EACR25-2164

Investigating the repurposing potential of cardiometabolic and hormonal medications for breast and ovarian cancer treatment using Mendelian randomisation

T. Bate1,2, K. Smith-Byrne3, R. Martin2,4, P. Haycock1,2, J. Yarmolinsky1,2,5
1University of Bristol, Medical Research Council Integrative Epidemiology Unit, Bristol, United Kingdom
2University of Bristol, Population Health Sciences, Bristol Medical School, Bristol, United Kingdom
3University of Oxford, Cancer Epidemiology Unit, Nuffield Department of Population Health, Oxford, United Kingdom
4University Hospitals Bristol and Weston NHS Foundation Trust and the University of Bristol, NIHR Bristol Biomedical Research Centre, Bristol, United Kingdom
5Imperial College London, Department of Epidemiology and Biostatistics, School of Public Health, London, United Kingdom
Introduction:

Approved cardiometabolic and hormonal medications targeting key hallmarks of cancer represent potential repurposing opportunities for cancer treatment. We aimed to investigate the repurposing potential of approved anti-hypertensive, lipid-lowering, anti-adiposity and anti-oestrogenic medications for breast and ovarian cancer treatment using the causal inference method Mendelian randomisation (MR).

Material and method:

We obtained or constructed germline genetic instruments from GWAS to proxy cardiometabolic traits (body mass index (BMI), low density lipoprotein cholesterol (LDL-c), systolic blood pressure (SBP)), and the following medication targets: anti-hypertensive (angiotensin-converting enzyme (ACE), beta-1 adrenergic receptor (ADRB1), sodium-chloride symporter (NCC)), lipid-lowering (3-hydroxy-3-methylglutaryl coenzyme A reductase (HMGCR), Niemann-Pick C1-Like 1 (NPC1L1), proprotein convertase subtilisin/kexin type 9 (PCSK9)), anti-adiposity (gastric inhibitory polypeptide receptor (GIPR)) or anti-oestrogenic (cytochrome P450 Family 19 Subfamily A Member 1 (CYP19A1)). The genetically proxied effects of these traits or targets on breast and ovarian cancer mortality were investigated using summary-level MR. Current indications for these medications, for example risk of coronary artery disease, stroke, type 2 diabetes and elevated SBP, were employed as positive control outcomes to assess validity of the genetic instruments. To assess risk of collider bias, summary-level MR was performed with cancer incidence outcomes and where required, adjustments for collider bias are ongoing.

Result and discussion:

Genetically proxied BMI lowering reduced breast cancer mortality (HR per SD decrease BMI: 0.89; 95% CI: 0.82-0.97), whilst genetically proxied HMGCR inhibition (HR per SD decrease LDL-c: 1.81, 95% CI: 1.15-2.83) and PCSK9 inhibition (HR per SD decrease LDL-c: 1.51, 95% CI: 1.05-2.18) increased risk of breast and ovarian cancer mortality, respectively. There was no strong evidence to support effects of any other traits or targets investigated. In general, there was evidence supporting effects of genetically proxied targets on their respective current indications, for example evidence supporting genetically proxied HMGCR inhibition decreasing risk of coronary artery disease (OR per SD decrease LDL-c: 0.67, 95% CI: 0.56-0.80).

Conclusion:

Aside from a protective effect of lower BMI on breast cancer survival, we did not observe strong evidence to support repurposing of anti-hypertensive, lipid-lowering, anti-adiposity or anti-oestrogenic medications for breast or ovarian cancer treatment. General limitations of applying MR to cancer prognosis outcomes including limited power of analyses, low heritability of cancer survival and treatment effects, in addition to collider bias may have affected analyses.