Int. J. Radiation Oncology Biol. Phys., Vol. 76, No. 3, pp. 735–740, 2010Copyright � 2010 Elsevier Inc.

Printed in the USA. All rights reserved0360-3016/10/$–see front matter

jrobp.2009.02.049

doi:10.1016/j.i

CLINICAL INVESTIGATION Prostate

SALVAGE RADIOTHERAPY FOR RISING PROSTATE-SPECIFIC ANTIGEN LEVELSAFTER RADICAL PROSTATECTOMY FOR PROSTATE CANCER: DOSE–RESPONSE

ANALYSIS

JOHNNY RAY BERNARD, JR., M.D.,* STEVEN J. BUSKIRK, M.D.,* MICHAEL G. HECKMAN, M.S.,y

NANCY N. DIEHL, B.S.,y STEPHEN J. KO, M.D.,* ORLAN K. MACDONALD, M.D.,z

STEVEN E. SCHILD, M.D.,x AND THOMAS M. PISANSKY, M.D.z

*Department of Radiation Oncology and yBiostatistics Unit, Mayo Clinic, Jacksonville, FL; Department of Radiation Oncology,zMayo Clinic, Rochester, MN; and xMayo Clinic, Scottsdale, AZ

Reprintion OncoFL 32224

Purpose: To investigate the association between external beam radiotherapy (EBRT) dose and biochemical failure(BcF) of prostate cancer in patients who received salvage prostate bed EBRT for a rising prostate-specific antigen(PSA) level after radical prostatectomy.Methods and Materials: We evaluated patients with a rising PSA level after prostatectomy who received salvageEBRT between July 1987 and October 2007. Patients receiving pre-EBRT androgen suppression were excluded.Cox proportional hazards models were used to investigate the association between EBRT dose and BcF. Dosewas considered as a numeric variable and as a categoric variable (low, <64.8 Gy; moderate, 64.8–66.6 Gy; high,>66.6 Gy).Results: A total of 364 men met study selection criteria and were followed up for a median of 6.0 years (range, 0.1–19.3 years). Median pre-EBRT PSA level was 0.6 ng/mL. The estimated cumulative rate of BcF at 5 years afterEBRT was 50% overall and 57%, 46%, and 39% for the low-, moderate-, and high-dose groups, respectively.In multivariable analysis adjusting for potentially confounding variables, there was evidence of a linear trend be-tween dose and BcF, with risk of BcF decreasing as dose increased (relative risk [RR], 0.77 [5.0-Gy increase];p = 0.05). Compared with the low-dose group, there was evidence of a decreased risk of BcF for the high-dose group(RR, 0.60; p = 0.04), but no difference for the moderate-dose group (RR, 0.85; p = 0.41).Conclusions: Our results suggest a dose response for salvage EBRT. Doses higher than 66.6 Gy result in decreasedrisk of BcF. � 2010 Elsevier Inc.

Biochemical failure, Dose response, Prostate bed, Prostate-specific antigen recurrence, Salvage.

INTRODUCTION

Randomized trials of definitive external beam radiotherapy

(EBRT) for localized prostate cancer have shown that

a higher radiation dose leads to improved prostate cancer out-

comes (1–5). Kuban et al. (1) updated the M. D. Anderson

experience of 301 men with T1b–T3 prostate cancer who

were randomly assigned to receive 70 or 78 Gy of EBRT.

Biochemical control was achieved in 76% of patients in the

70-Gy protocol vs. 83% of patients in the 78-Gy protocol,

and the 8-year actuarial estimates of freedom from failure

(biochemical or clinical) were 59% vs. 78%, respectively

(p = 0.04). Similarly, Zietman et al. (2) evaluated 393 men

with localized prostate cancer who were randomly assigned

to receive 70.2 or 79.2 Gy of EBRT. Biochemical control

at 5 years occurred in 79% of patients in the lower-dose pro-

tocol vs. 91% of patients in the higher-dose protocol; increas-

ing the radiation dose improved local control from 48% to

t requests: Steven J. Buskirk, M.D., Department of Radia-logy, Mayo Clinic, 4500 San Pablo Road, Jacksonville,. E-mail: buskirk.steven@mayo.edu

735

67% (p < 0.001). A Mayo Clinic retrospective review by

Vora et al. (6) reported 271 patients who received three-di-

mensional radiotherapy (3D-RT) with a median dose of

68.4 Gy (range, 66–71 Gy) and were compared with 145 pa-

tients who received intensity-modulated radiotherapy

(IMRT) with a median dose of 75.6 Gy (range, 70.2–77.4

Gy). The 5-year biochemical control rate was 74.4% with

lower-dose 3D-RT and 84.6% with higher-dose IMRT (p =

0.03).

The EBRT is also used after radical prostatectomy, either

adjuvantly or as salvage therapy for a rising prostate-specific

antigen (PSA) level, often referred to as biochemical failure

(BcF). Swanson et al. (7) concluded that a rising PSA level

postoperatively is most often attributable to the persistence

of cancer in the prostate fossa. This view is supported by

the observation that adjuvant EBRT reduces the risk of sub-

sequent metastatic disease relapse (7). A reduction in

Conflict of interest: none.Received Jan 21, 2009. Accepted for publication Feb 16, 2009.

736 I. J. Radiation Oncology d Biology d Physics Volume 76, Number 3, 2010

metastases and prostate cancer–related mortality was also

noted by Trock et al. (8) when salvage EBRT was used in pa-

tients with an elevated postoperative PSA level. These obser-

vations suggest that the persistence postoperatively of

prostate cancer in the prostate fossa results in a late wave

of metastases in some patients, as has also been suggested af-

ter definitive EBRT (9).

Salvage EBRT may be used postoperatively in patients

with a detectable PSA level, and several authors have re-

ported results with this treatment (8, 10, 11). Although

PSA levels decline after EBRT in a substantial proportion

of patients, PSA levels may rise subsequently in some pa-

tients. This pattern of PSA decline, often to an undetectable

level (11), with a delayed and steadily rising PSA profile sug-

gests the possibility that cancer persisted in the prostate fossa

despite salvage EBRT. It is a generally held view that lower

doses of EBRT should be administered in the salvage setting,

because the volume of cancer is presumed to be microscopic

and because patients are less tolerant of EBRT postopera-

tively. However, Sella et al. (12) found that the prostate fossa

mass averaged 1.4 cm (range, 0.8–4.5 cm) in diameter as

measured by endorectal magnetic resonance imaging, and

King and Kapp (13) suggested that the dose–response rela-

tionship of salvage and definitive EBRT are similar. To our

knowledge, no randomized trials of dose escalation for sal-

vage radiotherapy have been conducted.

Because the dose–response relationship of salvage EBRT

has not been studied sufficiently, we evaluated a cohort of

men treated at Mayo Clinic (14). The primary aim of this

study was to investigate whether an association exists be-

tween EBRT dose and BcF of prostate cancer. In an explor-

atory analysis, we also examined whether the relationship

between radiation dose and BcF is consistent for different

pre-EBRT PSA levels. It remains to be seen whether a ran-

domized trial will be conducted to study the question of

whether there will be a benefit to dose escalation in patients

with BcF postoperatively.

METHODS AND MATERIALS

This study was approved by the Mayo Clinic Institutional Review

Board. Patients were seen from July 1987 through October 2007 for

a detectable PSA level after radical prostatectomy. Information was

collected retrospectively from medical records. Data were obtained

from Mayo Clinic in Jacksonville, Florida (MCF), Mayo Clinic in

Rochester, Minnesota (MCR), and Mayo Clinic in Phoenix/Scotts-

dale, Arizona (MCA). Items of interest were preoperative PSA level,

pathologic tumor stage, Gleason score, DNA ploidy, surgical mar-

gin status, EBRT start date, patient age at EBRT start, pre-EBRT

PSA level, neoadjuvant and/or concurrent androgen suppression

(AS) associated with EBRT, EBRT prescribed dose, date of BcF,

date of last follow-up, and Mayo Clinic site. Patients who received

AS were excluded from this analysis.

The EBRT technique has been described previously (10). In sum-

mary, all patients were treated with EBRT using megavoltage (6- to

20-mV photons) linear accelerators at one of the three practice loca-

tions. EBRT treatment was planned and delivered to the estimated

region of the prostate bed. EBRT to the estimated region of the sem-

inal vesicle bed and remnants was done at the discretion of the

treating physician. Five patients also received treatment to the pelvic

lymphatic regions (45.0–50.4 Gy). The treatment volume was de-

fined by skeletal landmarks with or without surgical clip location

in the earlier years of treatment planning; by the mid 1990s, com-

puted tomography–based treatment planning was performed on all

patients. Retrograde urethrography was also performed in many pa-

tients. The median dose to the prostate bed was 64.8 Gy (range,

54.0–72.4 Gy).

A PSA value of 0.4 ng/mL and rising, as described by Amling

et al. (15), is currently used at our institution to define BcF after

surgery. We defined BcF after EBRT as a single PSA value of 0.4

ng/mL or higher, which had exceeded the post-EBRT nadir. The

date of the defining PSA value was considered the date of BcF with-

out backdating.

Numeric variables were summarized with the sample median,

minimum, 25th percentile, 75th percentile, and maximum. The

Kaplan-Meier method was used to estimate the cumulative rate of

BcF after the start of EBRT, censoring at the date of death or date

of last follow-up for patients who did not experience BcF. Cox pro-

portional hazards models were used to investigate the association

between EBRT dose and BcF; single variable Cox models were con-

sidered, as were multivariable models, adjusting for other possibly

confounding variables (preoperative PSA level, pathologic tumor

stage, Gleason score, margin status, patient age, and pre-EBRT

PSA level) simultaneously. An interaction between EBRT dose

and pre-EBRT PSA level was also considered in an analysis that

should be considered exploratory because of the low power to detect

such an interaction. Relative risks (RRs) and corresponding 95%

confidence intervals (CIs) were estimated. We considered EBRT

dose as a numeric variable to investigate a linear association with

BcF and also as a three-level categoric variable (<64.8, 64.8–66.6,

>66.6 Gy) to investigate a possible nonlinear association. Fisher’s

exact test or a Kruskal-Wallis rank sum test was used to compare pa-

tient characteristics according to dose category. Preoperative PSA

and pre-EBRT PSA levels were considered on the logarithm scale

in all Cox proportional hazards models because of their skewed dis-

tributions. Statistical analyses were performed using SPLUS (ver-

sion 8.0.1; Insightful Corporation, Seattle, Washington).

RESULTS

We identified 426 men who underwent salvage EBRT. A

final sample size of 364 patients (198 MCF, 106 MCR, 60

MCA) was analyzed after patients who received AS (n =

62) were excluded. Patients were placed in three groups

based on the EBRT dose: low dose (<64.8 Gy), moderate

dose (64.8–66.6 Gy), and high dose (>66.6 Gy). Patient char-

acteristics are shown in Table 1.

The median EBRT dose was 64.8 Gy (range, 54.0–72.4

Gy). The most common doses prescribed were 64.0 Gy

(n = 92; 25%); 66.6 Gy (n = 67; 18%); and 70.2 Gy (n =

57; 16%). There were 154 men (42%) in the low-dose group,

124 men (34%) in the moderate-dose group, and 86 men

(24%) in the high-dose group. Statistically significant differ-

ences in patient characteristics between radiation dose groups

were noted for age, pathologic tumor stage, preoperative PSA

level, pre-EBRT PSA level, and DNA ploidy (Table 1).

The median length of follow-up after the start of EBRT

was 6.0 years (range, 0.1–19.3 years), with 196 patients

(54%) experiencing BcF. Overall, the estimated cumulative

Table 1. Patient characteristics

Dose

VariableLow (<64.8 Gy)

(n = 154)Moderate (64.8–66.6 Gy)

(n = 124)High (>66.6 Gy)

(n = 86) p value

Age (y) 68 (44, 79) 68 (46, 81) 65 (50, 85) 0.005Pathologic tumor stage 0.03

T2 56 (37) 46 (38) 45 (53)T3a 52 (34) 52 (43) 26 (31)T3b 44 (29) 23 (19) 12 (14)T4 0 (0) 0 (0) 2 (2)

Margin 0.71Positive 82 (54) 66 (54) 50 (59)Negative 71 (46) 57 (46) 35 (41)

Gleason score 0.253–6 53 (43) 49 (42) 26 (31)7 52 (42) 44 (38) 41 (49)8–10 18 (15) 23 (20) 16 (19)

Preoperative PSA (ng/mL) 13.6 (1.1, 219.0) 7.4 (2.3-57.5) 7.5 (1.9, 47.0) <0.001Pre-EBRT PSA (ng/mL) 0.8 (0.1, 15.3) 0.5 (0.1, 14.4) 0.5 (0.1, 20.0) <0.001DNA ploidy <0.001

None 14 (11) 0 (0) 1 (2)Diploid 58 (46) 62 (67) 30 (63)Aneuploid 12 (9) 10 (11) 8 (17)Tetraploid 43 (34) 21 (23) 9 (19)

Abbreviations: EBRT = external beam radiotherapy; PSA = prostate-specific antigen.The sample medium (minimum, maximum) is given for numeric variables. The p values result from Fisher’s exact test or a Kruskal-Wallis

rank sum test. Information was unavailable for the following variables: pathologic tumor stage (n = 6), margin (n = 3), Gleason score (n = 42),preoperative PSA level (n = 32), and DNA ploidy (n = 96).

Rising PSA levels d J. R. BERNARD JR. et al. 737

rate of BcF at 1, 3, and 5 years after the start of EBRT was

20% (95% CI, 16–24%), 39% (95% CI, 33–44%), and

50% (95% CI, 44–55%), respectively (Fig. 1). At 5 years,

the estimated cumulative rates of BcF for the low-, moder-

ate-, and high-dose groups were 57%, 46%, and 39%, respec-

tively (Fig. 2, Table 2). In single variable analysis (Table 2),

there was strong evidence of a linear trend between EBRT

dose and BcF, with the risk of BcF decreasing as the

EBRT dose increased (RR, 0.70 [5.0-Gy increase]; 95%

CI, 0.56-0.87; p = 0.002), when dose was considered as a nu-

meric variable. In comparison with patients treated with a low

Fig. 1. Estimated cumulative rate of biochemical failure (BcF) afterthe start of external beam radiotherapy (EBRT). Dotted lines repre-sent 95% confidence intervals.

dose, there was a trend, although not statistically significant,

toward a decreased risk of BcF for patients treated with

a moderate dose (RR, 0.76; 95% CI, 0.56–1.04; p = 0.08),

whereas there was evidence of a decreased risk of BcF for pa-

tients treated with a high dose (RR, 0.60; 95% CI, 0.39–0.92;

p = 0.02).

In the multivariable analysis (Table 2), adjusting for preop-

erative PSA level, pathologic tumor stage, Gleason score,

margin status, patient age, and pre-EBRT PSA level, the ob-

served linear trend between increasing EBRT dose and de-

creased risk of BcF weakened slightly but still remained

(RR, 0.77 [5.0-Gy increase]; 95% CI, 0.59–1.01; p = 0.05),

whereas the observed trend toward a decreased risk of BcF

in patients treated with a high dose remained consistent

(RR, 0.60; 95% CI, 0.37–0.98; p = 0.04). The trend toward

a decreased risk of BcF in patients treated with a moderate

dose still did not reach statistical significance (RR, 0.85;

95% CI, 0.58–1.25; p = 0.41). We did not adjust for DNA

ploidy in the final multivariable analysis because information

concerning this variable was missing in many patients

(n = 96). However, in a sensitivity analysis adjusting for

DNA ploidy in addition to the aforementioned variables,

the associations between EBRT dose and BcF remained con-

sistent when considering dose as a numeric variable (RR,

0.71 [5.0-Gy increase]; p = 0.04) and as a categoric variable

(RR, 0.51 [high dose vs. low dose]; p = 0.04).

We also examined the association between EBRT dose

and BcF for varying levels of pre-EBRT PSA by stratifying

by the median pre-EBRT PSA value of 0.6 ng/mL (Fig. 3).

Fig. 2. Estimated cumulative rate of biochemical failure (BcF) afterthe start of external beam radiotherapy (EBRT) according to dose(low, <64.8 Gy; moderate, 64.8–66.6 Gy; high, >66.6 Gy).

Fig. 3. Estimated cumulative rate of biochemical failure (BcF) afterthe start of external beam radiotherapy (EBRT) according to dose(low, <64.8 Gy; moderate, 64.8–66.6 Gy; high, >66.6 Gy). (a)Patients with a pre-EBRT prostate-specific antigen (PSA) level of#0.6 ng/mL. (b) Patients with a pre-EBRT PSA level >0.6 ng/mL.

738 I. J. Radiation Oncology d Biology d Physics Volume 76, Number 3, 2010

There was no evidence of a difference in the relationship be-

tween dose and BcF for patients with a pre-EBRT PSA level

less than the median value (low) and a pre-EBRT PSA level

higher than the median value (high) (p = 0.68). However,

although statistical significance was not approached, we did

observe a stronger linear association between dose and BcF

in patients with high pre-EBRT PSA levels (RR, 0.69 [5.0-

Gy increase]; 95% CI, 0.47–0.99) compared with patients

who had low pre-EBRT PSA levels (RR, 0.85 [5.0-Gy in-

crease]; 95% CI, 0.57–1.28). Similarly, in comparison with

patients receiving a low dose, in patients receiving a high

dose, the decreased risk of BcF was more apparent in patients

with high pre-EBRT PSA levels (RR, 0.39; 95% CI, 0.18–

0.86) than in patients with a low pre-EBRT PSA level (RR,

0.84; 95% CI, 0.43–1.66). A larger sample size is needed

to properly examine the effect of dose on BcF for varying

pre-EBRT PSA levels.

DISCUSSION

Only a limited number of small retrospective studies have

shown a benefit of salvage EBRT doses higher than 64.8

Table 2. Association betwee

BcF

3-y Cumulative 5-y CumulativeEBRT Dose, Gy Rate, % (95% CI) Rate, % (95% CI)

Numeric variable5.0-Gy increase NA NA

Categoric variableLow: <64.8 47 (39–54) 57 (48–64)Moderate: 64.8–66.6 35 (25–43) 46 (36–55)High: >66.6 27 (16–37) 39 (25–51)

Abbreviations: BcF = biochemical failure; CI = confidence interval; EBtate-specific antigen; RR = relative risk.

Estimated RR and p values result from Cox proportional hazards modepathologic tumor stage, Gleason score, margin status, patient age, and pr

Gy (16–19); the largest of these studies included 122 patients

(19). In contrast, a pooled multi-institutional analysis of 1,540

patients (11), including those from MCR, did not identify

a dose–response relationship. However, Stephenson et al.(11) included patients treated with neoadjuvant and/or

n EBRT dose and BcF

Single variable model Multivariable model

RR (95% CI) p value RR (95% CI) p value

0.70 (0.56–0.87) .002 0.77 (0.59–1.01) .05

1.00 (reference) NA 1.00 (reference) NA0.76 (0.56–1.04) .08 0.85 (0.58–1.25) .410.60 (0.39–0.92) .02 0.60 (0.37–0.98) .04

RT = external beam radiotherapy; NA = not applicable; PSA = pros-

ls. Multivariable models were adjusted for preoperative PSA level,e-EBRT PSA level.

Rising PSA levels d J. R. BERNARD JR. et al. 739

concurrent short-term AS in their study, and they acknowl-

edged that the range of doses administered was relatively nar-

row. Nonetheless, the American Society for Therapeutic

Radiology and Oncology Consensus Panel has recommended

a dose ‘‘64 Gy or slightly higher’’ (20) for salvage EBRT, but

the group also stated that ‘‘the highest dose of radiation ther-

apy that can be given without morbidity is justifiable.’’

To supplement current data and to potentially clarify the

dose–response relationship, we evaluated a relatively large

cohort of men treated with salvage EBRT for a rising PSA

level after prostatectomy and also removed the potentially

confounding factor of AS. We evaluated EBRT dose as

(1) a numeric variable and

(2) a three-level category on the basis of the commonly used

doses of 64.8 and 66.6 Gy, allowing us to investigate lin-

ear and nonlinear associations with BcF.

In single variable analysis, we found strong evidence of

a general decreasing risk of BcF as the EBRT dose increased.

In evaluating low (<64.8 Gy), moderate (64.8–66.6 Gy), and

high (>66.6 Gy) EBRT doses, we observed a decreased risk of

BcF for patients receiving a high dose compared with those

receiving a low dose, but we did not observe a difference in

the risk of BcF between patients receiving low and moderate

doses. These results remained consistent after adjusting for

possible confounding factors in the multivariable analysis.

Anscher (21) suggested that 1 cm3 of prostate cancer vol-

ume would contain approximately 108 to 109 prostate cells.

This number of cells should produce a serum PSA level of ap-

proximately 3.5 ng/mL. Therefore, a PSA level of 0.35 ng/

mL after prostatectomy should represent the presence of

107 to 108 prostate cancer cells. Higher pre-EBRT PSA levels

would suggest an increased volume of disease. This supposi-

tion, coupled with the endorectal magnetic resonance imag-

ing findings of Sella et al. (12), supports the notion that the

dose–response relationship of salvage and definitive EBRT

are similar (13) and suggests that higher doses of radiation

would then be needed to eradicate all the cancer cells.

Although we did observe a stronger dose response in patients

with a high pre-EBRT PSA level (>0.6 ng/mL) compared

with other patients, this difference did not approach statistical

significance. However, power to detect such an interaction

was low in this study, and larger studies are needed to better

explore a possible increased dose response in patients with

a high pre-EBRT PSA level. These results, the overall results

of the present study, and the results of similar investigations

(16-19) provide evidence that the salvage EBRT dose recom-

mendation needs to be revisited.

Research evaluating radiation treatment planning and de-

livery to minimize toxicity with dose escalation is ongoing,

inasmuch as the tolerance of normal structures within or ad-

jacent to the prostate fossa may limit the EBRT dose that

may be given with minimal adverse effects. Feng et al.(22) observed a low rate of late genitourinary tract and gas-

trointestinal tract adverse events associated with adjuvant or

salvage EBRT (median dose, 64.8 Gy) in 959 men treated

between 1986 and 2004, a time when two-dimensional treat-

ment planning and delivery predominated. Cozzarini et al.(23) found that rectal dose–volume parameters affect the de-

velopment of hematochezia, and Fonteyne et al. (24) identi-

fied dose–volume constraints that may also provide useful

guidance. These observations suggest that inversely planned,

intensity-modulated EBRT, particularly with image-guided

localization (25), may allow the safe delivery of higher

EBRT doses. Indeed, this approach is already yielding prom-

ising results (26), and it should remain an avenue for active

research.

The results of this study are not definitive; the study’s ret-

rospective design may have introduced unforeseen biases.

However, this study has longer-term follow-up, does not

have the potentially confounding use of AS, and, to our

knowledge, represents the largest patient cohort to show

a benefit for higher doses of salvage EBRT. Considering

the results of our study, we conclude that it is appropriate

to consider doses higher than 66.6 Gy in the salvage setting

when normal-tissue dose constraints are met (23, 24). The re-

sults of this study should be validated by others; the most

meaningful future direction to assess these results would be

a randomized clinical trial (13).

REFERENCES

1. Kuban DA, Tucker SL, Dong L, et al. Long-term results of theM.D. Anderson randomized dose-escalation trial for prostatecancer. Int J Radiat Oncol Biol Phys 2008;70:67–74.

2. Zietman AL, DeSilvio ML, Slater JD, et al. Comparison of con-ventional-dose vs high-dose conformal radiation therapy in clin-ically localized adenocarcinoma of the prostate: A randomizedcontrolled trial. JAMA 2005;294:1233–1239. Erratum in:JAMA 2008;299:899–900.

3. Al-Mamgani A, van Putten WL, Heemsbergen WD, et al. Up-date of Dutch multicenter dose-escalation trial of radiotherapyfor localized prostate cancer. Int J Radiat Oncol Biol Phys2008;72:980–988.

4. Sathya JR, Davis IR, Julian JA, et al. Randomized trial comparingiridium implant plus external-beam radiation therapy with exter-nal-beam radiation therapy alone in node-negative locally ad-vanced cancer of the prostate. J Clin Oncol 2005;23:1192–1199.

5. Dearnaley DP, Sydes MR, Graham JD, et al. RT01 collabora-tors. Escalated-dose versus standard-dose conformal radiother-apy in prostate cancer: First results from the MRC RT01randomised controlled trial. Lancet Oncol 2007;8:475–487.

6. Vora SA, Wong WW, Schild SE, et al. Analysis of biochem-ical control and prognostic factors in patients treated with ei-ther low-dose three-dimensional conformal radiation therapyor high-dose intensity-modulated radiotherapy for localizedprostate cancer. Int J Radiat Oncol Biol Phys 2007;68:1053–1058.

7. Swanson GP, Hussey MA, Tangen CM, et al. SWOG 8794. Pre-dominant treatment failure in postprostatectomy patients is lo-cal: Analysis of patterns of treatment failure in SWOG 8794.J Clin Oncol 2007;25:2225–2229.

8. Trock BJ, Han M, Freedland SJ, et al. Prostate cancer-specificsurvival following salvage radiotherapy vs observation in men

740 I. J. Radiation Oncology d Biology d Physics Volume 76, Number 3, 2010

with biochemical recurrence after radical prostatectomy. JAMA2008;299:2760–2769.

9. Coen JJ, Zietman AL, Thakral H, et al. Radical radiation forlocalized prostate cancer: Local persistence of disease resultsin a late wave of metastases. J Clin Oncol 2002;20:3199–3205.

10. Buskirk SJ, Pisansky TM, Schild SE, et al. Salvage radiother-apy for isolated prostate specific antigen increase after radicalprostatectomy: Evaluation of prognostic factors and creationof a prognostic scoring system. J Urol 2006;176:985–990.

11. Stephenson AJ, Scardino PT, Kattan MW, et al. Predicting theoutcome of salvage radiation therapy for recurrent prostate can-cer after radical prostatectomy. J Clin Oncol 2007;25:2035–2041. Erratum in: J Clin Oncol 2007;25:4153.

12. Sella T, Schwartz LH, Swindle PW, et al. Suspected local recur-rence after radical prostatectomy: Endorectal coil MR imaging.Radiology 2004;231:379–385.

13. King CR, Kapp DS. Radiotherapy after prostatectomy: Is thereevidence for dose escalation out there? Int J Radiat Oncol BiolPhys 2008;71:346–350.

14. Buskirk SJ, Pisansky TM, Schild SE, et al. Salvage radiother-apy for rising prostate specific antigen after radical prostatec-tomy for prostate cancer: Dose response analysis [abstract].Int J Radiat Oncol Biol Phys 2006;66(Suppl 1):S322.

15. Amling CL, Bergstralh EJ, Blute ML, et al. Defining prostatespecific antigen progression after radical prostatectomy: Whatis the most appropriate cut point? J Urol 2001;165:1146–1151.

16. Valicenti RK, Gomella LG, Ismail M, et al. Effect of higher ra-diation dose on biochemical control after radical prostatectomyfor PT3N0 prostate cancer. Int J Radiat Oncol Biol Phys 1998;42:501–506.

17. Anscher MS, Clough R, Dodge R. Radiotherapy for a risingprostate-specific antigen after radical prostatectomy: The first10 years. Int J Radiat Oncol Biol Phys 2000;48:369–375.

18. Macdonald OK, Schild SE, Vora SA, et al. Radiotherapy formen with isolated increase in serum prostate specific antigen af-ter radical prostatectomy. J Urol 2003;170:1833–1837.

19. King CR, Spiotto MT. Improved outcomes with higher dosesfor salvage radiotherapy after prostatectomy. Int J Radiat OncolBiol Phys 2008;71:23–27.

20. Cox JD, Gallagher MJ, Hammond EH, et al. American So-ciety for Therapeutic Radiology and Oncology ConsensusPanel. Consensus statements on radiation therapy of prostatecancer: Guidelines for prostate re-biopsy after radiation andfor radiation therapy with rising prostate-specific antigenlevels after radical prostatectomy. J Clin Oncol 1999;17:1155.

21. Anscher MS. Adjuvant radiotherapy following radical prosta-tectomy is more effective and less toxic than salvage radiother-apy for a rising prostate specific antigen. Int J Cancer 2001;96:91–93.

22. Feng M, Hanlon AL, Pisansky TM, et al. Predictive factors forlate genitourinary and gastrointestinal toxicity in patients withprostate cancer treated with adjuvant or salvage radiotherapy.Int J Radiat Oncol Biol Phys 2007;68:1417–1423.

23. Cozzarini C, Fiorino C, Ceresoli GL, et al. Significant correla-tion between rectal DVH and late bleeding in patients treated af-ter radical prostatectomy with conformal or conventionalradiotherapy (66.6-70.2 Gy). Int J Radiat Oncol Biol Phys2003;55:688–694.

24. Fonteyne V, De Neve W, Villeirs G, et al. Late radiotherapy-in-duced lower intestinal toxicity (RILIT) of intensity-modulatedradiotherapy for prostate cancer: The need for adapting toxicityscales and the appearance of the sigmoid colon as co-responsi-ble organ for lower intestinal toxicity. Radiother Oncol 2007;84:156–163.

25. Schiffner DC, Gottschalk AR, Lometti M, et al. Daily electronicportal imaging of implanted gold seed fiducials in patients un-dergoing radiotherapy after radical prostatectomy. Int J RadiatOncol Biol Phys 2007;67:610–619.

26. De Meerleer G, Fonteyne V, Meersschout S, et al. Salvage in-tensity-modulated radiotherapy for rising PSA after radicalprostatectomy. Radiother Oncol 2008;89:205–213.