Balancing Clinical Trials, Real-World Data to Inform Evidence-Based Cancer Care
Katherine Reeder-Hayes, MD, MBA, MSc, Lineberger Comprehensive Cancer Center, UNC Chapel Hill, described the potential role of real-world evidence as a compliment to clinical trials as well as successful implementation of Big Data for real-world evidence generation.
Dr Reeder-Hayes began her discussion by reviewing the pros and cons of clinical trials in evidence accrual. While clinical trials are often considered the gold standard for efficacy testing in new drug development, they can be expensive to fund, are not a proper format in every setting, and may contribute to wasteful resource use. On the contrary, real-world evidence may be less expensive, more appropriate to some questions, and sufficient evidence for certain contexts. Furthermore, real-world evidence may conserve research resources by helping direct clinical trials.
As for real-world evidence from Big Data, Dr Reeder-Hayes explained that it may be used after clinical trials to test whether findings are adopted, alongside clinical trials to extend findings to broader populations, and in anticipation of clinical trials to inform the important problems, quantify effect sizes, and identify the right population of patients.
Dr Reeder-Hayes then highlighted three case studies to demonstrate integration between Big Data and clinical trial evidence. In the first of these case studies, the Big Data study design utilized SEER-Medicare data from 2005-2013 to conduct two sub-studies: disparities among trastuzumab use in patients with stage I-III HER2+ breast cancers from 2010-2011, and comparative effectiveness examination of all patients with stage I-III breast cancers treated with trastuzumab from 2005-2011. Researchers found that patterns of care changed dramatically in favor of TCH (docetaxel, carboplatin, and trastuzumab) compared with ACTH (doxorubicin, cyclophosphamide, paclitaxel, and trastuzumab) after presentation of the BCIRG 006 trial. Additionally, real-world studies uncovered racial disparities among the proportion of patients receiving trastuzumab.
In her concluding remarks, Dr Reeder-Hayes stressed the importance of a few takeaway messages: real-world evidence from Big Data can be used before, alongside, or after clinical trials to define the problem, extend the evidence base to specific populations, and improve clinical trial designs; and real-world evidence from Big Data requires expertise in handling and analyzing observational data with its unique challenges
Similarly to clinical trials, studies are best performed by experienced cross-disciplinary teams, she added. Studies are also most useful when they provide answers to the questions that are important to both doctors and patients.—Zachary Bessette