Abstract
According to a meta-analysis of 351 phase I trials, choosing a therapy based on the abnormalities that drive a tumor yields higher response rates than therapeutic selections that aren't biomarker-driven.
Choosing a therapy based on the abnormalities that drive a tumor, even in phase I clinical trials, is an independent predictor of response, according to an analysis presented at the annual meeting of the American Society of Clinical Oncology, June 3–7 in Chicago, IL.
“The usefulness of a biomarker-driven approach for treatment selection is still a matter of debate in the community,” said Maria Schwaederle, PharmD, of the Center for Personalized Cancer Therapy at the University of California, San Diego. “In the past, it was thought that phase I trials were not about efficacy and that their only purpose was to understand side effects. Our analysis really shows that this belief is completely outdated and that with biomarker selection—and especially with genomic biomarkers—we can reach high response rates.”
Schwaederle and her team conducted a meta-analysis of 351 phase I studies, involving 13,203 patients in total, that were published between January 1, 2011, and December 31, 2013. They compared efficacy endpoints such as response rates and progression-free survival in trials that involved personalized therapy, defined as using a biomarker-driven approach, with those that did not use personalized therapy.
Biomarker-selected treatment was significantly associated with higher response rates and progression-free survival, Schwaederle said. In the 58 studies that employed a personalized approach— for example, giving dabrafenib (Tafinlar; GlaxoSmithKline) to patients with BRAF mutations, or crizotinib (Xalkori; Pfizer) to patients with ALK+ tumors—the median response rate was 31% compared with just 5% in the trials that didn't consider biomarkers. The progression-free survival was 5.7 months among patients who received biomarker-based therapies compared with 2.95 months for those whose treatment wasn't biomarker-driven.
Could the difference be due to the use of targeted therapies? No, said Schwaederle. Comparing the personalized and nonpersonalized approaches in trials of targeted therapies, the researchers found response rates of 31.1% and 5.1%, respectively. “It's not just that targeted therapies are better, but that targeted therapies must be given to the right patients,” noted Schwaederle.
Of note, matching patients to therapy based on genomic biomarkers led to higher response rates than protein biomarkers—42% vs. 22.4%, respectively—a counterintuitive result. “It's really hard to explain this because the target, in the end, is a protein,” Schwaederle said. “We would expect to see better results with protein expression,” but understanding why this didn't happen “would require study on a case-by-case basis.”
“Precision medicine is here,” said Don Dizon, MD, who moderated a session at which the findings were presented. “We can use patient selection, using either changes in the tumor DNA, called genomic changes, or protein biomarkers, and do much better than we have done in the past.” –Suzanne Rose
