Primary and Metastatic Brain Tumours Assessed with the Brain and Torso [18F]FDG PET/CT Study Protocol—10 Years of Single-Institutional Experiences
, Andrzej Marszałek 3, Tomasz Piotrowski 1,4, Adrianna Medak 1, Katarzyna Pietrasz 2, Julia Wojtowicz 5, Hubert Szweda 4, Krzysztof Matuszewski 4 and Witold Cholewiński 1,2Round 2
Reviewer 1 Report
Response for the Reviewers
Dear authors,
thank you very much for this request and opportunity to review an article for your journal.
I think that this paper is very interesting and the topic is surely original. The idea to analyze the main international guidelines and compered these guidelines with the clinical experience derived from many papers and our experience is original and interesting. The type of acquisition (torso or whole body) in the clinical practice may be a dilemma for the nuclear medicine physicians and different point of view are present in literature.
Answer: We would like to sincerely thank the Reviewer #1 for a kind review.
However the methods are not clear, in particular the inclusion criteria need to be clarify both for articles and patients. Without these findings it is not so easy to correctly evaluate the results of this review. The discussion is very poor without specific explanation of the main results. The reason of the inclusion of the literature articles is not clear.
Answer: We would like to thank for the valuable remark. We improved the methodology and discussed the inclusion/exclusion criteria, using the flow chart as probably clearer. It is important for us to make our study available and understandable to all readers (pages: 2-5). We extended and significantly adjusted the method section. We extended discussion section (pages: 10-11).
Moreover, the conclusion are not supported by your data (0.01% of brain lesion detected with brain+dorso scan doesn't seem a great result).
Answer: Thank you for this highly valuable observation. We excluded from the final analysis all patients in whom the brain tumour has been previously mentioned/detected using other imaging methods or before the patient has been treated/managed in our institution. According with the patients’ medical record, we found all brain lesions that have been detected in this group. There are none other reports than mentioned in this study. We made sure to be as transparent and sincere in our analysis as it can be. We did not want to manipulate the data or to multiply the results. The study aimed to show that despite the method’s limitation, it is possible to detect clinically silent, unknown brain tumours. As previously mentioned, many nuclear medicine departments do not have access to other radiotracers than 18F-FDG and need to manage their patients. However, many of those departments exclude the brain region from the scanning protocol which decrease the possibility to evaluate those lesions and therefore – to administer the appropriate treatment. Nevertheless, accordingly with this remark, we adjusted our article, addressed above-mentioned issues. We definitely needed to explain the aims of the research using more details. If we mentioned all brain lesions detected, we would achieve nearly 1% (as epidemiological data). However, despite decreasing our own results value, we kept the results trustworthy. We were interested in describing ONLY brain tumours using the 18F-FDG PET/CT method and ONLY newly diagnosed. If the patient’s medical records mentioned some sort of brain lesion in the past, brain and torso scanning would be most likely performed in this particular subject in every institution, while brain region inclusion may be always considered as useful. Our study aimed to show and encourage to use the full scanning protocol in every condition in case the patient suffered from brain tumour occurrence. After reading this valuable remark we decided to adjust our manuscript and insert this explanation in the discussion section as well as in the introduction and methods. Moreover, the brain lesions occurrence in our database did not exceed 1% of all annually examined patients. However, after excluding patients in whom the brain lesion has been reported (mentioned in the medical records), has been followed by some symptoms (not clinically silent), and finally – transferred to another institution for the further management, we obtained lower results but ensured the data curation and reliability of the study.
We clarified the literature inclusion. We would like to use the official guidelines and reports rather than other studies. However, we included additionally researches describing good practice and stand-alone studies. We slightly extended the references section (suggested by the Reviewer #1, study protocol, statistical analysis).
METHODS
- The inclusion criteria of the articles are lacking. Which studies do you select? Studies with clinical suspect of brain lesions? Studies with "whole body" scan without any suspect? mixed? Please add some paragraphs with Search strategy, Study selection and Data abstraction. Besides, I don't understand why do you include external articles if nothing about them is described in the manuscript?
Answer: Thank you for the remark. We have adjusted inclusion/exclusion criteria, explained the study selection. We used all available database following to aim of the study. We used the literature to provide the data regarding the role of the 18F-FDG PET/CT studies in brain tumours diagnosis and compared guidelines/recommendations as well as other sources confirming the utilities of the study in brain region imaging. We described the possible limitations and used the terminology, proposed by the international societies. We are aware that the “whole body” scanning is more often used than a specific terminology. However, mentioning recommendations or describing the conclusions provided by guidelines, we decided to follow the nomenclature proposed by EANM to avoid confusion. In every nuclear medicine departments, the term “whole body” can be understood differently. External articles have been used to explain epidemiology, background of the study, the use of other than 18F-FDG radiopharmaceuticals, brain tumours characteristics, and others (pages:
- The inclusion criteria of your patients are lacking. What do you mean with "similar protocol"? Why did you performed torso and brain scan without clinical indications as you described?
Answer: The brain and torso protocol is a standardly performed acquisition protocol in our institution. After discovering clinically silent brain tumours, it became a daily routine. We believe that detecting even a few can be lifesaving for our patients and because of the limited access to other radiotracers, we are trying to make the available tools as best and as useful as we possibly can. We believe that there are multiple departments in similar or comparable position. However, our daily practice aims to help our patients, so we are obliged to improve the existing solutions to support patients’ management as best as we can. Following the Reviewer’s #1 remark, we explained more precisely patients inclusion criteria. We used a flow chart and text description. By “similar” study protocol, we meant single-phase brain and torso 18F-FDG PET/CT study. We improved the methods section (pages: 2-5).
- A table with the resume of the main technical features (type of scanner, uptake time, CT features, activities injected, and so on) of 24 article included is lacking. Besides, you should cite the articles that you used.
Answer: Thank you for the remark. We recorrected above-mentioned issue.
- "We chose the most often mentioned in the medical 113 literature statistical tests to obtain results and to provide conclusions" is too vague. Please clarify. Maybe, you should move paragraph 3.1 in the methods section.
Answer: Thank you for the remark, we adjusted this section and shifted it to Methods section.
- I suggest to add a table to resume the Results section to help the readers.
Answer: Thank you for suggestion, we added Table 2. to the manuscript.
- I suggest to explain better in the discussion the findings of this work and less the general characteristics of brain tumors....
Answer: We would like to thank Reviewer #1 for the suggestion, we improved the Discussion section (reordered, significantly extended).
- Nothing about the relationship between size of brain lesion and PET/CT is reported. This is a crucial point. We know that the resolution power of PET scanner is about 5 mm. Moreover, we have the partial volume effect. So the detection of brain lesions and the SUV values are influenced intrinsically by these two variables. write something about this point in the discusson section.
Answer: Thank you for the observation. The study was not constructed to evaluate FDG-PET as the method for brain lesion detection – the resolution of PET scanners and biology of brain lesion make FDG-PET/CT far behind the MRI and ceCT. So we are aware that the real number of brain lesion was most likely much higher than detected. The main conclusion of collected data analysis was that FDG-PET can detect important number of silent brain lesions even if in other scenarios this method is not recommended for brain tumor diagnosis.
- The reference 24 used as reference for suggesting SUV of 2.5 is not shareable. I suggest to delete this part. There is a strong debate in literature about the potential best cutoff of SUV and SUV is a factor affected by many parameters (body weight, technical features, extravasation, blood glucose level).
I believe excessive this point.
Answer: We followed the Reviewer #1 suggestion and removed paragraph and reference.
- Why do you write that your data support a change in "therapeutic management "? in you paper you didn't analysed the change in treatment or outcome in the patients.
Answer: Thank you for the observation. We decided to remove those elements as too brief. During the article’s revision we found that some phrases may confuse the reader. Describing the impact of the study on the therapeutic protocol demands much more space than a couple of sentences. Following this observation, we limited the volume of data to study objectives. According to available medical records, patients underwent treatment. In some of the cases, the clinical indication for the PET/Ct study performance was to assess the recurrence of previously treated disease. In some (presented in the study) we found no recurrence but brain primary or no local recurrence detectable but secondary tumour. Considering the differences between the sample, there is no easy way to present the exact and precise impact of the discovery on therapeutic management. Thus, we decided to remove those elements.
- You founded 0.01% of brain lesion among 14222 patients analysed. With these results underline the usefulness of brain scan is not shareable...
Answer: Thank you for the remark, we explained this issue considering previous remarks, we adjusted the discussion section to explain it to the readers.
REFERENCES
- You correctly spoke about not-FDG tracers for the study of brain, but I believe that you could add some more references about the usefulness of FDG in some brain cancers. Moreover, a recent review about different tracers is available and to cite: doi: 10.3390/ijms20194669.
Answer: We sincerely thank for the remarks. We adjusted the reference section and used the literature suggested by the Reviewer #1. Finally, we would like to sincerely thank for the extended and detailed review. The revision significantly contributed to the quality of the final version of the manuscript.
Reviewer 2 Report
Response for the Reviewers
Authors: We would like to sincerely thank the Reviewers for specific, detailed reviews. We tried to implement as many interesting and valuable data as possible. However, we understand the doubts raised by the Reviewers. Thus, all remarks have been analyzed, and respected. Once again, we would like to thank the Reviewers for the time and efforts, which significantly contributed to the quality of the paper.
After conducting this study nearly two years ago, we were able to collect all necessary data to present this study. After the detailed and thorough review, we substantially improved our manuscript. We presented data precisely, followed all Reviewers’ remarks. We improved language of the article. We inserted additional figures, presenting the data collection schemes. We addressed all remarks. We hope that the Reviewers’ will find the research eligible for publication.
Reviewer #2
This 10-years analysis is valuable to the field.
Answer: We would like to sincerely thank the Reviewer #2 for the kind opinion, and highly valuable remarks.
The major concerns of this analysis are
It is still not clear when FDG-PET brain scan should be indicated or to be included along with the torso scan, for metastases?
Answer: We would like to sincerely thank for the question. It is remains unclear. Authors still discuss the utilities of the the FDG-PET in brain region evaluation. We can observe a great debate in our country where brain and torso protocol is not used at all. We observed the discussion in many different countries/departments. The most common issue is lack of the report comparing the guidelines and authors’ perspective, supported by original studies. We decided to prepare the study (presented in this manuscript) and we spent nearly two years to prepare the dataset and collect all medical records + recommendations most suitable for this research. We have been keeping the database up-to-date, recorrected it several times. We made sure to use the official recommendations instead of published online materials with no DOI or any publishing coverage as much as it possible. We have been waiting with publishing materials to make sure we followed the newest materials published. We also considered the possibility that we will not submit our materials if the similar research will be published or the recommendations would be updated.
Primary brain tumors can also be non-FDG-avid. Without another tracer such as [C-11]methionine or [F-18]DOPA, how could this (cold spot) be certain to the primary brain tumor;
Answer: None of the tracers can certainly confirm that hot or cold spot is the primary brain tumor – this can be only done by the correlation of MRI and histology. Thus, in our opinion the added value of inclusion of brain area on the standard FDG-PET/CT protocol in oncology is the detection of abnormality in the brain which can be further evaluated by more specific methods.
The biggest concern is that among 14,222 patients in 10 years, how many patients with brain lesions are missed by FDG-PET given all the guidelines even though 155 were detected.
Answer: Thank you for the questions. First of all, we did not find any other mentioning about brain lesions in the medical records (which means that there were none or were not detected using other imaging tools before or after the PET/CT scanning). This does not give us certainty that we did not miss brain tumours. We are aware of the 18F-FDG PET limitations; however we believe that detecting brain lesions can be lifesaving, even when the cohort receiving help is not impressive. Nevertheless, we understand concern of the Reviewer #2, and we sincerely appreciate the opinion and doubts. We hope that after explanation and adjusting paper, we make the issue clearer.
Minor concerns are
- delayed scan was mentioned to differentiate inflammation, but no patient data/images were followed;
Answer: Thank you for the remark. We used the recommendations in which the possibility to perform delayed scanning is mentioned. However, we did not necessarily perform this protocol (delayed, dual-time-point) in those patients, because we extract the cohort in which we did not expected brain lesions. That is why the final result is 155. We evaluated brain tumours in more numerous group, however, the aim of this study was to show that the 18F-FDG PET study can be used to evaluate clinically silent brain tumours. Thus, to ensure the study reliability, we needed to exclude from the final results patients’ in whom: brain tumour has been detected/reported in other studies before the 18F-FDG PET study, in the medical records (in the past) or patients who were transferred to another institution (and we have no certainty that it was indeed the brain tumour).
- many "silent" primary brain tumors were detected from patients with other primary cancers, or CUPs? How was that certain, by biopsy or post-surgical histology?
Answer: In most of the examined patients, due to post-surgical histology. In patients in whom the histological examination was not available, we used the post-PET MRI. We adjusted the Methods, Results, and Discussion section to make sure the information was included, and it is clear. Thank you for this remark, we would like to make sure that the study is clear for the Reader, and we understand the confusion, which definitely needed adjusting the content of the manuscript.
Reviewer Comments
The authors tuned down the significance of their findings. The added scans can be some values.
Reviewer 3 Report
Response for the Reviewers
Authors: We would like to sincerely thank the Reviewers for specific, detailed reviews. We tried to implement as many interesting and valuable data as possible. However, we understand the doubts raised by the Reviewers. Thus, all remarks have been analyzed, and respected. Once again, we would like to thank the Reviewers for the time and efforts, which significantly contributed to the quality of the paper.
After conducting this study nearly two years ago, we were able to collect all necessary data to present this study. After the detailed and thorough review, we substantially improved our manuscript. We presented data precisely, followed all Reviewers’ remarks. We improved language of the article. We inserted additional figures, presenting the data collection schemes. We addressed all remarks. We hope that the Reviewers’ will find the research eligible for publication.
Reviewer #3
The topic is of interest and the manuscript is well written.
Answer: We would like to sincerely thank the Reviewer #3 for this kind opinion, we truly appreciate it.
It is not fully clear the meaning of brain and torso PET. Were a Brain and torso scan performed additionally to a whole-body PET/CT evaluation in 14222 subjects?
Do the authors think that performing a scan starting from the vertex could lead to similar results?
Answer: Thank you for posing this question. In fact, we performed one scanning: top/apex/vertex – mid-thigh. It can be also recognized as “whole body”. Because we used the official recommendations, we used the terminology mentioned and established by those recommendations. We provided “whole body” term in the article to avoid further confusions.
Reviewer Comments
The authors have properly replied to the requested criticisms.
Reviewer 4 Report
Response for the Reviewers
From an epidemiological and methodological point of view (I'm not a nuclear physician), I have several critical questions for the Authors:
- it's not clear if Cochrane methodology for systematic review has been followed, can you confirm it? How have you checked the risk of bias? Otherwise, is it a "narrative review"?
11- having mixed in an unique pot a review section and an experimental one is totally unacceptable from a Cochrane methodology
Answer: We would like to thank the Reviewer #4 for the remark. At first, the study has been constructed using the “original article” style. After consultations with the Editorial Office, it has been remodeled to the current style. We respected the expertise of the Editorial Office. Moreover, the 18F-FDG PET/CT study can be performed as: whole body (with or without brain or head region), total body, etc. The brain and torso 18F-FDG PET/CT protocol, mentioned in the EANM official guidelines seems to be an important issue, even after widely discussed limitations of the method in the brain tumours diagnosis. We need to be aware than many nuclear medicine departments do not have access to any other radiopharmaceutical than 18F-FDG and still be in a need to help their patients to recover. We analyzed all available recommendations and guidelines, mentioning brain 18F-FDG PET/CT scanning with no bias. We mentioned all possible and mentioned in the literature limitations of the study considering brain lesions’ evaluation and presented our own experiences. Accordingly with the global epidemiology, the brain tumours are very rare. To avoid any bias, we excluded all patients in whom the brain tumour has been suspected, reported in the past, clinically expected (i.e. neurological symptoms present). We were aware that exclusion criteria decrease the number of patients in whom our study was successful. However, we are also aware that detecting brain lesion demands therapeutic intervention and can be lifesaving. Nevertheless, some of nuclear medicine departments exclude brain region from the acquisition protocol, and do not have access to other radiopharmaceuticals, which significantly limit the possibility that patients’ receive help (it is especially important in those patients’ in whom the brain tumour was the only primary or metastatic malignant lesion). Authors still discuss the utilities of the FDG-PET in brain region evaluation. We can observe a great debate in our country where brain and torso protocol is not used at all. We observed the discussion in many different countries/departments. The most common issue is lack of the report comparing the guidelines and authors’ perspective, supported by original studies.
2- which search algorithm have you employed? Using MESH or not? If yes, which ones?
Answer: We used mainly MeSH (as well as JHP) and searched all available via NCBI, NLM databases. However, we achieved multiple results, so we needed to follow the same path as preparing the original database: manual. We needed to establish inclusion/exclusion criteria and check the positions (despite the keywords builder, we still received many positions which were not eligible for our study). We needed to collect them to make sure we have all necessary data to conduct the study. We decided to prepare the study (presented in this manuscript) and we spent nearly two years to prepare the dataset and collect all medical records + recommendations most suitable for this research. We made sure to use the official recommendations instead of published online materials with no DOI or any publishing coverage as much as it possible. We have been waiting with publishing materials to make sure we followed the newest materials published. We also considered the possibility that we will not submit our materials if the similar research will be published or the recommendations would be updated. After implementing the exclusion criteria for the literature search, we established the final list of references, updated after the review of our study. We used also the “external”, not considered recommendations materials to collect and present other data (i.e. epidemiology, tracers used for brain tumours diagnosis) to provide as many interesting and useful data to Readers as possible. To summarize, despite using specific descriptors, we needed to significantly filtered literature search results and spent nearly two years under building both literature and original databases.
3- "We evaluated 14,222 [18F]FDG PET/CT datasets" I'm unable to understand if you analyzed papers or datasets or patients? How have you obtained such an incredible number of datasets? Are you speaking about datasets or patients? maybe the latters
Answer: We would like to thank the Reviewer#4 for the observation. The database has been built in 10 years (2010-20, pages 3-4). In our institution, we perform many 18F-FDG PET/CT scans (15 556, excluded 1334 – page 4). We used term study-patient-dataset to avoid constant repetition of the same term. Each PET/CT study provides approximately 600 images (slices) of both CT (app. 250-300) and PET (app.250-300). That is why the term “dataset” is accurate. To evaluate each lesion, the clinician needs to assess multiple images.
4- "we performed a comprehensive analysis of the international societies’ recommendations", either have you analyzed already published reccomandations or datasets? coming from? mixing them? how?
10- "conclusion" I'm still unable to understand why and how a narrative review has been coupled to a so-called experimental section. Suming up official guidelines and personal experience does not add anything to the conclusion, which is quite arbitrary
11- having mixed in an unique pot a review section and an experimental one is totally unacceptable from a Cochrane methodology
Answer: Thank you for this remark. We used the recommendations to describe the utilities and limitations of the 18F-FDG PET/CT study in brain tumours diagnosis, to describe the “brain and torso” protocol (only mentioned in the recommendations). We used other sources to describe: the radiopharmaceuticals recommended for the brain tumours diagnosis (not always mentioned in the recommendations, especially describing only the 18F-FDG PET study). Datasets were original, single-institutional, unsponsored studies.
Before we agreed to remodel our article, we made an appropriate literature search and we saw Cochrane methodology-based articles, including both literature and original database. I.e.:
10.1111/j.1464-410X.2010.10032.x, and others, including more specific original data. Nevertheless, we understand the remark and we hope that after improving the manuscript, it is acceptable as an evidence-based review.
5- "When performing the necessary statistical analyses" this is totally unacceptable, you must write down a full stats methodology section
6- "Shapiro–Wilk test’s results to measure the variables’ distribution" Shapiro test does another job...
7- "H0, Ha, respectively" both these hyphotesis are undefined
8- "3.1. Statistical analyses" this section belongs to M'M, not to results!!!!!
9- "Z – test’s results" to do what? to prove what?
Answer: Following the remark of Reviewer #4, we provided the necessary data (page 5).
We decided to delete some of the elements of the paper due to some confusions it might bring. The final sample-size is insufficient to perform the analysis based on correlations: gender-brain tumours occurrence. We did not perform it, however some statements confused the Reader. Thus, we reviewed content and removed those elements. We placed the statistical analyses section in the Methods section.
12- the suppodsed chenge of management will be applied to an extremely small subcohort of patients. Maybe that a better clinical result could be reached by a more specific radiopharmaceutical compound, rather than 18F-FDG. The conclusions are then unreliable
Answer: Despite the small subcohort of patients, we need to be aware that those patients received help. If not the brain and torso scanning, they would not receive treatment or received it much later. It is especially important in this group, were the brain lesion was the only metastatic lesion or the only primary (restaging patients or staging in patients diagnosed as suspected of the tumour localized in other location than brain; we included only those patients in whom the brain tumour was clinically silent to make the study reliable; performing this protocol in patients in whom the brain lesion has been reported previously or which brought symptoms would enlarge the cohort but then, study would be biased. Moreover, more specific radiopharmaceuticals are not always easily accessible.
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
Dear authors,
thank you very much for this request and opportunity to review an article for your journal.
I think that this paper is very interesting and the topic is surely original. The idea to analyze the main international guidelines and compered these guidelines with the clinical experience derived from many papers and our experience is original and interesting.
The type of acquisition (torso or whole body) in the clinical practice may be a dilemma for the nuclear medicine physicians and different point of view are present in literature.
However the methods are not clear, in particular the inclusion criteria need to be clarify both for articles and patients. Without these findings it is not so easy to correctly evaluate the results of this review. The discussion is very poor without specific explanation of the main results.
The reason of the inclusion of the literature articles is not clear.
Moreover, the conclusion are not supported by your data (0.01% of brain lesion detected with brain+dorso scan doesn't seem a great result).
METHODS
- The inclusion criteria of the articles are lacking. Which studies do you select? Studies with clinical suspect of brain lesions? Studies with "whole body" scan without any suspect? mixed? Please add some paragraphs with Search strategy, Study selection and Data abstraction.
Besides, I don't understand why do you include external articles if nothing about them is described in the manuscript?
- The inclusion criteria of your patients are lacking. What do you mean with "similar protocol"? Why did you performed torso and brain scan without clinical indications as you described?
- A table with the resume of the main technical features (type of scanner, uptake time, CT features, activities injected, and so on) of 24 article included is lacking. Besides, you should cite the articles that you used.
- "We chose the most often mentioned in the medical 113 literature statistical tests to obtain results and to provide conclusions" is too vague. Please clarify. Maybe, you should move paragraph 3.1 in the methods section.
- I suggest to add a table to resume the Results section to help the readers.
- I suggest to explain better in the discussion the findings of this work and less the general characteristics of brain tumors....
- Nothing about the relationship between size of brain lesion and PET/CT is reported. This is a crucial point. We know that the resolution power of PET scanner is about 5 mm. Moreover, we have the partial volume effect. So the detection of brain lesions and the SUV values are influenced intrinsically by these two variables. write something about this point in the discussion section.
- The reference 24 used as reference for suggesting SUV of 2.5 is not shareable. I suggest to delete this part. There is a strong debate in literature about the potential best cutoff of SUV and SUV is a factor affected by many parameters (body weight, technical features, extravasation, blood glucose level). I believe excessive this point.
- Why do you write that your data support a change in "therapeutic management "? in you paper you didn't analysed the change in treatment or outcome in the patients.
- You founded 0.01% of brain lesion among 14222 patients analysed. With these results underline the usefulness of brain scan is not shareable.....
REFERENCES
- You correctly spoke about not-FDG tracers for the study of brain, but I believe that you could add some more references about the usefulness of FDG in some brain cancers.
Moreover, a recent review about different tracers is available and to cite: doi: 10.3390/ijms20194669.
Reviewer 2 Report
This 10-years analysis is valuable to the field.
The major concerns of this analysis are
- It is still not clear when FDG-PET brain scan should be indicated or to be included along with the torso scan, for metastases?
- Primary brain tumors can also be non-FDG-avid. Without another tracer such as [C-11]methionine or [F-18]DOPA, how could this (cold spot) be certain to the primary brain tumor;
- The biggest concern is that among 14,222 patients in 10 years, how many patients with brain lesions are missed by FDG-PET given all the guidelines even though 155 were detected.
Minor concerns are
4. delayed scan was mentioned to differentiate inflammation, but no patient data/images were followed;
5. many "silent" primary brain tumors were detected from patients with other primary cancers, or CUPs? How was that certain, by biopsy or post-surgical histology?
Reviewer 3 Report
The topic is of interest and the manuscript is well written.
It is not fully clear the meaning of brain and torso PET. Were a Brain and torso scan performed additionally to a whole-body PET/CT evaluation in 14222 subjects?
Do the authors think that performing a scan starting from the vertex could lead to similar results?
Reviewer 4 Report
From an epidemiological and methodological point of view (I'm not a nuclear physician), I have several critical questions for the Authors:
1- it's not clear if Cochrane methodology for systematic review has been followed, can you confirm it? How have you checked the risk of bias? Otherwise, is it a "narrative review"?
2- which search algorithm have you employed? Using MESH or not? If yes, which ones?
3- "We evaluated 14,222 [18F]FDG PET/CT datasets" I'm unable to understand if you analyzed papers or datasets or patients? How have you obtained such an incredible number of datasets? Are you speaking about datasets or patients? maybe the latters
4- "we performed a comprehensive analysis of the international societies’
recommendations", either have you analyzed already published reccomandations or datasets? coming from? mixing them? how?
5- "When performing the necessary statistical analyses" this is totally unacceptable, you must write down a full stats methodology section
6- "Shapiro–Wilk test’s results to measure the variables’ distribution" Shapiro test does another job...
7- "H0, Ha, respectively" both these hyphotesis are undefined
8- "3.1. Statistical analyses" this section belongs to M'M, not to results!!!!!
9- "Z – test’s results" to do what? to prove what?
10- "conclusion" I'm still unable to understand why and how a narrative review has been coupled to a so-called experimental section. Suming up official guidelines and personal experience does not add anything to the conclusion, which is quite arbitrary
11- having mixed in an unique pot a review section and an experimental one is totally unacceptable from a Cochrane methodology
12- the suppodsed chenge of management will be applied to an extremely small subcohort of patients. Maybe that a better clinical result could be reached by a more specific radiopharmaceutical compound, rather than 18F-FDG. The conclusions are then unreliable
