preheader BJMO

Significantly improved progression-free survival with cabozantinib in patients with previously treated radioiodine-refractory differentiated thyroid cancer

Results of the phase III pivotal COSMIC-311, evaluating single-agent cabozantinib in patients with radioiodine-refractory differentiated thyroid cancer who progressed after up to two prior vascular endothelial growth factor receptor-targeted therapies, indicate that cabozantinib significantly delays disease progression without any unexpected toxicities.

Patients with radioiodine (RAI)-refractory differentiated thyroid cancer (DTC) who progress on prior VEGFR-targeted therapies have a poor prognosis and limited therapeutic options. Cabozantinib is an inhibitor of VEGFR2, MET, AXL, and RET, which are all implicated in the pathogenesis of DTC. In line with this, cabozantinib showed clinical activity in patients with RAI-refractory DTC in early-phase studies. Based on these findings, the phase III COSMIC-311 study evaluated the efficacy and safety of cabozantinib versus placebo in patients with RAI-refractory DTC who had progressed during or after prior VEGFR-targeted therapy.

COSMIC-311 study design

In this double-blind, phase III trial, patients were randomised (2:1) to receive cabozantinib (60 mg once daily) or placebo. Patients randomised to placebo could cross over to open-label cabozantinib upon progression as confirmed by the blinded independent radiology committee (BIRC) per RECIST 1.1 criteria. In order to be eligible, patients had to have locally advanced or metastatic DTC, which is RAI-refractory or –ineligible. Patients (≥16 years) had to be pretreated with lenvatinib or sorafenib for DTC and have disease progression during or following treatment with up to two prior VEGFR inhibitors. Other inclusion criteria include an ECOG performance status of 0 to 1 and a serum thyroid-stimulating hormone level below 0.5 mIU/L. The co-primary endpoints were objective response rate (ORR) in the first 100 randomised patients and progression-free survival (PFS) in all randomised patients. A prespecified interim PFS analysis was planned for the intention to treat population at the time of the primary ORR analysis. Tumours were assessed every eight weeks for the first twelve months, then every twelve weeks per RECIST v1.1. Treatment was given until there was no longer clinical benefit or until intolerable toxicity.

Progression-free survival and objective response rates

As of August 19th, 2020, 125 patients had been randomised to the cabozantinib arm and 62 to the placebo arm. Median age of study participants was 66 year, 55% were female and 63% received prior lenvatinib.  Of note, patients in the cabozantinib arm had a higher percentage of bone and liver metastases, which are often associated with a poorer prognosis. After a median follow-up of 6.2 months, the primary endpoint of PFS was met at the planned interim analysis, with cabozantinib demonstrating a significant improvement over placebo (HR[96%CI]: 0.22[0.13-0.36], p< 0.0001). Since this interim analysis rejected the null hypothesis, this PFS is considered to be the final PFS and further patient enrolment was stopped. The median for PFS was not reached in the cabozantinib arm and was reported at 1.9 months in the placebo arm.  Cabozantinib was favoured over placebo for all pre-specified subgroups, including age below or above 65 years, prior lenvatinib treatment and the number of prior VEGFR tyrosine kinase inhibitors. The ORR was 15% for cabozantinib (all partial responses) as compared to 0% for placebo (p=0.028) but did not meet the prespecified criteria for statistical significance (p<0.01). Stable disease for at least sixteen weeks was obtained in 45% of patients in the cabozantinib arm vs. 27% in the placebo arm. At six months minimum follow-up, any reduction from baseline in target lesions was reported for 76% of patients treated with cabozantinib, as compared to only 29% of patients who received placebo. Finally, despite cross-over, a favourable OS trend was observed for cabozantinib vs. placebo (HR [95%CI]: 0.54 [0.27–1.11]).

The median duration of exposure was 4.4 months for cabozantinib and 2.3 months for placebo. Dose reductions due to any grade adverse events (AEs) occurred in 57% of patients with cabozantinib and in 5% of patients with placebo. Treatment discontinuations due to AEs not related to disease progression occurred in 5% of patients with cabozantinib vs. 0% with placebo. Treatment-emergent adverse events (TEAEs) of any grade with higher occurrences in the cabozantinib vs. placebo arm included diarrhoea (51% vs. 3%), hand-foot skin reaction (46% vs. 0%), hypertension (28% vs. 5%), fatigue (27% vs. 8%), and nausea (24% vs. 2%). Grade 3/4 AEs were experienced by 57% of patients with cabozantinib as compared to 26% of patients with placebo. No treatment-related deaths occurred in either arm.


Cabozantinib demonstrated a clinically and statistically significant improvement in PFS over placebo in patients with RAI-refractory DTC after prior VEGFR-targeted therapy without inducing unexpected toxicities. The ORR favoured cabozantinib over placebo, although this difference was not statistically significant. These results support cabozantinib as a potential new treatment option for previously treated RAI-refractory DTC.


Brose MS, et al. Cabozantinib versus placebo in patients with radioiodine-refractory differentiated thyroid cancer who have progressed after prior VEGFR-targeted therapy: Results from the phase 3 COSMIC-311 trial. Presented at ASCO 2021; Abstract 6001.

Speaker Marcia S. Brose

Marcia S. Brose

Marcia S. Brose, MD, PhD, FASCO, Abramson Cancer Center, University of Pennsylvania, Philadelphia, USA


See: Keyslides

Back to Top