As cancer treatment options multiply and pressure on the NHS intensifies, a new approach is emerging that replaces trial-and-error therapy selection with precision and speed. According to Jens Kelm, Co-Founder and CSO of PreComb, the convergence of Functional Precision Oncology and Artificial Intelligence is enabling clinicians to identify the most effective cancer treatments faster, reduce unnecessary toxicity and deliver more personalised, cost-effective care.
Each year in the UK, more than 100,000 patients begin chemotherapy. Any technological advance that helps select the right therapy faster has the potential not only to save lives, but also to significantly ease the pressure on an already over-stretched NHS. Yet despite enormous strides in analytics, diagnostics and genomics, oncologists have traditionally been forced to rely on a trial-and-error approach when selecting cancer treatments. Even today, most patients move through several lines of therapy before a drug or combination proves effective, exposing them to unnecessary toxicity, delaying meaningful clinical benefit and adding substantial costs to the healthcare system.
The challenge becomes clear when considering the scale and complexity of available treatment options. In England alone, more than 117 cancer drugs are in active use via the Cancer Drugs Fund and most patients require at least two drugs in combination. Every tumour behaves differently, even among those that share similar molecular features. Advances in genomics represented a major leap forward by allowing oncologists to identify targetable mutations, but only a small minority of patients – around 10% – present with such mutations. For the remaining 90%, selecting an effective therapy has largely remained a process of informed guesswork. This is costly not only in economic terms but also in terms of patient wellbeing, as each unsuccessful treatment forces patients to endure toxic side effects without therapeutic benefit, while tumours continue to grow and spread.
A new paradigm is now emerging, one that moves cancer care decisively away from trial and error. This transformation is being driven by the convergence of Functional Precision Oncology (FPO) and Artificial Intelligence (AI). Functional precision testing allows scientists to grow microtumours from a patient’s own cancer tissue and expose them to dozens of drugs or drug combinations. Rather than administering treatments blindly, clinicians can observe how tumour cells respond and use these insights to guide therapy selection. This approach relies not on theoretical predictions but on the direct, measurable behaviour of the tumour.
The real breakthrough comes when this functional data is combined with AI. Artificial intelligence enables the rapid analysis of vast volumes of drug-response data, imaging and multi-omic profiles generated during testing. By integrating functional results with genomic, proteomic and clinical information, AI creates a detailed portrait of each patient’s tumour, allowing more precise treatment decisions. Companies such as PreComb are developing platforms that automate and scale these processes, bringing personalised drug-response testing into routine clinical practice.
Crucially, the relationship between AI and functional testing is reciprocal. Functional testing provides AI with high-quality biological data, while AI accelerates analysis and improves reproducibility. Together, they create a virtuous cycle of continuous improvement.
The combined impact is already visible. Therapy decisions that once took weeks can now be informed within a fortnight, reducing ineffective treatments, improving outcomes and lowering costs. At a time when NHS cancer spending is rising by around 9% annually, the potential benefits are substantial.
The convergence of AI and functional testing offers a credible, evidence-driven path towards faster, more effective and more affordable cancer care. Few applications of AI are as powerful as those that save and extend human life.
