Tumour grade, treatment, and relative survival in a population-based cohort of men with potentially curable prostate cancer.

Abstract

There is insufficient information regarding the benefit of treatment with curative intent for men with localised poorly differentiated prostate cancer (PCa).
To evaluate relative survival in men with potentially curable PCa in relation to Gleason score (GS) and treatment as practiced in the community at large.
A population-based study including all men with localised PCa registered in Sweden's National Prostate Cancer Register.
Hormonal therapy, watchful waiting, and treatment with curative intent.
The ratio of observed deaths to expected deaths, determined from survival in the general male population of the same age, was assessed using Poisson regression analysis, with GS and treatment as covariates. Interaction between GS and treatment was tested in a multivariate Cox proportional hazard analysis.
A total of 31,903 men with potentially curable tumour (T1-T3, N0/NX, M0/MX, age <75 yr, and prostate-specific antigen [PSA] <20 ng/ml) were identified. GS was recorded for 28,454 of these men. Some 19,606 men (60.8%) were treated with curative intent, and 12,645 men (39.2%) were given either hormonal treatment or expectant management. The ratios between observed and expected survival gradually increased for men with GS 10, with GS to 3.3 for men treated conservatively and to 1.4 for men treated with curative intent. There was a significant interaction between GS and treatment, with a relatively greater benefit from treatment with curative intent for men with high-grade tumours. The results have to be interpreted with some caution, as there was no randomisation between the treatment groups.
Survival for men with well-differentiated tumours is close to that of the general population, regardless of treatment, but the outcome is dismal for men with poorly differentiated tumours, whichever treatment is applied. Nevertheless, men with poorly differentiated tumours benefit more from curative treatment than do men with well- differentiated tumours.

Full text

Page 1

Prostate Cancer
Tumour Grade, Treatment, and Relative Survival in a
Population-based Cohort of Men with Potentially Curable
Prostate Cancer
Sam Ladjevardia,*, Gabriel Sandblomb, Anders Berglundc, Eberhard Varenhorstd
aDepartment of Urology, University Hospital, Uppsala, Sweden
bDepartment of Surgery, University Hospital, Lund, Sweden
cOncology Centre, Uppsala University Hospital, Uppsala, Sweden
dDepartment of Urology, University Hospital, Linko ¨ping, Sweden
E U R OP E A N U R O L O G Y 5 7 ( 2 0 1 0 ) 6 3 1 – 6 4 0
available at www.sciencedirect.com
journal homepage: www.europeanurology.com
Article info
Article history:
Accepted March 2, 2009
Published online ahead of
print on March 10, 2009
Keywords:
Epidemiology
Conservative management
Curative management
Prostate cancer
Radical prostatectomy
Radiotherapy
Abstract
Background: There is insufficient information regarding the benefit of treatment with
curative intent for men with localised poorly differentiated prostate cancer (PCa).
Objective: To evaluate relative survival in men with potentially curable PCa in relation
to Gleason score (GS) and treatment as practiced in the community at large.
Design, setting, and participants: A population-based study including all men with
localised PCa registered in Sweden’s National Prostate Cancer Register.
Interventions: Hormonaltherapy,watchful waiting,and treatment with curativeintent.
Measurements: The ratio of observed deaths to expected deaths, determined from
survival in the general male population of the same age, was assessed using Poisson
regression analysis, with GS and treatment as covariates. Interaction between GS and
treatment was tested in a multivariate Cox proportional hazard analysis.
Results and limitations: A total of 31 903 men with potentially curable tumour (T1–T3,
N0/NX, M0/MX, age <75 yr, and prostate-specific antigen [PSA] <20 ng/ml) were
identified. GS was recorded for 28 454 of these men. Some 19 606 men (60.8%) were
treated with curative intent, and 12 645 men (39.2%) were given either hormonal
treatment or expectant management. The ratios between observed and expected
survival gradually increased for men with GS 10, with GS to 3.3 for men treated
conservatively and to 1.4 for men treated with curative intent. There was a significant
interaction between GS and treatment, with a relatively greater benefit from treatment
withcurativeintentformenwithhigh-gradetumours.Theresultshavetobeinterpreted
with some caution, as there was no randomisation between the treatment groups.
Conclusions: Survival for men with well-differentiated tumours is close to that of the
general population, regardless of treatment, but the outcome is dismal for men with
poorly differentiated tumours, whichever treatment is applied. Nevertheless, men with
poorly differentiated tumours benefit more from curative treatment than do men with
well- differentiated tumours.
# 2009 European Association of Urology. Published by Elsevier B.V. All rights reserved.
* Corresponding author. Department of Urology, University Hospital, 751 85 Uppsala, Sweden.
Tel. +46 18 611 00 00; Fax: +46 18 611 50 56.
E-mail address: Sam.ladjevardi@surgsci.uu.se (S. Ladjevardi).
0302-2838/$ – see back matter # 2009 European Association of Urology. Published by Elsevier B.V. All rights reserved.doi:10.1016/j.eururo.2009.03.007

Page 2

1. Introduction
The treatment of men with localised prostate cancer (PCa)
hasundergonerapiddevelopmentinrecentyears;however,
the favourable prognosis of these tumours, even if they are
left untreated, has made the management of men with
localised PCa acontroversial issue. In a Swedish randomised
controlled trial, Scandinavian Prostate Cancer Group 4
(SPCG-4), it was found that radical prostatectomy (RP) may
improve overall survival if the tumour is detected at an
early stage [1]. Nevertheless, RP and curative radiation
therapy (RT) may lead to overtreatment of men with slowly
progressing tumours that do not benefit from aggressive
treatment. The number of men needed to treat in order to
avoid one PCa death in SPCG-4 was estimated to be 20. In
another study based on men with favourable-risk PCa, it
was estimated that approximately 73 patients would
require radical treatment for each PCa death averted [2].
Similar results were also found in a study based on tumours
detected in the European Randomised Study of Screening
for Prostate Cancer [3]. A careful decision regarding the
management of these men is required to avoid over-
treatment. The outcome after RP or RT in men with poorly
differentiated tumours is even more uncertain, since SPCG-
4 only included men with well- and moderately differ-
entiated tumours. Men with poorly differentiated tumours
were not included in this study; it was considered unethical
to include them in a randomised study with expectant
management in one of the arms.
The aim of the present study is to describe relative
survival in a large, unselected, population-based cohort of
men with potentially curable PCa. A study regarding the
prognostic impact of prostate-specific antigen (PSA) in men
with localised PCa based on the same register has been
recentlypublished[4]. Inthepresent study,theanalysesare
based on the degree of differentiation and treatment.
2. Materials and methods
The study is based on Sweden’s National Prostate Cancer Register (NPCR)
[5]. The NPCR was started in 1996 as an extension of the Swedish Cancer
Register (SCR), which has been shown to have >98% completeness [6]. In
1996, the NPCR covered four of the six health care regions of Sweden; in
1997, it covered five regions, and since 1998, it has covered the whole of
Sweden. Because the NPCR relies on the SCR as a basic source of data,
virtually complete coverage of all cases of PCa diagnosed in the
population is ensured. In addition to data from the SCR, the NPCR also
includes data on tumour characteristics and treatment. Data regarding
treatmentin theNPCRarebasedonthefirst6moafter diagnosis.Data on
treatment following failure after first-line therapy are not registered in
the NPCR.
The aim of the NPCR is to provide a base for quality assurance of PCa
care and assessment of risk factors and outcomes. The accuracy of the
NPCR has been regionally validated, showing high accuracy of the
recorded data [7].
Personal registration numbers [8] are used to link data between the
tworegisters.Basicdata,includingdateofdiagnosisanddateofbirth,are
obtained from the SCR. Data on tumour stage according to the TNM
classification [9], the Gleason score (GS), or the World Health
Organisation (WHO) grade; PSA; treatment; and date of death were
obtained from the NPCR.
The process of assembling the cohort is shown in Fig. 1. The study is
based on men diagnosed 1 January 1996 to 31 December 2006. They
were followed until 31 December 2007.
2.1. Statistics
Relative survival was calculated by dividing the observed survival of the
cancer patients with the expected survival in the general male
population of corresponding age. In calculations of period estimates
of relative survival, the period methodology is applied to the expected
survivalratesin thesame wayasto theobserved survival rates[10].Data
on expected survival were obtained from Sweden’s statistical databases
[11].
Relative survival was estimated with stratification for treatment and
tumour differentiation. The tumours were defined as well differentiated
(GS 2–6 or WHO 1), moderately differentiated (GS 7 or WHO 2), and
poorly differentiated (GS 8–10 or WHO 3) [12]. This study was based on
men with tumours that could be considered potentially curable (ie, T1–
T3, NX/N0, M0/MX, PSA <20 ng/ml, and age <75 yr) [13]. In a separate
analysis, the overall survival was tested in a multivariate analysis Cox
proportional hazard analysis, with age, GS, and treatment as covariates.
We also tested whether there was an interaction between GS and
treatment.
3. Results
A total of 81 195 men were registered in the NPCR between
1996 and 2006. Mean age was 65.2 yr (standard deviation:
6.3 yr). Median follow-up time was 4.4 yr (range: 0–12 yr).
Overall, 31 903 men fulfilled the criteria for being
Fig. 1 – Flow chart of cohort assembly. PSA = prostate-specific antigen;
WHO = World Health Organisation; T4 = tumour classification 4;
N1 = regional lymph node metastasis; M1 = distant metastasis.
EUROPEAN UROL OG Y 57 (2 010) 631–640
632

Page 3

Table 1 – Baseline characteristics by treatment*
Conservative managementCurative managementTotal
Palliative treatmentsWatchful waiting Radical prostatectomyRadiotherapy
n
%
n
%
n
%
n
%
n
%
Age
<55
55–59
60–64
65–69
70–75
21
77
243
686
2183
0.7
2.4
7.6
21.4
68.0
183
647
1486
2544
4575
1.9
6.9
15.7
27.0
48.5
1310
2928
4259
3433
1020
10.1
22.6
32.9
26.5
7.9
234
788
1583
2104
1599
3.7
12.5
25.1
33.4
25.3
1748
4440
7571
8767
9377
4.1
11.1
20.3
27.1
37.4
Year of diagnosis
1996–2000
2001–2003
2004–2006
1195
1043
972
37.2
32.5
30.3
3281
2662
3492
34.8
28.2
37.0
2361
3953
6636
18.2
30.5
51.2
1580
2148
2580
25.0
34.1
40.9
8417
9806
13680
28.8
31.3
39.9
GS
2
3
4
5
6
7
8
9
10
Missing GS, WHO grade 1
Missing GS, WHO grade 2
Missing GS, WHO grade 3
11
26
61
243
829
896
331
190
29
173
297
124
0.3
0.8
1.9
7.6
25.8
27.9
10.3
5.9
0.9
5.4
9.3
3.9
116
168
787
1466
4004
1071
120
1.2
1.8
8.3
15.5
42.4
11.4
1.3
0.4
0.1
12.0
5.0
0.6
25
74
358
1549
6556
3140
478
127
0.2
0.6
2.8
12.0
50.6
24.2
3.7
1.0
0.1
2.5
2.0
0.4
9 0.1
0.8
2.5
9.9
38.3
29.1
6.6
3.3
0.5
3.3
4.4
1.2
161
318
1361
3883
13802
6942
1347
564
0.5
1.0
3.9
11.2
39.3
23.1
5.5
2.7
0.4
5.8
5.2
1.5
50
155
625
2413
1835
418
211
29
210
278
75
36
810 76
1132
471
322
257
54
1837
1303
30956
GS or WHO grade
GS 2–6, WHO 1
GS 7, WHO 2
GS 8–10, WHO 3
1343
1193
674
0.4
0.4
0.2
7673
1542
220
0.8
0.2
0.0
8884
3397
669
0.7
0.3
0.1
3462
2113
733
0.5
0.3
0.1
21362
8245
2296
0.7
0.3
0.1
PSA
<4
4–<10
10–<20
180
1248
1782
5.6
38.9
55.5
1488
5072
2875
15.8
53.8
30.5
1275
8215
3460
9.8
63.4
26.7
365
3287
2656
5.8
52.1
42.1
3308
17822
10773
9.2
52.0
38.7
T category
T0
T1 (abc missing)
T1a
T1b
T1c
T2
T3
TX
180.6
0.0
2.2
3.3
23.3
39.5
30.2
1.0
133
18
1163
444
4536
2540
504
1.4
0.2
12.3
4.7
48.1
26.9
5.3
1.0
46
13
166
133
7422
4727
364
79
0.4
0.1
1.3
1.0
57.3
36.5
2.8
0.6
23
5
39
80
0.4
0.1
0.6
1.3
39.5
38.0
19.6
0.5
220
37
1439
762
15200
10933
3071
241
0.7
0.1
4.1
2.6
42.0
35.2
14.5
0.8
1
71
105
748
1267
969
31
2494
2399
1234
3497
N category
N0
NX
147
3063
4.6
95.4
190
9245
2.0
98.0
4178
8772
32.3
67.7
2036
4272
32.3
67.7
6551
25352
17.8
82.2
M category
M0
MX
1243
1967
38.7
61.3
1982
7453
21.0
79.0
4398
8552
34.0
66.0
3512
2796
55.7
44.3
11135
20768
37.4
62.7
Mode of detection**
Screening
Symptoms
Other reasons
Missing information
389
1574
201
174
16.6
67.3
8.6
7.4
2233
3431
878
397
32.2
49.4
12.7
5.7
5275
4207
1260
648
46.3
36.9
11.1
5.7
1952
2463
514
367
36.9
46.5
9.7
6.9
9849
11675
2853
1586
33.0
50.0
10.5
6.4
Mean follow-up, yr (SD)4.6(2.7) 4.8 (2.9) 4.0(2.6)4.4 (2.6) 4.3 (2.7)
GS = Gleason score; SD = standard deviation; WHO = World Health Organisation.
* All men registered in the Swedish National Prostate Cancer Register with potentially curable prostate cancer (T1–T3, NX/NO, MX/MO, PSA <20 ng/ml, and age
?75 yr).
** Available from year 2000.
E U RO P E A N U R O L O GY 57 ( 2 0 1 0 ) 6 3 1 – 6 4 0
633

Page 4

considered as potential candidates for radical treatment
(T1–T3, N0/NX, M0/MX, age <75 yr, and PSA<20 ng/ml
[13]). All data presented in this report are based on this
group of men with localised tumours.
A total of 3449 (11.5%) patients had their diagnosis
before GS was included in the NPCR, and they were
classified by cytology. Tumour grades for these men were
WHO 1 (n = 1837; 5.8%), WHO 2 (n = 1303; 5.2%), and WHO
3 (n = 309; 1.5%).
A total of 27 228 men were diagnosed with PCa in year
2000 or later. From 2000 on, the basis for cancer diagnosis
was included in the register. PCa was detected by screening
in 9849 men (33%) and following presentation with clinical
symptoms in 11 675 men (50%). In 2853 men (10.5%), the
basis for cancer diagnosis was uncertain; in 1586 men
(6.4%), information was missing (Table 1).
The percentage of men undergoing radical treatment
was approximately the same for men with well-, moder-
ately, and poorly differentiated tumours, but there were
more men who received palliative treatment in the high-
grade group (Table 2).
The relative survival is shown in Fig. 2. Although the
relative survival of men with well-differentiated tumours
was >100% during the first 5 yr regardless of treatment, the
survival rate was dismal for men with moderately and
poorly differentiated tumours who were selected for
conservative management.
The ratio between observed and expected deaths by GS
and treatment in men diagnosed between 2000 and 2006 is
shown in Fig. 3. The ratio of patients treated with curative
intent varied little with tumour grade. For those who were
managed conservatively, the ratio was considerably higher
for men with poorly differentiated tumours than for those
with well-differentiated tumours.
In the multivariate Cox proportional hazards analysis,
treatment, age, and GS were found to significantly and
independently predict overall survival (Table 3). Further-
more, there were significant interactions between GS and
treatment, showing relatively greater benefit for survival
fromtreatmentwithcurativeintentformenwithhigherGS.
4.Discussion
The relative survival of men treated with radical intent and
those treated with expectant management diverged much
more in the intermediate- and the high-grade groups than
Fig. 2 – Relative survival: (a) Gleason score 2–6; (b) Gleason score 7; (c) Gleason score 8–10.
EUROPEAN UROL OG Y 57 (2 010) 631–640
634

Page 5

Table 2 – Distribution of primary treatment by Gleason score (GS)*
GS 2–6 GS 7 GS 8–10
PSA <4 PSA 4–<10 PSA 10–<20 PSA <4 PSA 4–<10 PSA 10–<20 PSA <4 PSA 4–<10 PSA 10–<20 Total
n
%
n
%
n
%
n
%
n
%
n
%
n
%
n
%
n
%
n
Total no. of cases2754– 13 204– 6291– 564– 4439– 3688– 153– 1018– 1229– 33 340
Conservative
management
Watchful waiting132848.2 427632.42069 32.9143 25.4 70015.8 69919 1711.1 969.4 107 8.7 9435
Palliative treatments
Bilateral orchiectomy
GnRH analogues
Antiandrogenes
Other palliative
treatments
95
5
36
8
46
3.4
0.2
1.3
0.3
1.7
588
26
281
176
105
4.5
0.2
2.1
1.3
0.8
660
33
350
184
92
10.5
0.5
5.6
2.9
1.5
43
4
28
7
4
7.6
0.7
5
1.2
0.7
433
33
267
104
28
9.8
0.7
6.0
2.3
0.6
717
57
462
155
43
19.4
1.5
12.5
4.2
1.2
42
6
27
4
4
27.5
3.9
17.6
2.6
2.6
227
19
151
39
14
22.3
1.9
14.8
3.8
1.4
405
32
276
74
23
33 3210
215
1878
751
359
2.6
22.5
6
1.9
Curative management
RP
RT (external)
RT (brachytherapy)
Other curative
treatments
1239
985
132
87
35
45.0
35.8
4.8
3.2
1.3
7941
5832
1240
725
144
60.1
44.2
9.4
5.5
1.1
3386
2067
946
332
41
53.8
32.9
15
5.3
0.7
353
241
70
35
62.6
42.7
12.4
6.2
1.2
3119
2039
753
275
52
70.3
45.9
17.0
6.2
1.2
2137
1117
751
229
57.9
30.3
20.4
6.2
1.1
92
49
26
15
2
60.1
32
17
9.8
1.3
656
344
219
75
18
64.4
33.8
21.5
7.4
1.8
683
276
306
92
55.6
22.5
24.9
7.5
0.7
19 606
12 950
4443
1865
3487 409
Missing information92 3.3 3993 1762.8 254.4 1874.2135 3.72 1.3 39 3.8 34 2.8 1089
GnRH = gonadotropin-releasing hormone; PSA = prostate-specific antigen; RP = radicall prostatectomy; RT = radiation therapy.
* All men registered in the Swedish National Prostate Cancer Register with potentially curable prostate cancer (T1[enT3, NX/NO, MX/MO, PSA <20 ng/ml, and age ?75 yr).
E U RO P E A N U R O L O GY 57 ( 2 0 1 0 ) 6 3 1 – 6 4 0
635

Page 6

in men with low-grade tumours, despite the fact that the
prognosisforalltreatments,includingRPandRT,wasworse
inthe high-grade group. This difference increased withtime
elapsed since diagnosis. Men with poorly differentiated
tumours should be given highest priority for radical
treatment, even though a relatively large percentage of
these men are not saved by the treatment. The relative
survival of men with highly differentiated tumours was
>100% during the first 5 yr, whichever treatment is given.
Beyond 5 yr after diagnosis, the relative survival fell very
slowly to <100% for men with highly differentiated
tumours that were managed conservatively.
With 31 903 men, this study represents the largest
population-based analysis to date. It reveals the relative
survival of men with clinically localised PCa, comparing
treatment with curative intent with conservative manage-
ment for low- and high-grade tumours. All men residing in
the area covered by the NPCR were included in the analysis,
which minimised the risk of patient selection. In contrast
with randomised controlled trials, which are usually
performed in centralised units with optimal resources
and the highest professional competence, the outcome of a
population-based register study, like the present one,
shows the effectiveness of the interventions as they are
practiced in the community at large. The external validity is
not limited by the selection mechanism that is associated
with a randomised controlled trial.
These data are derived from a population in which
screening is rarely performed (Table 1) and in which
conservative management of localised tumours is more
favoured than in North America. Most of the men registered
received their diagnosis before the first results from the
SPCG-4 trial were published [14]; this study has funda-
mentally changed the view on the management of PCa in
Scandinavia. Half of the men were diagnosed with PCa after
the occurrence of local symptoms. This may limit the
applicabilityofthestudy,sincemostmenwithlocalisedPCa
are now diagnosed with elevated PSA in the absence of
symptoms. Thus, the cohort studied may be composed of a
relatively larger proportion of men with more advanced
local growth than if assembled today, which may account
for some of the difference between active and conservative
treatment. In recent years, the use of PSA has led to the
detection of PCa at anearlier stage [15]. In the United States,
men with newly diagnosed PCa undergoing RP have GS ?8
in 8–21% of cases [16–18]. In the cohort presented in this
paper, the corresponding percentage for men treated with
RP was 27%. This difference may be explained by the fact
that a larger percentage of small indolent tumours remain
undetected, since screening is not widely practiced. This
may have affected theoutcomeof ourstudy, since menwith
low-grade tumours that would have been detected in the
United States and subsequently treated radically may not
have been diagnosed in the Swedish population.
Another possible source of confounding may be that
patients with only one core biopsy with GS 8 could be
treated differently than patients with several core biopsies
with the same GS. Despite the wide acceptance of GS as
Table 3 – Multivariate Cox proportional hazard analysis with
overall survival as the end point*
Model 1Model 2
HR95% CI HR 95% CI
CM
RP
RT
GS 2–6
GS 7
GS 8–10
Age at diagnosis
RP GS 7
RT GS 7
RP GS 8–10
RT GS 8–10
ref
0.36
0.54
ref
1.61
2.67
1.06
–
–
–
–
–
0.32–0.40
0.49–0.59
–
1.49–1.73
2.42–2.95
1.05–1.07
–
–
–
–
ref
0.41
0.62
ref
1.72
3.00
1.06
0.78
0.81
0.65
0.71
–
0.36–0.47
0.54–0.72
–
1.57–1.88
2.67–3.38
1.05–1.07
0.63–0.96
0.66–0.99
0.47–0.90
0.55–0.91
Model 3 Model 4
HR 95% CIHR 95% CI
CM
RP
RT
GS 2–6
GS 7
GS 8–10
Age at diagnosis
RP GS 7
RT GS 7
RP GS 8–10
RT GS 8–10
PSA 10–20 ng/ml
PSA 4–10 ng/ml
PSA <4 ng/ml
ref
0.37
0.54
ref
1.58
2.60
1.06
–
–
–
–
ref
0.82
0.96
–
0.33–0.41
0.49–0.59
–
1.46–1.70
2.35–2.87
1.05–1.06
–
–
–
–
–
0.77–0.89
0.86–1.08
ref
0.41
0.62
ref
1.69
2.92
1.06
0.78
0.81
0.65
0.71
ref
0.83
0.97
–
0.36–0.48
0.54–0.71
–
1.54–1.84
2.59–3.29
1.05–1.06
0.63–0.97
0.66–0.99
0.47–0.90
0.55–0.92
–
0.77–0.89
0.87–1.09
CI = confidence interval; CM = conservative management; GS = Gleason
score; HR = hazard ratio; PSA = prostate-specific antigen; RP = radical
prostatectomy; RT = radiotherapy.
* Model 1 includes age, GS, and treatment. In models 3 and 4, PSA is added
as a covariate. In models 2 and 4, the interaction between GS and
treatment is assessed by adding composite variables. The statistically
significant interaction between GS and treatment reflects the relatively
greater benefit from curative treatment for men with GS 7–10.
Fig. 3 – Ratio of observed deaths to expected deaths (patients diagnosed
2000–2006).
EUROPEAN UROL OG Y 57 (2 010) 631–640
636

Page 7

predictive factor, it does not provide a quantitative
assessment of the extent of tumour growth in the specimen.
Although a substantial number of men were followed
untildeathorforatleastadecade,themedianfollow-upwas
only 4.4 yr. The short median follow-up is explained mainly
bythefactthatthemajorityofmenwerediagnosedafteryear
2000. Although the number of men with substantial follow-
up was sufficient to make the Cox proportional hazard
analysis conclusive, a longer follow-up may have resulted in
tumourprogressioninmenwithwell-differentiatedtumours
and a more pronounced divergence between conservative
treatment and treatment with curative intent for this group.
Inpatientswithpoorlydifferentiatedtumours,therelative
survival estimate was well below 100%, reflecting excess
mortality among cancer patients compared with the general
population.Menwithwell-differentiatedtumours,however,
had a relative survival >100%, regardless of treatment
(Fig. 2a), at least during the first 5 yr. This paradoxically
highrelativesurvivalrateisprobablyexplainedbyaselection
ofmenwithlowercomorbidityandtumoursthatdonotlimit
the survival of their hosts, in whom tumours are revealed
despite few or minor symptoms. Men with similar tumours
but more concurrent disease are less likely to undergo an
examination in which indolent, nonlethal tumours would be
detected. Consequently, men with low-grade PCa appear to
have better survival than the general population.
Previous studies have shown that patients with low-risk
PCa are very likely to remain disease free after RP, RT, or
brachytherapy [19]. There are, however, no randomised
controlled studies comparing the treatment options among
patients with poorly differentiated cancer. RP in men with
poorly differentiatedtumoursisacontroversialissue.Onthe
one hand, the chance of achieving a tumour-free resection
border is much lower in men with poorly differentiated
tumours, making surgery less successful. This was one of the
reasonswhypoorlydifferentiatedtumourswerenotincluded
inSPCG-4.Ontheotherhand,thepoorsurvivalofthesemenif
managedconservativelymakesitworthwhiletogivethemat
leastasmallchancewithradicaltreatment.Wefoundthatthe
relative benefit in terms of improvement in relative survival
when comparing radical treatment with conservative man-
agementwashighformenwithhigh-gradetumours,despite
the higher risk for treatment failure. Despite these rather
discouraging figures for tumour-free resection border and
survivalinmenwithhigh-grade tumoursfollowing RPorRT,
thedismalprognosisiftreatedconservatively(Fig.2c)should
be kept in mind. Even a small chance of cure may seem, in
comparison, a much better alternative.
Although cancer-specific survival may be biased by
inaccurate registration of causes of death, there is little risk
of bias for relative survival if the expected survival is
determined from the background population from which
the study population is derived [20]. In the present study,
the study populations as well as the life tables for the
background population cover the entire country.
There are several reports discussing the results of RP and
RT in patients with organ-confined PCa, although none are
randomised. Tward et alpresented a large analysisincluding
60 290 men with clinically localised PCa in the Surveillance,
Epidemiology, and End Results (SEER) programme [21]. In
that study, the outcomes of patients who were treated with
surgery,RT, and nodefinitivetreatment werecompared,and
the authors concluded that PCa-specific mortality after 10 yr
waslowerforpatientswhoreceivedtreatmentwithintentto
cure. Tewari et al compared long-term survival in a retro-
spective cohort study of patients with PCa with GS ?8 who
were treated with conservative management, RT, and RP.
Theyinvestigated453menwithGS8–10,197(44%)ofwhom
weretreatedconservatively,137(30%)ofwhomreceivedRT,
and 119 (26%) of whom underwent RP. Using propensity
scoring analysis, median overall survival for conservative
therapy, RT, and RP was found to be 5.2, 6.7, and 9.7 yr,
respectively. Median cancer-specific survival was 7.8 yr for
conservative therapy and >14 yr for RT and RP. The risk of
cancer-specific death following RP was 68% lower than for
conservativetreatment(p<0.001)and49%lowerthanforRT
(p = 0.053) [22].
The American Urological Association guidelines are
based on a review of 13 888 articles, citations, and abstracts
reporting outcomes of PCa treatment in patients with
clinical stage T1 or T2 disease between 1991 to 2004 [23].
According to this review, the lack of evidence of superiority
of one therapy over another gives support to active
surveillance, RT, and RP as treatment options for patients
with high-risk disease. Nevertheless, active treatment is the
preferred option to avoid disease progression and death
from disease progression.
The risk of selection bias when assessing the outcome
after different treatments is a well-known fact and should
not be neglected. As shown in Table 3, there was a
significant difference in overall survival between patients
who underwent RP and RT. This difference, however, must
be interpreted with caution, since therapy decisions may be
based on several circumstances, particularly comorbidity,
that may serve as confounders. This is an insolvable
problem when comparing the outcomes of men undergoing
RP with those undergoing RT, since the latter is usually
considered the method of choice for men with high
comorbidity. Furthermore, RT as practiced today cannot
be compared with the way it was performed 10 yr ago. The
recent improvements in radiation techniques, with better
dose schemes and optimised doses, result in better survival
rates than were achieved a decade ago.
Straightforward comparisons of survival between differ-
ent treatment regimes in a nonrandomised study are always
biased by confounders, particularly comorbidity. No definite
conclusions can be drawn from the differences in survival
between the different treatments shown in Fig. 2. The
significant interaction between treatment and GS shown in
Table3,however,cannotbeattributedtocomorbidity,unless
there were an association between comorbidity and GS in
conjunction with treatment. In other words, although the
divergenceofrelativesurvivalratesafterdifferenttreatments
may be explained by confounding factors, those factors
cannot readily explain the more pronounced divergence in
survivalformenwithpoorlydifferentiatedtumoursandwith
well-differentiated tumours. Several confounding factors
mayinfluencethechanceofmenwithindolent oraggressive
E U RO P E A N U R O L O GY 57 ( 2 0 1 0 ) 6 3 1 – 6 4 0
637

Page 8

tumours to get their diagnosis as well as their subsequent
treatmentchoice;however,thesefactorswouldnotaffectthe
interaction between GS and treatment in survival.
5. Conclusions
In conclusion, our study shows that the outcome differs
little between conservative management and treatment
with curative intent in men with localised well- to
moderately differentiated tumours within a 10-yr period.
In contrast, for men with poorly differentiated tumours, the
treatment choice is crucial for the outcome. The high risk of
tumour progression despite treatment should not discou-
rage clinicians from offering these men surgery or RT.
Despite the lower rate of treatment success for these men,
the potential chance of benefit in terms of increased
survival after treatment is much higher than in men with
well- to moderately differentiated tumours.
Author contributions: Sam Ladjevardi had fullaccess 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: Ladjevardi, Sandblom, Varenhorst.
Acquisition of data: Varenhorst.
Analysis and interpretation of data: Ladjevardi, Sandblom, Berglund.
Drafting of the manuscript: Ladjevardi.
Critical revision of the manuscript for important intellectual content:
Sandblom, Varenhorst.
Statistical analysis: Berglund.
Obtaining funding: Varenhorst.
Administrative, technical, or material support: Varenhorst.
Supervision: Sandblom, Varenhorst.
Other (specify): None.
Financial disclosures: I certify 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, consultancies, 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: None.
References
[1] Bill-Axelson A, Holmberg L, File ´n F, et al. Radical prostatectomy
versus watchful waiting in localized prostate cancer: the Scandi-
navian Prostate Cancer Group-4 randomized trial. J Natl Cancer Inst
2008;100:1144–54.
[2] Klotz L. Active surveillance versus radical treatment for favorable-
risk localized prostate cancer. Curr Treat Options Oncol 2006;7:
355–62.
[3] Roemeling S, Roobol MJ, Postma R, et al. Management and survival
of screen-detected prostate cancer patients who might have been
suitable for active surveillance. Eur Urol 2006;50:475–82.
[4] Sandblom G, Ladjevardi S, Garmo H, Varenhorst E. The impact of
prostate-specific antigen level at diagnosis on the relative survival
of 28,531 men with localized carcinoma of the prostate. Cancer
2008;112:813–9.
[5] Varenhorst E, Garmo H, Holmberg L, et al. The National Prostate
Cancer Register in Sweden 1998–2002: trends in incidence, treat-
ment and survival. Scand J Urol Nephrol 2005;39:117–23.
[6] Cancer incidence in Sweden 2006. Regionalt Onkologiskt Centrum
Web site [in Swedish]. http://www.roc.se/cancerstat.asp. Accessed
27 June 2008.
[7] Sandblom G, Dufmats M, Olsson M, Varenhorst E. Validity of a
population-based cancer registerin Sweden—an assessmentof data
reproducibility in the South-East Region Prostate Cancer Register.
Scand J Urol Nephrol 2003;37:112–9.
[8] Population registration in Sweden. Skatteverket Web site. http://
www.skatteverket.se/blanketterbroschyrer/broschyr/info/717b.4.
39f16f103821c58f680008017.html. Accessed 9 July 2008.
[9] Sobin LH. TNM, sixth edition: new developments in general con-
cepts and rules. Semin Surg Oncol 2003;21:19–22.
[10] Dickman PW, Slogett A, Hills M, Hakulinen T. Regression models for
relative survival. Stat Med 2004;23:51–64.
[11] Statistikdatabasen [population statistics]. Statistics Sweden Web
site [in Swedish]. http://www.ssd.scb.se. Accessed 27 June 2008.
[12] Murphy GP, Busch C, Abrahamsson PA, et al. Histopathology of
localized prostate cancer. Consensus Conference on Diagnosis and
PrognosticParametersinLocalizedProstateCancer.Stockholm,Swe-
den, May 12–13, 1993. Scand J Urol Nephrol Suppl 1994;162:7–42.
[13] Aus G, Abbou CC, Bolla M, et al. EAU guidelines on prostate cancer.
Eur Urol 2005;48:546–51.
[14] Holmberg L, Bill-Axelson A, Helgesen F, et al. A randomized trial
comparing radical prostatectomy with watchful waiting in early
prostate cancer. N Engl J Med 2002;347:781–9.
[15] Amling CL, Blute ML, Lerner SE, Bergstralh EJ, Bostwick DG, Zincke
H. Influence of prostate-specific antigen testing on the spectrum of
patients with prostate cancer undergoing radical prostatectomy at
a large referral practice. Mayo Clin Proc 1998;73:401–6.
[16] Perrotti M, Rabbani F, Farkas A, Ward WS, Cummings KB. Trends in
poorly differentiated prostate cancer 1973 to 1994: observations
from the Surveillance. Epidemiology and End Results database. J
Urol 1998;160:811–5.
[17] Tefilli MV, Gheiler EL, Tiguert R, et al. Role of radical prostatectomy
in patients with prostate cancer of high Gleason score. Prostate
1999;39:60–6.
[18] Zincke H, Oesterling JE, Blute ML, Bergstralh EJ, Myers RP, Barrett
DM. Long-term (15 years) results after radical prostatectomy for
clinically localized (stage T2c or lower) prostate cancer. J Urol
1994;152:1850–7.
[19] Grossfeld GD,LatiniDM,LubeckDP,MehtaSS,CarrollPR.Predicting
recurrence after radical prostatectomy for patients with high risk
prostate cancer. J Urol 2003;169:157–63.
[20] Boer R, Ries L, Ballegooijen V, Feuer EJ, Legler J, Habbema D;
National Cancer Institute. Ambiguities in calculating cancer patient
survival: the SEER experience for colorectal and prostate cancer.
http://srab.cancer.gov/reports/tech2003.05.pdf. Published 2002.
Accessed 23 December 2008.
[21] Tward JD, Lee CM, Pappas LM, Szabo A, Gaffney DK, Shrieve DC.
Survival of men with clinically localized prostate cancer treated
with prostatectomy, brachytherapy, or no definitive treatment:
impact of age at diagnosis. Cancer 2006;107:2392–400.
[22] Tewari A, Divine G, Chang P, et al. Long-term survival in men with
high grade prostate cancer: a comparison between conservative
treatment, radiation therapy and radical prostatectomy—a propen-
sity scoring approach. J Urol 2007;177:911–5.
[23] Thompson I, Thrasher JB, Aus G, et al. Guideline for the manage-
ment of clinically localized prostate cancer: 2007 update. J Urol
2007;177:2106–31.
EUROPEAN UROL OG Y 57 (2 010) 631–640
638

Page 9

Editorial Comment on: Tumour Grade, Treatment,
and Relative Survival in a Population-based
Cohort of Men with Potentially Curable Prostate
Cancer
Andrea Gallina
Universita ` Vita Salute San Raffaele, Via Olgettina 60,
20132 Milan, Italy
gallina.andrea@hsr.it
Available guidelines state that radical prostatectomy
(RP) is the mainstay treatment for clinically localized
prostate cancer (PCa) in men with a life expectancy >10 yr
[1]. This guideline is based on several observations that
demonstrate a lack of significant difference in survival
between conservative management and definitive treat-
ment for PCa. All reports, however, included only a small
number of the most difficult population to manage:
patients who are diagnosed with poorly differentiated PCa
(Gleason score 8–10) [2].
The paper by Ladjevardi and colleagues [3] focuses on a
large cohort of patients treated with either curative intent
(RP or radiotherapy [RT]) or conservative management
(hormonal therapy or expectant management). The
authors elegantly tested the association between treat-
ment received and patients’ overall and cancer-specific
survival, highlighting the correlation with Gleason score.
This paper, albeit limited by a retrospective design,
demonstrated a significant decrease in the observed
survival in patients who were diagnosed with poorly
differentiated PCa and who were managed conservatively,
as compared with those who were treated with curative
intent. The implications of this study are important:
Patients with poorly differentiated PCa may deserve active
treatment, independent of age, since not only may
symptoms and local progression be limited but survival
is also improved when compared with palliative treat-
ment or observation.
It may be argued that curative treatment has several
sequelae that affect quality of life. As was recently
observed, however, hormonal treatment is not devoid of
complications [4]. Other studies are needed to confirm
those observations and to carefully stratify patients in
order to identify as many potentially curable patients as
possible, since current available information at diagnosis
is not sufficient for a correct definition of the risk of
harboring potentially life-threatening cancers [5]. Based
on these considerations, Ladjevardi et al’s paper has to be
considered as a basis for a potential new treatment
paradigm in patients who are diagnosed with poorly
differentiated PCa.
References
[1] Heidenreich A, Aus G, Bolla M, et al. EAU guidelines on prostate
cancer. Eur Urol 2008;53:68–80.
[2] Bill-Axelson A, Holmberg L, Ruutu M, et al. Radical prostatectomy
versus watchful waiting in early prostate cancer. N Engl J Med
2005;352:1977–84.
[3] Ladjevardi S, Sandblom G, Berglund A, Varenhorst E. Tumour
grade, treatment, and relative survival in a population-based
cohort of men with potentially curable prostate cancer. Eur Urol
2010;57:631–40.
[4] Isbarn H, Boccon-Gibod L, Carroll PR, et al. Androgen deprivation
therapy for the treatment of prostate cancer: consider both
benefits and risks. Eur Urol 2009;55:62–75.
[5] Chun FK-H, Suardi N, Capitanio U, et al. Assessment of pathological
prostate cancer characteristics in men with favorable biopsy fea-
tures on predominantly sextant biopsy. Eur Urol 2009;55:617–28.
DOI: 10.1016/j.eururo.2009.03.008
DOI of original article: 10.1016/j.eururo.2009.03.007
Editorial Comment on: Tumor Grade, Treatment,
and Relative Survival in a Population-Based
Cohort of Men with Potentially Curable Prostate
Cancer
Ofer Yossepowitch
Department of Urology, Rabin Medical Center,
Petah-Tikva, Israel 41000
oferyoss@netvision.net.il
The optimal management of clinically localized prostate
cancer (PCa) remains elusive. Radical prostatectomy (RP),
external-beam radiation therapy (EBRT), brachytherapy,
and active surveillance with delayed curative intervention
have all been proven effective for men with localized
disease. Primary androgen deprivation therapy (ADT),
which offers no cure, is also widely accepted as a
treatment alternative in this setting [1]. Until consensus
is reached regarding the optimal strategy, patients and
their physicians who are considering therapy for PCa will
continue to struggle with how to interpret results from
comparative retrospective analyses.
In the present article, the authors demonstrate a
survival advantage favoring curative interventions (RP
or EBRT) over expectant management (ADT or watchful
waiting) [2]. This benefit was noted in high-grade tumors
but not in low- or intermediate-grade tumors. While the
plea for local definitive therapy is justifiable [3] and,
intuitively, is appealing, several concerns must be
emphasized before endorsing a routine policy of encoura-
ging or discouraging men with PCa from having surgery or
radiotherapy. First, compared to men who had RP, those
whohad palliative therapywere mucholder and hadmore
advanced clinical stage and higher prostate-specific
antigen (PSA) levels (Table 1 in Ladjevardi et al [2]). Thus,
E U RO P E A N U R O L O GY 57 ( 2 0 1 0 ) 6 3 1 – 6 4 0
639

Page 10

the improved survival could simply reflect a more
favorable disease rather than the impact of therapy itself.
Second, the natural history and, particularly, the treated
history of low-grade PCa is remarkably protracted [4].
While adequately powered, the study lacks appropriate
follow-up to address a survival end point in the low-risk
setting. Third, analyzing overall survival rather than
cancer-specific mortality should rationally adjust for
patients’ comorbidities. The latter remains unaccounted
for in this paper.
Generally, an important limitation of nonrandomized
studies is the potential for biases associated with
unmeasured confounding variables, some of which are
certainly acknowledged in the manuscript. In addressing
this issue, the authors state, ‘‘In contrast with randomised
controlled trials, which are usually performed in centra-
lised units with optimal resources and the highest
professional competence, the outcome of a population-
based register study, like the present one, shows the
effectiveness of the interventions as they are practiced in
the community at large.’’ This assertion is undoubtedly
valid; however, one must wonder whether basing world-
classresearchonreal-worldpracticeisideal.Iwouldargue
that the optimal therapy for both high- and low-risk PCa
remains controversial and warrants proper assessment by
prospective, rigorous, and well-controlled research.
References
[1] Shahinian VB, Kuo YF, Freeman JL, Orihuela E, Goodwin JS.
Increasing use of gonadotropin-releasing hormone agonists for
the treatment of localized prostate carcinoma. Cancer 2005;103:
1615–24.
[2] Ladjevardi S, Sandblom G, Berglund A, Varenhorst E. Tumour
grade, treatment, and relative survival in a population-based
cohort of men with potentially curable prostate cancer. Eur Urol
2010;57:631–40.
[3] Yossepowitch O, Eggener SE, Serio AM, et al. Secondary therapy,
metastatic progression,and cancer-specific mortality in men with
clinically high-risk prostate cancer treated with radical prosta-
tectomy. Eur Urol 2008;53:950–9.
[4] Freedland SJ, Humphreys EB, Mangold LA, et al. Risk of prostate
cancer-specific mortality following biochemical recurrence after
radical prostatectomy. JAMA 2005;294:433–9.
DOI: 10.1016/j.eururo.2009.03.009
DOI of original article: 10.1016/j.eururo.2009.03.007
EUROPEAN UROL OG Y 57 (2 010) 631–640
640