DLL3 imaging for small-cell lung cancer
A drug that attacks small-cell lung cancer through a protein called DLL3 is already approved, but there is no approved scan that shows whether a patient's tumours still carry that protein. Doctors decide from a single biopsy taken from one place, which cannot show what the rest of the disease is doing.
Overview
Tarlatamab, a DLL3-directed bispecific T-cell engager, received accelerated FDA approval on 16 May 2024 and traditional approval on 19 November 2025, and DLL3-directed antibody-drug conjugates, CAR-T cells and a second T-cell engager (obrixtamig) are in trials. Patient selection rests on tissue immunohistochemistry from one archival or fresh biopsy. DLL3 is expressed in 85% of small-cell lung cancers at a 25% cut-off in the largest real-world series (895 of 1,050 patients), so most patients qualify on a single sample, but nothing in routine practice shows whether every metastatic site still expresses the target.
One imaging agent has human data. The zirconium-89-labelled anti-DLL3 antibody 89Zr-DFO-SC16.56, run investigator-initiated at Memorial Sloan Kettering under NCT04199741, was reported in Lancet Oncology in 2024: 18 patients, DLL3-avid tumours delineated on a single scan in 12 of 15 patients in the expansion cohort, uptake congruent with DLL3 immunohistochemistry in 15 of 16 patients with evaluable tissue, mean effective dose 0.49 mSv/MBq, one grade 1 allergic reaction and no grade 2 or worse adverse events. The trial is still recruiting with an estimated 105 patients. A gallium-68 DLL3 tracer opened in China in May 2026 (NCT07677787, 20 patients planned) and has not reported. A next-generation MSK construct, 89Zr-DFO-TDI-Y-010, with a lutetium-177 therapeutic pair, is preclinical.
The gap is therefore not biological plausibility, which a published 94% concordance with immunohistochemistry supports, but the absence of any industry-sponsored development path: the only clinical agent is an academic antibody with multi-day imaging kinetics, and no DLL3 imaging agent has been filed with any regulator.
Before starting a DLL3-directed T-cell engager, no test shows whether every metastatic site in a patient with small-cell lung cancer still expresses DLL3, or whether some lesions have lost it; the decision rests on one biopsy from one place.
- If almost every patient is DLL3-positive on biopsy, a scan that confirms it adds cost without changing management; the case rests on heterogeneity that has not been quantified prospectively.
- The market is tied to one therapy class in one small indication, so a companion-only agent carries the therapy's commercial risk without any standalone use to fall back on.
- A full antibody cannot be delivered at service scale; whoever wants this needs a same-day construct, and the published work does not provide one.
- Readout of the expansion cohort of NCT04199741, which would take the published series from 18 to a larger number of patients.
- First results from the gallium-68 DLL3 tracer at Xiamen (NCT07677787), the first non-antibody agent in humans.
- Any registered trial of the MSK next-generation 89Zr-DFO-TDI-Y-010 or its lutetium-177 pair, which would signal the field moving to a theranostic construct.
The five domains of a radiopharmaceutical due-diligence review, adapted from the Fryback-Thornbury hierarchy of diagnostic efficacy. Each carries the state of the public evidence and the fact behind it, not a score out of five: a single total hides the reason and stops telling candidates apart. Domains marked as carrying most weight are the ones that decide the case at this asset's stage.
DLL3 is expressed in 895 of 1,050 small-cell lung cancers (85%) at a 25% tumour-cell cut-off, and in 68% at a 75% cut-off, in the largest international real-world immunohistochemistry series.
One published first-in-human cohort of 18 patients with an antibody tracer; the trial continues to an estimated 105 patients. No second agent has reported human data.
A single scan delineated DLL3-avid tumour in 12 of 15 patients, and uptake agreed with tissue immunohistochemistry in 15 of 16 patients with evaluable tissue. No study yet relates scan findings to response on DLL3-directed therapy.
Mean effective dose 0.49 mSv/MBq with the highest absorbed organ dose in liver at 1.83 mGy/MBq; one grade 1 allergic reaction in 18 patients and no grade 2 or worse events.
No DLL3 imaging agent has been filed anywhere. The nearest precedent for a receptor-status tracer is 68Ga-DOTATATE, approved in 2016 on localisation performance against a reference standard rather than an outcome endpoint.
Red, amber or green on each review category, each with its reason, so an obvious no-go takes seconds rather than a meeting.
The only clinical agent is a full antibody needing imaging three to six days after injection, which does not fit a routine service. A same-day construct exists only preclinically.
Real but narrow: because DLL3 is expressed in most small-cell lung cancers, a scan changes management mainly in the minority with heterogeneous or lost expression, and that fraction has not been measured prospectively.
The lead antibody and its next-generation successor are institutionally owned at Memorial Sloan Kettering; freedom to operate for an independent construct is unexamined in public sources.
Sales would depend on the DLL3 therapy market, and small-cell lung cancer is a small, short-lived population with no standalone diagnostic question of its own.
Any of gallium-68, fluorine-18 or copper-64 would suit a small-molecule or fragment construct; the published antibody uses zirconium-89, which is cyclotron-made but established.
Only one academic agent has human data and one Chinese tracer has just opened; no company has a disclosed clinical DLL3 imaging programme.
- Memorial Sloan Kettering Cancer Centerphase 1/2 recruiting
89Zr-DFO-SC16.56, the only DLL3 tracer with published human data, and a preclinical successor pair 89Zr-DFO-TDI-Y-010 with 177Lu therapy.
https://clinicaltrials.gov/study/NCT04199741 - First Affiliated Hospital of Xiamen Universityrecruiting
A gallium-68 DLL3-targeted PET tracer; 20 patients planned, opened May 2026, no results published.
https://clinicaltrials.gov/study/NCT07677787 - Amgenapproved
Tarlatamab, the approved DLL3 T-cell engager that creates the demand; selection is by tissue immunohistochemistry, not imaging.
https://www.fda.gov/drugs/resources-information-approved-drugs/fda-grants-accelerated-approval-tarlatamab-extensive-stage-small-cell-lung-cancer - Boehringer Ingelheimactive, not recruiting
Obrixtamig, a second DLL3 T-cell engager in DLL3-positive small-cell lung cancer and other neuroendocrine tumours.
https://clinicaltrials.gov/study/NCT04429087
Similar pages
not linked directly; found by shared links- OpportunityHER3 imaging, and why three tracers stalled
Shares Immuno-PET, Zirconium-89, PET/CT and the tags companion diagnostic, lung.
- TargetCCK2 receptor (gastrin receptor)
Shares Small-cell lung cancer, Gallium-68, Lutetium-177 and the tag neuroendocrine.
- OpportunityClaudin-18.2 imaging when one biopsy says negative
Shares Immuno-PET, PET/CT, Gallium-68 and the tag companion diagnostic.
- IdeaCD8 PET to stop or switch immunotherapy early
Shares Immuno-PET, Zirconium-89.
- InstitutionShandong Cancer Hospital and Institute
Shares Small-cell lung cancer, PET/CT.
- CompanyAbdera Therapeutics
Shares DLL3, Small-cell lung cancer.
- IdeaTROP2 PET to choose and sequence TROP2 ADCs
Shares Immuno-PET, Zirconium-89, Gallium-68.
- OpportunityCAIX imaging of the indeterminate renal mass
Shares Immuno-PET, Zirconium-89, PET/CT, Gallium-68.