Menu Back to Poster-Presentations-Details

W-01: A Real World Investigation of Finasteride and the Risk of Prostate Cancer

Poster Presenter

Stephan Toomas Palm

Analyst, Clinical Sciences
Trinetx
United States

Objectives

Following clinical trial analyses of the association between finasteride and prostate cancer, this research explored the risk of developing prostate cancer with finasteride as compared to first-line treatment with an alpha-blocker in the real world using electronic medical records (EMR).

Method

EMRs from 39 healthcare organizations were analyzed using Trinetx to isolate 2 cohorts of patients with finasteride or tamsulosin after Jan. 2000. Prostate cancer outcomes were compared over a 10-year observation period after propensity scoring for age, diagnosis history, and baseline PSA levels.

Results

Within the dataset of 47 million patients, 813,808 had a diagnosis of benign prostatic hyperplasia (BPH). 230,184 patients had records of finasteride (F), and 777,602 patients had records of tamsulosin (T). Of these, 2,659 finasteride patients (with 95% of clinical facts from 2001 to 2018) and 11,890 tamsulosin patients (with 95% of clinical facts from 2003 to 2018) were identified with at least 3 years of medication duration, no prostate cancer prior to the medication use, and no documentation of the comparator medication. With the date of first medication as the index event, the medical histories of the cohorts were analyzed prior to this date. Cohorts were found to share similar characteristics, though they differed in age at the index event (T: M = 64.7 yrs., SD = 9.7; F: M = 68.0 yrs., SD = 9.7), pre-existing PSA values (T: M = 1.9 ng/mL, SD = 1.9; F: M = 2.7 ng/mL, SD = 2.6), and prevalence of some pre-existing diagnoses. Cohorts were propensity score matched 1:1 on these differences as well as specific diagnoses in pre-determined therapeutic areas (genitourinary, metabolic, cardiovascular, neoplastic, and mental) using a nearest neighbor greedy matching algorithm with a caliper of 0.25 times the SD. 2,636 matched patients were entered into an outcomes analysis for prostate cancer within 10 years following the index event. 3.9% of patients in the finasteride cohort and 7.1% patients in the tamsulosin cohort were identified with a diagnosis of prostate cancer within the observation period. RR = 0.55 RD = -3.2%, p < 0.0001, NNT = 31.

Conclusion

Finasteride has long been studied for its effect on the development and on the detection of prostate cancer. The Prostate Cancer Prevention Trial, published in 2003, found a decreased risk of prostate cancer with finasteride compared to placebo, but these results were found under arguably artificial clinical trial conditions, including only patients with normal PSA values at baseline. These findings have not been validated under real-world conditions prior to this study. Using electronic medical records, patients with finasteride were found to have a significantly lower risk of being diagnosed with prostate cancer than those with an alpha-blocker commonly prescribed as first-line treatment of BPH (tamsulosin). Real-world data provide a robust signal of a potential protective effect of finasteride against the development of prostate cancer. Additional Author: Manfred Stapff, MD, PhD