Impact of hMLH1 −93G>A (rs1800734) and hMSH2 1032G>A (rs4987188) Polymorphisms on Colorectal Cancer Susceptibility

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Dear Authors,

Your manuscript is really well-written, the material is novel and is very important for the cancer research field. The English is practically perfect, with a few exceptions:

• 136; There should be a gap between as P<0. and Data
• All tables should be numbered
• All values should have consistent significant digits
• 349; CRC Colorectal cancer - it has different font

Best, 

Your Reviewer

Author Response

We thank you for the valuable comments and time spent on our manuscript. The alterations were made throughout the manuscript.

Comment 1: 136; There should be a gap between as P<0. and Data

Response 1: The spacing at line 136 between P<0. and Data has been corrected.

Comment 2: All tables should be numbered

Response 2:  Thank you for noting this. The mistake was due to a technical issue during the submission process. Table numbers have now been correctly inserted into all captions in the revised manuscript.

Comment 3: All values should have consistent significant digits

Response 3: Numerical values have been revised to ensure consistent use of significant digits across all results.

Comment 4: 349; CRC Colorectal cancer - it has different font

Response 4: The font inconsistency noted at line 349 in “CRC Colorectal cancer” has been corrected for uniformity.

Thank you again for helping us improve the overall presentation and quality of the manuscript.

Reviewer 2 Report

Comments and Suggestions for Authors

This paper studies the relationship between two single nucleotide polymorphism, hMLH1 -93G>A and hMSH2 1032G>A and risk of colorectal cancer in Azerbaijani population. Study concludes that hMLH1 -93G>A polymorphism under recessive model offers protective role agains cancer risk in this population. No significant association was found for hMSH2 1032G>A.

The study is novel in Azerbaijani population although the rationale for selecting these SNPs could be strengthened. 

Methodology for most part is clear however details on how controls were selected can be improved.

Regarding the findings of hMLH1 -93G>A, I find the emphasis on the recessive model protective effect to be a bit problematic. Previous studies have shown inconsistent results on this. Functional data in the form of biological reasoning or references to support the protective effect is not stated anywhere in the paper. The possibility of a false positive may be further explored by statistical correction. Also, in the same table when grouped separately AA genotype shows somewhat contradictory results (1.474, p0.2) which requires some clarification.

Regarding hMSH2 1032G>A, results show a very low frequency of AA genotype (0.8%) which is a major limitation.

Author Response

Thank you for your careful review and helpful feedback on formatting and consistency. We have addressed all the noted issues as follows:

Comment 1: The study is novel in Azerbaijani population although the rationale for selecting these SNPs could be strengthened.

Response 1: Thank you for your thoughtful feedback. We have expanded the Introduction slightly to clarify our rationale for selecting these SNPs ( lines 65, 66; 78-81).

Comment 2: Methodology for most part is clear however details on how controls were selected can be improved.

Response 2: Thank you for your comment. The relevant section of the manuscript has been revised accordingly (lines 94-98).

Comment 3: Regarding the findings of hMLH1 -93G>A, I find the emphasis on the recessive model protective effect to be a bit problematic. Previous studies have shown inconsistent results on this. Functional data in the form of biological reasoning or references to support the protective effect is not stated anywhere in the paper. The possibility of a false positive may be further explored by statistical correction. Also, in the same table when grouped separately AA genotype shows somewhat contradictory results (1.474, p0.2) which requires some clarification.

Response 3: We truly appreciate this thoughtful comment, as it helped us identify an important issue. Unfortunately, during manuscript submission, an earlier version of the tables was mistakenly uploaded. We have now carefully reviewed and corrected all tables to ensure consistency with the main text and statistical results (Tables 2-5).

Comment 4: Regarding hMSH2 1032G>A, results show a very low frequency of AA genotype (0.8%) which is a major limitation.

Response 4: Thank you for this important observation. We acknowledge that the AA genotype of hMSH2 1032G>A was detected at a low frequency (0.8%) in our cohort, which limits the statistical power to evaluate its potential association with CRC risk. Similar low frequencies of this genotype have also been reported in other populations, as cited in the literature, indicating this may be a broader trend rather than a study-specific issue. Nonetheless, we have clearly stated this in the Discussion section as a limitation and emphasized the need for larger, multi-center studies to allow more conclusive interpretation of this variant’s role (lines 456, 457).

Once again, we would like to thank you for helping us improve the overall presentation and quality of the manuscript.

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript is well written, methods are appropriate.

However, I have several issues:
1) The number of patients is rather low for such a common cancer. Please provide the power of the study.

2) Please check the Hardy Wienberg equilibrium

3) Although authors' contribution is stated, the study of 2 PCR-RFLPs in 137+134 samples does not require 10 authors. 
4) Figure should be transferred to the results

5 ) Please provide table numbers in the table caption

Finally, as stated in the discussion, much larger studies have already been performed, and some conclusions have been reached; therefore, adding 137 samples to the previous thousands does not add anything new to the scientific field. 

Author Response

Please see the attachment. 

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Results of the study were revised, now showing hMLH1-93G>A polymorphism with increased risk of colorectal cancer under the recessive model (OR 1.78, CI=1.102-2.90, P=0.018).

While the results are biologically plausible and supported by other studies, the statistically significant findings were seen only in hMLH1-93G>A recessive model with modest increase in risk. The allele frequency difference is not statistically significant (p=0.064). Lack of statistical significance for almost all other associations does raise the concern of the study being underpowered.

Interpretation of recessive model significance should be balanced with nonsignificant findings with cautious interpretation of the results in discussion section in my opinion.

In depth comparison with studies with similar results might help in better contextualizing the findings. For instance, what were the odds ratio for those studies etc

For promoter region single nucleotide polymorphisms of hMLH1, how important would methylation status data be to support the overall findings?

Regarding hMSH2 result, 1032G>A variant is so rare that asserting no association sounds misleading.

Author Response

Comment 1: Results of the study were revised, now showing hMLH1-93G>A polymorphism with increased risk of colorectal cancer under the recessive model (OR 1.78, CI=1.102-2.90, P=0.018).

While the results are biologically plausible and supported by other studies, the statistically significant findings were seen only in hMLH1-93G>A recessive model with modest increase in risk. The allele frequency difference is not statistically significant (p=0.064). Lack of statistical significance for almost all other associations does raise the concern of the study being underpowered.

 Response 1:  We appreciate the reviewer’s thoughtful feedback. Due to these initially acknowledged limitations, we approached the study design with particular care to maximize validity within the available sample. We agree that the lack of statistically significant associations outside of the hMLH1 -93G>A recessive model may raise concern regarding the study’s statistical power. To address this, we performed a post-hoc power analysis using G*Power software. The analysis indicated that our study had approximately 70% power to detect an effect size equivalent to the observed odds ratio under the recessive model. While this is slightly below the commonly accepted threshold of 80%, it is considered moderately acceptable for exploratory genetic association studies, particularly in underrepresented populations.

We agree that the statistically significant association was observed specifically under the recessive model for the hMLH1 -93G>A polymorphism, while no significant association was found under other models or allele frequency comparisons. In response, we made a targeted adjustment to the Abstract and Conclusion to clarify that the observed association is specific to the recessive model, ensuring that the conclusion does not imply a broader overall association (lines 36-40; 443-446). We have retained the rest of the manuscript as is, as it already discusses these findings with appropriate caution and includes acknowledgment of the study’s limitations.

Comment 2: Interpretation of recessive model significance should be balanced with nonsignificant findings with cautious interpretation of the results in discussion section in my opinion.In depth comparison with studies with similar results might help in better contextualizing the findings. For instance, what were the odds ratio for those studies etc

Response 2:  We appreciate the reviewer’s thoughtful and constructive feedback. In response, we revised the Discussion section to ensure a more balanced interpretation of our results. Namely, we now clearly state that a statistically significant association was observed only under the recessive model, while other genetic models and the overall distribution did not yield significant differences. This clarification highlights the model-specific nature of our finding without overstating its implications (lines 378-403; 513-516). Additionally, we expanded the results by inclusion of direct comparisons with relevant studies.

Comment 3: For promoter region single nucleotide polymorphisms of hMLH1, how important would methylation status data be to support the overall findings?

Response 3:  Thank you for raising this important point. As noted in our limitations section, our study did not include methylation profiling. However, we think that assessing promoter methylation (e.g., via MethyLight or MS-MLPA) status would significantly strengthen the interpretation of promoter-region SNPs such as hMLH1 -93G>A. Promoter methylation of hMLH1 is a well-established mechanism leading to gene silencing in sporadic colorectal cancer, particularly in microsatellite instability-high (MSI-H) tumors. Given that the -93G>A variant lies within the promoter region, it is functionally relevant that it could influence transcriptional regulation via epigenetic mechanisms. However, as our study did not include methylation analysis, we felt it would be most appropriate to acknowledge this explicitly as a limitation rather than speculate in the main discussion.

Comment 4: Regarding hMSH2 result, 1032G>A variant is so rare that asserting no association sounds misleading.

Response 4:  We thank the reviewer for this important comment. We agree that the low number of individuals carrying the AA genotype limits the statistical power to assess the association between the hMSH2 1032G>A variant and CRC risk. To address this concern, we have added a clarifying sentence in the Results section, stating that although no statistically significant association was observed, the interpretation remains constrained by the very low number of AA genotype carriers, which may have limited the sensitivity to detect a true association (lines 319-322).

We would also like to emphasize that we approached the study design with due care and applied appropriate statistical tests throughout to ensure robust analysis. While we fully acknowledge the limitations posed by sample size, the findings presented reflect the data available, and we have interpreted them transparently. Importantly, our results remain consistent with previous reports that also did not establish a significant association for this rare variant, further supporting the reliability of our findings despite the limited frequency.

Once again, we thank the reviewer for their constructive comments, which have helped improve the clarity, depth, and interpretive quality of our manuscript!

Reviewer 3 Report

Comments and Suggestions for Authors

nothing further

Author Response

Response: Once again, we thank the reviewer for their constructive comments, which have helped improve
the clarity, depth, and interpretive quality of our manuscript.

Round 3

Reviewer 2 Report

Comments and Suggestions for Authors

This paper on two mismatch repair proteins has been revised to address comments with a more balanced interpretation of results. Would suggest to tone down the conclusion and expand the limitations section to include how future studies can be improved.

Author Response

 

Comment 1: This paper on two mismatch repair proteins has been revised to address comments with a more balanced interpretation of results. Would suggest to tone down the conclusion and expand the limitations section to include how future studies can be improved.

Response 1: Thank you. In response to the reviewer’s comment, we have revised the Conclusion to reflect a more balanced and measured interpretation of our findings (5. Conclusion section; lines 348-365). We have also expanded the Limitations (lines 335-347) paragraph to clearly outline the limitations of our study and suggest concrete ways future research can be strengthened. 

We are truly grateful for your thoughtful and rigorous feedback, which has helped us significantly improve the clarity, precision, and overall quality of the manuscript!