The Future of IO: Embracing the ‘Pioneer Spirit’ of Innovative Cancer Care

IO Learning sat down with Mohammad Elsayed, MD and Stephen B. Solomon, MD to discuss their recent paper on the future of IO.

The interventional oncology field is one that is ever-expanding and bringing fresh perspectives and innovative solutions for cancer care to the forefront of the health industry. IO Learning recently interviewed  Mohammad Elsayed, MD, interventional radiologist at Memorial Sloan Kettering, and Stephen B. Solomon, MD, Chief of Interventional Radiology at Memorial Sloan Kettering, Professor of Radiology at Weill Cornell Medical College, and the former president of the Society of Interventional Oncology, authors of a recent paper, ‘Interventional Oncology: 2043 and Beyond’, to get their insights on the potential innovations and advancements that may shape IO in 2043 and beyond. 

IO Learning: 

To start, you have noted the importance of a well-developed clinical and research infrastructure as a defining characteristic for the future of interventional oncology.  Can you elaborate on what makes a well-developed clinical and research infrastructure possible? 

Dr. Solomon:

Interventional radiology evolved from the field of diagnostic radiology where the infrastructure for clinical care of patients was not as strong as many of our colleagues in medicine or surgery. As the field of interventional oncology has grown, the necessity for seeing patients before their procedure for consulting with them, discussing the procedure and possible complications and alternatives, as well as seeing patients in follow-up has become apparent. In the past, when it comes to the field of interventional radiology, you might describe it in a way similar to what I believe Dr. Ernest Ring once said, a "You pick them. We stick them" field.  

But today—and in the future—I know there will be more and more clinical involvement of interventional radiologists and interventional oncologists. And so, the infrastructure that is needed is a well-supported IO clinic with time allocated in the interventional oncologists’ day or week for the clinic. Good care mandates the nursing support to deal with patients at home and mandates mid-level support to help us in the clinic, and support in the procedure rooms. I think the infrastructure that is currently seen in a surgical specialty needs to be applied to the field of interventional oncology.

Dr. Elsayed: 

That is probably the most important question in our subspecialty, and we could spend a whole hour discussing ways to further develop the infrastructure of IR and IO. 

In general, this involves a continuation of current activity in the IR community. This includes grassroots advocacy, like increasing awareness for procedures that are available for patients as well as investing in our training programs. Another major component of this is making sure that we have support from our medical institutions.

IO Learning: 

In your paper you discuss the potential for interventional radiologists to be the first point of contact for cancer patients in the future. Could you elaborate on how this shift could expedite cancer diagnosis and treatment initiation?

Dr. Elsayed: 

Sure. This review features the Rapid Diagnosis Program, which Dr. Solomon has been spearheading at our institution and advocating for it as a model at other centers. . 

In general, this is a program where patients who have suspected malignancy, either on imaging or due to other clinical features, are initially seen by an interventional oncologist who evaluates the imaging and, if appropriate, schedules a biopsy and other necessary tests. After that, the interventional oncologist facilitates a referral to the appropriate oncology specialist. This has expedited the time to diagnosis and also helped patients navigate this initial part of their cancer journey.

Dr. Solomon:

As Dr. Elsayed correctly stated, we at our institution have developed the infrastructure to be more active clinically with patients, not just doing procedures. We've gotten to the point where we can actually be the gateway for patients to the institution. One special example is the Rapid Diagnosis Program.

But I think that interventional oncologists offer several different procedures and techniques that lend themselves to being the gateway to the hospital and to the field, and one is diagnosis with biopsy. Another example is the care of a patient who has stage one renal cancer. 

It's very easy for a kidney cancer patient to be seen first by an interventional oncologist who can make the diagnosis with a biopsy, can discuss the treatment options of observation, ablation, and surgery just as it is for a surgeon or a urologist. And it can also be beneficial that the patient can be seen in follow-up IR clinic, whether the patient chooses observation or ablation. So, I think that the ability to have the clinical infrastructure to support this makes the field a stronger one.

IO Learning: 

Do you believe that AI or machine learning will be used to kind of not just diagnose rapidly, but even predict a future diagnosis for patients?

Dr. Elsayed: 

That's a great question. We already have certain developments, for instance, the LIRADs program for HCC where, based on imaging alone, we can make a cancer diagnosis with a high degree of certainty. Right now the model still heavily relies on tissue diagnosis. However, there could be a future where we use AI and big data to at least streamline this process reaching a diagnosis.  

Dr. Solomon:

I think we're going to learn a lot with AI. Our educational system, our medical experience up until now— it’s been on apprenticeship learning. And, I may learn everything I can learn from the patients and physicians around me, but that's the extent of my learning.

However, with AI, we hold the potential to acquire data from many institutions from around the world and pool that together to have an understanding that will be much stronger than any one physician or one institution can have themselves. And so we can learn a lot more and a lot faster by using all of the experiences of every physician around the world rather than just the local experience pools in which we currently operate.

IO Learning: 

Now, you mentioned the Rapid Diagnosis Program, and it has been cited as ‘a glimpse into the future of interventional oncology.’ How do you envision similar programs influencing cancer care on a broader scale?

Dr. Elsayed: 

What's important about the Rapid Diagnosis Program is that it reduces a lot of the burden on patients. It improves quality of life by decreasing the hardships required in seeing different specialists, obtaining fractured care, and basically having a prolonged experience in obtaining treatment. This is one of many opportunities for IR and IO to support patients. 

Our paper specifically mentions a greater IR inclusion in tumor board discussions. At some centers, the existing model is that a tumor board discussion happens, and then someone in that tumor board makes a referral to an interventional oncologist. This can lead to missed opportunities where patients are not evaluated for important procedures. This trend is now shifting to where interventional oncologists are becoming an essential part of the multidisciplinary tumor board, particularly for “liver” tumor boards. In the future, interventional oncologists will likely have an even greater presence in tumor boards focused on treating  cancers outside the liver. 

Finally, another component in which interventional oncology is going to have a greater presence is within broader oncology working groups and organizations responsible developing treatment guidelines and deciding how resources are allocated

Dr. Solomon:

The Rapid Diagnosis Program is a very exciting program. When a patient calls our hospital and says that they have many lesions on their imaging study and would  like to be seen at our cancer center, the triage center doesn't know whether this patient should be seen by a breast oncologist, a colon cancer oncologist, a lung cancer oncologist, etc. And so, historically, they might be sent to a surgeon because the surgeon was the person who would do surgery to get a tissue diagnosis. But that's not how we operate today.

Today, most diagnoses are made by needle biopsy done by interventional radiologists and oncologists. So, the visit to a surgeon to orchestrate the biopsy by the interventional oncologist is redundant. What would be faster, more streamlined, and effective would be for the interventional oncologist to see that patient who needs a diagnosis from the start. That's what the patient comes for, to get a diagnosis.

So we've set up this program where a patient who has multiple lesions on imaging and doesn't have a diagnosis is seen first by the interventional oncologist who makes a diagnosis. Then, the interventional radiologist refers them to the appropriate oncologist for appropriate care.

Having the infrastructure and the clinical model necessary to see patients has allowed us to play this important role in the institution. And I think that will be something that will be seen more and more throughout other institutions as well.

IO Learning: 

When thinking about  the integration of interventional oncology with palliative care teams, can you provide examples of specific procedures that could significantly enhance the quality of life for cancer patients?

Dr. Solomon:

So, we see interventional oncology as a specialty that cares for the patient throughout their cancer journey from the time of diagnosis, to possible definitive therapy, and to palliation. Palliation is a growing area where quality of life is seen as a critical part of the cancer patient's care. Whether it's draining pleural effusions, ascites, or doing nerve blocks or ablations of painful bone lesions, these are all considered palliative and improve a patient's quality of life. 

Dr. Elsayed: 

This is a really exciting area in which interventional radiologists are showing ‘we're here to stay.’ A lot of this is due to the groundbreaking work by many interventional radiologists, some that I've been very lucky to work with, in advancing palliative care. 

With regards to palliative care, our contributions can be divided into a few major subgroups. 

One is the palliation of pain. Cryoneurolysis and other ablative tools have a growing role in treating cancer pain, which can be particularly useful for reducing dependence on opioids and other medications. Nerve blocks also have a growing role in getting patients through procedures with less pain and sedation requirements.

Musculoskeletal interventions are also rapidly growing. We have certainly learned in the last few decades, with procedures like kyphoplasty and vertebroplasty, that we can improve patient outcomes in very important ways. We're finding that other musculoskeletal procedures, such as cementoplasty and screw fixation for pathologic fractures in the pelvis, can also help reduce pain and potentially even improve mobility.

We're also seeing more sophisticated interventions for treating patients with vascular, biliary, and lymphatic complications due to their cancer or following their cancer treatments. For instance, lymphatic leaks are increasingly not just managed by drainage catheters, but also by treating leaks at their source. With all the incredible work that is being done right now, these treatments are going to become even more sophisticated. 

IO Learning: 

How do you foresee the integration of precision oncology, particularly tumor biomarkers, influencing decision-making in interventional oncology procedures?

Dr. Elsayed: 

This is a good time to start with a definition of precision oncology, which can vary depending on who you ask.. In general, precision oncology refers to the use of tumor and patient-specific information to guide cancer therapies. We see it in practice today where, after biopsy, we analyze the tumors for certain genetic mutations that might be amenable to certain therapies. 

Just like medical oncologists, interventional oncologists can use precision oncology to guide various aspects of therapy. For instance, a study at our institution found that in patients with metastatic colorectal in the liver, presence of the KRAS mutation had worse outcomes after ablation. Information such as this can guide treatment selection, aggressiveness of therapy, and frequency of surveillance imaging. 

Another major development in the world of precision oncology is the use of “liquid biopsies.” This involves measuring small fragments of DNA from tumors, called circulating tumor DNA (ctDNA), in blood and other fluids throughout the body. This can tell us a number of things.  For instance, ctDNA can be used to guide chemotherapy treatment regimens and predict prognosis after surgical resection. Early data also suggests that ctDNA will have a role in interventional oncology, such as improving surveillance and predicting response to therapy. Finally, we can likely improve precision oncology by sampling tumoral vessels directly, which is a potential way that interventional oncologists can increase the utility of these kinds of tests. 

Dr. Solomon:

In our paper, we cited a few examples of the application of precision oncology, but there are many more where we are looking at genetics of patients and using those genetics as an indicator of who will be a good responder to a particular treatment and who may not be a good responder and be better off with an alternative treatment. I think that as we understand the genetics and mechanism of cancer, we'll have a better understanding of how to triage patients to different treatment options based on their genetics. 

One example of precision oncology in interventional radiology might be understanding that the margins may need to be larger in certain cancer types compared to other types. And so it might guide us in, for example, our ablation treatment size.

IO Learning: 

Could you elaborate on the potential of genetic editing in precision interventional oncology? How might this impact patient-specific treatment approaches?

Dr. Elsayed:

Gene editing is a hot topic, and will likely be a significant area of focus for interventional oncologists. For in depth discussion on this topic, I recommend readers take a look at some of the great work we referenced in our paper. There are a few ways in which gene editing can be used in the future with respect to interventional oncology. One major challenge of CRISPR and other gene editing tools is tissue selectivity. Among other factors, systemic delivery of gene editing machinery may not be feasible due to inadequate localization into tumors. Interventional oncologists may thus have a role in delivering these agents directly into tumors under imaging guidance. 

Certainly, we can think about other unique applications which are not well studied. For instance, we can modify tumors to express biomarkers that make them easier to see on imaging.  We can also modify tumors to make  them more sensitive to systemic or locoregional therapies. This is an exciting time for our specialty. 

Dr. Solomon:

Genetic editing is perhaps one of the more far-out applications that we discussed in our paper. But genetic editing is a technique that is growing, certainly in the science lab, and has had some limited applications outside in clinical medicine. 

But the idea that a genetic defect such as a sickle cell defect could be corrected by injecting and using genetic editing tools to edit the genetic abnormality makes you contemplate the possibility to apply this to certain cancers where we understand that genetic defects are leading to that cancer.

Interventional oncologists are the Federal Express or UPS treatments. Rather than injecting something intravenously and going throughout the body, we can indirectly target cancer by either image guided needle placement or intraarterial therapies. And by delivering it to the tumor, we can deliver the genetic editing tools directly to the tumor and perhaps correct defects and mitigate cancer growth. 

I think that intravenous delivery creates another burden—or another hurdle—that we need to scientifically figure out how to get around. That is, how do you not lose the tools as they float around in the bloodstream? If we start off by having it delivered to the tumor directly, we will be more likely to see the effect. So, I think that interventional oncologists will play an important role in this, and I think that we can use our imaginations and envision where genetic editing will go in the future.

IO Learning: 

How do you anticipate AI, deep learning, and robotics will enhance the precision and effectiveness of interventional oncology procedures?

Dr. Solomon:

So, we talked a little bit earlier about AI using data from around the world and sifting through that and finding trends and understanding what treatments work and don't work in particular situations. We are also using deep learning techniques for us to interpret imaging related to interventional oncology procedures. One example might be looking at a lung ablation where we can try to interpret the post-ablation imaging to identify recurrence and identify margins.

These are all potential applications of deep learning, and of course, robotics can improve our ability to deliver to different areas. There are robots that are used for percutaneous approaches and robots that are used for intravascular approaches that are coming online soon and will allow improved delivery more than likely as well. They also potentially can democratize interventional oncology so that advanced techniques are able to be disseminated throughout the world.

Dr. Elsayed: 

Another great question and topic that we could spend an hour trying to answer. Advances in these technologies have opened up a world of potential applications we haven't even considered yet. With so many brilliant minds working in this space, we are bound to see significant innovation in interventional oncology practice. 

There are some obvious areas where advances in technology will have a significant impact in the next two decades. Certainly, image interpretation will improve. AI, deep learning, and radiomics can improve tumor detection, predict response to therapies, optimize patient selection, and reveal other clinically actionable insights which are not readily apparent to the human eye. 

In the procedure room, AI can act as a copilot. For instance, it can streamline vessel navigation during transarterial therapies. AI will also make existing devices more sophisticated. For instance, deep-learning tools can make automatic image adjustments to reduce radiation dose to the operator and patient. The possibilities are endless. 

IO Learning: 

Augmented reality and robotics are both being introduced into the IO field as we speak. Firstly, would you say there is more promise to one or the other for IO procedures? 

Dr. Elsayed: 

In the near term, especially now that Apple introduced their new version of an augmented reality device, we'll likely see more widespread use of augmented reality over robotic devices. Ultimately, however, I think robotics will have an even greater influence on how we perform procedures.  Given the pace of development, it is a very likely future where we can use robotic devices to remotely perform procedures. 

There is also increasing interest in small robotic devices. For example, pioneering work by Dr. Oklu describes the use of nanobots, which can perform functions in the body without the need for catheters and wires. 

Dr. Solomon:

I think both of them are developing simultaneously. Augmented reality has some challenges with motion, but there are a lot of techniques that are being developed to counteract that. Robotics also has to deal with motion complications, but I think both are developing simultaneously and are exciting to follow. I think they're both tools that can help us with educating and training people and spreading the technology around the world.

IO Learning: 

Secondly, do you believe these tools can help with the turnover and retention of IO professionals, thus making IO procedures more accessible globally? 

Dr. Solomon:

I think that is a very important aspect of the future of interventional oncology. The success will not be because it can be practiced at only the most experienced facilities in the world. The success will be when these technologies and the field and practitioners are able to be distributed throughout the world so that everybody can benefit from the advances that have been made.

Dr. Elsayed: 

Our review talked about the use of virtual reality to simulate high fidelity learning experiences. This and other technologies will be useful for improving access to training and education. 

IO Learning: 

Do you believe AI will play a role in data acquisition and sharing to help create a more equitable and comprehensive care infrastructure for patients of all ethnicities, education backgrounds, and socioeconomic statuses? I know you talked about AI being utilized from a data perspective, but could it also help with equity and comprehensive care in general?

Dr. Solomon:

I think that it's understood that AI will include all of those aspects of patient data and will help us identify what are some of the problems and challenges of IO. We may even learn things that are not necessarily considered medical. It may be that the zip code that somebody is in is particularly challenging for something, and we may be able to devote more resources to that. So I think AI allows us to dissect through a lot of data and identify the gems there that provide information that can be useful.

Dr. Elsayed: 

The benefit of radiology is we have access to a large amount of data, imaging and otherwise, that we can leverage to identify how we can better serve our patients. It’s important that as these technologies are introduced, ensuring equitable healthcare delivery is a priority. I’m hopeful these tools will increase the scale at which we can reduce barriers to care and further personalize therapies. pushing. We are fortunate our field so many fierce advocates for equitable healthcare delivery.  And again, a very big topic that we could spend hours discussing, and there's a lot of experts already in the field like Dr. Judy Gichoya, who's looking at how imaging can be used to improve equity, for instance. What's nice about radiology is we have access to a lot of data, imaging, and otherwise that we can use to develop algorithms. And because of this, this can allow us to identify opportunities where there's gaps in how we're taking care of patients, and overall, it's going to help us personalize the way we treat our patients and really give the specific right treatments to the right patients.

IO Learning: 

What would you say is the most exciting innovation or advancement the field should be tracking currently? 

Dr. Solomon:

So that's a challenging one given that there's so many interesting areas in the field that are developing. I do think that immunology and intratumoral therapy will definitely be an exciting thing in the future, as well as the fact that interventional oncologists who deliver these potential advances to patients will hold an important role in the future.

I also think that the field of theranostics with targeted radio-isotopes and delivering them to tumors with genetic targets will also be very important. One example is, one of my colleagues, Dr. Etay Ziv, has been comparing intra-arterial vs intravenous delivery of Lutathera for neuroendocrine tumors and finding that he can deliver a much higher target to background signal through an intra-arterial approach versus an intravenous approach. And that's just one example of how this will be a real hope for the future with less toxicity systemically and more therapy on the target.

Dr. Elsayed: 

I'll break this up into the immediate term and the long term. Currently, as we know today, with image-guided ablation, one of the major challenges we currently have with image-guided ablation is difficulty in accurately assessing the ablation zone. I’m excited to see improvements in various tools that can more accurately assess the ablation margins, such as PET-guided interventions and ablation confirmation software. 

Right now, there's a lot of interesting work being done on using technology to more accurately assess the ablation margins. I’m excited to see how improving to develop ablation confirmation software as well as other imaging techniques. And so, I'm most excited in the immediate term about developing better standards for how we do ablations and making sure there's better outcomes with that regard. In the long term, I'm excited about seeing how the treatment of tumors outside the liver will evolve and change.

We do a lot of our locoregional therapy work in primary and secondary tumors in the liver, but as you can see, a lot of research is exploring how we can treat tumors outside the body. This will certainly change treatment algorithms and hopefully result in better, less invasive options for patients. 

What I’m most excited about is the general theme we have repeatedly touched upon in this interview—the trend towards personalized therapy and ultimately better outcomes for patients. This is an exciting time, as our ability to tailor the care plan to the patient will be important for the general themes we have highlighted in this interview.

Finally, just as a theme that we've had in this interview and in our paper, is the trend of automation and personalization. Automation will certainly make things easier for interventional oncologists and more efficient, but we'll also be able to personalize therapies and really, really find the right treatments for the right patients. And that's what I'm most excited about, is really just having better outcomes for patients.

IO Learning: 

And then, just to close out, are there any other additional concepts related to the future of IO that you would like to elaborate on for our readers? 

Dr. Solomon:

I think that what is essential for our field is for us to collaborate and create multi-institutional trials that will show us the value and identify the weaknesses of our treatments. Much like medical oncology has utilized the cooperative group system from the NCI to fund large trials across many institutions, we must similarly develop cooperative groups that will run multicenter trials to study interventional oncology.   It's only through learning in the clinical trial arena will we really be able to promulgate and change standards and practices. So we must set up the research infrastructure to allow us to do this.

Dr. Elsayed: 

The future of interventional oncology is very bright, given the pioneering spirit of the IR and IO community. A continued commitment to clinical excellence, research, and education will ensure that interventional oncology is well positioned for the future.

Source: Mohammad Elsayed ¹, Stephen B Solomon ¹, Interventional Oncology: 2043 and beyond, Pubmed doi: 10.1148/radiol.230139