Using AI to Optimize Therapeutic Choices and QOL in Patients With Cancer

In an interview with Peers and Perspectives in Oncology, Tim Showalter, MD, MPH, vice president of clinical development at ArteraAI and professor in the Radiation Oncology Department at the University of Virginia School of Medicine, discussed increasing patient quality of life (QOL) by using artificial intelligence (AI) and precision medicine to inform cancer treatment decisions.

Q: What initiatives are taking place to improve QOL in patients with cancer?

SHOWALTER: Improving the QOL for patients with cancer is of utmost importance. As our treatment options have expanded, so have the toxicities that are associated with those treatments.

A lot of the opportunities to improve QOL for patients with cancer starts at the very beginning, at the time of consultation. When I’m seeing a patient in the clinic, I’m thinking about choosing the right treatment options that will be most effective for that patient but also how to avoid adverse events [AEs] from treatments that are unlikely to benefit.

So I think the scientific efforts, like the work we’re doing in Artera that focus on precision medicine tools to better match patients with optimal therapies, are essential for improving the QOL of patients with cancer.

Q: In what ways is the cancer care landscape changing right now?

With precision medicine, we’re able to better tailor patients’ therapies. So we can match patients to therapies that are likely to have a high degree of benefit, and we can avoid using therapies for patients who are unlikely to benefit from them. To take an example from prostate cancer, we often wrestle with advising patients on whether or not to add hormone therapy to radiation therapy. While we know it benefits many patients, it does come with AEs like hot flashes and fatigue. The ArteraAI Prostate Test helps to identify patients who are most likely to benefit from therapy so we can concentrate delivery of that treatment to those patients.¹

Perhaps even more importantly, it helps us identify those patients who can be safely treated without that therapy. Expanding this point more broadly across the cancer continuum, the expansion of precision medicine tools to guide therapies is a trend that’s ongoing broadly. I see that translating into an immense QOL benefit for patients.

Q: How do you address the physiological impacts of cancer?

I see patients routinely in my clinical practice and have to counsel them about not only what treatment options are likely to maximize their cure but also about the overall physiological impacts over the short term and long term. I think it starts with a careful assessment and understanding of the cancer that the patient’s been diagnosed with, making that as thorough as possible.

That can include cancer tests such as imaging, precision medicine tools, and a thorough understanding of their current health conditions, their age, or their other ongoing health problems.

We use that information to help guide decisions to make sure that the level of aggressiveness and the therapies that are used for that patient strike the right balance between being effective while minimizing the harms of therapy.

I think that the initial assessment and counseling for the patient is probably the most important step. And then overall, as a community, supporting our patients, colleagues, and family members in dealing with a cancer diagnosis often involves encouraging healthy [habits].

Q: What do you think is the role of AI and precision medicine in increasing patient outcomes for survival and QOL?

AI plays a promising role in improving patient outcomes, both for survival and QOL. AI provides us a rapid, scalable, cost-effective solution to giving patients the direct insights that they need about their cancers. What’s exciting about this technology is that we can provide very decision-specific insights for a patient with cancer and their oncologists, and we can also bring this technology to serve many different cancers in the near future.

Because of the delivery system, which uses image files, AI can be used to make precision medicine available globally, within the reach of more patients. I think that’s the most exciting part. By better matching patients to the treatments that are likely to improve cure rates and are less likely to expose patients to unnecessary AEs, I think we can make a real impact for QOL.

Q: How would you recommend oncologists incorporate these things into their practice?

As oncologists, it’s important that we all remain familiar with the latest emerging evidence. Obviously, AI is making inroads into medicine as well as a variety of other aspects of our life. I think it’s worth taking the time to think about how to incorporate AI to directly serve our patients’ needs in the clinic. AI that’s applied to precision medicine is a great example where we can leverage the technology to provide actionable insights into cancer treatment decisions.

There may be many other opportunities to improve the patient experience, ranging across health care delivery steps and other aspects of care. So I’d encourage all my oncology colleagues to remain on the leading edge of the safe and effective adoption of AI to clinical practice.

Q: Are there any recent findings or upcoming trials you think represent this QOL increase?

In the work that I do, I’m most excited about the research efforts that Artera and our scientific collaborators have been a part of in terms of bringing AI technologies to treatment intensification or deintensification decisions for prostate cancer.

In particular, the work that we’ve been able to complete with NRG Oncology investigators on identifying patients who are likely to benefit from short-term androgen deprivation therapy added to radiation therapy, and we’ve developed a separate model to identify patients who benefit from long-term hormone therapy duration compared with short-term hormone therapy duration.^2,3

Those projects demonstrate that an AI approach can be used to analyze tumor specimens and help guide therapies, with a real impact both on cure rates and QOL considerations. As an oncologist, when I see a patient in the clinic, I want to understand directly how I can help this patient make the decision that’s in front of them. How can I help them benefit from the positive effects of additional cancer treatment like hormone therapy or help them avoid the negative AEs of a therapy that may not improve their outcomes? To me, these familiar examples are a testament to the clear impact of AI. I hope that as AI plays a broader role within precision medicine and oncology, we’ll see more algorithms and tools like the ones that we’ve built that go directly at the point of a patient’s clinical decision.

Q: What are you most excited about going forward in the cancer space for this new technology?

There are 2 points that are very compelling about the work that we do that I’d like to highlight. The first is that the availability of digital pathology and clinical outcomes data, combined with machine learning approaches, gives us the ability to very rapidly develop a whole armamentarium of high clinical utility tools that can directly serve treatment decisions. We’ve seen rapid development cycles for the tools that we’ve built in prostate cancer, and we hope to continue that push to make precision medicine globally available across cancer types.

The second point is that AI offers promise for improving the turnaround time for what we can offer to clinicians to serve their patients. One of the challenges of taking care of patients with cancer is that it can often be a major source of anxiety, particularly after diagnosis and before a treatment plan is in place.

AI approaches applied to digital pathology imaging provide us a pathway to precision medicine with a rapid turnaround, potentially even at point-of-care. That’s something that I value as a clinician because what we want to be able to do is help each patient we see get to the optimal decision as soon as possible.

References

1. ArteraAI prostate test. Artera. Accessed September 25, 2023. https://bit.ly/3LDBxhM

2. Spratt DE, Tang S, Sun Y, et al. Artificial intelligence predictive model for hormone therapy use in prostate cancer. NEJM Evid. 2023;2(8). doi:10.1056/EVIDoa2300023

3. Armstrong AJ, Liu VYT, Selvaraju RR, et al. Development and validation of an AI-derived digital pathology-based biomarker to predict benefit of long-term androgen deprivation therapy with radiotherapy in men with localized high-risk prostate cancer across multiple phase III NRG/RTOG trials. 2023;41(suppl 16):5001. doi:10.1200/JCO.2023.41.16_suppl.5001