This MR study, conducted in two independent PC datasets, presented strong genetic evidence supporting the potential of HMGCR inhibition and LDLR inhibition in lowering PC risk. The findings emphasize HMGCR as a promising drug target for PC treatment. However, no significant evidence was found for the impact of lipid traits and other lipid-lowering drug targets in reducing PC risk. The mediation analyses indicate that the mechanism through which HMGCR inhibitors reduce the risk of PC may, in part, be mediated by influencing BMI levels.
In previous MR studies on the associations between lipid traits and PC, no significant associations were found between LDL-C, TG, TC, HDL-C, and PC[19]. However, a causal association between TC and PC (OR 1.34, [95% CI, 1.02–1.76]; p = 0.03) in the Pancreatic Cancer Cohort Consortium1 consortium[20]; in addition, a causal association between LDL-C and PDAC (OR 1.16, [95% CI, 1.02–1.76]; p = 0.03) was found in the Pancreatic Cancer Cohort Consortium and the Pancreatic Cancer Case-Control Consortium[21]. The reasons for the non-consistent results may lie in the differences in data sources and the lack of adjustment for intrinsic risk factors. After we analyzed the results from both datasets, in line with previous research, no robust evidence was found to support a meaningful association between lipid traits and PC.
Cholesterol is involved in the growth, survival, and progression of cancer cells in malignant tumors, elevated cholesterol levels increase the risk of PC[22]. Furthermore, dysregulation of cholesterol metabolic pathways also affect cancer prognosis[23]. Statins have been identified as standard drugs for lowering cholesterol levels[24], including lovastatin, simvastatin, pravastatin, fluvastatin, etc., which exert their therapeutic effects by inhibiting the activity of HMG-CoA reductase encoded by the HMGCR gene, and statins have an inhibitory effect on a variety of tumors [25, 26]. Numerous studies support the clinical relevance of HMGCR inhibitors in PC. For example, in a transgenic mouse model, simvastatin was demonstrated to be an effective chemopreventive agent that delayed pancreatic intraepithelial neoplasia and inhibited the formation of PC[27]. Observational studies and clinical trials have indicated that statin use is associated with a reduced risk of PC and potentially enhances patient prognosis[28–30]. Our study demonstrates a previously unexplored correlation between HMGCR and PC at the genetic level. A positive causal relationship between BMI and PC has been identified in previous studies[31, 32], with high BMI levels being associated with increased mortality rates[33]. Through TSMR analysis, our study suggests that HMGCR potentially inhibits PC by regulating BMI levels. HMGCR, the target of statins known for their lipid-lowering effects[34], may modulate BMI and reduce the risk of developing PC. Additionally, statins also possess significant anti-inflammatory properties [35]. In obese patients, where obesity-induced fibrosis and inflammation drive PDAC progression and influence chemotherapy prognosis[31, 36], statins may mitigate the risk of PC by impacting inflammation levels. Although these mechanisms provide potential pathways for HMGCR inhibitors to reduce the risk of PC, the exact roles and interactions of these mechanisms need to be further explored and validated.
Furthermore, previous studies have shown that PCSK9 and LDLR are potential targets for cancer immunotherapy[37], while emphasizing the significant role of LPL in inflammation and obesity, positioning it as a promising target for chemoprevention and chemotherapeutic agents[38]. However, a recent MR study did not uncover any causal link between PCSK9 inhibitors and PC[39], aligning with the findings of our study. Although LDLR and PC did not show statistical significance in the SMR analysis in our study, statistical significance was revealed in the meta-analysis of the MR results of two independent PC datasets, so LDLR inhibitors may be applied in the targeted therapy of PC, and no causal association was observed between LPL and PC in our study, indicating the need for further research.
This study represents the first comprehensive analysis utilizing MR to investigate causal associations between lipids, lipid-lowering drug target genes, and PC. It was conducted using both the Discovery and the Replication Datasets, incorporating TSMR and SMR analyses to provide a higher level of evidence. Compared to clinical trials, which are associated with lengthy periods, high costs, and limited generalizability, drug-targeted MR offers a valuable approach for rapidly assessing drug applicability while overcoming these challenges. The study, however, has several limitations that should be acknowledged. Firstly, the relatively small number of PC cases may have affected the statistical power and effectiveness of the findings. Additionally, the data used in this analysis were limited to individuals of European ancestry, and thus, the extrapolation of the results to other ethnicities remains uncertain. It is important to further investigate the applicability in diverse populations. Moreover, genetic variation captures long-term changes in lipid levels, and the extent of its impact may not be directly comparable to the short-term effects of lipid-lowering drugs. Furthermore, due to the unavailability of individual data, stratified analyses could not be performed. Finally, although sensitivity analyses were conducted, the possibility of horizontal pleiotropy could not be completely ruled out.