Antimicrobial Resistance: Drug Pipeline Status and Global Mortality Projections

Executive Summary

The Global Research on Antimicrobial Resistance (GRAM) Project forecasts that bacterial AMR will cause 39 million deaths between 2025 and 2050, equating to three deaths every minute. Two forces are colliding to produce this trajectory: resistance is accelerating fastest in the regions least equipped to contain it, while the antibiotic pipeline is generating only modest innovation against the pathogens of greatest concern. In May 2026, member states at the 79th World Health Assembly adopted the updated Global Action Plan on Antimicrobial Resistance for 2026-2036, marking the highest-level political commitment to the issue yet. For decision-makers in healthcare, pharmaceuticals, and public health policy, the structural mismatch between where resistance is killing people and where new treatments are being developed represents both a governance failure and a commercial opportunity that existing incentive structures have consistently failed to close.

The Resistance Trajectory: Rate Matters As Much As Level

According to the CDC, AMR is now one of the world's most urgent public health problems, and as recently as 2019 it was associated with approximately 5 million total global deaths, including over 2.8 million annual infections in the US alone. These figures are a baseline, not a ceiling.

Trajectory, not just level: The GRAM project estimates that 1.91 million annual deaths will be directly due to AMR by 2050, a 67.5% increase from 1.14 million in 2021. The policy error most commonly made is treating current mortality as the measure of urgency. The trajectory, driven by the compounding of resistance genes across bacterial populations, means that each year of delayed action creates a steeper curve to reverse. The WHO GLASS 2025 report, drawing on over 23 million bacteriologically confirmed cases from 104 countries, is the most surveillance dataset ever assembled, and its findings on resistance rates in developing regions suggest that aggregate global figures systematically underweight the burden in low-surveillance environments.

Deaths due to MRSA increased most globally, leading directly to 130,000 deaths in 2021, more than doubling from 57,200 in 1990; among Gram-negative bacteria, resistance to carbapenems increased more than any other antibiotic type, rising from 127,000 to 216,000 between 1990 and 2021. Both trends spill directly into healthcare economics: MRSA predominantly drives hospital-acquired infection costs in high-income settings, while carbapenem-resistant Gram-negatives drive mortality in settings where intensive care options are limited.

The economic implications compound the clinical ones. The World Bank estimates that AMR could result in US$1 trillion in additional healthcare costs by 2050 and US$1 trillion to US$3.4 trillion in GDP losses per year by 2030. These projections translate directly into country-level fiscal risk, particularly for governments with large public healthcare commitments and aging populations, creating both economic and political pressure to fund AMR countermeasures that current commercial antibiotic markets cannot generate independently.

Where Self-Medication And Overuse Compound Systemic Risk

The driver picture is not monolithic. The misuse and overuse of antimicrobials in humans, animals, and plants are the main drivers in the development of drug-resistant pathogens. But the mechanisms differ sharply by region.

In India, a country that the Indian Council of Medical Research (ICMR) identifies as carrying one of the world's largest AMR burdens, Prime Minister Modi used his December 2025 Mann Ki Baat address to warn against indiscriminate antibiotic use, cautioning that growing resistance was making common infections such as pneumonia and urinary tract infections increasingly difficult to treat. The Times of India documented physicians treating patients who had previously self-medicated with antibiotics based on internet searches and prior prescriptions, only to develop resistant infections requiring intensive care. What is not being reported in many of these settings is the scale of antibiotic dispensing through informal channels, pharmacies without prescription requirements, and agricultural use, all of which represent invisible inputs to resistance that surveillance systems struggle to capture.

Japan presents the most rigorously documented counter-case. Live Science reported that as recently as 2005, national insurance data showed 60% of patients with non-bacterial upper respiratory infections in Japan were prescribed antibiotics, mostly broad-spectrum drugs such as third-generation cephalosporins, macrolides, and quinolones. Japan's 2016 National Action Plan on Antimicrobial Resistance, aligned with the WHO framework, produced measurable declines in prescription rates. The Japanese intervention demonstrates that stewardship policies can generate population-level results, but also reveals how slowly the system responds: Japan's elderly demographic, in which people over 65 make up approximately 30% of the population, means that AMR-related illness exposure is structurally embedded in its demographic profile regardless of prescription reform.

The European picture is more acute in a different vector. The Guardian's Peter Beyer commentary published in June 2026 identified drug-resistant gonorrhea as an early warning signal for broader AMR spread, noting that the traditional commercial model for antibiotic R&D has repeatedly failed to deliver the drugs most needed, particularly for low- and middle-income country populations where expected returns are lowest. Over 82 million people are infected with Neisseria gonorrhoeae annually, but the bacterium has developed resistance to almost all antibiotics, with a six-fold increase in resistant infections to ceftriaxone observed in some countries, particularly Cambodia and Vietnam.

The Pipeline: Genuine Progress And Structural Gaps

The treatment development picture entering mid-2026 contains genuine bright spots alongside persistent structural deficits.

On the approval side, December 2025 produced the most significant gonorrhea treatment advance in decades. The FDA approved zoliflodacin and gepotidacin for uncomplicated urogenital gonorrhea, marking the first oral treatments in nearly 40 years; clinical trials showed cure rates of 91% for zoliflodacin and 93% for gepotidacin.

Zoliflodacin's development, conducted under a not-for-profit model through the Global Antibiotic Research and Development Partnership (GARDP), demonstrates a different public-private partnership approach that prioritizes global health needs; Innoviva will market the drug in North America and Western Europe while GARDP pursues access in low- and middle-income countries.

Beyond gonorrhea, Scripps Research published findings in June 2026 in Nature Communications on a strategy to restore vancomycin effectiveness against VREfm, a hospital-acquired infection becoming increasingly common. The researchers focused on disabling a specific bacterial enzyme called secreted antigen A (SagA), present across a wide range of VRE strains, potentially making resistant bacteria vulnerable to vancomycin again without requiring new antibiotic development. Separately, Ars Technica reported a Nature study in June 2026 on the discovery of an "anti-biotin megacluster" that researchers describe as providing a paradigm for naturally evolved combination therapies, suggesting that antibiotic discovery should shift from isolating individual compounds to reconstructing native synergistic systems.

At the company level, Telum Therapeutics announced in July 2026 a EUR 18 million Series A financing to advance its lead program targeting hospital-acquired and ventilator-associated bacterial pneumonia caused by Acinetobacter baumannii, led by the AMR Action Fund. Acinetobacter baumannii is among the pathogens WHO designates as critical priority, with carbapenem-resistant strains now causing infections with extremely limited treatment options in hospital settings globally.

Short-term gain, long-term cost: Despite 90 new antibacterial agents in preclinical or clinical development as highlighted in WHO's 2025 antibacterial pipeline analysis, few clinical candidates target bacterial priority pathogens and even fewer are considered innovative. The FDA approvals of late 2025 are genuine milestones, but they address a single indication. The broader Gram-negative pathogen pipeline, which the GRAM project identifies as capable of averting 11 million deaths if successfully developed, remains structurally underfunded. Without additional investment, the pipeline is expected to continue to gradually decline, particularly from 2026 onwards when existing funding for late-stage Phase II and III studies is expected to decline; under a no-new-incentives scenario, the pipeline is expected to contain only 26 treatments in ten years, of which only 6 would be in late stages of development.

The broader systemic implication is that both economic and political dimensions of antibiotic R&D failure compound each other: the commercial model underproduces drugs targeting resistant Gram-negatives because profit margins are thin, while the not-for-profit model demonstrated by GARDP has proven it can work but lacks the scale to fill the entire pipeline gap.

Counterarguments

1. The 39-million-deaths forecast may be overstated due to model assumptions about non-intervention. The GRAM project's headline projection assumes a trajectory absent meaningful intervention. The same analysis projects that improved care and antibiotic access could avert 92 million deaths between 2025 and 2050, a figure nearly 2.5 times the projected AMR-attributable deaths. This suggests the forecast is highly sensitive to policy choices, and that the worst-case framing, which dominates media coverage, risks producing either fatalism or misallocated urgency. The evidence base genuinely supports a range of outcomes contingent on whether the GAP-AMR 2026-2036 generates funded implementation rather than political declaration.

2. The pipeline assessment may undercount non-traditional approaches that are already in late development. The WHO pipeline analysis focuses on traditional small-molecule antibiotics. BioTechniques reported in 2026 on AI-enhanced antibiotic discovery targeting multi-drug-resistant gonorrhea using deep learning to identify entirely new chemical structures with novel cellular pathway targets, reducing resistance probability. News-Medical reported in June 2026 that a drug repurposing screen of over 5,000 compounds uncovered antiviral properties in common antibiotics against hantavirus infections, suggesting that the utility of existing compounds may be broader than current labeling indicates. If phage therapy, microbiome interventions, and AI-guided repurposing are counted as part of the effective pipeline, the scarcity framing requires qualification.

3. Country-level incentive structures may undermine the WHO's One Health framing by treating AMR as a health-sector problem. The GAP-AMR's strength is its framing of AMR as a cross-sectoral challenge spanning human health, agriculture, and environment. But WHO GAP-AMR draft data shows that antibiotic use in the agrifood sector could increase by 29. levels by 2040, potentially reaching 142,481 tonnes. Agricultural ministries in major food-producing nations have consistently resisted binding restrictions on veterinary antibiotic use, and the One Health framework has no enforcement mechanism. If agricultural and environmental antibiotic use continues growing while clinical stewardship improves, net resistance drivers may not decline.

Indicators To Watch

The following table identifies observable signals that a risk officer or policy researcher can track to assess whether the AMR trajectory is improving or deteriorating.

Decision Relevance

Scenario A (~55%): Incremental progress, persistent access gaps. The not-for-profit pipeline produces two to three new approvals per decade targeting high-priority pathogens, but access to low-income countries remains delayed by commercial distribution models. Resistance rates in South Asia and sub-Saharan Africa continue rising above global averages. If you advise on healthcare procurement in emerging markets or hold positions in the AMR-adjacent biotech sector, accelerate supply agreements with GARDP-affiliated developers rather than waiting for commercial distributor coverage; the access gap is structural, not temporary. If you are a corporate risk officer in food and beverage or agriculture, assess your supply chain's antibiotic-use disclosure exposure as the WHO GAP-AMR 2026-2036 framework is moderate-to-high confidence to generate new mandatory reporting mechanisms at country level within three to five years.

Scenario B (~30%): Governance breakthrough, pipeline acceleration. The WHO GAP-AMR 2026-2036 generates genuine implementation financing, pull incentives for Gram-negative antibiotic development are adopted in major markets, and national action plans in high-burden countries move to funded implementation. If you are evaluating early-stage investment in antibiotic R&D, this scenario represents the clearest positive inflection for returns, particularly in Gram-negative platforms; pre-position in the companies whose pipeline targets align with the WHO priority pathogen list. If you hold health insurer or hospital group positions in high-income aging economies, this scenario reduces the compound surgical and oncological cost exposure that resistant infections impose on complex care pathways.

Scenario C (~15%): Accelerated resistance without pipeline catch-up. Agricultural antibiotic use continues rising, Gram-negative carbapenem resistance breaches thresholds in additional high-income settings, and the pipeline produces fewer than two new priority-pathogen approvals in the next five years. If you have operations in healthcare delivery sectors, stress-test your infection control protocols against a scenario where one or two current last-resort antibiotics see significant resistance upticks, as the Scripps vancomycin-resistance research and CIDRAP's pipeline reports both point to this as the direction of travel without intervention. If you lack direct healthcare exposure, monitor the AMR Action Fund's disbursement rate and the IFPMA's annual pipeline review as the earliest quantitative signals of which scenario is materializing.

• The WHO GLASS 2025 dataset draws on 104 reporting countries, but the highest-burden regions, including parts of South Asia and sub-Saharan Africa, have the weakest laboratory and reporting infrastructure. Resistance rates in these regions are moderate-to-high confidence underestimated by a margin that current methodology cannot precisely quantify.
• The GRAM Project's 39-million-deaths forecast is a modeled projection with a 95% uncertainty interval spanning 33 to 46 million deaths. It assumes no new broadly effective Gram-negative antibiotic class is developed; if the AI-guided and megacluster-based discovery strategies reported in mid-2026 yield clinical candidates within five to seven years, the upper bound becomes significantly less moderate-to-high confidence.
• Agricultural antibiotic use data is reported by national governments on a voluntary basis with limited independent verification. The 142,481-tonne-by-2040 projection in the WHO GAP-AMR draft is based on trend extrapolation and does not capture potential regulatory interventions that several major food-producing countries have signaled but not implemented.
• The access plans for zoliflodacin and gepotidacin in low- and middle-income countries remain incomplete, per the 2026 Access to Medicine Foundation AMR Benchmark. If GARDP's distribution strategy for zoliflodacin does not achieve regulatory approval in high-burden settings within two to three years, the clinical relevance of the December 2025 FDA approvals to global AMR mortality will be limited despite their scientific significance.
• This assessment does not extend to antifungal, antiviral, or antiparasitic resistance, which the WHO defines as part of the broader AMR challenge. The omission is scope-driven; fungal AMR in particular, driven by Candida auris spread, represents an adjacent risk that intersects with the same hospital-acquired infection pathways as bacterial resistance.