Background
Phase 3, open-label, 4-arm randomized controlled trial (SPIRIT). N=636 patients with newly diagnosed Ph+ CP-CML. Randomized to 4 arms exploring imatinib dose escalation and combinations with interferon-alpha (IFN-α2a) or cytarabine. Based on preclinical data suggesting IFN-α synergizes with imatinib to induce deeper molecular responses and immune-mediated CML control. Published NEJM 2010. Categorized as a negative trial because, despite an MR3 improvement with imatinib + IFN-α, toxicity precluded broad adoption.
Interventions and follow up
Arm A: Imatinib 400 mg PO once daily (reference arm)
Arm B: Imatinib 600 mg PO once daily (dose escalation)
Arm C: Imatinib 400 mg PO daily + pegylated IFN-α2a 90 mcg SC weekly
Arm D: Imatinib 400 mg PO daily + cytarabine 20 mg/m² SC days 14–28 of cycles 1–6
Primary endpoint: MR3 (MMR; BCR-ABL IS ≤0.1%) at 12 months
Median follow-up: 12 months (primary); 36 months (updated analysis)
Arm B: Imatinib 600 mg PO once daily (dose escalation)
Arm C: Imatinib 400 mg PO daily + pegylated IFN-α2a 90 mcg SC weekly
Arm D: Imatinib 400 mg PO daily + cytarabine 20 mg/m² SC days 14–28 of cycles 1–6
Primary endpoint: MR3 (MMR; BCR-ABL IS ≤0.1%) at 12 months
Median follow-up: 12 months (primary); 36 months (updated analysis)
Results
MR3 at 12 months (Arm C vs Arm A): 30.1% vs 14.5%, P<.001 (significant)
MR3 at 12 months (Arm B vs Arm A): 18.1% vs 14.5% (not significant)
MR3 at 12 months (Arm D vs Arm A): 17.9% vs 14.5% (not significant)
CCyR at 12 months: Similar across arms (~60–65%)
Treatment discontinuation (Arm C): 48% by 12 months, high toxicity-driven dropout
MR3 at 12 months (Arm B vs Arm A): 18.1% vs 14.5% (not significant)
MR3 at 12 months (Arm D vs Arm A): 17.9% vs 14.5% (not significant)
CCyR at 12 months: Similar across arms (~60–65%)
Treatment discontinuation (Arm C): 48% by 12 months, high toxicity-driven dropout
Adverse events
IFN combination (Arm C): Grade ≥3 AEs 61%, dominated by flu-like symptoms any grade 80%, depression 34%, grade ≥3 neutropenia 25%; quality of life significantly impaired
Other arms: Arm A grade ≥3 AEs 16%; Arm D (imatinib + Ara-C) grade ≥3 AEs 28% with cytarabine-related cytopenias; Arm B (imatinib 600) more GI toxicity (nausea, diarrhea)
Other arms: Arm A grade ≥3 AEs 16%; Arm D (imatinib + Ara-C) grade ≥3 AEs 28% with cytarabine-related cytopenias; Arm B (imatinib 600) more GI toxicity (nausea, diarrhea)
Conclusions
Imatinib + IFN-α2a achieved a statistically higher MR3 rate at 12 months versus imatinib alone (30.1% vs 14.5%) but with 61% grade ≥3 toxicity and 48% treatment discontinuation by 12 months. The modest molecular response improvement was not clinically adopted. Imatinib 600 mg and imatinib + cytarabine conferred no significant benefit.
Key Limitations
The MR3 benefit of imatinib + IFN, though statistically significant, was offset by unacceptable toxicity, the defining clinical lesson. High discontinuation in the IFN arm (48%) undermines interpretation of response data (per-protocol vs ITT divergence). Low-dose pegylated IFN was used; other doses might shift the benefit-risk profile. The trial predates 2nd- and 3rd-generation TKIs achieving far higher MR3 rates with better tolerability (e.g., nilotinib 55–73% MR3 at 12 months in ENESTnd). IFN + imatinib is occasionally used for TFR induction in small studies but is not standard. Arm B confirmed no dose-escalation benefit.
Clinical Context
SPIRIT confirmed that imatinib + IFN-α is too toxic for routine first-line use despite a molecular response signal, effectively closing the IFN combination pathway for CML frontline management. This was validated by the subsequent era of 2nd-generation TKIs (nilotinib, dasatinib) achieving much higher MR3/MR4 rates without IFN toxicity. IFN-α is still used in some specialized TFR-induction protocols but this is investigational. Imatinib 400 mg remains the benchmark comparator for CML frontline trials. ESMO-MCBS: not assigned.