Association of Cytotoxic T-Lymphocyte Antigen-4 (CTLA-4) Genetic Variants with Risk and Outcome of Cutaneous Melanoma

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This study investigated the correlation between many different germline SNVs of CTLA4 and skin melanoma, which unveiled many genotypes of CTLA4 may affect the proliferation, migration, apoptosis, necrosis of tumor cell, as well as survival outcomes of patients with skin melanoma. The methodology employed in the present is rational and appropriate, and the article is well-structured and logically organized. The study presented in this article makes a moderate contribution to precision oncology, but does not displayed enough clinical significance. As we know, CTLA4, as a famous immune checkpoint, has already been widely utilized as important target of ICI immunotherapy, including skin melanoma, so the first question is author just investigated the correlation of SNVs of CTLA4 with chemotherapy, why not link with ICI immunotherapy, particularly CTLA4-based immunotherapy. For example, are there many specific genotypes of CTLA4 mutation that would affect immunotherapy response in melanoma? I suggest author could complement this content. In addition, in terms of Transcription factor binding motif analysis, there is no figure showing this result, or any evidence about this analysis. Author should complement it.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This study evaluated the roles of the CTLA-4 SNVs, c.-1765C>T, c.-1661A>G, c.- 276 1577G>A, and c.-1478G>A, in the risk of cutaneous melanoma compared to health controls. There was also a assessment only looking at the included melanoma patients and relating the variants to clinicopathological aspects and also look at survival within the melanoma patients. In addition, the c.-1765C>T SNV was considered of great interest and was evaluated by complementary functional studies.

Concerning the introduction: it is a nice introduction with a nice figure. Only the last paragraph is too detailed. I do not believe every transition of amino-acid needs to be written out. 

Concerning the methods and results:

I do believe the control group of healthy controls is a biased one. It is from patients which were blood donor. This are usually more healthy individuals. I think this should be stated as a limitation in the discussion. They already state that there cutaneous melanoma group may not be representative. I agree with that. 

For the methods I do not understand why only look at genotypes, while I think it most logical to look at allelic level and only test additionally on genotype level if you only expect a strong homozygous effect. Either clarify, or change. 

In the multivariate analysis I do not understand why the cut-off for breslow is taken at 1,5 mm. This is not a common clinical cut-off as far as I know. In clincial stage Breslow and ulceration are already taken into account. Whey is ulceration not separately shown? Or only take stage or use both. In the text there is a distinction made between superficial versus non-superficial type. I think it is better to refer to 'superficial spreading', otherwise it is unclear if you mean a superficial Breslow. I prefer also the term 'lentigo maligna'. I do not understand why some multivariate analysis for genetic variants were not applicable? I am personally not a believer in that you need to perform a multivariate analysis as you do not expect the confounding factors to be related to the genotype (Mendelian randomization) . I know there is room for debate about this, so I am fine with showing multivariate analysis. 

You describe that some patients you did not have sufficient information on the follow-up. Can you show maybe in a supplement that these patients did not significantly differ on characteristics from the rest. To make sure there is no bias because of that. 

Discussion

I think generally speaking the results are presented a bit too strong. I do believe this is a biased patient population, but it is something interesting to further explore in a larger nationwide dataset. I think this work is more hypothesis generating. 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf