The Drug That Cures Itself Out of Business

In April 2019, a pharmaceutical company called Achaogen filed for Chapter 11 bankruptcy. [S6] This was not a story about a reckless startup that burned through capital chasing a drug that failed. Achaogen had a working drug — Zemdri, the trade name for plazomicin — that the FDA had approved eight months earlier for complicated urinary-tract infections caused by multidrug-resistant bacteria. [S6] The problem was what happened after approval. In Zemdri's first full year on the market, the drug generated $800,000 in sales against approximately $1.5 billion in development costs. [S6, S7] That gap is not a rounding error. It is the precise output of a structural trap built into the antibiotic business — one the pharmaceutical industry had recognized, priced correctly, and been quietly exiting for two decades before Achaogen even formed.

Two other antibiotic companies with FDA-approved products filed for bankruptcy the same calendar year: Melinta and Entasis. [S6] None of them had made a scientific error. All three outcomes were predictable.

The Revenue Geometry of a Cure

The trap begins with arithmetic.

When a cholesterol-lowering statin works, the patient takes it every day for the rest of their life — potentially for decades, generating continuous revenue for the manufacturer throughout. [S7] The drug manages an ongoing condition rather than resolving it, and its success guarantees sustained demand. The same logic holds for antihypertensives, diabetes medications, and disease-modifying therapies for autoimmune conditions. Chronic disease is commercially reliable: the patient stays alive, stays unwell, and stays on the drug.

An antibiotic inverts this structure. A standard course runs five to ten days. The patient recovers, stops buying the drug, and eliminates their own future demand in the act of being cured. [S7] A drug that works perfectly generates precisely one prescription per infection episode and then competes against itself by removing the clinical need for itself. Where chronic-disease drugs earn revenue continuously, a successful antibiotic is structurally self-terminating.

The expected net present value of a new antibiotic targeting gram-negative bacteria — the most clinically dangerous resistant organisms — is negative under current market conditions. [S7] This means that even a commercially successful antibiotic is likely to destroy shareholder value. No competent board, presented with that projection, would fund the program — and for several decades, most haven't. In the 1990s and early 2000s, pharmaceutical companies including Roche, Bristol-Myers Squibb, and Eli Lilly shut down their antibacterial research programs, citing economic rather than scientific rationales. [S8] What remained were smaller biotechs that lacked the capital to sustain the multi-decade development timelines that genuine antibiotic innovation demands. [S8] Achaogen, Melinta, and Entasis were exactly these companies. The major players had run the numbers and left.

The Stewardship Paradox

The second layer of the trap activates only after a company has committed to development — and it is perhaps the most structurally perverse feature of the antibiotic business.

New antibiotics are clinically precious precisely because bacterial populations have not yet developed resistance to them. They are, by definition, last-resort options. This is why antibiotic stewardship programs — now operating across most major hospitals in wealthy countries — actively restrict their use. [S7] The reasoning is medically sound: deploy a new antibiotic broadly and selection pressure will accelerate resistance, eroding the drug's efficacy faster than if it is held in reserve for patients with no other options.

For a drug developer, the consequence is grotesque. A company spends approximately $1.5 billion over a decade or more to bring a new antibiotic to market. [S7] It then discovers that public-health norms require it to actively suppress prescriptions of the product it needs to sell in high volume to recoup that investment. [S7] The drug's clinical novelty — the quality that makes it worth having — is precisely the quality that triggers its restriction. Hospitals are advised to reserve it. Stewardship pharmacists flag its overuse. Formulary committees add approval layers for each prescribed course.

The post-approval valley of death, not regulatory failure or clinical-trial collapse, is where antibiotic companies actually go under. [S6] Zemdri worked. It was never clinically deficient. It simply wasn't prescribed often enough for the business to survive — because prescribing it often enough would have accelerated the erosion of the resistance profile that made it valuable in the first place. [S6]

The Externality Nobody Pays For

The third layer is what makes the first two permanent.

Antibiotic resistance is a textbook tragedy of the commons. [S5] Each prescriber who administers a course of antibiotics captures the full clinical benefit — bacteria killed, infection cleared — while contributing a small, diffuse increment of selection pressure to bacterial populations globally. Each individual prescription is individually rational. The collective effect is the erosion of efficacy that no single prescriber bears and that no drug company can internalize. [S5]

Resistance can begin emerging within years of a drug's widespread deployment, sometimes reaching clinically significant prevalence within a single decade. [S5] But this timeline is long enough to sever cause from consequence for everyone with a financial stake in the present. The pharmaceutical company that bore the development cost cannot capture any economic return from preserving the drug's future efficacy. The prescriber who overprescribes faces no direct consequence when resistance hardens in a different patient, in a different ward, years later. Standard market mechanisms — including intellectual property protections — are structurally insufficient to correct an externality distributed across populations and time. [S5] There is no pricing mechanism for tomorrow's untreatable infection.

A Void That Never Officially Ended

The predictable result of this three-layer trap is an antibiotic discovery drought whose depth is routinely understated.

Most accounts mark the void as running from 1987 to approximately 2003, crediting linezolid (FDA-approved 2000) and daptomycin (FDA-approved 2003) with breaking it. The chemistry disagrees. Linezolid belongs to the oxazolidinone class, first described in 1978; daptomycin belongs to the lipopeptide class, reported in 1987. [S1] Every antibiotic approved since 2000 belongs to a chemical class whose scaffold was characterized before 1987. [S1] What ended around 2000 was the gap in clinical approvals. The structural discovery gap — the absence of genuinely new antibiotic chemotypes entering clinical use — by the stricter accounting remains unbroken. [S1]

This drought runs directly across the most profitable era in pharmaceutical history: the blockbuster decades of statins, antidepressants, biologics, and HIV antivirals. The companies that exited antibiotic research during this period were not under financial duress; they were thriving in other categories. The discovery void is a record of rational capital allocation, not scientific failure.

What Legislation Can and Cannot Fix

Congress reached for a policy lever in 2012. The GAIN Act — Generating Antibiotic Incentives Now — offered developers of qualifying antibiotics five additional years of market exclusivity, alongside fast-track designation and priority review. [S3] Annual antibiotic approval rates roughly doubled after the legislation passed, rising from about 0.8 per year before 2012 to 1.8 per year afterward. [S4]

Most rigorous assessments describe this outcome as disappointing. [S4] The mechanism is the time value of money. A five-year exclusivity extension benefits a company developing an incremental modification of an existing antibiotic — a project perhaps five to eight years from start to approval — far more than it benefits a company developing a genuinely novel scaffold, which may require fifteen or twenty years in development. [S4] The incentive shrinks in present value as the timeline lengthens. The perverse result: the GAIN Act may have reinforced rather than corrected the industry's bias toward incremental chemistry over foundational discovery, rewarding the modifications that least address the underlying problem. [S4]

The 2019 cluster of bankruptcies generated renewed policy urgency. The most-discussed structural fix is a subscription or "Netflix" payment model: governments pay a flat annual fee for access to an antibiotic regardless of volume dispensed, decoupling revenue from prescription count. [S6] A company could then support restrictive stewardship without destroying its own finances. A version has been piloted in the United Kingdom; it remains exceptional globally. [S6] Novel antibiotic research continues to emerge primarily from public grants and academic laboratories rather than commercial pipelines. [S4]

The Bill, Presented

In 2024, the WHO published its updated Bacterial Priority Pathogens List: 24 resistant organisms across 15 bacterial families, ranked by public health burden, treatability, and R&D pipeline gaps. [S2] Carbapenem-resistant Klebsiella pneumoniae ranked highest, with a composite urgency score of 84 percent. [S2] The critical tier is dominated by gram-negative bacteria — organisms that are biochemically difficult to kill and commercially unattractive to develop against: simultaneously the most dangerous target and the least viable investment. [S7]

Read as a policy document, the WHO list outlines research priorities. Read as a ledger, it itemizes four decades of correct market decisions compounding into crisis. Each pathogen in the critical tier reflects the rational behavior of pharmaceutical companies that ran the numbers on antibiotic investment, found the expected net present value negative, and moved capital accordingly. [S7] The WHO's 2024 critical tier is the invoice that pharmaceutical economics wrote across four decades and is now delivering to intensive care units worldwide.

The problem was never greed, negligence, or scientific failure. The market looked at antibiotics, ran the arithmetic accurately, and reached the right conclusion. The catastrophe is simply that it was right.