Assessing the impact of PCSK9 and HMGCR inhibitor on reproductive endocrine diseases: A drug-target Mendelian randomization analysis

Reproductive endocrine disorders (REDs), including polycystic ovary syndrome (PCOS), endometriosis, and infertility, have long been a cause of distress for women in their reproductive period. PCOS, a prominent endocrine disorder affecting women of reproductive age, presents a range of symptoms, such as polycystic ovaries, anovulation, and hyperandrogenism. [ 1 – 3 ] Nevertheless, the consequences of PCOS reach further than its effects on reproduction and hormonal balance. It often intertwines with obesity, disrupted lipid metabolism, and insulin resistance, thereby amplifying the likelihood of enduring complications like diabetes and cardiovascular disease. [ 4 , 5 ] Endometriosis, recognized as one of the most debilitating and enigmatic gynecological conditions, presents a myriad of troubling symptoms. These encompass dysmenorrhea, enduring pelvic pain outside of menstrual cycles, and dyspareunia. [ 6 , 7 ] Female infertility, characterized by the incapacity to achieve pregnancy following a span of 12 months of unprotected sexual intercourse, [ 8 ] significantly affects the quality of life for approximately one in 6 couples globally. [ 9 , 10 ] Given the considerable health, economic, and social implications imposed by these conditions, there is a pressing need to comprehend the underlying mechanisms and secure effective treatment options. Importantly, it has been firmly established that obesity plays a pivotal role in the elevated prevalence of PCOS. [ 11 – 13 ] Hyperlipidemia, a fundamental hallmark of PCOS, has been associated not only with the presence of hyperandrogenism but also with the intensification of hyperandrogenic manifestations in susceptible individuals. [ 14 , 15 ] These discoveries emphasize the importance of disruptions in lipid metabolism as crucial factors contributing to PCOS. The primary role of proprotein convertase subtilisin/kexin type 9 (PCSK9) lies in its capacity to serve as an inhibitory controller of lipid metabolism by means of its interaction with low-density lipoprotein receptors situated on cellular membranes. [ 16 , 17 ] Cholesterol assumes the role of a precursor in the synthesis of steroid hormones, while mevalonate plays a pivotal role in the biosynthetic pathway of cholesterol. [ 18 ] The transformation of HMG-CoA into mevalonate, facilitated by 3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMGCR), serves as the step that governs the rate in this pathway. [ 19 ] Remarkably, HMGCR can be specifically restrained by substances such as statins, which function as inhibitors of HMGCR. [ 20 ] Consequently, we have deduced that there exists a complex interplay between PCSK9 inhibitors, HMGCR inhibitors and PCOS. The HMGCR inhibitor (HMGCRi) demonstrated a reduction in the synthesis of ovarian androgens within the theca-interstitial cells of rodent ovaries. [ 21 ] Furthermore, the inclusion of simvastatin alongside oral contraceptives resulted in a decrease in circulating testosterone levels among women diagnosed with PCOS. [ 22 , 23 ] And HMGCRi improved insulin sensitivity and markers of inflammation in PCOS patients. [ 23 ] Atorvastatin reduced insulin resistance, hyperandrogenemia, C-reactive protein, and homocysteine in PCOS patients. [ 24 , 25 ] Thus, the utilization of an HMGCR inhibitor may potentially exert its influence on the pathogenesis of PCOS through a diverse range of mechanisms. Endometriosis is correlated with elevated levels of low-density lipoprotein cholesterol (LDL-C), non-high-density lipoprotein cholesterol, and triglycerides, as well as reduced levels of high-density lipoprotein cholesterol. [ 26 , 27 ] Individuals diagnosed with endometriosis face a heightened risk of developing hypercholesterolemia, and there exists a progressively greater likelihood of developing laparoscopically confirmed endometriosis in women affected by hypercholesterolemia. [ 28 ] Diet plays a significant role in endometriosis, impacting various aspects such as inflammation, prostaglandin metabolism, and estrogenic activity. [ 29 ] Therefore, it can be inferred that the utilization of PCSK9 inhibitors (PCSK9i) may potentially demonstrate a significant correlation with endometriosis. A growing body of evidence indicates a link between female infertility and disruptions in blood lipid profiles. Women facing infertility frequently exhibit heightened levels of triglycerides and LDL-C, while experiencing diminished levels of high-density lipoprotein cholesterol when compared to their healthy counterparts. [ 30 ] Henceforth, it follows that dyslipidemia could contribute to the onset and progression of female infertility. Nevertheless, further research is warranted to explore the specific effects of various lipid-lowering medications on female infertility. Within the realm of drug-target Mendelian randomization (MR) analysis, instrumental variables are employed to emulate the pharmacological inhibition of pharmacogenetic targets by leveraging genetic variations. Utilizing regression analysis, this methodology strives to elucidate the enduring impacts of drug utilization and bolster the establishment of causal inferences relating to the potential influence of said drug–gene targets on REDs. [ 31 , 32 ] In the present investigation, we procured recently published summary-level statistics from genome-wide association studies (GWAS) to delve into the causal association between genetically predicted PCSK9i, HMGCRi, and several REDs, namely, PCOS, endometriosis, and female infertility. This endeavor was undertaken through a drug-targeted MR analysis.

Investigation: Na Aru, Congyu Yang, Yuntian Chen. Software: Na Aru. Supervision: Jiaming Liu. Writing – original draft: Na Aru, Congyu Yang, Yuntian Chen. Writing – review & editing: Jiaming Liu.

The analysis utilized LDL-C summary statistics obtained from a genome-wide association study comprising 173,082 participants of European descent ( https://gwas.mrcieu.ac.uk/datasets/ieu-a-300/ ). By acquiring instrumental variables capable of targeting PCSK9 and HMGCR to mitigate LDL-C levels, it becomes feasible to simulate the effects of PCSK9i and HMGCRi. [ 33 ] Our selection criteria focused on single nucleotide polymorphisms located within 100 kilobases upstream or downstream of the PCSK9 and HMGCR gene regions that demonstrated significant associations with LDL-C levels. We implemented a linkage disequilibrium threshold ( r ² < 0.3) to ensure genetic variant independence, ultimately retaining 13 PCSK9-related and 8 HMGCR-related single nucleotide polymorphisms (SNPs) (Table S1, Supplemental Digital Content, https://links.lww.com/MD/P86 ). To verify the robustness of our findings, we replicated the analysis using an independent GWAS dataset containing 201,678 individuals of European-ancestry ( https://gwas.mrcieu.ac.uk/datasets/ieu-b-5089/ ), with results documented in Table S2, Supplemental Digital Content, https://links.lww.com/MD/P86 . Our MR analysis incorporated 3 primary disease outcomes, using coronary heart disease (CHD) as a positive control. All genetic datasets were restricted to European-ancestry populations to minimize population stratification bias. The CHD dataset included summary statistics from 63,746 affected individuals and 130,681 controls ( https://gwas.mrcieu.ac.uk/datasets/ieu-a-9/ ). For PCOS, we employed FinnGen Research’s 2021 release ( https://gwas.mrcieu.ac.uk/datasets/finn-b-E4_POCS/ ), which analyzed 16.4 million genetic variants across 642 clinically confirmed cases (based on ICD-9/10 criteria) and 118,228 unaffected controls. The endometriosis dataset, also from FinnGen ( https://gwas.mrcieu.ac.uk/datasets/finn-b-N14_ENDOMETRIOSIS/ ), comprised 8288 cases versus 68,969 controls. Finally, female infertility data were obtained from FinnGen ( https://gwas.mrcieu.ac.uk/datasets/finn-b-N14_FEMALEINFERT ), including 6481 cases and 68,969 controls of European descent. Given the established clinical use of PCSK9 and HMGCR inhibitors in CHD treatment, we employed CHD-related GWAS data as a reference standard to validate our genetic instruments. The analysis involved 3 key steps: (1) standardization of drug-target-associated genetic variants with outcome measures, (2) implementation of multiple MR approaches including MR-Egger, weighted median, inverse-variance weighted (IVW), simple mode, weighted mode, and MR-PRESSO analyses, and (3) prioritization of IVW results as the primary outcome due to its optimal statistical properties. [ 34 ] To assess potential heterogeneity and pleiotropy in our analyses, we implemented rigorous statistical evaluations. First, we quantified between-variant heterogeneity using Cochran Q statistic derived from both IVW and MR-Egger regression frameworks, considering P  > .05 as evidence of homogeneous genetic effects. Second, we evaluated horizontal pleiotropy through MR-Egger intercept testing, where an intercept term P -value > .05 suggested the absence of significant directional pleiotropic effects. [ 35 ] These complementary approaches provided comprehensive evaluation of key MR assumptions while maintaining appropriate type I error control. [ 35 ] To ensure valid causal inference through MR, our analysis adhered to 3 fundamental assumptions: (1) genetic instruments must strongly associate with the exposure (F-statistic > 10), (2) variants must be independent of known confounders, and (3) effects on outcomes must be mediated exclusively through the specified biological pathway (Fig. 1 ). To address potential pleiotropic bias, we systematically evaluated all SNPs against the GWAS Catalog ( www.ebi.ac.uk/gwas ), excluding those showing associations with secondary phenotypes ( P  < 5 × 10 -8 ). We further validated result robustness through leave-one-out sensitivity testing to identify influential variants. All analyses were implemented using the TwoSampleMR package (v0.5.7) in R (v4.2.1), incorporating established quality control procedures. [ 35 , 36 ] Research overview and design of drug-target Mendelian randomization analysis. CHD = coronary heart disease, HMGCR = 3-hydroxy-3-methylglutaryl-coenzyme A reductase, LDL-C = low-density lipoprotein cholesterol, PCOS = polycystic ovary syndrome, PCSK9 = proprotein convertase subtilis kexin 9, MR = Mendelian randomization, SNP = single nucleotide polymorphisms.

As anticipated, the results obtained through the IVW method revealed a significant reduction in the risk of CHD with the use of PCSK9 inhibitors (OR [95%]: 0.597 [0.467–0.727], P  = 6.57 × 10 -15 ), which aligns with the effect observed with HMGCR inhibitors (IVW: OR [95%]: 0.699 [0.555–0.842], P  = 9.85 × 10 −7 ) (Fig. 2 ). Table S3, Supplemental Digital Content, https://links.lww.com/MD/P86 illustrates the results obtained from MR-Egger, simple mode, weighted mode, and MR-PRESSO analyses. The analysis was also repeated using an alternative GWAS dataset, yielding similar findings (Table S4, Supplemental Digital Content, https://links.lww.com/MD/P86 ). The effect of PCSK9 and HMGCR inhibitor on coronary heart disease and reproductive endocrine diseases. Asterisk (*) represents MR analysis results after excluding SNPs associated with other phenotypes. CHD = coronary heart disease, CI = confidence interval, HMGCR = 3-hydroxy-3-methylglutaryl-coenzyme A reductase, nsnp = number of single nucleotide polymorphisms, OR = odds ratio, PCOS = polycystic ovary syndrome, PCSK9 = proprotein convertase subtilis kexin 9. The genetically predicted HMGCRi exhibited a noticeable protective effect against PCOS (IVW: OR [95%]: 0.321 [0.136–0.757], P  = .009), whereas PCSK9i did not have a significant effect on the risk of PCOS (IVW: P  = .240). Furthermore, the genetically predicted PCSK9i showed a negative correlation with the risk of endometriosis, as evidenced by both the IVW method (OR [95%]:0.791 [0.688–0.909], P  = .001) and the weighted median method (OR [95%]:0.788 [0.665–0.934], P  = .006). Conversely, HMGCRi did not exhibit statistical significance in relation to endometriosis (IVW: P  = .849; weighted median: P  = .722) (Fig. 2 ). The results obtained from other MR analysis methods are presented in Table S3, Supplemental Digital Content, https://links.lww.com/MD/P86 . However, neither PCSK9i nor HMGCRi exhibited a significant association with the risk of female infertility (Table S3, Supplemental Digital Content, https://links.lww.com/MD/P86 and Fig. 2 ). Moreover, we conducted a repeated analysis using an alternative GWAS dataset, and the outcomes yielded similar conclusions (Table S4, Supplemental Digital Content, https://links.lww.com/MD/P86 ). Consistent with MR assumptions, sensitivity analyses revealed no significant heterogeneity (MR-Egger intercept test, P  > .05) or horizontal pleiotropy (MR-PRESSO global test, P  > .05) in the PCSK9i and HMGCRi associations with REDs (Table S5, Supplemental Digital Content, https://links.lww.com/MD/P86 ). These findings were replicated in an independent GWAS dataset (Table S6, Supplemental Digital Content, https://links.lww.com/MD/P86 ), confirming the reliability of our results. Leave-one-out analysis further demonstrated that no single SNP disproportionately influenced the observed causal relationships. To address potential pleiotropic effects, we systematically screened all instrumental variables against the GWAS Catalog, identifying and excluding 4 pleiotropic SNPs associated with secondary traits (Tables S7 and S8, Supplemental Digital Content, https://links.lww.com/MD/P86 ). After their removal, the protective effect of PCSK9i against endometriosis remained robust (IVW OR = 0.97, 95% CI: 0.69–0.91, P  = .001). Similarly, HMGCRi maintained a significant inverse association with PCOS risk (IVW OR = 0.33, 95% CI: 0.14–0.76, P  = .019; Table S9, Supplemental Digital Content, https://links.lww.com/MD/P86 ). These results were further validated in sensitivity analyses using alternative GWAS summary statistics (Table S10, Supplemental Digital Content, https://links.lww.com/MD/P86 ), supporting the stability of our conclusions.

HMGCRi serve as the cornerstone of hyperlipidemia treatment and play a crucial role in both primary and secondary prevention of cardiovascular and cerebrovascular disorders. [ 33 – 37 ] In addition to their primary function of inhibiting HMG-CoA reductase and thereby reducing lipid levels, HMGCRi exert a variety of significant pleiotropic effects. These effects encompass improved endothelial function, increased expression of nitric oxide synthase, stabilization of coronary plaques, regression of atheromas, and diminished inflammation. [ 38 – 41 ] The novel approach of utilizing antibodies to inhibit PCSK9 represents a groundbreaking method for reducing LDL-C levels. Extensive clinical evidence has widely confirmed its effectiveness in promoting both the prevention and treatment of cardiovascular disease. [ 42 , 43 ] Beyond its impact on LDL-C levels, PCSK9i hold promise for exerting pleiotropic effects. These potential effects encompass enhancing the response of tumors to immune checkpoint therapy, suppressing platelet activation and thrombosis, and mitigating cell apoptosis. [ 44 – 46 ] The role of PCSK9i and HMGCRi in inflammation and endocrine function, particularly in relation to the pathogenesis of REDs, has garnered considerable attention in recent times. However, a comprehensive investigation of the causal relationship between PCSK9 and HMCGR inhibitors and the risk of REDs is still lacking. Through a drug-target MR analysis, we have discovered that HMGCRi demonstrates a significant reduction in the risk of PCOS, and PCSK9i may potentially decrease the risk of endometriosis. This finding not only deepens our understanding of HMGCRi and PCSK9i, but also provides essential theoretical guidance for the selection of lipid-lowering strategies. Pleasingly, our findings revealed a discernible protective impact of HMGCRi on PCOS. PCOS, which stands as the predominant endocrinopathy among women, manifests as the presence of polycystic ovaries, chronic anovulation, and excessive androgen levels. [ 47 , 48 ] The management of PCOS primarily focuses on symptomatic relief and encompasses lifestyle interventions as well as pharmacological measures, including the utilization of medications such as Metformin, oral contraceptives, and antiandrogens. [ 49 , 50 ] Regrettably, the management of PCOS poses a formidable challenge, and current interventions are unable to fully address the multitude of consequences associated with this syndrome. Recently, HMGCRi has emerged as a promising addition to the treatment repertoire for PCOS. Antoni J. Duleba et al demonstrated that inflammatory stimuli, specifically LPS and IL-1β, not only elevated androgen production but also instigated extensive alterations in gene expression. [ 51 ] RNA sequencing analysis revealed that both pro-inflammatory agents modulated numerous common genes, including those implicated in cholesterol biosynthesis, androgen biosynthesis, and innate immunity. Prior studies have suggested that inflammatory stimuli may upregulate HMG-CoA reductase-mediated cholesterol synthesis, potentially increasing resistance to HMGCRi. [ 52 , 53 ] These findings underscore a plausible mechanism through which anti-inflammatory treatments, which is HMGCR inhibitors, can effectively lower testosterone levels in women diagnosed with PCOS. [ 54 – 56 ] HMGCRi proves beneficial by diminishing the production of sex steroids, ameliorating dyslipidemia, and curtailing ovarian androgen synthesis through the inhibition of androgen production in thecal cells. [ 21 , 57 ] Studies have substantiated that HMGCRi exerts inhibitory effects on the proliferation of human theca-interstitial cells, both in ovaries affected by PCOS and those unaffected by the condition. [ 58 ] Continued utilization of HMGCRi has demonstrated noteworthy outcomes in ameliorating clinical manifestations and rectifying biochemical abnormalities among patients with PCOS. [ 59 ] Puurunen et al documented that HMGCRi led to elevated levels of fasting insulin and compromised insulin sensitivity over the course of 6 months of treatment. [ 60 ] And Raval et al ascertained the efficacy of HMGCRi in diminishing testosterone levels, whether employed as a standalone therapy or in conjunction with oral contraceptive pills. [ 61 ] Our investigation unveiled a safeguarding influence of HMGCRi against the susceptibility to PCOS, whereas PCSK9i failed to elicit a parallel response. This indicates that HMGCRi potentially mitigates the risk of PCOS via mechanisms unrelated to lipid reduction. The existing body of evidence strongly indicates that HMGCRi could potentially play a constructive role in diminishing circulating androgen levels among individuals with PCOS. A conceivable mechanism for the antiandrogenic properties of HMGCRi involves impeding the proliferation of human theca-interstitial cells, a phenomenon observed in both normal and PCOS-affected ovaries. [ 58 ] Previous studies have revealed the protective role of HMGCRi in other diseases, but further studies need to reveal their protective role in PCOS. Furthermore, our study has revealed that PCSK9i is associated with a diminished risk of endometriosis. Melo et al reported that over 50% of individuals afflicted with endometriosis experience dyslipidemia. [ 26 ] Inflammation plays a pivotal role in the pathogenesis of endometriosis. Research conducted on murine models has demonstrated that a high-fat diet significantly exacerbates endometriosis lesions, resulting in elevated levels of pro-inflammatory cytokines and oxidative stress. [ 62 ] Emerging evidence establishes that elevated triglyceride levels drive endometriosis progression through multifaceted pro-inflammatory mechanisms. Chronic hypertriglyceridemia induces macrophage activation and subsequent cytokine release, [ 63 ] which not only directly stimulates abnormal proliferation and invasion of endometrial cells but also initiates self-perpetuating inflammatory cascades that promote ectopic lesion formation. [ 64 ] At the vascular interface, triglyceride-mediated endothelial dysfunction [ 65 ] enhances leukocyte adhesion and systemic inflammation, while in adipose tissue, lipid-accumulated immune cells secrete prostaglandins and leukotrienes [ 66 ] that further disrupt hormonal signaling and immune homeostasis. This interconnected pathophysiology – spanning metabolic dysregulation, vascular dysfunction, and immune-endocrine cross-talk – provides a unified mechanistic explanation for the clinical association between dyslipidemia and endometriosis severity, suggesting potential therapeutic targets across these interrelated biological systems. The pro-inflammatory lipid mediators associated with elevated triglycerides not only exacerbate systemic inflammation but may also directly contribute to the severity of endometriosis symptoms, including chronic pelvic pain. Our findings demonstrate that PCSK9 inhibitors are associated with reduced endometriosis risk, while showing no such association for HMGCR inhibitors – suggesting distinct mechanistic pathways for these lipid-lowering therapies. Although the precise mechanisms remain to be fully elucidated, PCSK9i may exert protective effects through multiple potential pathways: (1) modulation of endometrial cell proliferation and apoptosis, (2) inhibition of ectopic cell migration and adhesion, and (3) attenuation of lipid-mediated inflammatory cascades. These observations position PCSK9 inhibition as a promising therapeutic strategy for endometriosis, particularly in patients with concurrent dyslipidemia. However, the dissociation between PCSK9i and HMGCRi effects underscores the need for further mechanistic studies, including in vitro investigations of endometrial stromal cells, animal models of lipid-inflammation–endometriosis interactions, and clinical trials evaluating PCSK9i’s potential to alleviate endometriosis symptoms. This research direction may yield novel insights into the metabolic-inflammatory axis of endometriosis pathogenesis while opening new therapeutic possibilities for this challenging condition. Our study did not establish a causal relationship between the inhibition of PCSK9 and HMGCR inhibitors and the risk of female infertility. Elevated concentrations of 4 lipid components in women were correlated with prolonged time-to-pregnancy. [ 67 ] Conversely, women exhibiting lower levels of total cholesterol, low-density lipoprotein cholesterol, and triglycerides demonstrated higher rates of pregnancy and live births. [ 67 ] Furthermore, Cai’s investigation uncovered an independent association between rising LDL-C levels and a decreased likelihood of ovulation, clinical pregnancy, and live birth. [ 68 ] Additionally, insufficient lipid concentrations prior to pregnancy were linked to reduced fertility in women. [ 67 ] The relationship between PCSK9/HMGCR inhibitors and female infertility remains complex and incompletely understood. While our MR analysis found no significant associations, conflicting evidence exists in the literature, with some studies suggesting protective effects of lipid-lowering therapies [ 69 ] while others show no LDL-C association. [ 70 ] This discrepancy may reflect the dual nature of these monoclonal antibody therapies: while potentially beneficial for metabolic aspects of infertility (particularly in PCOS/endometriosis through lipid pathway modulation), their immunogenic properties could theoretically impair fertility via antibody-mediated immune responses affecting reproductive tissues. The current evidence suggests that any influence of these inhibitors on female fertility likely involves a delicate balance between their metabolic benefits (reducing lipid-driven inflammation and hormonal dysregulation) and potential immunological consequences (disrupting implantation tolerance or ovarian function). This complexity underscores the critical need for targeted research, including mechanistic studies evaluating direct effects on reproductive tissues and carefully designed clinical trials assessing fertility outcomes in women receiving these therapies, particularly those with preexisting reproductive endocrine disorders. Several important limitations of our research warrant consideration. First, while MR provides robust evidence for causal inference, it cannot replace randomized controlled trials in establishing clinical efficacy, as it primarily evaluates the exposure–outcome relationship rather than direct therapeutic effects. This underscores the need for prospective clinical studies to validate the potential role of PCSK9 and HMGCR inhibitor in REDs management. Second, our analysis was restricted to European-ancestry populations due to limited GWAS data availability from other ethnic groups. Given known variations in genetic architecture and disease prevalence across populations, the generalizability of our findings requires verification through future multi-ethnic studies. We strongly recommend that subsequent research incorporate diverse population cohorts to assess potential ethnic-specific effects and ensure broader applicability of these pharmacological interventions. These important extensions will be critical for translating our genetic epidemiological findings into clinically actionable knowledge.

The GWAS summary statistics were obtained from the open-access MRC IEU database ( https://gwas.mrcieu.ac.uk/ ). All analyses were performed using R statistical software (version 4.2.1).