Autogene cevumeran (BNT122) by BioNTech: status and timeline

Autogene cevumeran (BNT122)

Stop cancer from coming back after surgery, starting with pancreatic cancer, where most patients relapse, by teaching each patient's T cells to recognise their own tumour's mutations.

PilotStage 3 of 5

Pancreatic Phase 2 (about 260 patients, enrollment closed) continues; colorectal Phase 2 stopped Aug 2026.

Updated 28 Aug 2026Checked 11 Oct0 updates this week

Milestones

Next · IMCODE003 primary completion (estimated)

Colorectal Phase 2 (BNT122-01) starts8 Mar 2021Complete.
Pancreatic Phase 1 results published in Nature10 May 2023Complete.
Randomized pancreatic Phase 2 IMCODE003 starts18 Oct 2023Complete.
3-year pancreatic follow-up published in Nature19 Feb 2025Complete.
Colorectal Phase 2 stopped after survival imbalanceMissed · 28 Aug 2026Missed.
IMCODE003 primary completion (estimated)Target Jan 2031Not yet reached.

Most important updates

Upcoming

Jan 2031IMCODE003 primary completion (estimated)in about 220 weeks (next)

Current obstacles

Colorectal failureGiven alone in colorectal cancer, it showed no benefit and a survival imbalance; it is unclear how much this says about the combination in pancreatic cancer.
Only half of patients respondIn Phase 1 only 8 of 16 patients made strong vaccine T cells; predicting or raising that share is unsolved.
Long wait for proofThe randomized pancreatic trial's main results are not expected until around 2031, per its registry entry.

Physics limits

Few targets in some tumoursVaccines need mutations that look foreign. Pancreatic and many colorectal tumours carry few, so there are fewer good targets and weaker responses.
Tumours hide and suppressCancer cells can stop showing the targeted fragments or build a shield of suppressive cells, so trained T cells may find nothing to kill.
Made-to-order takes timeSequencing and making a unique drug for each patient takes weeks (about 9 weeks from surgery to first dose in Phase 1), during which surviving cancer cells can grow.

How it works

4 parts

Sequence

Read the tumour's mutations

The removed tumour and normal blood are sequenced. Software picks up to 20 mutated protein fragments the immune system is likely to see.

Make

One batch per patient

mRNA for those fragments is made for that patient alone, packed in tiny fatty particles and infused into a vein, about 9 weeks after surgery in Phase 1.

Train

Raise killer T cells

Immune cells read the mRNA, display the fragments and train T cells. In Phase 1 responders these T cells reached up to 10% of all blood T cells.

Combine

With checkpoint drug and chemo

In pancreatic cancer it follows a dose of atezolizumab, a drug that releases the immune system's brakes, and is followed by FOLFIRINOX chemotherapy.

Spec sheet

Spec	Autogene cevumeran (BNT122)
Neoantigens per patient	Up to 20^R (reported)
Delivery	Intravenous mRNA-lipoplex, about 400 nm^R (reported)
Phase 1 pancreatic responders	8 of 16 vaccinated^R (reported)
Non-responder median recurrence-free	13.4 months (responders: not reached)^R (reported)
Vaccine T-cell clone lifespan	About 7.7 years (estimated median)^R (reported)
IMCODE003 enrollment	About 260 patients^R (reported)

^R reported by the company

Update log

4 updates

About BioNTech

The team behind Autogene cevumeran (BNT122)

BioNTech

mRNA vaccines, cancer immunotherapy, oncology, personalized medicine

BioNTech develops mRNA vaccines and immunotherapies for cancer and infectious disease. Founded 2008, public since 2019. In 2025 acquired CureVac and partnered with BMS on BNT327.

Publicly traded on Nasdaq (BNTX); 7,269 employees as of March 2026. Full-year 2025 revenue: €2.87 billion.
June 2025 BMS partnership: $1.5B upfront + $2B non-contingent + up to $7.6B in milestones for BNT327.
June 2025: Acquired CureVac for $1.25 billion to expand mRNA oncology pipeline.

biontech.com (opens biontech.com)Company profile →

Founded: 200818 yrs
Headquarters: Germany
Status: Public
Listed: BNTX (opens google.com)NASDAQ
Valuation: public
Staff: ~7,269Mar 2026
Last round: IPO (Nasdaq) · $150MOct 2019
Works in: BiotechRNA medicines
Coverage: 2 programs · 7 updateslatest 28 Aug 2026checked 10 Oct
Partners: Bristol Myers Squibb (opens biontech.com)Pfizer (opens pfizer.com)Genentech (opens gene.com)
People: Uğur Şahin (opens en.wikipedia.org)Co-founderÖzlem Türeci (opens en.wikipedia.org)Co-founderChristoph Huber (opens linkedin.com)Co-founder

Spec

Autogene cevumeran (BNT122)

Neoantigens per patient

Up to 20^R (reported)

Delivery

Intravenous mRNA-lipoplex, about 400 nm^R (reported)

Phase 1 pancreatic responders

8 of 16 vaccinated^R (reported)

Non-responder median recurrence-free

13.4 months (responders: not reached)^R (reported)

Vaccine T-cell clone lifespan

About 7.7 years (estimated median)^R (reported)

IMCODE003 enrollment

About 260 patients^R (reported)

Autogene cevumeran (BNT122)

Milestones

Most important updates

Upcoming

Current obstacles

Physics limits

How it works

Read the tumour's mutations

One batch per patient

Raise killer T cells

With checkpoint drug and chemo

Spec sheet

Update log

Fri 28 Aug

Wed 19 Feb 2025

Wed 18 Oct 2023

Wed 10 May 2023

About BioNTech

Current obstacles

Physics limits

How it works

Read the tumour's mutations

One batch per patient

Raise killer T cells

With checkpoint drug and chemo

Spec sheet

Update log

Fri 28 Aug

Wed 19 Feb 2025

Wed 18 Oct 2023

Wed 10 May 2023

About BioNTech