Points to Consider for Clinical Development of Drugs Intended for Treatment of Antimicrobial-resistant Gram-negative Bacterial Infections — PMDA Early Consideration Summary

In Plain Language

Antimicrobial-resistant Gram-negative bacterial infections can be difficult to treat because the bacteria are resistant to existing drugs and treatment options may be limited. This document sets out PMDA’s considerations for clinical development of drugs intended to treat these infections, including how to study them when there may be too few patients for a large, conventional trial. It discusses different approaches for drugs that directly act against the bacteria and drugs that do not have much direct antimicrobial activity but are used with other antimicrobial drugs. For clinical studies in patients with resistant infections, the document emphasizes randomized comparisons, clear evaluation within each infection area, and careful planning of the number and types of patients. These considerations aim to help developers assess efficacy and safety while recognizing the practical challenges of studying uncommon or locally different resistance patterns.

What This Guidance Covers

The document concerns the clinical development of drugs intended to treat antimicrobial-resistant gram-negative bacterial infections, abbreviated in the document as antimicrobial-resistant GNBI. It discusses development strategy, proposed indications at the time of an application for approval, and the design and analysis of clinical studies in patients with these infections. Its focus is on clinical evidence for efficacy and safety, particularly when patient numbers are limited or resistance patterns differ between Japan and other regions.

The document distinguishes between a new active pharmaceutical ingredient (API) expected to have antimicrobial activity against the target bacteria and a new API that has little or no direct antimicrobial activity and is used with approved antimicrobial drugs. It also addresses studies that include patients suspected of having antimicrobial-resistant GNBI, as well as the possibility of enrolling patients with infections from different representative disease areas in one study.

The document notes that pediatric drug development is important, but directs readers to other named materials for that topic. It does not set out a detailed pediatric development strategy in this document. It also does not provide a single universal trial design or sample-size target; instead, it describes considerations for selecting a design and justifying it in light of the development program and available patients.

Why It Matters

PMDA describes antimicrobial resistance as a major global public-health issue. Infections caused by resistant Gram-negative bacteria may have limited treatment options, creating a need to develop new effective antimicrobial drugs. The document identifies several known resistance mechanisms: production of β-lactamase, a decrease or deletion of porins, and/or overexpression of efflux pumps.

A central development challenge is that the resistant pathogen or resistance mechanism may be extremely rare. In addition, the types of resistance—for example, types of β-lactamase—may differ between Japan and other countries. These circumstances can make it difficult to plan and conduct a study with enough participants and a design capable of clearly establishing efficacy through a high-powered randomized comparison for non-inferiority or superiority.

PMDA also points to limits in the available comparable evidence. Studies may enroll patients in regions experiencing an epidemic at the time of the study, and a single study may include several disease areas. Comparable external data from patients with similar disease areas, causative bacteria, severity, and concomitant treatments may therefore be extremely limited. The document explains why an uncontrolled study could be difficult to interpret in this setting and why planning the comparison, disease-area assessments, and sample size carefully is important.

Key Recommendations

1. Plan clinical development around evidence in patients with resistant infections. Development should proceed on the premise that a clinical study in patients with antimicrobial-resistant GNBI will be conducted, following the considerations in the document, to confirm a favorable benefit-risk balance. If patient numbers are limited both in Japan and overseas, and resistance mechanisms or types differ between Japan and overseas, a multi-regional study including Japan should be planned to establish solid evidence about the investigational drug.

2. For a new API with expected direct antimicrobial activity, consider supplementary study in susceptible infections. In principle, studies in patients with antimicrobial-resistant GNBI should confirm efficacy in representative disease areas—body sites, such as organs or tissues, from which causative bacteria are frequently isolated—and evaluate safety using data from a certain number of patients. If it is difficult to conduct those resistant-infection studies with a sufficient sample size and design, developers should also examine efficacy in an exploratory manner in patients with antimicrobial-susceptible GNBI in one typical disease area relevant to the indications being developed. The document says these susceptible-infection studies should preferably take place before studies in patients with resistant infections. It explains that patients with resistant infections tend to have more severe disease manifestations, in part because their clinical management pathways are complicated by limited treatment options.

3. For a new API without much direct antimicrobial activity, consider whether susceptible-infection studies are needed. This situation includes, for example, a new β-lactamase inhibitor developed for use alongside approved antimicrobial drugs, either as a separately administered drug or in a co-formulated product. Such cases are usually considered less likely to need the susceptible-infection study described above. However, results from studies in susceptible infections may be useful if safety data for the investigational drug cannot be obtained in sufficient quantity from studies in resistant infections. If the dose of an approved antimicrobial drug used with the API exceeds the maximum dose approved in Japan, the safety acceptability of that dose level needs to be justified.

4. Base proposed target microorganisms and diseases on the clinical evidence. Microorganisms identified as causative bacteria in patients with resistant infections in clinical studies are, essentially, candidates for inclusion as target microorganisms in the proposed indication. In principle, the target diseases should be disease areas for which reasonable efficacy and safety have been confirmed from clinical-study results. The document describes a possible exception: a drug that would be regarded as a last resort in the treatment algorithm at development, such as a drug for carbapenem-resistant GNBI, may have an indication proposing various infections without limiting it to specific disease areas if it is expected to show a certain level of efficacy outside the studied areas. That expectation should be supported by pharmacokinetic characteristics, including tissue distribution, and clinical-study results in representative disease areas.

5. When combining disease areas in a study, define and assess them separately. If a sufficiently sized and designed study for each disease area is difficult because patient numbers are limited, it may be acceptable to enroll patients with infections from different representative disease areas caused by target microorganisms in the same study. The document gives possible combinations for carbapenem-resistant GNBI, including complicated urinary tract infection/pyelonephritis, hospital-acquired pneumonia/ventilator-associated pneumonia, and/or complicated intra-abdominal infection. Key inclusion and exclusion criteria and efficacy endpoints should be defined individually for each disease area. Because disease areas may have different criteria and endpoints, an overall pooled result using common efficacy criteria may be difficult to interpret. If efficacy differs across areas, developers should interpret pooled results carefully and evaluate the study comprehensively using results from each disease area as well as the merged results.

6. Consider enrollment before susceptibility results when that reflects intended use. In some indications, the document anticipates that treatment may begin in clinical practice before a drug-susceptibility test result is available. For investigational drugs expected to be used this way, it is appropriate to enroll patients suspected of having resistant GNBI and start the investigational drug without waiting for the test result. The primary efficacy analysis, however, should be conducted in the population confirmed to have antimicrobial-resistant GNBI.

7. Use randomized controlled studies with an appropriate control. Given the expected scarcity of comparable external information, the document says an uncontrolled study would be highly likely to produce results that are difficult to interpret. Clinical studies in patients with antimicrobial-resistant GNBI should therefore be conducted as randomized controlled studies, with a suitable control, such as standard treatment, chosen in light of the investigational drug’s intended clinical position. The document adds that statistical testing of non-inferiority or superiority against the control on the primary endpoint may provide the most convincing information to demonstrate efficacy.

8. Set sample size to support assessment across organisms, disease areas, and safety concerns. Study sites and the total target sample size should be selected in light of epidemiological information about regions where each target microorganism is epidemic, so that efficacy against a wide variety of target species can be studied. The target sample size should also be set for each disease area to obtain sufficient data supporting expected efficacy and safety there. If potential adverse events are anticipated from the characteristics of the study drugs or other available information, enough patients should be enrolled in view of the frequency of those events to evaluate safety appropriately.

9. If a confirmatory study is infeasible, explain the limits and justify the plan. When a study with a sufficient sample size and design cannot be conducted, and the planned study is not intended to confirm non-inferiority or superiority, developers should investigate feasible sample sizes using epidemiological information and other scientific considerations, and appropriately explain why a confirmatory study is infeasible. The clinical significance, validity, and limitations of the efficacy evaluation method should be clearly presented. Even when a feasibility-based sample size is justified, developers should address the difficulty of drawing definitive conclusions about efficacy and safety from a study enrolling extremely few patients, and thoroughly examine the number of participating sites and study duration. If alternative statistical decision-making methods are used, such as a Bayesian method borrowing external data, operating characteristics such as type I error rate and power should be provided and their validity considered.

In studies of resistant infections, safety should be clarified in the patients being studied. The document also says safety acceptability should be evaluated in light of expected benefits, including comparison with drugs in the control group, such as standard treatment.

Practical Implications

Interpretation: The document points toward a development plan that links the proposed indication to the people and infections actually studied. For sponsors, this means planning early for the possibility that resistance patterns, patient numbers, and available treatment options will affect both trial design and the eventual interpretation of evidence. A multi-regional plan may be relevant when patient numbers are limited and resistance types vary by region; the document specifically says such a study including Japan should be planned in that circumstance to establish solid evidence.

For clinical operations teams, studies that combine infection areas require more than one broad protocol category. Each disease area needs its own key entry and exclusion criteria and its own efficacy endpoint, according to the document. Site selection and the planned enrollment mix also need to take account of where target microorganisms are occurring, because the study should be able to examine efficacy across a wide variety of target species.

For statisticians, a pooled result should not automatically be treated as the whole efficacy story. Disease areas may differ in endpoints, eligibility criteria, and treatment effects, so the results by area need to be considered alongside the overall merged result. If the study cannot be confirmatory, the justification should explain the feasibility constraints and the meaning and limits of the planned efficacy assessment. If an alternative method, such as a Bayesian method borrowing external data, is used, the document calls for operating characteristics and consideration of their validity.

For safety and regulatory teams, the study plan should be able to characterize safety in patients with resistant infections and relate it to expected benefits and control-group treatments. Where a combination or co-formulated product uses an approved antimicrobial drug above its maximum dose approved in Japan, the safety acceptability of that dose level needs justification. These are interpretations of how the document’s considerations can inform planning; the document also recommends PMDA consultation for developers wishing to discuss individual clinical-study plans and clinical data packages.

Technical Details

The study-design logic in the document can be read as a sequence of decisions:

The document discusses non-inferiority or superiority testing as a potentially convincing primary analysis, but does not prescribe a universal margin, a fixed sample size, or one required statistical method for every development program. Where an alternative method is considered, the examples of operating characteristics given are type I error rate and power. The document says these should be provided under the target sample size, with their validity considered.

For interpretation of efficacy across several areas, the document distinguishes the merged results from the evidence within each individual disease area. If the size of the treatment effect differs across areas, pooling responder or non-responder results can make the overall result difficult to interpret. The study’s evaluation should therefore take account of both the separate disease-area results and the merged results.

Who Should Read It

This document is relevant to sponsors and developers planning clinical programs for drugs intended to treat antimicrobial-resistant Gram-negative bacterial infections. It is also useful for clinical and statistical teams designing studies, selecting endpoints and controls, planning enrollment across sites and infection areas, or assessing whether a proposed study can support a confirmatory conclusion.

Regulatory affairs professionals, medical writers, clinical research associates, investigators, and safety teams may find it useful when developing or reviewing the rationale for an indication, a study population, an analysis plan, or a safety evaluation. Its considerations are particularly pertinent when target pathogens or resistance mechanisms are uncommon, regional resistance patterns differ, treatment may start before susceptibility results are available, or the planned sample size is constrained.

Official PMDA Source

• Official title: “Points to Consider for Clinical Development of Drugs Intended for Treatment of Antimicrobial-resistant Gram-negative Bacterial Infections (Early Consideration)”
• Issuing body: Pharmaceuticals and Medical Devices Agency, Office of New Drug IV
• Date on cover: March 24, 2025
• Official PMDA PDF: Read the PMDA document
• PMDA list page: View PMDA Early Consideration documents

Source: Pharmaceuticals and Medical Devices Agency (PMDA) website, PMDA Early Consideration documents. This page was created by editing and summarising the PMDA document.

Consult the official document; this summary does not replace it. PMDA provides the English version for reference purposes, and the Japanese original prevails if the two versions differ. PMDA describes Early Consideration documents as reference information reflecting the Agency’s views at the time of publication; those views may change as scientific knowledge develops.

Related Clinical Biostats Resources

• All Clinical Biostats tutorials — Browse other educational summaries and tutorials across clinical research topics.
• Statistical Calculators — Access statistical calculators that may assist with clinical study planning and analysis.