BioNTech tries again with MabVax

The struggling US biotech MabVax Therapeutics is no more, but its technology lives on, with the revelation that an ADC coded BNT329 is about to go into clinical trials. BNT329, owned by BioNTech, appears to be based on an anti-CA19-9 antibody that BioNTech acquired from MabVax six years ago, before the latter went out of business.

This is one of five ADCs entering their first human studies, according to new listings on clinicaltrials.gov. These include an ADC version of AstraZeneca's anti-Steap2 Car-T therapy, plus bispecific conjugates from BeOne, Phrontline and Hangzhou DAC, signifying yet again how important the ADC modality is for biopharma at present.

CA19-9

BioNTech's BNT329 targets carbohydrate antigen sialyl-Lewis A (also known as CA19-9), an antigen present in gastrointestinal tumours including, most importantly, pancreatic cancer.

The group's involvement here dates back to 2019, when it acquired the antibody assets of MabVax, a US biotech that was running out of cash, and which subsequently entered bankruptcy proceedings. Among the assets was an anti-CA19-9 MAb coded MVT-5873 that was then in phase 1, and was subsequently taken forward by BioNTech as BNT321.

Development of BNT321 was terminated in January, but now comes disclosure that BNT329 is to enter the clinic next month. The new molecule uses the topoisomerase 1 inhibitor payload YL0010014, which also lies behind several BioNTech/MediLink ADCs, suggesting that BNT329 is the result of BioNTech applying MediLink's Tmalin linker-payload technology to BNT321.

According to OncologyPipeline the only other group to have worked on CA19-9 is the radiotherapeutics player Clarity Pharmaceuticals, but this work has stopped. MabVax/BioNTech had also earlier worked on radiopharmaceutical versions of BNT321.

Recently disclosed first-in-human studies*

Note: *projects newly listed on the clinicaltrials.gov database between 17 and 22 Sep 2025.

With bispecific ADCs seeing increasing buy-in, owing perhaps to interest in SyntImmune/Bristol Myers Squibb's izalontamab brengitecan, it's notable that three new clinical ADC entrants are bispecifics; two hit one of iza-bren's targets, EGFR, but combine this with activity at either B7-H3 (Phrontline's TJ101) or cMet (BeOne's BG-C0902).

BG-C0902 is clearly an attempt to improve on Johnson & Johnson's anti-EGFR x cMet Rybrevant, a naked MAb, following in the footsteps of AstraZeneca's similarly acting ADC tilatamig samrotecan. Meanwhile, Hangzhou DAC's DXC014 hits B7-H3 and PSMA, though targeting these two antigens together seems unusual.

AstraZeneca is going into the clinic with AZD0516, a monospecific anti-Steap2 ADC, about 18 months after AZD0754 entered human trials. AZD0754, an anti-Steap2 Car-T, is one of Astra's first internally developed cell therapy efforts, and the company also has a preclinical-stage anti-Steap2 T-cell engager, AZD6621, which is said to recruit CD8+ T-cells preferentially. No other industry players appear to be active in Steap2.

The T-cell engager modality also features among the latest first-in-human entrants, in the form of Normunity's NRM-823; this private US biotech closed a $75m series B round in January, and is keeping NRM-823's mechanism secret for now. However, it's known to have filed patents for MAbs acting on PLA2G10 and LY6K, so these are possible targets for NRM-823.