Cabozantinib improves tumour response in advanced HCC
Treatment with the tyrosine kinase inhibitor cabozantinib led to improved tumour response, in terms of target lesion regression, alpha-fetoprotein (AFP) response, and time to disease progression compared with placebo in patients with advanced hepatocellular carcinoma (HCC) who had prior treatment with sorafenib, according to the CELESTIAL* trial presented at ESMO Asia 2018.
Previously, survival data from the CELESTIAL trial has been reported in New England Journal of Medicine, which showed that cabozantinib significantly improved both overall survival (OS; hazard ratio [HR], 0.76; p=0.005) and progression-free survival (HR, 0.44; p<0.001) compared with placebo in patients with previously treated HCC.
The current secondary analysis of tumour response revealed that cabozantinib delayed time to progression (TTP) compared with placebo (median, 5.4 vs 1.9 months, HR, 0.41, 95 percent confidence interval [CI], 0.34–0.49). [ESMO Asia 2018, abstract 148O]
“TTP is similar to PFS for both arms, indicating that median PFS values are driven primarily by events of disease progression,” said Dr Thomas Yau from the University of Hong Kong.
“Similar PFS and TTP indicated that few patients died from non-PD** causes such as toxicity ... as TTP does not count patients who died from non-PD causes,” explained invited discussant Dr Stephen Chan from The Chinese University of Hong Kong.
Additional measures of tumour response
In the phase III, double-blind CELESTIAL study, 707 HCC patients who had previously been treated with sorafenib, with ECOG performance status ≤1, and Child-Pugh score A were randomized 2:1 to receive cabozantinib 60 mg once daily or placebo.
Among the evaluable patients with at least one post-treatment scan who were included in the secondary analysis, there were more patients in the cabozantinib arm who experienced any change from baseline in the sum of diameters (SOD) of tumour lesion compared with the placebo arm (47 percent vs 11 percent) — with 8 percent vs 1 percent of patients showing a best response of ≥30 percent reduction in SOD.
Reduction from baseline in serum AFP levels, a marker for HCC, was also greater in the cabozantinib arm vs the placebo arm (median, -26 percent vs +32 percent), with more cabozantinib-treated patients achieving a ≥50 percent reduction in serum AFP (23 percent vs 5 percent).
Among patients with baseline AFP ≥20 ng/mL (whereby 20 ng/mL was considered the upper limit of normal), the proportion of cabozantinib-treated patients who experienced any reduction in AFP levels was almost doubled that in the placebo arms (53 percent vs 23 percent). Furthermore, 9 percent of the patients in the cabozatinib arm had attained normal AFP levels compared with just 3 percent in the placebo arm.
Nonetheless, Chan questioned whether AFP response is clinically relevant, although it appears to be suggestive of signal of activity in the current study.
Noting that a proportion of patients (7–8 percent) in the placebo arm also experienced AFP drop of ≥50 percent, he suggested that researchers look closer into the details to find out the reason behind this.
“[To find out if] AFP response is really associated with better outcome, the researchers need to look into the association of AFP response with OS in the carbozantinib treatment arm. I will be very surprised if it is not associated with OS,” Chan commented.