Urologic Oncology: Seminars and Original Investigations 32 (2014) 303–308

Original article

Optimal timing of early versus delayed adjuvant radiotherapy following radical prostatectomy for locally advanced prostate cancer1 Keith J. Kowalczyk, M.D.a,*, Xiangmei Gu, M.S.b, Paul L. Nguyen, M.D.c, Stuart R. Lipsitz, Sc.D.b, Quoc-Dien Trinh, M.D.d,e, John H. Lynch, M.D.a, Sean P. Collins, M.D., Ph.D.f, Jim C. Hu, M.D., M.P.H.g a Department of Urology, Georgetown University Hospital, Washington, DC Center for Surgery and Public Health, Brigham and Women's Hospital, Harvard Medical School, Boston, MA c Department of Radiation Oncology, Lank Center for Genitourinary Oncology, Dana Farber Cancer Institute, Boston, MA d Vattikuti Urology Institute, Henry Ford Health System, Detroit, MI e Cancer Prognostics and Health Outcomes Unit, University of Montreal Health Center, Montreal, Canada f Department of Radiation Oncology, Georgetown University Hospital, Washington, DC g Institute of Urologic Oncology, Department of Urology, David Geffen School of Medicine at UCLA, Los Angeles, CA b

Received 16 June 2013; received in revised form 19 August 2013; accepted 1 September 2013

Abstract Objectives: Although post–radical prostatectomy (RP) adjuvant radiation therapy (ART) benefits disease that is staged as pT3 or higher, the optimal ART timing remains unknown. Our objective is to characterize the outcomes and optimal timing of early vs. delayed ART. Materials and methods: From the Surveillance, Epidemiology and End Results-Medicare data from 1995 to 2007, we identified 963 men with pT3N0 disease receiving early (o4 mo after RP, n ¼ 419) vs. delayed (4–12 mo after RP, n ¼ 544) ART after RP. Utilizing propensity score methods, we compared overall mortality, prostate cancer–specific mortality (PCSM), bone-related events (BRE), salvage hormonal therapy utilization, and intervention for urethral stricture. We then used the maximal statistic approach to determine at what time post-RP ART had the most significant effect on outcomes of interest in men with pT3N0 disease. Results: When compared with delayed ART in men with pT3 disease, early ART was associated with improved PCSM (0.47 vs. 1.02 events per 100 person-years; P ¼ 0.038) and less salvage hormonal therapy (2.88 vs. 4.59 events per 100 person-years; P ¼ 0.001). Delaying ART beyond 5 months is associated with worse PCSM (hazard ratio [HR] 2.3; P ¼ 0.020), beyond 3 months is associated with more BRE (HR 1.6; P ¼ 0.025), and beyond 4 months is associated higher rates of salvage hormonal therapy (HR 1.6; P ¼ 0.002). ART performed after 9 months was associated with fewer urethral strictures (HR 0.6; P ¼ 0.042). Conclusion: Initiating ART less than 5 months after RP for pT3 is associated with improved PCSM. Early ART is also associated with fewer BRE and less use of salvage hormonal therapy if administered earlier than 3 and 4 months after RP, respectively. However, ART administered later than 9 months after RP is associated with fewer urethral strictures. Our population-based findings complement randomized trials designed with fixed ART timing. r 2014 Elsevier Inc. All rights reserved. Keywords: Prostate cancer; Radical prostatectomy; Radiation therapy; SEER; Outcomes

1. Introduction Prostate cancer is the most commonly diagnosed solid malignancy in men and the second leading cause of deaths 1

Funding: This work was supported by a Department of Defense Prostate Cancer Physician Training Award (W81XWH-08-1-0283) presented to Dr. Hu. * Corresponding author. Tel.: þ1-202-444-4922; fax: þ1-202-444-7573. E-mail address: [email protected] (K.J. Kowalczyk).

1078-1439/$ – see front matter r 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.urolonc.2013.09.004

due to cancer in men in the United States [1]. Radical prostatectomy (RP) remains the most common treatment for clinically localized prostate cancer [2] with proven long-term cancer control compared with other treatment options [3]. However, despite recent stage migration secondary to prostate specific antigen (PSA) screening, up to one-third of men undergoing RP have locally advanced disease [4]. Although adverse tumor characteristics such as extraprostatic extension, seminal vesicle invasion, and positive

304

K.J. Kowalczyk et al. / Urologic Oncology: Seminars and Original Investigations 32 (2014) 303–308

surgical margins increase the risk of biochemical recurrence [5], level 1 evidence demonstrates that adjuvant radiation therapy (ART) lowers the risk of biochemical recurrence and [6,7] distant metastases [8], and improves overall survival [8] compared with observation. Although the Southwest Oncology Group (SWOG) 8794 trial randomized subjects to ART within 4 months of surgery vs. observation [8], the European Organisation for Research and Treatment of Cancer [6] and ARO 96-02/AUO AP 09/95 [7] trials randomized the subjects to radiation therapy (RT) within 90 and 81 days of surgery, respectively; these cut points were inherent to the study design, and the optimal timing for ART for maximal benefit remains unknown. In contrast, ART may interfere with postprostatectomy recovery and is associated with urethral stricture and decline in urinary and sexual function [9–11]. Extending the interval between RP and ART may attenuate the risk of long-term functional sequelae. Using a population-based approach, we examined the outcomes of early vs. delayed ART following RP for locally advanced prostate cancer. Additionally, we sought to determine the optimal timing for ART within 1 year of RP.

men receiving radiation therapy for spinal cord compression or palliation bone metastasis and 145 men with pathologically confirmed pelvic lymph node metastasis following RP. We also excluded 83 men who did not have continuous Medicare enrollment following diagnoses and 23 men with incomplete demographic data. We further limited our study to those men with pathologic T3 prostate cancer, thereby excluding 2928 men with pT2 or pT4 disease. In addition, we only examined men receiving postoperative pelvic radiation therapy within 1 year of RP, thus excluding another 898 men receiving RT later than 1 year after RP. Finally, 2 men who developed post-RP urethral stricture before RT were excluded, resulting in a final cohort that consisted of 963 men. Early ART was defined as radiation therapy o4 months postoperatively (n ¼ 419), whereas delayed ART was defined as radiation therapy 4 to 12 months postoperatively (n ¼ 544). A 4-month cut point was modeled to compare with randomized controlled trials (RCT) for ART vs. surveillance [8]. 2.3. Independent variables

2. Materials and methods 2.1. Data source Our study was approved by the Brigham and Women's Institutional Review Board; patient data were deidentified and the requirement for consent was waived. We analyzed the Surveillance, Epidemiology and End Results (SEER)-Medicare data that comprised a linkage of population-based cancer registries from 20 SEER areas covering approximately 28% of the U.S. population with Medicare administrative data [12]. Medicare provides health care benefits to most Americans who are 65 years or older. SEER-Medicare captures approximately 97% of incident cancer cases from participating regions and collects data such as patient demographics, tumor characteristics, and initial course of treatment [13]. 2.2. Study cohort We identified 32,419 men aged 465 years diagnosed with prostate cancer as their only malignancy from 2007 to 2009 who underwent RP from 1995 to 2009 using Current Procedural Terminology, Fourth Edition codes 55840, 55842, and 55845 for open retropubic RP; 55866 for minimally invasive RP; and 55810, 55812, and 55815 for perineal RP. Claims data were current through 2009. Men who were diagnosed at autopsy or death or had Medicare entitlement on the basis of end-stage renal disease were excluded. Men who were not enrolled in both Medicare A and B or without continuous Medicare enrollment through the study period were excluded as the outcomes may not be accurately captured. Of the 32,419 men undergoing RP during the study period, 5452 received RT following RP. We excluded 410

Age (66–69, 70–75, and 475 y) was obtained from the Medicare denominator file, while race (white, black, or other), geographic location, census tract measures of education level and median household income, population density (urban vs. rural), and marital status were obtained from the SEER database. Comorbidity was assessed using the Klabunde modification of the Charlson index based on inpatient, outpatient, and physician services the year before surgery [14]. 2.4. Dependent variables Under the SEER grading system, “well differentiated” corresponds with Gleason scores 2 to 4, “moderately differentiated” corresponds with Gleason scores 5 to 7, and “poorly differentiated” corresponds with Gleason scores 8 to 10. Gleason score 7 was moved from “moderately differentiated” to “poorly differentiated” with cases diagnosed after January 1, 2003. For this analysis, tumor grade was categorized into 2 groups based on the SEER grading system: well/moderately differentiated and poorly differentiated/unknown, similar to prior studies utilizing SEER [15]. Preoperative PSA levels were categorized within the SEER database as elevated, normal, and unknown. 2.5. Outcomes Primary outcomes assessed from the time of RP were provided in SEER and included overall mortality (OM), prostate cancer–specific mortality (PCSM), bone-related events, and use of salvage hormonal therapy (defined as androgen deprivation therapy (ADT) administered 41 y after RP, without any injections 6 months prior, so as to distinguish from hormonal therapy used in conjunction with

K.J. Kowalczyk et al. / Urologic Oncology: Seminars and Original Investigations 32 (2014) 303–308

ART). Bone-related events were determined by International Classification of Disease, Ninth Revision diagnosis codes 170.2, 170.6, 198.3, and 198.5 for bone metastasis and 733.10, 733.11-14, and 733.19 for pathologic fracture [16]. Additionally, postoperative intervention for urethral stricture was identified and compared using previously defined methods [17]. Urinary incontinence was not included as an outcome, as this is most accurately assessed with validated patient questionnaires. 2.6. Statistical analysis Demographic and tumor characteristics were compared using the χ2 test and were adjusted for using propensity score methods [18], which permit control for observed confounding factors that may influence both group assignment and outcome using a single composite measure and attempt to balance patient characteristics between groups. We used a logistic regression model to calculate the propensity (probability) of undergoing adjuvant vs. delayed ART based on all the covariates described earlier, and then weighted each subject's data based on the inverse propensity of being in 1 of the 2 treatment groups [19]. Covariate balance was checked after adjustment to ensure that there were no statistically significant differences. As the primary outcomes analyzed do not have an upper time limit and the length of follow-up varied, we compared the number of events per 100 person-years of follow-up after propensity weighting. The optimal cut points in the timing of ART for discriminating between men with and without the highest risk of failure (i.e., OM, PCSM, bonerelated events, and salvage hormonal therapy use) or intervention for urethral stricture were determined using the maximum statistic approach that determines at what point an outcome is statistically most likely to occur [20]. All statistical analyses were performed using SAS version 9.2 (SAS Institute, Cary, NC). All P values were 2-sided and considered significant at r0.05. 3. Results The sociodemographic characteristics are demonstrated in Table 1. Before propensity weighting, men were more likely to receive early ART between 1995 and 2004, whereas delayed ART was more common after 2005 (P ¼ 0.012). Although pathologic stage and grade were similar, late vs. early ART was more frequently associated with elevated preoperative PSA levels (P ¼ 0.030). Finally, concurrent ADT utilization along with ART was more common in men receiving delayed ART vs. early ART (41.1% vs. 29.3%; P o 0.001). Table 2 presents the outcomes of interest as both proportions and rates in events per 100 person-years. Overall, men receiving delayed ART had greater use of salvage hormonal therapy (24.4% vs. 17.9%; P ¼ 0.015)

305

but fewer interventions for urethral strictures (17.1% vs. 24.8%; P ¼ 0.003). There were no significant differences in OM, PCSM, bone-related events, or incontinence intervention. However, in adjusted analysis, early vs. late ART for pT3 disease was associated with improved rates of PCSM (0.47 vs. 1.02 deaths per 100 person-years; P ¼ 0.038) and less salvage hormonal therapy utilization (2.88 vs. 4.59 events per 100 person-years; P ¼ 0.001). Significant cut points for maximum ART benefit for pT3 prostate cancer are shown in Table 3. Delaying ART beyond 5 months was associated with worse PCSM (hazard ratio [HR] 2.3; P ¼ 0.020), beyond 3 months was associated with more bone-related events (HR 1.6; P ¼ 0.025), and beyond 4 months was associated with greater use of salvage hormonal therapy (HR 1.6; P ¼ 0.002). Finally, administering ART beyond 9 months was associated with fewer urethral strictures (HR 0.6; P ¼ 0.042).

4. Comment The optimal timing of ART for locally advanced prostate cancer following RP remains elusive. Initial findings from RCTs comparing immediate ART vs. observation revealed improved biochemical recurrence-free survival without overall survival differences [7,21,22]. An update of the SWOG 8794 trial in 2009 demonstrated fewer distant metastases and an 11% improvement in overall survival with ART with median follow-up of 12.6 years [8]. RCT study designs are limited in their ability to assess the optimal timing of ART because of fixed time points. Published RCTs used arbitrary cut points of 4 months [8], 90 days [21], and 81 days [7] for ART. However, the SWOG study also showed that earlier administration of radiation was associated with an increased risk of urethral stricture (17.8% vs. 9.5%) and total urinary incontinence (6.5% vs. 2.8%) [22]. Consequently, clinicians often delay starting adjuvant therapy by later than 4 months, as advocated by the SWOG trial, to allow continence recovery and to decrease the odds of radiation-related complications. However, it is unknown whether it is “safe” from a cancer control perspective to wait beyond these timeframes. Our study has several important findings. First, early vs. late ART was associated with improved PCSM for pT3 disease. To our knowledge, this is the first report of improved PCSM favoring early vs. late ART. The SWOG 8794 trial showed an improved median overall survival benefit of 1.7 years for ART vs. surveillance and estimated that 9.1 men needed treatment with ART to prevent 1 death [8]. However, prostate cancer–specific survival was not reported, and others have postulated that of the 22 deaths in the adjuvant RT arm, only 5 were due to prostate cancer [10]. Unlike the SWOG 8794 trial, we demonstrate that early ART for pT3 disease is associated with improved PCSM.

306

K.J. Kowalczyk et al. / Urologic Oncology: Seminars and Original Investigations 32 (2014) 303–308

Table 1 Baseline demographic and tumor characteristics of men receiving early vs. delayed adjuvant radiation therapy (ART) Variable

Before propensity weighting Early ART (n ¼ 420)

Year of diagnosis 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Age 66–69 70–74 75þ Charlson comorbidity index 0 1 2þ Race White Black Hispanic Asian/other Marital status Single Married Unknown Population density Metropolitan Nonmetropolitan Tumor grade Poorly/undifferentiated Well/moderately Pathologic tumor stage T3a T3b Preoperative PSA levels Elevated Normal Unknown Concurrent ADT No Yes

22 23 23 28 16 34 38 42 35 49 31 42 37

(5.2%) (5.5%) (5.5%) (6.7%) (3.8%) (8.1%) (9.1%) (10.0%) (8.3%) (11.7%) (7.4%) (10.0%) (8.8%)

P value Delayed ART (n ¼ 545)

Early ART

(3.3%) (2.8%) (2.9%) (4.2%) (4.0%) (9.5%) (7.7%) (9.4%) (7.0%) (11.4%) (12.8%) (13.4%) (11.6%)

0.012

256 (61.0%) 129 (30.7%) 35 (8.3%)

326 (59.8%) 176 (32.3%) 43 (7.9%)

0.863

336 (80.0%) 68 (16.2%) 16 (3.8%)

447 (82.0%) 84 (15.4%) 14 (2.6%)

336 24 33 27

452 31 40 22

18 17 17 22 16 38 35 40 32 48 44 50 42

(4.3%) (4%) (4%) (5.2%) (3.8%) (9.1%) (8.4%) (9.5%) (7.8%) (11.4%) (10.5%) (12%) (9.9%)

P value

Delayed ART (4.2%) (3.8%) (4%) (5.3%) (3.9%) (9%) (8.2%) (9.6%) (7.7%) (11.5%) (10.5%) (11.9%) (10.3%)

0.999

254 (60.6%) 130 (31%) 35 (8.4%)

332 (60.7%) 170 (31.2%) 44 (8.1%)

0.987

0.500

338 (80.6%) 67 (16%) 14 (3.4%)

442 (80.9%) 86 (15.7%) 18 (3.4%)

0.999

(82.9%) (5.67%) (7.3%) (4.0%)

0.391

343 23 32 21

447 31 42 26

(81.8%) (5.7%) (7.6%) (4.8%)

0.999

62 (14.8%) 348 (82.9%) 10 (2.4%)

83 (15.2%) 443 (81.3%) 19 (3.5%)

0.586

60 (14.3%) 348 (83.1%) 11 (2.6%)

80 (14.6%) 451 (82.5%) 16 (2.9%)

0.949

386 (91.9%) 34 (8.1%)

499 (91.6%) 46 (8.4%)

0.847

35 (8.4%) 383 (91.6%)

46 (8.5%) 500 (91.5%)

0.959

302 (71.9%) 118 (28.1%)

400 (73.4%) 145 (26.6%)

0.606

306 (73.2%) 112 (26.8%)

399 (72.9%) 148 (27.1%)

0.934

265 (63.1%) 155 (36.9%)

341 (62.6%) 204 (37.4%)

0.867

265 (63.2%) 154 (36.8%)

344 (63%) 202 (37%)

0.931

298 (71.0%) 23 (5.5%) 99 (23.6%)

409 (75.1%) 42 (7.7%) 94 (17.3%)

0.030

310 (74%) 27 (6.4%) 82 (19.7%)

403 (73.7%) 36 (6.6%) 108 (19.7%)

0.992

297 (70.7%) 123 (29.3%)

321 (58.9%) 224 (41.1%)

o0.001

270 (64.5%) 148 (35.5%)

350 (64.1%) 196 (35.9%)

0.902

(80.0%) (5.7%) (7.9%) (6.4%)

18 15 16 23 22 52 42 51 38 62 70 73 63

After propensity weighting

(81.9%) (5.5%) (7.6%) (4.9%)

23 21 22 29 21 49 45 53 42 63 58 65 56

ADT ¼ androgen deprivation therapy.

Second, utilizing the maximum statistic approach to determine at what point ART is most beneficial, we found that delay of ART for more than 5 months after RP for pT3 disease was associated with worse PCSM. This is a novel finding that has not been addressed in prior studies. Previous studies have not defined the optimal timing for post-RP ART and have instead utilized arbitrary cut points of 4 months in the limits of designing a randomized trial. The utilization of population-based data allows one to take a look at a large

subset of patients and determine when ART had the most benefit while limiting the side effects. Such cut points may allow maximizing functional outcomes without compromising cancer control. Similarly, a retrospective study by Trock et al. [23] found that the PCSM benefits of salvage radiation therapy occurred within 2 years after biochemical recurrence; however, the optimal timing of ART was not addressed. Third, although early ART was associated with improved PCSM, lower rates of bone-related events, and

K.J. Kowalczyk et al. / Urologic Oncology: Seminars and Original Investigations 32 (2014) 303–308

307

Table 2 Outcomes of early (o4 mo) vs. delayed (4–12 mo) adjuvant radiotherapy (ART) as raw proportion and measured in events per 100 person-years

Median follow-up years (Q1–Q3) Total person-years n (%) Overall mortality Prostate cancer–specific mortality Bone-related events Salvage hormonal therapy Urethral stricture

Early ART (n ¼ 419)

Delayed ART (n ¼ 544)

5.8 (3.8–8.7) 2691

5.1 (3.2–7.7) 3155

78 (18.6%) 11 (2.6%) 69 (16.4%) 75 (17.9%) 104 (24.8%)

83 (15.2%) 21 (3.9%) 94 (17.2%) 133 (24.4%) 93 (17.1%)

P value

0.191 0.365 0.795 0.015 0.003

Before propensity weighting Early ART (n ¼ 419) Rate per 100 person-years (number of events) Overall mortality 2.90 (78) Prostate cancer–specific mortality 0.53 (11) Bone-related events 2.82 (69) Salvage hormonal therapy 2.97 (75) Urethral stricture 4.98 (104)

Delayed ART (n ¼ 544)

P value

Early ART

Delayed ART

P value

2.63 0.92 3.29 4.62 3.44

0.511 0.126 0.123 0.005 0.014

2.54 0.47 2.75 2.88 4.74

2.83 1.02 3.33 4.59 3.65

0.479 0.038 0.240 0.001 0.090

lower rates of salvage hormonal therapy, our results suggest that delaying ART beyond 9 months postprostatectomy may lower the risk for urethral strictures. This is consistent with the SWOG 8794 results, which noted significantly more strictures with early ART [22]. Therefore, increased risk of urethral stricture should be discussed with men considering early ART and balanced with the potential benefits of improved PCSM and lower rates of bone-related events and salvage hormonal therapy. Our data suggest that for men with pT3 disease, it may be permissible to wait up to 5 months, if needed, for a patient to recover urinary and sexual function before administering ART without compromising prostate cancer–specific survival, although delaying ART beyond 9 months would lead to lower rates of urethral stricture. Fourth, men with pathologic pT3 prostate cancer experienced lower rates of bone-related events if ART was administered within 3 months of RP. Prevention of bone metastasis is a significant end point as this positively affects quality of life. Weinfurt et al. [24] noted significant declines in well-being with prostate cancer bone metastases and further declines in functional well-being following treatment for skeletal events. Fifth, early ART was associated with lower rates of salvage hormonal therapy, especially if administered within Table 3 Results of maximal statistic approach to calculate adjusted post–radical prostatectomy time cut points when adjuvant radiation therapy has the most significant benefit in men with pT3 prostate cancer

Overall mortality Prostate cancer–specific mortality Bone-related events Salvage hormonal therapy Urethral stricture

After propensity weighting

Months

Hazard ratio

P value

9 5 3 4 9

0.6 2.3 1.6 1.6 0.6

0.089 0.020 0.025 0.002 0.042

(83) (21) (94) (133) (93)

4 months of RP. These findings are consistent with those by the SWOG 8794 trial, which found lower use of ADT with early ART [8]. Avoiding hormonal therapy is beneficial to avoid the increased risk for diabetes and cardiovascular disease [25]. Additionally, hormonal therapy is associated with increased fatigue, hot flashes, and significant decreases in sexual function and overall quality of life [26,27]. Overall, our findings do suggest an overall benefit of earlier ART compared with delayed ART in men with pT3 prostate cancer. While data for postoperative PSA levels were not available to determine whether ART was given in a true adjuvant vs. salvage setting, all subjects were within 1 year of RRP and all had at least pT3 disease. Most of these men do not experience biochemical recurrence within 1 year of surgery, as shown in a recent large cohort study [28], therefore we believe these results closely reflect actual outcomes of men receiving ART with undetectable or at least low PSA values within 1 year of surgery. We do not expect these findings to displace level 1 evidence, nevertheless, we do feel they should encourage urologists recommending ART following RP to administer it within at least 5 months, with the caveat that there may be an increased risk of urethral stricture if administered within 9 months. There are no previous trials that have analyzed the exact timing of ART following RP, and we feel the data presented may be a starting point for urologists to determine the optimal timing. Our study must be interpreted in the context of the study design. First, we used an observational study design rather than a RCT and therefore we were unable to control for unmeasured confounders. Indeed, we defined radiation therapy within 12 months of RP as ART, but owing to the lack of precise PSA values, we could not rule out the possibility that some of the delayed radiation was being given as salvage therapy because of a rising PSA level. However, only 15% of men with locally advanced prostate

308

K.J. Kowalczyk et al. / Urologic Oncology: Seminars and Original Investigations 32 (2014) 303–308

cancer experienced biochemical recurrence within 12 months of RP in a recent large cohort study [28]. Although these men may be represented in our delayed ART cohort, biochemical recurrence may have been prevented in these men had they received early ART and may, in fact, support our findings that early ART leads to improved cancer control. Second, though there is underreporting of radiation therapy in some SEER registries [29], we used Medicare administrative data, which has high sensitivity in defining radiation therapy [30]. Third, our dataset lacks information on the type and dosage of radiation delivered, which may affect the outcomes. Fifth, because of heterogeneity in practice patterns, use of hormonal therapy after ART may not be an ideal surrogate for cancer control; however, hormonal therapy is associated with significant morbidity and greater costs.

[9] [10] [11]

[12]

[13]

[14]

[15]

5. Conclusions Initiating ART earlier than 5 months after RP for pT3 prostate cancer is associated with improved PCSM. Early ART is also associated with fewer bone-related events and less use of salvage hormonal therapy if administered within 3 and 4 months after RP, respectively. However, fewer urethral strictures are noted when ART is administered 9 months after RP. Our population-based findings complement level 1 evidence and provide cut points for the optimal timing of ART to improve cancer control while allowing for recovery of urinary and sexual function following RP. References [1] Jemal A, Siegel R, Xu J, Ward E. Cancer statistics, 2010 CA Cancer J Clin 2010;60:277–300. [2] Cooperberg MR, Lubeck DP, Meng MV, Mehta SS, Carroll PR. The changing face of low-risk prostate cancer: trends in clinical presentation and primary management. J Clin Oncol 2004;22:2141–9. [3] D'Amico AV, Whittington R, Malkowicz SB, et al. Biochemical outcome after radical prostatectomy, external beam radiation therapy, or interstitial radiation therapy for clinically localized prostate cancer. J Am Med Assoc 1998;280:969–74. [4] Cooperberg MR, Broering JM, Litwin MS, et al. The contemporary management of prostate cancer in the United States: lessons from the cancer of the prostate strategic urologic research endeavor (CapSURE), a national disease registry. J Urol 2004;171:1393–401. [5] Han M, Partin AW, Zahurak M, Piantadosi S, Epstein JI, Walsh PC. Biochemical (prostate specific antigen) recurrence probability following radical prostatectomy for clinically localized prostate cancer. J Urol 2003;169:517–23. [6] Van der Kwast T, Bolla M, Van Poppel H, et al. Identification of patients with prostate cancer who benefit from immediate postoperative radiotherapy: EORTC 22911. J Clin Oncol 2007;25:4178–86. [7] 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. [8] Thompson I, Tangen C, Paradelo J, et al. Adjuvant radiotherapy for pathological T3N0M0 prostate cancer significantly reduces risk of

[16]

[17]

[18] [19] [20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29] [30]

metastases and improves survival: long-term followup of a randomized clinical trial. J Urol 2009;181:956–62. Thompson I, Tangen C, Klein E. Is there a standard of care for pathologic stage T3 prostate cancer? J Clin Oncol 2009;27:2898–9. Koch MO. Should the patient with positive margins after radical prostatectomy receive adjuvant radiation? J Urol 2010;184:1838–9. Moinpour CM, Hayden KA, Unger JM, et al. 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. Potosky AL, Riley GF, Lubitz JD, Mentnech RM, Kessler LG. Potential for cancer related health services research using a linked Medicare-tumor registry database. Med Care 1993;31:732–48. Zippin C, Lum D, Hankey BF. Completeness of hospital cancer case reporting from the SEER Program of the National Cancer Institute. Cancer 1995;76:2343–50. Klabunde CN, Potosky AL, Legler JM, Warren JL. Development of a comorbidity index using physician claims data. J Clin Epidemiol 2000;53:1258–67. Wright JL, Dalkin BL, True LD, et al. Positive surgical margins at radical prostatectomy predict prostate cancer specific mortality. J Urol 2010;183:2213–8. Krupski TL, Smith MR, Chan Lee W, et al. Natural history of bone complications in men with prostate carcinoma initiating androgen deprivation therapy. Cancer 2004;101:541–9. Hu JC, Gu X, Lipsitz SR, et al. Comparative effectiveness of minimally invasive vs open radical prostatectomy. J Am Med Assoc 2009;302:1557–64. Rubin DB. Estimating causal effects from large data sets using propensity scores. Ann Intern Med 1997;127(8 Part 2):757–63. Robins JM, Hernán MÁ, Brumback B. Marginal structural models and causal inference in epidemiology. Epidemiology 2000;11:550–60. Mazumdar M, Glassman JR. Categorizing a prognostic variable: review of methods, code for easy implementation and applications to decision making about cancer treatments. Stat Medi 2000;19:113–32. Bolla M, van Poppel H, Collette L, et al. Postoperative radiotherapy after radical prostatectomy: a randomised controlled trial (EORTC trial 22911). Lancet 2005;366:572–8. Thompson I, Tangen C, Paradelo J, et al. Adjuvant radiotherapy for pathologically advanced prostate cancer: a randomized clinical trial. J Am Med Assoc 2006;296:2329–35. 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. J Am Med Assoc 2008;299:2760–9. Weinfurt KP, Li Y, Castel LD, et al. The significance of skeletalrelated events for the health-related quality of life of patients with metastatic prostate cancer. Ann Oncol 2005;16:579–84. Keating NL, O'Malley AJ, Smith MR. Diabetes and cardiovascular disease during androgen deprivation therapy for prostate cancer. J Clin Oncol 2006;24:4448–56. Herr HW, O'Sullivan M. Quality of life of asymptomatic men with nonmetastatic prostate cancer on androgen deprivation therapy. J Urol 2000;163:1743–6. Potosky AL, Knopf K, Clegg LX, et al. Quality-of-life outcomes after primary androgen deprivation therapy: results from the Prostate Cancer Outcomes Study. J Clin Oncol 2001;19:3750–5. Menon M, Bhandari M, Gupta N, et al. Biochemical recurrence following robot-assisted radical prostatectomy: analysis of 1384 patients with a median 5-year follow-up. Eur Urol 2010;58:838–46. Yu JB, Lackland AFB, Many S. NCI SEER public-use data: applications and limitations in oncology research. Oncology 2009;23:288–95. Virnig BA, Warren JL, Cooper GS, Klabunde CN, Schussler N, Freeman J. Studying radiation therapy using SEER-Medicare-linked data. Med Care 2002;40:49–54, [IV].