The Utility of Spectroscopic MRI in Stereotactic Biopsy and Radiotherapy Guidance in Newly Diagnosed Glioblastoma
, Karthik K. Ramesh 1,2,†, Anuradha G. Trivedi 1,2, Vicki Huang 1,2, Eduard Schreibmann 1, Brent D. Weinberg 3,4, Lawrence R. Kleinberg 5, Hui-Kuo G. Shu 1,4, Hyunsuk Shim 1,2,3,4 and Jeffrey J. Olson 4,6,*
Carine Chassain
Round 1
Reviewer 1 Report (Previous Reviewer 1)
Comments and Suggestions for AuthorsPlease see attached.
Comments for author File: 
Comments.pdf
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report (Previous Reviewer 2)
Comments and Suggestions for AuthorsComments for authors
Tomography – Manuscript ID: tomography-2788679 – V3 (received February 23, 2024) – “The Utility of Spectroscopic MRI in Stereotactic Biopsy and Radiotherapy Guidance in Newly Diagnosed Glioblastoma.”, by Abinand Rejimon, Karthik K. Ramesh, Anuradha G. Trivedi, Vicki Huang, Eduard Schreibmann, Brent D. Weinberg, Lawrence R. Kleinberg, Hui-Kuo G. Shu, Hyunsuk Shim and Jeffrey J. Olson.
The authors submitted a carefully-prepared revision, which satisfactorily addressed the remaining concerns.
Therefore, I recommend publication.
Author Response
We thank the reviewer for the acceptance of our publication.
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 AuthorsThe manuscript tomography_2788679 titled “The Utility of Spectroscopic MRI in Stereotactic Biopsy and Radiotherapy Guidance in Newly Diagnosed Glioblastoma” by Rejimon et al., describes the potential of using spectroscopic MRI (sMRI) in treatment of glioblastoma. I have several concerns about the manuscript (both minor and major), some of which I am listing below.
1. sMRI: Magnetic resonance spectroscopic imaging (MRSI) is an well-established method to measure neuro metabolite levels and obtain metabolite level maps over a large region of the brain. What is new about sMRI? It seems to be exactly the same as MRSI. It is not clear why the authors would introduce a new terminology instead of using the recognized terminology (MRSI). Perhaps sMRI has some difference from MRSI, in which case the authors should elaborate that. Otherwise replacing sMRI with MRSI is required, or some justification to use “sMRI” is needed.
2. I assume “RT” is referring to radiotherapy. Please mention that.
3. Pleas define CE and OS at their 1st mention.
4. In Abstract: “The median overall…was smaller (p=0.07)”: p=0.07 makes the comparison insignificant and so use of “smaller” is questionable.
5. In Introduction: “The average Cho/NAA…brain (0.45) or edema (0.77)”: The ratio quoted as 5.26 seems to be 5.18 from reference 19. In addition, I could not find the other ratios in reference 6 and 20. What exactly are these ratios representing? Surely, these are not the ratio of Cho and NAA levels, which for healthy adults is around 0.15. Are these ratios derived from the areas of the resonance peaks prior to correction for proton numbers (and relaxation)?
6. In Introduction: “T1w-CE and in T2w/FLAIR…detected by sMRI”: Please give a reference to justify this statement.
7. In Introduction: “In a study published…and ex-vivo fluorescence [12]”: What is meant by “high-resolution” here? Also, the 1st half of the sentence (especially “published by our 81 research group at Emory University in 2016”) needs to be rewritten.
8. In Introduction: “In another small cohort study…”: Please mention the cohort size.
9. In Introduction: “At Emory University …grade III tumors”: This sentence seems to be out of place unless a reference is added. Please rephrase this or delete this sentence.
10. In Introduction: Please remove the adjective “easily”.
11. In Materials and Methods: Please mention the model of the scanner.
12. In Materials and Methods: Please mention the acquisition resolution (prior to interpolation) and spatial coverage extent of the EPSI sequence.
13. In Materials and Methods: Please mention the acquisition parameters of the T1W-CE and T2W-FLAIR sequences including the resolution of each. Also, please replace TW-FLAIR by T2W-FLAIR in line 147.
14. In Materials and Methods: “Within BrICS, sMRI… a Cho/NAA ≥ 5X threshold”: (i) Why were sMRI images registered to both T1W-CE and T2W-FLAIR images? Surely, the resultant registered images will differ. Which one was used? (ii) Why was 5X used as threshold and not 2X? This is confusing.
15. In Materials and Methods: “Clinical tumor volumes … to the GTVs”: How were the margins to be added decided? The range (5-10 mm) will have a significant impact on the resulting CTVs.
16. In Materials and Methods: “To accommodate microscopic… treatment volumes (PTV1 and PTV2)”: Once again, how were the margins to be added decided? Adding 3 mm will have different effect from adding 5 mm, thus having a significant impact on the size of the PTVs.
17. In Materials and Methods: “Focal radiation doses… 60 Gy to PTV2”: Was this done on each subject? If so then it is not clear how this was achieved. There has to be quite a bit of overlap between PTV1 and PTV2. How were different radiation doses given to overlapping PTVs?
18. In Materials and Methods: Line 191: The sMRI lesion volumes were created with Cho/NAA ≥ 2x. Please explain how these were verified.
19. In Materials and Methods: “We then calculated… GTV2 or CTV2”: In line 171-172 PTV was mentioned to be the target volume. Was the target GTV or CTV or PTV?
20. In Results: Figure 2: “The 3D contour of Cho/NAA ≥ 5X is shown”: What about 2X? Which one is more important?
21. In Results: Line 258: How exactly (quantitatively) were high- and low-mismatch groups identified?
22. In Results: The high-mismatch group is supposed to larger differences have between Cho/NAA ≥ 2x and T1w-CE. However, this is not supported by the volumes reported in lines 283-285 for pre-RT conditions.
23. In Results: Figure 5: There are disagreements in the mentions of CTV and GTV in the figure and figure caption.
24. Part of the Discussion is repetitive (e.g. line 359).
25. In Discussion: “While the difference in… is clinically 380 significant”: While OS difference between the 2 groups was not significant (what was the p-value?), the median values were also not significant (p=0.07). Both these dampen the enthusiasm from the results’ perspective.
26. In Discussion: “The probability that recurrence… overlap percentages decrease”: If the probability of recurrence closer to the treatment volume is indeed low, then it is very possible that it is the only reason for lower overlap between pre-RT GTV2 and rCE volumes. In other words, the pre-RT GTV, as detected by T1W-CE receives the radiation and gets treated and the recurrence occurs at a different location causing the low overlap. On the other hand, it is possible that the volume that could have been detected by sMRI remains untreated and the recurrence happens in that area, resulting in higher overlap. This would suggest that perhaps sMRI is a biomarker for future recurrence, as mentioned in the Discussion. However, these results do not convincingly establish that sMRI can detect the tumor better for RT (for the reasons given at the beginning of this paragraph).
27. There is an inherent anomaly in the discussion presented in the manuscript. It suggests that sMRI is a better indicator of tumor volume than T1W-CE or T2W-FLAIR imaging. However, part of the reasoning is based upon overlap with a 2nd T1W-CE during progression. What about a 2nd sMRI? One needs to see the overlap of the 2 sMRIs also. However, it is very possible that sMRI is a better biomarker for future recurrence, which is not the main point of this manuscript.
In summary, the work can be a big step towards establishing sMRI as a meaningful tool in RT-guided treatment of glioblastoma but many questions and concerns remain.
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsComments for authors
Tomography – Manuscript ID: tomography-2788679 – “The Utility of Spectroscopic MRI in Stereotactic Biopsy and Radiotherapy Guidance in Newly Diagnosed Glioblastoma.”, by Abinand Rejimon, Karthik K. Ramesh, Anuradha G. Trivedi, Vicki Huang, Eduard Schreibmann, Brent D. Weinberg, Lawrence R. Kleinberg, Hui-Kuo G. Shu, Hyunsuk Shim and Jeffrey J. Olson.
In this manuscript, the authors analyze the value of the proton magnetic resonance spectroscopic imaging (MRSI) in glioma diagnostics and therapeutics planning. In a first part, they demonstrated clinical translatability by showing a biopsy case of a lower grade glioma patient. The biopsy target was delineated by MRSI. Then, they conducted a secondary analysis on 22 newly diagnosed glioblastoma (GBM) patients treated with belinostat. They investigated the relationship between under-radiated tumor areas identified by MRSI and overall survival. Their results revealed that patients with a lower volume of undertreated tumor detected by MRSI had significantly better survival outcomes, highlighting the potential benefits of integrating metabolite information from MRSI with treatment planning. Finally, they established the utility of MRSI in treating areas of future recurrence in patients with GBM. Then, they highlight the potential of advanced imaging techniques in improving the diagnostic and treatment strategies for GBM.
The study, albeit partially confirmatory, is original and potentially relevant as it investigates the contribution of the MRSI in the characterization of the spatial distribution of endogenous metabolites and its benefit in glioma grading and differentiating tumor from non-tumor.
Methods are more or less relevant. They aren’t major flaws or biases and conclusions are based on the data. The literature is up-to-date.
Authors report precisely author contributions, ethics board approval, disclosure of funding and conflicts of interest. There isn’t reason to suspect research misconduct.
Finally, the manuscript is well written and clearly presented. Here are some points the authors could consider to increase the impact of their work and relate it more closely to the published literature.
My main criticisms of this paper concern the MRSI method and some results (spectra in tumor and in non-tumor tissue for example, etc.) that should be presented to give more impact to the paper.
Thus, the authors must report precisely the MRSI method used. How did they overcome the challenges that make the acquisition of a perfect spectral data set difficult, including the homogeneity of both B0 and the RF field (B1), increased SAR with 3T magnetic field and thus limited B1 amplitudes, the intrinsically shorter T2 relaxation times, the stronger chemical shift displacement errors and lipid artifacts, the typically long measurement times, the need for an efficient water suppression module, and the need for automatic processing and quality assurance tools due to the usually higher resolutions?
Traditional MRSI uses phase encoding, in addition to volume preselection (e.g. PRESS, STEAM, (semi)LASER) to provide localization of each spectrum. Authors must specify their spectroscopic imaging method.
How did they deal with all the problems mentioned above? In particular, they must pay great attention to the management of chemical shift artifacts. Due to the fact that in a magnetic gradient field, spatial locations correspond to resonance frequencies, the images of metabolites with different resonance frequencies are spatially shifted. This phenomenon is referred to as chemical shift displacement error and it presents a challenge in magnetic resonance spectroscopy. The magnitude of the chemical shift displacement error increases with B0 due to the proportional frequency dispersion. How authors overcome this problem to provide accurate metabolites mapping?
How are the spectra analyzed? What do the Cho/NAA ratios represent? Is it the area under the curve, the amplitude of the signal? How is the contribution of macromolecules and lipids managed (especially at short TE)?
All of these points must be clarified and discussed before proposing spectroscopic imaging as a tool to improve the diagnostic and treatment strategies for GBM.
Classically, spectroscopic MRI is noted magnetic resonance spectroscopic imaging (MRSI), then “sMRI” should be “MRSI” throughout the manuscript.
It seems to me absolutely necessary to take these points into account before considering publishing these results.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsPlease see attached.
Comments for author File: Comments.pdf
Author Response
please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsComments for authors
Tomography – Manuscript ID: tomography-2788679 – V2 – “The Utility of Spectroscopic MRI in Stereotactic Biopsy and Radiotherapy Guidance in Newly Diagnosed Glioblastoma.”, by Abinand Rejimon, Karthik K. Ramesh, Anuradha G. Trivedi, Vicki Huang, Eduard Schreibmann, Brent D. Weinberg, Lawrence R. Kleinberg, Hui-Kuo G. Shu, Hyunsuk Shim and Jeffrey J. Olson.
The authors submitted a carefully-prepared revision, which satisfactorily addressed the remaining concerns.
Therefore, I recommend publication.
