Background: EP (4-6 cycles) is standard of care 1st-line therapy for metastatic SCLC, and no multinational studies have reported any improvement beyond that reported for EP. Evidence of an ongoing immune response to SCLC tumors suggests that immunotherapy that enhances this immune response to SCLC may enhance the clinical benefit of EP. Ipi, a fully human monoclonal antibody which blocks CTLA-4, augments antitumor immune responses. Because some responses to Ipi may differ from those observed with cytotoxic therapies, immune-related response criteria (irRC) were derived from WHO criteria to better capture response patterns observed with Ipi. A randomized Phase 2 study of Ipi with paclitaxel/carboplatin (PC) in pts with ED-SCLC showed significant improvement in progression-free survival (PFS), as measured by irRC, over PC alone in pts receiving Ipi and PC in a phased regimen (Ipi started after 2 cycles of PC). Furthermore, addition of Ipi did not exacerbate PC toxicity, and immune-related adverse events were managed using protocol-specific guidelines. This multicenter phase III study in pts with ED-SCLC (ClinicalTrials.gov identifier NCT01450761) will determine whether adding Ipi to EP increases OS vs EP alone. Methods: EP consists of 4 cycles of etoposide (100 mg/m², IV on Days 1-3 every 3 weeks [Q3W]) and cisplatin (75 mg/m², IV) or carboplatin (AUC=5, IV) once Q3W. Pts will be randomized to receive 4 doses of Ipi (10 mg/kg, IV) in Arm A or placebo in Arm B, Q3W during induction, starting after 2 cycles of EP (phased schedule). Eligible pts will then receive blinded study drug (Ipi in Arm A; placebo in Arm B) Q12W until disease progression or unacceptable toxicity. The primary objective is to compare OS. Secondary objectives are to compare OS in those who receive blinded study drug, compare PFS between study arms, and to estimate best overall response rate and duration of response. First-line ED-SCLC pts with ECOG performance ≤1 will be included. Pts with symptomatic CNS metastases or a history of autoimmune disease will be excluded. The study will randomize 1100 pts at a 1:1 ratio.