Diagnostic Radiopharmaceutical Trial Design: Is It Time to Change Nomenclature?

“What’s in a name? That which we call a rose by any other name would smell as sweet…”

— William Shakespeare

The field of nuclear medicine is going through a renaissance with significant investment by a biotechnology and pharmaceutical industry finally recognizing the benefits of the theranostic paradigm. Since the early days of nuclear medicine, the strength of molecular imaging has resided in the availability of a wide range of biologically relevant radiopharmaceuticals that can trace physiologic and biologic processes. One of the early pioneers in the field, Henry Wagner, was renowned for testing new tracers on himself. In his era, the passage of radiopharmaceuticals from successful proof-of-concept imaging to routine practice was rapid. It can be guaranteed that none of these early radiopharmaceuticals was manufactured under good manufacturing practice guidelines, and safety testing was rudimentary. Nevertheless, nuclear medicine grew and has been a safe and effective diagnostic modality with tracers available to evaluate a plethora of diseases.

Having been privileged to have met pioneers in nuclear medicine such as Henry Wagner and David Kuhl, who were involved in introducing new radiotracers that have benefited millions of patients, and having started my research career in an era when ethics submissions were considered wordy if they ran to more than a few pages, I lament the end of those times. Today, we face an ever more complex ethical review and regulatory process to clinically transition novel diagnostic radiopharmaceuticals through onerous and expensive regulatory approval and reimbursement pathways. Often, this process takes years or fails completely at a huge cost to society.

Why has research, in general, and radiopharmaceutical research, in particular, become so difficult? Regulators would contend that their governing principles are designed to protect the public from previously dangerous practices. For radiopharmaceutical research, this rhetoric is contradicted by the unparalleled safety profile of diagnostic tracer administrations to humans (1). In fact, the clear and present danger to patients is lack of access to the best diagnostic tests to characterize their disease and the costs of those investigations if they get to market through the need for the pharmaceutical industry to achieve a return on investment during the patent life of their product.

It has been recognized that the words we use influence the ways we think. Links between philology and epistemology that are culture-specific are the subject of considerable research (2). I believe that it is relevant to consider whether the nomenclature used to describe the clinical trial phases through which diagnostic radiopharmaceuticals pass has been an impediment to critical thinking about relevant design of such trials and, thereby, compromised the field. More specifically, is it sensible to apply the terminology relevant to drug development to radiopharmaceutical sciences? I contend that it isn’t.

The design of trials of new drugs is formulaic. Phase I trials test the safety of agents through dose escalation. Phase II trials test the efficacy of the agent and expand safety data in a selected population with the relevant disease. Phase III trials compare the efficacy of the new drug with existing standards, usually with the goal of replacing one with the other through demonstration of cost-effective benefit.

In phase I drug trials, dose escalation seeks to determine a maximum tolerated dose. Accordingly, eligibility criteria are generally very narrow, favoring participants with the lowest probability of experiencing adverse events caused by preexisting conditions rather than by the investigational agent itself. These participants are often rather atypical by virtue of having well-preserved organ function and a good performance status. For diagnostic radiopharmaceuticals, escalating the mass of the chemical entity or the activity administered to anywhere near the point of toxicity is never a relevant scientific objective. In fact, the goals are completely the opposite. Diagnostically, achieving high-quality images at the optimal imaging time using the lowest possible mass and radiation dose to the subject is desired. These objectives can now be achieved in the earliest stages of radiopharmaceutical development at extremely low radiation doses using microdosing quantities of investigational radiopharmaceuticals through the exquisite sensitivity of long–axial-field-of-view PET/CT (3).

Because diagnostic radiotracers are likely to be used across a broad spectrum of diseases and in the presence of multiple comorbidities, understanding, for example, the impact of renal or hepatic impairment, which are critical organs for handling of radiopharmaceuticals, on biodistribution, imaging characteristics, and radiation dosimetry, is critical. Accordingly, exclusion of such patients from early clinical evaluation of radiopharmaceuticals is nonsensical. Rather, how the imaging protocol should be adapted for such patients becomes important for the design of subsequent efficacy trials. For example, for radiopharmaceuticals with renal excretion, delayed imaging protocols may be appropriate. Alternatively, pharmacologic intervention (4), such as diuretic administration and fluid loading, may improve scan quality. The unique ability of nuclear medicine imaging to characterize the biodistribution of tracers over time allows an understanding of body regions within which uptake may mask detection of the disease process that the tracer seeks to detect. Uptake of [¹⁸F]FDG in the brain, which limits detection of brain metastases, and the hepatic metabolism of [¹⁸F]fluorothymidine, which restricts detection of liver metastases, are but 2 examples. The latter example also emphasizes the importance of testing novel agents in humans and not relying on animal data. Preclinical evaluation of this agent suggested low hepatic uptake, whereas humans have intense uptake in this organ (5). Because the tissue accumulation of radiopharmaceuticals reflects dynamic processes influenced by multiple factors, including perfusion, target affinity, washout rates, and metabolism, the assessment of novel radiopharmaceuticals should identify the timing of optimal contrast between pathologic tissues and normal organs. For agents with rapid tumor uptake and blood clearance, such as the small-molecule fibroblast activation protein inhibitor tracers (6), imaging within an hour of administration is desirable and has workflow advantages in busy PET departments. Conversely, for monoclonal antibodies, which are often species-specific, slow blood clearance and delayed accumulation in pathologic tissues necessitate imaging days after tracer administration as well as influencing the choice of radionuclide. Typically, ⁸⁹Zr is required to provide adequate statistical quality in the images obtained. Such information can also drive innovations in technology that enable late time point imaging, such as total-body PET/CT (7,8). Above all, preliminary studies need to demonstrate a favorable biodistribution for clinical evaluation of the disease process for which the tracer is being developed.

Although the aims, primary endpoints and trial eligibility criteria, for phase I drug development and first-in-human radiopharmaceutical trials are fundamentally different, radiopharmaceutical trial designs frequently mirror drug development with a focus on safety. For example, we have frequently seen prolonged QT intervals on an electrocardiogram being an exclusion criterion. This measure of ventricular repolarization may be relevant to drug trials but makes no sense for radiopharmaceuticals administered at microdosing levels, particularly when there is no evidence of specific or nonspecific cardiac uptake on preclinical evaluation. If members of risk-averse ethics committees propagate unreasonable constraints on rational radiopharmaceutical development by uncritically applying the familiar framework for the design of phase I trials of conventional drugs, might the solution be to remove the intellectual straitjacket of irrelevant nomenclature? Accordingly, I suggest renaming this first phase of diagnostic radiopharmaceutical development as a phase A (acquisition optimization) trial.

The primary aim of the next phase of conventional drug development is to assess efficacy, with the secondary aim being to expand safety data in a clinically relevant population. Study eligibility criteria are still generally narrow with respect to performance status and physiologic reserves. Primary endpoints for such trials are objective response rate and, more particularly for cancer, survival measures. For radiopharmaceutical development, the scientific objective is to determine whether the novel agent can detect or characterize disease. Although diagnostic accuracy is often focused on sensitivity and specificity, with trial design often preselecting patients with known disease to assess the former and individuals without disease or a low likelihood of disease to test the latter, in clinical practice, the most useful information is the negative predictive value in a patient with a high pretest probability of disease and the positive predictive value in a patient with a low pretest probability of disease. Definition of diagnostic performance in patient cohorts with varying pretest probabilities of disease is the essential next step in radiopharmaceutical development. This can be achieved through application of Bayesian principles with validation through a composite standard of truth that includes pathologic verification, when available, correlative imaging, response to image-guided therapeutic intervention, or clinical follow-up for an appropriate period. At this phase of development, it is useful to define the target population for further detailed evaluation. In oncology, this might include specific cancers or patients with specific imaging findings on conventional imaging. For example, this might include patients with lesions of low uptake of the standard-of-care tracer. Applying the same nomenclature to studies of conventional drugs and radiopharmaceuticals given such fundamentally different study objectives and procedural requirements invites confusion and inefficiency that benefits neither trial participants nor scientific advancement. Accordingly, I suggest renaming this phase of radiopharmaceutical development as a phase B (biologic relevance) trial.

When an investigational drug that has no or tolerable side effects along with signals of efficacy, phase III trials may be undertaken to compare the new drug with the current standard-of-care with the objective of achieving regulatory approval by demonstrating a superior and cost-effective therapeutic effect. In oncology, this generally requires that survival is increased. To maximize pharmaceutical company profit, the study population is generally designed with a view to achieving the broadest possible label. With well-established and widely available imaging paradigms, novel radiopharmaceuticals ideally seek to address an unmet need with respect to conventional imaging. Accordingly, demonstration of incremental diagnostic information is the objective when added to existing paradigms or, when the current paradigm has poor diagnostic utility, whether replacement can be achieved. The ProPSMA trial was such an example, addressing the suboptimal sensitivity of CT and bone scans for the detection of metastases in patients with prostate cancer (9). The primary goal of such studies is to establish whether the new radiopharmaceutical increases the rate of valid management decisions. In the early years of assessing the clinical impact of FDG PET in various cancers, we compared the management plan based on the available standard-of-care information before performance of FDG PET with the management plan after PET to impute the effect of incremental diagnostic accuracy on patient-important outcomes. We demonstrated high impact in a wide range of cancers, including lung cancer (10). Similar trial designs were recapitulated in the Australian Data Collection Project and the National Oncologic PET Registry in the United States (11) and have led to both routine use and reimbursement of FDG PET in oncology. As the dominant impact of FDG PET was to prevent futile surgical or radical radiotherapy treatment by detection of disease outside the proposed locoregional treatment field, the potential for the incremental diagnostic information to impact survival is limited. However, such standard endpoints for therapeutic intent drug trials are not reliable indicators of the patient-relevant benefits of enhanced diagnostic test performance because those metrics are almost entirely dependent on the efficacy of the treatment applied. Until there were effective treatments for metastatic melanoma, more accurate detection of metastatic disease was unlikely to improve patient outcomes. In the current era, early detection of metastatic disease improves the likelihood of response with multiple studies, including one from my former department (12), indicating that tumor volume is an important prognostic factor in the immune checkpoint therapy era. Similarly, in an article detailing Henry Wagner’s last Society of Nuclear Medicine Meeting highlights lecture, he related a similar story wherein rectilinear brain scanning resulted in a more accurate diagnosis and earlier surgical intervention but did not improve outcomes because surgery was ineffective (13). Presciently, he suggested rethinking how the patient benefits of molecular imaging studies are analyzed to ensure that all patient-important outcomes and not just survival metrics are incorporated into the ultimate value determination. In summary, given the different objective of novel therapeutic drugs and diagnostic radiopharmaceuticals, persisting with identical clinical trial designs and nomenclature represents an example of cognitive dissonance. To mitigate this anomalous and counterproductive status quo, I propose that preregistration radiopharmaceutical trials are renamed as phase C (clinical applicability) trials. A matrix of the differences in the trial framework is outlined in Table 1.

• © 2025 by the Society of Nuclear Medicine and Molecular Imaging.

• Received for publication January 8, 2025.
• Accepted for publication March 12, 2025.