EUROPEAN UROLOGY 62 (2012)

EUROPEAN UROLOGY 62 (2012) 472 487 available at www.sciencedirect.com journal homepage: www.europeanurology.com Platinum Priority Prostate Cancer Editorial by Judd W. Moul on pp. 488 490 of this issue Early Salvage Radiation Therapy Does Not Compromise Cancer Control in Patients with pt3n0 Prostate Cancer After Radical Prostatectomy: Results of a Match-controlled Multi-institutional Analysis Alberto Briganti a,y, *, Thomas Wiegel b,y, Steven Joniau c, Cesare Cozzarini d, Marco Bianchi a,e, Maxine Sun e, Bertrand Tombal f, Karin Haustermans g, Tom Budiharto g, Wolfgang Hinkelbein h, Nadia Di Muzio d, Pierre I. Karakiewicz e, Francesco Montorsi a, Hein Van Poppel c a Department of Urology, Vita-Salute University, San Raffaele Scientific Institute, Milan, Italy; b Department of Radiation Oncology, University Hospital Ulm, Ulm, Germany; c University Hospitals Leuven, Department of Urology, Leuven, Belgium; d Department of Radiotherapy, San Raffaele Scientific Institute, Milan, Italy; e Cancer Prognostics and Health Outcomes Unit, University of Montreal, Montreal, Canada; f Université Catholique de Louvain Department of Urology, Brussels, Belgium; g University Hospitals Leuven, Department of Radiology, Leuven, Belgium; h Department of Radiation Oncology, Charité Universitätsmedizin, Campus Benjamin Franklin, Berlin, Germany Article info Article history: Accepted April 26, 2012 Published online ahead of print on May 15, 2012 Keywords: Prostate cancer Radical prostatectomy Adjuvant radiation therapy Early salvage radiation therapy Abstract Background: Previous randomised trials demonstrated that adjuvant radiation therapy (art) improves cancer control in patients with pt3 prostate cancer (PCa). However, there is currently no evidence supporting early salvage radiation therapy (esrt) as equivalent to art in improving freedom from biochemical recurrence (BCR) after radical prostatectomy (RP). Objective: To evaluate BCR-free survival for art versus observation followed by esrt in cases of relapse in patients undergoing RP for pt3pn0, R0 R1 PCa. Design, setting, and participants: Using a European multi-institutional cohort, 890 men with pt3pn0, R0 R1 PCa were identified. Intervention: All patients underwent RP. Subsequently, patients were stratified into two groups: art versus initial observation followed by esrt in cases of relapse. Outcome measurements and statistical analyses: Propensity-matched analysis was employed, and patients were stratified into two groups: art versus observation and eventual esrt, defined as RT given at a postoperative serum prostate-specific antigen (PSA) 0.5 ng/ml at least 6 mo after RP. BCR, defined as PSA >0.20 ng/ml and rising after administration of RT, was compared between art and initial observation followed by esrt in cases of relapse using Kaplan-Meier and Cox regression methods. Results and limitations: Overall, 390 (43.8%) and 500 (56.2%) patients were treated with art and initial observation, respectively. Within the latter group, 225 (45.0%) patients experienced BCR and underwent esrt. In the postpropensity-matched cohort, the 2- and 5-yr BCR-free survival rates were 91.4% and 78.4% in art versus 92.8% and 81.8% in patients who underwent initial observation and esrt in cases of relapse, respectively ( p = 0.9). No differences in the 2- and 5-yr BCR-free survival rates were found, even when patients were stratified according to pt3 substage and surgical margin status (all p 0.4). These findings were also confirmed in multivariable analyses ( p = 0.6). Similar results were achieved when the cut-off to define esrt was set at 0.3 ng/ml (all p 0.5). y These authors contributed equally to the manuscript. * Corresponding author. Vita-Salute San Raffaele University, Urologic Research Institute, Via Olgettina 60, 20132 Milan, Italy. Tel. +39 022 643 7286; Fax: +39 022 643 7298. E-mail address: briganti.alberto@hsr.it (A. Briganti). 0302-2838/$ see back matter # 2012 European Association of Urology. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.eururo.2012.04.056

EUROPEAN UROLOGY 62 (2012) 472 487 473 Conclusions: The current study suggests that timely administration of esrt is comparable to art in improving BCR-free survival in the majority of pt3pn0 PCa patients. Therefore, esrt may not compromise cancer control but significantly reduces overtreatment associated with art. # 2012 European Association of Urology. Published by Elsevier B.V. All rights reserved. 1. Introduction Approximately 50% of pt3, node-negative prostate cancer (PCa) patients treated with radical prostatectomy (RP) will experience biochemical recurrence (BCR), particularly those with poorly differentiated disease and positive surgical margins (PSM) [1,2]. In this context, three prospective randomised trials demonstrated that adjuvant radiation therapy (art) relative to observation improves freedom from BCR [3 6]. Only one of these studies reported a positive effect for art on metastasis-free survival and overall survival [5]. In contrast, no trials included patients who underwent early salvage radiation therapy (esrt). As such, the existence of an equivalent evidence level is substantially lower with regard to salvage radiation therapy (SRT). Nonetheless, there is great interest in administering radiation therapy (RT) only at the time of prostate-specific antigen (PSA) recurrence because treatment would then be limited solely to recurring patients, and this would reduce treatment-related side effects. Moreover, concerns of overtreatment of art candidates may be diminished because a significant proportion of patients with adverse pathologic features might be cured by RP alone, with a low risk of long-term PCa-specific mortality [7,8]. The correct alternative to art would be to postpone it until the first occurrence of a rising PSA. In this context, retrospective observational data suggest that SRT has potential, given its association with durable response rates and cancer-control benefits [9 15]. Moreover, these reports have consistently showed that a low pretreatment serum PSA level was the strongest determinant of response after SRT. Unfortunately, the majority of patients treated with SRT in these studies had a PSA >0.5 ng/ml, therefore being irradiated in a late salvage scenario rather than in a true salvage setting [11,12]. Such delay might decrease the efficacy of SRT [9 15]. Given these existing limitations, we examined BCR-free survival of patients treated with art versus initial observation followed by esrt in cases of relapse. Our hypothesis is that given the administration of RT at low pretreatment PSA levels, esrt is comparable to art in improving freedom from BCR. 2. Materials and methods 2.1. Study population Between May 1991 and June 2007, 952 pt3pn0, M0, R0 R1 patients treated with RP and pelvic lymph node dissection at European tertiary care centres were considered. No patient received any neoadjuvant or adjuvant hormone therapy. Patients with a serum PSA >0.5 ng/ml at SRT were excluded from the analyses (n = 62). This discrimination resulted in 890 assessable patients. For this study, art was defined as immediate RT given within 6 mo after surgery with an undetectable PSA (<0.10 ng/ml), and esrt was defined as RT administered at a postoperative serum PSA 0.5 ng/ml at least 6 mo after RP. 2.2. Radiation therapy technique RT was defined as local radiation to the prostate and seminal vesicle bed alone; the pelvic lymph nodal area was not irradiated. All patients were treated with high-energy photon beams (10 25 MV) at conventional fractionation (1.8 2 Gy per fraction). Conventional nonconformal treatment was delivered, and rectangular or minimally blocked beams were used. Alternatively, a three-dimensional conformal RT was used. The clinical target volume (CTV) was delineated on computed tomography (CT) images and included the prostatic fossa and periprostatic tissue. Clinical findings, presurgery CT scans, and surgical clips guided the clinicians in defining CTV. The planned target volume was defined as CTV plus a 0.8- to 1-cm margin to account for organ motion and setup error. 2.3. Covariates and end points Patient age and serum PSA level at surgery, total dose of RT delivered, 2002 TNM pathologic stage and Gleason score, surgical margin status, serum PSA level at RT, and year of surgery were considered. BCR was defined as PSA >0.20 ng/ml and rising after administration of either art or esrt. Thus in patients undergoing initial observation and subsequent esrt because of a first BCR, the end point was considered as a further recurrence after esrt. Follow-up time was defined as the time (months) between surgery and BCR after RT or last follow-up. 2.4. Statistical analyses Independent-sample t tests and x 2 tests were used for comparisons of means and proportions, respectively. The x 2 trend test was used to evaluate the differences in treatment type over time. A one-to-one propensity score-matched analysis was used. Such a method can reduce residual confounding that may be associated with more conventional modelling [16]. Propensity scores were computed by modelling a logistic regression, with the dependent variable as the odds of undergoing art and independent variables such as Gleason score (<6 vs 6 vs 7 vs 8 vs 9), pathologic T stage (pt3a vs pt3b), and surgical margin status (positive vs negative). Subsequently, covariate balance between the matched groups was examined. Covariates between the two groups were considered equivalent, provided a standardised mean difference 10%. In the postpropensity-matched cohort, the Kaplan-Meier method was used to compare BCR-free survival of art versus esrt patients. These rates were computed at 2, 5, and 5 yr after surgery. The log-rank test was used to assess the statistical significance between the two groups. Subanalyses were performed within patients with negative surgical margins (NSM) or PSM with pt3a or pt3b disease as well as in patients with pt3a disease and NSM, pt3a disease and PSM, and pt3b disease and NSM. Multivariable Cox regression analysis was used to predict the risk of BCR in the postpropensity-matched cohort. Covariates comprised postoperative RT type (art vs esrt), RT dose, year of surgery (quartiles), patient age at surgery, preoperative serum PSA, pathologic

474 EUROPEAN UROLOGY 62 (2012) 472 487 Table 1 Descriptive characteristics of patients treated with radical prostatectomy for pt3n0 prostate cancer, stratified according to postoperative management, namely, adjuvant radiation therapy versus observation and early salvage radiation therapy (given at a prostate-specific antigen value =0.5 ng/ml) in cases of relapse Prepropensity score matching Postpropensity score matching Overall Adjuvant RT No adjuvant RT p value Adjuvant RT No adjuvant RT p value No. of patients (%) 890 (100.0) 390 (43.8) 500 (56.2) 390 390 Age at surgery, yr 0.1 0.4 Mean (median) 64.7 (65) 64.0 (64) 64.9 (65) 64.0 (64) 64.5 (65) IQR 61 69 60 68 61 69 60 68 61 69 Follow-up time, mo 0.001 <0.001 Mean (median) 54.3 (47) 71.9 (67) 40.6 (31) 71.9 (67) 41.0 (30) IQR 18 79 39 103 10 60 39 103 10 60 PSA at surgery 0.2 0.1 Mean (median) 13.8 (10) 14.2 (10) 13.4 (10) 14.2 (10) 13.2 (10) IQR 6.1 14.6 6.7 16.1 6.5 14.6 6.7 16.1 6.3 14.7 Dose of RT 0.05 0.1 Mean (median) 65.2 (66.0) 64.9 (65) 66.2 (66) 64.9 (65) 66 (66) IQR 60 70 60 70 66 66 60 70 66 66 Pathologic T stage, mean (median) 0.001 0.3 pt3a 645 (72.5) 261 (66.9) 384 (76.8) 261 (66.9) 274 (70.3) pt3b 245 (27.5) 129 (33.1) 116 (23.2) 129 (33.1) 116 (29.7) Surgical margin status, mean (median) <0.001 0.6 Negative 419 (47.1) 145 (37.2) 262 (52.4) 145 (37.2) 152 (39.0) Positive 471 (52.9) 245 (62.8) 238 (47.6) 245 (62.8) 238 (61.0) Pathologic Gleason score, mean (median) 0.003 0.1 <6 146 (16.4) 85 (21.8) 61 (12.2) 85 (21.8) 61 (15.6) 6 178 (20.0) 75 (19.2) 103 (20.6) 75 (19.2) 102 (26.2) 7 444 (49.9) 185 (47.4) 259 (51.8) 185 (47.4) 173 (44.4) 8 81 (9.1) 30 (7.7) 51 (10.2) 30 (7.7) 36 (9.2) 9 41 (4.6) 15 (3.8) 26 (5.2) 15 (3.8) 18 (4.6) Year of surgery, mean (median) <0.001 <0.001 1991 1998 145 (28.7) 172 (44.1) 83 (16.6) 172 (44.1) 74 (19.0) 1999 2000 122 (19.3) 98 (25.1) 74 (14.8) 98 (25.1) 64 (16.4) 2001 2003 118 (28.4) 92 (23.6) 161 (32.2) 92 (23.6) 131 (33.6) 2004 2007 128 (23.6) 28 (7.2) 182 (36.4) 28 (7.2) 121 (31.0) PSA at RT <0.001 <0.001 Mean (Median) 0.2 (0.1) 0.01 (0) 0.22 (0.2) 0.01 (0) 0.22 (0.2) IQR 0.0 0.28 0.0 0.0 0.1 0.3 0.0 0.0 0.1 0.3 RT = radiation therapy; IQR = interquartile range; PSA = prostate-specific antigen. Gleason score (<6 vs 6 vs 7 vs 8 vs 9), pathologic T stage (pt3a vs pt3b), and surgical margin status (positive vs negative). Finally, all the aforementioned steps were repeated in patients who received esrt administered at a serum PSA 0.3 ng/ml. All statistical tests were performed using the R statistical package (v.2.13.1), with a two-sided significance level set at p < 0.05. 3. Results Baseline characteristics are summarised in Table 1. Overall, 390 (43.8%) patients were treated with art, while 500 (56.2%) patients underwent an initial observation. Within the latter group, 225 (45.0%) patients experienced BCR after RP and consequently underwent esrt. Patients treated with art had higher rates of seminal vesicle invasion (33.1% vs 23.2%; p = 0.001) as well as higher rates of PSM (62.8% vs 47.6%; p < 0.001) than patients undergoing observation eventually followed by esrt. Moreover, patients receiving art had significantly longer mean follow-up (71.9 vs 40.6 mo; p = 0.001). Mean (median) RT dose was 64.9 (65) and 66.2 (66) Gy for those receiving art and esrt, respectively ( p = 0.05). A significant increase in the use of initial observation and esrt in cases of relapse was observed over time during the study period (all p < 0.001; Fig. 1). For the purpose of the study, 390 of 500 (78.0%) of patients who underwent initial observation were matched, with 390 treated with art. In the postpropensity-matched cohort, no statistically significant differences in terms of pathologic characteristics were found between the two groups (all p 0.1; Table 1), and covariate balance was achieved as demonstrated by a standardised mean difference 10%. Similarly, mean RT dose was comparable between those treated with art and those receiving esrt (64.9 vs 66 Gy, respectively; p = 0.1). Within the matched population, no statistically significant differences in terms of BCR-free survival rates were found between patients undergoing art and those managed with initial observation followed by esrt in cases of relapse (Fig. 2). Specifically, the 2- and 5-yr BCR-free survival rates were 91.4% and 78.4%, respectively, in patients who underwent art versus 92.8% and 81.8%, respectively, in patients who underwent observation followed by eventual esrt (hazard ratio [HR]: 0.9; p = 0.9). This finding was confirmed in subanalyses performed in patients with NSM or PSM with pt3a or pt3b disease as well as in patients with pt3a and NSM, pt3a and PSM, and pt3b and NSM (all p 0.4; Fig. 3 and Fig. 4). In multivariable analyses, pathologic Gleason score, preoperative PSA, and pathologic T stage emerged as significant

[(Fig._1)TD$FIG] EUROPEAN UROLOGY 62 (2012) 472 487 475 Fig. 1 Temporal trends in adjuvant radiation therapy (art) versus initial observation (obs) and eventual early salvage radiation therapy (esrt) use (A) within the overall prematch patient population; (B) after propensity score matching, when esrt was defined as radiation therapy administered at a prostate-specific antigen (PSA) =0.5 ng/ml; and (C) after propensity score matching, when esrt was defined as radiation therapy administered at a PSA =0.3 ng/ml.

476 [(Fig._2)TD$FIG] EUROPEAN UROLOGY 62 (2012) 472 487 Fig. 2 Kaplan-Meier estimates of biochemical recurrence free survival in patients receiving adjuvant radiation therapy (art) versus observation (obs) and early salvage radiation therapy (esrt; prostate-specific antigen at salvage radiation therapy =0.5 ng/ml) in cases of relapse (n = 390 given for the two arms separately; log-rank test: p = 0.9; hazard ratio: 0.9). BCR = biochemical recurrence. predictors of BCR (all p 0.03). Conversely, postoperative management (art vs esrt) and all other considered covariates namely, age, RT dose, year of surgery, and surgical margin status failed to reach statistical significance (all p 0.5; Table 2). In the second part of the analysis, matching was performed by lowering the PSA cut-off at esrt to 0.3 ng/ml. In this analysis, 390 of 469 (83.1%) of patients who underwent initial observation and eventual esrt were matched, with 390 treated with art. Following propensity-based matched analysis, no statistically significant differences in terms of pathologic characteristics were found between the two groups of patients (all p 0.1; Table 3). As such, no statistically significant differences with respect to BCR-free survival rates were found between patients undergoing art and patients who underwent observation and eventual esrt (Fig. 5). Specifically, the 2- and 5-yr BCR-free survival rates were 91.4% and 78.4% versus 93.1% and 81.9% in art and observation with eventual esrt patients, respectively (HR: 0.8; p = 0.7). This finding was confirmed in subanalyses performed in all patient subgroups (all p 0.6; Fig. 6 and Fig. 7). Similarly, in multivariable analyses predicting BCR, only pathologic Gleason score and T stage were significantly associated with higher risk of BCR (all p < 0.001), while neither postoperative management (art vs initial observation and/or esrt) nor any other covariates reached statistical significance (all p 0.06; Table 4). Table 2 Multivariable Cox regression analyses predicting biochemical recurrence in patients treated with radical prostatectomy for pt3n0, R0 R1 prostate cancer Variable Multivariable analysis HR (95% CI) Age at surgery 0.99 (0.95 1.04) 0.8 Postoperative management: art 1.0 (ref.) Observation and esrt * in case of relapse 0.91 (0.65 1.28) 0.6 Year of surgery 1991 1998 1.0 (ref.) 1999 2000 1.05 (0.69 1.59) 0.8 2001 2003 1.17 (0.75 1.81) 0.5 2004 2007 1.77 (0.99 3.10) 0.05 RT dose ** 1.06 (0.96 1.06) 0.8 Preoperative PSA 1.01 (1.0 1.02) 0.03 Pathologic Gleason score: <6 1.0 (ref.) 6 1.71 (0.86 3.41) 0.1 7 3.25 (1.81 5.86) <0.001 8 4.78 (2.37 9.62) <0.001 9 5.61 (2.41 13.01) <0.001 Surgical margin status: Negative 1.0 (ref.) Positive 1.01 (0.72 1.47) 0.9 Pathologic T stage: pt3a 1.0 (ref.) pt3b 2.48 (1.79 3.43) <0.001 HR = hazard ratio; CI = confidence interval; art = adjuvant radiation therapy; esrt = early salvage radiation therapy; RT = radiation therapy; PSA = prostate-specific antigen. * Defined as RT administered at a PSA 0.5 ng/ml. ** Analysis performed only in patients who underwent art or esrt. p

[(Fig._3)TD$FIG] EUROPEAN UROLOGY 62 (2012) 472 487 477 Fig. 3 Kaplan-Meier estimates of biochemical recurrence free survival in patients receiving adjuvant radiation therapy (art) versus observation (obs) and early salvage radiation therapy (esrt; prostate-specific antigen at salvage radiation therapy = 0.5 ng/ml) in cases of relapse in patients with (A) pt3a disease, (B) pt3b disease, (C) negative surgical margins, or (D) positive surgical margins. BCR = biochemical recurrence. 4. Discussion Postoperative RT administered either immediately after surgery (art) or at BCR may improve disease control and patient survival [5,9,12,13]. To date, three randomised trials have examined the rate of BCR-free survival in patients treated with art versus observation [3,4,6]. However, the real clinically relevant question is whether RT should be administered immediately in a blanket fashion or postponed until first occurrence of PSA rise. To date, this

478 [(Fig._3. (Continued).)TD$FIG] EUROPEAN UROLOGY 62 (2012) 472 487 Fig. 3. (Continued). question remains unanswered, because existing studies assessing the value of SRT were conducted in patients with advanced PSA recurrence (median PSA: 0.5 ng/ml) [4,5,11,12]. Moreover, of those with treatment failure after RP, only half actually received SRT [3]. As such, it has yet to be determined whether initial observation followed by esrt in cases of relapse is inferior to art. Given the absence of data comparing art and esrt, we sought to examine this issue. Following stringent inclusion criteria, no difference was found with regard to post-rt BCR between the two approaches (all p 0.4), even in subanalyses performed according to margin status and pt3 substages. To date, only a few studies have addressed a similar topic. Trabulsi et al, using a multi-institutional database of 211 art versus 238 SRT pt3 4N0 patients, showed that art significantly reduced the risk of long-term BCR after RP

[(Fig._4)TD$FIG] EUROPEAN UROLOGY 62 (2012) 472 487 479 Fig. 4 Kaplan-Meier estimates of biochemical recurrence free survival in patients receiving adjuvant radiation therapy (art) versus observation and early salvage radiation therapy (esrt; prostate-specific antigen at salvage radiation therapy =0.5 ng/ml) in cases of relapse in patients with (A) pt3a disease and negative surgical margins (NSM), (B) pt3a disease and positive surgical margins, and (C) pt3b disease and NSM. BCR = biochemical recurrence. compared with SRT (5-yr BCR-free survival rate: 73% vs 50%; p = 0.007) [17]. Similar results were reported by Ost et al, who showed a 5-yr BCR-free survival rate of 85% for art patients versus 65% for salvage intensity-modulated RT ( p = 0.002) [18]. However, when subgroup matching was performed in a cohort of 76 patients receiving either art or esrt (given at a PSA <0.5 ng/ml), no significant difference in the 3-yr BCR-free survival rates was found. Based on findings from a recent multi-institutional matched analysis comparing art and SRT in high-risk PCa patients, Budiharto et al. recommended art in all locally advanced PCa cases [19]. Despite the relevance of the

480 [(Fig._4. (Continued).)TD$FIG] EUROPEAN UROLOGY 62 (2012) 472 487 Fig. 4. (Continued). aforementioned studies, all were unfortunately hampered by important biases. First, none of these studies was able to account for the true denominator in the salvage group, namely, the proportion of patients with similar features [(Fig._5)TD$FIG] who were initially observed after RP and never recurred. The lack thereof heavily undermines the applicability of those results in clinical practice because in reality, as an alternative to immediate RT (art), initial observation is the Fig. 5 Kaplan-Meier estimates of biochemical recurrence free survival in patients receiving adjuvant radiation therapy (art) versus observation (obs) and early salvage radiation therapy (esrt; prostate-specific antigen at salvage radiation therapy =0.3 ng/ml) in cases of relapse (n = 390 given for the two arms separately; log-rank test: p = 0.7; hazard ratio: 0.8). BCR = biochemical recurrence.

EUROPEAN UROLOGY 62 (2012) 472 487 481 Table 3 Descriptive characteristics of patients treated with radical prostatectomy for pt3n0 prostate cancer, stratified according to postoperative management, namely, adjuvant radiation therapy versus observation and early salvage radiation therapy (given at a prostate-specific antigen value =0.3 ng/ml) in cases of relapse Prepropensity score matching Postpropensity score matching Overall Adjuvant RT No adjuvant RT p value Adjuvant RT No adjuvant RT p value No. of patients (%) 859 (100.0) 390 (45.4) 469 (54.6) 390 390 Age at surgery, yr: 0.1 0.2 Mean (median) 64.7 (65) 64.0 (64) 64.9 (65) 64.0 (64) 64.7 (64) IQR 61 69 60 68 61 69 60 68 61 69 Follow-up time, mo: <0.001 <0.001 Mean (median) 55.1 (48) 71.9 (67) 41.1 (32) 71.9 (67) 41.7 (33) IQR 19 80 39 103 10 60 39 103 10 60 PSA at surgery: 0.1 0.3 Mean (median) 13.4 (10) 14.2 (10) 12.2 (9) 14.2 (10) 12.2 (10) IQR 6.7 15.5 6.7 16.1 6.1 14.6 6.7 16.1 6.0 14.9 Dose of RT: 0.1 0.2 Mean (median) 65.1 (66.0) 65.0 (65) 66.1 (66) 65.0 (65) 66.0 (66) IQR 60 70 60 70 66 66 60 70 66 66 Pathologic T stage, mean (median): 0.001 0.1 pt3a 621 (72.3) 261 (66.9) 360 (76.8) 261 (66.9) 283 (72.6) pt3b 238 (27.7) 129 (33.1) 109 (23.2) 129 (33.1) 101 (27.4) Surgical margin status, mean (median): <0.001 0.1 Negative 395 (46.0) 145 (37.2) 250 (53.3) 145 (37.2) 171 (43.8) Positive 464 (54.0) 245 (62.8) 219 (46.7) 245 (62.8) 219 (56.2) Pathologic Gleason score, mean (median): 0.003 0.1 <6 141 (16.4) 85 (21.8) 56 (11.9) 85 (21.8) 56 (14.4) 6 172 (20.0) 75 (19.2) 97 (20.7) 75 (19.2) 97 (24.9) 7 429 (49.9) 185 (47.4) 244 (52.0) 185 (47.4) 191 (49.0) 8 77 (9.0) 30 (7.7) 47 (10.0) 30 (7.7) 31 (7.9) 9 40 (4.7) 15 (3.87) 25 (5.3) 15 (3.87) 15 (3.8) Year of surgery, mean (median): <0.001 <0.001 1991 1998 246 (28.6) 172 (44.1) 74 (15.8) 172 (44.1) 70 (17.9) 1999 2000 166 (19.3) 98 (25.1) 68 (14.5) 98 (25.1) 59 (15.1) 2001 2003 242 (28.2) 92 (23.6) 150 (32.0) 92 (23.6) 129 (33.1) 2004 2007 205 (23.9) 28 (7.2) 177 (37.7) 28 (7.2) 132 (33.8) PSA at RT: <0.001 <0.001 Mean (median) 0.06 (0) 0.01 (0) 0.17 (0.15) 0.01 (0) 0.16 (0.15) IQR 0 0 0 0 0.10 0.23 0 0 0.10 0.22 RT = radiation therapy; IQR = interquartile range; PSA = prostate-specific antigen. second option. Consequently, the lack of inclusion of recurrence-free patients may significantly bias the analyses in favour of art. To account for this limitation, we included patients who remained BCR free postoperatively within the esrt group. The proportion of patients within this group was 55%. Such a proportion is comparable to the rate of pt3 patients not recurring at 5 yr after RP enrolled in the observational arm of previous prospective randomised trials (51 58%) [3,6,20]. This also was true when such comparison was done according to each patient subgroup [4,20 22] (Table 5). Unfortunately, the median follow-up of our study did not allow us to reliably consider the proportion of patients developing BCR beyond the 5-yr follow-up time, which represents roughly 20% of all pt3 patients [5]. A second caveat that pertains to previous analyses relates to the lack of timely administration of postoperative RT. Specifically, some patients received art up to 13 mo after RP [17 19]. Such a long deferral could introduce a potential bias in favour of art, as these patients with persisting undetectable PSA are substantially less likely to develop recurrent disease, regardless of postoperative irradiation. We attempted to avoid such bias by restricting the time window of art administration to <6 mo from RP. Finally, a main limitation of previous analyses is the inclusion of patients with a PSA at SRT up to 2.0 ng/ml [19]. Administration of SRT at low pretreatment serum PSA levels represents a major determinant for better outcomes [15]. Recent data indicate that a PSA cut-off as low as 0.33 ng/ml represents an independent prognostic factor for achieving an undetectable PSA after SRT, providing an increased probability of a long-term durable response [15]. Accordingly, a second analysis was performed in the current study by lowering the PSA cut-off for the definition of esrt to 0.3 ng/ ml. Even when undertaking such an approach, no difference in BCR-free survival was found between art and esrt. Taken together, our data suggest that for patients with pt3n0 PCa, observation and esrt given in cases of relapse is comparable to art with respect to BCR, provided that RT is administered in a timely fashion. This is in line with existing recommendations by the European Association of Urology guidelines, which support either art or esrt given at a PSA <0.5 ng/ml in the presence of high-risk PCa following surgery [23]. However, further studies are needed to

482 [(Fig._6)TD$FIG] EUROPEAN UROLOGY 62 (2012) 472 487 Fig. 6 Kaplan-Meier estimates of biochemical recurrence free survival in patients receiving adjuvant radiation therapy (art) versus observation (obs) and early salvage radiation therapy (esrt; prostate-specific antigen at salvage radiation therapy =0.3 ng/ml) in cases of relapse in patients with (A) pt3a disease, (B) pt3b disease, (C) negative surgical margins, or (D) positive surgical margins. BCR = biochemical recurrence. confirm these findings in patients with multiple risk factors and according to the number and extent of PSM. From a clinical perspective, our results are highly pertinent and their implications timely. Indeed, approximately 50% of patients who were randomised to the observational arm in previous prospective randomised trials never recurred [3,6,20]. If art was administered in a blanket fashion following RP, such patients would have been overtreated, thus being unnecessarily exposed to treatment-related toxicity. Therefore, the potential restriction of SRT, provided that it is delivered at a PSA 0.5 ng/ml, to only patients recurring might potentially reduce

[(Fig._6. (Continued).)TD$FIG] EUROPEAN UROLOGY 62 (2012) 472 487 483 Fig. 6. (Continued ). overtreatment and treatment-related complications. Moreover, delaying RT might be associated with significant advantages with respect to postoperative functional outcomes. Since postoperative sexual function and urinary continence recovery occurs mostly within the first 12 mo following surgery, immediate RT may potentially have a negative impact on such recovery [24,25]. Recently, results from a randomised trial addressed such a hypothesis [26]. Although no difference with respect to postoperative erectile dysfunction (ED) was recorded between patients treated with RP alone and those receiving RP with art, it has to be noted that most patients (93% for both groups) were already affected by ED before treatment. Thus, despite the prospective, randomised design of the trial, such high

484 [(Fig._7)TD$FIG] EUROPEAN UROLOGY 62 (2012) 472 487 Fig. 7 Kaplan-Meier estimates of biochemical recurrence free survival in patients receiving adjuvant radiation therapy (art) versus observation (obs) and early salvage radiation therapy (esrt; prostate-specific antigen at salvage radiation therapy =0.3 ng/ml) in cases of relapse in patients with (A) pt3a disease and negative surgical margins (NSM), (B) pt3a disease and positive surgical margins, and (C) pt3b disease and NSM. BCR = biochemical recurrence. prevalence of baseline erectile function impairment may limit the validity of these results. Moreover, patients included in the trial were treated between 1991 and 1997. As such, they may not be representative of contemporary PCa patients treated with RP. Despite its strengths, our study is not devoid of limitations. First, data on treatment-related toxicity profiles between the two examined groups of patients were not available in the current databases. In this regard, we were unable to estimate treatment-related complications. Second,

[(Fig._7)TD$FIG] EUROPEAN UROLOGY 62 (2012) 472 487 485 Fig. 7. (Continued ). Table 4 Multivariable Cox regression analyses predicting biochemical recurrence in patients treated with radical prostatectomy for pt3n0, R0 R1 prostate cancer Variable Multivariable analysis OR (95% CI) Age at surgery 0.99 (0.95 1.03) 0.7 Postoperative management: art 1.0 (ref.) Observation and esrt * in case of relapse 0.87 (0.62 1.24) 0.5 Year of surgery: 1991 1998 1.0 (ref.) 1999 2000 1.05 (0.69 1.59) 0.8 2001 2002 1.16 (0.75 1.80) 0.5 2003 2006 1.47 (0.81 2.66) 0.2 RT dose ** 1.01 (0.93 1.07) 0.7 Preoperative PSA 1.01 (1.00 1.02) 0.06 Pathologic Gleason score <6 1.0 (ref.) 6 1.74 (0.87 3.46) 0.1 7 3.15 (1.76 5.64) <0.001 8 4.54 (2.23 9.28) <0.001 9 5.89 (2.57 13.48) <0.001 Surgical margin status: Negative 1.0 (ref.) Positive 0.97 (0.69 1.35) 0.9 Pathologic T stage pt3a 1.0 (ref.) pt3b 2.50 (1.80 3.46) <0.001 OR = odds ratio; CI = confidence interval; art = adjuvant radiation therapy; esrt = early salvage radiation therapy; RT = radiation therapy; PSA = prostate-specific antigen. * Defined as RT administered with a PSA 0.3 ng/ml. ** Analysis performed only in patients who underwent art or esrt. p the median RT dose delivered to both groups could be considered suboptimal. Indeed, a dose response effect has been demonstrated retrospectively in both the adjuvant [27,28] and salvage [29] settings. A 3% improvement in 5-yr BCR-free survival for every additional gray delivered in the postoperative setting has been postulated [28]. Moreover, a significant temporal trend towards the use of initial observation and esrt in cases of relapse was observed during the study period. Although this trend might have introduced a potential bias in the study, the effect of the year of surgery was accounted for in multivariable models. Despite this drawback, the type of RT approach did not represent a significant predictor of BCR. In addition, data on the overall population comorbidity profile as well as on PSA kinetics in the esrt group were not available in our multiinstitutional dataset. This limitation might have potentially introduced a patient-selection bias in the study. Furthermore, lack of information on clinical recurrence status within the multi-institutional database did not allow us to address differences in such an end point between the two treatment approaches. More importantly, the retrospective nature of the study may also lead to significant biases, such as the existence of unknown confounders, for which statistical modelling is unable to adjust. Accordingly, it is possible that such unmeasured confounders may have played a role in the physician decision to pursue initial observation in patients with similar disease characteristics. The aforementioned concerns will hopefully be addressed within the Radiotherapy and Androgen Deprivation in Combination After Local Surgery (RADICALS) trial conducted by the Medical Research Council. This prospective randomised study is designed to

486 EUROPEAN UROLOGY 62 (2012) 472 487 Table 5 Patient stratification according to pt3 substage, surgical margin status, and treatment type in the overall matched population (n = 780) Disease characteristics No. of patients (%) art, n (%) Observation, n (%) Patients not recurring within the observation group, n (%) Patients receiving esrt within the observation group due to biochemical recurrence after RP, n (%) pt3a any SM 535 (100) 261 (48.8) 274 (51.2) 158 (57.7) 116 (42.3) pt3b any SM 245 (100) 129 (52.7) 116 (48.3) 41 (35.3) 75 (64.7) NSM 297 (100) 145 (48.8) 152 (51.2) 97 (63.8) 55 (36.2) PSM 483 (100) 245 (50.7) 238 (49.3) 102 (42.9) 136 (57.1) pt3a with NSM 197 (100) 97 (49.2) 100 (50.8) 73 (73) 27 (27.0) pt3a with PSM 338 (100) 164 (48.5) 174 (51.5) 85 (48.9) 89 (51.1) pt3b with NSM 100 (100) 48 (48.0) 52 (52.0) 24 (46.2) 28 (53.8) art = adjuvant radiotherapy; esrt = early salvage radiotherapy; NSM = negative surgical margins; PSM = positive surgical margins; RP = radical prostatectomy; SM = surgical margins. determine the benefit of art compared with esrt as well as the potential role of concomitant androgen-deprivation therapy after PSA failure [30]. Until these results are available, art should be discussed with all patients who harbour pt3 PCa at RP, especially in cases of younger men with PSM [20]. In the meantime, the current report, although retrospective, is the largest study to confirm the equivalent efficacy of art versus esrt in a large cohort of patients using a matched-based approach. 5. Conclusions The current matched case-control study suggests that initial observation followed by esrt (given at a PSA 0.5 ng/ml) is comparable to art with respect to BCR-free survival in the majority of pt3pn0 PCa patients previously treated with RP. Therefore, esrt given at a low PSA level may not compromise cancer control while significantly reducing overtreatment associated with art. However, prospective randomised data still need to confirm our retrospective findings. Moreover, further studies are needed to confirm these results in patients with multiple risk factors and according to the number and extent of PSM. Author contributions: Alberto Briganti had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Van Poppel, Wiegel, Montorsi, Briganti. Acquisition of data: Wiegel, Briganti, Van Poppel, Montorsi, Hinkelbein, Cozzarini. Analysis and interpretation of data: Briganti, Bianchi, Sun, Joniau, Tombal, Budiharto, Di Muzio, Wiegel, Van Poppel, Montorsi, Karakiewicz. Drafting of the manuscript: Briganti, Bianchi, Sun, Van Poppel, Haustermans. Critical revision of the manuscript for important intellectual content: Van Poppel, Wiegel, Montorsi, Briganti, Karakiewicz. Statistical analysis: Briganti, Bianchi, Sun. Obtaining funding: None. Administrative, technical, or material support: None. Supervision: Van Poppel, Wiegel, Montorsi, Briganti, Karakiewicz. Other (specify): None. Financial disclosures: Alberto Briganti certifies that all conflicts of interest, including specific financial interests and relationships and affiliations relevant to the subject matter or materials discussed in the manuscript (eg, employment/affiliation, grants or funding, consultancies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: None. Funding/Support and role of the sponsor: None. References [1] Karakiewicz PI, Eastham JA, Graefen M, et al. Prognostic impact of positive surgical margins in surgically treated prostate cancer: multi-institutional assessment of 5831 patients. Urology 2005;66: 1245 50. [2] Swanson GP, Riggs M, Hermans M. Pathologic findings at radical prostatectomy: risk factors for failure and death. Urol Oncol 2007; 25:110 4. [3] Bolla M, van Poppel H, Collette L, et al., European Organization for Research and Treatment of Cancer. Postoperative radiotherapy after radical prostatectomy: a randomised controlled trial (EORTC trial 22911). Lancet 2005;366:572 8. [4] Thompson IM, Tangen CM, Paradelo J, et al. Adjuvant radiotherapy for pathologically advanced prostate cancer: a randomized clinical trial. JAMA 2006;296:2329 35. [5] Thompson IM, Tangen CM, Paradelo J, et al. Adjuvant radiotherapy for pathological T3N0M0 prostate cancer significantly reduces risk of metastases and improves survival: long-term followup of a randomized clinical trial. J Urol 2009;181:956 62. [6] Wiegel T, Bottke D, Steiner U, et al. Phase III postoperative adjuvant radiotherapy after radical prostatectomy compared with radical prostatectomy alone in pt3 prostate cancer with postoperative undetectable prostate-specific antigen: ARO 96-02/AUO AP 09/ 95. J Clin Oncol 2009;27:2924 30. [7] Bianco Jr FJ, Scardino PT, Eastham JA. Radical prostatectomy: longterm cancer control and recovery of sexual and urinary function ( trifecta ). Urology 2005;66:83 94. [8] Hull GW, Rabbani F, Abbas F, Wheeler TM, Kattan MW, Scardino PT. Cancer control with radical prostatectomy alone in 1,000 consecutive patients. J Urol 2002;167:528 34. [9] Cotter SE, Chen MH, Moul JW, et al. Salvage radiation in men after prostate-specific antigen failure and the risk of death. Cancer 2011;117:3925 32. [10] Katz MS, Zelefsky MJ, Venkatraman ES, Fuks Z, Hummer A, Leibel SA. Predictors of biochemical outcome with salvage conformal radiotherapy after radical prostatectomy for prostate cancer. J Clin Oncol 2003;21:483 9.

EUROPEAN UROLOGY 62 (2012) 472 487 487 [11] Stephenson AJ, Scardino PT, Kattan MW, et al. Predicting the outcome of salvage radiation therapy for recurrent prostate cancer after radical prostatectomy. J Clin Oncol 2007;25:2035 41. [12] Stephenson AJ, Shariat SF, Zelefsky MJ, et al. Salvage radiotherapy for recurrent prostate cancer after radical prostatectomy. JAMA 2004;291:1325 32. [13] Trock BJ, Han M, Freedland SJ, et al. Prostate cancer specific survival following salvage radiotherapy vs observation in men with biochemical recurrence after radical prostatectomy. JAMA 2008;299: 2760 9. [14] Vanuytsel L, Janssens G, Van Poppel H, Rijnders A, Baert L. Radiotherapy for PSA recurrence after radical prostatectomy. Eur Urol 2001;39:425 9. [15] Wiegel T, Lohm G, Bottke D, et al. Achieving an undetectable PSA after radiotherapy for biochemical progression after radical prostatectomy is an independent predictor of biochemical outcome results of a retrospective study. Int J Radiat Oncol Biol Phys 2009;73:1009 16. [16] Stukel TA, Fisher ES, Wennberg DE, Alter DA, Gottlieb DJ, Vermeulen MJ. Analysis of observational studies in the presence of treatment selection bias. JAMA 2007;297:278 85. [17] Trabulsi EJ, Valicenti RK, Hanlon AL, et al. A multi-institutional matched-control analysis of adjuvant and salvage postoperative radiation therapy for pt3-4n0 prostate cancer. Urology 2008;72: 1298 302. [18] Ost P, De Troyer B, Fonteyne V, Oosterlinck W, De Meerleer G. A matched control analysis of adjuvant and salvage high-dose postoperative intensity-modulated radiotherapy for prostate cancer. Int J Radiat Oncol Biol Phys 2011;80:1316 22. [19] Budiharto T, Perneel C, Haustermans K, et al. A multi-institutional analysis comparing adjuvant and salvage radiation therapy for high-risk prostate cancer patients with undetectable PSA after prostatectomy. Radiother Oncol 2010;97:474 9. [20] Van der Kwast TH, Bolla M, Van Poppel H. et al., EORTC 22911. Identification of patients with prostate cancer who benefit from immediate postoperative radiotherapy: EORTC 22911. J Clin Oncol 2007;25:4178 86. [21] Collette L, van Poppel H, Bolla M, et al. Patients at high risk of progression after radical prostatectomy: do they all benefit from immediate post-operative irradiation? (EORTC trial 22911) Eur J Cancer 2005;41:2662 72. [22] Swanson GP, Goldman B, Tangen CM, et al. The prognostic impact of seminal vesicle involvement found at prostatectomy and the effects of adjuvant radiation: data from Southwest Oncology Group 8794. J Urol 2008;180:2453 7. [23] Mottet N, Bellmut J, Bolla M, et al., EAU guidelines on prostate cancer. Part II: treatment of advanced, relapsing, and castrationresistant prostate cancer. Eur Urol 2011;59:572 83. [24] Ficarra V, Novara G, Artibani W, et al. Retropubic, laparoscopic, and robot-assisted radical prostatectomy: a systematic review and cumulative analysis of comparative studies. Eur Urol 2009;55:1037 63. [25] Mulhall JP, Bella AJ, Briganti A, McCullough A, Brock G. Erectile function rehabilitation in the radical prostatectomy patient. J Sex Med 2010;7:1687 98. [26] Moinpour CM, Hayden KA, Unger JM, et al., Southwest Oncology Group. Health-related quality of life results in pathologic stage C prostate cancer from a Southwest Oncology Group trial comparing radical prostatectomy alone with radical prostatectomy plus radiation therapy. J Clin Oncol 2008;26:112 20. [27] Cozzarini C, Montorsi F, Fiorino C, et al. Need for high radiation dose (70 Gy) in early postoperative irradiation after radical prostatectomy: a single-institution analysis of 334 high-risk, node-negative patients. Int J Radiat Oncol Biol Phys 2009;75:966 74. [28] King CR, Kapp DS. Radiotherapy after prostatectomy: is the evidence for dose escalation out there? Int J Radiat Oncol Biol Phys 2008;71:346 50. [29] King CR, Spiotto MT. Improved outcomes with higher doses for salvage radiotherapy after prostatectomy. Int J Radiat Oncol Biol Phys 2008;71:23 7. [30] Parker C, Sydes MR, Catton C, et al., RADICALS Trial Management Group. Radiotherapy and Androgen Deprivation in Combination After Local Surgery (RADICALS): a new Medical Research Council/ National Cancer Institute of Canada phase III trial of adjuvant treatment after radical prostatectomy. BJU Int 2007;99:1376 9.