The new math in clinical studies
ST. LOUIS, Mo.—Certara's Pharsight Consulting Services has developed a mathematical model of tumor growth inhibition, which when combined with baseline prognostic factors, predicts treatment effect with bevacizumab for patients with metastatic colorectal cancer. These results are now published online in the Journal of Clinical Oncology.
In an accompanying editorial, Dr. Michael Maitland, assistant professor of medicine at the University of Chicago Medicine, and his co-authors say they have identified time-to-tumor growth (TTG) as the "best metric to predict overall survival in metastatic colorectal cancer patients treated in the study." They predict that "if TTG proves a robust endpoint in this clinical setting, one advantage will be the reduced follow-up time needed for each patient on trial, and consequently, trials with this endpoint would likely have reduced costs and reach their conclusions sooner than more conventional PFS-based studies."
The Certara researchers estimated several tumor-size response metrics using longitudinal tumor-size models and data from two Phase III clinical trials, which compared bevacizumab with chemotherapy versus chemotherapy alone as first-line therapy for colorectal cancer. Trial participants included 923 Western and 203 Chinese patients. Baseline prognostic factors and the tumor-size metric estimates were assessed in multivariate models to predict overall survival. Multiple simulations of the Phase III studies were used to test the models' predictive capabilities.
Time-to-tumor growth proved to be the best metric for predicting overall survival. The proposed model worked equally well when predicting overall survival rates for the Western and Chinese patients, and could be used to support drug development decisions in either population.
Dr. René Bruno, managing director of Certara's Pharsight Consulting Services Europe and a senior author of the paper, says, "This approach of combining modeling with longitudinal tumor-size data may contribute to improved design and analysis of more informative early-stage clinical studies (Phase Ib, II). It could also enable researchers to select the most promising treatments and reduce the high attrition rate in Phase III oncology studies."
Dr. Robert Powell, former senior advisor at Roche China, and a co-author of the paper, adds, "It is important to know whether patients in a new market (e.g., China, India, Brazil) will be similar to patients in the original U.S. or EU New Drug Application (NDA). While pharmaceutical companies usually assume patients from different regions are the same, there is emerging evidence that they might be different with regard to efficacy, safety and dose response. This type of analysis helps better define Chinese colon cancer response relative to Western patients. Roche performed this combined Chinese and U.S. NDA study analysis to learn whether Chinese patients responded similarly to Western patients so they can use this information to plan future trials. Likewise, knowing these results will be important to local regulatory agencies such as the China Food and Drug Administration."
In his editorial, Maitland points out that "extensive investment in oncology drug discovery and development during the past decade has not been matched by similar innovation in clinical trial design during the same period, especially with regard to endpoint evaluation." He notes that alternative metrics to progression-free survival (PFS) for detecting the beneficial effects of cancer therapies are now an active area of research. "Innovations in clinical trial design could accelerate availability of effective new drugs and reduce the rate of failure in expensive late-phase development, and therefore reduce the overall costs of oncology drug development," he concludes.
Certara was formed by the acquisition and integration of Tripos, Simcyp and Pharsight Corp., and provides software and scientific consulting services to improve productivity and decision-making in drug discovery and development. Each Certara family brand focuses on a key phase within the drug discovery and development process; combined, they offer scientific modeling, analysis and simulation capabilities that can enable the cross-disciplinary approaches necessary for translational science initiatives.