EURURO Vol. 73 No. 1

14 148

treatment effects , these techniques are unable to fully
account for selection bias, and consequently residual
confounding remains Furthermore, there remain
important statistical limitations with respect to their ability
to determine differences in outcomes. To address such
residual confounding, Cooperberg et al performed an
elegant sensitivity analysis in which Kattan scores were
artificially increased for patients undergoing prostatec-
tomy. To show that surgery was not better than radiation,
there had to be an increase of more than 30 Kattan points,
which was considered unrealistic.
Other concerns with the meta-analysis of observational
studies include the relevance of the treatments included,
given recent advances in radiotherapy. However, examining
patients treated with dose-escalated intensity-modulated
radiotherapy (IMRT; >81 Gy) compared to radical prostatec-
tomy, Zelefsky et al found comparable results. Among
patients with high-grade PCa, Kishan et al found no
difference in overall survival between those treated surgi-
cally, those treated with EBRT and ADT, and those treated
with EBRT, brachytherapy boost, and ADT. While the authors
found lower rates of metastasis among men receiving
radiotherapy and ADT, this was confounded by short
follow-up (<5 yr) and co-administration of ADT.
This meta-analysis represents level 2a evidence,
although the limitations to account for unmeasured
confounding continue to be a problem for these studies
. Thus, despite a number of studies on this topic
), the question remains unresolved.
3.2. Global health-related QoL
A recent systematic review highlighted the importance of
patient-derived health-related QoL (HRQoL) assessment in
the evaluation of treatment outcomes among patients with
urologic cancers . While specific patient-reported
functional domains are of interest and more likely to reflect
treatment-related mechanisms, global HRQoL may be more
meaningful, despite limitations because of the ceiling effect
of these instruments. Three contemporary randomized
controlled trials assessed patient-reported outcomes (PROs)
including global HRQoL ). Among the ProtecT
cohort, Donovan et al demonstrated no differences in
physical health, mental health, anxiety, or depression
among men treated with surgery or radiotherapy. Lenner-
nas et al and Gilberti et al similarly found no
difference in overall measures of HRQoL whether patients
were treated with EBRT or brachytherapy when compared
to radical prostatectomy. PCa treatment may also affect the
QoL of patients’ spouses . Further work, including the
development of measures that overcome the ceiling effect,
is urgently needed in this area.
Table 1 – Key studies examining oncologic outcomes for treatment of localized prostate cancer with radiotherapy and radical
prostatectomy
Study Design Exposures Sample size Findings Limitations
Hamdy Randomized
controlled trial
RP vs EBRT + ADT 1098 No difference in PCSM
(p = 0.48) or OM (p = 0.87)
Underpowered
Over-representation
of low-risk patients
Lennernas Randomized
controlled trial
RP vs EBRT + BT
boost + ADT
89 No difference in PCSM Underpowered
Wallis Meta-analysis of
observational studies
RP vs RT (EBRT or BT) 95 791 Higher OM and PCSM among
patients treated with RT
Residual confounding
RP = radical prostatectomy; RT = radiotherapy EBRT = external beam RT; BT = brachytherapy; ADT = androgen deprivation therapy; PCSM = prostate cancer
specific mortality; OM = overall mortality.
Table 2 – Key studies examining functional outcomes for treatment of localized prostate cancer with radiotherapy and radical
prostatectomy
Study
Hamdy Lennernas Gilberti Resnick
Study design Randomized controlled trial Randomized controlled trial Randomized controlled trial Observational cohort study
Exposures RP vs EBRT + ADT RP vs EBRT + BT boost + ADT RP vs BT RP vs EBRT
Sample size 1098 89 174 1655
Findings
Global HRQoL Equivalent Equivalent Equivalent -
Incontinence Greater in RP Equivalent Equivalent Greater in RP at 2/5 yr
Equivalent at 15 yr
Erectile dysfunction Greater in RP Equivalent Greater in RP (short term)
Equivalent (long term)
Greater in RP at 2/5 yr
Equivalent at 15 yr
Bowel symptoms Greater in RT Equivalent Greater in RT (short term)
Equivalent (long term)
Greater in RT at 2/5 yr
Equivalent at 15 yr
Obstructive urinary
symptoms
Greater in RT Equivalent Greater in RT (short term)
Equivalent (long term)
EBRT = external beam radiotherapy; BT = brachytherapy; ADT = androgen deprivation therapy; RP = radical prostatectomy; RT = radiotherapy; HRQoL = health-
related quality of life.
E U R O P E A N U R O L O G Y 7 3 ( 2 0 1 8 ) 1 1 – 2 0 14