Rethinking oncology: zooming out on the world’s largest oncology conference

‘We cured the mouse,’ a prominent oncologist once joked as I interviewed him about some of his research.

It was a humorous reminder that what works in tightly controlled lab settings doesn’t necessarily make an impact in real patients.

That phrase came back to me this year as I followed the ASCO 2025 conference from afar. Every year, the meeting of the American Society of Clinical Oncology brings together tens of thousands of oncology professionals, showcasing cutting-edge science across a wide range of cancer types. This year’s theme, Driving Knowledge to Action: Building a Better Future, was all about turning scientific research into better outcomes for people with cancer.

I haven’t attended ASCO for a few years. When I did, it was as a communications professional supporting my clients (mostly pharmaceutical companies), or as a video journalist. This meant long, hectic days focusing on granular details, interviewing experts, and keeping an eye on news coverage.

There was very little time to interact with the science, and take it all in. I always wanted to take a step back, and look at the bigger picture painted by all that activity and all that energy. The deluge of information at the conference made the task seem Herculean, practically impossible.

This year, something changed. This year, I had AI.

I prompted Gemini to scan and synthesise the research presented at the conference: abstracts, presentations, commentary, and coverage by media and scientific platforms such as Medscape and OncLive.

I am conscious that AI has its limitations and biases, and that media coverage is heavily influenced by pharma’s PR activities. I tried to control for inaccuracies and biases, but I am not under the illusion that the summary I gathered is complete and true to the real-life event. Having said that, I am not looking for accuracy in detail. I am looking for patterns and themes, for broad brushstrokes rather than individual pixels.

ASCO being such a large and influential platform, I saw it as a proxy for what the field of oncology is doing at present. What does it focus on? What is the oncology community collectively producing?

The bird’s (A) eye view

A rigorous review process by the conference’s Scientific Programme Committee ensures that the research presented is of high scientific quality, relevant, and aligned with the conference’s theme: From knowledge to action: building a better future.

Through this process, 5,000 cancer research publications (abstracts) and presentations were selected this year. These covered advances in drug development, treatment regimens and interventions, and diagnostics for a wide spectrum of cancer types. Many of these studies were identified by experts as having the potential to significantly shift current clinical care paradigms.

From my AI-enabled scan of this rich body of research, three themes emerged:

1. Increasingly advanced, sophisticated treatments

As always, ASCO 2025 was a showcase for advanced, increasingly sophisticated solutions to cancer. Some notable examples were: antibody-drug conjugates — molecules that deliver toxic chemotherapy specifically to cancer cells; checkpoint inhibitors — molecules that reawaken the immune system to attack cancer; CAR-T — re-engineered immune cells that are trained to kill the patient’s very own cancer. Other innovations included powerful molecular diagnostics that detect minute amounts of tumour by-products in the blood; and AI-powered tools and algorithms to decode the ever-increasing complexity of data collected.

As a former scientist, I find these advances very exciting. There is no denying the intelligence, dedication and commitment of the oncology community in their quest to improve outcomes for people with cancer.

2. Deeper personalisation with precision medicine

Precision medicine aims for more precise, effective, and less toxic cancer treatments that can be tailored to each individual. It has been a dominant paradigm in oncology for quite some time, and this was reflected at ASCO this year too.

Precision medicine means targeting a person’s cancer based on its particular molecular characteristics. These are often genetic characteristics like a mutation, but can be any biological characteristic that can be reliably detected in blood, body fluids, tumours, or tissues.

These characteristics can be leveraged to develop tailored drugs, to determine if a person is likely to benefit from a particular treatment, or to measure their response to a given treatment.

At ASCO this year (and in previous years), numerous studies showcased the application of precision medicines in the treatment of a wide variety of cancer types, and also in guiding treatment decisions. AI tools were discussed in this context, supporting clinical trial design, enabling the monitoring of molecular responses to treatment, and assisting stratified therapy decisions.

3. The quest for more refined treatment regimens

The third theme I noted was the quest to optimise current treatments through the design of better, more efficacious treatment regimens. Examples include classic strategies like treatment combinations: combinations of immunotherapies, combinations of immunotherapies with chemotherapy, or cocktails of targeted therapies with chemotherapy. Other strategies included more novel approaches powered by AI or biomarkers (precision medicine again), as well as strategies that tailored therapies based on individual risk and response to treatment. The concept of de-escalation studies was a highlight, looking at the possibility of reducing the intensity or duration of treatment without compromising patient survival.

Now that I have drawn the broad brushstrokes of the bigger picture, I am going to try to answer my questions: What does this bigger picture tell us about the state of oncology today? What is this system trying to do? What is it doing instead?

The systemic view

1. Scientific excellence within a narrow paradigm

The dominant logic is clear: cancer is a molecular disease that needs to be addressed with molecular tools. Identify the molecular drivers of cancer and intervene with ever more targeted, more sophisticated tools. Therapies drug-antibody conjugates, checkpoint inhibitors, and AI-assisted matching are all part of this paradigm.

From a systems perspective, this reflects a deeply entrenched reductionist approach. The focus is at the level of parts: genes, pathways, proteins, even though it is well established that cancer is the result of the interaction of multiple factors at multiple levels: the cell, the tumour, the tumour’s environment, the patient and the environment in which they live.

Even non-molecular intervention studies are pulled back into this paradigm. The CHALLENGE trial showed that a structured exercise programme can reduce the risk of cancer and enhance survival in people with colon cancer. This is great, but can you guess what the researchers plan to do next? They plan to explore how exercise reduces cancer recurrence by studying blood samples of the patients who participated in the trial.

There is nothing wrong with that. I trained as a molecular biologist, and I would probably do the same if I were in their shoes. But my point is: the instinct to reduce the outcome back to blood markers shows how deeply ingrained the molecular mindset has become.

I don’t want to take away from the rigour and relevance of molecular studies. In fact, I applaud them. Reductionist thinking has led to spectacular advances in cancer care, but if the oncology field relies too heavily on it, it might end up stuck in a reinforcing feedback loop: the more we learn about molecules, the more narrowly we focus on molecules. The more narrowly we focus, the more importance we give to solving small, well-defined problems.

Such focus comes at a cost. We start to make decisions within a narrow context. We begin to tweak treatment combinations, to play with doses and timings. We stratify patients into smaller and smaller groups. We forget about the bigger picture — that those genes and proteins and rogue cancer cells exist within a complex setting that continuously evolves and adapts.

“We spend too much time tinkering with the numbers, when we should be thinking about the goals, rules, and mindsets of the system.” — Donella Meadows

2. Alignment with the pharmaceutical industry

What follows from the molecular-centric mindset is a heavy focus on molecular targeting products and pharmaceutical interventions.

I have worked with the pharmaceutical industry for over 12 years. I have witnessed first-hand how the people working in pharma care about making a real difference to people’s lives, so this is not a dig at pharma, or at anybody else for that matter.

But there is no denying that the pharma model is biased. It is a business model, not a healthcare model. It favours what can be measured and quantified (biomarkers, endpoints), what can be managed (clinical trials with limited duration and clean endpoints), what can be protected (intellectual property), and importantly, what can be monetised (ROI).

This creates another reinforcing loop in which research that fits inside molecular targeting, IP frameworks, and financial structures, is rewarded by more attention and more funding. As it is rewarded, it creates the incentive for more research within the mould, which gets further attention and further funding.

In parallel, the same loop suppresses the emergence of alternative paradigms like environmental drivers, or community-led care, not because they don’t work, but because they don’t fit in the dominant paradigms.

Another point to think about is the cost of adopting such medicines. The drug development process depends on advanced infrastructure, expensive biologics, and high-cost monitoring tools. All things that many health systems around the world cannot afford.

It’s really important to celebrate scientific achievements and innovation, but an equally important part of the conversation is about how to embed this innovation in real life, to produce tangible outcomes in the context of existing healthcare systems. Few talks addressed implementation science at the conference, or how to close the gap between what we know and what we do. Considering the theme of the conference was ‘From knowledge to action’, I think, those conversations should have been more central, more prominent and embedded in every session.

3. The illusion of precision medicine

Precision medicine is an elegant idea. You understand the molecular profile of a cancer, you design a drug that targets that profile, and the drug attacks the cancer cells while sparing healthy tissue. The result is a more effective treatment with fewer side effects.

But this model has a major limitation: resistance to treatment.

A tumour is a highly complex environment, and one that continues to change and evolve. When treated with a drug that targets one molecule, one node in that complex web of interactions, tumours almost always find ways to overcome it, and the drug stops working. So a new drug needs to be developed to address this resistance. But then, sooner or later, a new type of resistance appears (I’ve written about this before). This creates a continuous need for new drugs, which then create new forms of resistance, and so on, until the cancer becomes too advanced for treatment.

In systems thinking, this paradigm is called ‘Fixes that fail’. It describes situations where a quick fix is applied to address a problem’s symptom, and relieves the symptom. This gives a feeling of progress and resolution, but shifts attention away from the root causes of the problem. Unintended side effects (such as drug resistance) result from this approach and worsen over time, eventually undermining any long-term solutions.

What’s important to notice is not whether precision medicine works. It does. But when the lens is tightly focused on molecules and mutations, what is overlooked? What is missing?

Perhaps what is missing is another perspective. Instead of looking at resistance as treatment failure, what would happen if it were viewed as feedback instead? What if the language changed? Instead of talking about endpoints, which indicate the final stages of a process, how about thinking in terms of feedback in a cycle of learning and adaptation? Instead of thinking in a linear chain of cause and effect, how about embracing the complexity and messiness of a tumour?

4. There are cracks in everything. That’s how the light gets in.

Still, some of the research presented at ASCO this year hints at a broader, systems mindset.

Immunotherapy is one. The immune system is complex, capable of infinite variety and adaptation. Just like the tumour. Immune therapies that aim to work with the body’s immune system to produce a response against tumours hold a lot of promise. This can only happen if they are not reduced to traditional endpoints and statistical measures. New thinking is needed for these new tools. Yes, this introduces a level of complexity and variability, but it also opens up new possibilities.

Another example relates to the use of AI to optimise trial design, and the use of liquid biopsies and MRD (both markers of cancer characteristics as discussed in the precision medicine section) to guide treatment decisions. These tools allow for more and/ or faster real-time feedback from the patient’s body, which can then be used to adjust treatment. This introduces the possibility of a more dynamic approach that adapts based on how the patient is responding, rather than a static one rooted in statistics and best-practice guidelines.

I once asked a clinical oncologist how they dealt with the immense variability of cancer, considering that each patient is unique, and that each tumour is unique. Their response was that there was both a science and an art to it. Dynamic responses based on real-life feedback would be aligned with that logic: treatment becomes a series of quick test and learn cycles.

And aside from molecular interventions, the CHALLENGE trial, which showed that a structured exercise programme reduced the risk of cancer recurrence in patients with colon cancer, hints at what could happen if oncology opened up to other disciplines. Molecular interventions may work best out of silo.

Final thoughts

I wanted to paint the broad brushstrokes of ASCO, but I am aware of a major limitation: I did not attend the conference. I was not part of the thousands of conversations that took place in the halls, in the meeting rooms and coffee corners. My brushstrokes are perhaps more akin to a caricature, emphasising one part of ASCO, the part that I can access through my computer. But like a caricature, my aim is not to distort, but to highlight what I noticed.

What I noticed were the strengths and limits of the current oncology system: Rigorous science and a strong commitment to achieving better clinical outcomes. But an overall direction shaped by specific mindsets and incentives that reward certain kinds of advances and suppress others.

The focus on molecular targets and precision interventions dominated the conference. It is supported by a system of research, regulation, and commercialisation that favours measurable, scalable outputs. This makes sense, but it risks narrowing the field’s vision and missing other, potentially valuable, approaches.

If ASCO took a more intentional approach to broadening perspectives, it could become not only a platform for showcasing new drugs and technologies, but also a space for rethinking the very foundations of cancer care. This would mean creating room for research that sits at the margins, work that doesn’t easily fit into pharma’s product-driven model.

It would mean bringing in different types of thinkers to challenge assumptions and offer new frames, and encouraging the oncology community to engage with ideas like feedback loops, stocks and flows, and requisite variety. These concepts won’t replace clinical trials, but they might help the community to ask better questions and design solutions that consider the system, not just its parts.

Many already do, but their voices need to be louder. ASCO would present the ideal platform for that.

I will stop here, with a quote from renown systems thinker Russell Ackoff.

“The future is not about getting better at what we’ve always done. It’s about rethinking what needs to be done at all.”– Russell Ackoff

Thank you for reading. If you enjoyed this article, you might also enjoy my other articles on systems, futures and cancer: here and here. As always, I’d love to know your thoughts. let me know in the comments.