Background
Phase 3, randomized, open-label trial. 219 patients with intermediate-2 or high-risk primary myelofibrosis (MF), post-polycythemia vera MF, or post-essential thrombocythemia MF. Randomized approximately 2:1 (ruxolitinib ~146, BAT ~73). The pivotal European trial for ruxolitinib, conducted simultaneously with COMFORT-I.
Interventions and follow up
Arm A: Ruxolitinib orally twice daily (dose based on platelet count; n≈146)
Arm B: Best available therapy (BAT; investigator's choice: hydroxyurea, glucocorticoids, anagrelide, immunomodulatory drugs, other; n≈73)
Primary endpoint: Spleen volume reduction ≥35% at week 48 (by MRI/CT)
mFollow up: Median 12 months at primary analysis
Arm B: Best available therapy (BAT; investigator's choice: hydroxyurea, glucocorticoids, anagrelide, immunomodulatory drugs, other; n≈73)
Primary endpoint: Spleen volume reduction ≥35% at week 48 (by MRI/CT)
mFollow up: Median 12 months at primary analysis
Results
SVR ≥35% at week 48 (SVR35): 28% vs 0%, P<.001
SVR ≥35% at week 24: 32% vs 0%, P<.001
Median palpable spleen length change: −56% (ruxolitinib) vs +4% (BAT)
Median duration of response: not reached; 80% of responders still responding at median 12-month follow-up
QoL and symptom burden: improved in ruxolitinib arm
SVR ≥35% at week 24: 32% vs 0%, P<.001
Median palpable spleen length change: −56% (ruxolitinib) vs +4% (BAT)
Median duration of response: not reached; 80% of responders still responding at median 12-month follow-up
QoL and symptom burden: improved in ruxolitinib arm
Adverse events
Hematologic (grade 3–4): thrombocytopenia and anemia were the principal AEs, managed with dose reduction/interruption; 1 patient per arm discontinued for thrombocytopenia, none for anemia
Nonhematologic: rare and mostly grade 1–2
Transformation: 2 AML cases occurred, both in the BAT arm
Nonhematologic: rare and mostly grade 1–2
Transformation: 2 AML cases occurred, both in the BAT arm
Conclusions
Ruxolitinib produced marked, durable reductions in splenomegaly versus BAT in intermediate-2/high-risk myelofibrosis, with improved quality of life and manageable hematologic toxicity. Combined with COMFORT-I, these two trials supported the global regulatory approval of ruxolitinib as the first disease-modifying therapy for myelofibrosis.
Key Limitations
OS improvement was not demonstrated at primary analysis (too few events; follow-up too short). Open-label design (versus the blinded COMFORT-I) may introduce assessment bias for patient-reported outcomes. BAT was heterogeneous and observer-dependent. The 48-week primary endpoint (vs 24 weeks in COMFORT-I) was more conservative; absolute response rates were lower (28% vs 42%), partly reflecting longer follow-up and attrition. Extended (3–5 year) analyses later confirmed OS benefit.
Clinical Context
COMFORT-II, together with COMFORT-I, supported simultaneous FDA and EMA approval of ruxolitinib for MF. Long-term follow-up (3.5 years) showed significantly prolonged OS with ruxolitinib (median not reached vs 4.1 years, HR 0.48). These trials collectively established the rationale for JAK inhibitor therapy in MF, and ruxolitinib remains the foundational treatment against which newer agents (fedratinib, pacritinib, momelotinib) are compared, consistent with ESMO and ELN guidance.
References