Breast cancer is a devastating disease that affects millions of women worldwide. In recent years, there have been significant advancements in the treatment and management of breast cancer. One such development is the PREDICT prognostication tool, a web-based tool that estimates the risk of dying from breast cancer and the potential benefits of different adjuvant therapies.
Researchers at Case Western Reserve University School of Medicine in Cleveland recently conducted a study to evaluate the accuracy and performance of the PREDICT tool. They analyzed data from the National Cancer Database (NCDB), which included over 700,000 breast cancer patients in the United States. The study found that the tool provided decent accuracy in estimating overall survival (OS) rates.
The study found that the estimated median and mean 5-year OS rates ranged from 83.3% to 84.4%, compared to an observed rate of 89.7%. Similarly, the estimated 10-year OS rates ranged from 69.4% to 73.8%, compared to an observed rate of 78.7%. The performance of the tool was measured using area under the curve (AUC) values, which were 0.78 and 0.76 for survival at 5 and 10 years, respectively.
The researchers concluded that PREDICT is a clinically useful tool that can assist medical oncologists in decision-making and estimating clinical risk. It provides valuable information on the impact of adjuvant therapies on overall survival. However, they also acknowledged that no prediction model is perfect, and PREDICT is no exception.
This study is significant as it is the first to validate the PREDICT tool in a large U.S.-based dataset of breast cancer patients. Previous studies have validated the tool in other countries and populations, but this validation adds to its credibility and usefulness in the United States.
Limitations of the Study
Like any research study, this study had its limitations. The researchers noted that the NCDB does not provide detailed treatment data, such as the type or duration of adjuvant chemotherapy or endocrine therapy. This lack of information could potentially impact the accuracy of the tool’s predictions.
Moreover, the study excluded patients who received neoadjuvant therapy, which is the standard of care for certain types of breast cancer cases. This exclusion limits the generalizability of the study’s findings to a broader population of breast cancer patients.
In an accompanying commentary, Dr. Paul D.P. Pharoah of Cedars-Sinai Medical Center highlighted another important limitation of the PREDICT tool. He noted that the model’s performance is based on all-cause mortality rather than breast cancer-specific mortality. Since the goal of PREDICT is to estimate the benefits of systemic therapies on breast cancer-specific mortality, accurate estimation in this domain is crucial.
Dr. Pharoah also mentioned that it would be beneficial to test PREDICT’s assumption that if the prediction of all-cause mortality is reasonable, then breast cancer-specific mortality estimates will also be adequate. This would require data that is not currently available.
The PREDICT prognostication tool shows promise in estimating overall survival rates in breast cancer patients. This study adds to the growing body of evidence supporting the tool’s clinical usefulness. However, it is important to recognize its limitations and continue refining it to improve accuracy and applicability in different patient populations. With further advancements, PREDICT has the potential to become an invaluable tool in guiding treatment decisions for breast cancer patients.