Change in prostate tissue gene expression following finasteride or doxazosin administration in the medical therapy for prostatic symptoms (MTOPS) study
Hyo Young Choi; Kathleen C. Torkko; M. Scott Lucia; Khyobeni Mozhui; Won-Young Choi; Peter E. Clark; Jay H. Fowke
Scientific Reports2024
In plain terms
In a large US trial of drugs for an enlarged prostate, some men gave small prostate tissue samples before starting treatment and again later on. Researchers measured which genes were more or less active in these samples for 108 of the men. Compared with a dummy pill, finasteride changed the activity of about 400 genes, mostly turning down genes that respond to male hormones, while doxazosin, a different type of prostate drug, changed almost none. Among men on finasteride, 14 of the 15 whose symptoms stayed under control showed these gene changes, compared with 7 of the 13 whose symptoms got worse.
What it doesn’t show: It can't show whether these gene changes last after finasteride is stopped or happen anywhere other than the prostate: the samples came from older men with enlarged prostates during the trial, and the groups were small.
Summary (paraphrased)
To understand why some patients' symptoms persist despite treatment, the authors profiled gene expression in prostate transition-zone biopsies from 108 MTOPS trial participants before and after treatment. Finasteride produced a distinctive transcriptional signature: 398 genes changed expression relative to placebo (FDR < 0.05), broadly suppressing androgen-response, estrogen-response, and fatty-acid and amino-acid metabolic pathways, while doxazosin altered almost nothing. Crucially, the patients whose gene expression shifted (a definable molecular subgroup) were the ones most likely to respond clinically — nearly all finasteride responders showed the shift, versus only half of non-responders. The study demonstrates in humans that finasteride rewires tissue-level gene expression in a drug-specific way.
Evidence
Extracted from the full text; page numbers refer to the paper. The library’s own notes are labelled as such.
- Design
- Case–control studyNested case-control study within the MTOPS randomised, double-blind, placebo-controlled trial, using RNA sequencing of paired baseline and follow-up prostate biopsies
- Setting
- MTOPS trial (17 US clinical centres) and its NIDDK tissue biorepository; analysis led from the University of Tennessee Health Science Center
- Population
- Men aged 50 or over with moderate to severe urinary symptoms of benign prostatic hyperplasia (AUA score 8–30, low urine flow) in the MTOPS biopsy sub-study (1,198 consented). Within each arm, men whose condition progressed ("resistant" cases) were sampled 1:1 with men who did not ("responsive" controls) of similar age and symptom score. Mean age 62.7.
- Size
- 108 · 108 men with paired RNA-seq (finasteride 28, of whom 15 responsive and 13 resistant; doxazosin 28; combination 15; placebo 37), from 3,047 randomised in MTOPS
- Exposure
- Finasteride 5 mg/day, doxazosin up to 8 mg/day, or both, by random assignment
- Compared with
- Placebo arm for drug effects on gene expression; responsive vs resistant men within each arm
- Outcome
- Within-person change in gene expression (log2 follow-up/baseline) across 16,781 protein-coding genes in transition-zone biopsies; clinical progression as defined by MTOPS (a rise of 4 or more points in AUA score, acute urinary retention, recurrent urinary infection, renal insufficiency or incontinence)
- Follow-up
- Baseline and post-treatment biopsies; the paper does not state when the follow-up biopsy was taken (it cites MTOPS Year 5 tissue for an earlier analysis) and gives the trial duration as 5.5 years
Key results
- Versus placebo (FDR < 0.05), 398 genes changed with finasteride, 4 with doxazosin and 28 with the combination; 70% of the finasteride genes went down. The Discussion instead gives 416 genes for finasteride and 5 for doxazosin (p. 10); 416 is the union across the three drug arms (398 + 4 + 28 - 11 - 3, p. 3), and 5 is the doxazosin count at FDR < 0.1 (Fig. 1a) · p. 3, Fig. 1 (p. 4), p. 10
- Finasteride-changed genes were most enriched for the androgen-response hallmark (FDR = 2.4 × 10⁻¹⁶), then early oestrogen response, and for fatty-acid and amino-acid metabolism pathways, mostly down-regulated · p. 3
- Clustering on the 416 genes split patients into SC1 (clear changes, mostly finasteride or combination) and SC2 (little change, mostly placebo or doxazosin). In the finasteride arm, 14 of 15 responsive men (93.3%) vs 7 of 13 resistant men (53.8%) were SC1 (Fisher p = 0.02); in the combination arm all 5 SC2 men were responsive · pp. 3, 5, Fig. 2
- Using an 84% response rate from MTOPS, the paper estimates 86.8% of finasteride-treated men would be SC1, and that 90.1% of SC1 and 44.5% of SC2 would respond. Recomputing from 14/15, 7/13 and 0.84 gives 87.0%, 90.1% and 43.1%; the paper's values follow from the rounded inputs 0.93 and 0.54 (86.8%, 90.0%, 44.4%). The Discussion says 44.5% of SC2 men "developed clinical resistance" (p. 11), the reverse of p. 6 · pp. 5–6, Fig. 3, p. 11
- Within SC1 finasteride and combination patients (n = 31), larger baseline transition-zone volume was linked to resistance (Wilcoxon p = 0.007; 3.4 times the odds per 23 ml); top genes differing in change between resistant and responsive men were FKBP5 and SLC1A4 (28 genes at unadjusted p < 0.05) · pp. 8–10, Figs 5–6
Limitations the authors note
- Sample size was too small across treatment and response groups to support all subanalyses
- Newer BPH drugs could not be considered
- Bulk RNA-seq may mask differences between cell types
- Responsive vs resistant differences were not significant at FDR < 0.1, so a permissive p < 0.01 screen was used, and changes may be confounded by unknown factors
- Few non-white patients, and the association of race with resistance is unexplained
- Response in the combination arm may reflect doxazosin rather than finasteride
- Results are initial and need confirmation
Also worth weighing (library’s note)
- Participants were older men with BPH taking 5 mg/day; "resistance" means progression of urinary disease, not adverse effects
- No tissue was sampled after finasteride was stopped and no epigenetic marks were measured, so persistence after stopping cannot be assessed
- Case-control sampling over-represents progression (13 of 28 in the finasteride arm against an assumed 16% in the trial, p. 5), so within-sample proportions are not trial rates; the combination arm ended with 10 responsive and 5 resistant despite 1:1 sampling (Table 1)
- SC1 and SC2 were found by clustering on genes selected from the same patients, with no independent validation set
- Internal inconsistencies include 497 (p. 7, Fig. 4) vs 470 (p. 11) genes differing at baseline between SC1 and SC2, and FKBP5 and SLC1A4 described as both increasing and decreasing in responsive men (p. 8)
- Combination-arm estimates (53.3%, 86.1%, 98.2%; the Fig. 3b legend also says 9.1% resistant in SC2) do not follow from the stated inputs 0.5, 1 and 0.92, which give 54.0%, 85.2% and 100% (p. 6)
Funding: Not stated (MTOPS itself was NIDDK-sponsored, and tissue was used with NIDDK approval) Interests: None declared
What it can support (library’s note): That finasteride, unlike doxazosin, changes gene activity in the prostate transition zone of men with BPH during treatment, and that this change tracks clinical response. It cannot show whether such changes persist after stopping, occur in other tissues, or relate to adverse effects.
Why it’s in the corpus
Human evidence that finasteride, unlike doxazosin, changes gene expression in the prostate tissue of men with BPH during treatment, a transcriptional effect relevant to mechanism hypotheses for post-drug syndromes. It has no samples taken after stopping and no epigenetic measurements, so it does not show persistence, and its "resistance" means BPH progression, not adverse effects.
Related records
- Peer-reviewed paperPSSD Discord pick
Analysis of the finasteride treatment and its withdrawal in the rat hypothalamus and hippocampus at whole-transcriptome level
S Giatti, L Cioffi, S Diviccaro, R Piazza, R C Melcangi · Journal of Endocrinological Investigation
Whole-transcriptome (RNA-seq) analysis of hypothalamus and hippocampus in adult male rats treated with finasteride for 20 days and examined 24 h after last dose and 1 month after withdrawal. Treatment altered 186 genes in hypothalamus and 19 in hippocampus…
DISC-0122024PFSPSSD Discord picks