On September 16, 2026, Bristol Myers Squibb told Orum Therapeutics it was ending development of ORM-6151 — the CD33-targeting blood cancer candidate BMS had renamed BMS-986497 after acquiring it in 2023. Per Orum's regulatory filing two days later, the call came "after reviewing the clinical data" from the drug's Phase 1 trial in acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS) — well before the study's original primary-readout target of February 2027. For a molecule BMS paid $100 million upfront to acquire, walking away on an interim data look rather than a completed dataset is itself a signal: the benefit-risk profile evidently didn't clear the bar to keep funding the program to completion.
This is the second time in under eighteen months that a lead candidate built on Orum's Degrader-Antibody Conjugate (DAC) platform has been discontinued following a clinical safety and efficacy review — the first being ORM-5029, halted in April 2025 after a fatal liver toxicity event. This article breaks down what actually happened with ORM-6151, how the deal terms play out for both companies, what the DAC platform's second stumble means for the broader antibody-conjugate field, and why the underlying lesson — rigorous antibody backbone characterization before conjugation — applies well beyond this one program.
1. What Is ORM-6151 (BMS-986497)? A Degrader-Antibody Conjugate for AML and MDS
ORM-6151, later designated BMS-986497, was a degrader-antibody conjugate (DAC) that paired a CD33-targeting antibody with a payload that degrades GSPT1, a protein required for translation termination that cancer cells depend on to survive. CD33 is a well-established myeloid surface marker, which made it a logical targeting antigen for AML and higher-risk MDS — two blood cancers with limited treatment options once standard therapy fails. Orum Therapeutics originated the molecule and out-licensed — more precisely, sold — it to Bristol Myers Squibb in October 2023.
- Target: CD33 (antibody component) delivering a GSPT1 degrader (payload component)
- Indication: Relapsed or refractory AML and higher-risk MDS
- Trial design: Phase 1 monotherapy, plus double and triple combinations with azacitidine and venetoclax (NCT06419634)
- Originator: Orum Therapeutics; acquired by Bristol Myers Squibb in October 2023
- Status as of September 2026: Discontinued by BMS following a Phase 1 data review
2. Why BMS Walked Away After an Interim Phase 1 Data Review
Neither company has published a detailed breakdown of which findings — safety, efficacy, or both — drove the discontinuation. What's public is narrower but still telling: BMS's decision came from a review of the clinical data generated so far, not from a completed trial, and it landed roughly five months before the study's original February 2027 primary evaluation target. A sponsor that still saw a viable benefit-risk profile in the interim data would typically let a Phase 1 trial run to its planned readout, especially one already enrolling combination cohorts with azacitidine and venetoclax. Ending it early is the more informative signal here than any single disclosed adverse event.
Orum's own public response reinforces that reading. Rather than framing the decision as bad luck, the company said it intends to "understand why the clinical program was halted and incorporate any technical lessons into the development of subsequent candidates" — language that points toward a platform-level review, not a dismissal of the result as a one-off.
Technical Note
A DAC's payload doesn't need to be cytotoxic to cause dose-limiting toxicity. A GSPT1 degrader eliminates a protein that virtually every actively translating cell relies on, not just the tumor cell the antibody was meant to reach — which makes the antibody's targeting precision and the linker's payload retention especially consequential to the therapeutic window.
3. The Platform's Prior Red Flag: ORM-5029 and the April 2025 Liver Toxicity Signal
ORM-6151 was not the first Orum DAC candidate to run into trouble. ORM-5029, a HER2-targeting DAC for HER2-expressing solid tumors, was in its own U.S. Phase 1 trial when a serious adverse event was reported in November 2024: a patient died of liver failure. Orum discontinued ORM-5029 in April 2025 after what the company described as a comprehensive review of clinical safety, pharmacokinetic, and pharmacodynamic data. Investigators suspected the antibody-linker construct had released the GSPT1 degrader payload prematurely in the bloodstream, producing systemic exposure well beyond the intended tumor site — though a definitive root cause was never confirmed, and the company has said it modified antibody and linker structures in later-generation candidates as a result.
The two programs shared the same underlying degrader chemistry (GSPT1) and DAC architecture, differing mainly in the targeting antibody — HER2 for ORM-5029, CD33 for ORM-6151. That shared foundation is exactly why BMS's decision on ORM-6151, even without a confirmed mechanism, reads as more than an isolated data point: it's the platform's second discontinued lead asset inside eighteen months.
| Candidate | Antibody Target | Payload | Status | Reason / Outcome |
|---|---|---|---|---|
| ORM-5029 | HER2 (solid tumors) | GSPT1 degrader | Discontinued, April 2025 | Phase 1 SAE — patient death from liver failure; premature payload release suspected |
| ORM-6151 / BMS-986497 | CD33 (AML/MDS) | GSPT1 degrader | Discontinued by BMS, September 2026 | Benefit-risk profile did not support continued development after interim Phase 1 review |
| ORM-1153 | CD123 (hematologic malignancies) | Next-gen degrader, modified linker | FDA IND cleared; Phase 1 starting | Preclinical data suggest lower effective dose, wider therapeutic index |
4. The Deal Structure: Why Orum Keeps $100M, But Not the Rights
The financial mechanics of the wind-down matter as much as the science. BMS originally acquired ORM-6151 in October 2023 in a deal worth up to $180 million — a $100 million upfront payment plus up to $80 million in additional milestones. With the program discontinued, BMS is released from that remaining $80 million milestone obligation, while Orum is not required to return the $100 million it already banked.
What Orum does not get back is the asset itself. Because the 2023 agreement was structured as an outright Asset Purchase Agreement rather than a license, ownership transferred to BMS in full — there is no license-back or reversion clause to trigger on discontinuation. Orum CEO SJ Lee made this explicit in comments to the press:
"BMS acquired this program through an Asset Purchase Agreement. As such, there are no reversion rights to Orum."
In practice, ORM-6151 now sits discontinued inside BMS's pipeline rather than returning to Orum's hands for a potential re-partnering effort — a structurally different outcome than if the 2023 deal had been a license. Lee also noted Orum retains approximately $180 million in cash and does not expect the decision to trigger layoffs, an important distinction between a single asset's failure and a company-level crisis.
5. What This Means for the ADC/DAC Antibody-Conjugate Field
Two discontinued lead candidates from one platform inside eighteen months is a real setback for Orum specifically. It is a weaker signal about the antibody-conjugate field as a whole. Despite ORM-5029's failure in April 2025, Orum's underlying linker technology drew a separate collaboration from Vertex Pharmaceuticals worth up to $945 million later that year — capital that bet on the platform's chemistry even after its first clinical stumble. The broader ADC and DAC space has continued attracting large partnership deals through 2026, and Orum itself is not retreating from the mechanism: it's advancing ORM-1153, a CD123-targeting DAC for hematologic malignancies, with an FDA-cleared IND and dosing expected soon.
- Investor and partner capital is still flowing into degrader-antibody and antibody-drug conjugate platforms despite individual asset failures.
- Sponsors appear to be reviewing early clinical data faster and discontinuing marginal assets sooner, rather than carrying them to a full trial readout.
- Orum's stated response — attribute the failure to specific antibody-linker constructs and redesign rather than abandon the mechanism — is the more common industry pattern than platform-wide retreat.
- ORM-1153's preclinical profile (lower effective dose, wider preclinical therapeutic index) suggests the lessons from ORM-5029 and ORM-6151 are being applied at the molecule-design level, not just noted and set aside.
6. The Common Thread: Antibody Backbone Characterization Before Conjugation
Strip away the specific payload chemistry, and both Orum setbacks point to the same upstream variable: the behavior of the antibody-linker construct before and after it reaches its target. A conjugate's therapeutic window depends on the antibody binding its intended target with high specificity, minimal cross-reactivity against related antigens, and a linker that holds the payload — cytotoxin or degrader alike — until the construct is where it's supposed to be. When any of those properties is under-characterized going into the clinic, the payload's own potency becomes a liability rather than an asset, which is exactly the failure mode investigators suspected in ORM-5029 and the kind of risk any DAC, ADC, or bispecific conjugate program has to rule out before dosing patients.
That's the same discipline Sekbio applies to every custom antibody it develops, regardless of what the antibody is ultimately conjugated to or used for. Our monoclonal antibody development service runs each candidate through binding analysis (ELISA, Western blot, flow cytometry), SPR/BLI affinity kinetics, cross-reactivity and epitope mapping, and functional testing before it's considered a validated lead — the same categories of data that determine whether an antibody backbone is safe to build a conjugate program on. Combined with CHO and HEK293 expression platforms delivering lead candidates in 8–12 weeks, that gives biotech teams building next-generation antibody-conjugate programs — DAC, ADC, or bispecific — a characterized antibody backbone to start from, rather than an under-tested one to find out about later.
7. Frequently Asked Questions — ORM-6151, BMS, and the DAC Platform
What is ORM-6151 (BMS-986497)?
ORM-6151, later renamed BMS-986497, was a degrader-antibody conjugate (DAC) that paired a CD33-targeting antibody with a GSPT1-degrading payload. It was developed by Korean biotech Orum Therapeutics and was in Phase 1 testing for acute myeloid leukemia (AML) and higher-risk myelodysplastic syndrome (MDS), including as a triplet combination with azacitidine and venetoclax.
Why did Bristol Myers Squibb discontinue ORM-6151?
According to an Orum regulatory filing, BMS ended the program after reviewing the Phase 1 clinical data. Neither company has publicly specified whether safety or efficacy findings drove the call, but the decision came well before the trial's planned primary evaluation readout, originally expected in February 2027 — indicating BMS moved on an interim data review rather than waiting for the full dataset.
What happened with Orum's other DAC candidate, ORM-5029?
ORM-5029 was Orum's HER2-targeting DAC for solid tumors. In November 2024, a serious adverse event was reported in its U.S. Phase 1 trial; a patient died of liver failure. Orum discontinued the program in April 2025 after a review of clinical safety, pharmacokinetic, and pharmacodynamic data. Investigators suspected the antibody-linker construct may have released the GSPT1 degrader payload prematurely in the bloodstream, though a definitive cause was never confirmed.
Why doesn't Orum get the rights back to ORM-6151?
BMS acquired ORM-6151 in October 2023 through an outright Asset Purchase Agreement rather than a license. Orum CEO SJ Lee confirmed that structure means there are no reversion rights: even though BMS is discontinuing development and forfeiting its remaining $80 million in milestone payments, the asset itself does not automatically return to Orum. Orum keeps the $100 million upfront payment it already received.
What is a degrader-antibody conjugate (DAC), and how is it different from an ADC?
A conventional antibody-drug conjugate (ADC) links a targeting antibody to a cytotoxic small molecule that kills the cell it's delivered to. A degrader-antibody conjugate (DAC) instead links the antibody to a targeted protein degrader (TPD) payload, which eliminates a specific protein the cancer cell needs to survive rather than poisoning the cell outright. The antibody's job in both formats is identical: bind the target cell selectively and carry the payload there without releasing it prematurely.
What is Orum doing next after losing ORM-6151?
Orum is redirecting resources to ORM-1153, a CD123-targeting DAC for hematologic malignancies. The FDA has cleared its investigational new drug (IND) application, and Orum plans to dose its first patient soon in a relapsed/refractory AML trial. Preclinical data reportedly point to a lower effective dose and a wider margin between toxicity and efficacy than ORM-6151 showed. Orum's CEO has said the company holds roughly $180 million in cash and does not expect layoffs as a result of the BMS decision.
Does this setback mean the ADC/DAC antibody-conjugate field is losing investor confidence?
Not on the evidence so far. Despite two discontinued DAC candidates from the same company in under eighteen months, Orum's linker platform drew a separate collaboration from Vertex Pharmaceuticals worth up to $945 million in 2024, and the wider antibody-conjugate field — ADCs, DACs, and bispecific-format conjugates alike — continues to attract large partnership deals. What's changing is scrutiny: sponsors are reviewing early clinical data faster and walking away sooner when the benefit-risk profile doesn't hold up, rather than carrying a marginal asset to a full readout.
How does antibody backbone characterization relate to conjugate safety issues like these?
Whatever payload is attached, an antibody-conjugate's safety margin starts with the antibody: its binding specificity, its cross-reactivity against related targets, and how cleanly its linker chemistry holds the payload until the antibody reaches the intended cell. An antibody with unexpected off-target binding or an inadequately characterized linker interaction can turn a promising payload into a systemic toxicity signal. That's why rigorous upstream antibody characterization — affinity, specificity, cross-reactivity, and epitope mapping — is treated as a derisking step, not a formality, before any conjugation program moves into the clinic.
8. Summary
The ORM-6151 discontinuation is a data point worth sitting with beyond the headline:
- BMS discontinued ORM-6151 (BMS-986497) in September 2026 after reviewing interim Phase 1 data, months ahead of the trial's original February 2027 readout target.
- It's the DAC platform's second discontinued lead asset in under eighteen months, following ORM-5029's April 2025 halt after a fatal liver toxicity signal.
- The deal structure matters: because BMS bought ORM-6151 outright in 2023, Orum keeps the $100M upfront but has no reversion rights to the asset itself, even as BMS forfeits up to $80M in remaining milestones.
- The field isn't retreating — Orum's own linker technology drew a separate $945M Vertex collaboration in 2024, and the company is advancing a next-generation candidate, ORM-1153, built on lessons from both discontinuations.
- The underlying lesson generalizes: an antibody-conjugate's safety margin is set by the antibody backbone's specificity and linker stability as much as by the payload — which is why rigorous upstream antibody characterization is a derisking step for any DAC, ADC, or bispecific conjugate program.
If your team is developing an antibody backbone for a next-generation conjugate program and needs a fully characterized starting candidate, request a Sekbio antibody development quote to discuss affinity, specificity, and expression-platform requirements.