Longitudinal Diffusion MRI Characterizes Persistent Perivascular Diffusivity Asymmetry and White Matter Abnormalities After Cranioplasty for Decompressive Craniectomy

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

1. A major concern is the validity and interpretability of the DTI-ALPS index in the context of significant brain morphological changes after decompressive craniotomy. The ALPS index is based on specific anatomical assumptions regarding the orientation of projection and association fibers into the perivascular space. However, in patients with large skull defects, significant brain deformation, ventricular distortion, and changes in white matter structure are expected. Under these circumstances, these assumptions may no longer hold. Furthermore, recent views suggest that ALPS should be interpreted more cautiously and may reflect the ultrastructural characteristics of the white matter rather than being a direct indicator of glymphatic function. Under these circumstances, perioperative changes in brain morphology, cerebrospinal fluid distribution, and fiber orientation can significantly complicate ALPS measurements. Therefore, it remains unclear whether the observed differences truly reflect physiological changes in perivascular diffusion rate or are primarily caused by structural and geometric changes. This issue is particularly important in longitudinal study designs, as cranioplasty itself can induce geometric normalization, further complicating interpretation. The authors recommend a clearer recognition of this limitation and a corresponding adjustment of the physiological interpretation of ALPS. Therefore, the potential impact of brain morphological changes (brain displacement, ventricular deformation, changes in fiber orientation, etc.) on ALPS measurements needs to be clearly recognized and discussed in more detail.

2. Another important concern is the potential impact of cranioplasty-related imaging artifacts on diffusion MRI measurements. While PEEK implants are generally considered MRI-compatible, diffusion imaging using EPI is highly susceptible to geometric distortion, signal heterogeneity, and susceptibility-related artifacts. These effects can be particularly pronounced near cranioplasty sites. Importantly, concerns arise regarding the comparability of longitudinal measurements due to fundamentally different imaging conditions before and after cranioplasty (i.e., skull defect and implant material). Furthermore, the lack of susceptibility distortion correction (-rpe_none) may exacerbate these issues. This paper does not fully explore how these potential confounding factors affected ALPS measurements. Therefore, it is unclear whether the observed differences reflect biological changes or image-related artifacts. Given the lack of significant temporal changes in overall ALPS and white matter indices, interpretations of postoperative dynamic reorganization should be cautious.

3. It should be more clearly recognized that there is no significant association between ALPS and cognitive function or structural indices, and the clinical significance should be interpreted with caution.

Minor correction:
When abbreviations appear for the first time, their full names should be specified (see TBI).

Author Response

Dear Academic Editor and Reviewers,

Thank you very much for your careful evaluation and constructive comments on our manuscript. We have carefully revised the manuscript according to the reviewers’ suggestions and prepared a detailed point-by-point response.

Please see the attached Response to Reviewers file for the full response, including the corresponding page and line references in the revised manuscript.

Kind regards,
The Authors

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This study investigates longitudinal changes in perivascular diffusivity and white matter microstructure following cranioplasty, aiming to evaluate DTI-ALPS as a potential imaging biomarker of postoperative brain physiology. The authors used diffusion MRI in a cohort of patients undergoing cranioplasty after decompressive craniectomy, with preoperative and 3-month follow-up assessments, combined with TBSS, AFQ, and neuropsychological testing. They report that DTI-ALPS is reduced compared to controls, remains stable postoperatively, shows hemispheric asymmetry related to the defect side, and that white matter abnormalities persist without clear recovery. No robust associations between imaging metrics and cognition were identified.

These findings suggest a dissociation between physiological (perivascular diffusivity) and structural (white matter integrity) recovery processes after cranioplasty. The study contributes to the field by integrating glymphatic-related imaging proxies with advanced tract-based analyses in a longitudinal design, which is relatively novel in this clinical context. However, several important methodological and conceptual limitations reduce the strength of the conclusions, particularly regarding interpretation of ALPS as a glymphatic marker, technical limitations in diffusion processing, cohort heterogeneity, and limited longitudinal power.

Major Comments

1. The study includes a clinically heterogeneous population combining traumatic brain injury and malignant MCA infarction. These conditions differ substantially in pathophysiology, lesion distribution, and recovery trajectories. While both are indications for decompressive craniectomy, pooling them without stratified analysis introduces potential confounding and limits interpretability of the findings. No subgroup analyses are presented to assess whether results are consistent across etiologies. Additionally, there is a notable reduction from 43 baseline patients to 25 at follow-up. This attrition is substantial for a longitudinal study and may introduce selection bias. The manuscript does not evaluate whether patients lost to follow-up differ systematically from those retained, nor does it discuss how this may influence longitudinal conclusions.
2. The methodological framework is generally well described; however, several critical issues remain. First, diffusion preprocessing lacks susceptibility distortion correction due to absence of reverse phase-encoding or field maps. This is a significant limitation, particularly given that ALPS measurements are performed near ventricular regions where distortions are most pronounced. This limitation directly affects the validity of the primary outcome but is not sufficiently emphasized. Second, ALPS ROI placement was performed by two raters, but no inter-rater reliability metrics are reported. Given the sensitivity of ALPS to ROI positioning, the absence of quantitative reliability measures (e.g., ICC) weakens confidence in reproducibility. Third, while ALPS is correctly described as a proxy of perivascular diffusivity, the manuscript still frames interpretations in terms of glymphatic function. This conceptual leap is not fully justified by the methodology and should be more carefully constrained.
3. The use of linear mixed-effects models for longitudinal analysis is appropriate and a strength. Multiple comparison corrections (FDR, Bonferroni) are also applied. However, the study appears underpowered for several analyses, particularly longitudinal comparisons and correlation analyses. The absence of significant associations between imaging and cognition may reflect insufficient power rather than true absence of relationships. This limitation is not adequately discussed.
4. The interpretation of findings is generally logical but occasionally overextended. The conclusion that physiological and structural recovery follow different trajectories is presented as a key insight, yet it is largely based on negative findings (lack of ALPS change and lack of correlations). Alternative explanations such as limited sensitivity of ALPS, methodological constraints, or insufficient follow-up duration are not sufficiently considered. Similarly, references to glymphatic dysfunction should be tempered, as ALPS does not directly measure glymphatic flow or clearance.
5. Several analyses would strengthen the manuscript: Stratified analyses by etiology (TBI vs stroke) to address cohort heterogeneity, Assessment of differences between follow-up completers and non-completers, Inter-rater reliability metrics for ALPS ROI placement, and Sensitivity analyses addressing potential effects of diffusion distortions. Without these, the robustness of the conclusions remains limited.

Minor Comments

1. The Introduction is generally well written but somewhat dense and could be streamlined for clarity.
2. Terminology around “glymphatic system” and “perivascular diffusivity” should be used more consistently to avoid conceptual ambiguity.
3. Figure captions are informative, but some figures (particularly tract-based and node-wise plots) are visually dense and would benefit from clearer highlighting of key findings.
4. The neuropsychological results are extensively reported, but their clinical interpretation could be more concise and better integrated with imaging findings.
5. Some repetition is present between Results and Discussion sections, particularly regarding white matter abnormalities.

Author Response

Dear Academic Editor and Reviewers,

Thank you very much for your careful evaluation and constructive comments on our manuscript. We have carefully revised the manuscript according to the reviewers’ suggestions and prepared a detailed point-by-point response.

Please see the attached Response to Reviewers file for the full response, including the corresponding page and line references in the revised manuscript.

Kind regards,
The Authors

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

1. This paper explores the use of DTI-ALPS as a perioperative biomarker of brain function after cranioplasty in patients performing decompressive craniectomy. Although the manuscript addresses the relevant clinical issues, several points remain to be clarified.

   1. The diffusion MRI protocol did not include correction for susceptibility distortion.
   2.  The 3-month interval of follow-up was insufficient to capture clinically meaningful recovery of white matter structure, but allowed an overly strong conclusion of temporal dissociation between physiological reorganization and structural repair.
   3. Lack of significant relationships between imaging and cognitive measures after correction for multiple comparisons likely reflects the relatively small and modest sample.
   4. The TBSS results demonstrate widespread FA reductions, but the biological interpretation conflates chronic secondary degeneration with potentially reversible perioperative changes.
   5. The TBSS results report a mixed cohort of TBI and stroke patients with no subgroup analysis.
   6.  Manual delineation of regions of interest for ALPS was performed in a blinded manner, but without reporting of inter-observer reliability.
   7.  The authors should temper their claims regarding glymphatic function and reframe the discussion around perivascular diffusivity or tissue microstructure to maintain scientific rigor.

Author Response

Dear Academic Editor and Reviewers,

Thank you very much for your careful evaluation and constructive comments on our manuscript. We have carefully revised the manuscript according to the reviewers’ suggestions and prepared a detailed point-by-point response.

Please see the attached Response to Reviewers file for the full response, including the corresponding page and line references in the revised manuscript.

Kind regards,
The Authors

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Thank you for your sincere response.

Reviewer 3 Report

Comments and Suggestions for Authors

The author has provided effective responses to the raised questions, and publication is recommended.