Nightly vs on-demand sildenafil for penile
rehabilitation after minimally invasive
nerve-sparing radical prostatectomy: results of
a randomized double-blind trial with placebo
Christian P. Pavlovich, Adam W. Levinson*, Li-Ming Su†, Lynda Z. Mettee,
Zhaoyong Feng, Trinity J. Bivalacqua and Bruce J. Trock
James Buchanan Brady Urological Institute, Johns Hopkins Medical Institutions, Baltimore, MD, *Department of
Urology, Mount Sinai School of Medicine, New York, NY, and †Department of Urology, University of Florida College of
Medicine, Gainesville, FL, USA
Objectives
• To clarify the role of phosphodiesterase type 5 (PDE5)
inhibitors in post-prostatectomy penile rehabilitation
(PPPR).
• To compare nightly and on-demand use of PDE5 inhibitors
after nerve-sparing minimally invasive radical
prostatectomy (RP).
Patients and Methods
• We conducted a single-institution, double-blind,
randomized controlled trial of nightly vs on-demand 50-mg
sildenafil citrate after nerve-sparing minimally invasive RP.
• A total of 100 preoperatively potent men, aged <65 years,
with scores on the Erectile Function domain of the
International Index of Erectile Function (IIEF-EF) ≥26,
underwent nerve-sparing surgery.
• The patients were randomized to either nightly sildenafil
and on-demand placebo (nightly sildenafil group), or
on-demand sildenafil and nightly placebo (on-demand
sildenafil group; maximum on-demand dose six
tablets/month) for 12 months. Patients then underwent a
1-month washout period.
• Validated measures of erectile function (IIEF-EF score and
the Expanded Prostate Cancer Index Composite [EPIC])
were compared between treatment groups over the entire
13-month time course, using multivariable mixed linear
regression models.
Results
• The treatment groups were well matched preoperatively
(mean age 54.3 vs 54.6 years, baseline IIEF-EF score 29.4 vs
BJU Int 2013; 112: 844–851
wileyonlinelibrary.com
29.3, for the nightly vs the on-demand sildenafil groups,
respectively).
• No significant differences were found in erectile function
between treatments (nightly vs on-demand sildenafil) at any
single timepoint after RP, after adjusting for potential
confounding factors.
• When evaluated over all timepoints simultaneously, no
significant effects of treatment group (nightly vs on-demand
sildenafil) were found on recovery of potency, as assessed by
absolute IIEF-EF scores (P = 0.765), on percentage of men
returning to an IIEF-EF score >21 (P = 0.830), or on
IIEF-EF score recovery to a percentage of baseline value
(P = 0.778).
• When evaluated over all timepoints simultaneously, no
significant effects of treatment group were found on
secondary endpoints such as assessment of potency
(including EPIC item 59 response ‘erections firm enough for
intercourse’), attempted intercourse frequency or
confidence.
Conclusions
• Erectile recovery up to 1 year after RP does not differ
between previously potent men who use sildenafil nightly
compared to on-demand.
• This trial does not support chronic nightly sildenafil as
being any better than on-demand sildenafil for use in penile
rehabilitation after nerve-sparing minimally invasive RP.
Keywords
radical prostatectomy, erectile dysfunction, penile
rehabilitation, sildenafil citrate, phosphodiesterase type 5
inhibitor
© 2013 The Authors
BJU International © 2013 BJU International | doi:10.1111/bju.12253
Published by John Wiley & Sons Ltd. www.bjui.org
 Nightly vs on-demand sildenafil after nerve-sparing RP

Introduction comparing on-demand with nightly sildenafil use after

Phosphodiesterase type 5 (PDE5) inhibitors are considered to
be the first-line therapy for erectile dysfunction (ED) [1].
Radical prostatectomy (RP) is known to cause immediate ED,
after which recovery of potency typically occurs slowly or not
at all [2–5]. PDE5 inhibitors potentiate erections by increasing
cGMP levels, which promote corporal smooth muscle
relaxation and blood flow. Evidence, primarily from animal
studies, suggests that neuropraxia after RP may lead to
hypoxia, apoptosis, venous leak and fibrosis of the corpora
cavernosa; early PDE5 inhibitor administration may reverse or
minimize these effects [2,5–10]. While it is clear that PDE5
inhibitors promote erection on-demand, the data on their
rehabilitative role in humans are less conclusive [2,5,7,11–16].
Furthermore, PDE5 inhibitor dosing regimens for penile
rehabilitation remain ill-defined.
There have been two randomized controlled trials (RCTs)
comparing PDE5 inhibitors with placebo for
post-prostatectomy penile rehabilitation (PPPR) after open RP,
with conflicting results [11,12]; Padma-Nathan et al. [12]
reported an advantage to nightly sildenafil over placebo while
Montorsi et al. [11] found no rehabilitative advantage for any
vardenafil dosing regimen (nightly or on-demand) vs placebo,
although the on-demand group achieved the highest erectile
function (EF) scores. Given the lack of consensus regarding
PDE5 inhibitor usage for PPPR, we designed an RCT
minimally invasive RP, allowing both treatment groups some
sildenafil exposure.
Patients and Methods
Study Design
This was a double-blind, double-dummy with placebo,
single-institution RCT of nightly vs on-demand sildenafil after
nerve-sparing minimally invasive RP (either laparoscopic or
robot-assisted RP). Participants were randomized 1:1 to
receive nightly sildenafil 50 mg with on-demand placebo, or
on-demand sildenafil 50 mg (maximum six tablets/month)
with nightly placebo for 1 year, starting the day after
nerve-sparing minimally invasive RP (Fig. 1). Participants
were instructed to take their on-demand dose 1 h before
sexual activity, and their nightly dose 1 h before bedtime or
evening sexual activity.
The double-blind study period included mailed-in
quality-of-life assessments at 1, 3, 6, 9 and 12 months after
minimally invasive RP, and at 13 months after a washout
period, during which participants were not to use any
remaining study drug or other erectile aids. Sildenafil and
matching placebo pills were provided by Pfizer Inc, (New
York, NY, USA). Participants were asked to return unused
tablets for actual use assessments at 3-month intervals.

Fig. 1 Design of study and participant flow chart. *The ophthalmologist of a patient with baseline abnormal vision, but without change or adverse
effect during the study, requested discontinuation of the patient from protocol.

12-month double-blind
study period End of study assessment
Screening Randomization (Assessments at 1, 3, 6, 1-month washout (13 months)
9, 12 months)

Completed study (n = 36)

Nightly sildenafil 50 mg with No study drug, placebo or

on-demand placebo (n = 50) other erectile aids permitted
Discontinued study (n = 14)
Non-compliant, n = 7; lack of
efficacy, n = 3; side effects, n = 2;
withdrawn consent, n = 1;
102 Men with localized started radiation, n = 1
100 men receive nerve-sparing
prostate cancer (≤cT2a,
minimally invasive RP and are
Gleason score <8) and without
randomized
ED (IIEF-EF ≥26) screened
Completed study (n = 38)

On-demand 50 mg sildenafil No study drug, placebo, or

with nightly placebo (n = 50)
other erectile aids permitted
Discontinued study (n = 12)
Non-compliant, n = 5; lack of
efficacy, n = 3; withdrawn consent, n = 2;
started radiation, n = 1; stopped
per ophthalmologist request*, n = 1

© 2013 The Authors

BJU International © 2013 BJU International 845
Pavlovich et al.
Participants who took <80% of the nightly tablets were
considered off-protocol; they were nevertheless followed for
an intent-to-treat analysis. The study had >99% power to
detect an expected 1 SD (5-point) difference in Erectile
Function domain of the International Index of Erectile
Function (IIEF-EF) score at 12 months, and >80% power to
detect a 0.6 SD difference.
Study Participants
Participation was offered to men choosing to undergo
nerve-sparing RP who satisfied the following criteria: age <65
years, untreated prostate cancer < cT2b, biopsy Gleason score
<8, baseline IIEF-EF score ≥25/30, no PDE5 inhibitor use, and
presence of a steady sexual partner. Enrolled men with at least
one neurovascular bundle (NVB) preserved progressed to
randomization.
Study Assessments
Baseline data included demographic information and two
validated sexual health-related quality-of-life instruments: the
IIEF-EF and the Expanded Prostate Cancer Index Composite
(EPIC) questionnaire [17,18]. The primary outcome was EF
recovery over time, as assessed by IIEF-EF score. In addition,
the IIEF-EF score at each timepoint was expressed as a
percentage of the baseline score to assess the relative degree of
EF recovery. Secondary outcomes included the EPIC Sexual
Domain Summary, EPIC Sexual Function Subscale and EPIC
Sexual Bother Subscale scores, and specific items of the IIEF
(2, 6 and 15, and EPIC questionnaire (57, 59 and 63)
concerning erectile confidence, quality and intercourse
frequency. Potency was defined using EPIC item 59: ‘How
would you describe the usual QUALITY of your erections
during the last 4 weeks?’ A response of ‘4’ on a 1–4 scale, ‘firm
enough for intercourse’ indicated potency, while all other
responses indicated ED [19].
Based upon previous studies, which showed that the quality of
NVB preservation could be subjectively graded and correlates
with EF outcomes [20,21], surgeons prospectively recorded
NVB preservation quality as an additional variable. We used a
0–4 scale for each NVB, resulting in a nerve-sparing score
(NSS) ranging from 0 to 8, with higher scores representing
better preservation.
Patients returned at 3 and 13 months after RP. At other
assessments they were contacted by telephone and reminded
to send in their questionnaires. At each contact, enquiries into
possible adverse effects were made.
Statistical Analysis
A stratified randomization scheme was used to ensure
balanced allocation [22]. Sample size and power were
determined during the study design phase [23]. Demographic
© 2013 The Authors
846 BJU International © 2013 BJU International
factors and potential confounding variables were compared
between treatment groups using a t-test or non-parametric
alternative (continuous variables) or chi-squared tests
(categorical variables) to assess randomization success.
Analyses of treatment effect were based on intent-to-treat. The
primary analyses focused on IIEF-EF scores. Absolute IIEF-EF
scores and IIEF-EF recovery as a percentage of the baseline
score were compared between groups at each timepoint using
ANCOVA, and over all timepoints (longitudinal analysis) using
a linear mixed-effects model with random slopes and
intercepts and an unstructured variance–covariance matrix
[24]. The proportion of patients in each group achieving an
IIEF-EF score >21 after minimally invasive RP, considered a
threshold value for lack of ED or mild ED [11,25], was
compared at each timepoint using a chi-squared test, and
compared over all timepoints using logistic regression with a
generalized estimating equations framework, while adjusting
for confounding variables [26]. Secondary analyses concerned
EPIC questionnaire subscales and responses to single items
of the various validated instruments, and were similarly
compared at each timepoint and over all timepoints.
Interactions were modelled as cross-product terms to
determine whether the treatment effect differed by NSS or
age. Because some men did not return questionnaires at all
timepoints, we also analysed the data using multiple
imputation based on a Markov chain Monte Carlo simulation
to impute missing values [27]. All statistical tests were
two-sided, and all analyses were performed with SAS v9.2 (SAS
Institute, Cary, NC, USA).
Study Oversight
This investigator-initiated study was approved by our
institutional review board and designed, conducted, analysed
and written solely by the authors. Informed consent was
obtained from all subjects in the study.
Results
Study Population and Demographics
A total of 102 men were enrolled between 2006 and 2007; 100
progressed to randomization into two 50-man treatment arms.
The cohorts were well-matched (Table 1) and reasonably
healthy (23% hypercholesteremia, 29% hypertension, 1%
diabetes mellitus, 3% coronary artery disease – without
significant difference between treatment groups). A total of
98% of patients in each group underwent bilateral NVB
preservation, although the mean NSS was slightly higher in the
on-demand sildenafil cohort (7.1 vs 6.5, P = 0.033).
Sildenafil Usage and Adherence
Adherence was high at each visit, with lower total sildenafil
intake, as expected, in the sildenafil on-demand group (mean
 Nightly vs on-demand sildenafil after nerve-sparing RP

Table 1 Pre-treatment (baseline) characteristics of men randomized to nightly vs on-demand

sildenafil after RP.

Characteristic Nightly sildenafil On-demand sildenafil P

Mean (median; range) age range 54.3 (55; 42–63) 53.6 (54; 40–64) 0.520
Race, n (%) 0.749
White 45 (90) 44 (88)
Non-white 5 (10) 6 (12)
Mean (median; range) baseline IIEF-EF score 29.4 (30; 26–30) 29.3 (30; 26–30) 0.493
Mean (median; range) baseline EPIC sexual 81.0 (84.6; 48.1–96.2) 81.2 (84.6; 51.9–96.2) 0.605
domain summary score
Clinical stage, n (%) 0.614
T1c 37 (74) 40 (80)
T2a 13 (26) 10 (20)
Biopsy Gleason score, n (%) 0.790
6 41 (82) 42 (84)
7 9 (18) 8 (16)
Mean (median; range) preoperative PSA 4.7 (4.4; 0.6–14) 5.1 (5.1; 0.8–9.0) 0.307
Type of minimally invasive RP surgery, n(%) 0.476
Laparoscopic RP 40 (80) 37 (74)
Robot-assisted RP 10 (20) 13 (26)
NVB preserved, n (%) 1.00
1 1 (2) 1 (2)
2 49 (98) 49 (98)
Mean (median; range) NSS* 6.5 (6; 2–8) 7.1 (8; 3–8) 0.033

*NSS is the sum of the scores for right and left NVB quality, each individually graded by the attending surgeon on a 0–4
scale [21].

Fig. 2 Mean IIEF-EF scores by treatment arm during the double blind Fig. 3 Percentage of men who achieved potency, defined as IIEF-EF score
(months 1–12) and after washout (13-month) study periods. The boxed >21 1–13 months after RP, by treatment arm. The boxed legend shows the
legend shows the number of patients who filled out and returned the number of patients who filled out and returned the relevant questionnaire
relevant questionnaire at each timepoint. at each timepoint.

30 50
On-demand
On-demand
Nightly Nightly
40
Percent with IIEF-EF >21

20
Mean IIEF-EF

30

10 20

10
0
0 3 6 9 12 15
Months after RP 0
1 3 6 9 12 13
Months
On-demand (n): 50 48 39 38 38 30 38
Nightly (n): 49* 46 42 44 36 34 36 On-demand (n): 48 39 38 38 30 38
Nightly (n): 46 42 44 36 34 36

2227 mg; median 2650 mg; 4.4 50-mg pills/month) than the
sildenafil nightly group (mean 16 580 mg; median 17 325 mg; to age or IIEF-EF score at baseline, 1, 3 or 6 months (data not
28.8 50-mg pills/month). Not all participants returned all shown).
questionnaires at all timepoints; compliance ranged from 60 to
96% per timepoint, but was balanced between groups (Figs 2
Safety
and 3). Among men who did not return any questionnaires
after 6 months (n = 16), there were no differences between There was one serious adverse event, a pulmonary embolism
treatment groups with respect in a patient on nightly sildenafil. Overall, 26 patients

© 2013 The Authors

BJU International © 2013 BJU International 847
Pavlovich et al.

discontinued treatment (14 nightly, 12 on-demand; P = 0.648),
of whom two discontinued treatment because of side effects
(headache/blurred vision in one, blurred vision in another).
Primary Outcomes
Recovery of EF based on mean IIEF-EF score did not differ
significantly in univariate comparisons between nightly or
on-demand groups at any timepoint during treatment
(months 0–12 [Table 2]). After washout (at 13 months), the
patients in the nightly sildenafil group had lower mean (SD)
IIEF-EF scores (13.8 [9.9] vs 19.2 [9.8], P = 0.022; Fig. 2) than
patients in the on-demand group, but this difference was not
significant when adjusted for NSS (P = 0.071). Recovery of EF
based on the IIEF-EF percentage return to baseline score did
not differ by treatment group at any single timepoint
(Table 2). Because recovery of EF occurs over time, we
compared the entire time course of EF recovery between
treatment groups using linear mixed-effects models, including
covariates for age and NSS to adjust for any imbalance in
these factors. There was no significant treatment effect on
either IIEF-EF (P = 0.765) or on IIEF-EF recovery as a
percentage of baseline score through 13 postoperative months
(P = 0.778; Tables 3 and 4). The results did not change when
we repeated the analyses using multiple imputation to correct
for missing values (P = 0.725 and P = 0.721, respectively). The
effect of treatment did not vary within strata defined by age or
NSS (tests of interaction: P = 0.967 and P = 0.962,
respectively). Likewise, the percentage of men with IIEF-EF
score >21 did not differ significantly between treatment
groups at any timepoint, including after washout (Fig. 3), or
when evaluated over all timepoints simultaneously in a logistic
regression generalized estimating equations model (P = 0.830;
Table 5).
The summary NSS was the only factor that was consistently
found to have a significant association with EF outcomes in all
longitudinal multivariable models. A 1-unit increase in NSS
was associated with an absolute increase in IIEF-EF score of
1.65 (P = 0.005; Table 3), and a 5.4% increase in percentage
return to baseline IIEF-EF (P = 0.006; Table 4).
Secondary Outcomes
A number of other measures of sexual function were
evaluated at each individual timepoint and over all timepoints
in multivariable linear mixed models, including EPIC Sexual
Domain Summary Score, EPIC Sexual Function Subscale and
EPIC Sexual Bother Subscale scores, and items 2, 6 and 15 of

Table 2 Mean IIEF-EF scores and IIEF-EF percentage return to baseline for nightly vs on-demand
50-mg sildenafil treatment, 0–13 months after RP.

Months after RP IIE-EF score IIEF-EF % return to baseline

On-demand Nightly P On-demand Nightly P

0 months (baseline) 29.3 29.4 0.492 100 100 –

1 month 7.8 7.6 0.946 27 26 0.912
3 months 11.6 10.2 0.500 39 35 0.476
6 months 15.4 12.1 0.138 53 41 0.122
9 months 16.2 15.5 0.771 55 52 0.689
12 months 18.5 16.7 0.456 63 57 0.459
13 months (washout) 19.2 13.8 0.022 65 47 0.018

Table 3 Mixed model of nightly vs on-demand sildenafil treatment effect on IIEF-EF score, adjusted
for age, NSS and baseline IIEF-EF score, 1–13 months after RP.

Dependent variable Predictor Coefficient (95% CI) P

IIEF-EF score Treatment (nightly vs on-demand) 0.47 (−2.63, 3.58) 0.765

Baseline IIEF-EF score (continuous) 1.08 (−0.40, 2.55) 0.151
Age (years) −0.17 (−0.44, 0.09) 0.197
NSS (continuous) 1.65 (0.51, 2.78) 0.005

Table 4 Mixed model of nightly vs on-demand sildenafil treatment on IIEF-EF percentage return to
baseline, adjusted for age, and NSS, 1–13 months after RP.

Dependent variable Predictor Coefficient (95% CI) P

Percentage return to baseline Treatment (nightly vs on-demand) 1.52 (−9.06, 12.10) 0.778
IIEF-EF score Age (years) −0.68 (−1.57, 0.20) 0.128
NSS (continuous) 5.43 (1.56, 9.29) 0.006

© 2013 The Authors

848 BJU International © 2013 BJU International

Table 5 Logistic regression model of nightly vs on-demand sildenafil
treatment on binary IIEF-EF score (>21 vs ≤21), 1–13 months after RP.
Dependent Predictor Odds ratio
variable (95% CI)
IIEF-EF >21 vs Treatment (nightly vs 1.05 (0.65, 1.69)
IIEF-EF ≤21 on-demand)
Age (years) 0.97 (0.93, 1.01)
NSS (continuous) 1.25 (1.06, 1.46)
Model used generalized estimating equations to incorporate IIEF score over all time
points.
the IIEF, and items 57, 59 and 63 from the EPIC questionnaire.
None differed significantly between treatment groups
throughout the study, including potency defined by erection
quality, which was achieved by 53% of patients without any
sildenafil at the end of the study after 1-month washout. By
that point, 71% of study patients overall reported having
achieved satisfactory postoperative intercourse. Finally,
there was no effect of minimally invasive surgery type
(robot-assisted vs laparoscopic RP) on sexual function
outcomes, nor did accounting for study compliance alter any
result (data not shown).
All of the above analyses were repeated, excluding the two
patients who underwent unilateral, rather than bilateral
nerve-sparing. The results were unchanged (data not shown).
Discussion
While there is no doubt that on-demand PDE5 inhibitors
facilitate successful individual erections after RP, human
evidence for a sustained, rehabilitative improvement is limited
[2,5,7,11–15]. Nevertheless, PDE5 inhibitors are commonly
prescribed for nightly use as part of PPPR [5,7,28]. To our
knowledge, there have been three previous randomized trials
of exclusively oral PDE5 inhibitors that address PPPR: a small
highly selected non-placebo controlled study, and two larger
placebo-controlled RCTs that reached discordant conclusions
[11,12,15]. In the current trial, we found no benefit to nightly
sildenafil compared with on-demand sildenafil for PPPR,
consistent with data from the largest of the previous RCTs
[11].
A comparison of the current trial with the two preceding
placebo-controlled RCTs is warranted. Each trial enrolled
patients with similar baseline EF (mean baseline IIEF-EF score
29). Bilateral nerve-sparing was performed in essentially all
patients; however, while the other trials involved many
surgeons performing open RP at many institutions, the
current trial involved only three minimally invasive RP
surgeons at one institution. The authors of the other trials
chose a 9-month double-blind period followed by a 2-month
washout [11,12] compared with 12 months, followed by
1-month washout in the current trial. Only the current trial
Nightly vs on-demand sildenafil after nerve-sparing RP
had a true double-dummy design, with patients in both groups
receiving different regimens of both placebo and sildenafil.
P
The multicentre international sildenafil RCT reported by
Padma-Nathan et al. [12] was double-blind, with three parallel
0.830
fixed-dose treatment groups: placebo and 50 mg or 100 mg
0.103 sildenafil. That study used the IIEF-EF, but primary outcomes
0.008 were the percentage of patients who scored a sum of ≥8 on Q3
and Q4 of the IIEF-EF and answered ‘yes’ to ‘Over the past 4
weeks, have your erections been good enough for satisfactory
sexual activity?’ Trial enrolment ended prematurely because of
a ‘lack of treatment effect’, but 76 patients did complete the
trial. After washout, however, only one of 25 patients (4%) in
the placebo arm was potent, vs 14 of 51 patients (27%) in the
two sildenafil arms combined. This significant difference (P =
0.016) suggested a benefit of nightly sildenafil over placebo for
PPPR. That landmark study nevertheless had at least three
issues that weakened its findings [2,5,7,8,10,14,16]. First, the
number enrolled was well below the target sample size of 55
per group specified by the investigators [12]. Second, sildenafil
administration began 1 month after surgery, perhaps
diminishing potential rehabilitative effects. Third, the level of
EF recovery after washout, most notably in the placebo arm,
was far lower than expected, which could reflect the
consistency and quality of NVB preservation.
The endpoint of mean (SD) IIEF-EF score after washout is the
most comparable endpoint between the current trial and the
trial by Padma-Nathan et al. and was 13.1 (9.5) for nightly
sildenafil (50–100 mg dosing data combined) and 8.8 (7.0) in
their placebo arm [12], vs 13.8 (9.9) for nightly sildenafil and
19.2 (9.8) in the on-demand sildenafil in the current trial
(50 mg dosing). The Padma-Nathan et al. trial remains the
only placebo-controlled RCT to document a benefit of PDE5
inhibitor usage for PPPR in humans [12,15].
A second PDE5 inhibitor PPPR trial (REINVENT) was
published in 2008 by Montorsi et al. [11]. REINVENT was a
423-patient multi-institution, international, double-blind
placebo-controlled variable-dose RCT of placebo, nightly or
on-demand vardenafil. The primary outcome was percentage
of patients achieving IIEF-EF score >21. Results from this trial
did not support nightly PDE5 inhibitors over on-demand
dosing, and in fact found no improvement in EF after washout
for either protocol compared with placebo. This impressive
REINVENT trial was also subject to some limitations
[2,8,13,16]. The overall design was complicated, with three
arms and participant-controlled titration of dosing, which
muddied the analysis [5,8,13]. Drop-out rates were high,
ranging between 31 and 35% in the study arms. Although EF
may recover for up to 4 years [3,4], the active drug was only
given for 9 months. Further, the REINVENT trial lacked
quantification of drug usage, and had no defined limit in the
on-demand arm. It is possible that actual medication usage
may have been similar between the active arms [2,13,16]. In
© 2013 The Authors
BJU International © 2013 BJU International 849
Pavlovich et al.
the current trial, on-demand patients were limited to six
tablets/month, and usage assessments allowed us to be sure
that the two study arms represented different actual dosage
patterns.
It is possible to compare outcomes from the REINVENT trial
directly with the current trial as both assessed the percentage
of patients with IIEF-EF score >21 after washout. In the
REINVENT trial, 24.1% of patients using nightly and 29.1%
using on-demand vardenafil achieved an IIEF-EF score >21
after washout. The percentages in the comparable groups in
the current trial were 33.2 and 50%, respectively, albeit 2
months later. The authors of the REINVENT trial concluded
that their data argue against PPPR with nightly short-acting
PDE5 inhibitors; our current results, despite study differences,
concur with their main finding.
The current study has some notable strengths. Drug
administration began immediately after surgery, and usage was
assessed throughout. The quality of NVB preservation was
consistent and quantified using a NSS, which was again found
to be a significant predictor of EF recovery in multivariable
analyses, supporting previous published data [20,21].
Additionally, we used two complete validated instruments
(EPIC and IIEF) to account for the heterogeneity that exists
among patients defined as ‘potent’ by the IIEF [29]; neither
indicated an advantage for nightly dosing.
There are also limitations to the current study. A pure placebo
arm was not a part of our trial design for two reasons. First,
prevailing clinical practice at trial design was to recommend
nightly PDE5 inhibitor dosing after RP based on
contemporaneous data [12]. Second, we were concerned that
since PDE5 inhibitor use had become so prevalent after RP,
patients would not wish to enrol for fear of being randomized
to the placebo-only group. The study period in the current trial
was only 13 months; this was longer than similar trials [11,12]
but still short of the 18–24 months duration recommended by
some authors [7,8,13]. In addition, we used only 50 mg
sildenafil. While this dosage is potentially suboptimal for PPPR,
response rates in the Padma-Nathan trial were similar between
the 50- and 100-mg dosing groups [12], and it is better tolerated
than 100-mg dosing. Furthermore, we used a short-acting PDE5
inhibitor, but a longer-acting PDE5 inhibitor, such as tadalafil,
might perform differently [30]. Although subjects were
randomized, mean NSS was slightly higher in the on-demand
group, which could have confounded univariate analyses, but
this was adjusted for in multivariable analyses. Finally, 90% of
subjects were Caucasian, representative of our referral
population. Thus, results may not necessarily be generalizable.
Nevertheless, given our findings, we cannot recommend
nightly PDE5 inhibitors for PPPR as being any better than
on-demand PDE5 inhibitors, so in practice we recommend
early on-demand PDE5 inhibitors for patients who have
undergone RP.
© 2013 The Authors
850 BJU International © 2013 BJU International
In conclusion, in a RCT of sildenafil for PPPR, there was no
evidence of improved EF at any postoperative timepoint for
nightly vs on-demand 50-mg dosing. While these data do not
show any benefit to the chronic (nightly) administration of
sildenafil after RP compared with on-demand use, the
relatively favourable outcomes in terms of recovery of EF
overall continue to support the use of on-demand PDE5
inhibitors after nerve-sparing RP.
Acknowledgements
The authors wish to thank Pfizer Inc., and especially Dr Geri
Anastasio for supporting this study, providing study drug and
matching placebo, and for assistance in trial design. We also
thank Dr Jonathan Jarow for his support, analytical comments
and assistance in manuscript preparation.
This trial was supported with an independent
investigator-initiated grant from Pfizer Pharmaceuticals.
Financial disclosures: none declared.
Conflict of Interest
None declared.
References
1 Qaseem A, Snow V, Denberg TD et al. Hormonal testing and
pharmacologic treatment of erectile dysfunction: a clinical practice
guideline from the American College of Physicians. Ann Intern Med 2009;
151: 639–79
2 Magheli A, Burnett AL. Erectile dysfunction following prostatectomy:
prevention and treatment. Nat Rev Urol 2009; 6: 415–27
3 Glickman L, Godoy G, Lepor H. Changes in continence and erectile
function between 2 and 4 years after radical prostatectomy. J Urol 2009;
181: 731–5
4 Rabbani F, Stapleton AM, Kattan MW, Wheeler TM, Scardino PT.
Factors predicting recovery of erections after radical prostatectomy. J Urol
2000; 164: 1929–34
5 Hatzimouratidis K, Burnett AL, Hatzichristou D, McCullough AR,
Montorsi F, Mulhall JP. Phosphodiesterase type 5 inhibitors in
postprostatectomy erectile dysfunction: a critical analysis of the basic
science rationale and clinical application. Eur Urol 2009; 55: 334–47
6 McCullough AR, Levine LA, Padma-Nathan H. Return of nocturnal
erections and erectile function after bilateral nerve-sparing radical
prostatectomy in men treated nightly with sildenafil citrate: subanalysis of
a longitudinal randomized double-blind placebo-controlled trial. J Sex
Med 2008; 5: 476–84
7 Mulhall JP. The role and structure of a postradical prostatectomy penile
rehabilitation program. Curr Urol Rep 2009; 10: 219–25
8 Garcia FJ, Brock G. Current state of penile rehabilitation after radical
prostatectomy. Curr Opin Urol 2010; 20: 234–40
9 Rambhatla A, Kovanecz I, Ferrini M, Gonzalez-Cadavid NF, Rajfer J.
Rationale for phosphodiesterase 5 inhibitor use post-radical
prostatectomy: experimental and clinical review. Int J Impot Res 2008; 20:
30–4
10 Shindel AW. 2009 update on phosphodiesterase type 5 inhibitor therapy
part 1: recent studies on routine dosing for penile rehabilitation, lower
urinary tract symptoms, and other indications (CME). J Sex Med 2009; 6:
1794–808; quiz 93: 809–10

11 Montorsi F, Brock G, Lee J et al. Effect of nightly versus on-demand
vardenafil on recovery of erectile function in men following bilateral
nerve-sparing radical prostatectomy. Eur Urol 2008; 54: 924–31
12 Padma-Nathan H, McCullough AR, Levine LA et al. Randomized,
double-blind, placebo-controlled study of postoperative nightly sildenafil
citrate for the prevention of erectile dysfunction after bilateral
nerve-sparing radical prostatectomy. Int J Impot Res 2008; 20: 479–86
13 Mulhall JP. Does on-demand vardenafil improve erectile function
recovery after radical prostatectomy? Nat Clin Pract Urol 2009; 6: 14–5
14 Mulhall JP, Parker M, Waters BW, Flanigan R. The timing of penile
rehabilitation after bilateral nerve-sparing radical prostatectomy affects
the recovery of erectile function. BJU Int 2010; 105: 37–41
15 Bannowsky A, Schulze H, van der Horst C, Hautmann S, Junemann KP.
Recovery of erectile function after nerve-sparing radical prostatectomy:
improvement with nightly low-dose sildenafil. BJU Int 2008; 101: 1279–83
16 Novara G, Ficarra V. Re: Francesco Montorsi, Gerald Brock, Jay Lee, et al.
Effect of nightly versus on-demand vardenafil on recovery of erectile
function in men following bilateral nerve-sparing radical prostatectomy.
Eur Urol 2008; 54: 924–31. Eur Urol 2009; 55: e95–6; author reply e97–8
17 Rosen RC, Riley A, Wagner G, Osterloh IH, Kirkpatrick J, Mishra A.
The international index of erectile function (IIEF): a multidimensional
scale for assessment of erectile dysfunction. Urology 1997; 49: 822–30
18 Wei JT, Dunn RL, Litwin MS, Sandler HM, Sanda MG. Development
and validation of the expanded prostate cancer index composite (EPIC)
for comprehensive assessment of health-related quality of life in men with
prostate cancer. Urology 2000; 56: 899–905
19 Alemozaffar M, Regan MM, Cooperberg MR et al. Prediction of erectile
function following treatment for prostate cancer. JAMA 2011; 306:
1205–14
20 Moskovic DJ, Alphs H, Nelson CJ et al. Subjective characterization of
nerve sparing predicts recovery of erectile function after radical
prostatectomy: defining the utility of a nerve sparing grading system.
J Sex Med 2011; 8: 255–60
21 Levinson AW, Pavlovich CP, Ward NT, Link RE, Mettee LZ, Su LM.
Association of surgeon subjective characterization of nerve sparing
quality with potency following laparoscopic radical prostatectomy. J Urol
2008; 179: 1510–4
22 Meinert CL, Tonascia S. Clinical Trials: Design, Conduct, and Analysis.
New York: Oxford University Press, 1986
Nightly vs on-demand sildenafil after nerve-sparing RP
23 Lachin JM. Introduction to sample size determination and power analysis
for clinical trials. Control Clin Trials 1981; 2: 93–113
24 Littell RC, Pendergast J, Natarajan R. Modelling covariance structure in
the analysis of repeated measures data. Stat Med 2000; 19: 1793–819
25 Cappelleri JC, Siegel RL, Osterloh IH, Rosen RC. Relationship between
patient self-assessment of erectile function and the erectile function
domain of the international index of erectile function. Urology 2000; 56:
477–81
26 Zeger SL, Liang KY, Albert PS. Models for longitudinal data: a
generalized estimating equation approach. Biometrics 1988; 44:
1049–60
27 Shafer JL. Analysis of Incomplete Multivariate Data. New York: Chapman
& Hall, 1997
28 Teloken P, Mesquita G, Montorsi F, Mulhall J. Post-radical
prostatectomy pharmacological penile rehabilitation: practice patterns
among the international society for sexual medicine practitioners. J Sex
Med 2009; 6: 2032–8
29 Levinson AW, Ward NT, Sanda MG et al. Comparison of validated
instruments measuring sexual function in men. Urology 2010; 76:
380–6
30 Montorsi F, Nathan HP, McCullough A et al. Tadalafil in the Treatment
of Erectile Dysfunction following bilateral nerve sparing radical
retropubic prostatectomy: a randomized, double-blind, placebo controlled
trial. J Urol 2004; 172: 1036–41
Correspondence: Christian P. Pavlovich, Suite 3200, 301
Building, Johns Hopkins Bayview Medical Center, 4940
Eastern Avenue, Baltimore, MD 21224, USA.
e-mail: cpavlov2@jhmi.edu
Abbreviations: PDE5, phosphodiesterase type 5; PPPR,
post-prostatectomy penile rehabilitation; RP, radical
prostatectomy; IIEF-EF, Erectile Function domain of the
International Index of Erectile Function; ED, erectile
dysfunction; RCT, randomized controlled trial; EF, erectile
function; NVB, neurovascular bundle; NSS, nerve-sparing
score.
© 2013 The Authors
BJU International © 2013 BJU International 851