Men with an enlarged prostate (otherwise known as benign prostatic hyperplasia or BPH) commonly use a class of drugs called 5-alpha reductase inhibitors (5αRIs). 5αRIs reduce the size of enlarged prostates, alleviate the symptoms of BPH, and decrease the need for surgery. However, the hormonal changes associated with their use may alter their risk of anaemia (low levels of red blood cells or haemoglobin in the blood). To our knowledge, this potential drug side effect has not been thoroughly investigated in a real-world setting. We will, therefore, study this research question using the Clinical Practice Research Datalink (CPRD). Our study population will include all patients with a diagnosis for BPH. These patients will be classified as those treated with 5αRIs and those treated with α-blockers (an alternative therapy for BPH). We will then compare the number of men with anaemia between each treatment group. We will thus be able to determine if 5αRI use increases the chances of having anaemia compared with the α-blocker use among men with BPH.
5αRIs reduce prostate size among men with BPH by lowering levels of dihydrotestosterone, an androgen that promotes prostate growth. Previous studies have reported that decreased androgen levels are associated with an increased risk of anaemia; however, the anaemia effects of 5αRIs have not been previously assessed in a real-world setting. We will thus conduct a population-based retrospective cohort study using the CPRD to compare the risk of anaemia of 5αRIs versus that of α-blockers. We will identify all patients with incident BPH in the CPRD from April 1998 to March 2019. Using an active comparator, new-user approach, patients will enter the cohort on their first prescription for a 5αRI or an α-blocker. In our primary analysis, we will use an as-treated exposure definition and will follow patients until the development of anaemia, discontinuation of the study drug, initiation of the other study drug or combination therapy, or administrative censoring. Secondary outcomes will include different World Health Organization (WHO)-defined categories of anaemia severity: mild (129-110g/l), moderate (109-80g/l), and severe (<80g/l) anaemia. We will use Cox proportional hazards models with high-dimensional propensity scores and an inverse probability of treatment weighting to estimate the hazard ratio and 95% CIs for anaemia with 5αRIs versus α-blockers. In secondary analyses, we will examine the risk by molecule, by the duration of use, and whether any observed increased risk is reversible.
Health Outcomes to be Measured:
Anaemia (overall and by severity)