Background
Phase 3, international (175 sites, 29 countries), open-label, randomized (1:1). N=563 adults with ES-SCLC (ECOG 0–1) without progression after 4 cycles of platinum-etoposide, ≥3 concurrent with durvalumab; asymptomatic treated/untreated brain mets allowed (19.7% brain, 32.5% liver mets). Tarlatamab is a DLL3xCD3 bispecific T-cell engager. Stratified by brain/liver mets, response to induction (PR/CR vs SD), and platinum type.
Results
Interventions and follow up: Arm A: tarlatamab (1 mg step-up D1 C1, then 10 mg D8, D15 C1, then q2wk) + durvalumab 1500 mg q4wk (n=282)
Arm B: durvalumab 1500 mg q4wk (n=281)
Primary endpoint: OS from start of maintenance
Key secondary: investigator-assessed PFS (RECIST 1.1)
mFollow up: 15.9mo (Arm A) vs 16.1mo (Arm B); prespecified interim analysis, data cutoff July 7, 2026 (243 deaths)
Results: OS: median 21.1 (95% CI 20.8–NE) vs 12.7mo (11.2–14.8), HR 0.42 (95% CI 0.32–0.54), P<.001 (crossed interim boundary P<.008)
OS 6mo / 12mo: 91.5% / 79.4% vs 74.6% / 53.3%
PFS: median 7.4 (5.6–9.1) vs 1.9mo (1.9–2.0), HR 0.34 (95% CI 0.28–0.42), P<.001
PFS 6mo / 12mo: 53.4% / 38.4% vs 17.2% / 9.5%
ORR (measurable disease; n=235 vs 244): 26.4% (95% CI 20.9–32.5) vs 4.1% (2.0–7.4), RR 6.4 (3.4–12.2), P<.001
Median DOR: 12.7mo (7.4–NE) vs 5.6mo (3.6–NE)
Subgroups: benefit across prespecified subgroups incl. brain/liver mets (underpowered)
Arm B: durvalumab 1500 mg q4wk (n=281)
Primary endpoint: OS from start of maintenance
Key secondary: investigator-assessed PFS (RECIST 1.1)
mFollow up: 15.9mo (Arm A) vs 16.1mo (Arm B); prespecified interim analysis, data cutoff July 7, 2026 (243 deaths)
Results: OS: median 21.1 (95% CI 20.8–NE) vs 12.7mo (11.2–14.8), HR 0.42 (95% CI 0.32–0.54), P<.001 (crossed interim boundary P<.008)
OS 6mo / 12mo: 91.5% / 79.4% vs 74.6% / 53.3%
PFS: median 7.4 (5.6–9.1) vs 1.9mo (1.9–2.0), HR 0.34 (95% CI 0.28–0.42), P<.001
PFS 6mo / 12mo: 53.4% / 38.4% vs 17.2% / 9.5%
ORR (measurable disease; n=235 vs 244): 26.4% (95% CI 20.9–32.5) vs 4.1% (2.0–7.4), RR 6.4 (3.4–12.2), P<.001
Median DOR: 12.7mo (7.4–NE) vs 5.6mo (3.6–NE)
Subgroups: benefit across prespecified subgroups incl. brain/liver mets (underpowered)
Adverse events
Grade ≥3 (any): 46.8% vs 40.6%
Serious AEs: 41.8% vs 31.7%
Fatal AEs: 4.6% vs 4.7%
CRS: 49.6% vs 1.4%; mostly G1 (41.8%) or G2 (7.4%), one G3 (0.4%), no G4–5; median onset 22.5h after last dose; tocilizumab 3.9%, glucocorticoids 21.6%
ICANS: 1.1% (one G2; none ≥G3); broader neurologic events 9.6% vs 6.8% (G≥3 3.2% vs 1.4%)
Dysgeusia: 48.9% vs 2.5%
Hematologic G≥3: decreased neutrophil and white-cell count most common (tarlatamab arm)
Discontinuation: tarlatamab 5.7%; durvalumab 5.7% vs 5.8%
Serious AEs: 41.8% vs 31.7%
Fatal AEs: 4.6% vs 4.7%
CRS: 49.6% vs 1.4%; mostly G1 (41.8%) or G2 (7.4%), one G3 (0.4%), no G4–5; median onset 22.5h after last dose; tocilizumab 3.9%, glucocorticoids 21.6%
ICANS: 1.1% (one G2; none ≥G3); broader neurologic events 9.6% vs 6.8% (G≥3 3.2% vs 1.4%)
Dysgeusia: 48.9% vs 2.5%
Hematologic G≥3: decreased neutrophil and white-cell count most common (tarlatamab arm)
Discontinuation: tarlatamab 5.7%; durvalumab 5.7% vs 5.8%
Conclusions
Adding tarlatamab to durvalumab maintenance after first-line chemoimmunotherapy significantly improved OS (HR 0.42), PFS (HR 0.34) and ORR in ES-SCLC. CRS was frequent but mostly low grade, with ICANS rare. Strong candidate for a new first-line maintenance standard.
Key Limitations
Interim analysis with immature OS (median NE for upper CI; 15.9mo follow-up); open-label; comparator is durvalumab alone, not lurbinectedin + atezolizumab (IMforte) so no direct comparison; only patients without progression after induction enrolled; limited post-progression tarlatamab access in control arm (OS gain may partly reflect this being earlier use); PFS investigator-assessed; PRO analyses not formally tested and had declining data availability; Black patients underrepresented (1.4%); higher SAEs and hospitalization (43.3% vs 32.0%).
Clinical Context
Standard ES-SCLC first-line is platinum-etoposide + atezolizumab or durvalumab (IMpower133, CASPIAN) with PD-L1 maintenance; IMforte added lurbinectedin to atezolizumab maintenance (OS 13.2mo). Tarlatamab is approved as second-line SCLC (DeLLphi-304 OS benefit). First-line maintenance approval for this combination is not claimed here. Ongoing: DeLLphi-306 (LS-SCLC post-CRT), DeLLphi-312 (1L tarlatamab + durvalumab + chemo), PrE0512.
