A combined connectivity mapping and pharmacoepidemiology approach to identify existing medications with breast cancer causing or preventing properties

Abstract

Purpose: We applied a novel combined connectivity mapping and pharmacoepidemiological approach to identify medications that alter breast cancer risk.

Methods: The connectivity mapping process identified 6 potentially cancer-causing (meloxicam, azithromycin, rizatriptan, citalopram, rosiglitazone, and verapamil) and 4 potentially cancer-preventing (bendroflumethiazide, sertraline, fluvastatin, and budesonide) medications that were suitable for pharmacoepidemiological investigation. Within the UK Clinical Practice Research Datalink, we matched 45,147 breast cancer cases to 45,147 controls based on age, year, and general practice. Medication use was determined from electronic prescribing records. We used conditional logistic regression to calculate odds ratios (ORs) for the association between medication use and cancer risk after adjustment for comorbidities, lifestyle factors, deprivation, and other medication use.

Results: Bendroflumethiazide was associated with increased breast cancer risk (OR: 1.11; 95% CI: 1.06, 1.15); however the connectivity mapping exercise predicted that this medication would reduce risk. There were no statistically significant associations for any of the other candidate medications, with ever use ORs ranging from 0.93 (95% CI: 0.78, 1.11) for azithromycin to 1.16 (95% CI: 0.99, 1.37) for verapamil.

Conclusions: In this instance, our combined connectivity mapping and pharmacoepidemiological approach did not identify any additional medications that were substantially associated with breast cancer risk. This could be due to limitations in the connectivity mapping, such as implausible dosage requirements, or the pharmacoepidemiology, such as residual confounding.

Keywords: breast cancer; connectivity mapping; drug discovery; pharmacoepidemiology.