Advancing radiopharmaceutical development: Preclinical precision meets a new era of FDA oversight

Radiopharmaceuticals have rapidly evolved from a specialised therapeutic niche into a cornerstone of modern oncology. By combining molecular targeting with controlled radiation delivery, they add a new dimension to precision medicine, enabling both targeted treatment and quantitative imaging.

The clinical success of Lutathera for neuroendocrine tumours and Pluvicto for metastatic prostate cancer showed that radioligand therapies can achieve durable responses in patients who previously had limited options. Advances in isotope chemistry, conjugation design, and imaging analytics have broadened the field’s reach, drawing strong interest from established pharmaceutical companies and emerging biotech innovators.

Regulatory science is evolving in step with this innovation. The FDA now emphasises a continuum linking discovery, preclinical research, and clinical translation within a single, evidence-driven framework. Its 2025 draft guidance, Oncology Therapeutic Radiopharmaceuticals: Dosage Optimization During Clinical Development, introduces new expectations for how sponsors justify administered activity, optimise dosing, and model patient-specific exposure. The guidance moves beyond fixed-dose paradigms, encouraging individualised dosimetry and adaptive study designs supported by robust translational data.

This shift builds on earlier FDA frameworks that defined how biodistribution, dosimetry, and safety data could support early human evaluation. The emphasis has now moved toward integration, combining preclinical modelling, quantitative imaging, and early clinical data to inform dose justification in real time.

Together, these developments reflect a more dynamic regulatory philosophy. The FDA no longer treats non-clinical and clinical phases as separate stages, but as connected parts of a continuous translational process. Radiopharmaceutical sponsors are expected to demonstrate that preclinical models and clinical protocols share a common evidentiary thread tying biological mechanism to patient outcome through quantitative, reproducible data.

What the 2025 FDA guidance means for drug development

The 2025 draft guidance addresses a key challenge in radiopharmaceutical R&D: balancing effective tumour targeting with minimal radiation exposure to healthy organs. Traditional dose limits, derived from external beam radiation therapy, do not always apply to systemically administered radioligands. The new guidance recognises this and allows sponsors to justify higher administered activities or cumulative doses when supported by robust data.

This represents a meaningful shift from older paradigms. The FDA now invites sponsors to explore dose levels beyond conventional thresholds, provided modelling and empirical data show the doses are safe and therapeutically justified. To achieve this, developers should use early-phase trials supported by strong preclinical evidence and predictive models that represent human tumour and organ behaviour.

The guidance places renewed emphasis on quantitative dosimetry. Sponsors are encouraged to conduct individualised studies early, using imaging and modelling techniques to link preclinical biodistribution data with human dose projections. It also stresses transparent uncertainty reporting, including assumptions, variability, and limitations in dose estimation.

From a clinical design perspective, the agency favours adaptive, data-rich early-phase studies where cohorts test different administered activities or cycle frequencies. Imaging and biomarker data then guide the optimal balance between tumour control and organ safety. The guidance also underscores the importance of long-term follow-up, recognising that late radiation effects may take months or years to appear.

The overarching message is clear: radiopharmaceutical development must move beyond fixed paradigms toward patient-specific, data-driven optimisation. This requires a stronger bridge between preclinical and clinical evidence than ever before.

With radiopharmaceutical innovation accelerating and competition intensifying, developers are taking a closer look at how preclinical strategy drives long-term success. Investors, partners, and regulators now expect programmes to generate stronger translational evidence earlier in development, recognising that rigorous model design reduces downstream risk and shortens timelines.

This shift reflects a convergence of market expectations and scientific capability, rather than a purely regulatory response. In this environment, patient-derived xenograft (PDX) models have become critical, enabling teams to capture clinically relevant tumour behaviour, refine dose-response predictions, and build preclinical datasets that strengthen confidence before first-in-human studies.

PDX tumours preserve the architecture, receptor expression, and microenvironment of human cancers. For radiopharmaceutical development, this realism enables meaningful evaluation of tumour-to-organ dose ratios, uptake kinetics, and therapeutic margins. When biodistribution and dosimetry are performed in PDX models, the resulting data better predicts how compounds behave in patients, supporting confidence in non-traditional dosing strategies.

For example, a sponsor developing a somatostatin receptor-2 (SSTR2)–targeted radioligand can use a panel of neuroendocrine PDX models to assess variability and consistency of uptake across clinically relevant receptor densities. This data helps justify isotope choice, linker chemistry, and administered activity per cycle, aligning with the FDA’s expectations for data-driven dose justification.

Likewise, in developing next-generation PSMA-targeted alpha emitters, PDX models from heavily pretreated prostate tumours can estimate tumour-to-kidney ratios under different isotopic conditions. If results show stable safety margins and sustained tumour retention, they provide a rational basis for exploring higher administered doses in early trials.

PDX models also underpin the personalised dosimetry concept championed by the FDA. Studying multiple PDX tumours representing different molecular subtypes or resistance phenotypes helps identify patterns that may guide individualised dosing later in development. This data contributes to the mechanistic understanding of radiation distribution and variability the agency now expects, advancing radiopharmaceuticals toward true precision medicine.

Integrating PDX, imaging, and omics in the new era of development

The FDA’s guidance aligns with a broader trend: integrating imaging, modelling, and multi-omic profiling across the continuum of drug discovery. Its call for product-specific dosimetry and dose–response modelling mirrors modern R&D practices combining molecular analytics with computational prediction.

When paired with PDX models, these tools create a powerful feedback loop between discovery and translation. Linking proteomic or transcriptomic data from PDX tumours with PET/CT imaging reveals how receptor expression correlates with radioligand uptake and treatment efficacy. Such correlations guide patient selection and adaptive dosing strategies, fulfilling the FDA’s vision of a radiopharmaceutical ecosystem that is both data-driven and biologically grounded.

The broader implications for the industry

The 2025 draft guide establishes a regulatory environment that is both flexible and demanding. It rewards scientific rigour, translational modelling, and proactive dialogue with regulators. For developers, investment in biologically relevant models, quantitative dosimetry, and imaging infrastructure is no longer optional. It is fundamental.

Programmes relying solely on cell line xenografts or simplified dosimetry risk being viewed as insufficiently predictive. In contrast, studies leveraging PDX models, advanced imaging, and mechanistic modelling will be better positioned to meet FDA expectations for dose justification and safety evaluation. Ultimately, this framework aims to accelerate timelines by reducing uncertainty, not lowering standards.

Radiopharmaceuticals are transforming oncology, and regulatory science is evolving to match their sophistication. The FDA’s 2025 draft guidance on dosage optimisation marks a pivotal moment, setting clear expectations for how sponsors should approach dose finding, safety modelling, and individualised treatment planning.

Within this framework, patient-derived xenografts play a central role. Their ability to capture human tumour heterogeneity and preserve clinically relevant biology allows them to bridge preclinical dosimetry with patient outcomes. PDX models help developers craft smarter, defensible dosing strategies that align with FDA expectations while de-risking clinical translation.

As the field advances, success will depend on integrating high-fidelity tumour models, quantitative imaging, and molecular data into cohesive development ecosystems. Companies that embrace this approach will meet evolving regulatory standards and deliver safer, more effective radiopharmaceutical therapies to patients faster and with greater confidence.

References

1. U.S. Food and Drug Administration. Oncology Therapeutic Radiopharmaceuticals: Nonclinical Studies and Labeling Recommendations. August 2019. 
2. U.S. Food and Drug Administration. Oncology Therapeutic Radiopharmaceuticals: Dosage Optimization During Clinical Development (Draft Guidance). August 2025. 
3. U.S. Food and Drug Administration. Exploratory IND Studies. January 2006. 

About the author

Denis R. Beckford-Vera, PhD, is head of radiopharmacology at Champions Oncology. A radiopharmaceutical R&D leader with 20+ years of experience advancing novel imaging agents and targeted radiotherapies from discovery to clinic, he specialises in immunoPET and alpha/beta-emitting therapeutics, with recognised achievements in CD46- and CD33-directed radioimmunotherapies. Beckford-Vera has built and led high-performing teams, established laboratories, and guided programmes through IND-enabling studies and first-in-human trials. He holds a PhD in Nuclear Chemistry, a miniMBA from Rutgers, and executive training from Wharton. He is the author of more than 25 peer-reviewed publications, four book chapters, and two US patents, blending deep scientific expertise with strategic leadership in precision oncology.