Free Salivary Amino Acid Profile in Breast Cancer: Clinicopathological and Molecular Biological Features

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The objective of the article is to determine the potential information for studying a profile of salivary amino acids involved in the metabolic processes of patients with breast cancer. The study included 116 patients with 14 breast cancers, 24 patients with benign breast disease, and 25 healthy controls. They mentioned that saliva is a unique biological fluid that contains information about the human body, is non-invasive and can be easily performed in any clinical setting. They conclude that the amino acid profile of saliva allows them to distinguish significant differences in the amino acid profiles sufficient to decide over a tree for the classification of breast cancer samples, non-malignant breast pathologies and healthy controls with a sensitivity of 85.3. %, which shows the potential of using saliva amino acid profile for the diagnosis of breast cancer, even in the early stages. There is no doubt that these types of studies can contribute to the development of personal medicine.

 

Minor comments

 

The manuscript is well organized and supported with results from the literature and his own work. However, the keywords metabolome and diagnosis must be rethought. As they describe it, the metabolome refers to the complete set of molecules or metabolites (such as carbohydrates, lipids, amino acids, organic acids, nucleic acids, hormones,  fatty acids, vitamins, co-factors, pigments, antibiotics, and other signaling molecules) found within a biological sample. Instead of a metabolome they should refer to an amino acid profile for the quantitative content of the 26 amino acids in the saliva that they determined. And since they are not diagnosing, but simply checking if they can classify and distinguish between breast cancer samples, non-malignant breast pathologies and healthy controls, diagnosis should be avoided as an already established as a cleared or approved amino acids based test.

 

Mayor comments

 

Why did not they perform the statistics to determine the sample size. Where the hypothesis, alpha error, beta error, statistical power, variability, losses in the study, effect size, etc. are reflected. It is necessary to carry out the calculation corresponding to the type of study and that support their results. And to match 1:1 cases and controls, otherwise the analysis is not very accurate.

 

Saliva samples were collected during hospitalization strictly before the start of treatment. What about the way the non-malignant breast pathologies and healthy controls were collected? What were the considerations?

 

Regarding the determination of the expression of the receptors for estrogen, progesterone, HER2 and Ki-67. What about the non-malignant breast pathologies and healthy controls. At least they could have made comparisons with the non-malignant breast pathologies or in any case with the adjacent tissues to understand if there is any contribution in the correlation, they make with the diagnosis of breast cancer and the detection in early stages.

 

Concerning to the determination of the amino acid composition of saliva. How many repetitions performed on each sample?

 

Are the values ​​presented in their table averages? It is understood that they do not reflect the individual concentrations of amino acids, and that they only observe the relative changes in the compared concentrations. A supplementary table is needed with any determination or variable from each sample (amino acids, BMI, receptors, stage, menopausal status, age, etc.), that were used in their analysis. And to explain how do they make comparisons, without considering that the controls are not matched in proportion to the number of cases or variables?

 

Why in the discussion they no longer considered any contrast with the antecedents? Sugimoto et al [24]; Cheng et al. [25]; Zhong et al. [26]; or Murata et al. [27]. In the way it is currently presented, it is not possible to review the accuracy of the results. 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Dear Authors!

Thank you for submitting your manuscript "Features of the salivary amino acid profile depending on the molecular biological subtype of breast cancer" for publication.

The selected topic of analyzing salivary amino acids is, in my opinion, a very good decision. It could help identify potential risks and benefits before starting the therapy.

Please find my comments below:

1. In Materials and Methods, page 3, line 107 ff: Please rewrite the section for clarity. It is redundant in some parts.

2. As above: Please add information on the AA kit, including which amino acids were present and at what concentrations.

3. Statistical analyses and the descriptions are thorough and detailed. However, I miss chromatograms and MS traces. Please add MS Base Peak Chromatograms (BPC) for standard amino acids used to generate the calibration curve, the BPC for AA that were found to be significantly modified in breast cancer versus healthy controls as described on page 5, line 162 ff. Of course, not all amino acids need to be shown, just the significantly over or under-expressed from an exemplary case.

4.  On page 7, line 192 ff, you write: "... it is interesting that for constructing the tree..." Please explain why!

Thank you and kind regards.

Comments on the Quality of English Language

Please let a native English speaker check the English style and grammar.  

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

What the authors present is a progress of their research. The authors addressed the previous comments. For them to be able to identify more forceful significant statistical differences between their groups and comparisons, it was suggested that they perform a sample size calculation. However, they argue that the number of variables when comparing different subgroups completely coincided, and that the number of patients in each subgroup diverged slightly. That it is impossible to compose groups in such a way that the number of healthy controls always coincides with the number of patients, since they divide into a large number of subgroups. That a detailed comparison does not make sense, since the molecular biological subtype of breast cancer is not indicated in the literature, the need for which they emphasize in the conclusions of their manuscript, and that the relative change in amino acid concentrations differ among the different molecular biological subtypes of breast cancer. Therefore, they compromise to continue their study, they plan to expand each subgroup to 50 people; and that currently are being collected, but that for rare subtypes this takes time.

 

One way or the other their findings suggests that these amino acids may characterize the presence of the cancer pathology. Although this will not be validated until the number of patients in their study increases. Since as they themselves mention the concentration of individual amino acids in saliva in breast cancer compared with healthy controls varies significantly due to small sample sizes.

 

The current version aims to show that changes in the amino acid profile depend on the molecular biological characteristics of the tumor. And therefore, they proposed the construction of a decision tree for the classification of samples of breast cancer, non-malignant breast pathologies and healthy controls which shows the potential of using the amino acid profile of saliva for the diagnosis of breast cancer, including in the early stages.

 

I share the authors' enthusiasm that the study of salivary amino acid profile from all molecular biological subtypes of breast cancer promise the potential utility of their combinations for diagnostic purposes according to the amino acid values.  Just don't have to leave aside that the profile may show the result patients classified correctly, false negatives or false positive classification, with an accuracy of 75%, that should be improved.