Bispecific antibodies
Two arms, two targets: a tumour antigen with CD3 to pull in T cells, two checkpoints at once, or two tumour receptors so escape through one is blocked by the other. The format (BiTE, DuoBody, 2:1, ImmTAC) sets half-life and dosing. The grid below is target against second target: 23 by 17 from 60 medicines, 12 combinations approved, 27 in development, 0 tried and stopped, 355 untried in this corpus.
Antibody-drug conjugatesRadioligandsCAR cell therapiesTCR-T cell therapiesBispecific antibodiesDegraders and molecular gluesSmall-molecule inhibitorsNaked antibodiesCytokinesCancer vaccinesOncolytic virusesOther cell therapies
Approved 12In development 27Tried and stopped 0Recorded, state unclear 0Untried here 35523 target rows x 17 second target columns. Click a cell to filter the table.
Every tried combination
39 combinations
| Medicines | Records | Stopped, and why | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| PD-1 | Second target not recorded | In development | 0 | 3 | 0 | - | - | - | ||
| PD-1 | VEGF / VEGFR | Approved | 1 | 1 | 0 | - | - | |||
| PD-1 | TIM-3 | In development | 0 | 2 | 0 | - | - | |||
| PD-1 | TIGIT | In development | 0 | 1 | 0 | - | - | |||
| PD-1 | PD-L1 | In development | 0 | 1 | 0 | - | - | |||
| EGFR | MET | Approved | 1 | 3 | 0 | - | ||||
| EGFR | TRRAP | In development | 0 | 1 | 0 | - | ||||
| EGFR | HER3 | In development | 0 | 1 | 0 | - | - | |||
| EGFR | HER2 | In development | 0 | 1 | 0 | - | ||||
| EGFR | LGR5 | In development | 0 | 1 | 0 | - | - | |||
| CD20 | CD3 | Approved | 4 | 1 | 0 | Complete response letter for STARGLO combination (regional heterogeneity); approved in EU and 1 more recorded reason | ||||
| CD20 | Second target not recorded | In development | 0 | 1 | 0 | - | ||||
| BCMA | CD3 | Approved | 3 | 1 | 0 | - | ||||
| BCMA | CD38 | In development | 0 | 1 | 0 | - | ||||
| DLL3 | CD3 | Approved | 1 | 2 | 0 | - | ||||
| DLL3 | CD47 | In development | 0 | 1 | 0 | - | - | |||
| PD-L1 | VEGF / VEGFR | In development | 0 | 1 | 0 | - | - | |||
| PD-L1 | TIGIT | In development | 0 | 1 | 0 | - | ||||
| PD-L1 | TRRAP | In development | 0 | 1 | 0 | - | ||||
| PD-L1 | CD137 (4-1BB, TNFRSF9) | In development | 0 | 1 | 0 | Genmab has decided to discontinue further clinical development of acasunlimab following strategic portfolio prioritization. The decision was not related to saf… | ||||
| CD19 | CD3 | Approved | 1 | 2 | 0 | - | ||||
| Claudin 18.2 | CD3 | In development | 0 | 2 | 0 | - | - | |||
| Claudin 18.2 | CD47 | In development | 0 | 1 | 0 | - | - | |||
| CTLA-4 | PD-1 | Approved | 1 | 1 | 0 | - | ||||
| EpCAM | CD3 | Approved | 1 | 1 | 0 | - | ||||
| HER2 | HER2 | Approved | 1 | 0 | 0 | - | - | |||
| HER2 | HER3 | Approved | 1 | 0 | 0 | - | - | |||
| BCMA x GPRC5D | CD3 | In development | 0 | 1 | 0 | - | ||||
| CD25 (IL-2 receptor alpha) | PD-1 | In development | 0 | 1 | 0 | - | - | |||
| FAP | Mesothelin | In development | 0 | 1 | 0 | - | - |
(showing the first 30)
Stopped, and why
0 medicines recorded as stopped; 3 recorded reasons, including stopped studies of medicines still in development- AcasunlimabIn development
Genmab has decided to discontinue further clinical development of acasunlimab following strategic portfolio prioritization. The decision was not related to safety concerns.
- GlofitamabApproved
Complete response letter for STARGLO combination (regional heterogeneity); approved in EU
- OdronextamabApproved
CRLs for DLBCL and FL, confirmatory trial enrolment status
Proposed, not yet tried
12 proposals from the open pipeline, scored and searched (docs/OPEN-PIPELINE.md); hypotheses, not records of a medicineClinical evidence 9Preclinical 1No public evidence 2
| Score | Target | Second target | Status | Rationale | Evidence | Now in the corpus |
|---|---|---|---|---|---|---|
| 81 | PD-L1 | CTLA-4 | Clinical evidence | PD-L1 x CTLA-4 bispecific. PD-L1 blockade is approved across many cancers (atezolizumab, durvalumab) and CTLA-4 blockade both alone (ipilimumab, tremelimumab) and inside a PD-1 x CTLA-4 bispecific (cadonilimab). The corpus has PD-1 x CTLA-4, PD-1 x VEGF and PD-L1 x VEGF bispecifics (cadonilimab, ivonescimab, pm8002) but no PD-L1 x CTLA-4. The PD-1 x CTLA-4 bispecific (cadonilimab) is approved and shows the format can deliver dual checkpoint blockade with less colitis than the two antibodies; PD-L1 x CTLA-4 would concentrate CTLA-4 blockade in PD-L1-rich tumour tissue rather than lymph nodes, the same avidity logic applied to the ligand side.Caveat: It needs to show that tumour-localised CTLA-4 blockade beats the approved PD-1 x CTLA-4 bispecific or the durvalumab plus tremelimumab pair on toxicity or response, a high bar in cancers where those already work.Proposed by OnCo open pipeline, searched 2026-09-23. | Untried | |
| 78 | PD-1 | LAG-3 | Clinical evidence | PD-1 x LAG-3 bispecific. PD-1 blockade is approved across cancers (pembrolizumab, nivolumab) and LAG-3 blockade is approved in a fixed-dose combination with PD-1 blockade (relatlimab-nivolumab). The corpus has PD-1 x CTLA-4, PD-1 x TIM-3 and PD-1 x TIGIT bispecifics (cadonilimab, azd7789, rilvegostomig) but no PD-1 x LAG-3 bispecific, the pair with the strongest clinical validation. The fixed-dose relatlimab plus nivolumab combination is approved (relatlimab-nivolumab), so the biology is proven; a single bispecific molecule would bind cells that co-express both receptors, exhausted tumour-infiltrating T cells, with higher avidity than two separate antibodies and could be dosed as one product.Caveat: A PD-1 x LAG-3 bispecific reached late-stage trials and was discontinued; the new molecule would need a reason to differ, such as different epitopes, Fc design or a population defined by LAG-3 expression.Proposed by OnCo open pipeline, searched 2026-09-23. | Untried | |
| 75 | Folate receptor alpha | CD3 | Preclinical | Folate receptor alpha x CD3 bispecific. Folate receptor alpha is validated by an approved ADC (mirvetuximab-soravtansine) with two more in phase 3 (rinatabart-sesutecan, luveltamab-tazevibulin), and the CD3 arm by approved solid-tumour engagers (tarlatamab). The corpus holds an ovarian engager against MUC16 (ubamatamab) but none against folate receptor alpha. Folate receptor alpha is dense and homogeneous in high-grade serous ovarian cancer and largely absent from normal tissue except the kidney tubule and choroid plexus; ovarian cancer's ascites is rich in T cells, which an engager could recruit, and the corpus already has an ovarian engager against MUC16 (ubamatamab).Caveat: Ovarian cancer's microenvironment is immunosuppressive and checkpoint blockade has failed there; an engager may need a costimulatory partner or intraperitoneal delivery to work.Proposed by OnCo open pipeline, searched 2026-09-23. | Untried | |
| 73 | CEACAM5 | CD3 | Clinical evidence | CEACAM5 x CD3 bispecific. CEACAM5 has two ADCs in phase 3 in the corpus (precemtabart-tocentecan, tusamitamab-ravtansine) and the CD3 arm is validated in solid tumours (tarlatamab). No corpus T-cell engager targets CEACAM5, though it is the classic marker of the largest immunotherapy-resistant cancer in the corpus (colorectal). CEACAM5 is dense on colorectal and lung adenocarcinoma with limited normal expression apart from colonic epithelium; microsatellite-stable colorectal cancer has no working immunotherapy, so an engager that brings T cells in is the right shape of idea, and the shed antigen (CEA) in serum is the sink to design around.Caveat: Soluble CEA in serum can soak up the engager and colonic epithelium expresses the antigen; a 2+1 format that needs membrane-bound density is the plausible design, and an earlier attempt found modest activity.Proposed by OnCo open pipeline, searched 2026-09-23. | Untried | |
| 73 | HER2 | CD3 | Clinical evidence | HER2 x CD3 bispecific. HER2 is validated across antibodies, ADCs and bispecifics (trastuzumab, trastuzumab-deruxtecan, zanidatamab) and the CD3 arm across seven approved engagers (blinatumomab, teclistamab, tarlatamab). The corpus has no HER2 x CD3 T-cell engager, the obvious cell for the most validated antigen and the most validated effector arm. HER2 is the most validated solid-tumour antigen, but T-cell engagers have no antigen-density threshold and HER2 is expressed at low levels on many normal epithelia; a low-affinity or avidity-dependent (2+1) format is what a HER2 x CD3 needs to spare normal tissue.Caveat: Normal-tissue HER2 will be hit unless the CD3 arm is tuned to need high antigen density; the first-generation attempts were limited by exactly this.Proposed by OnCo open pipeline, searched 2026-09-23. | Untried | |
| 73 | Somatostatin receptor 2 | CD3 | Clinical evidence | Somatostatin receptor 2 x CD3 bispecific. SSTR2 is validated by an approved radioligand (lutathera) and the CD3 arm in neuroendocrine cancer by an approved DLL3 engager (tarlatamab). No corpus T-cell engager targets SSTR2. SSTR2 is restricted and dense on neuroendocrine tumours, and tarlatamab proves engagers work in neuroendocrine-lineage cancer; SSTR2 on pancreatic islets, pituitary and gut endocrine cells predicts endocrine toxicity, and a GPCR is a hard antibody target.Caveat: An SSTR2 x CD3 engager reached phase 1 and was discontinued; endocrine on-target effects and the difficulty of engaging a GPCR remain, so a new construct would need to show why it differs.Proposed by OnCo open pipeline, searched 2026-09-23. | Untried | |
| 73 | TROP2 | CD3 | No public evidence | TROP2 x CD3 bispecific. TROP2 carries three approved ADCs (sacituzumab-govitecan, datopotamab-deruxtecan, sacituzumab-tirumotecan) and the CD3 arm is validated in solid tumours by tarlatamab (tarlatamab). No corpus T-cell engager targets TROP2. TROP2 is on normal skin, oesophagus and other epithelia; an ADC tolerates this because payload delivery scales with antigen density, but a CD3 engager kills at low density, so the window is narrow unless the CD3 arm is masked or avidity-gated.Caveat: Normal-epithelium expression predicts on-target toxicity; a protease-masked or 2+1 avidity format would be needed to have any window.Proposed by OnCo open pipeline, searched 2026-09-23. | Untried | |
| 70 | PSMA | CD3 | Clinical evidence | PSMA x CD3 bispecific. PSMA is validated by an approved radioligand (pluvicto) and the CD3 arm by seven approved engagers (blinatumomab, teclistamab). The corpus has prostate T-cell engagers against STEAP1 and KLK2 (xaluritamig, pasritamig) and a PSMA x CD28 costimulator (jnj-87189401) but no PSMA x CD3 engager, so the cell is untried in the corpus. PSMA is tumour-restricted enough for a radioligand and the corpus already holds prostate engagers against STEAP1 (xaluritamig) and KLK2 (pasritamig) and a PSMA x CD28 costimulatory bispecific (jnj-87189401); the prostate microenvironment's T-cell exclusion and cytokine release at active doses are the obstacles.Caveat: Several PSMA x CD3 engagers reached the clinic and were stopped for cytokine release or limited activity; a new attempt needs a reason to differ, such as half-life extension, subcutaneous step-up dosing or pairing with the PSMA x CD28 costimulator.Proposed by OnCo open pipeline, searched 2026-09-23. | Untried | |
| 69 | Nectin-4 | CD3 | No public evidence | Nectin-4 x CD3 bispecific. Nectin-4 is validated by an approved ADC (enfortumab-vedotin) and the CD3 arm by tarlatamab in solid tumours (tarlatamab). No corpus T-cell engager targets Nectin-4. Nectin-4 is dense in urothelial cancer with skin as the main normal site; bladder cancer responds to immunotherapy, so T cells are present, and enfortumab plus pembrolizumab already shows the antigen and immune activation can be combined.Caveat: Skin expression of Nectin-4 would produce on-target dermatitis from an engager as it does from the ADC; whether that is tolerable at active doses is unknown.Proposed by OnCo open pipeline, searched 2026-09-23. | Untried | |
| 67 | CD123 | CD3 | Clinical evidence | CD123 x CD3 bispecific. CD123 is validated by an approved ADC (pivekimab-sunirine) and an approved toxin fusion (tagraxofusp), and the CD3 arm by blinatumomab (blinatumomab). No corpus T-cell engager targets CD123. CD123 covers leukaemic stem cells and is uniformly high in BPDCN, a disease where tagraxofusp already validates the antigen; capillary leak and cytokine release are the toxicities, and BPDCN's rarity limits trials.Caveat: Earlier CD123 x CD3 engagers showed activity mainly in a TP53-mutant subgroup and were limited by cytokine release; a new one would need to define its population first.Proposed by OnCo open pipeline, searched 2026-09-23. | Untried | |
| 67 | CD33 | CD3 | Clinical evidence | CD33 x CD3 bispecific. CD33 is validated by an approved ADC (gemtuzumab-ozogamicin) and the CD3 arm by blinatumomab in the neighbouring lymphoid leukaemia (blinatumomab). No corpus T-cell engager targets a myeloid antigen, so the CD33 x CD3 cell is untried in the corpus. AML is the natural haematological extension of blinatumomab's success in ALL; the problems are low T-cell numbers in AML patients, myeloid-cell CD33 causing cytokine release, and the leukaemia's immunosuppressive niche.Caveat: Several CD33 x CD3 engagers reached phase 1 with modest response rates and cytokine release; the case for a new one rests on half-life extension and use in measurable residual disease rather than bulk relapse.Proposed by OnCo open pipeline, searched 2026-09-23. | Untried | |
| 61 | CD38 | CD3 | Clinical evidence | CD38 x CD3 bispecific. CD38 is validated by two approved antibodies (daratumumab, isatuximab) and the CD3 arm by three approved myeloma engagers (teclistamab, elranatamab, talquetamab). The corpus has a trispecific that includes CD38 (isb-2001) but no CD38 x CD3 bispecific. Myeloma engagers against BCMA and GPRC5D work, and CD38 is dense on plasma cells; but CD38 is on activated T cells and NK cells, so an engager would attack its own effectors, and the corpus already holds a BCMA x CD38 x CD3 trispecific (isb-2001) that uses CD38 as an avidity arm.Caveat: An earlier CD38 x CD3 was stopped after phase 1, and CD38 on T cells makes fratricide a structural problem; the trispecific route, where CD38 adds avidity rather than being the sole target, may be the only viable form.Proposed by OnCo open pipeline, searched 2026-09-23. | Untried |
Not placed
5 of 65 medicines in this format (8%)These medicines belong to the format but their target is on no record the engine reads: the targets field is empty, the target lists no such drug, the trials linked to it name no target of its own, and the modality, mechanism and summary text name none the corpus knows. They are listed here rather than placed by guesswork; adding the target to the record puts them in the grid on the next build.
How this grid is read from the records
Each medicine's parts come from its own record: the targets field first, else the target records that list the medicine, else a sibling conjugate that shares its antibody name, else the trials linked to it, else the target names the corpus knows found in its modality, mechanism or summary text. The two arms come from the targets field, with the effector arm (CD3, CD28, CD137, CD47) taking the column; where a record names one arm, the other is read from the modality or mechanism text, and the format (BiTE, DuoBody, 2:1, ImmTAC) from the same text. Every part carries the fields it was read from and a confidence in the JSON file.
What the states mean, and do not
Approved: a medicine in the cell has an approval row or a regulator's approved entry and is not recorded as withdrawn. In development: a recruiting, active or planned trial, a development-phase record status, or active studies in the ClinicalTrials.gov index, and no approval. Tried and stopped: every medicine in the cell is withdrawn, negative or historic, or its only trial evidence is a terminated, withdrawn or negative study, or its approval was withdrawn. Untried: no medicine in this corpus combines the two parts.
A cell is coloured by its strongest medicine; the table shows each medicine with its own state. The reasons quoted under Stopped, and why are the records' words, including the registry's whyStopped text where the sponsor wrote one, and include stopped studies of medicines that are still in development elsewhere. Nothing here is medical advice.