The clinical response to regorafenib does not depend on tumor mutations. Among patients with metastatic colorectal cancer who participated in the phase 3 CORRECT (Colorectal Cancer Treated With Regorafenib or Placebo After Failure of Standard Therapy) study, an analysis of tumor specimens for KRAS and PIK3CA mutations did not predict clinical benefit in the patients assigned to regorafenib compared with placebo, said Michael Jeffers, PhD. He presented the results of the study at the 2013 Gastrointestinal Cancers Symposium.
“There was a trend toward regorafenib benefit regardless of KRAS mutational status,” said Jeffers, lead investigator of the analysis. “We found the same thing when we looked at PIK3CA mutations.”
He and colleagues conducted a retrospective biomarker analysis on DNA isolated from baseline plasma and archival tumor tissue specimens to determine if a correlation existed between mutational status and the clinical response to regorafenib in the CORRECT trial. The KRAS, PIK3CA, and BRAF genes were analyzed.
Mutational analysis of plasma showed mutant KRAS in 69% of samples and mutant PIK3CA in 17%. Mutational analysis of tumor tissue demonstrated mutant KRAS in 59% of specimens and mutant PIK3CA in 12%. The concordance in KRAS results between archival tissue versus fresh plasma was 79%. Concordance of PIK3CA results was 88%.
The small percentage of patients who had wild-type KRAS in archival tumor tissue but mutated KRAS in fresh tumor tissue “may have acquired KRAS-mutant tumor cells following therapy with anti-epidermal growth factor receptor, a phenomenon that has been described in colorectal cancer,” said Jeffers, associate director of Clinical Oncology Biomarkers at Bayer HealthCare Pharmaceuticals, Montville, New Jersey. Mutational analysis of fresh plasma may therefore be a more accurate reflection of current mutational status, he indicated.
"The study overall showed that regardless of biomarker subgroup, regorafenib seems to have benefit,” he said. The hazard ratio for overall survival favored regorafenib in patients with mutant and wild-type KRAS and PIK3CA, consistent with the overall study population. In addition, the hazard ratio for progression-free survival was <1.0 in all biomarker subgroups that were assigned to regorafenib compared with placebo.
Owing to low patient numbers, subgroups based on BRAF mutational status were not analyzed. BRAF mutations were detected in 3.4% of the plasma samples and 1.5% of the tumor tissue samples analyzed.
“Colorectal cancer has a high prevalence of KRAS mutations as well as PIK3CA mutations as part of this disease. Since the low-hanging fruit didn’t seem to work, we’ll go on to other things,” said Jeffers. Thus far, “we haven’t come across anything to differentiate responders from nonresponders. We have the tumor samples; we can do immunohistochemistry and run expression analysis. We haven’t done either of those yet. We have looked at plasma proteins but haven’t published those data yet.”
Jeffers M, Van Cutsem E, Sobrero AF, et al. Mutational analysis of biomarker samples from the CORRECT study: correlating mutation status with clinical response to regorafenib. Presented at: 2013 Gastrointestinal Cancers Symposium; January 24-26, 2013; San Francisco, CA. Abstract 381/Poster A49.