Independent educational summary. This page summarises “Points to Consider in the Clinical Development of In Vivo Diagnostics (Early Consideration)” in plain language. It was created by Clinical Biostats by editing and summarising the document; it is not an official PMDA publication and is not legal or regulatory advice. Consult the official document. PMDA provides the English version for reference purposes, and the Japanese original prevails if the two differ.
In Plain Language
In vivo diagnostics are products administered to a person to help diagnose a disease or assess a related condition, and the document addresses how their clinical value can be evaluated. It says developers should show both that the examination produces accurate information for its intended purpose and that the information matters for patient care. The document discusses development strategies, including when overseas trial results might contribute to evidence for use in Japan, and considerations for planning confirmatory trials. It describes choices about comparison methods, accuracy measures, image review and how to explain the benefit of the diagnostic examination. The central reason is that the right evidence depends on what the product is meant to do and how its results affect clinical decisions.
What This Guidance Covers
The document covers clinical development of in vivo diagnostics, including diagnostic radiopharmaceuticals. It defines these as medicinal products administered directly to the human body—for example, through blood vessels, the gastrointestinal tract, the respiratory tract or subcutaneous tissue—to diagnose or support diagnosis of diseases. It identifies three broad categories: radiopharmaceuticals used in nuclear medicine examinations such as PET and SPECT, contrast agents, and functional imaging/diagnostic agents.
Its focus is on issues often discussed when planning clinical trials: overall development strategy and the clinical data package, confirmatory-trial design and endpoints, image evaluation, and explaining clinical significance. The document says it supplements the “Guideline for Clinical Evaluation of Diagnostic Radiopharmaceuticals” by focusing on these issues. It does not state a specific product category or clinical situation that is excluded from its considerations.
Why It Matters
As the document explains, no guidelines had been issued that specifically addressed methods for the clinical evaluation of in vivo diagnostics. Development strategies were therefore considered case by case through individual consultations. The document also notes that global development has advanced, including through multi-regional clinical trials (MRCTs) and use of foreign clinical trial data, raising development questions not originally envisaged in the cited diagnostic radiopharmaceuticals guideline.
The document responds by setting out considerations that can help structure those discussions. It emphasizes that showing diagnostic accuracy alone is not the whole evaluation: developers should also explain the clinical significance of the information obtained. Its considerations reflect currently available knowledge and may change as new findings emerge.
Key Recommendations
Assess both accuracy and clinical significance. The clinical utility of an in vivo diagnostic should be demonstrated from two perspectives: the accuracy of the information obtained in relation to the examination’s intended purpose, and the clinical significance of that information. The document addresses accuracy through confirmatory-trial considerations and clinical significance through the policy for explaining patient benefit.
Consider global development early when appropriate. To support timely availability of effective and safe in vivo diagnostics in Japan, participation in MRCTs should be considered at an early development stage if deemed appropriate. When planning MRCT participation, developers should consider ethnic factors that may affect diagnostic performance, based on the “Basic Principles on Global Clinical Trials” and related notifications.
Consider how foreign confirmatory-trial evidence could be used. If development has progressed ahead overseas and foreign confirmatory-trial results are already available, the document considers it possible to conduct exploratory domestic trials with designs similar to the foreign confirmatory trials. These trials can confirm whether comparable diagnostic performance and safety can be expected in Japanese subjects as in non-Japanese subjects. If this is confirmed, a clinical data package may be constructed with the foreign confirmatory-trial results as the primary evidence.
For certain products, consider a strategy relying primarily on foreign confirmatory data. This may be considered for diagnostics whose principle involves specific accumulation in target sites, where no other extrinsic or intrinsic ethnic factors affecting diagnostic performance are anticipated, and pharmacokinetic similarity between Japanese and non-Japanese subjects has been confirmed—for example, in Phase I studies. The document says this strategy may apply to imaging-based diagnostics such as nuclear medicine radiopharmaceuticals and contrast agents. Even in these cases, assessing diagnostic performance in Japanese subjects in Phase I or exploratory trials is considered useful for explaining performance in the Japanese population.
Plan for interpretation of images in Japanese practice. Because clinical experience in Japanese medical practice during development is expected to be limited, it is recommended that images from foreign confirmatory trials be evaluated by image readers within a Japanese clinical setting. Based on those evaluations, training materials for image interpretation and assessment should be developed. The document recommends consulting PMDA about development plans that use foreign confirmatory-trial data.
Define the intended clinical use and choose a suitable confirmatory design. The intended clinical use needs to be clearly defined—for example, assessing a patient’s condition, such as disease presence or absence or disease stage, or identifying tissue or lesions for surgery or biopsy. Developers should consider the product’s clinical positioning relative to existing diagnostic methods. If established methods provide comparable information, a comparative study is recommended. Within-subject and parallel-group designs may both be considered, with the choice determined by product characteristics. If a comparative study is not feasible, a single-arm study evaluating performance against a predefined threshold may be acceptable; the threshold’s validity should then be adequately justified using prior product trial results, trial results for established diagnostic methods and relevant literature.
Choose endpoints that show performance against a suitable reference. The document recommends diagnostic performance based on the Standard of Truth (SOT) as the primary endpoint in confirmatory trials. It generally recommends both sensitivity and specificity as primary endpoints, while recognizing that intended use can justify sensitivity alone as the primary endpoint, with specificity assessed secondarily. In other situations, clinical utility may be shown with primary endpoints other than sensitivity and specificity. If both sensitivity and specificity are not primary endpoints, the appropriateness of the development plan should be discussed with PMDA.
Match the unit of endpoint analysis to the clinical decision. For products detecting multiple lesions or sites in one patient, define the primary endpoint at the patient level or lesion/site level according to which better informs clinical decisions. When the objective is to diagnose the patient’s overall condition, patient-level evaluation is generally appropriate as the primary endpoint and lesion/site-level evaluation as a secondary endpoint. When the objective is to localize target tissue or lesions accurately, lesion/site-level evaluation is generally appropriate as the primary endpoint.
Support objective and reproducible image interpretation. It is appropriate to appoint at least two independent evaluators who are not affiliated with investigational sites; three or more is preferable. If assessments differ, the final judgment should be determined by an objective method, such as majority voting. Evaluation manuals and pre-study training are recommended to reduce variation among evaluators.
Explain the clinical benefit in terms of people undergoing the examination. Examples include enabling more appropriate treatment selection and supporting safe, appropriate surgery through accurate diagnosis and precise identification of tissue or lesions. How to demonstrate benefit depends on current medical practice, including the availability of existing diagnostic methods and established treatment options. Where clinical guidelines or literature establish sufficient medical consensus that use of the diagnostic enables optimal treatment selection and facilitates the safe and appropriate performance of surgical procedures, additional trials to demonstrate the investigational product’s clinical benefits may not be necessary. If the clinical significance of interventions based on the diagnostic result is not established in the current treatment algorithm, the utility of those interventions—for example, improvement in prognosis compared with existing treatments—should be evaluated in clinical trials. Those results are then needed to demonstrate the clinical significance of performing the examination.
Practical Implications
Interpretation: A development plan can begin by writing down the intended clinical use in terms of the decision the examination is meant to inform. That purpose helps determine which established diagnostic method, if any, is an appropriate comparator; whether the trial should compare methods; and whether an endpoint should summarize performance for each patient or for each lesion or site. It also helps explain why a particular accuracy measure is important for the intended decision.
Interpretation: Sponsors considering overseas evidence should distinguish between showing that foreign results may be relevant and confirming that performance and safety can be expected to be comparable in Japanese subjects. The document describes exploratory domestic trials as one possible way to make that confirmation when foreign confirmatory results already exist. It also identifies pharmacokinetic similarity as a condition for considering a strategy based primarily on foreign confirmatory-trial data in the particular setting it describes.
Interpretation: Trial teams may need to coordinate endpoint definitions, reference methods and image-review procedures before enrollment or assessment begins. Independent evaluators, written manuals and pre-study training address the document’s concern with objectivity, reproducibility and variability in image interpretation. If multiple lesions are assessed within a person, a prespecified patient-level or lesion/site-level primary endpoint can keep the analysis aligned with the clinical objective.
Interpretation: The evidence for clinical significance should be considered separately from evidence of accuracy. A product might provide accurate information, but the development rationale still needs to explain how that information contributes to care. Whether further clinical trials are needed to show benefit depends, as the document states, on the medical-practice context and on whether the clinical significance of decisions prompted by the diagnostic result is already established.
Technical Details
Diagnostic performance measures. Sensitivity and specificity are measures used to evaluate diagnostic performance. The document generally recommends setting both as primary endpoints to demonstrate that accurate information can be obtained. Depending on intended use, sensitivity alone may instead be the primary endpoint and specificity a secondary endpoint. Its example is a setting in which complete lesion detection is clinically important even if some false positives occur, such as support for biopsy or screening when existing diagnostic methods provide the final diagnosis.
Threshold-based single-arm studies. When a comparative study is not feasible, a single-arm study against a predefined threshold may be acceptable. The threshold should not be treated as self-justifying: its validity should be adequately justified by considering prior clinical trial results for the product, clinical trial results for established diagnostic methods and relevant literature.
Unit of evaluation. A product that detects multiple sites in a patient raises a choice between patient-based and lesion/site-based evaluation. The document ties this choice to the intended diagnostic objective and to what would inform physicians’ decisions. Overall diagnosis of a patient’s condition generally points to a patient-based primary endpoint; accurate localization generally points to a lesion/site-based primary endpoint.
Ethnic factors. The document’s footnote distinguishes intrinsic factors, including physiological functions such as circulatory dynamics and renal function, distribution and expression levels of target molecules, height and weight, from extrinsic factors. The latter include diagnostic criteria and algorithms, pathological classification systems, equipment and software settings, image-reading and interpretation procedures, training and reader or assessor proficiency, and surgical techniques and proficiency. These are factors to consider when they may affect diagnostic performance.
Image assessment. The recommendation is for at least two independent evaluators who are not affiliated with investigational sites, with three or more preferable. For discrepant assessments, an objective method such as majority voting can determine the final judgment. Manuals and pre-study training are recommended to reduce variation; the document does not specify a particular statistical model or numerical performance threshold for these procedures.
Who Should Read It
This document is relevant to sponsors and development teams planning clinical studies of in vivo diagnostics, including diagnostic radiopharmaceuticals, contrast agents and functional imaging/diagnostic agents. It will also be useful to statisticians and clinical research professionals defining study designs, reference methods, endpoints and image-review processes; regulatory teams considering the role of foreign clinical data or planning consultation with PMDA; and investigators and image evaluators involved in clinical assessment. It is particularly pertinent when intended use, trial geography, reference standard, or the connection between diagnostic results and patient benefit is being planned.
Official PMDA Source
- Official title: Points to Consider in the Clinical Development of In Vivo Diagnostics (Early Consideration)
- Issuing body: Pharmaceuticals and Medical Devices Agency; Office of New Drug II
- Date on the cover: March 12, 2026
- PMDA PDF: Official PMDA PDF
- PMDA list page: 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, and the Japanese original prevails if there is any inconsistency with the English translation.
Related Clinical Biostats Resources
- All Clinical Biostats tutorials — Browse other educational summaries and tutorials, including those on clinical trial design and diagnostic development.
- Statistical Calculators — Access calculators that may support statistical work in clinical research.
Key Takeaways
- Clinical utility involves both accurate information for the intended examination and the clinical significance of that information.
- Study design, reference method and endpoint should fit the diagnostic’s intended use and the clinical decision it supports.
- Foreign confirmatory-trial results may contribute to a Japanese data package in the circumstances described, with attention to relevant ethnic factors and, where appropriate, domestic evaluation.
- Image review should be organized to support independent, objective and reproducible assessments.
- Whether clinical-benefit trials are needed depends in part on whether the significance of the diagnostic-informed treatment or procedure is already established in medical practice.