Selected Serum Biomarkers (Leptin, Chromogranin A, CA19-9, CEA) in Patients with Pancreatic Neuroendocrine Neoplasm and Associations with Metabolic Syndrome

Round 1

Reviewer 1 Report

Thank you for this interesting study, the design is not entirely clear (so please clarify) but I gather that it is a clinical prospective study (not observational?) to investigate SBMs in people with neuroendocrine tumour of the pancreas. Please clarify on the design. Please decide on the primary and secondary objectives of this study and then reflect this in the title. As the title is also misleading, unless to investigate the associations between MS and SBMs in this cohort of people was your primary objective?

The statistical methods are also unclear. Please can I suggest that you clearly specify in the abstract and in the methods section which methods you use (the correlation techniques) to investigate the association, and why these were the methods of choice for his study? Please define which statistical technique was used to calculate AUC? I suggest a table could be included listing the relationships that were investigated (what with what?  as there is a lot!).

An important in my mind concept is that the relationships of SBMs and cancer are complex and depend on parson’s characteristics (as you rightly point out in your secondary analysis), their gender and sex for example. Please can I request that you consider multivariable analysis (as your analysis as only univariable) to explore these interactions. It is unclear why the subgroups analysis were only undertaken in the cancer group? Was the control group too small? This limits your study considerably.

To this extent, the purpose of the subgroup analyses is unclear. Please could you clearly state why you have undertaken these analyses other than just to explore the differences? If it is just to show the differences in the biomarkers due to BMI, sex and MS, then to my view too  much attention is placed in the paper on these analysis, and although interesting, not necessarily very novel and cannot directly help differentiate cancer from non-cancer and it is unclear what the purpose of these analysis is. In general the objectives of the paper should be defined clearly (please list them, that would help a lot) at the end of the introduction. Otherwise, the introduction is superbly written, clearly and informatively, thank you.

The abstract however, is unclear and could be misleading.  It is unclear from the abstract if it is about malignant neoplasm or non-malignant (this has a lot of implications?), for simplicity could you call it cancer? Also as mentioned above, please be clearer and more concise in the abstract on what relationships were explored (with what) and why.  And also please be clear that the subgroup analysis were only performed for people with cancer and clarify why was that. Abstract in general took substantial amount of time to read and I was not clear until after I read the whole paper (still not clear on what the exact objectives as the objectives change through the paper). The authors have done a lot of analysis so it is important to structure the paper in a clear way from the start so I will suggest another iteration on redrafting the manuscript for clarity.

There are some misleading statements (which does not recognise the limitations of the study design and statistical analyses) such as:

- the study was designed to assess the accuracy of the SBMs in PanNENs as a diagnostic

- AUC analysis of CEA and CgA could differentiate PanNENs from controls…

Please clearly state the design of the study in methods (not in results), discuss the limitations of the design, and also please clearly reflect the study design in the objectives.

Some other comments:

 

- with that many statistical tests undertaken, I suggest reducing statistical significance level to 0.01 otherwise you may be risking many type1 errors

- please provide in methods, which statistical technique was used to calculate AUC

- in your demographics summary table, please provide the subgroup split for the controls, for information even if you are not using it

- please explain in methods, why were the healthy controls admitted to the unit

- please provide units for all the SBMs, weight, bmi etc at the first mention in the results section. It just needs tidying up.

- You have abbreviated PanNENpts but then you write it in full. In general, I would advise against this abbreviation as it just obstructs the reading. Similarly, PanNEN, especially in the abstract I feel that this abbreviation doesn’t add anything so please remove it at least from the abstract. I am not an expert in this so may not understand the biochemistry of the cancer that you study but for ease for the reader could you call it pancreatic endocrine cancer? But please feel free as I realise that this suggestion may be incorrect. What is TM? In general, the methods section is so abbreviation heavy that it is hard to read. Please limit the abbreviations to minimum.

- BMI of <25 is not always a "normal" weight, for example because people in this group could be underweight. I am not suggesting splitting it any further, but I am suggesting please do not call it normal weight but people with BMI <25 as this group could encompass many weight statuses.

Thank you for giving me the opportunity to read this interesting paper. It is a good contribution however my take is that re-planning and reworking the manuscript for clarity of the objectives, methods and results would improve it. Overall, I hope my comments are helpful.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

This is a report about the Pancreatic neuroendocrine neoplasm.
The article is original and very well written. I think that it is a good contribution to this subject. By the way I have some concern, and I would like to have some comments for the results section and  the discussion in order to improve the manuscript.

Authors should indicate in the results the number of patients with jaundice. In these patients, CA 19.9 levels may be falsely elevated

 

In the discussion the authors should underlying the problem of the metabolic syndrome especially for patients that will undergoing surgery for resection.

Surgery is one of the best alternatives to increase survival in these patients, but it is not without complications.

Various risk factors for POPF (postoperative fistu) after pancreatectomy have been recognizes, as high body mass index (BMI). 

The authors should cite this meta-analysis in which fistula risk, metabolic syndrome, and BMI are also discussed. (https://pubmed.ncbi.nlm.nih.gov/35525852/).

 

 

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Thank you to all the authors for providing all the edits and for all the explanations to questions. I have learnt a lot reviewing this manuscript. I have one final comment, please could you return to my comment about the method to calculate AUC, please could you clarify in the methods section what method was used to calculate AUC and write one or two sentences why? In other words, what method was used to undertake predictions, was it logistic regression (quite commonly used) or was it a different binary classifier? 

The authors stated that "AUC was calculated as the area under the receiver-operating characteristic (ROC) curve, using ROC analysis". ROC is a graphical visualisation of AUC. Please provide more clarity on the method/statistical model that was used to undertake predictions/classification (differentiation between the two groups of patients), also if and how was the split for training and validation performed and please consider the limitations of that as well. 

 

Author Response

Please see the attachment

Author Response File: Author Response.pdf