Journal of Cancer Research and Clinical Oncology

, Volume 139, Issue 1, pp 147–157

Adjuvant and salvage radiotherapy after prostatectomy: outcome analysis of 307 patients with prostate cancer

Authors

    • Radioterapia, Azienda Ospedaliero-Universitaria di Careggi
  • Silvia Scoccianti
    • Radioterapia, Azienda Ospedaliero-Universitaria di Careggi
  • Sara Cassani
    • Radioterapia, Azienda Ospedaliero-Universitaria di Careggi
  • Samantha Cipressi
    • Radioterapia, Azienda Ospedaliero-Universitaria di Careggi
  • Donata Villari
    • Clinica Urologica IIUniversity of Florence
  • Alberto Lapini
    • Clinica Urologica IUniversity of Florence
  • Calogero Saieva
    • Molecular and Nutritional Epidemiology UnitISPO (Cancer Research and Prevention Institute)
  • Tommaso Cai
    • Department of UrologySanta Chiara Hospital
  • Maurizio Pertici
    • Radioterapia, Azienda Ospedaliero-Universitaria di Careggi
  • Irene Giacomelli
    • Radioterapia, Azienda Ospedaliero-Universitaria di Careggi
  • Lorenzo Livi
    • Radioterapia, Azienda Ospedaliero-Universitaria di Careggi
  • Marco Ceroti
    • Molecular and Nutritional Epidemiology UnitISPO (Cancer Research and Prevention Institute)
  • Giulio Nicita
    • Clinica Urologica IIUniversity of Florence
  • Marco Carini
    • Clinica Urologica IUniversity of Florence
  • Giampaolo Biti
    • Radioterapia, Azienda Ospedaliero-Universitaria di Careggi
Original Paper

DOI: 10.1007/s00432-012-1309-9

Cite this article as:
Detti, B., Scoccianti, S., Cassani, S. et al. J Cancer Res Clin Oncol (2013) 139: 147. doi:10.1007/s00432-012-1309-9
  • 290 Views

Abstract

Aim

In men with adverse pathology after radical prostatectomy, the most appropriate timing to administer radiotherapy (RT) remains a topic of debate. We analyzed in terms of efficacy, prognostic factors and toxicity the two therapeutic strategies: immediate postoperative radiotherapy (PORT) and salvage radiotherapy (SART).

Materials and methods

Between January 1995 and November 2010, 307 patients underwent adjuvant or salvage radiotherapy, after prostatectomy.

Results

In the PORT group, 42 patients (20.7 %) had biochemical failure, with a median time to biochemical failure of 1.8 years; two parameters (age at diagnosis and PSA pre-RT) resulted to be significant at the survival analysis for overall survival (p = 0.003 and p = 0.046, respectively). In the SART group, 33 patients (31.7 %) had biochemical relapse; sixteen patients died of prostate cancer; postoperative hormones therapy, conformal radiotherapy and level of PSA pre-RT >1.0 ng/ml resulted to be significant at the survival analysis, p = 0.009, p = 0.039 and p = 0.002, respectively.

Conclusion

Our study is limited by its retrospective and nonrandomized design. As such, decisions to treat with adjuvant or salvage radiotherapy and the time to initiate therapy were based on patient preference and physician counseling. Our recommendation is to suggest adjuvant radiotherapy for all patients with adverse prognostic factors and to reserve salvage radiotherapy for low-risk patients, when the biochemical recurrence occurs.

Keywords

Prostate cancerAdjuvant radiotherapySalvage radiotherapyProstatectomy

Introduction

Radical prostatectomy (RP) is a standard of care treatment option for prostate cancer, the most commonly diagnosed malignancy in men (Anscher 2004; Hagan et al. 2004; Macdonald et al. 2004). This procedure is a potentially curative treatment alternative to definitive radiation for patients with organ confined disease.

Approximately 27–53 % of men with clinically staged T1 or T2 prostate cancers will develop a detectable serum prostate-specific antigen (PSA) level within 10 years of prostate removal, with rates increasing with pathologically staged T3 or T4 disease (Macdonald et al. 2004).

When a biochemical recurrence occurred, in the absence of a detectable recurrence, it is not easy to distinguish local recurrence in the prostatic bed from distant metastases (Anscher 2004). Approximately 60–70 % of these patients, however, if left untreated would develop overt distant metastases within 10 years (Pound et al. 1999).

Both pathologic pT3 disease and positive surgical margins adversely affect surgical outcomes; these unfavorable features are associated with an increased risk of biochemical relapse, local recurrence, distant metastasis and death from prostate cancer after radical prostatectomy.

The basic premise underlying adjuvant radiotherapy (RT) is that local recurrence antecedes systemic, metastatic spread in many patients in whom radical prostatectomy failed.

Adjuvant radiation therapy in high-risk patients has been shown to increase rates of biochemical control with increased local control rates and disease-free survival. Controversy remains concerning when adjuvant RT should be administered postoperatively (Bolla et al. 2005; Schild 1998, 2001; Catton et al. 2001).

Debate on the optimal management of patients with pathologic T3 or a positive resection margin after radical prostatectomy has continued in recent years.

Recently, three phase III studies reported that postoperative adjuvant radiotherapy improved PSA relapse-free or clinical progression-free survival for patients with pT3 or positive resection margins, in comparison with observation (Bolla et al. 2005; Swanson et al. 2005; Wiegel et al. 2009). The European Association of Urology has also issued a guideline for the management of pT3 or positive surgical margins following RP (Heidenreich et al. 2011).

Materials and methods

This study is a retrospective analysis of patients (Stage pT2–pT4 pN0M0) who received postoperative radiation therapy (RT) at the Radiotherapy Unit, Careggi Hospital in Florence.

A total of 307 patients, who had undergone post-prostatectomy radiotherapy between January 1995 and November 2010, were included in this study. All patients underwent radical prostatectomy and pelvic lymph node dissection. No patients with positive lymph nodes were included in the study. No patient was thought to have metastatic disease based on pre-radiotherapy imaging. We focused on those patients who received adjuvant radiation therapy (PORT), defined as patients who were irradiated within 6 months of their prostatectomy and who had an undetectable serum PSA at the start of radiation therapy and on salvage radiation patients (SART) who were referred for radiation therapy because of a persistent postoperative serum PSA or patients who manifested a PSA recurrence after a period of undetectable PSA.

More in details, 307 patients were analyzed, 203 treated in the adjuvant setting and 104 treated with salvage radiotherapy; their median age was 65.8 years at the time of diagnosis (range 41–83 years). None previously had RT, cytotoxic medications or orchiectomy.

At the histopathological examination, 45 patients (14.7 %) presented an intracapsular disease, and 262 patients (85.3 %), an extracapsular disease; in 125 patients, resection margins were positive, and almost half patients (44 %) showed a Gleason Score >7. The mean value of PSA at the time of diagnosis was 20.4 ng/ml (range 0–128); in particular, the majority of patients (66.1 %) had a level of PSA value between 0 and 20 ng/ml, 79 (25.7 %) patients had a PSA value between 21 and 40 ng/ml and only a few patients (8.2 %) showed level of PSA greater than 40 ng/ml (p < 0.001).

The mean postoperative PSA was 0.36 ng/ml for the whole group (range 0–4.5), PSA being <0.2 ng/ml for more than 60 % of patients. The mean PSA before RT was 0.9 ng/ml (range 0–22.1) for the whole group. The majority of patients (81.4 %) had not then carried out a postoperative hormonal treatment.

Almost all the patients underwent three-dimensional treatment planning (3D-CRT) for post-prostatectomy RT. The planning computed tomography scan was performed with 3-mm slices, with the patient in the supine position and using a leg immobilization system (Combifix-Sinmed, Civco, Kalona, IA, USA). The dose was prescribed at the isocenter according to the International Commission of Radiation Units and Measurements recommendations, and conventional fractionation was used (2 Gy/fraction, five fractions weekly). RT was delivered with a four-field technique and 10–18 MV photons. The clinical target volume (CTV) of RT was limited to the prostatic bed and periprostatic tissue, ensuring adequate coverage of the vesicourethral anastomosis, but no attempt was made to treat regional pelvic lymph nodes. The planning treatment volume (PTV) included the clinical target volume plus a 10-mm margin in all directions. The mean total radiation dose was 66.7 Gy (range 60–74) in 2 Gy fractions, for both PORT and SART.

Two hundred and thirty-one patients (75.2 %) received a total dose ≤66 Gy; the majority of them (85.2 %) has been dispensed a PORT. Seventy-six patients were treated with more than 66 Gy patients, mostly receiving salvage radiotherapy.

More in details, patient’s and treatment’s characteristics were reported in Table 1.
Table 1

Distribution of 307 prostate cancer cases according to RT intent and selected individual clinical characteristics

Characteristic

PORT (n = 203)

SART (n = 104)

All (n = 307)

p*

Age at diagnosis (years)

    

 Mean ± SD

65.1 ± 7.3

67.0 ± 6.0

65.8 ± 7.0

0.047

 Range

41–83

49–78

41–83

 

 ≤67 years

121 (59.6)

53 (51.0)

174 (56.7)

 

 >67 years

82 (40.4)

51 (49.0)

133 (43.3)

0.15

PSA at diagnosis (ng/ml)

    

 Mean ± SD

19.2 ± 37.5

22.7 ± 17.3

20.4 ± 32.1

<0.001

 Range

0–71

1–128

0–128

 

 0–20

166 (81.8)

37 (35.6)

203 (66.1)

 

 21–40

19 (9.4)

60 (57.7)

79 (25.7)

 

 >40

18 (8.9)

7 (6.7)

25 (8.2)

<0.001

Postoperative stage

    

 pT2

22 (10.8)

23 (22.1)

45 (14.7)

 

 pT3–pT4

181 (89.2)

81 (77.9)

262 (85.3)

0.008

Gleason score

    

 ≤6

44 (21.7)

25 (24.0)

69 (22.5)

 

 7

77 (37.9)

26 (25.0)

103 (33.5)

 

 >7

82 (40.4)

53 (51.0)

135 (44.0)

0.069

Margin status

    

 Negative

102 (50.2)

80 (76.9)

182 (59.3)

 

 Positive

101 (49.8)

24 (23.1)

125 (40.7)

<0.001

Postoperative PSA (ng/ml)

    

 Mean ± SD

0.10 ± 0.28

0.85 ± 0.52

0.36 ± 0.52

 

 Range

0–2.65

0.01–4.5

0–4.5

<0.001

 <0.2

184 (90.6)

7 (6.7)

191 (62.2)

 

 0.2–1.0

15 (7.4)

61 (58.7)

76 (24.8)

 

 >1.0

4 (2.0)

36 (34.6)

40 (13.0)

<0.001

Preoperative hormones

    

 No

175 (86.2)

86 (82.7)

261 (85.0)

 

 Yes

28 (13.8)

18 (17.3)

46 (15.0)

0.41

Postoperative hormones

    

 No

173 (85.2)

77 (74.0)

250 (81.4)

 

 Yes

30 (14.8)

27 (26.0)

57 (18.6)

0.017

Pre-RT PSA (ng/ml)

    

 Mean ± SD

0.47 ± 1.73

1.73 ± 3.19

0.90 ± 2.4

 

 Range

0–18.6

0–22.1

0–22.1

<0.001

Radiation dose (Gy)

    

 Mean ± SD

66.2 ± 4.1

66.8 ± 4.1

66.7 ± 4.2

 

 Range

60–74

60–74

60–74

<0.001

 ≤66

173 (85.2)

58 (55.8)

231 (75.2)

 

 >66

30 (14.8)

46 (44.2)

76 (24.8)

<0.001

3D-CRT

    

 No

3 (1.5)

6 (5.8)

9 (2.9)

 

 Yes

200 (98.5)

98 (94.2)

298 (97.1)

0.035

Data in parentheses are percentages

PORT postoperative RT, SART salvage RT, SD standard deviation, RT radiotherapy, HT hormonal therapy

p* = p value from Chi square test or Mann–Whitney Test, as appropriate

During RT, patients were visited by the physicians every 2 weeks, and acute adverse effects of RT were scored, according to the Radiation Therapy Oncology Group (RTOG) scale. Late GI and GU morbidity were prospectively assessed using the National Cancer Institute Expanded Common Toxicity criteria version 2. All patients were followed up in an outpatient basis after the end of radiotherapy every 3 months for the first 2 years, every 6 months for the third- and fourth year and annually thereafter. Their evaluation included the assessment of treatment-related toxicity, a physical examination including rectal examination, blood count and biochemical analyses including the prostatic-specific antigen. Radiological studies were obtained based only on complaints of the patient or when believed clinically to be indicated by the radiation oncologist. Patients were evaluated for biochemical failure, clinical local recurrence and distal metastases. A patient was considered to have a biochemical failure if he had two consecutive measurements of serum PSA increase >0.2 ng/mL. The time between the two PSA measurements was required to be at least 30 days. Additional treatment after biochemical failure was at the discretion of the patient’s radiation oncologist. No uniform criteria for further intervention were applied. The time to biochemical relapse was calculated from the last day of radiotherapy to the time of the first increase of PSA. The follow-up time was calculated from the last day of radiotherapy. The goal of the study was to assess the impact of postoperative radiation therapy on biochemical control.

Statistical analyses

The end of RT was used as the start of observation for the survival analyses. In our survival analysis, we considered as events the deaths for all causes (Overall Survival, OS). For the overall survival (OS) analysis, the survival time was calculated from this date to the date of death or to the last follow-up for the patients resulted to be alive. We also calculated the disease-free survival (DFS) as the time from the end of RT to the occurrence of biochemical recurrences. The crude probability of death or biochemical recurrence was estimated by using the Kaplan–Meier method, and differences between patient groups were assessed by the log rank test. Estimated relative risks of dying or local relapse occurrence were expressed as hazard ratios (HR) and their corresponding 95 % confidence intervals (95 % CI). Univariate Cox regression models were used to evaluate the effect of each specific parameter. Multivariate Cox regression with stepwise selection was performed to identify the major significant death or biochemical recurrence predictors. Statistical results were considered significant at a p value <0.05. All statistical tests were performed using the SPSS for windows software (version 8.1).

Cox proportional hazard regression analysis of possible prognostic factors for relapse was performed. Included for this analysis were age at initial diagnosis, PSA level before treatment (0–20 ng/mL vs. 21–40 ng/mL vs. >40 ng/mL), pathologic postoperative stage (pT2 vs. T3 and pT4), Gleason score (≤6 vs. =7 vs. >7), resection margin status, use of androgen deprivation therapy, postoperative RT PSA level (<0.20 ng/mL vs. 0.20–1 ng/mL vs. ≥1 ng/mL), RT intent (PORT vs. SART) and radiation dose (≤66 Gy vs. >66 Gy).

For the disease-free survival or overall survival analysis, the potential predictors included in each univariate and multivariate model were chosen from the following patient, tumor and treatment-related factors: age, stage, preoperative PSA, use of androgen deprivation therapy (ADT), postoperative/pre-RT PSA level, margin status, Gleason score, RT intent (PORT vs. SART), radiation dose.

Results

At the time of analysis, the median follow-up was 4.9 years (range 1.3–13.3 years) from the end of RT. At the time of analysis, 276 patients were alive, 68 patients were alive with biochemical failure, 26 patients were dead of cancer and five patients were dead of other causes. Out of 203 patients who received elective postoperative radiation therapy, 42 patients (20.7 %) had biochemical failure at the time of analysis, with a median time to biochemical failure of 1.8 years; they all started hormones treatment. They developed bony metastases, and ten patients out of them died of prostate cancer. All of them had adverse prognostic factors at the initial pathological specimens (positive margins and extracapsular extension). Out of 104 patients who received salvage radiotherapy, 33 patients (31.7 %) had biochemical relapse; sixteen patients died of prostate cancer. See Table 2 for more details.
Table 2

Events occurring in the follow-up of 307 prostate cancer cases

Variable

PORT (n = 203)

SART (n = 104)

All (n = 307)

p*

Follow-up (years)

    

 Mean ± SD

3.3 ± 2.3

4.5 ± 2.5

3.7 ± 2.4

<0.001

 Range

1.3–13.3

1.6–11.3

1.3–13.3

 

Died

    

 N (%)

15 (7.4)

16 (15.4)

31 (10.1)

0.044

Biochemical relapse

    

 N (%)

42 (20.7)

33 (31.7)

75 (24.4)

0.036

Time to failure (years)

    

 Mean ± SD

1.8 ± 1.4

2.0 ± 2.1

1.9 ± 1.7

0.82

 Range

0.2–5.0

0.1–7.9

0.1–7.9

 

PORT postoperative RT, SART salvage RT

p* = p value from Chi square test or Mann–Whitney Test, as appropriate

Overall survival rates for the entire cohort were 69.8 % at the time of analysis.

At univariate analysis age, postoperative hormone therapy and PSA pre-RT resulted statistically significant (p = 0.01, p 0.02, p < 0.001, respectively, see Table 3). At multivariate Cox regression analysis with stepwise selection, only two parameters (age at diagnosis over 67 years and PSA pre-RT >1.0 ng/ml) emerged as independent death predictors (HR = 2.37, CI 1.13–4.96, p = 0.022 and HR = 3.51, CI 1.73–7.11, p < 0.001, respectively).
Table 3

Overall survival of 307 prostate cancer cases according to selected individual characteristics: number of patients, number of deaths, survival (%), log rank test and HR and 95 % CI from univariate Cox regression models

Variable

No. of patients

Deaths

OS (%)

Log rank test

HR (95 % CI)*

Age at diagnosis

     

 ≤67 years

174

11

73.1

0.01

1

 >67 years

133

20

66.8

2.56 (1.22–5.35)

PSA at diagnosis (ng/ml)

     

 0–20

203

19

63.8

0.61

 

 21–40

79

8

85.8

 

 >40

25

4

81.3 ≈ 0

 

Postoperative stage

     

 pT2

45

4

77.8

0.70

 

 pT3–pT4

262

27

69.6

 

Gleason score

     

 ≤6

69

8

67.0

0.15

 

 7

103

6

71.6

 

 >7

135

17

79.7

 

Margin status

     

 Negative

182

19

74.8

0.11

 

 Positive

125

12

56.4

 

Postoperative PSA

     

 <0.2

191

15

54.8

0.17

 

 0.2–1.0

76

8

82.4

 

 >1.0

40

8

76.6

 

Preoperative hormones

     

 No

261

26

69.1

0.71

 

 Yes

46

5

78.9

 

Postoperative hormones

     

 No

250

20

73.8

0.02

1

 Yes

57

11

59.9

2.34 (1.12–4.90)

Pre-RT PSA

     

 <0.2

172

9

67.5

<0.001

1

 0.2–1.0

74

6

75.2

1.44 (0.51–4.04)

 >1.0

61

16

65.9

4.23 (1.87–9.57)

RT intent

     

 PORT

203

15

54.7

0.39

 

 SART

104

16

80.5

 

Radiation dose (Gy)

    

 ≤66

231

19

56.9

0.60

 

 >66

76

12

73.5

 

3D-CRT

     

 No

9

3

66.7

0.29

 

 Yes

298

28

68.1

 

* HR and 95 % CI from univariate regression model for parameters with significant log rank test at K–M survival

In the subgroup of patients treated with PORT, two parameters (age at diagnosis and PSA pre-RT) resulted to be significant at the survival analysis for OS (p = 0.003 and p = 0.046, respectively). At multivariate Cox regression analysis, only age at diagnosis emerged as independent death predictor (HR = 5.02, CI 1.58–15.93, p = 0.006).

In the SART group, three parameters (postoperative hormones therapy, conformal radiotherapy and level of PSA pre-RT >1.0 ng/ml) resulted to be significant at the survival analysis (p = 0.009, p = 0.039 and p = 0.002, respectively). At multivariate Cox regression analysis, only level of PSA pre-RT >1.0 ng/ml emerged as independent death predictor (HR = 6.64, CI 1.89–23.33, p = 0.003).

Disease-free survival rates for the entire cohort were 55.6 % at the time of analysis.

At univariate analysis PSA at initial diagnosis, Gleason score and PSA pre-RT resulted statistically significant (p = 0.01, p 0.009, p < 0.0001, respectively, see Table 4). At the multivariate Cox regression analysis for DFS, four parameters emerged as independent death predictors: level of PSA at diagnosis >40 ng/ml (p = 0.007; HR 2.36; 1.26–4.42), Gleason Score >7 (p = 0.0008; HR 2.22; 1.39–3.54), positive resection margins (p = 0.027; HR 1.72; 1.07–2.79) and level of PSA pre-RT >1.0 ng/ml (p > 0.0001; HR 3.35; 2.09–5.35).
Table 4

Disease-free survival of 307 prostate cancer cases according to selected individual characteristics: number of patients, number of relapses, DFS (%), log rank test and HR and 95 % CI from univariate Cox regression models

Variable

No. of patients

Relapses

DFS (%)

Log rank test

HR (95 % CI)*

Age at diagnosis

     

 ≤67 years

174

40

53.5

0.37

 

 >67 years

133

35

59.1

 

PSA at diagnosis

     

 0–20

203

37

72.3

0.01

1

 21–40

79

26

48.9

1.60 (0.97–2.65)

 >40

25

12

20.5

2.50 (1.30–4.81)

Postoperative stage

     

 pT2

45

6

84.1

0.08

 

 pT3–pT4

262

69

50.9

 

Gleason score

     

 ≤6

69

17

43.4

0.009

1

 7

103

16

73.6

0.70 (0.35–1.39)

 >7

135

42

50.4

1.63 (0.93–2.88)

Margin status

     

 Negative

182

46

57.1

0.23

 

 Positive

125

29

58.7

 

Postoperative PSA

     

 <0.2

191

36

68.3

0.08

 

 0.2–1.0

76

25

44.8

 

 >1.0

40

14

60.5

 

Preoperative hormones

     

 No

261

58

61.1

0.10

 

 Yes

46

17

35.1

 

Postoperative hormones

     

 No

250

62

54.5

0.90

 

 Yes

57

13

60.7

 

Pre-RT PSA

     

 <0.2

172

28

65.3

<0.0001

1

 0.2–1.0

74

17

53.9

1.32 (0.72–2.42)

 >1.0

61

30

23.4

3.49 (2.08–5.84)

RT intent

     

 PORT

203

42

66.4

0.29

 

 SART

104

33

47.2

 

Radiation dose (Gy)

     

 ≤66

231

52

65.6

0.51

 

 >66

76

23

52.8

 

3D-CRT

     

 No

9

2

74.1

0.44

 

 Yes

298

73

52.7

 

* HR and 95 % CI from univariate regression model for parameter with significant log rank test at K–M survival

Analyzing the DFS in the PORT group (203 patients, 42 relapses), five parameters resulted to be significant at survival analysis: level of PSA at diagnosis (p = 0.046); status of margins (p = 0.003); postoperative PSA (p = 0.008); preoperative use of hormone therapy (p = 0.014); and level of PSA pre-RT (p = 0.001). At multivariate Cox regression analysis, the following parameters persisted as significant independent relapse predictors: positive margins (p = 0.002; HR 2.76; 1.45–5.26), postoperative levels of PSA between 0.2 and 1.0 ng/ml (p = 0.028; HR 2.70; 1.12–6.54); preoperative use of hormonal therapy (p = 0.017; HR 2.37; 1.17–4.79); and PSA pre-RT >1.0 ng/ml (p = 0.0006; HR 3.58; 1.72–7.43).

However, in the SART group (104 patients, 33 relapses), two parameters resulted statistically significant at survival analysis for DFS: Gleason Score (p = 0.01) and PSA pre-RT (p = 0.006). At multivariate Cox regression analysis with stepwise selection, Gleason Score >7 (p = 0.01; HR = 2.61; 1.23–5.53) and PSA pre-RT >1.0 ng/ml (p = 0.005; HR = 2.57; 1.32–5.40) persisted as significant independent relapse predictors.

Only one patient suffered grade 3 acute urinary and rectal toxicity, whereas none developed acute grade 4 urinary and bowel relapse-free survival in the whole series. The treatment was very well tolerated; all patients completed the planned treatment. Of the 307 patients, 85 suffered from grade 1 or grade 2 cystitis, and 38 patients developed G1–G2 proctitis during radiotherapy treatment. No patients complaint severe late toxicity in the follow up.

Discussion

PORT

Approximately 35 % of patients undergoing a radical prostatectomy for prostate cancer will experience biochemical recurrence within 10 years of surgery. High Gleason score, extracapsular extension, seminal vesicle invasion and positive surgical margins are identified as risk factors for biochemical recurrence after prostatectomy are (Kaplan and Meier 1958; Wheeler et al. 1998; Swindle et al. 2005; Anscher and Prosnitz 1991).

In the current manuscript, we report a retrospective analysis of the impact of PORT and SART on survival after prostatectomy; our results, according with the published data, have shown that postoperative RT offers better disease control after surgery for prostate cancer.

Three randomized studies and several retrospective studies have suggested that patients with unfavorable prognostic factors are at a higher risk of local and/or distant failure after prostatectomy and would be expected to benefit most from postoperative RT.

In 2005, Bolla et al. (2005) first published the results of a large randomized trial, by the European Organisation for Research and Treatment of Cancer (EORTC), among 1,005 patients enrolled to immediate postoperative radiotherapy (502 patients) or to a wait-and-see policy (503 patients). Eligible patients had extracapsular extension and/or positive surgical margins, with pathological negative lymph nodes.

The authors showed that, after a median follow-up of 5 years, immediate radiotherapy after radical prostatectomy improves biochemical progression-free survival (74 vs. 52.6 %) and local control in patients with high risk factors, compared to observation policy. They recommended further follow-up to assess the effect of biochemical progression-free survival on distant metastases and overall survival. The authors reported a cumulative incidence of grade 3 toxicity at 5 years of only 4.2 %.

Later, Van der Kwast et al. (2007) explored the relationship between prognostic factors for recurrence and the magnitude of the benefit from immediate postoperative radiotherapy. This subset analysis, obtained after a central review pathology of 552 patients out of 1,005 enrolled in the EORTC randomized trial, showed that adjuvant radiotherapy reduces the risk of biochemical recurrence, specifically in those with positive surgical margins, whereas those with negative margins (irrespective of other risk factors) in general do not seem to benefit.

In the current study, we did observe an effect from the margin status for disease-free survival in the subgroup of patients treated with PORT, but we did not observe any advantages in the SART patients.

Similar increases in PSA progression-free survival of approximately 20 % were reported in the other two randomized studies.

In the American study, among 425 men with pathologically advanced prostate cancer, Swanson found a 10-year biochemical disease-free survival of 47 versus 23 % in the observation arm, with a longer median follow-up of 9.7 years. The median metastasis-free survival time was slightly improved (14.7 years for RT vs. 13.2 years for observation). Moreover, adjuvant postoperative RT also significantly lengthened the interval to the initiation of androgens deprivation therapy; after 5 years, respectively, 21 and 10 % of the patients in the observation and RT groups had received hormonal therapy (p < 0.001). Additional follow-up is needed to assess the effect of RT on overall survival not significantly improved. Adverse effects were more common with radiotherapy versus observation (23.8 vs 11.9 %), including rectal complications (3.3 vs. 0 %), urethral strictures (17.8 vs. 9.5 %) and total urinary incontinence (6.5 vs. 2.8 %) (Swanson et al. 2005).

Wiegel et al. (2009) reported a biochemical disease-free survival of 81 % in patients with pT3 prostate cancer and an undetectable postoperative PSA (<0.1 ng/mL) compared with 60 % for the observation arm, after a median follow-up of 40 months. On univariate analysis, Gleason score more than 6 and less than 7, a PSA >10 ng/mL before RP, tumor stage and positive surgical margins were predictors of outcome. Compared with both other studies, the rate of adverse effects in the German study was low: there was only one event of grade 3 urinary toxicity with no grade 4 events; altogether, the cumulative rate of adverse effects for bladder and rectum was 21.9 % in the RT arm and 3.7 % in the wait-and-see group.

The reason might be the use of 3D-CRT for all patients in the ARO 96–02/AUO AP 09/95 trial, but not in the former trials.

In fact, it is known that three-dimensional planning radiotherapy reduces acute and late effects, especially for doses greater than 60 Gy (Dearnaley et al. 1999).

In the current study, 95 % of the patients underwent three-dimensional treatment planning; compared with other studies, the rate of adverse effects in our patients was significantly low: only one patient, receiving conventional high-dose radiotherapy suffered grade 3 acute urinary and rectal toxicity, whereas none developed acute grade 4 urinary and bowel toxicity.

The significant local response to radiation therapy is remarkable considering that the radiation doses used (60 to 64 Gy) are suboptimal by today’s standard. Since the results were still significantly positive, further improve in local control would likely may achieve with higher radiation doses or adding other adjuvant treatment.

In the current study, only 12.7 % of PORT patients showed a biochemical relapse with a 5-year PSA recurrence survival of 86.9 %; although longer follow-up is needed to confirm, our results seems better than what reported in the randomized trials, probably due to the high median RT dose (66 Gy) and due to fact that almost all the patients analyzed had undetectable postoperative PSA.

In fact, it is known that postoperative PSA levels represent a prediction of subsequent outcomes. The EORTC trial mandated an undetectable PSA for enrolling patients; as previously reported, the 5-year PSA failure-free rate in radiation patients was 74 %. Again in the SWOG study, the 5-year PSA failure-free rate for patients with an undetectable PSA was higher than observation patients and similar to the Bolla’s results. Also, in the SWOG trial, radiation patients with a post-prostatectomy PSA of more than 0.2 and ≤1.0 ng/mL had a PSA failure-free rate of 34 % at 5 years (compared with 77 % for the ≤0.2 ng/mL group).

In the present series, PSA at initial diagnosis (p = 0.046), postoperative PSA (p = 0.008) and pre-RT PSA (p = 0.001) showed statistical significance at the univariate analysis; moreover, postoperative PSA (p = 0.028) and pre-RT PSA (p = 0.0006) maintained significance at the multivariate analysis.

Concerning the independent prognostic factors for biochemical failure, the available studies in the literature are relatively unanimous on the role of the Gleason score (Petrovich et al. 2002), reporting that a Gleason score ≥8 is a predictor for progression and survival.

Conversely, in the current series, the Gleason score was not an independent risk factor for progression or overall survival in multivariate analysis, although it approached statistical significance and it showed significance in the SART group (p = 0.0008).

The prognostic impact of positive surgical margins on biochemical failure is the subject of much discussion in the literature. In the present study, surgical margin status was an independent predictor for disease recurrence at the multivariate analysis (p = 0.002). These findings are sustained by most of the studies available in the literature.

Positive surgical margins were the strongest predictor of biochemical disease-free survival, in the review pathology data of specimens from participants in the 22911 EORTC trial (Van der Kwast et al. 2007; Dearnaley et al. 1999; Petrovich et al. 2002; Pettus et al. 2004; Palisaar et al. 2002).

Actually, which is the optimal treatment that must be proposed after prostatectomy is not clear; we know that there are benefits of receiving adjuvant treatment, but we do not know if we should propose adjuvant RT alone or adjuvant hormone therapy alone or both. No study is available in the literature comparing different adjuvant treatments after RP.

The findings of the present study suggest that patients who presented risk factors for recurrence, after radical surgery (positive surgical margins and detectable postoperative PSA), may benefit from adjuvant RT combined with hormonal therapy.

The findings of the present study are in accordance with several prospective randomized controlled trials showing that early hormonal therapy combined with RT can significantly improve survival in patients with locally advanced prostate cancer (Bolla et al. 1997; Lawton et al. 2001).

Ten years ago, Corn et al. (1999) published a secondary analysis of RTOG trial 8531, to evaluate the effect of immediate androgen suppression in conjunction with radiotherapy versus radiotherapy alone on a group of men after prostatectomy. With a median follow-up of 5 years, the estimated progression-free survival rate was 65 % for the men who received combination therapy and 42 % for those treated by RT alone with hormones reserved for relapse (p = 0.002). However, no differences were evident in the rates of clinically diagnosed local progression, distant relapse or overall survival between the two groups.

Choo et al. (2009) reported the efficacy of a combined approach of postoperative radiotherapy plus 2 years of androgen suppression, for 78 patients with pT3 disease and/or positive surgical margins after radical prostatectomy. With a median follow-up from RT of 6.4 years, relapse-free rates at 5 and 7 years were 94.4 and 86.3 %, respectively, and survival rates were 96 % at 5 years and 93.1 % at 7 years.

These results seems superior to the ones reported in the randomized trial of postoperative radiotherapy but the study shows some shortcomings, as the authors correctly reported, in particular the low number of patients enrolled (only 78), short follow-up (median follow-up from RT of 6.4 years) and, lastly, the use of neoadjuvant androgen ablation therapy in 10 % of patients.

So until additional prospective randomized data will not be available, androgen deprivation therapy cannot yet be viewed as a standard of care in addition to postoperative radiotherapy.

The ongoing RTOG trial 9601 is comparing adjuvant RT with or without bicalutamide (150 mg daily for 2 years) for prostatectomized patients who are found to have pathologic Stage T3 disease.

The ongoing Radiation Therapy Oncology Group 0621 Phase II study is addressing the issue of adjuvant RT combined with hormones and chemotherapy.

SART

At the present time, there are no published randomized trials that compare adjuvant versus salvage RT. Although the three PORT trials encouraged salvage treatment in those who failed after observation, the parameters for salvage treatment were not predetermined. There is a recently activated study designed to evaluate the issue of salvage versus adjuvant RT and the role of androgen deprivation therapy in such a setting: the RADICALS (Radiotherapy and Androgen Deprivation in Combination after Local Surgery) (Parker et al. 2007). Post-prostatectomy patients are randomized to early radiotherapy or to delayed treatment (when there are two consecutive rises with a PSA >0.1 ng/mL or three consecutive rises).

A few SART series indicate a long-term free relapse of 30–50 % but only one recent series showed a significant impact on survival. Trock et al. (2008) compared the outcome in 238 men treated with SART to that in 397 with PSA recurrence after surgery who did not receive salvage treatment. SART was associated with a significant increase in cancer specific survival but this benefit was limited to men with PSA doubling time less than 6 months who received SART within 2 years.

Trabulsi et al. (2008) reported on 449 patients with extra capsular prostate cancer, 211 patients receiving adjuvant radiotherapy and 238 patients receiving salvage radiotherapy. PORT significantly reduced the risk of long-term biochemical progression after RP compared with SART (5-year biochemical control was 73 % after adjuvant radiotherapy, compared with 50 % after SART; p = 0.007).

Recently, Geinitz et al. (2012) performed a retrospective analysis of 96 men treated with salvage radiotherapy for intermediate- or high-risk prostate cancer, with more than 50 % of them having positive surgical margins. After SART, 66 % of patients reached a PSA nadir of <0.2 ng/mL but the 5-year biochemical free relapse survival was only 35 %.

One of the largest mono-institutional studies analyzing the outcome of patients treated with salvage RT is the one reported by Neuhof et al. in 2007. In their study, at the univariate analysis, five parameters were statistically significant predictors of PSA recurrence: preoperative PSA level, pathologic stage, Gleason score, tumor grading and pre-RT PSA level. On multivariate analysis, only Gleason score and pre-RT PSA level were found to be independently predictive of PSA recurrence (Neuhof et al. 2007).

Recently, Ohri et al. (2012) performed a review of published series reporting treatment outcomes following salvage radiotherapy. Radiobiological models demonstrate the interaction between pre-RT PSA, RT dose and bPFS. On multivariate analysis, bPFS increased with salvage RT dose by 2.5 % per Gy and decreased with pre-RT PSA by 18.3 % per ng/mL (p < 0.001).

The optimal dose for salvage radiotherapy is still a matter of debate.

For salvage RT, the American Society for Therapeutic Radiology and Oncology (ASTRO) consensus guidelines recommend using the highest radiation dose that can be delivered with acceptable morbidity and suggest a minimum of 64 Gy.

Tomita et al. (2009) analyzed 51 patients with a persistent or rising PSA >0.20 ng/ml at some point after surgery, who were treated with salvage radiotherapy (median dose of 60 Gy). After a median follow-up of 36 months, the 3-year bNED was 55.1 %, and at the multivariate analysis, PSADT >3.0 months, Gleason score ≤7 and RT dose ≥60 Gy were significantly associated with improved bNED.

In our patients, we did not observe a statistical significance for radiation dose in terms of OS and DFS in the SART group.

On multivariate analysis, among 59 patients treated with salvage RT by Quero et al. (2008), at a median dose of 66 Gy, only pre-RT PSA ≥1.0 ng/ml predicted relapse. The authors concluded that RT should be given earlier after biochemical relapse to obtain optimal biochemical control.

Indeed, a few retrospective studies on salvage RT showed that doses >64 Gy resulted in better biochemical control than lower doses. The much larger study by Bernard suggested that doses greater than 66.6 Gy improve bNED.

Actually, a phase III trial is needed to address the question of optimal dose for salvage radiotherapy.

When deciding whether to administer PORT or SART, the potential benefits of treatment must be weighed against potential treatment toxicity. Postoperative radiation is associated with late urinary and gastrointestinal toxicity in 10–20 % of patients. The SWOG trial described proctitis and rectal bleeding in 3.2 % of patients, urethral stricture in 17 %, urinary incontinence in 6.5 % and a higher overall rate of adverse events (Swanson et al. 2005). Similarly, SART analysis revealed diarrhea in 31 % of patients and proctitis in 41 % (Jung et al. 2007).

Conclusion

Our study is limited by its retrospective, nonrandomized design. As such, decisions to treat with adjuvant or salvage radiotherapy and the time to initiate therapy were based on patient preference and physician counseling. Thus, they were subject to inherent selection bias. Immediate postoperative and salvage RT have not been compared in a prospective controlled trial. Since only retrospective data are available, no conclusions could be drawn.

Supporters of salvage RT alone argue that the benefit of this technique when applied early appears to be similar to that of postoperative RT, and moreover, salvage RT avoids giving postoperative RT to patients who do not need it.

Anyway, even though postoperative RT means treating patients who could not otherwise have developed a relapse, one potential side effect of salvage RT is that it loses patients whose recurrence could have been prevented by postoperative RT. Only randomized trials will be able to validly compare postoperative RT and salvage RT; therefore, inclusion in the ongoing studies of this type should be encouraged. Our recommendation is to suggest adjuvant RT for all patients with adverse prognostic factors. For low-risk features, a wait-and-see policy is reasonable, with salvage RT delivered early when a biochemical recurrence occurs.

Conflict of interest

We declare that we have no conflict of interest.

Copyright information

© Springer-Verlag 2012