The 3 Titans of Healthcare Progress: Machine Learning, Precision Medicine, Combination Products
In the last couple of years, there has been an enormous amount of interest in combining medical products to create more effective treatments for various diseases. Combination products typically involve a mix of drugs, biological products, and/or devices that work together to provide targeted therapies known that are classified under the umbrella of precision medicine. As precision medicine becomes increasingly important in healthcare, engineers have pinpointed an opportunity to leverage machine learning to optimize these combination products and improve the quality of life for patient and boost positive outcomes in treating/regulating diseases.
Combination products are particularly useful in the instance where utilizing a single drug or device may not be enough to effectively treat a disease or condition. For example, a medical device that combines a drug with a device may be used to deliver the drug directly to the affected area, increasing its efficacy. Similarly, creating cocktails of multiple drugs that target different aspects of a disease can provide a more comprehensive treatment.
Progress Shown
One medical subject field where combination products are particularly important is in the treatment of cancer. Cancer is an extremely complex disease, and it usually demands multiple therapies to effectively target the inflicted area while preventing its spread. For example, chemotherapy drugs are often used in combination to improve their effectiveness and reduce side effects. Likewise, precision medicine therapies that specifically target mutations that drive certain cancers may be used in combination to provide a more efficient and effective treatment.
As precision medicine continues to cement its footprint in healthcare, a window of opportunity to use machine learning to optimize combination products and improve patient outcomes. Precision medicine is the practice of tailoring medical treatment to individual patients based on their unique genetic, environmental, and lifestyle factors. By taking a more personalized approach to treatment, precision medicine can improve patient outcomes and reduce healthcare costs.
How does Machine Learning Provide Value
Machine learning is a subset of artificial intelligence that is especially useful in precision medicine because it can identify relationships and patterns in large amounts of complex data. Machine learning algorithms can be trained to analyze patient data, such as genetic and biomarker data, in order to identify the most sound choice in combination products for a particular patient. For instance, machine learning algorithms could be used to quickly, and accurately identify combinations of drugs that are most effective for patients with specific genetic mutations.
Furthermore, artificial intelligence and machine learning (AI/ML) make it possible to identifying effective combination products, machine learning can also be used to optimize dosing and reduce side effects, while maintain, increasing, or decreasing potency. Machine learning allows for medical professionals to analyze data to understand how different patients respond to different doses of drugs; machine learning algorithms can also identify the optimal dose for a particular patient. By analyzing data on the side effects of multiple drugs, machine learning algorithms can clearly pinpoint the combinations of drugs that are less likely to cause adverse effects to patients in specific scenarios.
While the use of combination products and machine learning in precision medicine is still in its early stages, there are already several promising technologies that can be used to improve the quality of care and life for patients. The combination of the cancer drug pembrolizumab with the chemotherapy drug pemetrexed has been shown to work better together than either drug alone in treating cell lung cancer. A second example is a machine learning algorithm developed by the company Owkin that has helped identify a combination of two drugs that may be effective in treating pancreatic cancer.
Combination products and machine learning algorithms still have several challenges that will need to be addressed as technologies become more and more integrated. The first challenge is obtaining large amounts of high-quality data to train the machine learning algorithms in order to accurately represent data. The data must be diverse and representative of different patient populations to ensure that the treatments constructed can reach a wide variety of patients..
The second challenge is the ability and requirement to validate the accuracy, robustness, and precision of combination products making use of machine learning algorithms in clinical trials. While the data is promising early on, these approaches need to become more rigorous and numerous studies are needed to confirm their safety.
Despite these challenges, by continuous research and tailoring treatment to individual patients, we can improve patient outcomes and reduce healthcare costs in the long run. Refining these technologies will provide more effective and personalized treatments for a wide range of diseases and conditions.
Endnotes
Winstead, Edward. “Pembrolizumab Improves Survival in Advanced Triple-Negative Breast Cancer.” National Cancer Institute, 9 Aug. 2022, https://www.cancer.gov/news-events/cancer-currents-blog/2022/pembrolizumab-triple-negative-breast-cancer-improves-survival. Accessed 15 Feb. 2023.