Superbugs Kill a Million People a Year. Curing Them Is a Failing Business.

In June 2018, a California biotech called Achaogen won FDA approval for Zemdri, a new antibiotic for multidrug-resistant urinary tract infections — exactly the kind of drug the World Health Organization had been begging the industry to build. Ten months later, Achaogen filed for Chapter 11. The company did everything right: real science, a cleared regulatory bar, a drug targeting carbapenem-resistant bacteria on the WHO's priority list. The market's verdict was bankruptcy anyway.

That was seven years ago, and every investor in the sector remembers it. It is the single cleanest explanation for why the drugs that underwrite all of modern medicine — surgery, chemotherapy, transplants, childbirth — are quietly running out.

The pipeline is shrinking while the bugs accelerate

In March, the Access to Medicine Foundation published its 2026 Antimicrobial Resistance Benchmark, the industry's periodic report card. The headline finding: the number of antimicrobial projects in development at large pharmaceutical companies has fallen 35% in five years, from 92 to 60. Only five of those are for children under five, the group most vulnerable to resistant infections.

The demand side is moving in the opposite direction. More than a million people die each year directly from drug-resistant infections, and resistance contributes to roughly four million deaths — figures the foundation expects to roughly double by 2050. The World Health Organization reported in October 2025 that one in six laboratory-confirmed bacterial infections worldwide was resistant to antibiotic treatment in 2023, and that more than 40% of the antibiotics it tracks lost potency against common bloodstream, gut, urinary tract, and sexually transmitted infections between 2018 and 2023.

Against that, the supply side has consolidated to almost nothing. The Benchmark found just three large pharma companies still meaningfully invested in antimicrobial R&D: GSK, with 30 active projects, plus Japan's Shionogi and Otsuka. Pfizer, which shared the top spot with GSK in 2021, has fallen back. AstraZeneca isn't ranked at all — it has no antibiotic portfolio.

Briefings like this land in members' inboxes before the market prices them in. Join free →

Why curing people doesn't pay

The economics are structurally broken in a way no other drug class shares. By standard industry estimates, a new antibiotic takes roughly a decade and well over a billion dollars to develop — comparable to an oncology drug. But where a new cancer therapy is prescribed as widely as possible at six-figure prices, a new antibiotic is deliberately locked in the cabinet. Stewardship — the medically correct practice of reserving novel antibiotics for infections nothing else can treat — means the better and newer the drug, the less it sells. A course lasts days, not years. Revenue models built on volume collapse when the product's entire clinical value lies in not being used.

The cost of that broken model lands everywhere else in medicine. Infection is the second-biggest killer of cancer patients; Lord Ara Darzi, who chairs the Fleming Initiative on antimicrobial resistance, made the point at the Benchmark's launch that cutting-edge oncology is pointless if patients then die of infections that were treatable a decade ago.

What makes this maddening is that the science is working. In December, the FDA approved two new antibiotics in the same month: Innoviva's Nuzolvence, the first new oral treatment for gonorrhea in decades, and GSK's Blujepa for uncomplicated urinary tract infections and urogenital gonorrhea. The Benchmark counts seven more promising late-stage programs — from GSK, Otsuka, Shionogi, BioVersys, F2G, Innoviva, and Venatorx — aimed at the deadliest resistant pathogens. The bottleneck is not chemistry. It is the absence of a buyer.

The money arriving from outside the market

Because the market won't pay, philanthropy and governments are being forced to. Last month the Gates Foundation, Novo Nordisk Foundation, and Wellcome launched Gr-ADI, a $60 million consortium funding 18 research teams across 17 countries to attack Gram-negative bacteria — the deadliest resistance driver — using AI-enabled discovery, with all data shared openly. It is the first tranche of a $300 million joint health-R&D partnership. CARB-X, the Boston-based nonprofit accelerator, opened its 2026 funding round in February. The prize for getting this right is enormous: research published in The Lancet estimates a steady supply of new Gram-negative antibiotics could avert 11.1 million deaths over the next 25 years, against resistance-driven economic losses the UN puts at more than $1 trillion globally.

But grants fix discovery, not the market. The structural fixes are payment reform. The UK moved first: since 2022, the NHS has paid Pfizer and Shionogi fixed annual subscription fees for access to two critical antibiotics — paying for existence rather than volume, like a Netflix contract — and expanded the scheme in 2024 with a budget of £100 million a year. Washington has its own version drafted: the PASTEUR Act, which would pay developers of critical new antimicrobials $75–300 million per year in federal subscription contracts. It was reintroduced in June with bipartisan sponsors in both chambers. It has been introduced every Congress since 2020 and has never once reached a floor vote — while the US logs 2.8 million resistant infections and roughly 35,000 deaths a year.

Which sets up the real question for investors: if a G7 government ever flips from paying per pill to paying for existence — and one already has — a small set of listed companies gets repriced overnight. The names, the handicapping, and what they're worth if Washington never acts, below.

The rest of this briefing is free — it just requires a free AlphaBriefing account: the five listed names with real antimicrobial exposure and what each is actually worth owning for, the PASTEUR trade handicapped honestly, and the one adjacent industry that gets paid whether Congress acts or not.

Create your free account → — 30 seconds, no card.