Using AI and Predictive Algorithms to Help Advanced Cancer Patients Identify the Best Treatment Options

Advanced cancer patients, individuals that have a <50% chance of 5 year-survival and have exhausted standard of care options, are often seeking for innovative therapies. These drugs may be investigational and only available via clinical trials. Unfortunately, >95% of cancer patients are excluded owing to exclusion criteria, geographic limitations, timing issues, or simply being at the wrong institution at the wrong time. Alternatively, they might be trying off-label treatments or novel combination therapies via a wide variety of means (right-to-try, expanded access, novel combination therapy, novel indications, etc..). Identifying the right cancer treatment is difficult. Cancer is highly personal, and what works for one person might not work for someone else. Unfortunately, unless a paper is published or a case series documented, most of the learning from these n-of-1 experiments is lost. Finally, oncologists often don't have enough time to help patients find the best treatment options due to the increasing demand of oncology services from factors including an aging and growing population. In fact, ASCO estimates that there will be a shortage of >2200 oncologists by 2025.

We will discuss how artificial intelligence (AI) and natural language processing (NLP) tools can be utilized to analyze Real World Data (RWD) and other data sources to empower patients and their oncologists to make more informed and effective treatment decisions. Furthermore, collecting the safety and efficacy information of these therapies can also be beneficial for researchers of investigational drugs, new therapeutic combinations, or novel indications. This RWD can increase the ability to identify causal inferences between treatments and outcomes.

Learning Objectives:

1. Identify challenges advanced cancer patients and community oncologists face in finding treatment options

2. Evaluate how artificial intelligence could empower more advanced cancer patients and their treating oncologists with the best treatments options they should consider

3. Explore how Real World Data (RWD) on the safety and efficacy of these treatment options may be beneficial for researchers of investigational drugs, new therapeutic combinations, or novel indications