Quick background: oncology dose optimization

• Traditionally, early development of oncology products has focused on identifying the maximum tolerated dose, or MTD, of an investigational product. However, this “highest dose” approach ignores several factors, including target interactions and off-target toxicities, longer-term toxicities, and safety considerations outside of dose-limiting toxicities. Additionally, this approach generally lacks applicability to new treatments that are more targeted in their effects, including immunotherapies and targeted small-molecule therapies. This incompatibility spawned a movement focused on optimizing dose selection in oncology drug development.
• In 2021, the FDA’s efforts on dose optimization were formalized with the formation of the Oncology Center of Excellence’s Project Optimus. At the time, OCE Director RICHARD PAZDUR remarked that the field of oncology had long suffered from a cultural bias that “more is better” when it comes to dosing oncology products. Pazdur added that OCE is “going to start making it more of a requirement” that companies conduct dose-finding to facilitate optimization and that it is in companies’ “best interest” to follow this approach. The agency subsequently partnered with the American Society of Clinical Oncology in 2022 to host a two-day workshop on dose optimization strategies and methods. [Read AgencyIQ’s analysis of the workshop here.]
• The FDA’s work shifted from conversation to policy with the development of a guidance document on dose optimization for oncology products. The January 2023 draft version emphasized that dosage optimization efforts should occur before approval. Sponsors should generally select doses for a clinical trial based on “relevant nonclinical and clinical data” and conduct randomized, parallel dose-response trials when feasible. The draft guidance received robust feedback from stakeholders, with a mix of support of and opposition to the agency’s proposals. The FDA ultimately finalized the foundational guidance in August 2024. [Read AgencyIQ analyses here and here.] As explained by the associated Federal Register notice, the agency incorporated some stakeholder feedback, including an enhanced focus on product-specific advice and “clarity on the collection and analysis of key pharmacologic data.” The final version preserves the take-home messages of its draft, clarifying that the that the agency can place a deficient protocol on clinical hold.

Tailoring recommendations for radiopharmaceuticals

• The FDA issued a new draft guidance document in August 2025 with recommendations specific for radiopharmaceuticals therapies, or RPTs. The agency instructed readers to consider its contents alongside the broader oncology dose optimization guidance. The agency also previously finalized a guidance document in 2019 addressing two other aspects of RPT development: nonclinical studies and certain aspects of product labeling.
• The 2025 draft guidance defines RPTs in relation to other products. “RPTs, typically administered systemically to treat cancer, share characteristics with both external beam radiation therapy (EBRT) (e.g., cell killing through ionizing radiation) and systemic drug therapy (e.g., systemic biodistribution and pharmacologic targeting),” it says.
• RPTs represent an emerging precision oncology modality. Theranostic products, which target cancer cells to receive radiation for both diagnostic purposes via imaging and to treat the tumor, are an important and growing RPT application. A 2020 review in The Lancet Oncology further explains that theranostic radiopharmaceuticals may carry alpha- or beta-emitting particles directly to the cancer, “which is achieved by attaching targeting ligands (small molecules, peptides, or antibodies) to chelators that complex radioisotopes [require] for systemic delivery.” Other strategies for delivering a radioactive payload to an imaging and treatment target include directly tagging ligands with radioactive iodine, as well as direct RPT delivery via intra-arterial catheter, the authors explained. Examples of approved RPTs include Novartis’ two products Pluvicto (lutetium-177 vipivotide tetraxetan) for certain metastatic castration-resistant prostate cancers and Lutathera (Lu-177 dotatate) for certain gastroenteropancreatic neuroendocrine tumors.
• The draft guidance document describes inherent challenges in dose-finding studies for RPTs. RPT administration has historically adhered to established ERBT limits, but this strategy does not account for inherent differences in the treatments such as the dose rate, distribution and timing of doses. In addition, traditional dose-finding studies, such as identifying an MTD, are not easily applicable for RPTs because these studies rely on a finite toxicity window that does not translate to the potential long-term or cumulative effects of radiation-based therapies.
• The FDA outlined expectations for dose-finding trials, which should generally enroll populations with limited life expectancy and limit relative exposure in patients with lower morbidity and mortality risk. Trials should account for, but not categorically exclude, patients who were previously treated with other radiation-based therapies. The document notes that the informed consent for dose-finding trials of RPTs should describe the “the potential risks of long-term, late-onset, and cumulative radiation toxicity and that such information may not be well-characterized for a product during initial development stages.”
• The trial design should aim to reduce the risk of toxicity, according to the draft guidance. RPT trials should include a fixed number of administration cycles and prespecify and justify a limit that corresponds to “radiation absorbed doses to critical organs,” since there’s a difference between the radiation dose administered and that absorbed by the body. The guidance explains that this limit can be “product- and indication-specific,” and “may be revised based on emerging data.” The document says that the limit should be based on the amount of uncertainty regarding the RPT’s activity, with more flexibility afforded in cases including established products with substantial clinical data. When the trial moves into human participants, the draft advises that the starting dose should be informed by EBRT literature and generally should not exceed those limits. In cases where early data suggest limited toxicity and that a higher dose may be optimal, the guidance says sponsors should request a meeting with the agency and be prepared to discuss “detailed initial data, including safety, preliminary efficacy dosimetry, and pharmacokinetic data.”
• The ultimate selection of the radiation dose for trials intended to support a marketing application should be based on comprehensive data analyses. The data collected should align with the intended treatment duration (for example, multiple cycles), include enough participants and compare multiple doses. Similar to the advice in the FDA’s more general guidance on dose optimization for oncology products, the new RPT document emphasizes the importance of randomized dose-response trials, saying “Randomized dose-response trials may be particularly important when evaluating a dosage(s) that exceeds EBRT-defined organ tolerances, or a dosage(s) that corresponds to the MTD of an RPT.” The guidance emphasizes the importance of considering lower dosages for participants who have received prior RPTs considering cumulative toxicity. The draft guidance also outlines the safety monitoring and dosimetry data that should support RPT dosing.
• The document remains in draft form. It garnered significant feedback from stakeholders. Several comments addressed the guidance’s considerations for exceeding EBRT organ-tolerance limits. Eli Lilly expressed support for the suggested approach, which it said, “reflects a welcome shift away from overly restrictive EBRT-based dosage constraints, paving the way for innovation and more effective RPT dosing strategies that better balance safety with therapeutic potential.” The American Association for Cancer Research agreed that crossing this threshold could be appropriate in certain circumstances, writing, “To enhance transparency and efficiency, FDA could describe additional data types (i.e., beyond clinical safety and efficacy results) that may justify such dosing, including model-based predictions or translational findings.” Some stakeholders requested further flexibility; Johnson and Johnson wrote that formal meetings to discuss doses exceeding EBRT could add friction, saying, “We believe that trials can be prospectively designed to include safeguards via appropriate study designs that are submitted to an IND and thus would negate the need for formal meetings with FDA.”

Now, a new white paper builds off the draft guidance

• Friends of Cancer Research published a document titled “Optimizing Dosing for Radiopharmaceutical Therapies” on Sept. 1. The white paper, developed by a working group made up of pharmaceutical, radiopharmaceutical and academic medicine stakeholders, with participation from Friends representatives, explicitly acknowledges the FDA’s August 2025 draft guidance publication. However, the paper explains that “additional operational guidance may help sponsors translate these principles into practical trial designs, escalation strategies, and evidence-generation approaches.” The paper first offers a detailed discussion of its scope and relevant terminology, which received a relatively light overview in the FDA’s guidance. The white paper also explains how RPT dose optimization can align with principles previously articulated by OCE’s Project Optimus.
• The white paper describes an alternative perspective on the role of EBRT-derived reference points. The FDA’s draft guidance says that sponsors can propose doses to exceed EBRT organ tolerances “when there is adequate rationale” and the dose is justified and discussed with the FDA in formal meetings. On the other hand, the white paper describes reliance for RPT development on EBRT-derived limits as misplaced due to fundamental differences in delivery, distribution and timing. Taken together, these differences affect the exposure-outcome relationship in the two modalities in a fashion that is likely to widen the therapeutic window for RPT drug delivery, where tumor response may be seen further in advance of toxicity to normal tissue – as explained in Figure 1 of the white paper. The relationships “are less well defined and may vary across patients due to differences in administered activity, biodistribution, tumor and normal tissue uptake and retention, radionuclide properties, PK, and targeting efficiency,” the figure’s caption explains. “EBRT-derived thresholds should therefore be treated as contextual reference points rather than fixed escalation boundaries,” the paper says, later explaining, “Generating empirical data across a range of administered activity levels – including levels beyond traditional reference points – is necessary to characterize RPT-specific therapeutic index and support evidence-based dosage optimization. Biologically informed, dosimetry-informed, or biodistribution-informed escalation strategies may also be appropriate when supported by a clear scientific rationale.”
• The paper offers a robust discussion of how RPT-specific dosage optimization can be operationalized through careful administered activity escalation. Figure 2 in the white paper summarizes a conceptual framework for a progressive transition involving four key elements: Characterization of exposure-response and exposure-toxicity relationships, integrating multiple data sources and evaluating patient-to-patient variability. “EBRT-derived reference points are expected to remain an anchor for starting dose selection in RPT development,” the paper says. “As RPT-specific data accumulate, dosage decisions should transition from EBRT-anchored assumptions toward evidence-based optimization informed by observed relationships among administered activity, dosimetry, safety, and therapeutic activity. In this framework, EBRTs provide context for starting dose selection but do not define subsequent escalation boundaries.”
• Administered activity escalation warrants additional operational considerations to address and balance safety risks with development goals. Table 3 in the white paper maps benefit-risk and evidence generation considerations for a range of development contexts, explaining how the authors see evidence generation informing future development in each context. For example, the paper proposes that escalation will likely go above EBRT-derived ranges in the development context of late-stage or refractory disease, as opposed to earlier-line treatment or conditions with longer survival curves where longer-term and more mature data will likely be available. Additionally, prior RPT exposure “should generally not be considered exclusionary,” the authors wrote, supporting the approach with data from prior clinical studies. “Prior radiation exposure alone may not be a reliable predictor of toxicity risk,” they concluded.
• The paper sets out a framework of “data domains” to guide dose escalation decisions, based on a structured review process and predefined criteria. These domains include clinical safety observations; PK and pharmacodynamic, systemic exposure, and safety biomarkers; dosimetry and radiation exposure patterns; imaging that shows biodistribution and treatment response; and “emerging clinical signals of therapeutic activity,” which could include reduced symptoms or improved functional status. Another consideration is the potential for delayed or time-dependant toxicities, though the white paper acknowledges that “Very long toxicity observation periods may be operationally impractical during dose escalation studies and may extend beyond the expected survival of many patients with advanced malignancies.”
• The paper proposes that sponsors leverage “layered” safety strategies that combine short-term monitoring and long-term follow up. “Layered monitoring strategies can combine standard [dose-limiting toxicity]-based escalation rules with scheduled safety reviews, interim analyses, or protocol-defined checkpoints to evaluate longitudinal safety data without unnecessarily delaying escalation,” it explains. From a governance perspective, the paper generally supports an adaptive approach where emerging data can inform ongoing decisions. A practical example: “trial governance may pair a primary escalation framework with protocol-defined review rules that apply observed toxicity rates from later cycles as triggers for cohort expansion, de-escalation, dosage modification, or stopping.” Table 4 in the paper outlines the authors’ envisioned layered safety monitoring and governance of dose escalation, encapsulating much of the paper’s previous discussion.
• Evidence generated through this framework should be structured to enable decision-making, and the paper provides a question-based example in its Appendix 1. The authors urge sponsors to implement “granular, transparent reporting” to benefit both internal decision-making and the broader field. “Because RPT-specific exposure-outcome relationships remain sparsely characterized, the collective body of evidence assembled across academia and industry will allow the entire field to better inform escalation and dosage optimization decisions with greater confidence than any single program can achieve in isolation.”
• The paper concludes with four near-term focus areas: (1) establishing shared principles for administered activity escalation in RPT development, (2) advancing totality-of-evidence decision-making, (3) harmonizing prospective evidence generation and (4) supporting iterative refinement of RPT-specific dosage optimization frameworks. “Together, these steps can help shift RPT development from inappropriate reliance on historical EBRT-derived reference points toward evidence-based approaches tailored to the characteristics of each therapy,” the paper states.

Analysis

• OCE is ostensibly working on incorporating stakeholder feedback into the final version of its guidance on the topic. The progress on that effort is unclear, since OCE just published a 2026 guidance agenda in September containing only two items, neither of which is the radiopharmaceutical draft guidance. However, the coalition led by Friends of Cancer Research is working to expand on gaps left by the draft, offering a rich source for the agency to potentially pull from should it look to add more detail to its recommendations.
• The robust framework for safety monitoring and dose escalation proposed in the draft may provide a structured framework that could allow sponsors to move beyond the current draft’s recommendation to seek FDA consultation for each instance where sponsors propose to exceed EBRT-established dose limitations. The white paper’s proposed framework incorporates widely employed regulatory concepts such as benefit-risk and clinical context of use and may inform the final guidance’s approach. A shift here could result in more predictability and flexibility for developers.
• What’s next? The paper will be the focus of a forthcoming Sept. 15 public webinar featuring a variety of industry players. During the event, “expert panelists will examine fit-for-purpose strategies for administered activity escalation and dosimetry-informed development that integrate emerging clinical, dosimetry, imaging, and biological data to support regulatory decision-making.”

To contact the author of this item, please email Amanda Conti ( aconti@agencyiq.com).
To contact the editor of this item, please email Holland Johnson ( hjohnson@agencyiq.com) or Kari Oakes (koakes@agencyiq.com).

Key Documents and Dates

• White paper: Optimizing Dosing for Radiopharmaceutical Therapies, September 2026
• Public Webinar, September 15, 2026

https://home.agencyiq.com/article/000001a0-8273-da3f-a5b2-bffff5150000?subType=analysis&articleSource=aiq-analysis