In 1891, an American oncologist, William Coley, had a eureka moment. He realised when some cancer patients developed certain bacterial infections, their cancer went into remission – and started injecting patients’ tumours with bacteria to treat them. His work became the basis of what we now know as immunotherapy.
After effectively disappearing for most of the 20th century, overshadowed by the rise of radiotherapy and chemotherapy, immunotherapy has come roaring back in recent decades. Training patients’ immune systems to fight off cancer has become a frontline treatment against many forms of the disease. However, there’s a problem. While immunotherapy can be highly effective against certain types of blood cancers, it’s far less effective against solid tumours with dense structure that hard to penetrate – and unfortunately, solid tumours account for the vast majority of global cancer deaths.
Seeing what humans could not: discovery through AI
The problem of this huge unmet clinical need is not a lack of immune firepower, but visibility. Traditional research methods rely on classic cell markers and human‑defined assumptions, which often cause rare or unconventional immune cells to be overlooked. At The Chinese University of Hong Kong (CUHK), a team led by Professor Patrick Tang Ming-kuen from the Department of Anatomical and Cellular Pathology in the Faculty of Medicine took a fundamentally different approach. They applied advanced AI models to analyse millions of single cells from a vast array of solid tumours, including lung, liver, pancreas and more. This unbiased data‑first strategy unexpectedly revealed a previously unreported type of T‑cell – one that consistently appeared across different solid tumours but had remained invisible to conventional analysis.
“By using single-cell RNA-sequencing, we can view the genetic signature of every individual cell, rather than just a broad average,” says Professor Tang. “When we paired this high-definition data with unbiased, AI-driven analysis, we removed the human lens entirely. The AI serves as an objective witness, identifying immune cell states based purely on data rather than preconceived expectations.”
Super T-cell with built‑in GPS and battle gear
What they’ve found opens up whole new vistas of opportunity where treating cancer is concerned. Our bodies, it turns out, produce a type of T-cell – white blood cells that are vital components of our immune systems – with some very special characteristics. Dubbed “Super” T-cells by the team, not only can they pinpoint precisely where solid tumours are, but also wipe them out, even inside solid tumours, which are particularly challenging places for immune cells to flourish.
“Our research unexpectedly revealed that ‘Super’ T-cells use a sophisticated sensing mechanism driven by tissue inflammation signatures,” says Dr Max Chan, Postdoctoral Fellow from the Department of Anatomical and Cellular Pathology. “While conventional T-cells are often excluded from the dense, inflamed environment of a solid tumour, ‘Super’ T-cells are programmed to recognise and thrive within these conditions. By leveraging these tissue-sensing capabilities, they have the brave to overcome the physical and biological barriers that typically render standard immunotherapies ineffective.”
As a result of their inflammation sensors, the cells it has discovered are kind on the body: they precisely target tumours without causing any damage to healthy tissue. This means any medication made from these cells won’t cause what’s known as a cytokine storm – a dangerous flood of normally harmless immune signalling molecules at lethal concentrations, which has been responsible for many COVID-19 death, for example.
From discovery to scale: turning cells into a therapy
Discovery of “Super” T-cells is just the kick start of an innovation of immunotherapy – they’ve also worked out how to produce lots of them from patient’s own blood, in just a handful of days. It holds out the hope that their discovery can achieve a safe and effective immunotherapy for solid tumours, with the potential to help millions.
To speed the process, the team has set up a company, CELLmeric, which has its own cell engineering platform. It avoids the usual approach, where a modified virus is used as a way of delivering genetic material. By doing so, says Professor Tang, “we achieve a significantly higher level of clinical safety and avoid the complexities of viral vector manufacturing. This allows us to produce high-quality, stable and safe therapeutic cells at a more affordable price, making life-saving cell therapy accessible to a much broader patient population.”
That therapy’s effectiveness will depend on the inflammatory landscape a particular type of cancer. So far, it appears particularly potent against some of the big beasts of the solid tumour world, such as the leading cancer types: lung, pancreas, and liver. The team is currently collaborating with institutions in Chinese mainland and Australia to better determine the range of cancers that can benefit most and move forward to clinical trials.
“While clinical translation is a multi-year endeavour involving rigorous regulatory milestones, we are moving as efficiently as possible toward delivering a transformative treatment for patients.” says Professor Tang.
This journey is already well underway, supported by prestigious recognition, including the Excellence Award at the National Disruptive Technology Innovation Competition, Gold Medal at the Asia Exhibition of Innovations and Inventions, Silver Medal at the International Exhibition of Inventions Geneva, and direct funding from the Ministry of Science and Technology. Crucially, the project has been empowered by the State Key Laboratory of Translational Oncology (CUHK), the Innovation and Technology Support Programme (Mid-stream, theme-based) from the Innovation and Technology Commission of the Government of the Hong Kong Special Administrative Region of the People’s Republic of China, General Research Fund from Research Grant Council and the Faculty’s Passion for Perfection Scheme.
