Beyond approval: Building a coordinated, adaptive evidence ecosystem for novel antibiotics
– by Ursula Theuretzbacher

The views and opinions expressed in this article are solely those of the original author(s) and do not necessarily represent those of GARDP, their donors and partners, or other collaborators and contributors. GARDP is not responsible for the content of external sites.

Rising antimicrobial resistance in Gram-negative bacteria has intensified the need for new antibiotics that overcome established resistance mechanisms with minimal or no cross-resistance to existing agents. For clinicians caring for critically ill patients, the priority is timely access to effective therapies against difficult-to-treat pathogens, especially carbapenem-resistant (CR) Gram-negative bacteria. Recent approvals of agents targeting these organisms demonstrate that regulatory innovation can accelerate access. At the same time, they expose a structural limitation: at the time of launch, the available evidence is often insufficient to guide optimal use in the very populations most in need.

New antibiotics developed to address unmet medical needs often lack sufficient evidence of added clinical benefit at the time of approval to justify high costs.

Limited patient availability requires expedited trials

The major limitations in current antibiotic clinical development for narrowly defined populations such as infections caused by CR Acinetobacter, CR Enterobacterales or CR Pseudomonas aeruginosa, are patient availability for clinical efficacy trials as well as methodological limitations. Even large multinational programs struggle to enrol substantial numbers of patients and lack any formal hypothesis testing.^1,2 As a result, regulators accept smaller datasets and larger non-inferiority margins under expedited pathways designed for serious infections with limited treatment alternatives. These pathways allow smaller, faster development programs, but inevitably leave greater uncertainty at approval. New antibiotics developed to address unmet medical needs often lack sufficient evidence of added clinical benefit at the time of approval to justify high costs.^3,4 Approval is typically accompanied by specific postmarketing requirements and commitments. Postmarketing requirements are mandatory commitments and address specific regulatory uncertainties related to safety, efficacy, pharmacokinetics, resistance or specific patient populations (e.g., paediatric patients). However, most product-specific studies are unlikely to generate the broader evidence needed to guide optimal clinical use, thereby shifting a substantial proportion of evidence generation into the post-approval phase.

At the time of approval, critical clinical questions often remain unresolved. Efficacy and safety in key patient subgroups such as profoundly immunocompromised hosts, patients receiving renal replacement therapy or extracorporeal membrane oxygenation, individuals with extreme obesity or cachexia, and patients with very high illness severity scores, are often underexplored. The presence of polymicrobial infections and regional variation in background therapy and standard of care further complicate interpretation. Limited geographic representation may hinder extrapolation to regions with distinct resistance epidemiology and standards of care. A significant challenge is the persistent delay in generating and incorporating paediatric data into drug labels. Consequently, similar evidence-generation strategies are needed for paediatric and adult populations alike.⁵ In all these groups, Important uncertainties regarding clinical positioning and use in high-risk populations remain.

In high-income countries, off-label use begins immediately, antimicrobial stewardship policies vary across institutions, and strong preferences regarding drug positioning emerge despite the absence of robust comparative data.

The recent approvals of cefiderocol and sulbactam–durlobactam illustrate these challenges. Cefiderocol was initially approved for complicated urinary tract infections and subsequently for hospital-acquired (HAP) and ventilator-associated bacterial pneumonia (VAP). Its greatest clinical relevance lies in its activity against CR Gram-negative organisms. Although the pathogen-focused CREDIBLE-CR study supported its use in patients with limited treatment options, the small sample size and numerically higher mortality among patients with Acinetobacter infections generated uncertainty regarding optimal clinical positioning.⁶ The recently published open-label GAME CHANGER trial did not confirm this signal, but important questions remain regarding first-line use, combination therapy, and comparative effectiveness.⁷ Despite the limited dataset, divergent and strongly held views regarding its role in therapy emerged rapidly, reflecting the vacuum created by the absence of coordinated comparative evidence.

Similarly, sulbactam–durlobactam was approved on the basis of a single pivotal non-inferiority trial with a limited number of patients infected by CR Acinetobacter.⁸ Although the study met its prespecified non-inferiority margin and demonstrated a favourable renal safety profile compared with colistin, regulators explicitly acknowledged the limited sample size and residual uncertainty because of the serious unmet medical need.

Both approvals addressed serious unmet need. Both also left important questions unanswered regarding comparative positioning, performance in clinically relevant subgroups, and generalizability to broader critically ill populations. Nevertheless, clinicians must extrapolate beyond the available data when managing complex cases.

Instead of organizing research around individual products, platforms should be structured around prioritized clinical syndromes and resistance phenotypes, such as pneumonia or bloodstream infection due to CR Gram-negative bacteria.

Post-approval: A critical phase for learning

Conventional post-authorisation studies are unlikely to resolve these uncertainties efficiently. They are often slow to initiate, underpowered for rare pathogens and designed primarily to fulfil specific regulatory obligations rather than to answer clinically relevant and especially comparative effectiveness questions. Observational case series and local registries frequently lack standardized comparators and are highly vulnerable to confounders and unsuspected biases.⁹ They may be published years after widespread clinical uptake. As a result, post-approval knowledge accumulates slowly, inconsistently and is frequently supported by evidence of limited quality and strength. The critical early phase after approval—when uncertainty is greatest and use is expanding—is rarely leveraged for structured, prospective learning.

A more coherent strategy would reframe post-approval research as a coordinated, pathogen- or indication-focused learning strategy embedded in routine care.¹⁰ Instead of organizing research around individual products, platforms should be structured around prioritized clinical syndromes and resistance phenotypes, such as pneumonia or bloodstream infection due to CR Gram-negative bacteria. Within a standing master protocol using an adaptive design, multiple approved agents could be evaluated concurrently. Treatment arms could be added or discontinued as standards evolve, facilitating head-to-head comparisons and efficient allocation of scarce patients. Bayesian adaptive methodologies are particularly suited to small populations, enabling iterative updating of treatment probabilities while preserving statistical rigor. Such platforms require durable infrastructure and multinational collaboration. Importantly, several prominent initiatives already provide structural foundations for this model.

Post-approval networks are operational and sustainable

The European Clinical Research Alliance on Infectious Diseases (Ecraid) provides a pan-European infrastructure for high-quality investigator-initiated infectious disease trials, facilitating rapid study initiation and cross-border enrolment through harmonized governance and trial support.¹¹

ADVANCE-ID (Advancing Clinical Evidence in Infectious Diseases) complements this model in Asia and collaborations in Europe with Ecraid by strengthening clinical research capacity and enabling multinational studies across regions with diverse resistance epidemiology. The ongoing TREAT-GNB platform trial demonstrates its potential for coordinated evidence generation.¹²

In the Americas, the Antibacterial Resistance Leadership Group (ARLG) provides established infrastructure for multicentre antibacterial clinical research and has demonstrated the feasibility of innovative trial designs, including platform approaches.¹³

In parallel, public post-approval resistance surveillance networks¹⁴ have documented the rapid emergence and detection of resistance to newly introduced antibiotics. These findings highlight the limitations of company-initiated surveillance programs conducted primarily to satisfy regulatory requirements.¹⁵ Such programs frequently fail to capture real-world prescribing patterns and are insufficient to guide clinical practice.

In the era of antimicrobial resistance, responsible stewardship must extend beyond prudent prescribing to encompass a collective commitment to structured, collaborative learning.

Next steps: A global model for strategic alignment?

Collectively, existing research infrastructures demonstrate that multinational, academically led networks in infectious diseases are operational and sustainable. The next step is strategic alignment around standing, adaptive, post-approval platforms focused specifically on resistant pathogens. To support post-approval clinical evidence generation, academic leaders, sponsors and regulators should converge on shared master protocols within these networks on a global basis. Governments and philanthropies should prioritize maintenance of such platforms over fragmented, underpowered single-centre studies. Public–private collaboration remains important, but scientific governance and data transparency must remain under independent oversight to preserve credibility.

Implementation of a global model requires cultural change. Clinicians and antimicrobial stewardship teams must prioritize participation in coordinated research platforms over anecdotal or unstructured use of new agents. Professional societies can facilitate consensus on priority research questions, endorse participation in adaptive platforms as best practice, and integrate emerging findings into dynamic clinical guidelines. National funders should incentivize the use of harmonized protocols within coordinated research networks instead of supporting isolated, underpowered clinical studies. International coordination is indispensable for rare pathogens, as no single country can enrol sufficient numbers of patients to generate robust comparative data.

Funding remains a critical barrier to clinical evidence generation, particularly for small companies developing novel antibiotics. The costs of evidence generation should be shared across industry, public funders, and healthcare systems. Such an approach will require substantial international coordination and collaboration among stakeholders.

Cefiderocol and sulbactam–durlobactam illustrate that expedited approval pathways can deliver urgently needed therapies to patients with few alternatives but inevitably leave important evidence gaps. Coordinated, adaptive post-approval research embedded within multinational clinical networks offers a pragmatic mechanism to reconcile rapid access with robust evidence generation. In the era of antimicrobial resistance, responsible stewardship must extend beyond prudent prescribing to encompass a collective commitment to structured, collaborative learning.

References

1. Carmeli Y, Cisneros JM, Paul M, Daikos GL, Wang M, Torre-Cisneros J, et al. (2025) Aztreonam-avibactam versus meropenem for the treatment of serious infections caused by Gram-negative bacteria (REVISIT): a descriptive, multinational, open-label, phase 3, randomised trial. Lancet Infect Dis. 25:218-230.
2. Wunderink RG, Giamarellos-Bourboulis EJ, Rahav G, Mathers AJ, Bassetti M, Vazquez J, et al. (2018) Effect and Safety of Meropenem-Vaborbactam versus Best-Available Therapy in Patients with Carbapenem-Resistant Enterobacteriaceae Infections: The TANGO II Randomized Clinical Trial. Infect Dis Ther. 7:439-455.
3. Mitra-Majumdar M, Powers JH, Brown BL, Kesselheim AS. (2022) Evidence at time of regulatory approval and cost of new antibiotics in 2016-19: cohort study of FDA approved drugs. BMJ Medicine. 1:e000227.
4. Yahav D, Tau N, Shepshelovich D. (2021) Assessment of Data Supporting the Efficacy of New Antibiotics for Treating Infections Caused by Multidrug-resistant Bacteria. Clin Infect Dis. 72:1968-1974.
5. Selig, D, Aminu, F, Cammarata, S, Chen, T, Dolak, L, Duprez, S, et al. (2025) Landscape of Post-Marketing Requirements Under the Pediatric Research Equity Act for Antibiotics from 2009–2024. Antibiotics; 14, 583.
6. Bassetti M, Echols R, Matsunaga Y, Ariyasu M, Doi Y, Ferrer R, et al. (2021) Efficacy and safety of cefiderocol or best available therapy for the treatment of serious infections caused by carbapenem-resistant Gram-negative bacteria (CREDIBLE-CR): a randomised, open-label, multicentre, pathogen-focused, descriptive, phase 3 trial. Lancet Infect Dis. 21:226-240.
7. Paterson DL, Sulaiman H, Liu PY, Chatfield MD, Yilmaz M, Salmuna ZN, et al. (2026) Cefiderocol versus standard therapy for hospital-acquired and health-care-associated Gram-negative bacterial bloodstream infection (the GAME CHANGER trial): an open-label, parallel-group, randomised trial. Lancet Infect Dis. 2026; 26:148-159.
8. Kaye KS, Shorr AF, Wunderink RG, Du B, Poirier GE, Rana K, Miller A, et al. (2023) Efficacy and safety of sulbactam-durlobactam versus colistin for the treatment of patients with serious infections caused by Acinetobacter baumannii-calcoaceticus complex: a multicentre, randomised, active-controlled, phase 3, non-inferiority clinical trial (ATTACK). Lancet Infect Dis. 23:1072-1084.
9. Hess DR. (2023) Observational Studies. Respir Care. 68:1585-1597.
10. Mahar RK, McGlothlin A, Dymock M. Barina B, Bonten M, Bowen A, et al. (2025) Statistical documentation for multi-disease, multi-domain platform trials: our experience with the Staphylococcus aureus Network Adaptive Platform trial. Trials. 26, 49.
11. Hassoun-Kheir N, van Werkhoven CH, Dunning J, Jaenisch T, van Beek J, Bielicki J et al. (2022) Perpetual observational studies: new strategies to support efficient implementation of observational studies and randomized trials in infectious diseases. Clin Microbiol Infect. 28:1528-1532.
12. Mo Yin, National University of Singapore, et al. (2025) Optimising TREATment for Severe Gram-Negative Bacterial Infections (TREAT-GNB). Accessed 21/07/2026
13. Chambers HF, Cross HR, Souli M, Evans SR, Patel R, Fowler VG, et al. (2023) The Antibacterial Resistance Leadership Group: Scientific Advancements and Future Directions, Clin. Infect. Dis. 77, Suppl. 4:S279–S287.
14. Garlasco J, Arieti F, Morra M, Tebon M, Ortiz D, Pezzani MD, et al. (2025) The Emerging Resistance Index: tracking early resistance to new antibiotics. The Lancet Infect. Dis. 26:e219-e227.
15. Sader HS, Rhomberg P, Fuhrmeister AS, Mendes RE, Flamm RK, Jones R. (2019) Antimicrobial Resistance Surveillance and New Drug Development. Open Forum Infect Dis. 2019; 6. S5-S13.