397 476
Improvements with mirabegron were maintained in pre-
defined subcohorts of treatment-naı¨ve, treatment-experi-
enced, and older patients, as well as after matching to
controls for potentially confounding differences in baseline
characteristics. The findings of a sensitivity analysis that
tested assumptions around the date of discontinuation
were similar to those of the base-case analysis. Adherence,
assessed in terms of MPR with a fixed denominator, was
significantly greater with mirabegron than with all anti-
muscarinics in the overall study population, although these
benefits appeared to be limited mainly to treatment-naı¨ve
patients in both matched and unmatched analyses. Overall,
our findings suggest that persistence and adherence with
mirabegron are statistically superior to those with other
antimuscarinics in a large UK primary care population. The
clinical significance of increased adherence has been
highlighted recently in a prospective study showing that
women who adhered to OAB medication had significantly
greater improvements in urinary symptoms than non-
adherent women did
[24].
Our observations are consistent with other recent real-
world studies comparing mirabegron to antimuscarinic
agents under different health care systems in the US, Japan,
and Canada, which reported significantly better persistence
with mirabegron
[15–17]. In the North American studies,
both of which were based on retrospective claims data, the
relative risk of discontinuation with mirabegron versus
tolterodine ER was similar to the present study (HR: Canada,
1.44; USA, 1.64; UK, 1.56)
[15,16]. Persistence with
mirabegron versus tolterodine was also greater in treat-
ment-naive and treatment-experienced cohorts in both
studies
[15,16], as in the present study. The smaller
Japanese study, based on medical records, reported a
significantly improved 12-mo persistence rate with mir-
abegron compared to tolterodine (38% vs 20%)
[17], as in the
present study (38% vs 20%). In addition, two noncompara-
tive studies of persistence with mirabegron in UK popula-
tions have recently been reported
[25,26], but cannot be
directly compared with our study because of a shorter
duration of follow-up in one study (6 mo)
[25]and many
patients (37%) received mirabegron in combination with
antimuscarinics in the other
[26].
In our study, discontinuation of antimuscarinics was less
common among men than among women and generally
occurred within 1–3 mo, compared to a median of 5.6 mo
with mirabegron. We were unable to examine the reasons
for discontinuation in our study as these data are not
contained within the CPRD database. However, data from a
large US survey (
>
5000 respondents) suggested that the
most common reasons for discontinuation of antimuscari-
nics were treatment not working as expected, switching to a
new medication, coping without medication, and side
effects
[27]. Other reasons described in the literature
included inadequate patient counselling resulting in
unrealistic patient expectations
[28], cost
[27,29,30],
unwillingness to take long-term treatment
[27] ,and
proactive treatment holidays, all of which may have
occurred in our study. It is conceivable that the initial
separation of the Kaplan-Meier curves
( Fig. 2) is attributable
to differences between mirabegron and antimuscarinics in
the occurrence of bothersome anticholinergic side effects,
notably dry mouth,
[9,10]. The time to onset of adverse
events with antimuscarinics is approximately 1 wk
[31]and
fits with this early difference between groups. The gradient
of the curves was generally comparable after 3 mo,
suggesting that reasons for later discontinuations may
have been common to both drug classes. Further efforts are
needed to better understand the reasons for discontinua-
tion of OAB medications and how to support patients so that
they achieve long-term compliance.
This study has many design strengths including a large
population from the CPRD database that is broadly
representative of the UK general population
[21]and the
use of matching to control for baseline imbalances, notably
the proportion of treatment-experienced patients, which
was greater in the mirabegron cohort at baseline. The
unmatched analysis was used as the primary analysis
because, after applying the stringent inclusion/exclusion
criteria, it was uncertain if there would be adequate patient
numbers for matching. The main study limitations are its
retrospective design and the use of prescription-event rather
than patient-derived data (eg, patient diaries) to estimate
outcomes. Although patients with
<
12-mo follow-up after
treatment initiation were excluded (
n
= 6078;
Fig. 1) to
support the assessment of all study objectives and end-
points, these patients could have theoretically been included
in the analysis of time to discontinuation. The inability to
capture reasons for discontinuation of treatment, as well as
any potential health benefits resulting from increased
persistence in terms of symptom severity and health-related
quality of life (HRQoL), were other limitations of this study. It
should also be noted that there was variation in the
prescribing of mirabegron relative to tolterodine ER in the
early months of the study selection period (
Supplementary
Fig. 2
), possibly because of the UK launch of mirabegron
(February 2013) and the release of updated treatment
guidelines to include mirabegron
[32] .The relatively later
availability of mirabegron and its positioning within UK
guidelines
[32]may also have contributed to the higher
proportion of treatment-experienced patients in this group.
Analyses of health care resource use and associated costs
from the present study will be reported separately. Large
prospective observational studies of mirabegron are ongo-
ing in the USA (PERSPECTIVE;
https://clinicaltrials.gov/ct2/ show/NCT02386072) and Europe (BELIEVE;
https:// clinicaltrials.gov/ct2/show/NCT02320773), which include
persistence and HRQoL outcomes in a real-life setting. These
studies may help to better understand the benefits of
improved treatment persistence, reasons for discontinua-
tion, and why men in our study were less likely to
discontinue treatment than women.
5.
Conclusions
Patients receiving mirabegron remain on treatment for
significantly longer and have significantly better 12-mo
persistence and adherence rates compared to tolterodine ER
and other antimuscarinics commonly prescribed for OAB in
E U R O P E A N U R O L O G Y 7 2 ( 2 0 1 7 ) 3 8 9 – 3 9 9
397