The largest infectious disease congress in Europe wrapped up this week in Munich. Four days of data, press releases, pipeline corners, and poster sessions. Tens of thousands of clinicians, researchers, and industry representatives. Here is what actually mattered.
A company of roughly sixty people in Edmonton, Alberta, presented two fundamentally different weapons against the same class of bacterial target — penicillin-binding proteins — at the same congress. One is already in Phase 3 with real efficacy data against carbapenem-resistant infections. The other is a first-in-class molecule built on chemistry that was discovered in 1986 and has never, until now, made it anywhere near a patient.
Fedora Pharmaceuticals is not a name that appears in pipeline surveys or Wall Street analyst notes. It has no stock ticker, no celebrity CSO, no splashy venture announcements. What it has are two drug candidates that, together, represent more genuine innovation against Gram-negative superbugs than most pharmaceutical companies ten times its size have produced in a decade.
Two weapons
The first is nacubactam, a beta-lactamase inhibitor with a difference: it doesn’t just protect a partner antibiotic from destruction — it also directly attacks PBP2, an essential enzyme in the bacterial cell wall. That dual mechanism makes the combination of cefepime-nacubactam or aztreonam-nacubactam effective against bacteria that have learned to resist every standard carbapenem.
At ESCMID, the Integral-2 Phase 3 results confirmed high efficacy against carbapenem-resistant Enterobacterales across multiple infection types — the patients for whom the current standard of care is already failing. Aztreonam-nacubactam achieved a 72.3% composite response rate against CRE. The oral presentation was selected for the podium. A poster on in vitro characterisation was designated top-rated — top 2% of all ESCMID submissions. Meiji Seika, Fedora’s Japanese partner, has filed for approval in Japan. Fedora holds the US rights.
The second weapon is stranger and, potentially, more important.
FPI-2119 is a derivative of lactivicin, a natural product first isolated from Empedobacter lactamgenus by Takeda researchers in 1986. Lactivicin does something no other natural product does: it binds penicillin-binding proteins without being a beta-lactam. No beta-lactam ring means no vulnerability to beta-lactamases — the enzymes that bacteria have spent billions of years evolving to destroy penicillins, cephalosporins, and carbapenems. Forty years of chemistry couldn’t solve lactivicin’s poor penetration into Gram-negative bacteria. Fedora did.
Eight posters and an oral presentation at ESCMID confirmed FPI-2119’s activity against Pseudomonas aeruginosa, Enterobacterales, Acinetobacter baumannii, Neisseria gonorrhoeae, Campylobacter, Salmonella, and Shigella. Beta-lactamase stability was confirmed against E. coli carrying a broad range of resistance genes. No measurable cardiotoxicity. Pharmacokinetics consistent with once- or twice-daily dosing. Mouse UTI models showed dose-dependent bacterial clearance. Karen Bush, one of the world’s foremost experts on beta-lactamases, said the data shows FPI-2119 “has the potential to become a valuable tool to fight today’s infectious diseases.”
FPI-2119 is preclinical. No human has received it. The MIC numbers from the oral presentation are not yet in the public domain. But the breadth of activity — spanning the three most dangerous Gram-negative pathogens plus gonorrhea and enteric infections — from a single non-beta-lactam scaffold is, if confirmed in humans, the kind of advance the field has been waiting for since the 1960s.
One lab. Sixty people. Two chemical classes. One targeting the same essential machinery through entirely different molecular architectures.
Who else showed up
BioVersys, roughly eighty people in Basel, dosed the first patient in its Phase 3 CRAB trial during congress week itself — a company so small that its CEO presented at a session on how funding shapes antibiotic innovation. Their five posters confirmed that BV100, an IV rifabutin formulation, overcomes the rpoB mutations that defeat rifampicin, and showed synergy with polymyxin B even against polymyxin-resistant isolates. A second pipeline compound, BV500, targets Mycobacterium abscessus under a Shionogi collaboration.
Shionogi itself brought real-world cefiderocol data: the CIRCE study from Spain showed 68% clinical cure at day 14 and 83% survival at day 28 in MBL-producing Enterobacterales infections — the hardest-to-treat Gram-negative cases. Microbiological eradication was 85% in bloodstream infections.
Wockhardt, in Mumbai, wasn’t at ESCMID but its shadow was. Zaynich — cefepime-zidebactam, another PBP2 enhancer like nacubactam — is under FDA priority review with a decision expected by mid-2026. The first NCE from an Indian pharmaceutical company to reach a US NDA. Nacubactam and Zaynich share a mechanism but different development strategies: one fixed-dose, one flexible combination. Both could reach the US market within two years.
Who didn’t
Roche had zero presentations on zosurabalpin, the tethered macrocyclic peptide that was going to be the first new Gram-negative antibiotic class in fifty years. Phase 3 was announced for “early 2026.” It is now late April. No first-patient-first-visit announcement. No NCT registration. No enrollment confirmation. At the largest European infectious disease congress, where BioVersys had five presentations, nacubactam had ten, and FPI-2119 had ten, zosurabalpin had none.
Roche’s presence at ESCMID 2026 was focused entirely on diagnostics.
This may mean nothing — timelines slip, Phase 3 preparations are complex, data embargoes exist. But the absence is now the data point. The one Big Pharma company remaining in novel Gram-negative antibiotic development presented nothing on its flagship molecule at the field’s flagship congress.
The pattern
Fedora. BioVersys. Wockhardt. Shionogi. The companies carrying the antibiotic pipeline are not the ones you would expect. They are mid-sized Japanese firms, Swiss biotechs with fewer employees than a single Roche division, Indian generics companies that decided to build a research arm, and a sixty-person Canadian outfit that licensed a molecule to Roche for $750 million in potential milestones, watched the relationship end, took the rights back, found a Japanese partner, and showed up at ESCMID with two weapons instead of one.
This is what the antibiotic pipeline actually looks like in 2026. Not a crisis of science — the molecules exist. A crisis of who shows up to develop them.
ESCMID Global 2026 ran April 17–21 in Munich. This dispatch covers press releases and company announcements from the congress. MIC data from FPI-2119 oral presentations and detailed Integral-2 efficacy breakdowns remain embargoed or unpublished as of April 20, 2026. This is Post #30. Previously: The Reckoning (azithromycin’s triple failure), The Broken Economics (market failure), The Phage Paradox (commercialization crisis).