An accurate understanding of what the future may look like is important for patients with solid tumors facing the end of their lives. It helps patients understand the severity of their disease and understand treatment options. Now, new technology may give patients more control over their care and end-of-life plan. Researchers from Huntsman Cancer Institute contributed to a that may help to one day improve a patient’s decision-making process by calculating a more accurate prognosis. professor in the College of Nursing, Anna Beck, MD, researcher and director of supportive oncology and survivorship, and oncology clinical pharmacist at Huntsman Cancer Institute explain how advanced technology can equip patients to make decisions that match their goals.
What are solid tumors?
Solid tumors are a form of cancer that typically start as a solid mass in an organ. Cancers that start with a solid tumor can include breast, colon, and pancreatic cancers. Solid tumors usually clump in an organ system before spreading. The study we did focused on solid tumors, as opposed to liquid or blood tumors like leukemias and lymphomas, which may behave differently.
What is advanced cancer?
When solid tumors spread to other organs, doctors call this "advanced" or "metastatic" cancer. Patients with advanced cancer were the focus of our study. The goal of treating patients with advanced cancer is usually not to cure the cancer, but rather extend life or improve quality of life. This can make cancer treatment decisions challenging over time.
How does an accurate prognosis impact treatment options?
Patients with their oncologist must make decisions about whether a new treatment makes sense for them. Understanding the severity of their cancer and what treatments are available can help a patient make better decisions. A patient's understanding of what may lie ahead should match their oncologist’s view. This is especially important when someone has limited time or treatments may not achieve the success they are hoping for. Honoring a patient’s wishes while giving them the best treatment is the goal.
Why is this research important?
Patients in their last six months of life may choose to stop treatments like chemotherapy to focus on quality of life. Sometimes continuing treatments with a low chance of being successful for the patient can harm the patient more than it is helping them.
What is machine learning and how was it used in this study?
Machine learning allows a computer to find patterns in data that can answer a question like how likely it is that a patient will survive the next 6 months. A computer can learn from patient data found in electronic health records. Our experts used the records of thousands of patients with solid tumors who were treated at Huntsman Cancer Institute. The information included things like blood test results, weight, and vital signs.
What results came from your research?
Using the computer’s learned algorithm, we were able to use patient records to identify those with ‘low’ versus ‘likely’ chance of survival in six months. With this new information, we were able to transparently rank factors which influenced the prediction, something that a human has not been able to do on their own. One surprising finding was that the type of cancer was less predictive than non-cancer features. This means that factors related to lab tests, weight, and vital signs seemed to be more predictive than the specific cancer type when it came to chance of survival in the next six months.
How might this impact patient care?
This algorithm may help oncologists to better estimate and communicate prognosis. It may also help ensure patients access supportive care and other resources sooner and focus on what matters most when approaching the last six months of life. We hope this research will further support Huntsman Cancer Institute’s culture of shared decision-making and doing what is best for the patient. The more transparent we can be with patients about their disease and what their future holds, the more we give them autonomy to be a part of the decision-making process.