Expert Provides Insight on Biomarker Research in Thyroid Cancer

Marcia S. Brose, MD, PhD

An analysis of patients with differentiated thyroid cancer (DTC) enrolled in the phase III DECISION trial failed to show biomarkers that were predictive of response to sorafenib (Nexavar), according to results presented at the 2016 ASCO Annual Meeting.

In the analysis, researchers examined baseline serum thyroglobulin (n = 403) and candidate plasma biomarkers (n = 395) to identify prognostic and/or predictive biomarkers in the 417 patients enrolled in the DECISION study.

Data showed that elevated thyroglobulin, VEGF-A, VEGF-C, TGF-β1, and low E-cadherin levels were associated with a poor prognosis in DTC. However, none of the biomarkers were able to predict benefit from sorafenib.

In the double-blinded DECISION trial, 417 patients with locally recurrent or metastatic progressive radioiodine (RAI)-refractory disease were randomized to receive sorafenib or placebo. Results showed that the median progression-free survival with sorafenib and placebo was 10.8 months and 5.8 months, respectively.

Moreover, 42% of patients in the sorafenib arm had stable disease for more than 6 months compared with 33% with placebo. The trial’s findings led to the FDA approval of sorafenib in this setting in November 2013.

OncLive: You’re an author on the DECISION analysis, which did not show that biomarkers determine response to sorafenib. Can you discuss these findings?

In an interview with OncLive, an investigator on the biomarker analysis and the lead author of the DECISION trial, Marcia S. Brose, MD, PhD, who is an associate professor of Otorhinolaryngology, Head and Neck Surgery, at the Hospital of the University of Pennsylvania, sheds light on the biomarker assessment and the overall developments with biomarker research in the field of DTC.Brose: Biomarkers are, of course, very interesting to everybody in all types of thyroid cancer. We wanted to find a biomarker to show who did well and didn’t do well. We actually published data from our phase II study with sorafenib (Nexavar) that patients who had higher levels of ATK in the nucleus were less likely to get responses. That is good to know and, in this case, if you had [elevated] thyroglobulin or high levels of VEGF-A, you had a poor prognosis with differentiated thyroid cancer.

Where is the field of thyroid cancer currently at regarding biomarker development?

Because these biomarkers don’t predict for response, none of them have shown that, while they are interesting and might be associated with improved outcomes in some patients, none of them have risen to the level where they would be clinically meaningful. None of them are at the point where we would select patients based on any of the biomarkers that were studied in that poster.There is a benefit in DECISION for all patients. It is true that some patients do better, some do worse, but all patients do better when compared with placebo. Since biomarkers might help us to predict who is going to have a longer or less benefit—and that might be helpful for treatment planning long-term— that might start to have a role in that way first.

None of them are to the point where we would use this therapy or not use that therapy. The only situation where that is the case is a study that we did that showed that vemurafenib (Zelboraf) is an active agent for BRAF-positive thyroid cancer, which was presented at the 2013 European Cancer Congress. That is a biomarker that is clinically meaningful.

With differentiated thyroid cancer, I do believe we should be genotyping everyone at this point because we have enough targeted therapies that, while sorafenib and lenvatinib (Lenvima) don’t discriminate based on biomarkers, they are beneficial to all patients. Additional agents will be available that may target specific secondary or other mutations in tumors.

We routinely genotype patients when we treat them. They will do well with sorafenib and lenvatinib but, with a BRAF mutation, I know I have another option. We are at the point where we think that biomarkers are important. We are not at the point where they are FDA-approved agents.

However, they are certainly active and would be viable options for patients after the FDA options are used up. Since most of these patients are fairly healthy, many of them will be ready and able and need third- and fourth-line therapies. The biomarkers will probably help us to find options for them in the third- and fourth- line settings that they otherwise would not have.

Pena CE, Wilhelm S, Meinhardt G, Brose MS, et al. Biomarkers of prognosis in patients with differentiated thyroid cancer: results from the DECISION trial. J Clin Oncol. 2016;34 (suppl; abstr 6059).