571 692
1.
Introduction
Alpha-blockers (ABs) and 5-alpha reductase inhibitors (5-
ARIs) have an established role in treating lower urinary tract
syptoms (LUTS) attributed to benign prostatic hyperplasia
(BPH)
[1–6]. Recently, a new AB and drugs in other classes
approved for different indications have shown promise in
this setting. The purpose of our reviewwas to determine the
comparative effectiveness and safety of medications newly
used in the last 10 yr for LUTS attributed to BPH, both as
single agents and in combination.
2.
Evidence acquisition
2.1.
Protocol
We developed an a priori written protocol, together with a
technical report that incorporated input from key stake-
holders, a multidisciplinary Technical Expert Panel, and
public comment (available at the Agency for Healthcare
Research and Quality [AHRQ] website
http://effectivehealth care.ahrq.gov/index.cfm/search-for-guides-reviews-and- reports/?productid=2067&pageaction=displayproduct).
2.2.
Eligibility criteria
Based on our Population, Interventions, Comparisons,
Outcomes, Timing, and Setting criteria (Supplementary
data) we included randomized controlled trials (RCTs) that
tested comparative effectiveness of treatments involving
newer drugs in men aged 45 yr with LUTS attributed to
BPH. We defined these newer drugs as those that were Food
and Drug Administration (FDA) approved for BPH since
2008 or which, though not FDA approved for BPH, have been
studied for the treatment of BPH since 2008 and were
selected through a formal process of stakeholder involve-
ment (Supplementary data). Comparators included medi-
cations FDA approved for BPH before 2008. Included RCTs
were at least 1 mo in duration with no minimum sample
size. We additionally searched for large (
n
100 patients),
longer-term ( 1 yr duration) observational studies to
assess long-term or rare treatment associated harms only.
We limited inclusion to English language articles.
The primary predefined outcomes of interest were
changes reflecting clinically important differences (Supple-
mentary data) in validated measures to assess LUTS
(International Prostate Symptom Score [I-PSS]: score ranges
0–35 with higher scores indicating more severe symptoms;
or American Urological Association Symptom Index scores),
prostate-related bother or quality of life (QoL; I-PSS QoL
question; BPH/LUTS impact scale), as well as rates of disease
progression and/or treatment failure (prevention/delay of
need for surgical intervention; acute urinary retention
[AUR]). We also assessed common and serious medication
adverse effects (AEs).
2.3.
Information sources and literature search
We searched Ovid Medline, Ovid Embase, and the
Cochrane Central Register of Controlled Trials with filters
for study design (Supplementary data), to identify
relevant RCTs published through June 20, 2016. We also
searched for relevant systematic reviews and other key
references. Lastly, we searched the Clinical Trials
( www. clinicaltrials.gov) and the FDA
( www.fda.gov/Drugs )websites to identify additional completed and ongoing
studies.
2.4.
Study selection process, data extraction, and risk of bias in
studies
Two independent investigators screened titles and
abstracts to identify studies meeting the eligibility criteria.
Data were extracted by one investigator and reviewed and
verified for accuracy by a second investigator. Risk of bias
(RoB) of eligible studies was assessed using AHRQ
guidance by one investigator and reviewed by a
second
[7].
2.5.
Synthesis of results
We assessed clinical and methodological heterogeneity and
variation in effect size to determine the appropriateness of
pooling data
[8]. When three or more trials reported similar
comparisons and outcomes, data were pooled using a
Hartung, Knapp, Sidik, and Jonkman method
[9]random
effects model for proportion of I-PSS responders or mean
changes in I-PSS scores in Stata (StataCorp., College Station,
TX, USA). We pooled other outcomes in RevMan (RevMan,
Spartanburg, SC, USA)
[10]and converted DerSimonian-
Laird random effects confidence intervals to Hartung,
Knapp, Sidik, and Jonkman confidence intervals using
an excel spreadsheet provided in Inthout et al
[9]. We
assessed between study variance with Tau
2
and measured
the magnitude of heterogeneity with the
I
2
statistic. If
substantial heterogeneity was present (ie,
I
2
70%), we
stratified the results to assess treatment effects based on
patient or study characteristics and/or explored sensitivity
analyses
[8,11]. We pooled across different ABs unless
there were at least three trials for a given agent. We
interpreted efficacy and comparative effectiveness using
established thresholds indicating clinical significance
(Supplementary data).
For the body of evidence from RCTs, we rated our
confidence in the estimates of effect for the primary
outcomes as overall strength of evidence (SOE) as high,
moderate, low, or insufficient (Supplementary data)
[12]. For observational studies, we did not formally assess
SOE, but provided descriptive information in narrative
form.
3.
Evidence synthesis
3.1.
Search results
Our literature search identified 1139 references, of which
124 were selected for full-text review
( Fig. 1). This
process mapped to 43 unique RCTs. In addition, we
E U R O P E A N U R O L O G Y 7 1 ( 2 0 1 7 ) 5 7 0 – 5 8 1
571