A UK Biobank Study on Genetic Variants in Pattern-Recognition Receptor (PRR) Signaling Indicates Self-Perpetuatin Inflammation of Cholesteatoma

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

v

I think this paper has some worthy for readers after minor revision.

Novelty: First large-scale genetic study (n=678 cholesteatoma cases) leveraging the UK Biobank for this disease.

1. Clear hypothesis: Moves the field from microbial-centric to genetic-driven self-perpetuating inflammation.
2. Methodological framework: Use of GRS and ΔGRS between cases and population provides a robust comparative approach.
3. Clinical relevance: Suggests possible preventive strategies and biologic therapies, addressing the unmet need for medical treatment of cholesteatoma.

  Typographical errors:

• “Variants s in IL6” (line 141) → should be “Variants in IL6”.
• “an informed diagram” (line 73) → better phrased as “a schematic diagram”.

  References: Some duplicates (e.g., Kuczkowski et al. cited twice) could be consolidated.

  Abbreviations: Ensure consistent use (e.g., IL1A vs. Ila typo in Table 2).

  English style: While clear overall, some sentences are overly long and could be tightened for readability.

Figures and Data Presentation

• Figure 1 (signaling scheme) and Scheme 1 (GRS calculation) are helpful, but resolution and labeling need improvement for clarity.
• Table 1 is very long (147 SNPs); summarizing key variants with the strongest effect would be more reader-friendly, with the full table moved to Supplementary Material.
• Table 2 (ΔGRS results) is clear and well-structured, but could benefit from ranking genes by effect size in descending order.

Author Response

please see attached file

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript explores genetic variants associated with cholesteatoma using UK Biobank data and proposes that self-perpetuating inflammation may be driven by polymorphisms in downstream PRR signaling genes. While the topic is potentially relevant and the dataset substantial, the manuscript suffers from serious methodological, statistical, and interpretational shortcomings that undermine its scientific validity.

1. Introduction and Contextual Framing

• The introduction is overly brief and fails to adequately contextualize the disease or justify the study’s rationale. It does not clearly define the clinical burden or global prevalence of cholesteatoma, which is essential for understanding the significance of the research.
• The statement “Several studies have demonstrated upregulation of PRR-related pathways in cholesteatoma” is made without citation. Such claims must be supported by references to peer-reviewed literature.
• The manuscript does not explain why PRR signalling is a particularly compelling target for genetic investigation in cholesteatoma, nor does it distinguish between primary genetic drivers and secondary inflammatory responses.

2. Methodological Transparency

• The authors claim that β-coefficients were calculated using allele frequencies from the UK Biobank, but do not explain the statistical model used. It is unclear how β values (which typically derive from logistic regression models) were obtained from raw allele frequencies alone. Without a clear regression framework, the validity of these coefficients is questionable.
• The manuscript implies that odds ratios (ORs) were derived from allele frequency ratios using the formula, but this oversimplifies the statistical relationship and ignores confounding factors and population structure.
• There is no mention of whether genotype data were coded additively (e.g., 0/1/2 for risk alleles), nor whether Hardy-Weinberg equilibrium was assessed.

3. Statistical Analysis and Validation

• The manuscript lacks any formal statistical testing. No p-values, confidence intervals, or correction for multiple comparisons are provided. This omission makes it impossible to assess whether the observed differences are statistically significant or biologically meaningful.
• The ΔGRS values presented in Table 2 are not accompanied by the underlying GRSchol and GRSpop scores, nor by any indication of variance or significance testing. Without these, the ΔGRS values are uninterpretable.
• There is no correction for key covariates such as age, sex, BMI, or ancestry, which are critical in genetic association studies. The absence of population stratification analysis raises concerns about confounding.
• No ROC curve analysis, regression modeling, or predictive performance metrics are presented, despite the manuscript’s implication that GRS could be used for risk stratification.

4. Data Presentation and Formatting

• Table 1 spans three pages and lists 147 variants with raw β values. This level of detail is excessive for the main manuscript and should be moved to supplementary material.
• Variant IDs are presented in a non-standard format (e.g., “3_38138612_C_CGGCGG”), which is difficult to interpret. It would be more appropriate to use rsIDs, along with chromosomal location and risk allele annotation.
• Scheme 1 is referenced but not clearly explained or labelled. Its purpose and content should be clarified.
• Decimal points are incorrectly formatted using commas instead of periods, which is inconsistent with English-language scientific conventions.
• The list of abbreviations is incomplete; gene names and other key terms should be included.

5. Interpretation and Conclusions

• The manuscript draws strong conclusions about the role of genetic variants in sustaining inflammation in cholesteatoma, yet these are not supported by robust statistical evidence.
• The suggestion that cholesteatoma is driven by “self-perpetuating inflammation” due to genetic predisposition is speculative and not substantiated by the data presented.
• The results section is minimal, consisting of only a few lines of text, one figure, and two tables. This is insufficient to support the manuscript’s claims.

Author Response

please see attached file

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for your detailed replies and for the revisions you have made to the manuscript. I appreciate the inclusion of clinical relevance and incidence data in the introduction, the addition of supporting references, the improved formatting of tables and figures, and the clarification of your methodological approach. The introduction of bootstrap confidence intervals and the explicit discussion of limitations are also positive steps forward.

However, several major concerns remain:

• The study still does not employ logistic regression or other regression models adjusted for relevant covariates such as age, sex, BMI, smoking status, socioeconomic deprivation, or ancestry. These data are available in the UK Biobank and are essential to avoid confounding. Without such analyses, the results remain largely descriptive and hypothesis‑generating.

• The Results section is still minimal, consisting of only a few lines of text, one figure, and one table. A more extensive narrative interpretation of the findings, highlighting specific variants and their effect sizes, is needed to support the claims.

• The suggestion that cholesteatoma is driven by “self‑perpetuating inflammation” due to genetic predisposition remains speculative. Stronger statistical evidence is required before such a conclusion can be justified.

• Although you explained why Hardy–Weinberg equilibrium was not assessed, some form of validation or reference to UK Biobank QC procedures would strengthen confidence in the data.

In summary, while the manuscript has improved in clarity and presentation, it remains primarily hypothesis‑generating. For publication, I strongly recommend performing covariate‑adjusted regression analyses and expanding the Results section with more detailed interpretation. Only then will the conclusions be sufficiently supported by robust statistical evidence.

Author Response

Thank you for your patience while we completed the requested revisions. We are pleased to inform you that our biostatistician has now provided the regression analyses.

Notably, we were unable to perform regression analyses using the total genetic risk score, as the bioinformatics tools of the DNAnexus platform used to access UK Biobank data repeatedly resulted in code execution failures, a problem that has also been noted by other researchers in various diseases. Instead, we selected individual polymorphisms with increased β-coefficients,  thus having the greatest contribution to the genetic score and the greatest possibility of introducing bias. In accordance with the reviewer’s strong recommendation, we conducted covariate-adjusted regression analyses to ensure adequate statistical support for our findings and expanded the Results section with a more detailed interpretation of these analyses. As we mention in our Methods “In addition, we conducted covariate-adjusted regression analyses utilizing sex, age, deprivation and tobacco use as independent variables. Variants with increased β-coefficients of the genes that exhibited the largest differences in genetic risk scores were utilized for the regression analyses”, in our Results “Examples of the regression analyses in selected genetic variables are shown in table 2 and more specifically the association of rs139240442 (IL6) and rs140764737 (TREM10) with demographic, lifestyle, and socioeconomic variables. None of the examined factors showed a statistically significant association with these genetic variants”   and in our Discussion “Recent epidemiological studies using UK Biobank data have reported associations between cholesteatoma and factors such as sex, age, socioeconomic deprivation, and tobacco-related mental and behavioral disorders [40]. In the present study, exploratory regression analyses were therefore performed for selected genetic variants with the highest β-coefficients within genes exhibiting the largest differences in genetic risk scores—variants that contribute most strongly to the overall genetic burden and thus have the greatest potential to introduce confounding. These analyses did not reveal statistically significant associations between the examined polymorphisms and the evaluated demographic, lifestyle, or socioeconomic factors. Given the limited sample size and retrospective design, these findings should be interpreted with caution and warrant validation in larger, independent cohorts.”

Despite the reported limitations, we believe that the addition of covariate-adjusted analyses substantially strengthens the robustness of our results and provides more convincing support for the concept of a self-perpetuating inflammatory process in cholesteatoma, a debilitating yet relatively neglected disease.

yours sincerely,

Round 3

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have adequately addressed all major concerns raised in previous rounds. The manuscript has improved substantially in clarity, methodological transparency, and presentation. Although certain methodological limitations and statistical weaknesses remain, these are inherent to the retrospective design and the constraints of the UK Biobank dataset, and the authors now acknowledge them appropriately.