Nogapendekin alfa inbakicept met its lymphocyte surrogate but lost its first-line NSCLC Phase 3

What was tried

ImmunityBio ran QUILT-2.023 (NCT03520686), a Phase 3, open-label, randomized, four-cohort study of nogapendekin alfa inbakicept (N-803, brand name Anktiva) added to standard first-line care for advanced or metastatic non-small cell lung cancer (ClinicalTrials.gov, NCT03520686). N-803 is an interleukin-15 superagonist complex, an IL-15 N72D mutein bound to an IL-15 receptor alpha sushi-domain IgG1 Fc fusion (Nogapendekin alfa inbakicept: First Approval, 2024). The control arms used checkpoint inhibitor regimens such as pembrolizumab with or without chemotherapy, and the experimental arms added subcutaneous N-803 (ClinicalTrials.gov, NCT03520686). Primary endpoints were progression-free survival at 24 months and the change in absolute lymphocyte count over 27 weeks (ClinicalTrials.gov, NCT03520686). The study began on 2018-05-18 and reached an actual enrollment of 102 before termination, with a primary completion date of 2025-10-13 and no posted results in the registry (ClinicalTrials.gov, NCT03520686).

The biological hypothesis

IL-15 drives the proliferation and survival of CD8 T cells and natural killer cells through the shared IL-2 and IL-15 receptor beta and common gamma chains, while sparing the regulatory T cells that IL-2 expands (Nogapendekin alfa inbakicept: First Approval, 2024). N-803 was engineered to deliver that signal with a longer half-life and stronger receptor engagement than native IL-15, which in preclinical models expanded effector lymphocytes and produced durable antitumor immunity (ALT-803 IL-15 superagonist, 2016,, and). The first-line NSCLC logic was complementary: checkpoint blockade releases the brake on T cells, and an IL-15 superagonist supplies the proliferative push, so adding N-803 should deepen and extend responses. The same biology later supported the drug's approval with BCG in bladder cancer (, and).

What actually happened

QUILT-2.023 produced a statistically significant and sustained increase in absolute lymphocyte count for N-803 plus checkpoint inhibition versus checkpoint inhibition alone (P = 0.0065, OncLive conference coverage, 2024). The trial then wound down. The registry records termination because no active subjects remained, with results to be posted within a year of the primary completion date (ClinicalTrials.gov, NCT03520686). ImmunityBio had by then secured FDA approval of N-803 with BCG in BCG-unresponsive non-muscle-invasive bladder cancer in April 2024 (doi:10.1158/1078-0432.CCR-25-1231) and moved its first-line NSCLC effort into a redesigned Phase 3 with tislelizumab. The openFDA FAERS database held 28 reports naming Anktiva, 14 of them serious, with top terms neutropenia, chills, anaemia, and pyrexia, a profile consistent with cytokine-driven cytopenias and constitutional effects (openFDA FAERS, queried 2026-06-02). In Open Targets, the IL15 association with non-small cell lung carcinoma scored 0.0216, literature-only at 0.178, with no human genetic evidence (Open Targets, IL15 ENSG00000164136).

Failure mechanism, best guess

This was a reprioritization, not a safety stop or a clean efficacy refutation. The trial hit its pharmacodynamic surrogate, a higher lymphocyte count, but a lymphocyte rise is a mechanism check, not a clinical outcome. The 102-patient final enrollment across four cohorts left the progression-free survival comparison thin, and a small open-label Phase 3 spread across regimens cannot deliver a registration-grade first-line result on its own. The deeper issue is translational: systemic cytokine agonism reliably expands effector cells in blood, yet converting that expansion into durable tumor control in first-line NSCLC has been hard for the IL-15 class. With an approval in hand in bladder cancer and a cleaner NSCLC design available, the rational move was to stop an underpowered four-cohort study and redeploy.

How to prevent this next time

Two quantitative tools would have set a firmer bar before a Phase 3 commitment.

First, explicit power calculations tied to the clinical endpoint rather than the surrogate. A four-cohort, 102-patient open-label study is well powered to detect a lymphocyte-count shift but underpowered for a first-line progression-free survival difference.

Second, a Bayesian predictive-probability framework that updates the probability of a clinical win from the surrogate readout.

PP(success)=∫θ​P(final success∣θ,interim data)⋅π(θ∣interim data)dθ

Feeding the significant lymphocyte signal (P = 0.0065) into a predictive model with a weak prior linking that surrogate to survival would have quantified how little the surrogate moved the probability of a progression-free survival benefit, and would have justified either a focused single-regimen design or an earlier stop.

The single highest leverage change would have been to power the trial on a pre-specified progression-free survival hazard ratio in one regimen, with a Bayesian predictive-probability gate that translated the lymphocyte surrogate into an explicit clinical-success probability before expanding to four cohorts.

What this means for similar programs

IL-15 superagonists remain pharmacologically active and have a real approval in bladder cancer, but their systemic-immunotherapy thesis in solid tumors still lacks a clean first-line survival result. Programs in this class should separate mechanism confirmation from efficacy testing, using small biomarker studies to prove lymphocyte expansion, then committing to a single, adequately powered regimen for the clinical endpoint. Surrogate endpoints like absolute lymphocyte count belong in early signal-finding, not as co-primaries that can make an underpowered Phase 3 look successful.

Open questions

What were the progression-free survival and overall survival results in QUILT-2.023, and will the posted data show any clinical separation? Does the lymphocyte-count increase correlate with response in any cohort? Will the redesigned NSCLC Phase 3 with tislelizumab use a single regimen and a clinical primary endpoint?

Sources

1. • ClinicalTrials.gov, NCT03520686 (Phase 3 open-label randomized four-cohort design, enrollment 102, primary endpoints progression-free survival at 24 months and change in absolute lymphocyte count, start 2018-05-18, primary completion 2025-10-13, termination reason, no posted results). - Nogapendekin alfa inbakicept: First Approval, 2024,(IL-15 superagonist structure and mechanism). - IL-15 Superagonist NAI in BCG-Unresponsive NMIBC, 2023,- FDA Approval Summary: nogapendekin alfa inbakicept with BCG, 2025,(April 2024 approval). - ALT-803 IL-15 superagonist enhances effector subpopulations, 2016,- ALT-803 durable antitumor immunity, 2016,- OncLive conference coverage of QUILT-2.023, 2024 (statistically significant absolute lymphocyte count increase, P = 0.0065). - Open Targets Platform: IL15 (ENSG00000164136) association with non-small cell lung carcinoma 0.0216, literature datatype 0.178, no genetic datatype. - openFDA FAERS, queried 2026-06-02 (28 Anktiva reports, 14 serious, top terms neutropenia, chills, anaemia, pyrexia). Available from: https://clinicaltrials.gov/study/NCT03520686.