Use of antihypertensive drugs and breast cancer risk: a two-sample Mendelian randomization study
Curation statements for this article:-
Curated by eLife
Evaluation Summary:
Results from observational studies examining an association between the antihypertensive drugs and the risk of breast cancer reported inconsistent findings. This study uses a two-sample Mendelian randomization approach (MR), which overcomes the limitation of observational studies by using genetic variants as a proxy for modifiable exposures, to investigate the relationship between the use of antihypertensive medication and breast cancer risk. Using publicly available data and including a comprehensive assessment of antihypertensive drugs, the authors identified two SNPs that were associated with breast cancer risk. While the findings suggest that antihypertensive medication use may be associated with breast cancer risk, there are some methodological issues that need to be addressed.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
Abstract
Background
Observational studies regarding the correlation between the use of antihypertensive medication and the risk of breast cancer (BC) reported inconsistent findings. We performed a two-sample Mendelian randomization using instrumental variables to proxy changes in gene expressions of antihypertensive medication targets to interrogate this.
Methods
We assessed the association between single-nucleotide polymorphisms (SNPs) and drug targetable gene expression with expression quantitative trait loci in blood. Further, we investigated association between the SNPs and BC risk with genome-wide association study summary statistics. We then confirmed the hits from Mendelian randomization with tissue-specific analyses along with additional sensitivity assessments (horizontal pleiotropy, colocalization, multiple tissue enrichment etc.).
Results
The overall BC risk was decreased 16% with one standard deviation (SD) increase of SLC12A2 gene expression in blood (odds ratio, 0.86, 95% confidential interval, 0.78-0.94). This signal was further confirmed in estrogen receptor positive (ER+) BC (0.85, 0.78-0.94). In addition, one SD increase in expression of PDE1B in blood was associated with 7% increased risk of ER+ BC (1.07, 1.03-1.11). We detected no evidence of horizontal pleiotropy for these associations and the probability of the causal variants being shared between the gene expression and BC risk was 81.5%, 40.5% and 66.8%, respectively. We failed to observe any significant association between other targeted genes and BC risk.
Conclusions
Use of antihypertensive medications that target SLC12A2 and PDE1B is associated with increased and decreased BC risk, respectively.
Funding
This work was supported by the Swedish Research Council [2018-02400 to K.S., 2020-01175 to J.S., 2021-01187 to J.J.], Cancerfonden [2017 CAN2017/340 to J.J.], Crafoordska Stiftelsen [to J.J.], MAS Cancer [to J.J.], ALF funding from Region Skåne [to J.J. and K.S.]. The funding body was not involved in the design of the study and collection, analysis, and interpretation of data and in writing the manuscript.
Article activity feed
-
Evaluation Summary:
Results from observational studies examining an association between the antihypertensive drugs and the risk of breast cancer reported inconsistent findings. This study uses a two-sample Mendelian randomization approach (MR), which overcomes the limitation of observational studies by using genetic variants as a proxy for modifiable exposures, to investigate the relationship between the use of antihypertensive medication and breast cancer risk. Using publicly available data and including a comprehensive assessment of antihypertensive drugs, the authors identified two SNPs that were associated with breast cancer risk. While the findings suggest that antihypertensive medication use may be associated with breast cancer risk, there are some methodological issues that need to be addressed.
(This preprint has been reviewed …
Evaluation Summary:
Results from observational studies examining an association between the antihypertensive drugs and the risk of breast cancer reported inconsistent findings. This study uses a two-sample Mendelian randomization approach (MR), which overcomes the limitation of observational studies by using genetic variants as a proxy for modifiable exposures, to investigate the relationship between the use of antihypertensive medication and breast cancer risk. Using publicly available data and including a comprehensive assessment of antihypertensive drugs, the authors identified two SNPs that were associated with breast cancer risk. While the findings suggest that antihypertensive medication use may be associated with breast cancer risk, there are some methodological issues that need to be addressed.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)
-
Reviewer #1 (Public Review):
In this study, the authors examine the relationship between the use of antihypertensive medication and breast cancer risk using a Mendelian randomization approach.
The strengths of the methods are that they use a two-sample design and have access to publicly available data for this. Several sensitivity analyses were carried out, as well as assessments of the assumptions in studies that use instrumental variables. In addition, the genetic instrument was validated with systolic blood pressure. However, there is some difficulty in following all that was done. Figure 1 does not give complete information. If SNPs are the instrumental variables, why aren't they listed as the "exposure" for the MR for target genes and breast cancer? In section 5 in the methods, the exposure is also indicated as gene expression (and …
Reviewer #1 (Public Review):
In this study, the authors examine the relationship between the use of antihypertensive medication and breast cancer risk using a Mendelian randomization approach.
The strengths of the methods are that they use a two-sample design and have access to publicly available data for this. Several sensitivity analyses were carried out, as well as assessments of the assumptions in studies that use instrumental variables. In addition, the genetic instrument was validated with systolic blood pressure. However, there is some difficulty in following all that was done. Figure 1 does not give complete information. If SNPs are the instrumental variables, why aren't they listed as the "exposure" for the MR for target genes and breast cancer? In section 5 in the methods, the exposure is also indicated as gene expression (and not SNPs). It is also unclear how gene expression can easily stand in for medication use. A diagram (or preferably a DAG) could provide more information to understand the different steps in this MR study.
The authors carried out several analyses and identified two SNPs that were associated with breast cancer risk, suggesting that antihypertensive medication use may be associated with risk.
The results give further information on understanding the mechanisms of breast cancer. Antihypertensive medication is essential for many people, thus, the results would not really lead to less use. That said, the specific genes (and in particular their expression) are related to specific drugs. Thus, there may need to be more work to decide if some drugs are contraindicated in people with an increased risk of breast cancer.
-
Reviewer #2 (Public Review):
Strengths of this manuscript include the comprehensive assessment of 23 potential antihypertensive targets against breast cancer risk, using a two-sample Mendelian randomization framework which permitted the authors to maximise statistical power for their analyses. The principal limitation of this work is that the authors' claims (for the only finding that survives a Bonferroni correction for multiple testing) do not appear to be justified based on the data that they present. The authors have concluded that an effect of an SNP on breast cancer risk is mediated via SLC12A2 (a putative antihypertensive drug target) when this finding could simply reflect the effect of this variant on the expression of a neighbouring gene (CTC-228N24.3) based on data presented by the authors.
The authors show that their SLC12A2 …
Reviewer #2 (Public Review):
Strengths of this manuscript include the comprehensive assessment of 23 potential antihypertensive targets against breast cancer risk, using a two-sample Mendelian randomization framework which permitted the authors to maximise statistical power for their analyses. The principal limitation of this work is that the authors' claims (for the only finding that survives a Bonferroni correction for multiple testing) do not appear to be justified based on the data that they present. The authors have concluded that an effect of an SNP on breast cancer risk is mediated via SLC12A2 (a putative antihypertensive drug target) when this finding could simply reflect the effect of this variant on the expression of a neighbouring gene (CTC-228N24.3) based on data presented by the authors.
The authors show that their SLC12A2 eQTL (rs17764730) is more strongly associated with CTC-228N24.3 (P~0.00) expression than SLC12A2 expression (P+3.2x10-86). Using top eQTLs across respective genes as instruments, they then show similar MR associations of genetically-proxied whole blood SLC12A2 expression (beta=0.15, se=0.039, p=1.05x10-4) vs genetically-proxied whole blood CTC-228N24.3 expression (beta=0.05, se=0.014, p=1.15x10-4) and similar colocalisation estimates for SLC12A2 (H4=81.5%) and CTC-228N24.3 (H4=77%). Notably, the top SLC12A2 eqtl is in perfect LD (r2=1.00) with the top CTC-228N24.3 eQTL (rs6888037) so these results are not surprising.
What these analyses tell me is that one cannot clearly conclude that the effect of this variant on breast cancer risk is mediated via SLC12A2 expression. Unless the authors can provide additional (strong) evidence to support this effect being mediated via SLC12A2, I'm afraid the results presented in this manuscript do not support the conclusions reached by the authors.
-
