The Chemo-Gut Pilot Study: Associations between Gut Microbiota, Gastrointestinal Symptoms, and Psychosocial Health Outcomes in a Cross-Sectional Sample of Young Adult Cancer Survivors

Round 1

Reviewer 1 Report

Although it's an interesting relevant study and the authors have generated an ample amount of data, still, the novelty and innovation are lacking as it is well known now that during chemotherapy gut micro-biome got disturbed and the patients also went through a lot of depression and stress.

Few comments

1. The presentation of the paper is not proper, look like a thesis work rather than a manuscript.
2. In the introduction, the authors are discussing dysbiosis and the immune system, this part can be shortened as the authors have not shown any data about the immune system.
3. In the introduction, the authors should describe mental health, etc.
4. The section named “ Procedure” should come after the “Demographics and clinical data” section, followed by the “ Psychological and Gastrointestinal Outcomes” section of the methodology
5. How the increased population of Selenomondales, loneliaceae, and Intestinibacter in survivors was associated with poor mental health, authors need to explain this point. Authors need to explain the role of these bacteria specifically in gut health and mental health
6. Also, authors can also discuss the relationship between microbial Metabolites and their relation to gut and mental health.
7. Moreover on, the basis of this study, what are the recommendations that can be given to cancer patients so that they can overcome these side effects.

Author Response

Comments and Suggestions for Authors

Although it's an interesting relevant study and the authors have generated an ample amount of data, still, the novelty and innovation are lacking as it is well known now that during chemotherapy gut micro-biome got disturbed and the patients also went through a lot of depression and stress.

Response: We thank the reviewer for their constructive comments. While this statement is correct regarding what is known about the effects of chemotherapy on the gut microbiota during treatment, we wish to reiterate and clarify the 3 key novel components of our study. First, our study was conduced exclusively with young adults while most other studies investigate the gut microbiota in older adults. Second, we specifically focused on the effects that chemotherapy had on the gut microbiota after treatment, not during treatment (which is already known). Third, this is the only study that has comprehensively measured the gut microbiota, GI and psychosocial outcomes using a battery of standardized questionnaires within a single study. These factors clearly demonstrate the novelty of this study and we have added a statement within the introduction to further reiterate and clarify this (see lines: 132 - 136). We have also addressed the novelty of our findings throughout the discussion (e.g. lines: 532 – 534; 581 - 584)

 

1. The presentation of the paper is not proper, look like a thesis work rather than a manuscript.

Response: If the reviewer could please be more specific regarding the presentation of the paper, we would be happy to address this concern. However, we formatted the manuscript according to the formatting requirements set out by MDPI. We have revised the manuscript and made sure its presentation is consistent with journal articles, and also addressed the comments regarding reorganizing parts of the paper below, which should also address this comment.

 

2. In the introduction, the authors are discussing dysbiosis and the immune system, this part can be shortened as the authors have not shown any data about the immune system.

Response: We appreciate this feedback and have removed some of the content related to immune mechanisms and other content that was not pertinent to the present study from the introduction. It was included because our original plan was to measure cytokines, but this had to be removed due to COVID restrictions. Please see the revised manuscript with tracked changed for specific revisions.

 

3. In the introduction, the authors should describe mental health, etc.

Response: We have included more detail regarding mental health as it pertains to specific domains of psychosocial health that we measured in this study, while also clarifying psychosocial health as follows    “Chemotherapy is associated with acute and chronic adverse changes in psychosocial health, but potential underlying mechanisms require elucidation. Mental health, a dimension of psychosocial health (i.e. psychological, behavioral, emotional, and social factors), in this context refers to one’s emotional, psychological, and social well-being [21]” (lines: 90 - 93)

4. The section named “ Procedure” should come after the “Demographics and clinical data” section, followed by the “ Psychological and Gastrointestinal Outcomes” section of the methodology

Response: This has been reorganized in the methods section to be consistent with this request.

 

5. How the increased population of Selenomondales, loneliaceae, and Intestinibacter in survivors was associated with poor mental health, authors need to explain this point. Authors need to explain the role of these bacteria specifically in gut health and mental health

Response: While we appreciate that this is an interesting question, our intention was not to examine whether these specific bacteria were related to GI or psychosocial health outcomes. The purpose of this aspect of our analysis was only to compare the differential taxonomic abundance between survivors and controls, to determine whether differences exist in taxonomic composition between the groups. Although we later ran correlations with the most abundant ASVs in our sample in relation to GI and psychosocial health outcomes, only Intestinibacter was identified as being correlated with cognitive function and depressive functional interference, and we expand on this finding within our discussion (lines: 601 – 602; 670 - 673)

 

6. Also, authors can also discuss the relationship between microbial Metabolites and their relation to gut and mental health.

Response: While this would be an interesting topic of discussion, since we did not measure any metabolites we decided not to speculate on their involvement. However, we agree this is an interesting area for future investigation and we have added suggestion of this in the “Implications and future directions” section of the discussion (lines: 722 - 724)

 

7. Moreover on, the basis of this study, what are the recommendations that can be given to cancer patients so that they can overcome these side effects.

Response: We agree that focusing on knowledge translation with recommendations for cancer patients/survivors is an important aspect of this work, however as this work is still early we believe it is prudent to be cautious with the recommendations made. Nevertheless, we wish to reiterate that we already suggested the following recommendations for addressing potential GI and psychosocial symptoms:

• Using probiotics to address some of the GI and potentially also psychosocial symptoms (lines: 728 - 730)
• For researchers, developing integrative interventions that use probiotics and other nutrition-based strategies (lines: 732 - 734)
• The first 6 months to potentially 1-year post-treatment may be an especially critical time to intervene using microbiota-based interventions (lines: 730 - 732)

Reviewer 2 Report

The manuscript gives a comprehensive and well presented report on associations between gut microbiota, gastrointestinal and psychosocial health outcomes in a cross-sectional sample of young adult cancer survivors. For me, especially interesting were findings that show that changes in gut microbiota persist beyond treatment. I have some minor suggestions for improvement of the manuscript.

 

1. P-values in terms “p<.05” in the abstract are too general. Exact values are needed or the p-values could be omitted (for example, it can just be written that results are statistically significant).
2. I appreciated comprehensive and excellently written overview of the subject in the Introduction and Discussion, but in my opinion it is too long for such paper. Large parts could be omitted. However, I think the text can remain unless shortening is a demand from the editors.
3. For the controls healthy individuals were chosen of the same age. Isn’t there to be expected that they would have different psychosocial health conditions than cancer survivors independent from treatment and gut microbiota? Did authors consider to include the third group – cancer survivors without treatment affecting gut microbiota?
4. Line 305: In the last part of the sentence “…and resided in a metropolitan area.” It appears the percentage is missing or the authors meant something like “… and all but one control resided in a metropolitan area.”
5. Please correct use of decimal places throughout the text. For example in paragraph in lines 309-319 all percentages are per one decimal place except in “...although nearly half (47%) were diagnosed with more advanced (stage III or IV) cancers…”, please use “47.0%”. Furthermore, when SD is telling the mean is not very precise, the second decimal place is irrelevant. For example in line 310 “30.12 (SD= 5.96)” should be replaced with “30 (SD= 6)” or at least with “30.1 (SD= 6.0)”.
6. On Figure 2D the x axis could be drawn up to 5% or 6% instead up to 8% so the bars would be more visible.
7. Not all abbreviations are explained. For example C.I. is first mentioned in line 372.
8. Lines 380-382: This sentence belongs to main text not to figure’s description: “Compared to healthy controls, survivors had significantly more belly pain (**p= < 0.001), gas and bloating (**p= 0.003), constipation (**p= 0.003), and diarrhea (*p= 0.042) (group n’s = 13 – 18).” Similarly at Figure 4 remove the sentence in lines 394-397 and accordingly with all the others Figures’ descriptions (but this additional explanations should remain in supplementary figures). You can however, write some number (for example p-values) on the graphs in case you would like to emphasize the results.
9. Looking at Figures 5A and 5B I wonder if there might be an issue of multiple testing. But authors already covered the issue in detail in lines 670-679. However, many readers read only Abstract and Conclusions, so I would appreciate if a (very short) warning would also be included in the Conclusions.
10. For the Figures 5A, 5B, SuppB and SuppC the group N’s are given to 10-17 and 10-18. Do I understand it correctly that for every psychosocial and every GI outcome there were at least half of individuals having it? Also in healthy controls? Could you provide the N for each calculated rho?
11. Also all the correlation’s rhos could be written in heatmap not just the significant ones. But I do not insist on this.
12. In the Figures’ description please write that Spearman’s rho is presented.

 

Author Response

The manuscript gives a comprehensive and well presented report on associations between gut microbiota, gastrointestinal and psychosocial health outcomes in a cross-sectional sample of young adult cancer survivors. For me, especially interesting were findings that show that changes in gut microbiota persist beyond treatment. I have some minor suggestions for improvement of the manuscript.

Response: We thank the reviewer for their insightful and detailed feedback.

Specific Comments:

1. P-values in terms “p<.05” in the abstract are too general. Exact values are needed or the p-values could be omitted (for example, it can just be written that results are statistically significant).

Response: As the word limit for the abstract is only 200 words, we have updated the abstract as suggested to simply state “results are statistically significant”.

 

2. I appreciated comprehensive and excellently written overview of the subject in the Introduction and Discussion, but in my opinion it is too long for such paper. Large parts could be omitted. However, I think the text can remain unless shortening is a demand from the editors.

Response: We have shortened some sections of the introduction and discussion with an intention to improve the readability of the paper (please see tracked changes for specific revisions).

 

3. For the controls healthy individuals were chosen of the same age. Isn’t there to be expected that they would have different psychosocial health conditions than cancer survivors independent from treatment and gut microbiota? Did authors consider to include the third group – cancer survivors without treatment affecting gut microbiota?

Response: Based on the literature it was hypothesized that the young adult survivors would have poorer psychosocial health, but GI symptoms in young adult cancer survivors have previously not been explored, nor compared to healthy individuals. However, based on the results from a survey study we recently completed showing that 51% of survivors experience moderate to severe GI symptoms persisting up to an average of 2.5 years post treatment, we hypothesized that survivors in this study would also have increased GI symptoms. Regarding the third group, we presume the reviewer means cancer survivors who did not receive chemotherapy (i.e. hypothesized to affect the gut microbiota)? In this case, we did consider this and tried to recruit patients who had received other treatments exclusive of chemotherapy (e.g. surgery), however we were not able to recruit a group of such patients as we used convenience sampling (due to the COVID pandemic we could not recruit in clinic) and all potential survivors who were interested in the study had previously received chemotherapy. We were hoping to have that other group of survivors with treatments other than chemotherapy, but were not able to recruit anyone to that group. We have added a note on this within the manuscript (lines: 151 – 154). As well, at present we do not know which treatments could reliably be said not to affect the gut microbiota. Therefore, this presents a further challenge in recruiting to this third group.

4. Line 305: In the last part of the sentence “…and resided in a metropolitan area.” It appears the percentage is missing or the authors meant something like “… and all but one control resided in a metropolitan area.”

Response: We have rephrased that sentence to clarify that all but one control resided in a metropolitan area.

5. Please correct use of decimal places throughout the text. For example in paragraph in lines 309-319 all percentages are per one decimal place except in “...although nearly half (47%) were diagnosed with more advanced (stage III or IV) cancers…”, please use “47.0%”. Furthermore, when SD is telling the mean is not very precise, the second decimal place is irrelevant. For example in line 310 “30.12 (SD= 5.96)” should be replaced with “30 (SD= 6)” or at least with “30.1 (SD= 6.0)”.

Response: The results have been updated to reflect this request.

6. On Figure 2D the x axis could be drawn up to 5% or 6% instead up to 8% so the bars would be more visible.

Response: We have adjusted the x-axis as requested to enhance the visibility of the bars in figure 2D.

7. Not all abbreviations are explained. For example C.I. is first mentioned in line 372.

Response: We have clarified abbreviations throughout the manuscript and added a list of abbreviations at the end of the paper for quick reference.

8. Lines 380-382: This sentence belongs to main text not to figure’s description: “Compared to healthy controls, survivors had significantly more belly pain (**p= < 0.001), gas and bloating (**p= 0.003), constipation (**p= 0.003), and diarrhea (*p= 0.042) (group n’s = 13 – 18).” Similarly at Figure 4 remove the sentence in lines 394-397 and accordingly with all the others Figures’ descriptions (but this additional explanations should remain in supplementary figures). You can however, write some number (for example p-values) on the graphs in case you would like to emphasize the results.

Response: We have updated the figure descriptions to be consistent with this request.

 

9. Looking at Figures 5A and 5B I wonder if there might be an issue of multiple testing. But authors already covered the issue in detail in lines 670-679. However, many readers read only Abstract and Conclusions, so I would appreciate if a (very short) warning would also be included in the Conclusions.

Response: we added the following statement to the conclusion in acknowledgment of this: “However, given that we did not control for multiple comparisons in this exploratory study these results must be interpreted with caution.” (lines: 742 - 743)

 

10. For the Figures 5A, 5B, SuppB and SuppC the group N’s are given to 10-17 and 10-18. Do I understand it correctly that for every psychosocial and every GI outcome there were at least half of individuals having it? Also in healthy controls? Could you provide the N for each calculated rho?

Response: For all psychosocial outcomes except pain behaviour and PTSD symptoms, where some participants reported no symptoms and thus no score (noted in methods, lines: 279 - 280), all participants had an outcome score for each measure. With the exception of the Impact of Live Event scale and the PROMIS pain behaviour scale which allow for a score of 0, PROMIS measures allow for each participant to receive a score, which can be very low (indicating few/no symptoms good health – as was typically the case for our control group) or high score (indicating severe symptoms). For our correlation analysis we excluded any correlation that did not have at least n=10. We detailed our decision for this within the methods section in (lines: 287 - 290). We did not include the n for each rho as this would be very cumbersome given the multiple correlations and instead have reported the range of n’s.

 

11. Also all the correlation’s rhos could be written in heatmap not just the significant ones. But I do not insist on this.

Response: We decided not to include all of the correlation rho’s in the heatmap as this creates a very messy looking heatmap, potentially making it more difficult to interpret, and we wanted to keep the focus on the significant rho’s only.

12. In the Figures’ description please write that Spearman’s rho is presented.

Response: we have added this to the figures.

Reviewer 3 Report

This manuscript is well organized and is a fair assessment of the data that was able to be collected with COVID-19 modifications to the protocol. The authors do a good job of interpreting data and placing them in the context of current literature. The topic of gut-brain communication is of interest in the oncology community.

Specific Comments:

1. “Homeostatic dysregulation” is not a common phrase (line 36). Would suggest explaining or rephrasing.
2. With the consideration that microbiome refers to the collection of genes and microbiota to the collection of organisms, it is suggested to review instances of their use through the manuscript and confirm they are used properly.
3. Confirm methods from literature are reported correctly. For example, line 48 indicates Loman, et al. treated mice every day with paclitaxel, which is not accurate based on the paper.
4. The shift from separated survivor groups to combined survivor groups is a bit jarring. Providing rationale for their combination in figures 2-5 would add to cohesiveness of manuscript
5. In graphical representation of data, it is standard to make controls the left-most group.
6. Indicate what error bars represent in figure legends.
7. Figure 4 is separated, either make 2 different figures, or combine into one. Similar for figure 5.
8. If figure 5 heatmaps are for only the survivor group, please only put n=(#) for that group, not for survivors and controls.

Author Response

This manuscript is well organized and is a fair assessment of the data that was able to be collected with COVID-19 modifications to the protocol. The authors do a good job of interpreting data and placing them in the context of current literature. The topic of gut-brain communication is of interest in the oncology community.

Response: We thank the reviewer for their time and constructive feedback on our manuscript.

 

Specific Comments:

1. “Homeostatic dysregulation” is not a common phrase (line 36). Would suggest explaining or rephrasing.

Response: This has been rephrased.

2. With the consideration that microbiome refers to the collection of genes and microbiota to the collection of organisms, it is suggested to review instances of their use through the manuscript and confirm they are used properly.

Response: we have reviewed our usage of microbiota throughout the manuscript to ensure it is consistent and have updated any discrepancies. All such changes are noted with “tracked changes” in the manuscript. Notably, while some of the studies discussed herein use programs, such as PICRUSt, to infer the metabolic capacity of the microbiome, we have focused our discussion of the findings from these studies to the taxonomic characterization of the microbiota using 16S rRNA sequencing and therefore refer to microbiota throughout the manuscript as opposed to the microbiome. We have added a sentence to clarify this (see lines: 43 - 47)

 

3. Confirm methods from literature are reported correctly. For example, line 48 indicates Loman, et al. treated mice every day with paclitaxel, which is not accurate based on the paper.

Response: Thank you for this observation. We have double checked the methods and revised accordingly.

 

4. The shift from separated survivor groups to combined survivor groups is a bit jarring. Providing rationale for their combination in figures 2-5 would add to cohesiveness of manuscript.

Response: Since beta diversity analysis revealed no significant differences between groups, we decided that merging the survivor group for the correlation analysis was warranted, which also served to increase the power for the analyses. For the analyses of GI and psychosocial outcomes, there is insufficient evidence in the literature suggesting that time off treatment warrants stratification of the survivors group. Moreover, another study we recently completed with N=317 cancer survivors showed that time off treatment did not significantly impact GI symptoms or mental health. As such, we decided not to stratify the survivor group for all analyses. A brief statement clarifying this has been added (lines: 386 - 392).

 

5. In graphical representation of data, it is standard to make controls the left-most group.

Response: As MDPI made set no specific requirements regarding this, and other clinical work published by MDPI has similarly presented control data as the right-most group (e.g. Doege et al. 2021 https://doi.org/10.3390/cancers13112754; Maurer et al., 2021 https://doi.org/10.3390/cancers13081854), we feel that the presentation of our data is consistent with the journal and other clinical studies comparing cancer cohorts and controls.

 

6. Indicate what error bars represent in figure legends.

Response: this has been added to all figures.

 

7. Figure 4 is separated, either make 2 different figures, or combine into one. Similar for figure 5.

Response: These figures have been separated as suggested and are shown using tracked changes.

 

8. If figure 5 heatmaps are for only the survivor group, please only put n=(#) for that group, not for survivors and controls.

Response: This has been revised accordingly to indicate the range of n’s each group.

Round 2

Reviewer 1 Report

All the suggestions have been addressed very nicely by the authors and now manuscript can be accepted.