Background
Phase III, double-blind, placebo-controlled RCT. 292 patients with chronic liver disease (CLD) and thrombocytopenia (platelet count 40–80 × 10⁹/L) scheduled for elective invasive procedures. Included cirrhosis from any etiology. Excluded patients with a history of portal vein thrombosis (PVT).
Interventions and follow up
Arm A: Eltrombopag 75 mg/day orally for up to 14 days pre-procedure
Arm B: Placebo
Primary endpoint: Proportion avoiding platelet transfusion before and during procedure
mFollow up: 28 days post-procedure
Arm B: Placebo
Primary endpoint: Proportion avoiding platelet transfusion before and during procedure
mFollow up: 28 days post-procedure
Results
Transfusion avoidance: 72% vs 19%, P<.001
Portal vein thrombosis: 6 (4%) vs 1 (1%), P=.09 — trial stopped early by DSMB
Platelet count increase ≥20×10⁹/L: 77% vs 12%
Portal vein thrombosis: 6 (4%) vs 1 (1%), P=.09 — trial stopped early by DSMB
Platelet count increase ≥20×10⁹/L: 77% vs 12%
Adverse events
Main adverse events: Grade ≥3 AEs: 18% vs 16%. PVT occurred in 6 patients in eltrombopag arm vs 1 placebo — driven by this imbalance, the trial was stopped early by the Data Safety Monitoring Board. Thrombosis at other sites also numerically higher in eltrombopag arm. No increase in major bleeding.
Conclusions
Eltrombopag significantly increased platelet counts and reduced transfusion need in CLD patients pre-procedure; however, it was stopped early due to a 4-fold excess of portal vein thrombosis in the treatment arm. This safety signal precluded FDA approval for this indication.
Key Limitations
Key Limitations: Trial terminated early — underpowered for definitive PVT risk assessment, but signal was strong enough to halt development. Cirrhotic patients have complex hemostasis (pro- and anticoagulant factor imbalances) that TPO agonists may disrupt unpredictably. The PVT excess may reflect the known prothrombotic state of cirrhosis exacerbated by platelet count elevation. Comparator trials (ADAPT-1/2 with avatrombopag) did not show PVT excess, possibly due to drug-specific differences or lower platelet targets.
Clinical Context
Eltrombopag is NOT approved for CLD-related thrombocytopenia due to PVT risk. In contrast, avatrombopag (ADAPT-1/2) and lusutrombopag (L-PLUS 2) are FDA approved for this indication. This trial is board-critical: it illustrates why drug class is not a class effect — different TPO agonists have different thrombotic profiles in cirrhosis. ESMO-MCBS not applicable.
References
References: Afdhal NH et al, NEJM 2012 (ELEVATE)