E U R O P E A N U R O L O GY 7 3 ( 2 0 18 ) 8 9 0 – 8 9 6

available at www.sciencedirect.com
journal homepage: www.europeanurology.com

Platinum Priority – Prostate Cancer

Editorial by Scott Eggener on pp. 897–898 of this issue

Survival Outcomes of Men with Lymph Node-positive Prostate

Cancer After Radical Prostatectomy: A Comparative Analysis of
Different Postoperative Management Strategies

Karim A. Touijer a,*, Robert Jeffery Karnes b, Niccolo Passoni c, Daniel D. Sjoberg c, Melissa Assel c,
Nicola Fossati d, Giorgio Gandaglia d, James A. Eastham a, Peter T. Scardino a, Andrew Vickers c,
Cesare Cozzarini e, Francesco Montorsi d, Alberto Briganti d
a b
Urology Service, Department of Surgery, Memorial Sloan Kettering Cancer Center, New York, NY, USA; Department of Urology, Mayo Clinic Rochester,
Rochester, MN, USA; Department of Epidemiology & Biostatistics, Memorial Sloan Kettering Cancer Center, New York, NY, USA; d Division of Oncology, Unit
c

of Urology, IRCCS Ospedale San Raffaele, Vita-Salute San Raffaele University, Milan, Italy; e Department of Radiotherapy, IRCCS Ospedale San Raffaele, Vita-
Salute San Raffaele University, Milan, Italy

Article info Abstract

Article history: Background: Optimal management of patients with lymph node metastasis (LNM) after radical
Accepted September 22, 2017 prostatectomy (RP) remains undefined.
Objective: We evaluated the association between three different management strategies and
survival in prostate cancer with LNM after RP.
Associate Editor: Design, setting, and participants: We analyzed data of 1338 patients with LNM after RP from
Matthew Cooperberg three tertiary care centers. Three hundred and eighty-seven patients (28%) were observed, 676
(49%) received lifelong adjuvant androgen deprivation therapy (ADT), and 325 (23%) received
adjuvant external beam radiation therapy (EBRT) and ADT. Three hundred and sixty-eight men
Keywords: were followed for more than 10 yr.
Prostatic neoplasms Outcome measurements and statistical analysis: Primary outcome measure was overall sur-
vival (OS). Secondary outcomes were cancer-specific survival (CSS) and other-cause mortality.
Lymphatic metastasis Kaplan-Meier methods were used to visualize OS for the three treatment groups. Cox propor-
Prostatectomy tional hazards regression was utilized to compare OS and CSS among the three groups.
Prognosis Results and limitations: ADT + EBRT was associated with better OS than ADT alone (hazard ratio
[HR]: 0.46, 95% confidence interval [CI]: 0.32–0.66, p < 0.0001) or observation (HR: 0.41, 95% CI:
Adjuvant radiotherapy 0.27–0.64, p < 0.0001). Higher-risk patients benefited more from ADT + EBRT than lower-risk
Antineoplastic agents patients. Ten-year mortality risk difference between ADT + EBRT, observation, or ADT alone
Hormonal ranged from 5% in low-risk patients to 40% in high-risk patients. Adjuvant ADT + EBRT was also
associated with better CSS than observation or ADT alone (p < 0.0001), ADT had better CSS
Mortality compared to observation (HR: 0.64, 95% CI: 0.43–0.95, p = 0.027). However, ADT was associated
with an increased risk of other-cause mortality (HR: 3.05, 95% CI: 1.45–6.40, p = 0.003)
compared with observation, resulting in similar OS between ADT and observation (HR: 0.90,
95% CI: 0.65–1.25, p = 0.5). While selection bias might remain, its effect would operate in the
opposite direction to our findings.
Conclusions: In men with LNM after RP, ADT + EBRT improved survival over either observation
or adjuvant ADT alone. This survival benefit increases with higher-risk disease.
Patient summary: Lymph node metastasis following radical prostatectomy is associated with
poor survival outcomes. However, we found that adjuvant androgen deprivation therapy with
external beam radiation therapy improved survival in these patients.
© 2017 European Association of Urology. Published by Elsevier B.V. All rights reserved.

* Corresponding author. Urology Service, Department of Surgery, Memorial Sloan Kettering Cancer
Center, 1275 York Avenue, New York, NY 10065, USA. Tel. +1 646 422 4486; Fax: +1 212 988 0768.
E-mail address: touijerk@mskcc.org (K.A. Touijer).

http://dx.doi.org/10.1016/j.eururo.2017.09.027
0302-2838/© 2017 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 GY 7 3 ( 2 018 ) 8 9 0 – 8 9 6
1. Introduction
The presence of lymph node metastasis (LNM) after radical
prostatectomy (RP) is a poor prognostic sign. The long-term
risk of death from prostate cancer is substantially increased,
and is estimated to range between 20% and 42% [1,2]. How-
ever, the optimal management of patients diagnosed with
LNM at the time of RP is not clearly established. The debate
centers on the question of appropriate timing and type of
secondary therapy.
Results from the Eastern Cooperative Oncology Group
randomized clinical trial (ECOG 3886) comparing immedi-
ate lifelong androgen deprivation therapy (ADT) to obser-
vation, made a strong argument in favor of early initiation of
ADT [3]. However, concerns about the trial design and
disease characteristics of the enrolled patients limit the
applicability of this trial's findings to contemporary
prostate cancer patients.
Therapeutic decisions in the management of patients with
LNM after RP are driven by physicians’ preference or dictated
by institutional standards. This heterogeneity results in
various therapeutic strategies including observation followed
by treatment only if biochemical recurrence develops [4],
adjuvant long-term ADT [5], or a combination of adjuvant ADT
and external beam radiation therapy (EBRT) [6]. It is currently
unknown what benefit either of these adjuvant treatments
provide compared to observation, particularly, given the
known morbidity associated with ADT and EBRT [7–11].
We analyzed data from three centers in the USA and
Europe to quantify differences in survival in patients with
LNM after RP managed with observation, lifelong adjuvant
ADT, or a combination of adjuvant ADT + EBRT.
2. Patient and methods
2.1. Study design
The study was approved by the Institutional Review Board from the three
participating institutions: Memorial Sloan Kettering Cancer Center
(MSK), New York, NY, Mayo Clinic, Rochester, MN, and San Raffaele
Hospital, Milan, Italy. This was a retrospective analysis comparing
survival outcomes of node-positive patients after RP performed between
1988 and 2010 at one of three institutions, according to whether they
were managed by observation, received lifelong adjuvant ADT, or
adjuvant ADT + EBRT. The choice of treatment was primarily driven by
practice patterns or standard of care at each institution. Overall, 84% of
the patients managed by observation after RP were treated at MSK, while
73% of those managed by lifelong adjuvant ADT were treated at the Mayo
Clinic, and 83% of the patients who received adjuvant ADT + EBRT were
treated at San Raffaele Hospital.
Observation consisted of no treatment until patients experienced
biochemical recurrence (prostate specific antigen [PSA]  0.1 ng/ml with
subsequent confirmatory rise, PSA  0.2 ng/ml for the Milan patients,
and PSA  0.4 ng/ml for the Mayo clinic patients) at which point patients
were treated. ADT alone was the most commonly used (72%), followed by
salvage EBRT (13%), ADT plus EBRT (8.9%), and ADT plus chemotherapy
(2.1%). The remaining 4.7% were enrolled in clinical trials. Only 24.52% of
patients in the observation group had a detectable PSA between 42 d and
100 d after RP.
Adjuvant ADT consisted of either bilateral orchiectomy or luteinizing
hormone releasing hormone agonist. ADT was generally intended to be
891
life-long. When combined with EBRT, the median duration of ADT was
5.9 yr (interquartile range [IQR]: 3.55–8.91), 10% of the patients received
ADT for less than 1 yr, 15% for 1–3 yr, and 75% for more than 3 yr.
Adjuvant radiation therapy consisted of local radiation to the
prostatic  seminal vesicle bed and pelvic lymph nodes area (whole-
pelvis radiotherapy [WPRT]). All patients were treated with high-energy
photon beams (6–18 mV) at conventional fractionation (1.8–2 Gy/
fraction), at a median dose of 68 Gy (IQR: 66, 70). A three-dimensional
conformal (3D-CRT) approach was always used up to 2002. Intensity-
modulated radiation therapy (IMRT) was gradually introduced since
2003.
The clinical target volumes (CTVs) were drawn on computed
tomography images and always included the prostatic bed (CTV1). In
case of seminal vesicle invasion (pT3b), the seminal vesicle bed was also
identified (CTV2).
The corresponding planned target volumes (PTVs) were defined as
CTVs plus a 0.8-cm margin, with the exception of the cranio-caudal
direction where a 1 cm margin was applied. The lymph-nodal area was
contoured only in patients treated with IMRT and included obturator,
hypogastric, internal and external iliac, as well as presacral lymph-nodes.
The corresponding PTV lymph node was generated by isotropic
expansion of the lymph-nodal area of 7 mm.
In patients treated with 3D-CRT, adjuvant radiation therapy consisted
of a four-field (anteroposterior [AP]-posteroanterior plus latero-lateral
[LL]) WPRT at a median dose of 50 Gy (IQR: 45–50 Gy), followed by a
three-field (AP plus LL) boost to PTV1 at a median dose of 68 Gy (IQR: 56–
72 Gy). The median dose to PTV2, if present, was 60 Gy. WPRT was
irradiated by means of static 6-MeV 5–7 field IMRT (always 50.4 Gy), but
the consequential boost to prostatic bed was always delivered with 3D-
CRT (AP + LL) up to a median dose of 70 Gy (IQR: 68–72).
From the respective institutional databases, 1471 patients with LNM
after RP and extended pelvic lymph node dissection for prostate cancer
were identified. The extended pelvic lymph node dissection included the
external iliac, hypogastric, and obturator fossa nodes. None of the
patients received preoperative ADT, EBRT, or chemotherapy. Treatment
was considered adjuvant if it was administered within 6 mo of surgery. A
total of 33 patients were treated more than 6 mo after RP and thus were
excluded from the analysis. In addition, 50 patients with missing
information on the type of adjuvant therapy were also excluded, leaving
1388 men eligible for analysis.
2.2. Study end points
The primary outcome measure was overall mortality. Secondary
outcomes were cancer-specific mortality. Information about date of
death was obtained from the Social Security Death Index notification for
MSK and Mayo Clinic. Data on the date and cause of death for patients
treated in Milan were obtained from the Italian National Civil Registry.
Data on development of metastasis were only available from two of the
three sites.
2.3. Statistical analysis
Our primary goal was to assess differences in overall and cancer-specific
mortality among patients receiving observation, adjuvant ADT, and
adjuvant ADT + EBRT following RP. Kaplan-Meier methods were used to
visualize overall survival for the three treatment groups. Cox propor-
tional hazards regression was utilized to compare overall and cancer-
specific survival (CSS) among the three groups. The Cox models were
adjusted for pathologic Gleason score (6 vs 7 vs 8–10), tumor stage (pT2–
pT3a, pT3b, pT4), surgical margin status, number of positive nodes
removed (1 vs 2 vs 3+), and age. As sensitivity analysis, we repeated the
cancer-specific mortality model utilizing a competing risk regression
model with death from other causes as the competing risk. To assess
892 E U R O P E A N U R O L O G Y 7 3 ( 2 0 18 ) 8 9 0 – 8 9 6

whether the estimates of treatment effect on survival were confounded Table 2 – Multivariable Cox proportional hazards regression
by heterogeneity in practice patterns among the three institutions, we predicting the hazards for overall survival and cancer-specific
survival
repeated the survival regression models using only data from Milan as
this center treated many patients with ADT only and ADT + EBRT after RP. Variable Hazard ratio 95% confidence p value
All analyses were conducted using Stata 13 (Stata Corp., College Station, interval
TX, USA).
Overall survival (ADT vs ADT + EBRT, p < 0.0001)
Observation Ref. Ref. Ref.
ADT 0.90 0.65, 1.25 0.5
3. Results
ADT + EBRT 0.41 0.27, 0.64 <0.0001
Cancer-specific survival (ADT vs ADT + EBRT, p = 0.0003)
Our cohort included 1388 men with LNM who experienced Observation Ref. Ref. Ref.
321 all-cause deaths and 171 cancer-specific deaths. Median ADT 0.64 0.43, 0.95 0.027
ADT + EBRT 0.26 0.15, 0.44 <0.0001
follow-up time was 69 mo (IQR: 39–126) among survivors,
with 368 men followed for more than 10 yr. There were All models are adjusted for pathologic Gleason Grade (6 vs 7 vs 8–10),
several differences between treatment groups (Table 1). tumor stage (pT2–pT3a, pT3b, pT4), surgical margin status, number of
positive nodes removed (1 vs 2 vs 3+), and age.
Men in the adjuvant ADT group were older. Men receiving
ADT = androgen deprivation therapy; EBRT = external beam radiation
observation and ADT + EBRT had higher rates of Gleason therapy; Ref. = reference.
score (8–10) than men receiving ADT only. Men receiving
ADT + EBRT also had higher pathologic stage than ADT and
observation patients (14% ADT + EBRT, 4.0% ADT, and 6.7% Fig. 1. These results suggest that any oncologic benefit from
observation had pT4 disease) and had a higher positive lifelong adjuvant ADT may be offset by an increase in risk of
margin rate (72%). death from other causes. Multivariable competing risk
Men treated with ADT + EBRT had a lower risk of any regression with death from other causes as the outcome and
cause mortality than observation (HR: 0.41, 95% CI: 0.27– cancer death as the competing event demonstrated that
0.64, p < 0.0001) or ADT only (HR: 0.46, 95% CI: 0.32–0.66, lifelong adjuvant ADT was associated with an increased risk
p < 0.0001; Table 2). There was no significant difference in of death from other causes compared to observation (HR:
survival between men treated with adjuvant ADT and 3.05, 95% CI: 1.45–6.40, p = 0.003).
observation (HR: 0.90, 95% CI: 0.65–1.25, p = 0.5). Sensitivity analysis using data from the Milan center
Men treated with adjuvant ADT had lower rates of only, differences between ADT versus ADT + EBRT were very
prostate cancer death than men managed with observation similar to that found in the primary analysis. For example,
only (HR: 0.64, 95% CI: 0.43–0.95, p = 0.027; Table 2). Similar the CSS HR for ADT + EBRT versus ADT was 0.39 (95% CI:
results were observed when using competing risk analysis, 0.18–0.82) using the data from Milan compared with 0.41
with the HR being 0.63 for comparing men treated with ADT utilizing all data. This suggests that differences between
versus observation (95% CI: 0.42–0.95, p = 0.027). Differ- institutions do not explain the observed differences
ences between groups in overall and CSS are shown in between treatments.

Table 1 – Demographic and clinical characteristics of the analytic cohort

Observation (n = 387; 28%) ADT (n = 676; 49%) ADT + EBRT (n = 325; 23%) p valuea

Institution
Milan 30 (7.8%) 153 (23%) 271 (83%) <0.0001
Mayo 30 (7.8%) 496 (73%) 46 (14%)
MSK 327 (84%) 27 (4.0%) 8 (2.5%)
Age 62 (57–67) 66 (60–70) 65 (59–69) <0.0001
Pre-RP PSA, ng/ml (n = 1381) 9 (5–15) 14 (8–27) 15 (8–31) <0.0001
Pathological Gleason score (n = 1379)
6 7 (1.8%) 119 (18%) 31 (10%) <0.0001
7 194 (51%) 336 (50%) 140 (43%)
8–10 178 (47%) 220 (33%) 154 (47%)
Pathological stage
pT2–pT3a 217 (56%) 247 (37%) 72 (22%) <0.0001
pT3b 144 (37%) 402 (59%) 206 (63%)
pT4 26 (6.7%) 27 (4.0%) 47 (14%)
Total no. of nodes removed (n = 1387) 15 (10–21) 12 (9–17) 17 (12–22) <0.0001
No. of positive nodes
1 244 (63%) 367 (54%) 140 (43%) <0.0001
2 72 (19%) 145 (21%) 80 (25%)
3+ 71 (18%) 164 (24%) 105 (32%)
Positive surgical margin (n = 1385) 126 (33%) 373 (55%) 235 (72%) <0.0001

Statistics presented are median (interquartile range) or frequency (percent).

ADT = androgen deprivation therapy; EBRT = external beam radiation therapy; MSK = Memorial Sloan Kettering Cancer Center; PSA = prostate-specific antigen;
RP = radical prostatectomy.
a
Continuous variables compared using the Kruskal-Wallis test, and categorical variables compared using the Chi-square test.
 E U R O P E A N U R O L O GY 7 3 ( 2 018 ) 8 9 0 – 8 9 6 893
[(Fig._1)TD$IG]

Fig. 1 – (A) Kaplan-Meier curve for overall survival and (B) cancer-specific survival stratified by treatment group.
A: Black line is androgen deprivation therapy plus external beam radiation therapy, blue line is ADT, and red line is observation.
B: Black line is androgen deprivation therapy plus external beam radiation therapy, blue line is androgen deprivation therapy, and red line is
observation.
ADT = androgen deprivation therapy; EBRT = external beam radiation therapy; RP = radical prostatectomy.

As men treated with adjuvant ADT + EBRT have a lower
mortality rate than men treated with adjuvant ADT or
observation, we sought to quantify and to assess whether
the survival benefit of ADT + EBRT is uniform for all patients
or whether patients with higher-risk disease benefit more
from the additional adjuvant therapy. Men treated with ADT
or observation had similar overall survival. Combining these
two groups, we built a Cox model to predict all-cause
mortality within 10 yr using postoperative Gleason score
(6 vs 7 vs 8+), pathology T stage (pT2–pT3a vs pT3b vs pT4),
surgical margin status, and the number of positive nodes
(1 vs 2 vs 3+; Table 3). This served as our model for
predicting risk of death without adjuvant ADT + EBRT. Using
locally weighted Kaplan-Meier estimates for risk of all-
cause mortality at 10 yr, we calculated the observed risk of
death by the predicted risk assuming no adjuvant ADT
+ EBRT. The difference between the observed and the
predicted risk of death among patients receiving ADT
+ EBRT is the survival benefit conferred by the additional
therapy (Fig. 2).
To aid in clinical decision-making, the Cox prediction
model was converted into a point system that quantifies the
survival benefit of adjuvant ADT + EBRT compared to no
adjuvant treatment. We multiplied the coefficient estimates
from the Cox model by 4 and rounded to nearest integer
value to determine the points associated with each
predictor (Table 3). Patients are assigned points based on
their cancer characteristics and the sum represents the
patient's risk of death within 10 yr (Table 4).
4. Discussion
Although the presence of LNM is associated with an
increase in prostate cancer mortality [1], the optimal

Table 3 – Cox model predicting all-cause mortality for men receiving either adjuvant androgen deprivation therapy only or no adjuvant
therapy

Hazard ratio 95% Confidence interval p value Coefficient Coefficient*4 Points

Postoperative Gleason
6 Ref. Ref. Ref. +0
7 1.13 0.79–1.62 0.5 0.12 0.48 +0
8+ 1.99 1.37–2.89 0.0003 0.68 2.75 +3
Pathology T stage
pT2-pT3a Ref. Ref. Ref. +0
pT3b 1.30 0.98–1.70 0.065 0.26 1.04 +1
pT4 2.07 1.25–3.42 0.005 0.72 2.91 +3
Positive surgical margin 1.72 1.33–2.23 <0.0001 0.54 2.16 +2
No. of positive nodes
1 +0
2 0.99 0.73–1.34 0.9 –0.01 –0.04 +0
3+ 1.09 0.81–1.46 0.6 0.08 0.34 +0

Twelve patients with missing information on surgical margin status and pathologic Gleason were omitted from the multivariable analyses.
Ref. = reference.
894 E U R O P E A N U R O L O G Y 7 3 ( 2 0 18 ) 8 9 0 – 8 9 6
[(Fig._2)TD$IG]
with different outcomes depending on the tumor grade and
number of lymph nodes involved. While some could be
cured with RP alone, over two-thirds will need further
therapy and a third will ultimately succumb to prostate
cancer [4]. Long-term survival data on patients with LNM
demonstrated prognostic differences based on Gleason
score, number of metastatic lymph nodes, and positive
surgical margins and stratified node positive patients in
low-, intermediate-, and high-risk groups [12].
Results from ECOG 3886 showed a survival benefit in
favor of immediate hormonal therapy over observation
[3]. However, it is not clear whether this finding is
generalizable to node-positive prostate cancer patients in
the modern PSA era. The majority of the patients enrolled in
this trial were unscreened, and in the observation arm,
Fig. 2 – Observed versus the predicted risk of death among patients hormonal therapy was withheld until osseous metastasis
receiving ADT + EBRT. developed. Therefore neither the tumor characteristics nor
Black line is the difference in risk of all-cause mortality after radical
prostatectomy by treatment among men with lymph node metastasis the management strategy are applicable to contemporary
with and without adjuvant androgen deprivation therapy plus external patients. ECOG 3886 reported more deaths from causes
beam radiation therapy. The grey dashed lines are the 95% confidence
other than prostate cancer in the immediate ADT group than
intervals. Positive differences suggest a survival benefit for patients
receiving androgen deprivation therapy + external beam radiation in the deferred ADT group. While the findings in the present
therapy. A dotted grey line has been added at zero, for reference. study are in the same direction as those reported by
ADT = androgen deprivation therapy; EBRT = external beam radiation
Messing et al. [3] the magnitude of the positive effect of
therapy; RT = radiation therapy.
immediate ADT in reducing cancer specific mortality is not
the same. This is likely due a difference in the threshold to
Table 4 – Predicted risk of 10-yr all-cause mortality by adjuvant trigger salvage therapy (biochemical recurrence vs distant
treatment group based on pathological Gleason, pathological metastasis) or perhaps the type of salvage therapy used.
stage, margin status, and the number of positive nodes
Previous reports have shown a potential benefit of
Points n (%) 10-yr risk of death (%) Difference (%) adjuvant radiation therapy and ADT in subset of men with
LNM after RP [6]. A comparative analysis of ADT versus
Without ADT With ADT
+ EBRT + EBRT radiation therapy combined with ADT in patients with LNM
showed a beneficial impact on survival in favor of radiation
0 234 (17) 17.0 12.3 4.8
plus ADT, particularly in men with no more than two
1 175 (13) 21.6 14.1 7.5
2 133 (10) 27.5 16.2 11.3 positive nodes, nonorgan-confined, and intermediate- to
3 360 (26) 32.4 18.5 13.9 high-grade disease; this study lacked an observation arm
4 131 (9.5) 35.7 19.0 16.7 [2]. We did not find the number of metastatic nodes as a
5 98 (7.1) 46.0 22.6 23.4
predictor of response to radiation therapy. This difference is
6 197 (14) 53.2 25.0 28.3
8 57 (4.1) 71.1 31.3 39.8 likely due to differences in statistical methodologies used.
Our data demonstrated superior overall and CSS in
ADT = androgen deprivation therapy; EBRT = external beam radiation
patients treated with surgery and adjuvant ADT + EBRT than
therapy.
surgery alone or surgery and adjuvant ADT. Given the
prognostic heterogeneity of node-positive patients we
management for men with LNM is not clearly established. developed a risk-adjusted tool based on the patient and
Various management options are used including either disease characteristics to quantify the overall survival
observation with initiation of therapy after biochemical benefit of administering adjuvant EBRT in combination
recurrence, adjuvant ADT, or adjuvant radiation therapy in with ADT. We believe this tool would be useful to design a
combination with ADT. Since there is no robust evidence to clinical trial to evaluate the true benefit of adjuvant EBRT
support one approach over the other, the choice of either of and ADT in this setting.
these options is often determined by the treating physician, The magnitude of survival benefit increases as the
or prescribed in accordance with institutional guidelines pathological characteristics of the disease worsen. This
and protocols. finding suggests that despite advanced pathologic features
Long-term data of men with LNM managed conserva- and presence of LNM, timely maximal local cancer control
tively with observation after surgery, reserving treatment by way of surgery and radiation therapy can in some
only for those who develop biochemical recurrence, showed instances avert the disease progression and demise from
that up to 30% remain free of biochemical recurrence and prostate cancer. Trock et al [13] have similarly shown a
65% remain free of distant metastasis at 10 yr after RP greater impact of salvage postoperative radiotherapy in
[4]. These data provide insight on the natural course of men with worse disease characteristics (PSA doubling
surgically treated prostate cancer with LNM and suggest time < 6 mo, Gleason  8), regardless of the lymph node
that patients with LNM represent a heterogeneous group status.
 E U R O P E A N U R O L O GY 7 3 ( 2 018 ) 8 9 0 – 8 9 6
Findings from the present study argue against the belief
that the presence of pelvic LNM invariably equates to
systemic spread. Indeed, the study provides evidence that
within the heterogeneous group of patients with LNM, it is
those with the most adverse pathological features that
benefit the most from more local therapy in the form of
adjuvant radiation. We have seen herein that the survival
advantage of adjuvant EBRT persisted despite a higher rate
of positive surgical margins in this group (72%) than the
groups treated with either adjuvant ADT (55%) or observa-
tion (33%). However, whether similar results obtained by
the administration of adjuvant EBRT can be paralleled using
a radiation therapy approach given at the time of early
recurrence cannot be addressed by our analyses. Such an
approach would eventually limit the number of patients
potentially overtreated by EBRT.
The results from this study provide insight as to the
management of node-positive disease, by recognizing the
importance of local therapy in patients with the most
adverse pathological features and providing a decision
making tool to estimate the survival benefit of adjuvant
radiation therapy across the node positive disease risk
spectrum. However, we do recognize that these results
require confirmatory studies.
Selection bias is always a concern in an observational
study. We found that men who received ADT + EBRT had a
worse prognostic profile than men who were observed or
received adjuvant ADT alone. Yet the ADT + EBRT group had
better overall and CSS. If we make the common assumption
that unmeasured confounders are correlated with mea-
sured covariates, then any selection bias in the study would
operate in the opposite direction to our findings. Hence it is
unlikely that our finding favoring ADT + EBRT was related to
selection bias.
Another possible source of bias is that treatments were
strongly correlated with institution. For instance, more
than 80% of patients undergoing observation were treated
at MSK, with a similar proportion of those undergoing
ADT + EBRT being treated at the Milan site. Institutional
differences in ascertainment or attribution could not
cause bias as our outcome is overall survival. Differences
in surgical technique are also unlikely as all sites
are high-volume specialist tertiary referral centers. It is
also unlikely that differences can be ascribed to covari-
ates, for instance, if there was a tendency to upgrade at
one center. This is because the univariate results (Fig. 1)
are in a similar direction to the multivariable results
(Table 2).
5. Conclusions
Our data suggest that in men with LMN adjuvant ADT
+ EBRT confers long-term overall survival advantage than
either observation or ADT alone. The absolute difference in
survival increases as disease characteristics worsen, imply-
ing a greater value in maximizing local oncologic control
even in the face of LNM and locally advanced disease. With
external validation, these data would provide the rationale
for a risk-adjusted management of node-positive prostate
895
cancer and informed selection criteria for future clinical
trials.
Author contributions: Karim A. Touijer had full access to all the data in the
study and takes responsibility for the integrity of the data and the accuracy
of the data analysis.
Study concept and design: Touijer, Karnes, Briganti.
Acquisition of data: Passoni, Touijer, Briganti, Karnes, Fossati, Gandaglia.
Analysis and interpretation of data: Sjoberg, Assel, Vickers, Briganti,
Karnes, Touije.
Drafting of the manuscript: Touijer, Briganti, Karnes.
Critical revision of the manuscript for important intellectual content:
Briganti, Karnes, Montorsi, Cozzarini, Sjoberg, Vickers, Eastham,
Scardino, Touijer.
Statistical analysis: Sjoberg, Assel, Vickers.
Obtaining funding: Touijer.
Administrative, technical, or material support: None.
Supervision: Touijer.
Other: None.
Financial disclosures: Karim A. Touijer certifies that all conflicts of
interest, including specific financial interests and relationships and
affiliations relevant to the subject matter or materials discussed in the
manuscript (eg, employment/affiliation, grants or funding, consultan-
cies, honoraria, stock ownership or options, expert testimony,
royalties, or patents filed, received, or pending), are the following:
None.
Funding/Support and role of the sponsor: The research was supported by
The Sidney Kimmel Center for Prostate and Urologic Cancers and an NIH/
NCI Cancer Center Support Grant (P30 CA008748). The funders of the
study had no role in study design, data collection, data analysis, data
interpretation, or writing of the report.
References
[1] Eggener SE, Scardino PT, Walsh PC, et al. Predicting 15-year prostate
cancer specific mortality after radical prostatectomy. J Urol
2011;185:869–75.
[2] Abdollah F, Karnes RJ, Suardi N, et al. Impact of adjuvant radiother-
apy on survival of patients with node-positive prostate cancer. J Clin
Oncol 2014;32:3939–47.
[3] Messing EM, Manola J, Sarosdy M, Wilding G, Crawford D, Trump D.
Immediate hormonal therapy compared with observation after
radical prostatectomy and pelvic lymphadenectomy in men with
node-positive prostate cancer. N Engl J Med 1999;341:1781–8.
[4] Touijer KA, Mazzola CR, Sjoberg DD, Scardino PT, Eastham JA. Long-
term outcomes of patients with lymph node metastasis treated
with radical prostatectomy without adjuvant androgen-deprivation
therapy. Eur Urol 2014;65:20–5.
[5] Boorjian SA, Thompson RH, Siddiqui S, et al. Long-term outcome
after radical prostatectomy for patients with lymph node positive
prostate cancer in the prostate specific antigen era. J Urol
2007;178:864–70.
[6] Da Pozzo LF, Cozzarini C, Briganti A, et al. Long-term follow-up of
patients with prostate cancer and nodal metastases treated by
pelvic lymphadenectomy and radical prostatectomy: The positive
impact of adjuvant radiotherapy. Eur Urol 2009;55:1003–11.
[7] Braga-Basaria M, Dobs AS, Muller DC, et al. Metabolic syndrome in
men with prostate cancer undergoing long-term androgen-depri-
vation therapy. J Clin Oncol 2006;24:3979–83.
[8] Hershman DL, Unger JM, Wright JD, et al. Adverse health events
following intermittent and continuous androgen deprivation in
patients with metastatic prostate cancer. JAMA Oncol 2016;
2:453–61.
896 E U R O P E A N U R O L O G Y 7 3 ( 2 0 18 ) 8 9 0 – 8 9 6
[9] Lapi F, Azoulay L, Niazi MT, Yin H, Benayoun S, Suissa S. Androgen
deprivation therapy and risk of acute kidney injury in patients with
prostate cancer. JAMA 2013;310:289–96.
[10] Levine GN, D’Amico AV, Berger P, et al. Androgen-deprivation
therapy in prostate cancer and cardiovascular risk: a science advi-
sory from the American Heart Association, American Cancer Socie-
ty, and American Urological Association: endorsed by the American
Society for Radiation Oncology. Circulation 2010;121:833–40.
[11] Saylor PJ, Karoly ED, Smith MR. Prospective study of changes in the
metabolomic profiles of men during their first three months of
androgen deprivation therapy for prostate cancer. Clin Cancer Res
2012;18:3677–85.
[12] Moschini M, Sharma V, Zattoni F, et al. Risk stratification of pN+
prostate cancer after radical prostatectomy from a large single
institutional series with long-term follow-up. J Urol 2016;
195:1773–8.
[13] Trock BJ, Han M, Freedland SJ, et al. Prostate cancer-specific survival
following salvage radiotherapy vs observation in men with bio-
chemical recurrence after radical prostatectomy. JAMA 2008;
299:2760–9.