Review Reports

Reviewer 1

Reviewer comments

Author response

Overall, this is a well-written, comprehensive, and timely review that addresses one of the most challenging aspects of targeted alpha therapy, namely image-based dosimetry for ²²⁵Ac. The manuscript demonstrates a strong command of the physics, radiochemistry, and clinical workflow, and it successfully integrates recent advances in quantitative SPECT/CT, daughter redistribution, and dosimetric modeling. The figures are informative, and the clinical relevance is clearly articulated. The following minor comments are intended to further strengthen clarity and synthesis, rather than to address fundamental weaknesses.

We thank the reviewers for their thoughtful and constructive feedback. Their comments were highly appreciated and have helped improve the clarity, balance, and overall quality of the manuscript. We believe the revisions made in response to these suggestions have strengthened the review and enhanced its relevance for the field.

1.

While the review thoroughly describes multiple clinical and phantom-based imaging protocols, a concise comparative summary table highlighting key methodological differences (e.g., use of ²²¹Fr vs. ²¹³Bi photopeaks, number of imaging time points, handling of decay-chain disequilibrium) would improve readability and help readers rapidly contextualize the relative strengths and limitations of different approaches

Table 1 has been revised to improve clarity and focus. The radionuclide column was removed, as the table title already specifies ²²⁵Ac. Details on the gamma camera manufacturer and the number of projections have been omitted for conciseness. From Delker et al., the ¹⁷⁷Lu protocols were removed to emphasize information specific to ²²⁵Ac. The table now provides more detailed and relevant parameters, including photopeak energies, energy window widths, imaging time points, and decay-characteristic energies, enhancing its utility for readers interested in quantitative ²²⁵Ac imaging.

2.

The manuscript appropriately adopts an RBE value of approximately 5 for α-particles, consistent with current practice. However, a brief acknowledgment of the biological and microdosimetric uncertainties associated with assuming a fixed RBE—particularly with respect to dose rate, subcellular distribution, and tissue-specific sensitivity—would further strengthen the biological interpretation of the reported absorbed doses.

This has been addressed on page 23“Harmonisation of Activity Quantification for ²²⁵Ac Dosimetry – Provisional and Subject to Ongoing Validation and page 25, at the two highlighted sections, under the headings “Microdosimetry in Actinium-225 Targeted Alpha Therapy”

3

The discussion of the ¹³⁴Ce/¹³⁴La PET in vivo generator as a theranostic surrogate for ²²⁵Ac progeny behavior is a notable strength of the review. To maintain balance, the authors may consider more explicitly emphasizing known limitations, such as potential underestimation of tumor uptake for slowly internalizing vectors and the constraints of extrapolating PET-based surrogates to quantitative α-dosimetry.

The limitation are indicated on page 26, highlighted in yellow for clarity.

“At the same time, certain limitations should be considered. PET-based surrogate imaging may underestimate tumor uptake for slowly internalizing or non-internalizing vectors, and the indirect nature of this approach may limit direct extrapolation of PET-derived ac-tivity distributions to quantitative α-dosimetry.”

4

The manuscript convincingly motivates multi–time-point SPECT/CT imaging for accurate dosimetry. A short clarification on minimum clinically acceptable imaging schedules (e.g., reduced time-point or mixed-model approaches) could further enhance translational relevance, particularly for centers with limited scanner availability or patient tolerance constraints

This has been addressed under the “SPECT/CT Imaging Acquisition” section on page 13.

“Acquisition protocols for quantitative SPECT/CT imaging used in dosimetry, including scan times, projections, and energy windows, are described in detail in the subsequent section. Accurate ²²⁵Ac dosimetry typically relies on multi–time-point imaging, with three time points sufficient to capture temporal activity distributions during the initial therapy cycle [70, 71]. For subsequent cycles, a single time-point acquisition can provide a reason-able compromise between quantitative accuracy and patient burden. Hybrid approaches, in which a single SPECT/CT acquisition is complemented by multiple planar scans at additional time points, offer a practical strategy to maintain dosimetry reliability while accommodating clinical constraints. Establishing such minimum clinically acceptable im-aging schedules is particularly relevant for centers with limited scanner availability or patients with restricted tolerance for repeated imaging,”

5

The discussion on the limitations of extrapolating EBRT-derived organ tolerance doses to ²²⁵Ac therapy is appropriate and scientifically sound. The authors could strengthen this section by briefly highlighting which organs are most affected by this uncertainty (e.g., kidneys, salivary glands) and by emphasizing the current lack of robust, α-emitter–specific clinical dose–toxicity thresholds.

Addressed on page 16. The discussion now emphasizes that consideration of organs at risk, including the kidneys, salivary glands, and red marrow, is critical in ²²⁵Ac therapy, as these tissues may constrain the number of safely deliverable treatment cycles. The current lack of α-emitter–specific clinical MTD introduces additional uncertainty, underscoring the importance of careful interpretation of absorbed doses and potential cumulative toxicity in these critical organs.

Reviewer 2

Reviewer comments

Author response

The manuscript presents a review on the challenges and methodologies of image-based dosimetry for Actinium-225 in targeted alpha therapy (TAT). The authors have compiled an overview of production routes, decay properties, clinical imaging protocols, dosimetry workflows, and future directions. The manuscript is an important addition to the literature, and I request the authors address my concerns before acceptance:

We sincerely appreciate the reviewer’s careful reading and thoughtful guidance. Their constructive suggestions have been extremely helpful in refining the manuscript, enhancing its clarity, and ensuring that key points are clearly communicated. We are grateful for their time and valuable input.

1

The manuscript is lengthy and has repetitive content in many sections. Sections could be streamlined for better readability. For instance, the dosimetry workflow is described in multiple sections with overlapping content.

The term “workflow” has been removed from pages 3, 13 and to reduce repetition. 

2

The discussion of relative biological effectiveness (RBE) is spread across multiple sections. Consolidating this into a dedicated subsection would improve readability.

RBE discussion under the “Clinical Quantitative Imaging and Dosimetry Workflow for Actinium-225” on Page 11 has been removed.

On page 15, a clarifying statement was included to indicate that RBE is discussed in the section ‘Relative Biological Effectiveness Considerations and Biologically Effective Dose.

 …“To account for the increased biological effectiveness of alpha particles, absorbed doses are commonly weighted using an RBE factor of approximately five (see section on “Relative biological effectiveness Considerations and Biological Effective Dose” for further detail).

3

In Figure 6, the text describing steps 1–8 is dense. A simplified schematic or bullet-point summary of the core dosimetry steps would enhance clarity.

Core steps included on Page 12 as numbered bullets.

4

The term “voxel S-value (VSV) kernels” is introduced without sufficient explanation for a general readership.

Explanation included on Page 15. Voxel S-value (VSV) kernels represent the mean absorbed dose delivered to a target voxel per radioactive decay in a source voxel within a homogeneous medium. By convolving these precomputed kernels with the spatial distribution of time-integrated activity from imaging, voxel-level absorbed dose maps can be efficiently generated.

5

In Table 1, some entries are incomplete ( “Not specified” for several parameters). Where possible, please provide typical values or ranges based on the literature.

 Page 17 all attainable information has been included and the table updated.

6

The table caption should clearly state that it includes both clinical and phantom studies.

Page 17, The table title reflects the content, encompassing both phantom and clinical studies. It is labeled as “Summary of Clinical and Phantom ²²⁵Ac SPECT Imaging Protocols for Dosimetry”, and this distinction is also indicated within the table itself.

7

Add a column for reconstruction method (e.g., OSEM, Q.Clear) as this significantly impacts quantification.

Column for reconstruction is included.

8

In some places, “SPECT/CT” is written as “SPECT-CT” or “SPECT-CT.” Please harmonize.

Harmonized.

9

The abbreviation “TIA” is used for “time-integrated activity,” but later “TIAC” appears for “time-integrated activity coefficient.” Define these clearly upon first use.

The terms time-integrated activity (TIA) and time-integrated activity coefficients (TIACs) are defined upon first use on page 12.

Time-integrated activity coefficient (TIAC) defined on page 21

10

In production routes and waste management section, the details were detached from the central theme of image-based dosimetry. Consider condensing these or explicitly linking them to dosimetry challenges.

Following text has been included on Page 5. “The ²²⁵Ac production route via ²²⁹Th generators, ²³²Th spallation, or ²²⁶Ra irradiation determines radionuclide purity, specific activity, and co-produced isotopes such as ²²⁷Ac, all of which directly impact SPECT/CT-based image quantification. Co-emitted gamma photons from impurities influence counting statistics and require careful gamma window selection, while variations in specific activity affect signal intensity and optimal imaging time points. Additionally, waste streams containing these radionuclide impurities must be carefully characterized, decay-stored, and handled in compliance with regulatory requirements, as residual long-lived isotopes not only pose radiological hazards but can also compromise the accuracy and reproducibility of quantitative imaging and patient-specific dosimetry.

11

While challenges are mentioned (e.g., low photon yield, recoil effects), a dedicated summary of current limitations and how they affect clinical implementation would strengthen the review.

A summary has been included on Page 19.

“Clinical implementation of image-based dosimetry for ²²⁵Ac remains limited by low administered activities, restricted photon statistics, and complex decay schemes, which affect quantitative robustness. While SPECT using ²¹³Bi and ²²¹Fr emissions enables patient-specific dosimetry, protocol optimization and validation remain essential, and surrogate PET approaches provide mechanistic insight but do not directly yield absorbed α-dose. These limitations directly influence treatment planning, including uncertainty in organ-at-risk dose constraints and the number of safely deliverable therapy cycles.”

12

The manuscript mentions AI and CZT cameras as future directions but could better discuss their current readiness level and barriers to clinical adoption.

Challenges for CZT are addressed on Page 26.

AI-based methods have demonstrated clear potential to improve quantitative ²²⁵Ac dosimetry through projection denoising, image registration, automated segmentation, and voxel-level TACs, but wider clinical adoption is currently limited by the need for standardized training datasets, regulatory validation, and seamless integration into clinical workflows. CZT-based gamma cameras offer improved sensitivity and energy resolution for low-count α-emitter imaging, however, their routine clinical adoption remains constrained by system availability, cost, and technical challenges such as septal penetration, high-energy photon handling, and detector dead-time effects. Continued technical optimization and multicenter validation are required before these technologies can be broadly implemented in clinical dosimetry practice.

Reviewer 3

Reviewer comments

Author response

The manuscript is well written and free of serious issues.

1.

However, Table 1 on page 15 should be Table 2, and the 225Ac on line 298 should be in superscript.

The correction has been made on page 17, and the associated in-text citation has been updated accordingly.

2

Other improvements to the manuscript could be made by adding a list of abbreviations, as well as providing clearer explanations of the various targeting vectors, e.g. Ac-J594, Ac YS5, Ac-hu11B6 or Ac-SibuDAB on lines 56 and 61.

The list of abbreviation has been included and the full names of the vectors specified on Page 2.

3

It would also be beneficial to include figures showing the structure of the complexes used, such as EDTA, HEHA, DTPA, DOTPA and DOTATATE.

The complete names of the complexes have been included. We thank the reviewer for this valuable suggestion, which could indeed enhance the clarity and completeness of the discussion. However, given the scope and focus of the current manuscript on image-based dosimetry

rather than radiochemical design, it could be beyond the current remit to include structural figures of these complexes for this manuscript.