Using computational tools, researchers have developed a novel method to assess which patients with metastatic triple-negative breast cancer may benefit from immunotherapy, according to a recent study published by Arulraj et al in the Proceedings of the National Academy of Sciences (PNAS).
Background
Immunotherapy is typically administered to try to boost the body’s own immune system to attack cancer cells; however, only some patients respond to treatment. “It’s really important that we identify those patients for whom it will work, because the toxicity of these treatments is high,” stressed lead study author Theinmozhi Arulraj, PhD, a postdoctoral fellow at Johns Hopkins University.
To identify patients likely to respond to immunotherapy, previous studies have examined whether the presence or absence of certain cells—or the expression of various molecules in the tumor—can indicate response to immunotherapy. These predictive biomarkers are useful in selecting the most appropriate treatment options.
“Unfortunately, existing predictive biomarkers have limited accuracy in identifying patients who will benefit from immunotherapy,” explained senior study author Aleksander Popel, PhD, Professor of Biomedical Engineering and Oncology at the Johns Hopkins University School of Medicine. “Moreover, a large-scale assessment of characteristics that predict treatment response would require the collection of tumor biopsies and blood samples from many patients and would involve performing several assays, which is very challenging,” he added.
In a prior study, published by Arulraj et al in Science Advances, the researchers used an in-house modeling framework and developed a computational model with a special focus on late-stage breast cancer.
Study Methods and Results
In the recent study, the researchers employed a mathematical model known as quantitative systems pharmacology to generate 1,635 virtual patients with metastatic, triple-negative breast cancer. They then performed treatment simulations with the immunotherapy agent pembrolizumab and fed the data into powerful computational tools—including statistical and machine learning–based approaches—with the goal of determining biomarkers that accurately predict treatment response. The researchers focused on identifying which patients would and would not respond to treatment.
Using the partially synthetic data produced by the virtual clinical trial, the researchers evaluated the performance of 90 biomarkers alone and in double, triple, and quadruple combinations. They found that pretreatment biomarkers from tumor biopsies or blood samples taken prior to treatment initiation had limited ability to predict treatment outcomes. However, on-treatment biomarkers taken following the initiation of treatment seemed to be more predictive of outcomes. Of note, the researchers discovered that some commonly used biomarker measurements such as the expression of the PD-L1 molecule and the presence of lymphocytes in the tumor performed more effectively when assessed prior to treatment initiation compared with after treatment initiation.
Further, the researchers analyzed the accuracy of measurements that did not require invasive biopsies—including immune cell counts in the blood—in predicting treatment outcomes. They revealed that some blood-based biomarkers performed comparably to tumor- or lymph node–based biomarkers in identifying a subset of patients who responded to treatment, suggesting a potentially less-invasive approach to predict treatment response. They indicated that measurements of changes in tumor diameter can be readily obtained by computed tomography scans also may prove predictive. “This, measured very early within 2 weeks of treatment initiation, had a great potential to identify who would respond if the treatment were continued,” emphasized Dr. Popel.
To validate the findings, the researchers performed a virtual clinical trial with patients selected based on change in tumor diameter at 2 weeks following treatment initiation. “The simulated response rates increased more than twofold—from 11% to 25%—which is quite remarkable. This emphasizes the potential for noninvasive biomarkers as an alternative, in cases where collecting tumor biopsy samples is not feasible,” underlined Dr. Arulraj.
Conclusions
Collectively, the findings may shed light on how to better select patients with metastatic breast cancer for immunotherapy. The researchers hope the results of their study can help to design future clinical studies and their novel method could be replicated in other cancer types.
“Predictive biomarkers are critical as we develop optimized strategies for triple-negative breast cancer, so as to avoid overtreatment in patients expected to do well without immunotherapy and undertreatment in those who do not respond well to immunotherapy,” detailed co–study author Cesar Santa-Maria, MD, MSCI, Associate Professor of Oncology and a breast medical oncologist at the Johns Hopkins University Sidney Kimmel Comprehensive Cancer Center. “The complexities of the tumor microenvironment make biomarker discovery in the clinic challenging, but technologies leveraging in-silico [computer-based] modeling have the potential to capture such complexities and aid in patient selection for therapy,” he underscored.
Disclosure: The research in this study was supported by the National Institutes of Health and in part by the National Science Foundation. For full disclosures of the study authors, visit pnas.org.
