Medicine, Technology and the End of Cancer

The following text is a translation of the article "Medicine, Technology and the End of Cancer" by BioNTech founders Dr. Uğur Şahin and Dr. Özlem Türeci, published in The New York Times on December 3, 2023.

*

Artificial intelligence may hold the key to unlocking personalized cancer vaccines.

Decades ago, as young doctors on a cancer ward at a university hospital in Germany, we often had to deliver a heartbreaking message to our patients: "We have no treatment options left to offer you." At the time, chemotherapy and radiation therapy were our main weapons against advanced cancer. Although sometimes effective, they often fell short of the cure that so many hoped for.

Today, the landscape of cancer treatment has changed profoundly. We have witnessed the emergence of new drugs that target tumor growth with great precision. Some cancers once considered untreatable can now be controlled, even at advanced stages. However, for most patients with metastatic cancer, the promise of a highly effective therapy remains an unattainable dream.

So why does a cure remain the exception for patients with advanced cancer, despite billions of dollars being invested in cancer research each year?

The answer lies in the nature of cancer . Over time, cancer arises from random genetic changes in healthy cells, known as mutations . Mutations can be triggered by lifestyle choices, inherited predispositions, existing long-term health conditions, or exposure to hazardous chemicals. As these changes accumulate, mutated cells turn into cancerous entities.

This randomness has two consequences:

• First, each cancer is as unique as the individual it affects, meaning that even among people diagnosed with the same type of cancer, only a subset shares common mutations.
• Second, each tumor is a complex tapestry of billions of different cells, constantly learning to adapt, evade the immune system , and resist the treatment strategies we throw at it.

But could we harness another personal, continually learning force to defeat these elusive enemies: our own immune system?

The body's defense system consists of billions of cells with remarkable capabilities. Among them are nature's vigilant sentinels, constantly patrolling our bodies: T cells . However, evolution has programmed our immune system to fight external threats, viruses and bacteria, rather than mutations arising from within. Consequently, only a small fraction of mutations attract the attention of our immune defenses.

Now imagine a future with on-demand personalized cancer vaccines . These would be custom-designed to precisely communicate each patient's unique cancer mutations, acting like a "Wanted" poster that instructs immune cells to launch multipronged attacks against the tumor.

What was only a vision at our patients' bedsides 30 years ago is now the focus of intensive clinical trials conducted by us and others.

Accumulated advances in science and technology across various fields have provided a tailwind.

(An explanation of how the cancer vaccine being developed at BioNTech works)

Decoding the genetic makeup of each patient's cancer from a small biopsy, at high resolution and within hours, has become possible by reading DNA with next-generation sequencing technologies . Newly available, widespread computing power helps us process the enormous amounts of data this sequencing generates. We apply advanced data analyses, and artificial intelligence (AI) algorithms help us identify the mutations we consider most relevant. These mutations then become the basis of the "Wanted" poster delivered to cells through messenger RNA (mRNA) in the vaccine. RNA, nature's oldest messenger, is our preferred vaccine platform. It can be rapidly designed, customized, and manufactured within a few weeks. After all, time is of the essence for cancer patients.

Personalized cancer vaccine candidates are currently undergoing rigorous testing, and access is limited to controlled clinical trials. In small numbers of patients, these trials have shown that it is possible to activate and expand T cells that recognize tumor cells based on selected mutations—a vital prerequisite for the immune system to fight cancer. Melanoma and pancreatic cancer trials have recently shown the potential benefit of personalized mRNA vaccines in reducing the risk of cancer recurrence after surgery. Researchers are also conducting larger clinical trials comparing these personalized vaccines with current standards of care in certain cancers. Data from these trials in the coming years will provide information about the safety and efficacy of personalized cancer vaccines under development.

Advances in technology and science often occur in parallel silos, but when they converge, the results can be groundbreaking. The combination of mRNA and artificial intelligence exemplifies such a convergence and could herald a new era in medicine, laying the groundwork for transformative, patient-centered medical solutions.

We believe AI will continue to play an increasingly important role in developing personalized cancer vaccines and medicines. AI algorithms can rapidly analyze large genomic datasets and identify patterns and correlations that traditional methods might miss. This rapid, precise analysis is particularly useful when identifying relevant cancer mutations among the many genetic variations in a patient's tumor. As global cancer databases expand, predictions for each patient's mutation selection and vaccine design will likely keep improving.

The molecular properties of highly versatile mRNA enable rapid, scalable production in miniaturized, parallelized, automated manufacturing facilities. Parallelized manufacturing breaks up the traditional assembly line, allowing different products to be produced simultaneously. For us, these characteristics of mRNA-based manufacturing lay the foundation for truly personalized medicines to become widely available one day, enabling scientific knowledge enhanced by AI to be put into practice promptly. Combining these technologies could allow rapid adjustments to personalized mRNA vaccines as a patient's tumor mutation profile evolves.

In medical research, we believe the marriage of technologies such as AI and mRNA is ushering in a transformative era comparable to Moore's Law for microchips.

Just as Moore's Law predicted exponential growth in computing power, we are now witnessing a similar acceleration that could address urgent, unmet medical needs. But there is a difference: the pace at which new knowledge emerges and technological barriers are overcome is outstripping our ability to develop and approve new therapies within the traditional regulatory and procedural systems of existing drug development, clinical practice, and care paradigms.

Scientists, governments, public health sectors, and societies are therefore being called on to prevent a growing gap between what could be possible with current knowledge and technological advances and the treatment options offered to patients.

With all this in mind, how far are we from a medical future with personalized cancer vaccines?

It is still early. Like any new treatment, personalized cancer vaccine candidates must pass through the stages of clinical development and demonstrate superior efficacy compared with existing therapies. We expect the first approval and adoption of such therapies for selected cancers by 2030. The following decade could bring a technology-driven transformation that makes personalized therapies widely available and affordable.

Cancer is a deeply personal disease, and it is time for its treatment to be personal too.