Imaging Spectrum of the Developing Glioblastoma: A Cross-Sectional Observation Study

Round 1

Reviewer 1 Report

I congratulate with Authors for their interesting and well written work. The birth and growth of glioblastoma is an intriguing process, as we often lack longitudinal studies. Indeed, the content of the manuscript reflects my personal experience of those rare cases where a scan is acquired at early stages: a vague cortical T2 hyperintensity is probably the first alteration in rising GBM. Usually, the scan is performed after a seizure, which may reflect early cortical involvement in these cases.

However, direct cortical involvement is uncommon to most “typical” GBMs, which are most likely diagnosed at later stage. Probably, the sequence of radiological alterations is similar in those, more frequent, cases where GBM is primary seen in subcortical white matter which may be less eloquent to patients until a significant volume is reached and a typical GBM is already seen at MRI. Nonetheless, we cannot exclude that the imaging pattern described in this manuscript is specific to those GBM which arise in the cortex or in close proximity to it. I wish Authors could discuss this point in the manuscript.

Other minor points:

1)      Please better define the definition of solid enhancement and of nodular enhancement: as far as I understood Authors use nodular when a smaller “dot” of enhancement is seen. However, I would clarify the terms in the Methods section.

2)      Figure 4, I do not know if the figures that were attached to the version of manuscript for revision are at lower quality, in which case I ask mercy to Authors. However, “my” Figure 4 was a low quality one, which prevented to zoom and have a closer look to the signal alterations we are looking for.

Moreover, I could never really appreciate the suggested pattern T2 hyper towards T2 iso-density, as a quite discrete tumor was already evident at follow-up scans in the panel. Could you please identify a better case to fully appreciate these changes?

3)      Table 2: Oedema rows are misaligned, please correct.

Author Response

Comments and Suggestions for Authors

I congratulate with Authors for their interesting and well written work. The birth and growth of glioblastoma is an intriguing process, as we often lack longitudinal studies. Indeed, the content of the manuscript reflects my personal experience of those rare cases where a scan is acquired at early stages: a vague cortical T2 hyperintensity is probably the first alteration in rising GBM. Usually, the scan is performed after a seizure, which may reflect early cortical involvement in these cases.

The authors thank the reviewer for taking the time to read and critique the manuscript.

However, direct cortical involvement is uncommon to most “typical” GBMs, which are most likely diagnosed at later stage. Probably, the sequence of radiological alterations is similar in those, more frequent, cases where GBM is primary seen in subcortical white matter which may be less eloquent to patients until a significant volume is reached and a typical GBM is already seen at MRI. Nonetheless, we cannot exclude that the imaging pattern described in this manuscript is specific to those GBM which arise in the cortex or in close proximity to it. I wish Authors could discuss this point in the manuscript.

The following paragraph has been added to the discussion:

The majority of the 84 GBM lesions in this study originated from the cerebral cortex. Cortical origin of GBM is consistent with a murine model which used fluorescent labelling with single-cell resolution to track tumorigenesis; this model provided strong evidence for GBM formation in grey matter [32]. Most patients in this cohort presented with seizure presumably secondary to neoplastic infiltration into the cortex. Other patients presented with focal neurologic deficits with the tumour developing in eloquent areas of the brain [33]. However, the current research has also shown that cortical origin is not exclusive to GBM. GBMs have also been reported to originate and recur in the subventricular zone, and it is possible that tumours arising in the cortex are due to secondary outward migration of abnormal brain tumour cells [34, 35], providing an explanation for lesions that develop in deep and subcortical white matter.  Indeed, an interesting observation in this current study was the development of GBM at the olfactory bulb in four patients that had received prior radiotherapy for an earlier remote GBM. Human GBM cells injected into immunodeficient nude mice have been shown to preferentially migrate to olfactory bulbs [36]. Additionally, olfactory bulbs have been implicated as a radioresistant niche for GBM cells [37]. Conversely, olfactory bulbs are also a rich source of neural stem cells [38], and the generation of a new, de novo GBM from local cellular mutations, rather than a result of cell migration and proliferation from a primary lesion, cannot be discounted.

Other minor points:

• Please better define the definition of solid enhancement and of nodular enhancement: as far as I understood Authors use nodular when a smaller “dot” of enhancement is seen. However, I would clarify the terms in the Methods section.

This has been defined in the methods section as suggested. Section 2.2.2. (number 4)

2)      Figure 4, I do not know if the figures that were attached to the version of manuscript for revision are at lower quality, in which case I ask mercy to Authors. However, “my” Figure 4 was a low quality one, which prevented to zoom and have a closer look to the signal alterations we are looking for.

Moreover, I could never really appreciate the suggested pattern T2 hyper towards T2 iso-density, as a quite discrete tumor was already evident at follow-up scans in the panel. Could you please identify a better case to fully appreciate these changes?

A new Figure 4 with 300 dpi has been uploaded. All images are now at least 300 dpi.

A new box has been inserted into Figure 2 containing enlarged images to better demonstrate the tumour containing T2 signal that is isointense to grey matter (T2-iso).

• Table 2: Oedema rows are misaligned, please correct.

Corrected.

Reviewer 2 Report

Overall, the idea of manuscript is average, furthermore, the contribution is not adequate at the moment. The manuscript needs significant work.  

1.      What are the limitations of the related works?

2.      Are there any limitations of this carried out study?

3.      How to select and optimize    the user-defined parameters in the proposed model?

4.      There are quite a few abbreviations are used in the manuscript. It is suggested to use a table to host all the frequently used abbreviations with their descriptions to improve the readability

5.      Explain the evaluation metrics and justify why those evaluation metrics are used?

6.      Some sentences are too long to follow, it is suggested that to break them down into short but meaningful ones to make the manuscript readable.  

7.      The title is pretty deceptive and does not address the problem completely.

8.      The related works section is very short and no benefits from it. I suggest increasing the number of studies and add a new discussion there to show the advantage. 

9.   Use Anova test to record the significant difference between performance of the proposed and existing methods.

The paper needs minor revision.

Author Response

The authors thank the reviewer for taking the time to read the manuscript.

1. What are the limitations of the related works? These are stipulated within the discussion (Section 4)

2. Are there any limitations of this carried out study? These are stipulated at the end of the discussion (Section 4)

3. How to select and optimize the user-defined parameters in the proposed model? No model has been proposed. 

4. There are quite a few abbreviations are used in the manuscript. It is suggested to use a table to host all the frequently used abbreviations with their descriptions to improve the readability The authors are happy to add a list of abbreviations but add that all abbreviations are explained in the text.

5. Explain the evaluation metrics and justify why those evaluation metrics are used? This is an observational study of routine clinical imaging with little evaluation metrics. Statistics has been provided to compare patient cohorts and these have been stated in the methods section. 

6. Some sentences are too long to follow, it is suggested that to break them down into short but meaningful ones to make the manuscript readable.  The manuscript has been re-read and edited. 

7. The title is pretty deceptive and does not address the problem completely. The title describes the study but has been modified to take into account the comments from reviewer 4.

8. The related works section is very short and no benefits from it. I suggest increasing the number of studies and add a new discussion there to show the advantage.  There is little in the literature from which to compare to the current work. Studies available have been addressed and the discussion has been lengthened following peer review.

9. Use Anova test to record the significant difference between performance of the proposed and existing methods. The authors have used statistically appropriate methods. 

Reviewer 3 Report

Fascinating work. Very pleasant to read, and with an adequate and recent literature review supported by 38 references

Congratulations to the authors.

We suggest, however, some aspects that could be improved...

- In the abstract, it lacks possible future work taking into account the results of the work presented,

- At the end of the "Introduction" chapter (line 79), they should mention the structure of the document as well as a brief description of what is covered in each of the article sections,

- Regarding the Imaging review, the computed tomography studies protocols need to be clarified.

- The work would gain a lot if more neuroradiologists were part of the  MRI and CT studies review team.

- Images have very long captions. We advise that part of the description of imaging findings appear in the text and not in the captions,

- The text size of the tables must be smaller than the text of the document,

-Many results are presented in the "results" section that later does not appear as relevant in the conclusion. The conclusion section is very summarized and without the perspective of the pertinence of carrying out future work

- At the end of the document it is not very clear the contribution of the different authors

Good job

Author Response

Comments and Suggestions for Authors

Fascinating work. Very pleasant to read, and with an adequate and recent literature review supported by 38 references

Congratulations to the authors.

We suggest, however, some aspects that could be improved...

The authors thank the reviewer for taking the time to read and critique the manuscript.

1. In the abstract, it lacks possible future work taking into account the results of the work presented.

The following sentences have been added to the end of the abstract.

Red flags for non-TRA GBM lesions are cortical/subcortical CT hyperdense/T2 iso-intense/low ADC. Future research correlating this imaging spectrum with pathophysiology may provide insight into GBM growth patterns.

2. At the end of the "Introduction" chapter (line 79), they should mention the structure of the document as well as a brief description of what is covered in each of the article sections

The authors are happy to add this but would like to defer to the editor as such a paragraph is not stated as necessary in the instructions for authors.

3. Regarding the Imaging review, the computed tomography studies protocols need to be clarified.

The following sentence has been added to methods (section 2.2)

Archetypally, pre-operative imaging was performed at the presenting hospital, including the tertiary referral centre. Consequently, pre-operative CT and MR imaging protocols varied with the imaging centre. CT was typically performed using Siemens and GE Medical systems machines with a range of KV 120-140.

4. The work would gain a lot if more neuroradiologists were part of the MRI and CT studies review team.

The authors confirm that neuroradiologists were part of the imaging review team. This is stated in the methods section.

5. Images have very long captions. We advise that part of the description of imaging findings appear in the text and not in the captions

The authors would like to defer to the editor for advice. The authors feel that the Figure legends are comprehensive and allow for the Figures to stand alone from the main manuscript – as suggested in the instructions for authors.

6. The text size of the tables must be smaller than the text of the document

Table text reduced as suggested 

7. Many results are presented in the "results" section that later does not appear as relevant in the conclusion. The conclusion section is very summarized and without the perspective of the pertinence of carrying out future work

The authors thank the reviewer and have altered the conclusion to the following (red new text):

Developing GBM appear to display a spectrum of imaging features moving through phases of T2 hyperintensity to CT hyperdensity, T2-iso, reduced diffusivity and variable enhancement before showing TRA GBM. The mean time of progression to TRA GBM from T2 hyperintensity and from CT hyperdensity being 140 days and 69 days respectively. A CT cortical/subcortical hyperdense lesion that shows concomitant T2 iso and reduced diffusivity, irrespective of the presence or absence of enhancement, should be red flags for non-typical GBM and trigger urgent diagnosis and treatment. Future research correlating this imaging spectrum with pathophysiology may provide insight into GBM growth, particularly examining for possible differences that may exist in molecular biology between the features described in this cohort and those lesions of typical radiological appearances of GBM.

8. At the end of the document it is not very clear the contribution of the different authors

Bold type set has been added to categories of author contribution adding clarity.

Good job

Reviewer 4 Report

The manuscript deals with the probability of detecting changes in MRI images that may warn of the incipient development of a GBM even before the appearance of the Typical Radiological Appearance (TRA).

In the Materials and Methods section, one aspect should be clarified: if a single evaluation has been made for each patient, this is not a longitudinal study. In other words, if no follow-up has been performed (either prospectively or retrospectively), it is a cross-sectional study, even if the data collection took eight years. This is the typical example of making an assessment (on one occasion) of each patient, or making two or more assessments; in the first case we say that a picture has been taken (= cross-sectional study) and in the second a film has been made (longitudinal study, because there is follow-up). As the data are presented, it seems that there is only one check-up for each patient, so the study should be presented as a cross-sectional study and not as a retrospective longitudinal study.

Sections 2.2.1 and 2.2.2 are called the same. It would be convenient to give them different names depending on what is being assessed. It seems that in the first section they refer to the already standardized TRA and in the second to the incipient imaging patterns they intend to identify. The first heading could be referred to as Imaging Categorization and the second as Early Image Profiles or Incipient Image Patterns.

Based on Tables 1 and 2, perhaps the authors have not taken full advantage of the information available to them. As a recommendation, we suggest that they could make Kaplan-Meier survival curves according to the categories or profiles of early images in section 2.2.2. They could also make comparisons between these categories and the frequencies in each category in terms of the presence of headache, neurological deficit, epileptic seizures and cerebral hemorrhage.

In the discussion, it would be advisable to comment that another limitation of the study is that the images were only evaluated by a single observer. In the event of any personal bias (due to the subjectivity of a visual observation), this could lead to a systematic error. It would have been appropriate for more than one expert radiologist to have assessed the images and for the inter-observer Kappa index to show a high degree of agreement (close to 1).

Author Response

The authors thank the reviewer for taking the time to review the manuscript. 

The manuscript deals with the probability of detecting changes in MRI images that may warn of the incipient development of a GBM even before the appearance of the Typical Radiological Appearance (TRA).

In the Materials and Methods section, one aspect should be clarified: if a single evaluation has been made for each patient, this is not a longitudinal study. In other words, if no follow-up has been performed (either prospectively or retrospectively), it is a cross-sectional study, even if the data collection took eight years. This is the typical example of making an assessment (on one occasion) of each patient, or making two or more assessments; in the first case we say that a picture has been taken (= cross-sectional study) and in the second a film has been made (longitudinal study, because there is follow-up). As the data are presented, it seems that there is only one check-up for each patient, so the study should be presented as a cross-sectional study and not as a retrospective longitudinal study.

The authors thank the reviewer for the clarification. The term longitudinal has been removed from the manuscript, in favour or cross-sectional study.

Sections 2.2.1 and 2.2.2 are called the same. It would be convenient to give them different names depending on what is being assessed. It seems that in the first section they refer to the already standardized TRA and in the second to the incipient imaging patterns they intend to identify. The first heading could be referred to as Imaging Categorization and the second as Early Image Profiles or Incipient Image Patterns.

Thank you. The subheadings have been changed as suggested.

Based on Tables 1 and 2, perhaps the authors have not taken full advantage of the information available to them. As a recommendation, we suggest that they could make Kaplan-Meier survival curves according to the categories or profiles of early images in section 2.2.2. They could also make comparisons between these categories and the frequencies in each category in terms of the presence of headache, neurological deficit, epileptic seizures and cerebral hemorrhage.

            The authors did not want to over-reach with the statistical analysis in this observational study. It is with respect that the authors feel that the appropriate level of statistical analysis has been performed at this level but we accept that statistical analysis on future prospective data using the suggested methods will be warranted.

In the discussion, it would be advisable to comment that another limitation of the study is that the images were only evaluated by a single observer. In the event of any personal bias (due to the subjectivity of a visual observation), this could lead to a systematic error. It would have been appropriate for more than one expert radiologist to have assessed the images and for the inter-observer Kappa index to show a high degree of agreement (close to 1).

            Agreed – the limitation has been added to the discussion.

Round 2

Reviewer 2 Report

The paper can be accepted.

The paper can be accepted.