A Novel Risk Prediction Model for Checkpoint Inhibitor-Related Autoimmune Toxicities
Abstract
Scientific Objective and Rationale: Recent research has shown the effectiveness of immunotherapy treatments in managing patients with both incurable and curable cancers. These treatments turn the patient’s immune system against cancer cells, resulting in impressive and long-lived responses to cancer treatment in many patients. While incredibly effective for some, these treatments do not work for all patients and they are unfortunately associated with toxicities arising from the immune system attacking the patient’s own body. So-called autoimmune toxicities can range from mild and self-limited, to severe and life threatening. Early data show that many of these toxicities can be chronic and life-long, an area of rising importance as patients are treated with these therapies in earlier stage, curable disease. Unfortunately, little is known about how to predict which patients receiving immunotherapy treatments will develop significant autoimmune toxicities related to their treatments. Furthermore, rates of autoimmune toxicities reported for these drugs are based off of studies that were too small to detect the actual occurrence of these rare side effects. Additionally, patient populations included in these studies are often not representative of patients receiving these therapies in the real world. In order to better understand risks of autoimmune toxicities from cancer immunotherapies, we will study their rates in a large database of patients who received cancer immunotherapies within the Veterans Health Administration hospital system. Based off of the patient characteristics of those that developed significant autoimmune toxicities, we will develop a tool capable of detecting patients at increased risk of autoimmune toxicities. This will allow treating physicians and patients to be better informed of the risks of therapy and tailor treatment strategies and surveillance techniques to better monitor their treatment and improve outcomes. PI Career Goals in Cancer Research: Dr. Keller is a junior faculty member at Washington University in St. Louis and a physician at the St. Louis Veterans Affairs Medical Center. His career goal is to improve outcomes for cancer patients receiving immunotherapy treatments by better understanding the risks for individual patients undergoing these treatments. To better understand these risks, Dr. Keller will use advanced statistical techniques that harness the powers of computer machine learning and neural networks to better analyze the available data and draw more accurate conclusions from the available data sources. Applicability of the Research: This research is novel and immediately impacts patient care. The product of the research is a tool that will allow physicians to assess risks of autoimmune toxicities in cancer patients receiving immunotherapy treatment via the assessment of easily available clinical characteristics. This will enable quick and simple risk calculation that will inform both patients and physicians as to the risks inherent in immunotherapy treatment for a given patient. Given that immunotherapies are approved for many cancer diagnoses, this will be useful to large numbers of patients. Benefit to Service Members: The research proposed is of substantial benefit to Service members. The risk prediction tool obtained will have been derived from a population of Unites States military Veterans and will be highly applicable to the diverse racial and economic population of present and past Service members. With an aging population and the evolving use of immunotherapies in early stage cancer patients, many more patients will be treated with these agents each year. As such, knowledge of their risks and appropriate counseling of patients is a crucial public health issue.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Mar 10, 2021
- Source ID
- W81XWH2010785
Entities
People
- Jesse Keller
Organizations
- United States Army