Utilizing Data Science to Tailor Treatment for Men with Metastatic Prostate Cancer
Abstract
Rationale: One in eight American men will be diagnosed with prostate cancer during their lifetime. Most men who die of prostate cancer have advanced disease. The most common treatments for advanced prostate cancer are testosterone-blocking drugs: enzalutamide (ENZA) and a combination of abiraterone acetate + prednisone (AAP). These drugs increase the chances for survival, but they may increase blood pressure and cause serious heart problems, such as heart attack. The risk appears particularly high for people who already have high blood pressure or heart problems, which many patients with prostate cancer do, especially Black patients. Due to safety concerns, patients having these health problems are usually not included in the clinical studies for drug approval. As a result, we do not know what happens to patients who are too sick to be part of the drug studies. In the U.S., more than half of individuals over the age of 65 have high blood pressure which is not be controlled by medication or have had serious heart problems. If these patients are diagnosed with prostate cancer and treated with these new medications, they may have worse treatment outcomes than published results. One study found that the chance of being hospitalized after AAP increased by nearly seven-fold if the patient has uncontrolled high blood pressure or heart attack before taking the drug. Since high blood pressure, heart problems, and prostate cancer are common among Service Members, Veterans, and Black men, it is important to understand the impact of the patient’s physical condition and race on the outcomes of AAP vs. ENZA to reduce the chances of having undesirable effects from cancer therapies. This is especially important for Black men, because they are more likely to die from prostate cancer and other diseases. Although AAP and ENZA have shown similar survival benefits, they pose different amounts of risk in men with unchecked high blood pressure or heart problems. It is, however, a challenge to choose the most beneficial treatment because relevant data are not available comparing treatment risks. Currently, there are no clear clinical guidelines about how to choose one over the other to maximize the wellness of the patients. Our team will identify a large number of patients with lethal prostate cancer treated in the Veterans Administration Health System using a new approach called the natural language processing algorithm. We will then use data mining techniques to provide new information about what to expect based on the health condition of the patients. The findings of our study will help guide doctors who are prescribing treatment for prostate cancer. They will also help patients answer common questions such as: (1) Which medicine works better for me? (2) What is the risk of side effects that may lead to hospitalization or death? and (3) What symptoms should I look for? It is important to learn more about the outcomes of these drugs, particularly the risks they may pose for some individuals. Objective and Aims: The two major objectives of this study are: (1) to answer the question, What would be the treatment outcomes for someone like me and (2) to compare the risk of serious health problems and death after taking AAP vs. ENZA for patients with different health conditions. Applicability: Our study will provide new information that can have immediate applications: (1) help patients and doctors make better treatment choice to reduce undesirable side effects; (2) help care teams understand and anticipate potential problems; (3) help identify actionable intervention to improve outcomes; and 4) have better treatment outcomes and reduce the need to go to hospital. We expect that this novel approach will reduce undesirable treatment side effects and result in better outcomes, thereby reducing the differences in mortality between Whites and Blacks. The new approach used in this study will also help future studies
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Dec 28, 2022
- Source ID
- W81XWH2210581
Entities
People
- Grace Lu-yao
Organizations
- Thomas Jefferson University
- United States Army