CLN-978 left B-cell lymphoma after three patients and reopened in three autoimmune indications

What was tried

Cullinan Therapeutics ran NCT05879744, study CLN-978-001, an open-label Phase 1 dose escalation and dose expansion of CLN-978 in adults with relapsed or refractory B-cell non-Hodgkin lymphoma. Part A escalated dose in sequential cohorts. Part B was to expand at a selected dose in three groups: diffuse large B-cell lymphoma after at least two prior lines including a CD20 antibody and an anthracycline, follicular lymphoma grade 1 to 3a after at least two prior lines, and other B-cell histologies.

Eligibility required an ECOG performance status of 2 or better and a documented CD19-positive B-cell neoplasm by WHO classification, covering diffuse large B-cell, high-grade B-cell, primary mediastinal large B-cell, follicular, mantle cell and marginal zone lymphoma. The primary endpoints were safety and tolerability and definition of a dose regimen through dose-limiting toxicities, both over 24 months. Response rate, complete response, duration of response and pharmacokinetics were secondary.

The study opened on 31 May 2023. Primary completion and completion are both recorded as 1 June 2024, with three participants enrolled. The posted reason reads "Strategic business decision". The record was updated to terminated on 22 October 2026. No results have been posted.

The biological hypothesis

CLN-978 is a 65-kDa multispecific antibody construct that binds CD19 on B cells, CD3 epsilon on T cells, and human serum albumin. The preclinical characterisation, published in the Journal for ImmunoTherapy of Cancer in 2023, reports picomolar affinity for CD19 and nanomolar affinity for CD3 epsilon, with the albumin-binding domain added to extend serum half-life.

The design addressed two known limits of the class. Blinatumomab, the approved CD19 by CD3 engager, has a short half-life and requires continuous intravenous infusion, and low CD19 density is a recognised escape route after CD19-directed therapy. The same paper reports that redirected human T cells eliminated target cells carrying fewer than 300 CD19 copies per cell, and that primates showed a long serum half-life, deep and sustained B-cell depletion, and reduced cytokine release when the construct was given subcutaneously.

The target is among the best validated in oncology. CD19 is a coreceptor for the B-cell antigen receptor complex that lowers the threshold for downstream signalling, and Open Targets classes it as a surface antigen with approved-drug tractability. The association between CD19 and non-Hodgkin lymphoma scores 0.405, built from a clinical datatype score of 0.621 and a literature score of 0.897, with no genetic association datatype. Fourteen clinical-stage candidates are recorded against CD19, six at approval.

What actually happened

The trial dosed three patients and stopped. On 16 April 2024 Cullinan announced that it was halting the lymphoma study and expanding into autoimmune disease, renaming itself from Cullinan Oncology to Cullinan Therapeutics and raising 280 million dollars in a private placement the same day.

The reported observations from those three patients, as described by Fierce Biotech on the day of the announcement, were favourable. CLN-978 was clinically active at the 30 microgram starting dose. Both patients with detectable B cells at baseline had rapid, deep and sustained B-cell depletion. Two patients had cytokine release syndrome, in both cases grade 1.

No maximum tolerated dose was established, no recommended Phase 2 dose was selected and no response data were generated, because Part B never opened. The hypothesis set out in the 2023 preclinical paper, that affinity tuning would reach low-CD19 tumours other agents miss, was never tested in a patient.

The asset did not stop. Three Phase 1 trials of subcutaneous CLN-978 are now recruiting: systemic lupus erythematosus (NCT06613360, opened January 2025, 24 planned), rheumatoid arthritis (NCT06994143, opened May 2025, 37 planned) and Sjogren disease (NCT07041099, opened October 2025, 36 planned). Ninety-seven planned autoimmune participants against three actual oncology participants is the shape of the decision, computed from those four registry records.

Failure mechanism, best guess

The archetype is strategic reprioritisation, and here the label is the whole explanation rather than a cover for something else. Nothing in the public record points to a safety or efficacy problem in lymphoma. What changed was the comparator. By early 2024 the case for CD19-directed therapy in autoimmune disease had been made in print, by a lupus case series in Nature Medicine in 2022, a follow-up series in the New England Journal of Medicine in 2024 and myositis and systemic sclerosis data in Cell in 2024. Against that, a late-line lymphoma setting already served by approved CAR-T products, a CD19 antibody-drug conjugate and a CD19 antibody offered a crowded path to a smaller prize.

The cost is a specific piece of unrecovered information. Three patients were exposed, and the question the trial was built to answer, whether affinity tuning reaches low-antigen tumours, went unanswered. A dose escalation stopped at 30 micrograms generates an exposure record and no dose.

How to prevent this next time

No endpoint-level data exist, so no quantitative reanalysis is offered. Two levers apply, and both are portfolio levers rather than trial levers.

The first is indication sequencing priced before first dose. The autoimmune case for CD19 depletion was visible in the literature from 2022, a year before this trial opened. A programme that may carry an asset into autoimmune disease should say so before enrolling oncology patients, so the oncology stage is either sized to answer its own question or not run.

The second is a minimum-information stopping rule. Terminating a dose escalation at the starting dose returns almost nothing to the field. Committing in advance to complete escalation through to a recommended Phase 2 dose, or to publish what was learned, converts a halted trial into a usable result for the three people who took part.

A third observation belongs here even though the sponsor alone did not control it. The registry record was not updated to terminated until 22 October 2026, two and a half years after the public announcement. Anyone screening the registry for terminated CD19 programmes in that window saw an active trial.

The single highest leverage change would have been pre-committing to complete the dose escalation to a recommended Phase 2 dose before any portfolio decision could close the study.

What this means for similar programs

The destination is not empty ground. Within the Claidex graph, CD19 now carries five recorded failures, and the four that preceded this one are all in autoimmune indications: LY3541860 in rheumatoid arthritis and in relapsing multiple sclerosis, both efficacy failures of non-depleting CD19 inhibition, and KYV-101 in lupus nephritis and in diffuse cutaneous systemic sclerosis, both strategic withdrawals. The modified risk score for CD19 is 74, at the top of the orange band, with saturation contributing 14.5 of 15 and archetype severity 17.8 of 25.

Depleting pharmacology distinguishes CLN-978 from the LY3541860 failures, so the read-across is not direct. What does hold is crowding. Seventeen distinct programmes now sit against this target across Open Targets and the Claidex graph, and the indications CLN-978 moved into are the same ones in which other CD19 agents have already stopped.

Open questions

Will the three patients dosed in NCT05879744 appear in any publication, and will the low-CD19 hypothesis be tested anywhere? Does the subcutaneous half-life-extended format carry a different cytokine release profile at autoimmune-range doses than the two grade 1 events at 30 micrograms suggest?

Sources

1. • ClinicalTrials.gov, NCT05879744, CLN-978 in relapsed or refractory B-cell non-Hodgkin lymphoma, record accessed 25 October 2026: https://clinicaltrials.gov/study/NCT05879744 - ClinicalTrials.gov, NCT06613360 (systemic lupus erythematosus), NCT06994143 (rheumatoid arthritis) and NCT07041099 (Sjogren disease), accessed 25 October 2026: https://clinicaltrials.gov/study/NCT06613360 - Meetze K and colleagues, CLN-978, a novel half-life extended CD19/CD3/HSA-specific T cell-engaging antibody construct with potent activity against B-cell malignancies with low CD19 expression, Journal for ImmunoTherapy of Cancer, 2023,- Correction to the above, Journal for ImmunoTherapy of Cancer, 2023,- Mackensen A and colleagues, Anti-CD19 CAR T cell therapy for refractory systemic lupus erythematosus, Nature Medicine, 2022,- Muller F and colleagues, CD19 CAR T-cell therapy in autoimmune disease, a case series with follow-up, New England Journal of Medicine, 2024,- Wang W and colleagues, Allogeneic CD19-targeted CAR-T therapy in patients with severe myositis and systemic sclerosis, Cell, 2024,- Schett G and colleagues, CAR T-cell therapy in autoimmune diseases, Lancet, 2023,- Baeuerle PA and Wesselborg S, T-cell-engaging bispecific antibodies in cancer, Lancet, 2023,- Falchi L and colleagues, Bispecific antibodies and CAR-T cells, dueling immunotherapies for large B-cell lymphomas, Blood Cancer Journal, 2024,- The current state of bispecific antibodies and T-cell directed therapy in NHL, Cancers, 2025,- Masson G, Cullinan rebrands, halts lymphoma trial to reflect autoimmune refocus, Fierce Biotech, 16 April 2024: https://www.fiercebiotech.com/biotech/cullinan-rebrands-halts-lymphoma-trial-reflect-autoimmune-focus - Cullinan Therapeutics announces strategic expansion into autoimmune diseases, GlobeNewswire, 16 April 2024: https://www.globenewswire.com/news-release/2024/04/16/2863448/0/en/Cullinan-Therapeutics-Announces-Strategic-Expansion-into-Autoimmune-Diseases.html - Open Targets Platform, CD19 (ENSG00000177455) and non-Hodgkin lymphoma (MONDO_0018908) association, tractability and clinical candidate records, queried 25 October 2026: https://platform.opentargets.org - openFDA drug adverse event endpoint, queried 25 October 2026 for CLN-978 and for blinatumomab: https://api.fda.gov/drug/event.json.