A Strategy for Success

The Life Sciences Industrial Strategy is the UK government’s attempt to identify areas of potentially sustainable change in healthcare and research. It came about as a result of the country’s vote to leave the European Union. In the course of evaluating the assets that we as a country possess, the National Health Service (NHS) realized that life science is a £64 billion industry here – a jewel in the UK’s crown. Academia and industry are huge; we have three of the top 10 universities in the world and a hotbed of spinout companies. I think, as a country, we have a great deal to offer – and I think digital and computational pathology will be an ongoing strength. The UK is a well-respected country in terms of its academic and research capabilities. In many countries, digital and computational pathology still have low adoption rates. We now have the opportunity to develop those technologies, use them to take our pathology service to the next level, and then even market them. If we start developing and validating pathology algorithms, we can then not only improve our own healthcare services, but also potentially help other countries improve theirs.

Building Rome

Change is not going to happen overnight. The fact that the government is now making digitization a priority is already a huge step; that means they will start to look at the funding and organization that will be necessary to roll out a fully digital pathology service. At the moment, I understand that they are looking at piloting a program across five to six hubs before a complete rollout. It will be important to select the right hubs – ones that are already Centers of Excellence in digital pathology – so that they can try to implement at pace and scale. To pathologists and laboratory medicine professionals, I would say, “Watch this space.” I think things will become much clearer in the next three to six months – more announcements, more ways to get involved… This is just the start!

We have opportunities to completely change clinical practice – and the driving forces to make it happen, too. Pathology is what we call a “greying profession,” with a lot of pathologists due to retire in the next few years and vacancy rates of up to 25 percent. But with a growing, aging population and an increasing disease burden, the role of the pathologist will only become more important over time. The Royal College of Pathologists estimates that around 20 million slides are examined each year in the UK (1) and that the increasing demand for pathology services as indicated by the Keele Pathology Laboratory Benchmarking Programme exceeds the growth in the number of consultant pathologists each year (2).

The most valuable resource in any pathology lab is the pathologist – so we need to use them and their skills, knowledge and experience in whatever way is most beneficial for patient outcomes and service sustainability. And, in my opinion, that means the growing adoption of computational pathology practices.

Digital pathology technology (for instance, in terms of scanning) has improved greatly over the last decade. I think we are now at the stage radiology achieved about 15 years ago when picture archiving and communication systems (PACS) were widely introduced. Those machines revolutionized the way that care was delivered, particularly with respect to turnaround times, and I think digital and computational pathology will do the same. These tools enable pathologists to use their skills in a wiser and more targeted way. PACS didn’t happen overnight; it was rolled out over a period of time, with test cases prior to full implementation. These “digital exemplars” will help the NHS establish best practices and tackle the practical realities of digitization.

If you couple the move to digital with artificial intelligence (AI) and machine learning, pathologists will be able to focus their efforts on the cases that really require their time and attention – not on a barrage of ordinary negative results. If we can have AI tools screen out cases that don’t require a pathologist’s eye, we should see faster diagnoses, more efficient use of pathology resources, and closer attention to more complex cases that need specialist care. I think the rise of computational pathology will really take the field into the future and ensure that we have sustainable pathology services for as long as we need them. Ideally, pathologists will get involved as soon as possible – not only will it help them to keep up with the changes, but it will enrich their working lives, ease their strain, and improve patient care, too.

Are our jobs safe?

If you look at the raft of companion diagnostics coming down the road, differential diagnosis is becoming more difficult. Some say that, as our molecular abilities improve, we no longer need histology – but I don’t believe that at all. We need every aspect of pathology as much as ever, and that means we need pathologists. We need not just data, but interpretation as well. Technology is not the answer to all of our problems; it’s a tool that enables pathologists to do their job in the best possible way.

There are already a number of algorithms that can help with diagnosis and treatment decision-making; we simply need to ensure that we can adopt them. Technology advances so fast that sometimes ethics and regulations lag behind – but we need to catch up in that respect. Algorithms are the tools that will free up pathologists’ time to prioritize appropriately. They are the aids that will help decide which slides need to be referred to specialists. Obviously, this will take some time; we have to not only develop and refine these algorithms, but also train and validate them. The more an algorithm is used, the better the machine becomes at implementing it. They are not always right, but the more data they acquire, the more precise they become. It’s no different to a shopping algorithm that suggests products based on your previous purchases and browsing habits!

A blue-sky future

The ideal digital pathology laboratory of the future would be holistic; it would bring together all of the pieces of information – histopathology, molecular pathology, sequencing – for pathologists. With a complete picture of the patient, they’ll be able to take a much bigger role on the multidisciplinary team, and in patient care in general. They become the guardians of advice and knowledge.

What else do I hope we’ll have in our blue-sky future labs? I’d like the ability to see and link databases. The pathologist could examine the patient’s histological, molecular and genomic status, then search for similar patients and see what courses of treatment have been most effective. The possibilities are endless – and that’s the great thing. I think pathology is going through a very difficult time at the moment; there are more and more demands being put upon pathologists. But I believe that, if we embrace new digital and computational technologies, we have a bright and rosy future ahead. If technology can free pathologists to do the things that really matter, how great is that?