Prostate cancer specific mortality after 5α-reductase inhibitors medication in benign prostatic hyperplasia patients: systematic review and meta-analysis

Figures & data

Figure 1. Flow chart.

Table 1. Baseline characteristics.

Author	Study period	Area	Number of patients	Actual number of patients	Number of case/control	Kind of 5-ARI	Inclusion criteria	Use of claim data	Study type
Sarkar et al. 2019 [26]	180mo.	Western	80875	80875	8587/72288	finasteride, dutasteride	Patients who were diagnosed as prostate cancer during study period	1	Retrospective cohort
Azoulay et al. 2015	120mo.	Western	13892	13892	574/13318	finasteride, dutasteride	Patients who were diagnosed as prostate cancer during study period	1	Retrospective cohort
Pinsky et al. 2013	36mo	Western	8231	6706	3299/3407	dutasteride	Male who had 1 negative prostate biopsy with PSA 3-10ng/ml (2.5-10ng/ml in under 60 years)	0	Prospective cohort
Rompay et al. 2019	240mo.	Western	24012	16248	4571/11677	finasteride, dutasteride	Over 40 years male who newly started 5-ARI	1	Retrospective cohort
Wallner et al. 2016 [10]	180mo.	Western	174895	174895	25388/149507	not mentioned	Over 50 years male who were prescipted medication for BPH/LUTS	1	Retrospective cohort
Murtola et al. 2017	96mo.	Western	80454	78615	9316/69299	finasteride, dutasteride	Over 55 years male who newly started 5-ARI	1	Retrospective cohort
Milonas et al. 2016	36mo.	Western	220	191	85/106	dutasteride	40-80 years male who had suspected prostate cancer due to elevated PSA or abnormal DRE	0	RCT
Murtola et al. 2013	168mo.	Western	1315	1016	115/901	finasteride or dutasteride	Patients who were get radical prostatectomy during study period	0	RCT
Murtola et al. 2016	36mo.	Western	6537	4209	908/3301	finasteride or dutasteride	Patients who were diagnosed as prostate cancer during study period	0	RCT
Thompson et al. 2013 [3]	40mo.	Western	18880	18880	9423/9457	finasteride	Male who had PSA ≤ 3ng/ml	0	RCT
Kjellman et al. 2012	204mo.	Western	3791	3005	199/2806	finasteride	Male who were prescripted 5-ARI or a-blocker	1	Retrospective cohort

Figure 2. Forest plot diagram showing the effect of 5α-reductase inhibitors on cancer specific mortality and all-causes mortality.

Table 2. Effects of moderators on cancer specific mortality and all-causes mortality.

	Cancer specific mortality							All-causes mortality
Variables	k	Coef.*	HR	95% CI Low	95% CI High	P^†	P^‡	k	Coef.*	HR	95% CI Low	95% CI High	P^†	P^‡
No. of patients	6	<0.001	–	<-0.001	<0.001	–	0.818	6	<0.001	–	<0.001	<0.001	–	0.022
Age	6	−0.018	–	−0.072	0.036	–	0.516	6	−0.005	–	−0.034	0.025	–	0.757
Study duration (months)	6	0.002	–	−0.001	0.005	–	0.276	6	0.001	–	−0.001	0.002	–	0.215
Follow up (years)	5	0.032	–	−0.134	0.198	–	0.704	5	0.055	–	−0.205	0.315	–	0.679
Types of 5-ARI						0.953	–						0.775	–
Finasteride	1	–	0.930	0.759	1.139			1	–	1.020	0.967	1.076
Finasteride or Dutasteride	4	–	0.959	0.673	1.365			4	–	1.027	0.891	1.185
Dutasteride								1	–	0.930	0.800	1.081
No agents name	1	–	0.850	0.718	1.007
Study design						0.555	–						0.992	–
RCT	1	–	0.968	0.739	1.268			2	–	1.019	0.966	1.075
Cohort	5	–	0.790	0.603	1.035			4	–	1.005	0.882	1.145
Claim data		–				0.555	–		–				0.434	–
Yes	5	–	0.968	0.739	1.268			3	–	1.034	0.884	1.209
No	1	–	0.790	0.603	1.035			3	–	1.008	0.959	1.060
Medication after PC diagnosis		–				0.561	–		–				0.813	–
Yes	3	–	0.994	0.665	1.486			3	–	0.990	0.810	1.210
No	3	–	0.878	0.774	0.995			3	–	1.012	0.963	1.064
Additional medication		–				0.214	–		–				0.005	–
Yes	2	–	1.064	0.612	1.849			2	–	0.984	0.918	1.055
No	4	–	0.873	0.783	0.973			4	–	1.101	1.052	1.151

k, number of observations. * Regression coefficient. † P-value of meta-ANOVA for categoriacal moderators. ‡ P-value of random effect meta-regression using restricted maximum likelihood (REML) for countinuous moderators .

Figure 3. Funnel plots for publication bias.

Figure 4. Traffic light plot on Risk of bias.  

Table 3. Assessment of methodological quality.  

Bias domain	Bias due to confounding	Bias due to selection of participants	Bias in classification of inteventions	Bias due to deviations from intented intervetions	Bias due to missing data	Bias due to measurement of outcomes	Bias in selection of the reported results
Evaluation standard for each bias domain	Low :	Low :	Low :	Low :	Low :	Low :	Low :
	Matched control, adjusted HR	Participants' PSA value and cancer risk are displayed.	Because all studies are on 5-ARI.	Single center	Claimed data	HR is calculated, median follow up period is presented	HR, survival rate and cause of death are presented
	Moderate :	Moderate :		Moderate :	Moderate :	Moderate :	Moderate :
	i) Matched control, unadjusted HR ii) Unmatched control, adjusted HR	PSA value is not displayed		Multicenter, claimed data	Not claimed data, description of the follow up method	Only HR No HR, median follow up period and survival rate are presented	HR isn’t presented
	Serious :	Serious :			Serious :	Serious :	Serious :
	Unmatched control, no HR	No description of selection			Not claimed data, no description of the follow up method	HR, median follow up period and survival rate aren’t presented	HR and survival rate aren’t presented
	Critical :
	No control
Sarkar 2019	Moderate	Low	Low	Moderate	Low	Low	Low
	(Unmatched control, adjusted HR)	(Participants' PSA value and cancer risk are displayed)		(Claimed data)	(Claimed data)	(HR, median follow up)	(HR, survival rate)
Azoulay 2015	Moderate	Low	Low	Moderate	Low	Low	Low
	(Unmatched control, adjusted HR)	(Participants' PSA value and cancer risk are displayed)		(Claimed data)	(Claimed data)	(HR, median follow up)	(HR, survival rate)
Pinsky 2013	Serious	Low	Low	Moderate	Moderate	Moderate	Moderate
	(Unmatched control, no HR)	(Participants' PSA value and cancer risk are displayed)		(PCPT, REDUCE trial data)	(Not claimed data, description of the follow up method)	(Survival rate is presented, No HR and median follow up period)	(survival rate)
Rompay 2019	Moderate	Moderate	Low	Moderate	Low	Low	Low
	(Unmatched control, adjusted HR)	(PSA value is not displayed)		(Claimed data)	(Claimed data)	(HR, median follow up)	(HR, survival rate)
Wallner 2016	Low	Moderate	Low	Moderate	Low	Low	Low
	(Matched control, adjusted HR)	(PSA value is not displayed)		(Claimed data)	(Claimed data)	(HR, median follow up)	(HR, survival rate)
Murtola 2017	Serious	Moderate	Low	Moderate	Low	Moderate	Low
	(Unmatched control, Unadjusted HR)	(PSA value is not displayed)		(Claimed data)	(Claimed data)	(Only HR)	(HR, survival rate)
Milonas 2016	Serious	Low	Low	Low	Serious	Moderate	Moderate
	(Unmatched control, no HR)	(Participants' PSA value and cancer risk are displayed)		(Single center)	(Not claimed data, no description of the follow up method)	(Only survival rate)	(survival rate)
Murtola 2013	Moderate	Low	Low	Low	Moderate	Moderate	Low
	(Unmatched control, adjusted HR)	(Participants' PSA value and cancer risk are displayed)		(Single center)	(Not claimed data, description of the follow up method)	(Only HR)	(HR, survival rate)
Murtola 2016	Moderate	Low	Low	Low	Moderate	Low	Low
	(Unmatched control, adjusted HR)	(Participants' PSA value and cancer risk are displayed)		(Single center)	(Not claimed data, description of the follow up method)	(HR, median follow up)	(HR, survival rate)
Thompson 2013	Moderate	Moderate	Low	Moderate	Moderate	Moderate	Low
	(Unmatched control, adjusted HR)	(PSA value is not displayed)		(PCPT trial data)	(Not claimed data, description of the follow up method)	(Only HR)	(HR, survival rate)
Kjellman 2012	Serious	Moderate	Low	Moderate	Low	Low	Low
	(Unmatched control, Unadjusted HR)	(PSA value is not displayed)		(Claimed data)	(Claimed data)	(HR, median follow up)	(HR, survival rate)

Sarkar RR, Parsons JK, Bryant AK, et al. Association of treatment with 5α-reductase inhibitors with time to diagnosis and mortality in prostate cancer. JAMA Intern Med. 2019;179(6):812–819.

 PubMedGoogle Scholar

Wallner LP, DiBello JR, Li BH, et al. 5-alpha reductase inhibitors and the risk of prostate cancer mortality in men treated for benign prostatic hyperplasia. Paper Presented at: Mayo Clinic Proceedings. 2016;91(12):1717–1726.

 PubMedGoogle Scholar

Thompson IM, Goodman PJ, Tangen CM, et al. The influence of finasteride on the development of prostate cancer. N Engl J Med. 2003;349(3):215–224.

 PubMed Web of Science ®Google Scholar