Posts

Curious Dr. George | Plumbing the Core and Nibbling at the Margins of Cancer

The Challenges of Using Artificial Intelligence to Improve Cancer Treatment

Razelle Kurzrock, MD, and Jeff Shrager, PhD

In a previous post, CureMatch co-founder Razelle Kurzrock, MD, told us all about her company’s artificial intelligence (AI) platform that matches patients with treatments based on their cancer’s molecular profile. Here, AI expert Jeff Shrager, PhD, responds, and Kurzrock offers a rebuttal.

Shrager is Co-Founder and Director of Research at xCures, and was formerly Director of Research at Cancer Commons. He is also an Adjunct Professor of the Symbolic Systems Program at Stanford University. Email: jshrager@stanford.edu.

Kurzrock is Director of the Center for Personalized Cancer Therapy and the Rare Tumor Clinic at U.C. San Diego, and Co-Founder and Board Member of CureMatch, Inc. Email: razelle@curematch.com.

Shrager: Whereas I applaud Dr. Kurzrock and CureMatch for their efforts to apply machine learning in precision oncology, I want to offer a bit of a heads-up.

Whereas it is certainly true that “we live in the ‘big data’ generation,” two senses of that term are often conflated. Google and Facebook have enormous datasets with many independent observations across relatively few features. Medical data, especially at the molecular level, is exactly the opposite, having relatively few independent observations across an enormous feature space. Moreover, the settings in which modern AI (i.e., machine learning) has seen successes are those where there are either existence proofs of a solution, which can be drawn upon as a teacher, (e.g., self-driving cars, where even 16-year-olds drive cars adequately well), in closed systems for which we have excellent simulators (e.g., astrophysics), domains in which the roles are static (e.g., games), or in which experiments are basically free (games again, or any domain with a good simulator).

Medicine is completely different: We have essentially no simulations, medical experiments are extremely costly, we lack good treatments (which is why we’re bothering with this at all), and the treatment space changes rapidly. You can’t just teach your robot doctor to cure cancer by observing good doctors curing cancer, because there are no such doctors and cures—there may be some better and some worse doctors, but as far as I know, there isn’t one that can cure cancer “adequately well” who you can use as a guide; indeed, there may be no cure for cancer at all.

Heads up! Machine-learning applications in domains like medicine, where there are small numbers of samples that range over very high dimensionality feature spaces, and with the above-enumerated limitations, are exceedingly prone to getting stuck in non-optimal minima, preferring solutions that work well enough, over exploring solutions that might work better than the ones that have been observed or tried. The way out of this problem is active learning: Rather than taking the apparently best action in all cases, one must balance the strength of belief in one’s rankings against the information gain of trying something new. Doing this requires having a global view of the whole medical (or at least oncological) space, and working out some very difficult “statistico-ethical” questions. Indeed, this is what the clinical trial system is striving to do, although it is doing so horribly inefficiently, and will basically never get there. We can solve this problem, but it requires a much broader AI approach than simply treating each patient in accord with a locally-optimal solution.

(This commentary abbreviates the argument made in much greater detail in a paper I wrote with my colleagues at xCures last year for The Journal of Law, Medicine & Ethics: Is Cancer Solvable? Towards Efficient and Ethical Biomedical Science.) 

Kurzrock: I would like to thank Dr. Shrager for highlighting two excellent points pertaining to the use of AI in routine oncology practice and the inefficiency of clinical trials—I fully agree with him. Allow me to provide some brief comments.

First, I concur that current AI-containing software platforms are certainly not sophisticated enough to be “robot doctors” that could treat cancer. Indeed, decision-support platforms like CureMatch’s BionovTM are not here to replace oncologists. They are necessary tools that help oncologists process immensely complicated data, such as that revealed by next-generation sequencing of tumors. Decision-support platforms are rule-based systems that enable evaluation of complex information by utilizing prior knowledge, akin to the dimensional origami model Dr. Shrager referenced in his earlier work.

Moreover, some of the work that lends confidence to the decision-support platforms are clinical trials. I agree with Dr. Shrager’s point regarding clinical trials’ extreme inefficiency, the fact that they are indispensable to clinical oncology research, and the concept that new clinical strategies are needed, especially to address the questions raised by today’s precision medicine that utilizes complex molecular diagnostics. For example, the prospective cross-institutional I-PREDICT study demonstrated the value of customized, matched combination therapies (rather than scripted monotherapies) and of a matching score similar to that used by BionovTM. Other efforts, such as obtaining real-world data via a Master Observational Trial are also unique approaches that enhance the clinical trial process.

***

Copyright: This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Curious Dr. George | Plumbing the Core and Nibbling at the Margins of Cancer

Using Artificial Intelligence to Match Combination Targeted Therapies in Oncology

A Q&A with Razelle Kurzrock, MD, Director of the Center for Personalized Cancer Therapy and the Rare Tumor Clinic at U.C. San Diego, and Co-Founder and Board Member of CureMatch, Inc. Email: razelle@curematch.com

Q: The new understanding of many cancers brought about by molecular testing has led to a whole new field: precision oncology, which emphasizes targeted and immunotherapy. While promising, and sometimes spectacularly successful, targeted monotherapy has limitations. The evolution of targeted and immunotherapy by combinations of drugs offers new scientific options for cancer patients. But there are so many new molecular findings, so many new investigational drugs or drugs newly approved by the U.S. Food and Drug Administration (FDA), and so few appropriate patients, that matching patients to best drug combinations can be a mathematical nightmare. What have you and your company CureMatch to offer to help with this dilemma?

A: Thank you for this excellent question. As you correctly noted, tumors, even those that share the same histologic origin, are highly heterogenous and unique at the molecular level. Therefore, the existing paradigm of treating all cancer patients based on their tumor’s tissue of origin, even by adding minimal biomarker stratification criteria, has proven largely inadequate. The advent of molecular diagnostics allows for improved patient stratification during therapy selection; however, most patients are still treated with monotherapies, which ultimately perform poorly.

Early data show that individualized matched combination therapies targeting most of a patient’s druggable aberrations are associated with improved outcome. However, selecting the “right” combination in routine oncology practice could be challenging. The average oncologist is pressed for time, seeing approximately 350 new patients annually and up to 100 patients per week. To complicate matters, even if an oncologist wanted to rationally combine only the approximately 300 FDA-approved “oncology-specific” drugs, there would be estimated 45,000 possible two-drug and approximately 4.5M three-drug combinations. Even molecular tumor boards found in some academic centers rely largely upon expert knowledge and experience to tailor personalized combination treatment strategies for hundreds of patients with unique molecular profiles. Clearly, the drug selection process is rapidly outpacing human capabilities, and software tools are needed to help with data analytics.

Bionov™, a rule-based artificial intelligence platform developed by CureMatch, utilizes the latest data available on targeted, immuno-oncology, hormone therapy, and cytotoxic agents. Bionov™ employs an algorithm that matches patients with monotherapy and multidrug regimens based on their available tumor “omic” profile. Drug regimens provided in the Bionov™ report are ranked using a predictive “Bionov™ score” that reflects the degree to which a given regimen matches the patient’s molecular profile.

To generate our database, we curated all FDA-approved drugs relevant to oncology for their biological impact on their targets. Recently, we added oncology drugs that have been approved by the European Medicines Agency (EMA) to our database, and FDA/EMA-approved drugs are kept up-to-date based on their respective labeling changes. Further, we researched and curated preclinical and clinical literature pertaining to the efficacy of these drugs, including drug toxicities and contraindications. The CureMatch scientific team conducts literature reviews on a routine basis to ensure drug efficacy is kept current.

Our methodology has been validated in several studies, and I will highlight two of them here. First, in a retrospective meta-analysis of 70 exceptional responders for whom molecular profiling data was available, Bionov™ correctly ranked the response to all treatment regimens (including failed regimens) with 84% sensitivity and 77% specificity. This analysis demonstrates how the Bionov™ algorithm is able to discriminate, solely on the basis of the molecular fingerprints of a patient’s cancer, treatment regimens that favor a positive outcome from those that are more likely to be associated with an unsuccessful response. The second study I want to highlight is a prospective clinical trial: our group found that a higher matching score (similar to the Bionov™ score) was an independent predictor of increased disease control rate, prolonged progression-free (PFS), and overall survival rates. Furthermore, PFS was significantly improved in 75% of patients treated with combination therapies based on high matching scores.

We live in the “big data” generation. As the patient progresses through their treatment journey, massive amounts of actionable molecular information are generated, and clinical oncologists may not be entirely prepared to effectively utilize it. We believe that predictive analytics models—such as Bionov™—can provide an alternative framework for modern clinical practice, collaborating with and empowering oncologists in their decision-making process.

***

Copyright: This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.