Mendelian randomization: a novel approach for the prediction of adverse drug events and drug repurposing opportunities

Using genetic data to strengthen causal inference in observational research

Causal inference is essential across the biomedical, behavioural and social sciences.By progressing from confounded statistical associations to evidence of causal relationships, causal inference can reveal complex pathways underlying traits and diseases and help to prioritize targets for intervention. Recent progress in genetic epidemiology - including statistical innovation, massive genotyped data sets and novel computational tools for deep data mining - has fostered the intense development of methods exploiting genetic data and relatedness to strengthen causal inference in observational research. In this Review, we describe how such genetically informed methods differ in their rationale, applicability and inherent limitations and outline how they should be integrated in the future to offer a rich causal inference toolbox.

Genetic Insights for Drug Development in NAFLD

Drug development is a costly, time-consuming, and challenging endeavour, with only a few agents reaching the threshold of approval for clinical use. Therefore, approaches to more efficiently identify targets that are likely to translate to clinical benefit are required. Interrogation of the human genome in large patient cohorts has rapidly advanced our knowledge of the genetic architecture and underlying mechanisms of many diseases, including nonalcoholic fatty liver disease (NAFLD). There are no approved pharmacotherapies for NAFLD currently. Genetic insights provide a powerful and new approach to infer and prioritise candidate drugs, with such selection avoiding myriad pitfalls, while defining likely benefits. In this review, we discuss the prospects and challenges for the optimal utilisation of genetic findings for improving and accelerating the NAFLD drug discovery pipeline.

GWAS Identifies Risk Locus for Erectile Dysfunction and Implicates Hypothalamic Neurobiology and Diabetes in Etiology

Erectile dysfunction (ED) is a common condition affecting more than 20% of men over 60 years, yet little is known about its genetic architecture. We performed a genome-wide association study of ED in 6,175 case subjects among 223,805 European men and identified one locus at 6q16.3 (lead variant rs57989773, OR 1.20 per C-allele; p = 5.71 × 10⁻¹⁴), located between MCHR2 and SIM1. In silico analysis suggests SIM1 to confer ED risk through hypothalamic dysregulation. Mendelian randomization provides evidence that genetic risk of type 2 diabetes mellitus is a cause of ED (OR 1.11 per 1-log unit higher risk of type 2 diabetes). These findings provide insights into the biological underpinnings and the causes of ED and may help prioritize the development of future therapies for this common disorder.

Using the MR-Base platform to investigate risk factors and drug targets for thousands of phenotypes

Mendelian randomization (MR) estimates the causal effect of exposures on outcomes by exploiting genetic variation to address confounding and reverse causation. This method has a broad range of applications, including investigating risk factors and appraising potential targets for intervention. MR-Base has become established as a freely accessible, online platform, which combines a database of complete genome-wide association study results with an interface for performing Mendelian randomization and sensitivity analyses. This allows the user to explore millions of potentially causal associations. MR-Base is available as a web application or as an R package. The technical aspects of the tool have previously been documented in the literature. The present article is complementary to this as it focuses on the applied aspects. Specifically, we describe how MR-Base can be used in several ways, including to perform novel causal analyses, replicate results and enable transparency, amongst others. We also present three use cases, which demonstrate important applications of Mendelian randomization and highlight the benefits of using MR-Base for these types of analyses.

Using the MR-Base platform to investigate risk factors and drug targets for thousands of phenotypes

Mendelian randomization (MR) uses genetic information to strengthen causal inference concerning the effect of exposures on outcomes. This method has a broad range of applications, including investigating risk factors and appraising potential targets for intervention. MR-Base has become established as a freely accessible, online platform, which combines a database of complete genome-wide association study results with an interface for performing Mendelian randomization and sensitivity analyses. This allows the user to explore millions of potentially causal associations. MR-Base is available as a web application or as an R package. The technical aspects of the tool have previously been documented in the literature. The present article is complimentary to this as it focuses on the applied aspects. Specifically, we describe how MR-Base can be used in several ways, including to perform novel causal analyses, replicate results and enable transparency, amongst others. We also present three use cases, which demonstrate important applications of Mendelian randomization and highlight the benefits of using MR-Base for these types of analyses.

Association of common gene variants in glucokinase regulatory protein with cardiorenal disease: A systematic review and meta-analysis

Background

Small-molecules that disrupt the binding between glucokinase and glucokinase regulatory protein (GKRP) in the liver represent a potential new class of glucose-lowering drugs. It will, however, take years before their effects on clinically relevant cardiovascular endpoints are known. The purpose of this study was to estimate the effects of these drugs on cardiorenal outcomes by studying variants in the GKRP gene (GCKR) that mimic glucokinase-GKRP disruptors.

Methods

The MEDLINE and EMBASE databases were searched for studies reporting on the association between GCKR variants (rs1260326, rs780094, and rs780093) and coronary artery disease (CAD), estimated glomerular filtration rate (eGFR), and chronic kidney disease (CKD).

Results

In total 5 CAD studies (n = 274,625 individuals), 7 eGFR studies (n = 195,195 individuals), and 4 CKD studies (n = 31,642 cases and n = 408,432 controls) were included. Meta-analysis revealed a significant association between GCKR variants and CAD (OR:1.02 per risk allele, 95%CI:1.00–1.04, p = 0.01). Sensitivity analyses showed that replacement of one large, influential CAD study by two other, partly overlapping studies resulted in similar point estimates, albeit less precise (OR:1.02; 95%CI:0.98–1.06 and OR: 1.02; 95%CI: 0.99–1.04). GCKR was associated with an improved eGFR (+0.49 ml/min, 95%CI:0.10–0.89, p = 0.01) and a trend towards protection from CKD (OR:0.98, 95%CI:0.95–1.01, p = 0.13).

Conclusion

This study suggests that increased glucokinase-GKRP disruption has beneficial effects on eGFR, but these may be offset by a disadvantageous effect on coronary artery disease risk. Further studies are warranted to elucidate the mechanistic link between hepatic glucose metabolism and eGFR.

Using Human 'Experiments of Nature' to Predict Drug Safety Issues: An Example with PCSK9 Inhibitors

INTRODUCTION

When a new drug enters the market, its full array of side effects remains to be defined. Current surveillance approaches targeting these effects remain largely reactive. There is a need for development of methods to predict specific safety events that should be sought for a given new drug during development and postmarketing activities.

OBJECTIVE

We present here a safety signal identification approach applied to a new set of drug entities, inhibitors of the serine protease proprotein convertase subtilisin/kexin type 9 (PCSK9).

METHODS

Using phenome-wide association study (PheWAS) methods, we analyzed available genotype and clinical data from 29,722 patients, leveraging the known effects of changes in PCSK9 to identify novel phenotypes in which this protein and its inhibitors may have impact.

RESULTS

PheWAS revealed a significantly reduced risk of hypercholesterolemia (odds ratio [OR] 0.68, p = 7.6 × 10^-4) in association with a known loss-of-function variant in PCSK9, R46L. Similarly, laboratory data indicated significantly reduced beta mean low-density lipoprotein cholesterol (- 14.47 mg/dL, p = 2.58 × 10^-23) in individuals carrying the R46L variant. The R46L variant was also associated with an increased risk of spina bifida (OR 5.90, p = 2.7 × 10^-4), suggesting that further investigation of potential connections between inhibition of PCSK9 and neural tube defects may be warranted.

CONCLUSION

This novel methodology provides an opportunity to put in place new mechanisms to assess the safety and long-term tolerability of PCSK9 inhibitors specifically, and other new agents in general, as they move into human testing and expanded clinical use.

Using Mendelian Randomization to Decipher Mechanisms of Bone Disease

PURPOSE OF REVIEW

This review summarizes the basic principles of Mendelian randomization (MR) and provides evidence for the causal effect of multiple modifiable factors on bone outcomes.

RECENT FINDINGS

Several studies using MR approach have provided support for the causal effect of obesity on bone mineral density (BMD). Strikingly, studies have failed to prove a causal association between elevated 25(OH) D concentrations and higher BMD in community-dwelling individuals. The MR approach has been successfully used to evaluate multiple factors related to bone mineral density variation and/or fracture risk. The MR approach avoids some of the classical observational study limitations and provides more robust causal evidence, ensuring bigger success of the clinical trials. The selection of interventions based on genetic evidence could have a substantial impact on clinical practice.

Causal Inference in Cancer Epidemiology: What Is the Role of Mendelian Randomization?

Observational epidemiologic studies are prone to confounding, measurement error, and reverse causation, undermining robust causal inference. Mendelian randomization (MR) uses genetic variants to proxy modifiable exposures to generate more reliable estimates of the causal effects of these exposures on diseases and their outcomes. MR has seen widespread adoption within cardio-metabolic epidemiology, but also holds much promise for identifying possible interventions for cancer prevention and treatment. However, some methodologic challenges in the implementation of MR are particularly pertinent when applying this method to cancer etiology and prognosis, including reverse causation arising from disease latency and selection bias in studies of cancer progression. These issues must be carefully considered to ensure appropriate design, analysis, and interpretation of such studies. In this review, we provide an overview of the key principles and assumptions of MR, focusing on applications of this method to the study of cancer etiology and prognosis. We summarize recent studies in the cancer literature that have adopted a MR framework to highlight strengths of this approach compared with conventional epidemiological studies. Finally, limitations of MR and recent methodologic developments to address them are discussed, along with the translational opportunities they present to inform public health and clinical interventions in cancer. Cancer Epidemiol Biomarkers Prev; 27(9); 995-1010. ©2018 AACR.

Evaluating the potential role of pleiotropy in Mendelian randomization studies

Pleiotropy, the phenomenon of a single genetic variant influencing multiple traits, is likely widespread in the human genome. If pleiotropy arises because the single nucleotide polymorphism (SNP) influences one trait, which in turn influences another ('vertical pleiotropy'), then Mendelian randomization (MR) can be used to estimate the causal influence between the traits. Of prime focus among the many limitations to MR is the unprovable assumption that apparent pleiotropic associations are mediated by the exposure (i.e. reflect vertical pleiotropy), and do not arise due to SNPs influencing the two traits through independent pathways ('horizontal pleiotropy'). The burgeoning treasure trove of genetic associations yielded through genome wide association studies makes for a tantalizing prospect of phenome-wide causal inference. Recent years have seen substantial attention devoted to the problem of horizontal pleiotropy, and in this review we outline how newly developed methods can be used together to improve the reliability of MR.

Risk of Neuropsychiatric Adverse Effects of Lipid-Lowering Drugs: A Mendelian Randomization Study

BACKGROUND

Recent studies have highlighted the possible risk of neuropsychiatric adverse effects during treatment with lipid-lowering medications. However, there are still controversies that require a novel genetic-based approach to verify whether the impact of lipid-lowering drug treatment results in neuropsychiatric troubles including insomnia, depression, and neuroticism. Thus, we applied Mendelian randomization to assess any potential neuropsychiatric adverse effects of conventional lipid-lowering drugs such as statins, proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors, and ezetimibe.

METHODS

A 2-sample Mendelian randomization study was conducted based on summary statistics from genome-wide association studies for lipids, insomnia, depression, and neuroticism. Single-nucleotide polymorphisms located in or near drug target genes of HMGCR, PCSK9, and NPC1L1 were used as proxies for statins, PCSK9 inhibitors, and ezetimibe therapy, respectively. To assess the validity of the genetic risk score, their associations with coronary artery disease were used as a positive control.

RESULTS

The Mendelian randomization analysis showed a statistically significant (P <.004) increased risk of depression after correcting for multiple testing with both statins (odds ratio=1.15, 95% CI: 1.04-1.19) and PCSK9 inhibitor treatment (odds ratio =1.19, 95%CI: 1.1-1.29). The risk of neuroticism was slightly reduced with statin therapy (odds ratio=0.9, 95%CI: 0.83-0.97). No significant adverse effects were associated with ezetimibe treatment. As expected, the 3 medications significantly reduced the risk of coronary artery disease.

CONCLUSION

Using a genetic-based approach, this study showed an increased risk of depression during statin and PCSK9 inhibitor therapy while their association with insomnia risk was not significant.

A critical appraisal of pharmacogenetic inference

In essence, pharmacogenetic research is aimed at discovering variants of importance to gene-treatment interaction. However, epidemiological studies are rarely set up with this goal in mind. It is therefore of great importance that researchers clearly communicate which assumptions they have had to make, and which inherent limitations apply to the interpretation of their results. This review discusses considerations of, and the underlying assumptions for, utilizing different response phenotypes and study designs popular in pharmacogenetic research to infer gene-treatment interaction effects, with a special focus on those dealing with of clinical effects of drug treatment.

Bone mineral density and risk of type 2 diabetes and coronary heart disease: A Mendelian randomization study

Background: Observational studies have demonstrated that increased bone mineral density is associated with a higher risk of type 2 diabetes (T2D), but the relationship with risk of coronary heart disease (CHD) is less clear. Moreover, substantial uncertainty remains about the causal relevance of increased bone mineral density for T2D and CHD, which can be assessed by Mendelian randomisation studies.  Methods: We identified 235 independent single nucleotide polymorphisms (SNPs) associated at p<5×10 ^-8 with estimated heel bone mineral density (eBMD) in 116,501 individuals from the UK Biobank study, accounting for 13.9% of eBMD variance. For each eBMD-associated SNP, we extracted effect estimates from the largest available GWAS studies for T2D (DIAGRAM: n=26,676 T2D cases and 132,532 controls) and CHD (CARDIoGRAMplusC4D: n=60,801 CHD cases and 123,504 controls). A two-sample design using several Mendelian randomization approaches was used to investigate the causal relevance of eBMD for risk of T2D and CHD. In addition, we explored the relationship of eBMD, instrumented by the 235 SNPs, on 12 cardiovascular and metabolic risk factors. Finally, we conducted Mendelian randomization analysis in the reverse direction to investigate reverse causality. Results: Each one standard deviation increase in genetically instrumented eBMD (equivalent to 0.14 g/cm ²) was associated with an 8% higher risk of T2D (odds ratio [OR] 1.08; 95% confidence interval [CI]: 1.02 to 1.14; p=0.012) and 5% higher risk of CHD (OR 1.05; 95%CI: 1.00 to 1.10; p=0.034). Consistent results were obtained in sensitivity analyses using several different Mendelian randomization approaches. Equivalent increases in eBMD were also associated with lower plasma levels of HDL-cholesterol and increased insulin resistance. Mendelian randomization in the reverse direction using 94 T2D SNPs or 52 CHD SNPs showed no evidence of reverse causality with eBMD. Conclusions: These findings suggest a causal relationship between elevated bone mineral density with risks of both T2D and CHD.