptors (TCRs), triggering an immune response to attack the cancer cells.

In summary, for immune activation against neoantigens, the recognition process must involve the MHC (HLA-TCR), the neoantigen peptide, and the T cell receptor (TCR) working in synergy.

As shown above, the immune activation of neoantigens depends on the binding affinity between the neoantigen peptide and the Major Histocompatibility Complex (MHC), as well as how well the resulting complex is recognized by T cell receptors (TCRs), which is the key factor. Our prediction model is designed to analyze these interactions by learning from a database of peptide binding affinities and immune activation. This enables the model to rapidly predict the immunogenicity of neoantigen candidates and their binding strength.

Ultimately, our goal is to quickly identify the most effective neoantigens for each patient’s immune response and apply them to develop truly personalized cancer vaccines.

Advancing Cancer Treatment with Data Technology

Personalized cancer vaccines are set to redefine the paradigm of cancer research. Their safety and efficacy have already been validated in multiple clinical trials, and leading mRNA technology pioneers such as Moderna and BioNTech are actively advancing research in this field. These vaccines have demonstrated effectiveness against specific cancers like melanoma, and they are expected to be applied to a broader range of cancers, including lung, colorectal, and pancreatic cancer, in the future.

With the rapid advancement of artificial intelligence and big data, neoantigen prediction models are also evolving. In particular, the adoption of deep learning technology, which learns patterns from data, and cloud computing, which processes large-scale datasets, is enabling faster and more accurate predictions. As a result, these advancements are establishing themselves as core technologies for personalized treatment.

As mRNA (messenger RNA) technology and AI-driven prediction models continue to advance, treatment efficiency and accessibility will improve, ultimately benefiting more patients. We hope that our AI-based neoantigen prediction model will serve as a breakthrough in overcoming the limitations of existing treatments and pave the way for a new future in cancer therapy.

Written by Seihwan Jeong (Professional, Cancer Vaccine Team, Drug Discovery), LG Chem.