EURURO Vol. 71 No. 4

535 692

1. Introduction
Metastatic castration-resistant prostate cancer (mCRPC) is
characterised by disease progression, despite medical or
surgical castration, and unfavourable prognosis . Before
chemotherapy, several hormonal manipulation strategies
existed, including antiandrogen initiation, antiandrogen
withdrawal, switching antiandrogens, or higher antiandro-
gen doses. Second-line hormonal therapy with high-dose
(150 mg/d) bicalutamide, as and where approved, produced
prostate-specific antigen (PSA) reductions of 50% in up to
45% of patients with nonmetastatic CRPC However,
since 2010, five additional therapies have been approved for
mCRPC, including enzalutamide and abiraterone .
Understanding the effect of various therapeutic strate-
gies upon health-related quality of life (HRQoL) has become
an important therapeutic goal in clinical studies. Instru-
ments used to evaluate HRQoL include the Functional
Assessment of Cancer Therapy–Prostate (FACT-P) and Brief
Pain Inventory, Short-form (BPI-SF) in trials of enzaluta-
mide and abiraterone , and the European Quality
of Life 5-Domain Scale (EQ-5D) in a trial with enzalutamide
.
HRQoL benefits have been observed with enzalutamide
versus placebo and abiraterone/prednisone versus predni-
sone alone in men with mCRPC before and after chemother-
apy The randomised phase 2 TERRAIN trial
compared the efficacy of enzalutamide and bicalutamide
in patients with mCRPC who progressed on luteinising
hormone receptor hormone agonist/antagonist therapy
or after bilateral orchiectomy (ClinicalTrials.gov,
NCT01288911) Enzalutamide significantly reduced
the risk of prostate cancer progression or death versus
bicalutamide. A key exploratory objective of TERRAINwas to
compare treatment effects of enzalutamide versus bicalu-
tamide on patient-reported outcomes (PROs) through
prospective prespecified analyses.
This report examines the impact of enzalutamide versus
bicalutamide on HRQoL in men with asymptomatic or
mildly symptomatic mCRPC in TERRAIN.
2. Patients and methods
2.1. Study design and patients
HRQoL was assessed during TERRAIN, a multinational, phase 2,
randomised, double-blind study of enzalutamide versus bicalutamide
in 375 patients with mCRPC. Methodological details of the primary
TERRAIN study design, eligibility criteria, and conduct have been fully
described elsewhere Patients were randomised (1:1) via an
interactive voice and web response system to enzalutamide 160 mg/d
or bicalutamide 50 mg/d, in addition to androgen deprivation therapy
(Supplementary data). The study protocol was approved by the ethics
committees of participating institutions and conducted in accordance
with the Declaration of Helsinki. Patients provided written informed
consent before enrolment.
2.2. HRQoL assessments
Data for the current posthoc analysis were prospectively collected with
validated patient self-reported questionnaires for HRQoL using FACT-P
and EQ-5D and for pain using BPI-SF ). Both FACT-P and EQ-5D
have pain-related questions, but these have not been analysed
separately. PROs were assessed at baseline (BPI-SF at screening; EQ-
5D/FACT-P at 1 wk) and every 12 wk until study drug discontinuation.
HRQoL deterioration and pain progression were defined on the basis
of changes in scores (vs baseline) previously shown to be clinically
meaningful to patients ). Pain progression endpoints were: (1)
average pain progression, (2) worst pain progression, and (3) pain
interference progression.
Skeletal-related events are a component of disease progression, but
the low incidence (enzalutamide 11.4%; placebo 10.5%) in TERRAIN
asymptomatic/mildly symptomatic mCRPC patients precluded any
robust inferential assessment of this parameter.
2.3. Statistical analyses
Questionnaire completion was defined as 1 question answered at an
assessment time point. Adjusted completion rate at every assessment
time point for all questionnaires was calculated as number of patients
completing 1 question, divided by the total number of patients
available (alive and still on the study drug) at that time point.
PRO analyses were performed on the full analysis set (all patients
randomised). Analysis of HRQoL change from baseline and deterioration
included only patients with baseline and 1 postbaseline score.
To estimate longitudinal changes from baseline, the primary analysis
used a mixed effects model for repeated measures (MMRM) . MMRM
analysis uses all available data and assumes that missing observations
are missing at random (MAR). In HRQoL studies, missing data are often
not MAR (eg, patients experiencing increased toxicity and disease
progression/death are more likely to miss assessments). If the
probability of missingness depends on unobserved HRQoL scores, then
missing data are considered nonignorable, or missing not at random
. To address this possibility, a second analysis utilised a pattern
mixture model (PMM) via sequential modelling with multiple imputa-
tion when imputation varies by reason for treatment discontinuation
(Supplementary Table 2).
Both models included the covariates: (1) treatment group, (2)
bilateral orchiectomy or receipt of luteinising hormone receptor
hormone agonist/antagonist therapy started before or after the diagnosis
of metastases, (3) Eastern Cooperative Oncology Group score at baseline,
(4) age, (5) baseline HRQoL value, and (6) time. The prespecified time
lack of multiple comparisons corrections, and unknown effects of anxiety/depression
on HRQoL.
Conclusions: In patients with asymptomatic/mildly symptomatic mCRPC, enzalutamide
provides HRQoL benefit versus bicalutamide.
Patient summary: Enzalutamide treatment was associated with better health-related
quality of life in several domains versus bicalutamide in asymptomatic/mildly symptomatic
metastatic castration-resistant prostate cancer. This likely relates to previously reported
lower rates of symptomatic disease progression.
# 2016 European Association of Urology. Published by Elsevier B.V. All rights reserved.
E U R O P E A N U R O L O G Y 7 1 ( 2 0 1 7 ) 5 3 4 – 5 4 2 535