Background
CELESTIAL: phase 3, double-blind, placebo-controlled trial in 707 patients with advanced hepatocellular carcinoma, Child-Pugh A, previously treated with sorafenib and up to two prior systemic therapies, randomized 2:1 to cabozantinib or placebo.
Interventions and follow up
Arm A: Cabozantinib 60 mg PO daily
Arm B: Placebo
Primary endpoint: OS
mFollow up: NR
Arm B: Placebo
Primary endpoint: OS
mFollow up: NR
Results
OS: 10.2mo vs 8.0mo (Arm A vs B); HR 0.76, 95% CI 0.63-0.92; P=.005
PFS: 5.2mo vs 1.9mo (Arm A vs B); HR 0.44, 95% CI 0.36-0.52; P<.001
PFS: 5.2mo vs 1.9mo (Arm A vs B); HR 0.44, 95% CI 0.36-0.52; P<.001
Adverse events
Overall: Grade ≥3 events 68% vs 36% (Arm A vs B)
Dermatologic/vascular: Palmar-plantar erythrodysesthesia 17% vs 0%; hypertension 16% vs 2% (Arm A vs B)
Hepatic/constitutional/GI: Increased AST 12% vs 7%; fatigue 10% vs 4%; diarrhea 10% vs 2% (Arm A vs B)
Dermatologic/vascular: Palmar-plantar erythrodysesthesia 17% vs 0%; hypertension 16% vs 2% (Arm A vs B)
Hepatic/constitutional/GI: Increased AST 12% vs 7%; fatigue 10% vs 4%; diarrhea 10% vs 2% (Arm A vs B)
Conclusions
Cabozantinib improved OS and PFS versus placebo in previously treated advanced hepatocellular carcinoma.
Key Limitations
High rate of grade ≥3 toxicity with frequent dose reductions; Child-Pugh A only; mixed second/third-line population; conducted prior to first-line immunotherapy era.
Clinical Context
CELESTIAL led to FDA and EMA approval of cabozantinib for previously treated HCC (2019). ASCO/ESMO list cabozantinib among second-line options after prior systemic therapy in Child-Pugh A disease.
References
Abou-Alfa GK et al, NEJM, 2018, PMID: 29972759
Bruix J et al, Lancet, 2016 (RESORCE), PMID: 27932229
Bruix J et al, Lancet, 2016 (RESORCE), PMID: 27932229