The head of Cleveland Clinic’s Cancer Institute explains why the fight against cancer is a story of real, measurable progress.
There are few who know as much about cancers as Alex Adjei, head of the Cleveland Clinic’s Cancer Institute. A respected and experienced professor of oncology, after completing his doctorate at the University of Alberta in Edmonton, Canada, he was chief resident in the Department of Medicine at Howard University in Washington, DC, and completed a clinical fellowship at the Johns Hopkins University School of Medicine and Hospital in Baltimore. The winner of numerous awards, he is also editor-in-chief of the Journal of Thoracic Oncology and the JTO Clinical and Research Reports and has authored more than 300 peer-reviewed publications.
Here, he talks to Healthcare Today about the progress made in cancer research, why oncologists have embraced AI and how we can diversify participation in clinical research.
“It’s important to recognise that advances have been made across all areas of cancer care.”
Are we moving fast enough in cancer research, or rather, are we moving as quickly as we can?
It’s remarkable how much progress has been made. It hasn’t been fast enough, of course, but the advances are real.
When I began my career in the late 1990s, for example, a patient with metastatic lung cancer had only one chemotherapy option. It was poorly effective – tumours shrank in just 17-20% of patients, and survival beyond a year was rare. Today, it’s a completely different picture. We no longer think of lung cancer as a single disease. Instead, we look at a patient’s genetic profile and tailor therapies accordingly. As a result, many people with metastatic lung cancer are now living five years or more.
The same is true for other cancers. Kidney cancer and melanoma once had virtually no effective treatments – chemotherapy was both toxic and ineffective. Now, for renal cell cancer, we have multiple targeted oral drugs. For melanoma, we have a whole suite of immunotherapies that are transforming outcomes.
Whether you see the glass as half-full or half-empty depends on your perspective, but overall, I think the fight against cancer is a story of real, measurable progress.
Which areas of oncology research are closest to delivering transformative changes for patients?
It’s important to recognise that advances have been made across all areas of cancer care.
Take imaging, for instance. Not long ago, detecting a brain metastasis required a tumour of around a centimetre, because we relied on CT scans. Now, with MRI, we can identify lesions as small as five millimetres. That change alone has transformed what’s possible.
Treatment techniques have also evolved. In the past, a patient with even one or two brain metastases would undergo whole-brain radiation. Today, with gamma knife surgery, we can target and destroy individual tumours precisely – even multiple lesions – and repeat the procedure if necessary.
Surgical approaches have advanced as well. Now, minimally invasive techniques use small incisions and scopes, allowing patients to go home within two days. Similarly, prostatectomy used to be a painful, debilitating operation. Today, robotic surgery means the surgeon controls the instruments from a console, and the patient can leave hospital the next day.
Systemic therapies themselves have also been transformed by immunotherapy and targeted treatments aimed at the genetic drivers of tumours.
Progress has been made at every stage. The difficulty is that many cancers are still diagnosed late, and that’s where most deaths occur.
“The reality is that for the majority of cancers, there are still no actionable genetic mutations.”
Precision oncology and genomic profiling are becoming more routine. Are we getting close to having personalised cancer treatment as standard?
We are getting closer, but human biology is far more complex than we once assumed. In the early days, the expectation was that by sequencing the genome, identifying mutations, and targeting them, we would be able to treat most cancers. The reality is that for the majority of cancers, there are still no actionable genetic mutations. In many cases, there isn’t a single driver mutation we can target – instead, there are multiple mutations working together.
Targeted, personalised therapies are very common in lung cancer but far less so in colorectal or prostate cancers, where there are only a handful of mutations we can exploit. This has led to a growing recognition that focusing solely on genes isn’t enough. We need to look at proteins as well, because it is the proteins that actually drive the biology.
The problem is that proteomics has proved enormously difficult. There are thousands of amino acids in the blood, producing extremely complex protein profiles. This is where artificial intelligence may represent the next frontier.
One particularly important application is in risk stratification. Today, when someone undergoes surgery for cancer, we often give them chemotherapy or immunotherapy afterwards for up to a year, just in case. With AI and proteomics, the first step in truly personalised therapy will be personalised risk. We will be able to say: this patient is cured, no further treatment needed, while that patient is at high risk of recurrence and should receive additional therapy. That would be a major breakthrough.
Enthusiasm for AI within the healthcare sector is mixed. Is it fair to say that oncologists are at the forefront of AI and are embracing it in a way that not everybody seems to be?
I think you are probably right, and it comes back to what I said earlier about the complexity of cancer and the limits of human perception and intellect in making sense of it. Take pathology, for example. Traditionally, a pathologist studies slides under a microscope, deciding whether a sample shows cancer and, if so, which markers it has. It is remarkable how well this system has worked, given how subjective it is. Now, with digital pathology and AI, we can train algorithms on thousands of samples. That gives us the potential not only to identify cancer more accurately, but also to distinguish between tumours that are likely to recur and those that are not.
The same is true in radiology. In the US, for example, women begin screening mammograms from the age of 40, which creates a workload with which radiologists struggle to keep up AI could be trained to, at the very least, rule out normal scans.
AI is also transforming monitoring. One area we are working on involves wearable devices, which continuously record temperature, pulse, respiration and oxygen saturation and transmit the data to a central system. The goal is to learn which combinations of readings predict that a patient is about to deteriorate. This is crucial because many of the novel cancer therapies we use, such as CAR T-cell therapy, can trigger severe complications. At present, we keep patients in hospital for observation, but with these bio-buttons, we could instead allow them to go home while still being closely monitored.
And then there is the concept of the digital twin, which is still aspirational but could fundamentally change how we treat cancer. The idea is to build a detailed virtual model of each patient by integrating data from multiple sources. All of this information, combined with records of how other patients with similar profiles have responded to different treatments, would be fed into AI systems and the computer could then predict which treatment is most likely to work for them.
AI is what will allow us to integrate all these parameters and make genuinely predictive models. That is why oncologists are so excited about its potential.
“Pragmatic trial design, simplified eligibility, and decentralisation through digital tools are making it far easier for patients.”
What needs to change in the way we run trials to speed up innovation and include more diverse patient populations?
This is a question that’s very close to my heart, because it’s an area I’ve been working on extensively: how do you diversify participation in clinical research? It’s not just about social justice or equity, though those are important. Scientifically, it also makes sense. People around the world have different diets, lifestyles and habits, and diseases can present differently across populations.
There are several challenges. One is human capital: do we have enough trained physicians and researchers to carry out the necessary studies? Another challenge lies in the way we design trials. Traditionally, clinical research has relied on strict eligibility criteria. These rigid standards create a kind of paradox: we only recruit the marathon runners of health, yet when the drug is approved, it will be used by people with real-world conditions. That mismatch is problematic.
One promising approach is the pragmatic trial. Here, criteria are kept minimal, and oncologists treat patients as they normally would, with data collected as part of routine care.
Another key area is decentralisation. Participation in clinical trials is often hindered by the time commitment and travel required. We’re now reducing this burden through innovations such as remote consent. With secure platforms like DocuSign, patients can review information at home.
These kinds of changes – pragmatic trial design, simplified eligibility, and decentralisation through digital tools – are making it far easier for patients to take part. And that, in turn, is vital if we are to generate research that truly reflects the diversity of the populations we aim to serve.
How do you balance the speed of innovation with the cost pressures that exist in every healthcare system?
That’s a difficult challenge, and it requires everyone – clinicians, governments, pharmaceutical and device companies, NGOs, and patient advocates – to work together to find solutions. The costs are enormous. Some progress is being made: for instance, virtual care reduces the need for travel, which lowers costs for participants. As I just mentioned, pragmatic trials also help by avoiding frequent, expensive imaging requirements like MRIs.
AI and other technologies could simplify and shorten trials further, though that’s still in development.
This is crucial because the cost of developing a drug directly affects its price. Innovations in trial design and technology could reduce expenses for both companies and health systems.
