Validation of Preoperative Neoadjuvant Bevacizumab Therapy for Newly Diagnosed Glioblastoma via Comparative Analyses with Propensity Score Matching

Round 1

Reviewer 1 Report (Previous Reviewer 1)

Comments and Suggestions for Authors

Thank you for the response. The revisions address my main concerns and acknowledge the study's limitations. I have no further comments.

Reviewer 2 Report (Previous Reviewer 3)

Comments and Suggestions for Authors

The article has been improved sufficiently.

Reviewer 3 Report (Previous Reviewer 4)

Comments and Suggestions for Authors

The authors have done a good job in refining the conclusions based on these data and are to be commended for their efforts of unearthing data for this relevant clinical question. 

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This study examines whether neoadjuvant bevacizumab administered before resection is associated with improved perioperative functional outcomes and favorable imaging response in glioblastoma, and whether these changes translate into OS differences compared with patients who did not receive NeoBev. The authors report a better postoperative KPS in the NeoBev group despite no clear survival advantage. To strengthen confidence in these findings, the manuscript requires more consistent reporting (including percentages and statistical tests), correction of figure mislabeling, and analyses that account for the significantly greater extent of resection in the NeoBev group, as well as clarification of whether imaging response–survival associations are treatment-specific. Please find my comments below:

Major comments:

• The authors mention that baseline characteristics were similar between the NeoBev and Control groups except for the extent of resection, significantly higher in the NeoBev group (p<0.01). I am supposing that p-values were calculated using percentages instead of raw numbers. It is recommended to include the percentages for all values relative to the total number of patients in each group in Tables I and II.
• There is a mislabeling in Figure 1. Panel A in Figure 1 indicates PFS. However, the figure indicates OS. Similarly, panel B indicates a cut-off of 55%. However, the figure shows a cutoff of 37%.
• The conclusion that NeoBev improves post-operative function may be confounded by the significantly higher extent of resection in the NeoBev group, since EOR is a key determinant of post-op KPS and may also lie on the causal pathway (NeoBev enabling greater resection). Although baseline KPS appears similar between groups, please clarify that the timing of KPS assessments (preoperative and postoperative) is comparable between groups, given surgery is scheduled 21–30 days after NeoBev. In addition, report the within-patient change from baseline to postoperative KPS for each group (baseline vs post-op, and difference in KPS), with statisitcal tests. It is also recommended to perform an adjusted analysis (e.g., ANCOVA/regression or ordinal model) including extent of resection, plus a sensitivity analysis stratified by extent of resection. The discussion should texplicitly consider EOR as a confounder/mediator.
• There is also a mislabeling in Figure 4 with the text. The authors mention that no significant differences in survival were observed between groups in Figure 4A and 4B but p values are <0.01 and referring to the 37% cutoff.
• As presented in Figure 4, it’s unclear whether the reported association between imaging response (e.g., tumor volume reduction) and survival is specific to NeoBev or simply reflects the general prognostic value of reduced tumor burden regardless of treatment. Please compare NeoBev and control patients directly (e.g., stratified analyses and a formal interaction test) to assess whether NeoBev modifies the imaging–survival relationship.

Minor comments:

-Please include the statistical test in each figure legend.

- Please increase the font size in every figure so that it is easier to read.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Authors presented a good paper on bevacizumab application as a neoad agent. however groups are inhomogeneous in matter of number and type of tumors (e.g. differences between IDH mutant and WT should be considered because it could be an enormous bias in survival parameters). Authors are invited to subanalyze each group.

The analysis of bevacizumab response should be quantitatively analyzed. Adding a pre- and post op PET scan analysis could maximize the results and give strength to the paper.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The study provides a comprehensive evaluation of the effectiveness of neoadjuvant bevacizumab (neoBev) therapy in newly-diagnosed glioblastoma (GB), making a clear distinction between patients undergoing standard chemoradiotherapies and those who used neoadjuvant therapy by matching patients using a propensity score approach to focus on patients undergoing standard treatment. The write-up provides significant strength to data by stating:
The manuscript describes a well-organized outcome assessment on the basis of performance status, extent of resection, and volumetric MRI response, focusing on T1Gd contrast-enhanced response without underestimating the importance of T1Gadolinium contrast reduction in place of FLAIR to avoid the pseudo-response criteria by showing a significant impact on predicting progression-free and overall survival by a reduction of ≥37% in T1Gd criteria by ROC analysis, while FLAIR reduction was less specific to promote pseudo-response criteria by concurrently confirming the effectiveness of neoBev therapy in improving patients’ function and resectability while failing to improve universal survival, especially in patients undergoing gross total resection therapy by showing a superior progression-free survival in patients undergoing standard therapy alone in a significant manner.
The report provides strength to its findings by pointing out its multicenter and propensity-matched data, while being somewhat weak by being a retrospective and smaller sample size approach to focus on molecular malignant profiling in patients. However, this article provides insight to clinical outcomes and practice by recognizing GB patients with lower initial KPS and those within eloquent regions to focus on the effectiveness of therapy in GB patients by recommending a potentially significant volumetric measure involving patients undergoing T1Gadolinium reduction in place of FLAIR reduction to avoid pseudo-response criteria among GB patients.

The authors need to do some suggested changes to make the manuscript better:


Although there is an explanation of the inclusion criteria, additional information on how candidates in the neoBev group were identified from a clinical perspective (for example, the location of the tumors, the level of symptoms) would have helped. So, Explain selection criteria for NeoBev group properly.

Enhance Molecular Profiling: Adding MGMT methylation status, TERT, and EGFR information (which was absent in several instances) may improve the prognostic and treatment information.

Justify the 21–30 Day Surgical Window: This point can be explained by providing more evidence with specific references to the particular time period mentioned.

Report Adverse Events: The perioperative complications, like wound healing, thromboembolism, should be included for assessing safety, which has particular relevance for anti-VEGF agents.

External Validation of T1Gd ≥37% Threshold: Suggest external validation of the ≥37% cutoff in an independent prospective study, perhaps using radiomic/perfusion imaging features.

Think about Prospective Trial Design: Propose an adaptive/biomarker-guided trial in which patients can be stratified on the basis of eloquence, edema, and/or VEGF expression

Supplement Surgical Data: Incorporating information on operative time, blood loss, and resection margins would prove the conclusion that operability has been enhanced.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

Pre-treatment with bevacizumab of newly diagnosed patients likely to have a glioblastoma is presented as a potential way to improve clinical outcomes. The core message of this paper is that bevacizumab may increase resectability in selected patients and improve initial performance scores, although not resulting in improved PFS or OS.

I recommend the authors for the detailed and hard work. However, these conclusions are not substantiated by the current analysis, which has some major flaws, due to fundamental issues with the design.

First, the small sample size is the most important consideration with effects on multiple levels: selection over a longer period of time, the patient population is highly selected. Further, because of the low sample size, the propensity score matching can only rely on a limited number of potential prognostic factors, increasing the likelihood of residual confounding. For example, data on MGMT-promotor methylation, preoperative tumor size, T2-Flair volume, steroid use are missing from the comparison without explanation.

Second, the comparison of baseline factors of the matched cohorts can be ameliorated by addition of steroid use, tumor volumes, waiting time from first MRI to surgery, other procedural effects.

Third, as the authors also acknowledge, the data driven cut-off determination in the T1-Gd analysis, is a a major limitation, and in my opinion should be removed prior to publication. A reader may perceive this method as a way to find significant p-values.

Overall clinical considerations: the patients that would be selected for neoBev would be patients with larger tumors with more edema and more clinical deficits. Patients that would now be pre-treated with steroids to accomplish oedema reduction and functional improvement before a decision to operate is taken. In these cases, delaying surgery to allow for the bevacizumab “washout” period would pose a risk of tumor growth, performance deterioration and worse outcome.

Given the potential harm that is induced by such a strategy, it is my opinion that more reliable positive outcome data from prospective trials should be presented before any patient is offered neoBev outside clinical trials.

Author Response

Please see the attachment

Author Response File: Author Response.pdf