Abstract
Background
In the DeLLphi-304 trial, at a median follow-up of <12 months, tarlatamab significantly improved overall survival (OS), progression-free survival (PFS), and patient-reported outcomes (PROs), and demonstrated a favorable safety profile versus chemotherapy. We report efficacy and safety outcomes after extended follow-up in patients with small-cell lung cancer (SCLC).
Patients and methods
In this phase 3, open-label trial, patients with SCLC whose disease had progressed on or after platinum-based chemotherapy were randomized (1:1) to receive either tarlatamab or chemotherapy (topotecan, lurbinectedin, or amrubicin) until disease progression, unacceptable toxicity, or death. The primary endpoint was OS. Secondary endpoints included investigator-assessed PFS, objective response, disease control, duration of response per RECIST v1.1, PROs, and safety.
Results
As of data cutoff (18 February 2026), median OS was 13.6 and 8.4 months with tarlatamab and chemotherapy, respectively (median follow-up, 23.3 and 22.8 months). Median PFS was 4.2 and 3.2 months, respectively (median follow-up, 22.2 and 21.9 months). Kaplan-Meier estimated OS rates at 12/24 months were 55.9%/30.3% with tarlatamab versus 39.0%/14.9% with chemotherapy; corresponding PFS rates at 12/24 months were 20.6%/14.2% versus 5.9%/2.3%. Objective response rate was 35.4% with tarlatamab, including six confirmed complete responses versus 20.4% with chemotherapy; at data cutoff, ongoing responses were observed in 31.1% (28/90) versus 3.8% (2/52) of responders, respectively. Sustained disease control for ≥52 weeks was observed in 24.4% versus 4.3%. Treatment-related adverse events led to treatment discontinuation in 3.2% versus 6.6%; grade ≥3 events decreased over time, with no serious TRAEs with onset ≥12 months. CRS and ICANS occurred early in treatment and were predominantly low grade.
Conclusions
With extended follow-up, tarlatamab continues to demonstrate sustained and durable clinical benefit and a favorable long-term tolerability profile versus chemotherapy, in patients with SCLC after first-line platinum-based chemotherapy, supporting its use as the standard of care in this setting.
Keywords: BiTE®, bispecific T-cell engager, delta-like ligand 3, immunotherapy, small-cell lung cancer, tarlatamab
INTRODUCTION
Tarlatamab, a delta-like ligand 3 (DLL3)-targeted bispecific T-cell engager (BiTE®) immunotherapy,1 significantly improved overall survival (OS) compared with standard chemotherapy (topotecan, lurbinectedin, or amrubicin) in patients with small-cell lung cancer (SCLC) in the second-line (2L) setting.2 At a median follow-up of <12 months in the primary analysis of DeLLphi-304 trial, the median OS achieved with tarlatamab was 13.6 months (95% confidence interval [CI], 11.1 to not reached) compared with 8.3 months with chemotherapy (95% CI, 7.0–10.2).2 Tarlatamab also demonstrated a significant progression-free survival (PFS) benefit, improved patient-reported outcomes (PROs), including dyspnea and cough, and showed a favorable safety profile versus chemotherapy.2, 3 Compared with chemotherapy, tarlatamab was associated with lower incidence of both grade ≥3 adverse events (AEs) and AEs resulting in treatment discontinuation.2 With this extended follow-up of the DeLLphi-304 study, we report the durability of clinical benefit and characterize the long-term efficacy and safety of tarlatamab versus chemotherapy.
PATIENTS AND METHODS
DeLLphi-304 was a multinational, phase 3, open-label trial comparing tarlatamab with chemotherapy in patients with SCLC progressing on or after platinum-based chemotherapy (NCT05740566).2 Patients were randomized to receive either tarlatamab or chemotherapy until disease progression, unacceptable toxicity, or death. The study design, eligibility criteria, patient stratification, treatment administration, and ethical considerations have been described previously.2 Patients aged ≥18 years with ECOG performance-status of 0–1 and stable asymptomatic brain metastasis were eligible. The primary endpoint was OS. Secondary endpoints included investigator-assessed PFS, PROs, objective response, disease control, duration of response (DOR) according to RECIST v1.1, and safety.
RESULTS
Patients
Between May 31, 2023, and July 30, 2024, 688 patients were evaluated, of whom 509 eligible patients were randomly assigned to receive either tarlatamab (n = 254) or chemotherapy (n = 255). Demographic and baseline disease characteristics of the patients have been described previously.2
As of the data cutoff date (February 18, 2026), median duration of tarlatamab exposure (range) was 18.2 weeks (0.1–129.0) and 10.9 weeks (0.1–118.0) with chemotherapy. Seventy-six patients (29.9%) in the tarlatamab group and 42 patients (16.5%) in the chemotherapy group remained in the study; 39 (15.4%) and 4 (1.6%), respectively, were still receiving treatment.
Efficacy
Median follow-up for OS was 23.3 months in the tarlatamab group and 22.8 months in the chemotherapy group. At data cutoff, 388 deaths had occurred including 176 in the tarlatamab and 212 in the chemotherapy group. Median OS with tarlatamab was 13.6 months (95% CI, 11.9–15.8) versus 8.4 months (95% CI, 7.2–10.2) with chemotherapy (stratified hazard ratio [HR], 0.62; 95% CI, 0.51–0.76; Figure 1A). Kaplan-Meier estimated OS rates at 12, and 24 months were 55.9% and 30.3%, respectively, with tarlatamab, versus 39.0% and 14.9% with chemotherapy.
Figure 1. (A-C): OS (A), PFS (B), and DOR (C) with tarlatamab versus chemotherapy at extended follow-up.
aSurvival curves and median values were estimated using Kaplan-Meier (KM) method and 95% CI of median were estimated using log-log transformation of KM survival estimate by Brookmeyer and Crowley (1982) method.8
bHazard ratios and 95% CIs were estimated using a stratified Cox proportional hazards model; a hazard ratio <1.0 indicates a better survival of tarlatamab arm.
Tick marks indicate censored observations. Total number of patients censored: (A) tarlatamab, n = 78; chemotherapy, n = 43; (B) tarlatamab, n = 43; chemotherapy, n = 37; (C) tarlatamab, n = 29; chemotherapy, n = 4.
Abbreviations: CI, confidence interval; DOR, duration of response; FU, follow-up; HR, hazard ratio; OS, overall survival; PFS, progression-free survival.
Median follow-up for PFS was 22.2 months in the tarlatamab group and 21.9 months in the chemotherapy group. Median PFS was 4.2 months (95% CI, 3.0–4.4) with tarlatamab and 3.2 months (95% CI, 2.9–4.2) with chemotherapy (stratified HR, 0.69; 95% CI, 0.57–0.84; Figure 1B). Kaplan-Meier estimated PFS rates at 12, and 24 months were 20.6% and 14.2%, respectively, with tarlatamab, compared with 5.9% and 2.3%, with chemotherapy. Prespecified subgroup analyses of OS and PFS are shown in Figure S1 and S2. The proportion of patients who received at least one subsequent anticancer therapy was 53.2% in tarlatamab and 57.0% in chemotherapy. None in the tarlatamab group received subsequent DLL3-directed therapy versus 11.1% in the chemotherapy group, including 10.2% who received tarlatamab (Table S1). Investigator-assessed objective response was observed in 90 patients (35.4%) in the tarlatamab group, including 6 confirmed complete responses and 84 partial responses, versus 52 patients (20.4%) with partial responses in the chemotherapy group (Table 1; Figure 1C). Median DOR was 7.1 months (95% CI, 4.5–12.4) in tarlatamab group and 5.5 months (4.2–5.7) in chemotherapy group. The Kaplan-Meier estimated DOR at 12 months among confirmed responders was 41.6% and 10.1% for tarlatamab and chemotherapy, respectively. Sustained disease control for ≥52 weeks was observed in 62 patients (24.4%) with tarlatamab and in 11 patients (4.3%) with chemotherapy (Table 1). At data cutoff, responses remained ongoing in 28 of 90 responders (31.1%) in the tarlatamab group and 2 of 52 responders (3.8%) in the chemotherapy group. The incidence of anti-tarlatamab antibody development was 9.4% (22/235). Of these 22 patients, 11 (50.0%) had transient antibody responses that reverted to antibody-negative status by the last on-study assessment. The incidence of anti-tarlatamab neutralizing antibodies was 4.3% (10/235); 5 of these patients (50.0%) had transient neutralizing antibody responses.
Table 1:
Objective responses per RECIST in tarlatamab and chemotherapy groups
| Best overall responsea, n (%) | Tarlatamab (N = 254) |
Chemotherapy (N = 255) |
|---|---|---|
| Objective response rateb | 90 (35.4) | 52 (20.4) |
| Complete response | 6 (2.4) | 0 (0.0) |
| Partial response | 84 (33.1) | 52 (20.4) |
| Stable disease | 83 (32.7) | 112 (43.9) |
| Progressive disease | 56 (22.0) | 50 (19.6) |
| Not evaluable | 1 (0.4) | 1 (0.4) |
| No post-baseline scan | 24 (9.4) | 40 (15.7) |
| Disease control rateb | 173 (68.1) | 164 (64.3) |
| Sustained disease control (≥52 weeks) | 62 (24.4) | 11 (4.3) |
| Kaplan-Meier estimates for DORc, % (95% CI) at 12 months | 41.6 (31.3–51.5) | 10.1 (3.7–20.3) |
Assessment of disease response was determined using RECIST v1.1 guidelines.
Exact CIs were calculated using Clopper Pearson method.6
95% CIs were estimated using Kalbfleisch and Prentice (1980) method.7
Abbreviations: CI, confidence interval; DOR, duration of response; RECIST, Response Evaluation Criteria in Solid Tumors.
Safety
Grade ≥3 treatment-emergent adverse events (TEAEs) occurred in 59.9% and 80.7% of patients in tarlatamab and chemotherapy groups, respectively; 27.8% and 62.7% had treatment-related AEs (TRAEs; Table S2). TEAEs and TRAEs by preferred term and grade are presented in Table S3-S4). The most common serious TRAEs with tarlatamab were cytokine release syndrome (CRS; 17.1%), pyrexia (4.4%), and immune effector cell-associated neurotoxicity syndrome (ICANS; 3.6%; Table S5). In the tarlatamab group, the incidence of grade ≥3 TRAEs decreased over time, from 21.0% with onset <3 months to 12.8% at 3–12 months and 6.5% at ≥12 months (Figure 2; Table S6-S7). TRAEs leading to treatment interruption or discontinuation were lower with tarlatamab than chemotherapy across intervals. No serious TRAEs with onset ≥12 months were reported. CRS, ICANS, and pyrexia occurred predominantly early with tarlatamab and were largely grade 1–2, except one fatal ICANS event (Figure 2, Figure S3), while hematologic TRAEs were more frequent with chemotherapy (Table S4). Few CRS events occurred thereafter, including one serious event with onset >12 months during rechallenge, while no ICANS events were observed after the early treatment period (Table S8-S9). One patient had recurrence of grade 3 CRS. Treatment-related dysgeusia occurred more frequently with tarlatamab (23.8%; 60/252) than with chemotherapy (1.2%; 3/244); all events were grade 1–2 (Table S4). Dysgeusia incidence decreased over time with continued treatment (Figure 2; Table S7). With the extended follow-up, no new safety signals were reported.
Figure 2. (A-B): Temporal pattern of most common TRAEsa with tarlatamab (A) and chemotherapy (B).
aTRAEs with incidence of ≥15% in the treatment arms are presented. These durations refer to time from the start of the treatment.
Abbreviations: CRS, cytokine release syndrome; TRAE, treatment-related adverse event.
DISCUSSION
With additional follow-up of approximately one year after the primary analysis, tarlatamab continued to demonstrate sustained survival benefits over chemotherapy in the phase 3 DeLLphi-304 trial. Tarlatamab reduced the risk of death by 38% with persistent separation of Kaplan-Meier curves and was associated with higher landmark OS and PFS rates versus chemotherapy. The estimated 24-month OS rates of 30.3% with tarlatamab versus 14.9% with chemotherapy, and the corresponding PFS rates of 14.2% versus 2.3%, suggest a prolonged benefit from tarlatamab in a clinically relevant subset of patients. Results of prespecified subgroup analyses were generally consistent with the overall population, including patients with poor prognostic factors such as a chemotherapy-free interval of <90 days or brain metastases.2 Responses were more durable with tarlatamab, with a 12-month DOR of 41.6% versus 10.1% with chemotherapy. Sustained disease control for ≥52 weeks was achieved in higher proportion of patients with tarlatamab compared with chemotherapy. Although a similar proportion of patients in both treatment groups received subsequent anticancer therapy, approximately 10% of the chemotherapy group received subsequent DLL3-directed therapy, primarily tarlatamab. The incidence of anti-tarlatamab antibody development was low. Given the limited sample size, the association between anti-tarlatamab antibody development and the efficacy and safety of tarlatamab could not be determined and further studies are warranted.
With extended follow-up, tarlatamab maintained a favorable safety profile versus chemotherapy, with no new safety signals identified. Treatment-related toxicities generally decreased over time, including grade ≥3 TRAEs, with no serious TRAEs with onset ≥12 months reported. The incidence of low-grade CRS and ICANS events early in treatment was consistent with the primary analysis.2
These extended follow-up data further support the durable clinical benefit of tarlatamab in previously treated SCLC as reported earlier.4, 5 The survival benefit observed at the primary analysis2 was maintained with longer follow-up, with no new safety signals identified.
In conclusion, extended 23-month follow-up of the phase 3 DeLLphi-304 study reaffirms durable clinical benefit and manageable safety profile of tarlatamab versus chemotherapy, supporting tarlatamab as the standard of care for patients with SCLC following first-line platinum-based chemotherapy.
Supplementary Material
Highlights:
Tarlatamab reduced the risk of death by 38% versus chemotherapy after approximately 23 months’ follow-up.
Responses were longer with tarlatamab; 12-month DOR was 41.6% versus 10.1% with chemotherapy.
Tarlatamab maintained a favorable long-term safety profile, with toxicities decreasing and no new safety signals.
ACKNOWLEDGEMENTS
Medical writing support was provided by Manoj Goyal, Ph.D., and Lakshmi Narendra Bodduluru, Ph.D. (of Amgen Inc.), and graphical design assistance was provided by Vinod Goshamahal, MBA (of Amgen Inc.). Authors are thankful to Edith Sonnenberg for her invaluable contribution to this study.
FUNDING
This study is funded by Amgen Inc.
Declaration of Interest Statement
The author is an Editorial Board Member/Editor-in-Chief/Associate Editor/Guest Editor for this journal and was not involved in the editorial review or the decision to publish this article.
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:
CMR declares financial support from Amgen; consultant and or advisor role at AbbVie inc., Amgen, AstraZeneca (AZ), Daiichi Sankyo (DS), Genentech, Merck & Co., and Novartis; received licensing Fees/royalties for DLL3-directed therapies. M-JA received payment or honoraria for lectures, presentations, speakers bureaus, manuscript writing, or educational events from AZ, Lilly, Roche, Merck-Sharpe & Dohme (MSD), Takeda, ONO, Novartis, Amgen, Merck Serono, Yuhan, and DS; reports consult and or advisory role for AZ, Roche, Lilly, MSD, Takeda, ONO, Alpha Pharmaceutical, Amgen, Merck Serono, Pfizer, Yuhan, Arcus Ventures, DS, Voronoi, and Genexin; and received research funding from Yuhan. MS reports consultant and or advisory role at Amgen, AZ, Bristol-Myers Squibb (BMS), Roche, Gilead Sciences, GlaxoSmithKline (GSK), Janssen Pharmaceuticals, MSD, Novartis Pharma, Regeneron Pharmaceuticals, and Sanofi; invited speaker at Amgen, BMS, GSK, Janssen, Lilly, MSD, and Roche; institutional research/grant support from BMS, and Janssen Pharmaceuticals; and principal investigator, and steering committee member at Amgen and Janssen Pharmaceuticals. FHB reports consulting/advisory fees or personal fees from Amgen, AZ, Bayer, Blueprint Medicines, Boehringer-Ingelheim (BI), BMS, Guardant Health, Janssen, MSD, Novartis, Pfizer, Roche/Genentech, Takeda, and several other oncology companies; received research funding/institutional support from multiple industry sponsors, including AZ, Amgen, BMS, BI, Genentech/Roche, MSD, and Novartis; received honoraria, speakers' fees, travel support, or participation on advisory boards for several pharmaceutical companies involved in thoracic oncology development program.
GM reports advisory and consultation fees and or honoraria and payment for testimony from AZ, BMS, MSD, Janssen, Roche, BI, Takeda, Novartis, Amgen, GSK, Pierre Fabre, and Sanofi; received travel and accommodation fees from AZ, BI, BMS, MSD, Janssen, GSK, Pierre Fabre, Roche, Takeda, Novartis, Amgen, and Sanofi; Served as member of data safety monitoring body or advisory board at Novartis, Roche, BMS, MSD, AZ, Takeda, Janssen, GSK, Amgen, BI and Sanofi; leadership or fiduciary role in European Society for Medical Oncology working groups (Educational Publication Working Group, Adolescents and Young Adults working group) UNPAID; PI at Novartis, Roche, BMS, MSD, AZ, GSK, Amgen, Gilead and Sanofi. TY reports institutional research funding for clinical trials from AbbVie, Amgen, Astellas Pharma, AZ, BI, BMS, Chugai Pharmaceutical, DS, Eli Lilly, Merck, MSD and Novartis; and honoraria for advisory and/or speaking activities from Amgen, AZ, BI, BMS, Chugai Pharmaceutical, DS, Eli Lilly, MSD, Novartis, Ono Pharmaceuticals, Pfizer, Takeda Pharmaceuticals and Thermo Fisher Scientific. BCC reports consulting fees from ArriVent, AZ, BI, BMS, CJ Biosciences, Cyrus Therapeutics, MSD, Yuhan, Amgen, Gilead, Johnson & Johnson (J&J), and GSK; grants from CJ Bioscience, Cyrus, ImmuneOncia, J&J, JINTSbio, MSD, Yuhan, Dong-A ST, and Ligachem bioscience; royalties and licensing fees from Champions Oncology, Crown Bioscience, Imagen, PearlRiver Bio GmbH (PDX, PDC Licensing Contract – not patent); payment or honoraria from ASCO, AZ, Guardant, Roche, ESMO, IASLC, Korean Cancer Association, Korean Society of Medical Onoclogy, Korean Society of Thyroid-Head and Neck Surgery, Korean Cancer Study Group, Novartis, MSD, The Chinese Thoracic Oncology Society, Pfizer, Zailab; member of advisory board or data safety monitoring board st KANAPH Therapeutic Inc, Cyrus therapeutics, J INTS Bio; member of the board of directors at J INTS BIO; owns stocks options at J INTS BIO, KANAPH Therapeutics Inc, and also founder of DAAN Biotherapeutics. SA declares speaker and lecture fees form MSD AG, AZ, Roche, BMS, Merck Serono, Takeda, Medwise consulting, Amgen, BI, Pfizer, DS and Medison; travel support from AZ, Roche, MSD, and Pfizer; served as member of Advisory Boards at Roche, BeOne, Amgen, Bayer, AZ, MSD, J&J, BMS, Medison, Racura Oncology, Daiichi-Sankyo, Merck Serono, and BI. SRP reports consulting/advisory roles for AbVie Inc., Blueprint Medicines Corporation, Cogent Biosciences Inc., Regeneron Pharmaceuticals Inc., Janssen Pharmaceuticals Inc., and Revolution Medicines; and travel/accommodation support from Revolution Medicines. PR declares receipt of speaking and lecture fees from Abbott, Amgen, AZ, DS, Lilly Spain, and Thermo Fisher; consulting or advisory role for Amgen, Daiichi Sankyo, IGES Pharma, BI; travel and accommodation expenses from AZ, Amgen, BMS, Kiowa Kyrin, MSD, PharmaMar, Roche, and Thermo Fisher. LP-A reports grants from AZ, BMS, MSD, and Pfizer; consulting fees and/or personal fees from Amgen, AZ, Bayer, BMS, Daiichi Sankyo, GSK, Janssen, Lilly, Merck, Mirati, MSD, Novartis, Pfizer, PharmaMar, Roche, Sanofi, and Takeda; honoraria for speakers bureau from AZ, Janssen, and Merck, Amgen, Bayer, BMS, DS, GSK, Lilly, Mirati, MSD, Novartis, Pfizer, PharmaMar, Roche, Sanofi, Regeneron, Gilead, Beigene, and Takeda; travel support from J&J, AZ, BioNTech and Onclive, and Roche, owns stocks from Altum Sequencing, Stab Therapeutics. JM reports payment of honoraria for lectures, prsentations, and speakers’ bureaus from Merck, AZ, Pfizer, BMS, DS, MSD, Roche, and Novartis; and grants/contracts from AZ, Roche, and Pierre Fabre Pharmaceuticals. MG reports an advisory role for AbbVie; speaker/advisory activities and travel support from Amgen; advisory activities from Astellas Pharma, Gilead Sciences and Novartis; speaker activities from AZ, BMS and Takeda Oncology; consultancy for DS; research/project support from Roche; travel support from AZ, Roche, Regeneron Pharmaceuticals and Sandoz; and principal-investigator activities for Janssen Pharmaceuticals, Merck, Pfizer and Regeneron Pharmaceuticals. AL, LS, UD, and BZ declares no conflicts to disclose. SH, BA, and UP are Amgen employees and report stock options in Amgen Inc.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ClinicalTrials.gov Identifier: NCT05740566.
Data Sharing Statement
Qualified researchers may request data from Amgen clinical studies. Complete details are available at the following: https://wwwext.amgen.com/science/clinical-trials/clinical-data-transparency-practices/clinical-trial-data-sharing-request.
REFERENCES
- 1.Giffin MJ, Cooke K, Lobenhofer EK et al. AMG 757, a Half-Life Extended, DLL3-Targeted Bispecific T-Cell Engager, Shows High Potency and Sensitivity in Preclinical Models of Small-Cell Lung Cancer. Clin Cancer Res 2021; 27 (5): 1526–1537. [DOI] [PubMed] [Google Scholar]
- 2.Mountzios G, Sun L, Cho BC et al. Tarlatamab in Small-Cell Lung Cancer after Platinum-Based Chemotherapy. N Engl J Med 2025; 393 (4): 349–361. [DOI] [PubMed] [Google Scholar]
- 3.Mountzios G, Lammers PE, Sun L et al. Patient-reported outcomes with tarlatamab in extensive-stage small cell lung cancer after platinum-based chemotherapy: results from the phase 3 DeLLphi-304 trial. Lung Cancer 2026; 220. [DOI] [PubMed] [Google Scholar]
- 4.Dowlati A, Hummel H-D, Champiat S et al. Sustained Clinical Benefit and Intracranial Activity of Tarlatamab in Previously Treated Small Cell Lung Cancer: DeLLphi-300 Trial Update. Journal of Clinical Oncology 2024; 42 (29): 3392–3399. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Sands J, Cho BC, Ahn MJ et al. OA10.03 Tarlatamab Sustained Clinical Benefit and Safety in Previously Treated SCLC: DeLLphi-301 Phase 2 Extended Follow-up. Journal of Thoracic Oncology 2024; 19 (10, Supplement): S30–S31. [Google Scholar]
- 6.Clopper CJ, Pearson ES. The Use of Confidence or Fiducial Limits Illustrated in the Case of the Binomial. Biometrika 1934; 26 (4): 404–413. [Google Scholar]
- 7.Kalbfleisch JD, Prentice RL. The statistical analysis of failure time data: John Wiley & Sons; 2002. [Google Scholar]
- 8.Brookmeyer R, Crowley J. A Confidence Interval for the Median Survival Time. Biometrics 1982; 38 (1): 29–41. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Qualified researchers may request data from Amgen clinical studies. Complete details are available at the following: https://wwwext.amgen.com/science/clinical-trials/clinical-data-transparency-practices/clinical-trial-data-sharing-request.