e u r o p e a n u r o l o g y 5 4 ( 2 0 0 8 ) 1 0 7 – 1 1 7

avai lab le at www.sciencedi rect .com

journa l homepage: www.europeanurology.com

Prostate Cancer

Poor Overall Survival in Septa- and Octogenarian Patientsafter Radical Prostatectomy and Radiotherapy for ProstateCancer: A Population-Based Study of 6183 Men

Claudio Jeldres a, Nazareno Suardi a,b, Jochen Walz a,c, Fred Saad a,Georg C. Hutterer a,d, Naeem Bhojani a, Shahrokh F. Shariat e,Paul Perrotte a, Markus Graefen f, Francesco Montorsi b, Pierre I. Karakiewicz a,*aCancer Prognostics and Health Outcomes Unit, University of Montreal Health Center, Montreal, QC, CanadabDepartment of Urology, Vita-Salute University San Raffaele, Milan, ItalycDepartment of Urology, University Medical Centre Eppendorf, Hamburg, GermanydDepartment of Urology, Graz Medical University, Graz, AustriaeDepartment of Urology, University of Texas Southwestern Medical Center, Dallas, TX, USAfMartini-Clinic, Prostate Cancer Centre, Hamburg, Germany

Article info

Article history:Accepted October 19, 2007Published online ahead ofprint on October 30, 2007

Keywords:ComorbidityProstate cancerRadical prostatectomyRadiotherapySurvival analysis

Abstract

Objectives: The life expectancy of candidates for attempted curative therapy of pros-

tate cancer should not be inferior to 10 yr. We examined the rate of 10-yr survival in

septa- and octogenarians treated for prostate cancer with either attempted curative

external beam radiotherapy (EBRT) or radical prostatectomy (RP).

Methods: Within a population-based cohort of 17,570 EBRTor RPpatients, 6183 men aged

70 yr or older were treated with either RP (n = 1591) or EBRT(n = 4592) and represented the

focus of crude survival analyses. Age and Charlson Comorbidity Index represented

covariates. To control for prostate cancer-specific mortality, we repeated the analyses in

a subset of 2704 men (RP, n = 881; EBRT, n = 1823) who had no clinical evidence of disease

relapse of prostate cancer.

Results: Overall actuarial 10-yr survival was 38.5% (RP 59.3% vs. EBRT 30.3%, p < 0.001)

versus 36.5% in those without clinical evidence of disease relapse (RP 63.8% vs. EBRT

22.6%, p < 0.001). In multivariate Cox regression models, EBRT was associated with a

2.1-fold (p < 0.001) and 2.9-fold (p < 0.001) higher risk of mortality relative to RP in all

men and in men without clinical evidence of disease relapse, respectively.

Conclusions: Forty percent of septa- and octogenarian men who are selected for RP do

fe expectancy to warrant attempted curative therapy. Even more

en who receive EBRT die before reaching the 10-yr mark. These

e the need for more stringent EBRT and RP selection criteria, if the

strikingly, 70% of m

findings may indicat

not have adequate li

overtreatment.

ociation of Urology. Published by Elsevier B.V. All rights reserved.

goal is to minimise

# 2008 European Ass

* Corresponding author. Cancer Prognostics and Health Outcomes Unit,University of Montreal Health Center (CHUM), 1058, rue St-Denis, Montreal, Quebec,Canada, H2X 3J4. Tel. +1 514 890 8000 35336; Fax: +1 514 227 5103.E-mail address: pierre.karakiewicz@umontreal.ca (P.I. Karakiewicz).

0302-2838/$ – see back matter # 2008 European Association of Urology. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.eururo.2007.10.038

e u r o p e a n u r o l o g y 5 4 ( 2 0 0 8 ) 1 0 7 – 1 1 7108

1. Introduction

Radical prostatectomy (RP) and external beamradiotherapy (EBRT) represent two commonly usedattempted curative treatment modalities for indivi-duals with localised prostate cancer (PCa) [1].Conservative management with delayed androgendeprivation or immediate androgen deprivationrepresent the main alternatives, especially in olderindividuals [2,3].

Favorable outcome may be expected in themajority of men, as long as the PCa grade is eitherlow or intermediate and as long as the host’s lifeexpectancy (LE) does not allow uninterrupted tumorprogression over several decades [4–6]. The combi-nation of favourable PCa grade and limited LE, whichusually is related to advanced age, indicates low riskof PCa-specific mortality.

The importance of tumor grade has been high-lighted in several recent contributions [7–10]. Inter-estingly, 81% of contemporary men treated with RPat a tertiary care institution between 1992 and 2004were diagnosed with well-differentiated or moder-ately differentiated tumors [11]. Similarly, 71% ofmen treated with EBRT between 1988 and 1995harboured well-differentiated or moderately differ-entiated PCa at diagnosis [12]. Therefore, on thebasis of tumor grade criteria, the majority of

Table 1 – Descriptive table for 6183 patients aged 70 yr or olderradiation therapy

Variables

All patients Radic

Total 6183 (100%)

Age at treatment (yr)

Mean (median) 74 (73)

Range 70–95

70–74 3962 (64.1%)

75–79 1611 (26.1%)

80–84 457 (7.4%)

�85 153 (2.5%)

Charlson Comorbidity Index

Mean (median) 1.9 (2.0)

Range 0.1–12

0 1750 (28.3%)

1–2 2515 (40.7%)

3–4 1308 (21.2%)

5–6 479 (7.7%)

>6 131 (2.1%)

Follow-up time (yr)

Mean (median) 5.2 (4.8)

Range 0.1–15.5

Deaths during follow-up 3267 (52.8%)

Actuarial survival (yr)

Median (mean) 7.4 (7.7)

contemporary men treated with attempted curativetherapy (RP or EBRT) harbour low- or intermediate-risk PCa, which may pose a minimal threat to thehost’s LE and may be overtreated with definitivetherapy [6,13]. To address the issue of potentialovertreatment, we examined the crude survivalafter attempted curative therapy in a population-based cohort of 6183 elderly men (�70 yr) treatedwith either RP or EBRT.

2. Methods

2.1. Study cohort

The Quebec Health Plan represents the exclusive health

insurer in the province of Quebec. Its database allows

ascertainment of all health services covered by the plan and

provided in the province of Quebec. These services include all

treatment modalities for PCa, including attempted curative

EBRT, RP, bilateral orchiectomy, as well as hormonal therapy.

Moreover, the health plan relies on International Classification of

Diseases, 9th Revision (ICD-9) codes and their respective dates,

which allow defining of baseline Charlson Comorbidity Index

(CCI) scores [14,15]. Finally, the health plan provides crude

survival data for all enrollees.

The health plan database allowed us to identify men aged

70 yr or older treated with either RP or attempted curative

EBRT, between January 1, 1989, and December 31, 2000, who

were diagnosed with PCa (ICD9-185-9). Each record included

treated with either radical prostatectomy or external beam

Number of patients

al prostatectomy External beam radiation therapy

1591 (25.7%) 4592 (74.3%)

72 (71) 75 (74)

70–89 70–95

1414 (88.9%) 2548 (55.5%)

144 (9.1%) 1467 (31.9%)

27 (1.7%) 430 (9.4%)

6 (0.4%) 147 (3.2%)

1.3 (1.0) 2.1 (2.0)

0–8 0–12

628 (39.5%) 1122 (24.4%)

673 (42.3%) 1842 (40.1%)

227 (14.3%) 1081 (23.5%)

58 (3.6%) 421 (9.2%)

5 (0.3%) 126 (2.7%)

7.3 (7.1) 4.4 (4.1)

0.1–15.5 0.1–15.5

560 (35.2%) 2707 (59.0%)

12.1 (10.7) 5.7 (6.5)

e u r o p e a n u r o l o g y 5 4 ( 2 0 0 8 ) 1 0 7 – 1 1 7 109

the type of treatment, the date of RP or of the first EBRT

treatment, age at treatment, vital status up to July 31, 2004, and

the pretreatment CCI score. The health plan records contain

no information on tumor stage or grade, or on prostate-

specific antigen levels either before or after attempted curative

therapy. The analyses targeted 6183 (35.2%) evaluable patients

of 17,570 men of all ages, who were treated with EBRT

(n = 7892) or RP (n = 9678) during the study period.

Cause-specific mortality is not provided in the database.

Therefore, we controlled for the potential effect of PCa-specific

mortality by repeating all analyses within a subset of

individuals, who did not receive any secondary therapy.

These patients were considered to have no clinical evidence of

PCa disease relapse because virtually all clinically significant

Fig. 1 – (a) Overall survival in 6183 patients aged 70 yr or older

radiation therapy. (b) Overall survival (stratified according to tre

with either radical prostatectomy (n = 1591) or external beam ra

PCa relapses will be treated with some form of secondary

therapy, at least in the form of androgen deprivation [16]. This

approach is particularly true in the free and universal health

care access that is available to all Quebec residents. Therefore,

it can be safely assumed that virtually all men who died

without receiving any secondary therapy had not succumbed

to PCa.

Secondary therapy was defined as EBRT after RP. Alter-

natively any type of hormonal therapy after either RP or

attempted curative EBRT was also considered secondary

therapy. Hormonal therapy consisted of either medical

castration, including steroidal or nonsteroidal antiandrogens,

or of bilateral orchiectomy. No cases of salvage RP were

recorded in the study population.

treated with either radical prostatectomy or external beam

atment type) in 6183 patients aged 70 yr or older treated

diation therapy (n = 4592).

Fig. 2 – Kaplan-Meier analyses addressing time to secondary therapy in patients treated with either radical prostatectomy

(n = 1591) or external beam radiation therapy (n = 4592).

Table 2 – Univariable and multivariable analysesaddressing overall mortality in 6183 men 70 yr or oldertreated with either radical prostatectomy or externalbeam radiation therapy

Predictors Univariable Multivariable

RR; p value RR; p value

Age at treatment (yr) —; <0.001 —; <0.001

75–79 vs. 70–74 1.8; <0.001 1.5; <0.001

80–84 vs. 70–74 3.3; <0.001 2.5; <0.001

�85 vs. 70–74 5.1; <0.001 3.6; <0.001

Charlson Comorbidity Index —; <0.001 —; <0.001

1–2 vs. 0 1.4; <0.001 1.3; <0.001

3–4 vs. 0 2.2; <0.001 1.8; <0.001

5–6 vs. 0 3.0; <0.001 2.3; <0.001

>6 vs. 0 4.7; <0.001 3.3; <0.001

Secondary therapy (yes vs. no) 0.7; <0.001 0.7; <0.001

Treatment type (EBRT vs. RP) 2.7; <0.001 2.1; <0.001

RR, rate ratio; EPRT, external beam radiation therapy; RP, radical

prostatectomy.

The dashes imply that no rate ratio is associated with the overall

effect of age or with the effect of Charlson Comorbidity Index.

Instead, the subcategories of both variables have specific rate

ratios.

e u r o p e a n u r o l o g y 5 4 ( 2 0 0 8 ) 1 0 7 – 1 1 7110

2.2. Statistical analyses

The first set of analyses addressed the overall survival in the

entire cohort. Subsequently, all analyses were restricted to

patients whose secondary therapy consisted exclusively of

androgen deprivation. Finally, all analyses were repeated in

patients unexposed to secondary therapy. In all models,

therapy type (RP vs. EBRT) represented the predictor of

interest. Covariates included age at treatment, comorbidities,

and presence or absence of secondary therapy delivery. The

statistical analyses relied on chi-square, independent sample t

test, Kaplan-Meier, life table, and Cox regression analyses. All

statistical tests were performed with the use of S-PLUS

Professional, version 1 (MathSoft Inc, Seattle, WA, USA). All

tests were two-sided with a significance level set at 0.05.

3. Results

Between 1989 and 2000, 6183 patients 70 yr of age orolder were diagnosed with PCa and treated witheither RP (1591, 25.7%) or EBRT (4592, 74.3%). Table 1shows the descriptive statistics of this cohort. Themedian actuarial survival for the entire cohort was7.4 yr (mean, 7.7). After RP the median actuarialsurvival was 12.1 yr (mean, 10.7) vs. 5.7 yr (mean, 6.5)after EBRT (log-rank, p < 0.001).

Fig. 1 a shows the overall survival plot for thecohort of 6183 men. According to life table analyses,the 10-yr survival probability was 38.5%. At 10 yrafter RP, overall survival was 59.3 versus 30.3% afterEBRT (Fig. 1 b; log-rank, p < 0.001). Fig. 2 shows theactuarial time to secondary therapy according to

treatment type (RP vs. EBRT) and indicates that at 10yr 83.6% of RP patients were free of secondarytherapy versus 69.9% for EBRT patients (log-rank,p < 0.001).

Table 2 shows the univariable and multivariableCox regression models addressing survival in theoverall population. In multivariable analyses, the

Table 3 – Univariable and multivariable analysesaddressing overall mortality in 5997 men 70 yr or oldertreated with either radical prostatectomy or externalbeam radiation therapy whose secondary therapyconsisted of only hormonal ablation

Predictors Univariable Multivariable

RR; p value RR; p value

Age at treatment (yr) —; <0.001 —; <0.001

75–79 vs. 70–74 1.8; <0.001 1.4; <0.001

80–84 vs. 70–74 3.2; <0.001 2.5; <0.001

�85 vs. 70–74 5.0; <0.001 3.6; <0.001

Charlson Comorbidity Index —; <0.001 —; <0.001

1–2 vs. 0 1.4; <0.001 1.3; <0.001

3–4 vs. 0 2.2; <0.001 1.8; <0.001

5–6 vs. 0 3.0; <0.001 2.3; <0.001

>6 vs. 0 4.7; <0.001 3.2; <0.001

Treatment type (EBRT vs. RP) 2.8; <0.001 2.2; <0.001

Secondary therapy (yes vs. no) 0.8; <0.001 0.7; <0.001

RR, rate ratio; EPRT, external beam radiation therapy; RP, radical

prostatectomy.

The dashes imply that no rate ratio is associated with the overall

effect of age or with the effect of Charlson Comorbidity Index.

Instead, the subcategories of both variables have specific rate

ratios.

Fig. 3 – (a–c) Kaplan-Meier analyses in the cohort of patients

whose secondary therapy consisted of only hormonal

ablation or no therapy at all (n = 5997) according to delivery

or no delivery of secondary therapy.

e u r o p e a n u r o l o g y 5 4 ( 2 0 0 8 ) 1 0 7 – 1 1 7 111

risk of overall mortality was 2.1-fold higher for thosetreated with EBRT versus RP ( p < 0.001). Advancedage at treatment, higher CCI score, and delivery ofsecondary therapy represented significant andindependent predictors of overall mortality in,respectively, univariable and multivariable analyses(all p values < 0.001, Table 2).

The survival of patients whose secondary therapyconsisted exclusively of androgen-ablation therapyor of no therapy at all (n = 5997) is shown in Fig. 3 aand is subsequently stratified (RP vs. EBRT) accord-ing to primary treatment type (Fig. 3 b–c). In theoverall cohort, as well as in the EBRT cohort,androgen-ablation therapy exerted a protectiveeffect (both p < 0.001) on survival. Conversely, nosurvival difference was recorded in the RP cohort( p = 0.9). The overall survival benefit of hormonaltherapy (rate ratio, 0.7, p < 0.001) is also shown in themultivariable analysis (Table 3).

Of the overall cohort of 6183 patients, 2704(43.7%) had no clinical evidence of PCa relapse(Table 4) during the study follow-up. Of these,881 (32.6%) underwent RP and 1823 (67.4%) receivedEBRT. Their follow-up ranged from 0.1 to 15.5 yr(mean, 4.9; median, 4.4). The overall medianactuarial survival was 7.2 yr (mean, 6.6). The medianactuarial survival times after RP and EBRT were,respectively, 12.5 yr (mean, 11.0) and 3.4 yr (mean,5.4) (log-rank, p < 0.001). RP patients were younger(t test, p < 0.001) and had fewer comorbidities(chi-square, p < 0.001). Fig. 4 a shows the overall

survival within this cohort of 2704 men. Accordingto life table analyses, 10-yr survival probability was36.5% (RP 63.8% vs. EBRT 22.6%; p < 0.001, Fig. 4 b).

Table 5 shows univariable and multivariable Coxregression models addressing overall survival in therestricted cohort of 2704 men, who had no clinicalevidence of PCa relapse. In multivariable analyses

Table 4 – Descriptive table for a subset of 2704 patients treated with either radical prostatectomy or external beamradiation therapy, who were not exposed to any secondary therapy

Variables Number of patients

All patients Radical prostatectomy External beam radiation therapy

Total 2704 (100%) 881 (32.6%) 1823 (67.4%)

Age at treatment (yr)

Mean (median) 74.2 (73) 72.1 (71) 75.1 (74)

Range 70–93 70–89 70–93

70–74 1720 (63.6%) 777 (88.2%) 943 (51.7%)

75–79 689 (25.5%) 78 (8.9%) 611 (33.5%)

80–84 225 (8.3%) 20 (2.3%) 205 (11.2%)

�85 70 (2.6%) 6 (0.7%) 64 (3.5%)

Charlson Comorbidity Index

Mean (median) 2.2 (2) 1.4 (1) 2.6 (2)

Range 0–12 0–8 0–12

0 658 (24.3%) 337 (38.3%) 321 (17.6%)

1–2 1051 (38.9%) 371 (42.1%) 680 (37.3%)

3–4 642 (23.7%) 135 (15.3%) 507 (27.8%)

5–6 269 (9.9%) 33 (3.7%) 236 (12.9%)

>6 84 (3.1%) 5 (0.6%) 79 (4.3%)

Follow-up time (yr)

Mean (median) 4.9 (4.4) 7.2 (6.9) 3.8 (3.4)

Range 0.1–15.5 0.1–14.3 0.1–15.5

Deaths during follow-up 1540 (57.0%) 276 (31.3%) 1263 (69.3%)

Actuarial survival (yr)

Median (mean) 6.6 (7.2) 12.5 (11.0) 3.4 (5.3)

e u r o p e a n u r o l o g y 5 4 ( 2 0 0 8 ) 1 0 7 – 1 1 7112

the rate of mortality was 2.9-fold higher for thosetreated with EBRT versus RP ( p < 0.001). Advancedage and higher CCI score also represented signifi-cant and independent predictors of overall mortalityin, respectively, univariable and multivariable ana-lyses (both p < 0.001).

Finally, Fig. 5 a–b stratifies survival according toage categories (70–74 vs. 75–79 vs. 80–84) andtreatment modality, after the restriction to indivi-duals unexposed to secondary therapy (n = 2704),with the intent of assessing the survival rates at10 yr. Interestingly, 62.5%, 44.0%, and 19.1% of RPpatients aged 70–74, 75–79, and 80 or older survivedbeyond 10 yr. The corresponding proportions ofEBRT patients were 37.9%, 24.8%, and 9.4%.

4. Discussion

In Canada, the remaining LE of a 70-, 75- and80-year-old male is, respectively, 13, 10, and 7 yr [17].However, in many septa- and octogenarians, comor-bidities may undermine these LE estimates. ShorterLE may in turn question the validity of attemptedcurative therapy for localised PCa [18]. Currently, a10 yr or longer LE is considered the minimumrequirement for attempted curative therapy[1,19,20]. Patients may be overtreated if they die ofnon-PCa causes prior to the 10-yr threshold. We

used this premise in our analyses and consideredpatients who did not survive beyond the 10-yrthreshold as overtreated. Cause-specific mortalitywas not available in the health plan database;therefore, we repeated all analyses in a subset ofpatients who had no clinical evidence of PCa relapsebecause they were not exposed to any form ofsecondary therapy. In the province of Quebec accessto health care is free and universal. Therefore, deathfrom PCa after failure of attempted curative therapywithout being at least exposed to androgen-depri-vation therapy prior to death it is virtually incon-ceivable [16,21]. Consequently, it can be assumedthat the restricted analysis effectively controlled forany possible PCa-related deaths.

The first analysis addressed overall survival in theentire cohort of 6183 men aged 70 yr or older treatedwith attempted curative RP (n = 1591) or EBRT(n = 4592). The second analysis excluded patientswhose secondary therapy did not consist of andro-gen deprivation, which resulted in 5997 men (RP,n = 1410, EBRT, n = 4587). This analysis addressedoverall survival according to either delivery or nodelivery of androgen deprivation as secondarytherapy. The final analysis was restricted to 2704men (RP, n = 881, EBRT, n = 1823) who had no clinicalevidence of PCa relapse after either RP or EBRT. Theend point consisted of overall survival. In allanalyses, patients treated with EBRT were older

Fig. 4 – (a) Overall survival in a subset of 2704 patients treated with either radical prostatectomy or external beam radiation

therapy, who were not exposed to any secondary therapy. (b) Overall survival (stratified according to treatment type) in a

subset of 2704 patients treated with either radical prostatectomy (n = 881) or external beam radiation therapy (n = 1823),

who were not exposed to any secondary therapy.

e u r o p e a n u r o l o g y 5 4 ( 2 0 0 8 ) 1 0 7 – 1 1 7 113

and had more comorbidities than their counterpartstreated with RP. In all analyses, EBRT was associatedwith lower survival rates. In the first analysis, whichincluded all patients regardless of secondary ther-apy status, only 30% of men treated with EBRTsurvived the 10-yr mark versus 59% of men treatedwith RP. When the population was restricted to 2704men with no clinical evidence of PCa relapse (thirdanalysis), only 23% of men treated with EBRTsurvived beyond the 10-yr mark versus 64% ofmen treated with RP. When data were furtherstratified according to age, between 19% and 62%

of RP patients survived the 10-yr mark versus 9–38%of EBRT patients. Taken together, these findingsindicate that an important proportion of septa- andoctogenarians will be overtreated. The rate ofovertreatment is particularly important in EBRTpatients. Nonetheless, scrutiny should be applied toRP and EBRT candidates older than 70 yr.

Our survival rate in the EBRT cohort of menaged 70 and above (n = 1823) who were unexposed tosecondary therapy (22.6% at 10 yr), is in agreementwith the observations of Albertsen et al [6], whoshowed a 30% survival rate in patients with

Fig. 5 – (a) Kaplan-Meier and life table analyses addressing

10-yr survival in radical prostatectomy and external beam

radiation therapy according to three different age

categories (70–74 vs. 75–79 vs. 80–84 years) in the whole

cohort. (b) Kaplan-Meier and life table analyses addressing

10-yr survival in radical prostatectomy and external beam

radiation therapy patients according to three different age

categories (70–74 vs. 75–79 vs. 80–84 yr) in patients

unexposed to secondary therapy.

Table 5 – Univariable and multivariable analysesaddressing overall mortality in a subset of 2704 men70 yr or older treated with either radical prostatectomy orexternal beam radiation therapy, who were not exposedto any secondary therapy.

Predictors Univariable Multivariable

RR; p value RR; p value

Age at treatment (yr) —; <0.001 —; <0.001

75–79 vs. 70–74 2.1; <0.001 1.5; <0.001

80–84 vs. 70–74 3.0; <0.001 1.8; <0.001

�85 vs. 70–74 4.5; <0.001 3.5; <0.001

Charlson Comorbidity Index —; <0.001 —; <0.001

1–2 vs. 0 1.5; <0.001 1.3; <0.001

3–4 vs. 0 2.6; <0.001 2.0; <0.001

5–6 vs. 0 4.0; <0.001 2.7; <0.001

>6 vs. 0 5.4; <0.001 3.5; <0.001

Treatment type (EBRT vs. RP) 3.9; <0.001 2.9; <0.001

RR, rate ratio; EPRT, external beam radiation therapy; RP, radical

prostatectomy.

The dashes imply that no rate ratio is associated with the overall

effect of age or with the effect of Charlson Comorbidity Index.

Instead, the subcategories of both variables have specific rate

ratios.

e u r o p e a n u r o l o g y 5 4 ( 2 0 0 8 ) 1 0 7 – 1 1 7114

moderately differentiated PCa treated with delayedintervention. The novelty of our findings consists ofthe fact that only a minority of elderly EBRT patientsqualify for attempted curative EBRT. For example,only 25% of men aged 75–79 treated with EBRT andunexposed to secondary therapy survived the 10-yr

mark. Similarly, a non-negligible proportion of RPpatients are overtreated. For example, only 44% ofmen aged 75–79 treated with RP and unexposed tosecondary therapy survived the 10-yr mark. There-fore, scrutiny should be used for septagenarianswho are being considered for attempted curativetherapy. For octogenarians such treatment choiceshould be considered in only truly exceptional cases.A recently validated nomogram can be used toidentify patients who could be overtreated owing toinsufficient life expectancy [22].

It is also of interest to note that androgen-ablationtherapy appears to improve overall survival in EBRTpatients, but has no effect in RP-treated men (Fig. 3).This finding is contrary to the recently reporteddetrimental survival effect of androgen-deprivationtherapy [23]. Several explanations may be proposed.One may stipulate that the survival benefit ofhormonal therapy in EBRT patients may relate tolesser treatment efficacy relative to RP. However,unavailability of EBRT dosages and the absence oftumor characteristics prevent us from makingdefinitive conclusions.

Interestingly, our findings demonstrated thatsurvival of patients treated with RP and EBRT differs,even after adjustment for age and comorbidities.Men treated with EBRT had a 2.1- to 2.9-fold higherrisk of mortality than their RP counterparts. Fowleret al [24], in a substantially smaller nonrandomisedseries of RP (n = 138) and EBRT (n = 138) patients,found a similar observation, namely a 3.8-fold

e u r o p e a n u r o l o g y 5 4 ( 2 0 0 8 ) 1 0 7 – 1 1 7 115

higher age-adjusted risk of mortality ( p = 0.02) inEBRT versus RP patients. Taken together, our and theFowler et al data suggest that neither age norcomorbidity represent sensitive enough metrics tocompletely and accurately account for the survivaldiscrepancy between RP and EBRT patients. There-fore, additional characteristics, such as for examplePCa-specific comorbidity indices, should be consid-ered to better predict other-cause mortality inelderly patients subjected to attempted curativePCa treatment modalities. Such a profile could bemore specific and more accurate than the CCI. Thesimilarity between our findings and those of Fowleret al also validates our observations, relative to menfrom the United States [17].

The present study is not devoid of limitations.Lack of cancer-specific mortality represents one ofthe main limitations, which is a limitation of otherstudies [25]. We circumvented this limitation byrestricting our analyses to patients unexposed tosecondary therapies, in whom we demonstratedvirtually the same trends as in the entire cohort.This adjustment might overestimate the effect ofPCa-specific mortality. Only a small fraction ofpatients exposed to secondary therapies, such asandrogen deprivation, will succumb to PCa. Forexample, in a series of 752 RP patients with a medianfollow-up of 11 yr, only 6.6% died of PCa, and PCa-specific mortality was even lower (2%) in anotherlarge series with long-term follow-up [26,27]. Lack ofclinical and pathological characteristics of treatedPCa represents another potential weakness of ourstudy. Clinical data would have allowed furtheradjustment in multivariable models. However, giventhe very low rate of PCa-specific mortality, it isunlikely that clinical PCa characteristics would havestrongly affected the observed survival rates. More-over, treatment selection is not based on only LE andcancer characteristics. Quality-of-life considera-tions, patient and physician preferences, and treat-ment availability all add to the complexity oftreatment selection. The administrative nature ofthe data did not allow consideration of thesevariables in our analyses. Finally, we do not haveconfirmatory data stating that all men with evi-dence of meaningful relapse received secondarytherapy. However, the free and universal access tohealth care in the Province of Quebec allows all mento receive androgen-ablation therapy. Moreover, in asurvey performed during our study period, 90% ofCanadian urologists declared they used not onlyandrogen-ablation therapy but combined androgen-ablation therapy [16]. Therefore, it is unlikely thatany patient with clinically meaningful relapsewould not be offered or would not receive secondary

therapy in at least the form of androgen-ablationtherapy. Despite these limitations, our series pro-vides a valuable critical look at the use of attemptedcurative treatment modalities in elderly patientswith PCa.

5. Conclusions

Forty percent of septa- and octogenarian men whoare selected for RP do not have adequate LE towarrant attempted curative therapy. Even morestrikingly, 70% of men who receive EBRT die beforereaching the 10-yr mark. These findings indicate theneed for more stringent EBRT and RP selectioncriteria if the goal is to minimise overtreatment.

Conflicts of interest

The authors have nothing to disclose.

Acknowledgements

Pierre I. Karakiewicz is partially supported by theUniversity of Montreal Health Center Urology Associ-ates, Fonds de la Recherche en Sante du Quebec, theUniversity of Montreal Department of Surgery, andthe University of Montreal Health Center (CHUM)Foundation. Jochen Walz is partially supported by theVereinigung Norddeutscher Urologen.

References

[1] Aus G, Abbou CC, Bolla M, et al. EAU guidelines on prostate

cancer. Eur Urol 2005;48:546–51.

[2] Bubolz T, Wasson JH, Lu-Yao G, Barry MJ. Treatments for

prostate cancer in older men: 1984–1997. Urology 2001;58:

977–82.

[3] Roemeling S, Roobol MJ, de Vries SH, et al. Active surveil-

lance for prostate cancers detected in three subsequent

rounds of a screening trial: characteristics, PSA doubling

times, and outcome. Eur Urol 2007;51:1244–51.

[4] Aus G, Robinson D, Rosell J, Sandblom G, Varenhorst E.

Survival in prostate carcinoma–outcomes from a prospec-

tive, population-based cohort of 8887 men with up to

15 years of follow-up: results from three countries in

the population-based National Prostate Cancer Registry

of Sweden. Cancer 2005;103:943–51.

[5] Johansson JE, Andren O, Andersson SO, et al. Natural

history of early, localized prostate cancer. JAMA 2004;

291:2713–9.

[6] Albertsen PC, Hanley JA, Fine J. 20-year outcomes follow-

ing conservative management of clinically localized pros-

tate cancer. JAMA 2005;293:2095–101.

e u r o p e a n u r o l o g y 5 4 ( 2 0 0 8 ) 1 0 7 – 1 1 7116

[7] Chun FK, Briganti A, Lebeau T, et al. The 2002 AJCC pT2

substages confer no prognostic information on the rate of

biochemical recurrence after radical prostatectomy. Eur

Urol 2006;49:273–9.

[8] Chun FK, Steuber T, Erbersdobler A, et al. Development

and internal validation of a nomogram predicting the

probability of prostate cancer Gleason sum upgrading

between biopsy and radical prostatectomy pathology.

Eur Urol 2006;49:820–6.

[9] Simmons MN, Stephenson AJ, Klein EA. Natural history of

biochemical recurrence after radical prostatectomy: risk

assessment for secondary therapy. Eur Urol 2007;51:

1175–84.

[10] Chun FK, Briganti A, Shariat SF, et al. Significant upgrad-

ing affects a third of men diagnosed with prostate cancer:

predictive nomogram and internal validation. BJU Int

2006;98:329–34.

[11] Freedland SJ, Mangold LA, Walsh PC, Partin AW. The

prostatic specific antigen era is alive and well: prostatic

specific antigen and biochemical progression following

radical prostatectomy. J Urol 2005;174:1276–81.

[12] Shipley WU, Thames HD, Sandler HM, et al. Radiation

therapy for clinically localized prostate cancer: a

multi-institutional pooled analysis. JAMA 1999;281:

1598–604.

[13] Miller DC, Gruber SB, Hollenbeck BK, Montie JE, Wei JT.

Incidence of initial local therapy among men with lower-

risk prostate cancer in the United States. J Natl Cancer Inst

2006;98:1134–41.

[14] Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new

method of classifying prognostic comorbidity in longitu-

dinal studies: development and validation. J Chronic Dis

1987;40:373–83.

[15] D’Hoore W, Sicotte C, Tilquin C. Risk adjustment in out-

come assessment: the Charlson comorbidity index. Meth-

ods Inf Med 1993;32:382–7.

[16] Ramsey EW, Elhilali M, Goldenberg SL, et al. Practice

patterns of Canadian urologists in benign prostatic hyper-

plasia and prostate cancer. Canadian Prostate Health

Council. J Urol 2000;163:499–502.

Editorial Comment on: Poor Overall Survival inSepta- and Octogenarian Patients after RadicalProstatectomy and Radiotherapy for ProstateCancer: A Population-Based Study of 6183 MenMichael J. BarryMedical Practices Evaluation Center,Massachusetts General Hospital,Boston, MA, USAmbarry@partners.org

Evidence is accumulating that from a populationperspective, early detection and attempted curativetreatment of prostate cancers found by screening

[17] Statistics Canada J. Complete life table, Quebec, 2000–

2002: MALES wwwstatcanca 2000 to 2002;Catalogue No.

84-537-XIE:27–29.

[18] Hall WH, Jani AB, Ryu JK, Narayan S, Vijayakumar S. The

impact of age and comorbidity on survival outcomes and

treatment patterns in prostate cancer. Prostate Cancer

Prostatic Dis 2005;8:22–30.

[19] Middleton RG. The management of clinically localized

prostate cancer: guidelines from the American Urological

Association. CA Cancer J Clin 1996;46:249–53.

[20] Scardino P. Update: NCCN prostate cancer Clinical

Practice Guidelines. J Natl Compr Canc Netw 2005;

3(Suppl 1):S29–33.

[21] Porter CR, Gallina A, Kodama K, et al. Prostate cancer-

specific survival in men treated with hormonal therapy

after failure of radical prostatectomy. Eur Urol 2007;52:

446–54.

[22] Walz J, Gallina A, Saad F, et al. A nomogram predicting 10-

year life expectancy in candidates for radical prostatec-

tomy or radiotherapy for prostate cancer. J Clin Oncol

2007;25:3576–81.

[23] Loblaw DA, Virgo KS, Nam R, et al. Initial hormonal

management of androgen-sensitive metastatic, recur-

rent, or progressive prostate cancer: 2006 update of an

American Society of Clinical Oncology practice guideline.

J Clin Oncol 2007;25:1596–605.

[24] Fowler Jr JE, Terrell FL, Renfroe DL. Co-morbidities and

survival of men with localized prostate cancer treated

with surgery or radiation therapy. J Urol 1996;156:1714–8.

[25] Thompson Jr IM, Tangen CM, Paradelo J, et al. Adjuvant

radiotherapy for pathologically advanced prostate cancer:

a randomized clinical trial. JAMA 2006;296:2329–35.

[26] Porter CR, Kodama K, Gibbons RP, et al. 25-year prostate

cancer control and survival outcomes: a 40-year radical

prostatectomy single institution series. J Urol 2006;176:

569–74.

[27] Roehl KA, Han M, Ramos CG, Antenor JA, Catalona WJ.

Cancer progression and survival rates following anato-

mical radical retropubic prostatectomy in 3,478 consecu-

tive patients: long-term results. J Urol 2004;172:910–4.

may do more harm than good among older men.Arguably the most important piece of evidence is asubgroup analysis of the Scandinavian ProstateCancer Group 4 trial of radical prostatectomy versuswatchful waiting, which showed no benefit amongmen over age 65, even when diagnosed withoutscreening [1]. Among older men diagnosed byscreening, over-diagnosis and lead time would meanthe number who would need to be treated to preventa prostate cancer death would be very high indeed.Jeldres and colleagues document that a substantialminority of prostate cancer patients aged 70 or oldertreated with radical prostatectomy and the majority

treated with radiotherapy in Quebec were notdestined to live 10 yr due to competing hazards ofmortality [2]. These men were unlikely to havebenefited from treatment and undoubtedly suffereda considerable toll of side effects, including incon-tinence and erectile dysfunction [3].

However, translating these insights into the careof individual patients is not easy. In all likelihood,an occasional man with a life expectancy <10 yrwill eventually die of something other thanprostate cancer if he is screened and treated, eitherbecause he harbored an aggressive cancer that wasstill curable or because he ‘‘beat the odds’’ andlived much longer. But how much morbidity, interms of dealing with a cancer diagnosis and theside effects of unnecessary treatment, is it worthcausing among large numbers of men for thisoccasional victory? Although better informeddecisions about therapy, as suggested by Jeldresand colleagues, are desirable, I believe betterinformed prostate-specific antigen (PSA) screeningdecisions are even more important. In the UnitedStates, where PSA testing is as widespread as ‘‘fast

food,’’ more informed patients choose PSA screen-ing less often [4].

References

[1] 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.

[2] Jeldres C, Suardi N, Walz J, et al. Poor overall survival in

septa- and octogenarian patients after radical prosta-

tectomy and radiotherapy for prostate cancer: a popula-

tion-based study of 6183 men. Eur Urol 2008;54:107–17.

[3] Potosky AL, David WW, Hoffman RM, et al. Five year

outcomes after prostatectomy or radiation therapy for

prostate cancer: the Prostate Cancer Outcomes study.

J Natl Cancer Inst 2004;96:1358–67.

[4] Volk RJ, Hawley ST, Kneuper S, et al. Trials of decision

aids for prostate screening: a systematic review. Am J

Prev Med 2007;33:428–34.

DOI: 10.1016/j.eururo.2007.10.039

DOI of original article: 10.1016/j.eururo.2007.10.038

e u r o p e a n u r o l o g y 5 4 ( 2 0 0 8 ) 1 0 7 – 1 1 7 117