Researchers at the Johns Hopkins Kimmel Cancer Center and the Johns Hopkins University School of Medicine have devised a method to identify which patients with metastatic triple-negative breast cancer may benefit from immunotherapy, utilizing computational tools. Their findings were published on October 28 in the Proceedings of the National Academy of Sciences.
Immunotherapy aims to enhance the immune system’s ability to target and destroy cancer cells. However, not all patients respond to this type of treatment, as noted by the lead author of the study, Dr. Theinmozhi Arulraj, a postdoctoral fellow at Johns Hopkins: “It’s crucial to pinpoint patients who are likely to benefit since these treatments can be highly toxic.”
Research has been conducted to determine if certain cell types or specific molecules expressed in tumors can signal whether a patient might respond to immunotherapy. These molecules, called predictive biomarkers, are essential for determining the most suitable treatment for patients, according to senior author Dr. Aleksander Popel, a professor of biomedical engineering and oncology at Johns Hopkins University School of Medicine.
“Unfortunately, current predictive biomarkers are not very accurate at identifying patients who will see positive results from immunotherapy,” Popel explains. “Additionally, assessing various characteristics that predict treatment response would necessitate collecting tumor biopsies and blood samples from many patients, along with conducting several assays, which is a significant challenge.”
The research team used a mathematical technique known as quantitative systems pharmacology to create 1,635 virtual patients with metastatic triple-negative breast cancer and conducted treatment simulations using the immunotherapy drug pembrolizumab. They analyzed this data with advanced computational methods, including statistical and machine-learning approaches, to discover biomarkers that can reliably predict treatment responses, concentrating on which patients would respond positively or negatively.
Through the analysis of the synthetic data generated by their virtual trial, the researchers evaluated 90 biomarkers individually and in various combinations. They found that pretreatment biomarkers, which are obtained before therapy begins, had limited effectiveness in predicting patient outcomes. In contrast, biomarkers measured during treatment (on-treatment biomarkers) were more effective in forecasting results. Interestingly, certain commonly utilized biomarkers, like the expression of PD-L1 and the presence of lymphocytes in tumors, showed better predictive power when measured before treatment rather than after it commenced.
The team also examined non-invasive measurements, such as immune cell counts in the bloodstream, to see if they could effectively predict treatment outcomes, finding that some blood-based biomarkers were as effective as tumor or lymph node-based biomarkers in identifying patients who would respond to treatment. This could signify a less invasive method for predicting treatment efficacy.
Changes in tumor size can be captured via CT scans and may also serve as predictive indicators, Popel suggests: “Measuring this within the first two weeks of starting treatment shows great promise in determining who might benefit from continuing therapy.”
To confirm their findings, the researchers conducted a virtual clinical trial with patients based on the change in tumor size two weeks after treatment initiation. “The simulated response rates more than doubled—from 11% to 25%—which is quite significant,” Arulraj comments. “This highlights the potential of non-invasive biomarkers as alternatives when obtaining tumor biopsy samples is impractical.”
“Predictive biomarkers are vital as we develop enhanced strategies for treating triple-negative breast cancer to avoid giving unnecessary treatment to those who are unlikely to benefit, while also ensuring that those who do not respond well to immunotherapy receive appropriate care,” states co-author Dr. Cesar Santa-Maria, an associate professor of oncology and a breast cancer medical oncologist at Johns Hopkins Kimmel Cancer Center. “Understanding the complexities of the tumor microenvironment makes discovering biomarkers in the clinical setting challenging, but technologies that use in-silico modeling can help capture these complexities and assist in choosing the right patients for therapy.”
Overall, these new insights provide a clearer understanding of how to optimally select patients with metastatic breast cancer for immunotherapy. The researchers believe these findings will inform the design of future clinical trials and could potentially be applied to other types of cancer.
Previously, the team created a computational model focusing specifically on late-stage breast cancer, where the tumor has metastasized to different areas of the body. This research was published in Science Advances last year and utilized data from several clinical and experimental studies for thorough validation.
This current study received support from the National Institutes of Health (grant R01CA138264) and was partly conducted at the Advanced Research Computing at Hopkins facility, funded by the National Science Foundation under grant OAC1920103.
Co-authors of the study include Hanwen Wang, Atul Deshpande, Ravi Varadhan, Elizabeth Jaffee, and Elana Fertig from Johns Hopkins, along with Leisha Emens from Kaiser Permanente in South Sacramento, California.
Dr. Popel serves as a consultant for Incyte and J&J/Janssen, as well as being a co-founder and consultant for AsclepiX Therapeutics, and he receives research funding from Merck. These agreements are managed by The Johns Hopkins University in line with its conflict-of-interest policies.
