The Role of Adjuvant Radiotherapy in Pathologically Lymph Node-Positive Prostate Cancer

Transcription

1 The Role of Adjuvant Radiotherapy in Pathologically Lymph Node-Positive Prostate Cancer Naresh Jegadeesh, MD 1 ; Yuan Liu, PhD 2,3,4 ; Chao Zhang, PhD 2,3,4 ; Jim Zhong, MD 1,2 ; Richard J. Cassidy, MD 1,2 ; Theresa Gillespie, PhD 2,5 ; Omer Kucuk, MD 2,6 ; Peter Rossi, MD 1,2 ; Viraj A. Master, MD, PhD 2,6 ; Mehrdad Alemozaffar, MD 2,7 ; and Ashesh B. Jani, MD 1,2 BACKGROUND: Postoperative management of prostate cancer with lymph node involvement (LNI) is controversial. Retrospective evidence supports the selective use of radiotherapy (RT) after extended pelvic lymph node dissection. It is unclear whether this is generalizable to practice in the United States, where extended dissection is uncommon. The authors identified patients with LNI who potentially could derive a survival benefit with adjuvant RT plus androgen-deprivation therapy (ADT). METHODS: Patients with N1M0 prostate adenocarcinoma who underwent radical prostatectomy (RP) and subsequently received ADT from 2003 through 2011 were identified from the National Cancer Database. Kaplan-Meier analyses, log-rank tests, and multivariable Cox proportional hazards regression were performed using overall survival (OS) as the primary outcome. RESULTS: In total, 906 of 2569 eligible patients (35.3%) received RT, and RT was more frequently received by patients who were diagnosed in later years, had fewer positive lymph nodes, had involved surgical margins, and were aged <65 years (all P <.05). The 5-year OS rate was 87% versus 82% in those who received RT versus those who did not (P 5.007). After propensity score matching, 826 patients remained in each cohort. RT retained an association with OS (5-year OS: 88% vs 81%; P 5.009; hazard ratio, 1.43; 95% confidence interval, ; P 5.008). No interaction was identified between the effect of RT on OS across tested strata of total lymph nodes examined, lymph node ratio, total number of positive lymph nodes, margin status, Gleason score, and prostate-specific antigen. CONCLUSIONS: RT plus ADT was associated with improved OS after RP in patients with LNI. These results may help guide therapy in the absence of randomized evidence. Cancer 2017;123: VC 2016 American Cancer Society. KEYWORDS: androgen-deprivation therapy, extended pelvic lymph node dissection, lymph node-positive prostate cancer, radical prostatectomy. INTRODUCTION Prostate adenocarcinoma (ACP) is the most commonly diagnosed noncutaneous malignancy in males within the United States. Approximately 12% of these patients, or 26,500 in total, had regional lymph node involvement (LNI) at diagnosis in This is a heterogeneous group with ultimately poor cancer-related outcomes. 2,3 Despite the prevalence of LNI, there is a dearth of high-level evidence available to guide genitourinary oncologists in its optimal management. Pathologic LNI after radical prostatectomy (RP) is a clinical scenario of increasingly relevance for several reasons. LNI has declined significantly in the era of routine prostate-specific antigen (PSA) screening. 4 With recent US Preventative Task Force recommendations against routine PSA screening, there is reasonable expectation that LNI at diagnosis may again become more prevalent. 5,6 Before routine PSA screening, early modern prostatectomy series estimated a rate of LNI nearing 30%. 7 There are evolving indications for RP, as clinical practice guidelines increasingly accept RP as a potential first-line therapy for select patients with high-risk ACP. These patients are at high risk for harboring LNI. 8 In addition, conventional imaging techniques are inadequate for definitively assessing preoperative lymph node staging. 9 Corresponding author: Naresh Jegadeesh, MD, Department of Radiation Oncology, Emory University, 1365 Clifton Road NE, Suite T104, Atlanta, GA 30322;. Fax: (404) ; njegadeesh@emory.edu 1 Department of Radiation Oncology, Emory University, Atlanta, Georgia; 2 Winship Cancer Institute, Emory University, Atlanta, Georgia; 3 Department of Biostatistics and Bioinformatics, Emory University, Atlanta, Georgia; 4 Rollins School of Public Health, Emory University, Atlanta, Georgia; 5 Department of Surgery, Emory University, Atlanta, Georgia; 6 Department of Medical Oncology, Emory University, Atlanta, Georgia; 7 Department of Urology, Emory University, Atlanta, Georgia Jegadeesh s current affiliation: John H. Stroger, Jr. Hospital of Cook County, 1969 W Ogden Ave, Chicago, IL 60612; naresh.jegadeesh@cookcountyhhs.org See related editorial on pages 387-9, this issue. The data used in the study are derived from a deidentified National Cancer Database file. The National Cancer Database is a joint project of the Commission on Cancer of the American College of Surgeons and the American Cancer Society. The American College of Surgeons and the Commission on Cancer have not verified and are not responsible for the analytic or statistical methodology used or for the conclusions drawn from these data by the investigator. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Additional supporting information may be found in the online version of this article DOI: /cncr.30373, Received: April 27, 2016; Revised: July 20, 2016; Accepted: July 25, 2016, Published online November 8, 2016 in Wiley Online Library (wileyonlinelibrary.com) 512 Cancer February 1, 2017

2 Adjuvant RT in pn-positive Prostate Ca/Jegadeesh et al Traditional dogma recognized LNI as a proxy for occult metastatic disease and thus probably not suitable for curative management. However, mounting retrospective evidence does suggest an improvement in outcomes in some men who receive aggressive locoregional therapy In the high-risk, lymph node-negative, postprostatectomy setting, multiple randomized trials have identified a benefit with the addition of adjuvant radiotherapy (RT) We identified only 1 trial that investigated postprostatectomy adjuvant ADT for patients with N1M0 disease, 17 and no trials evaluating local therapy have been conducted in this setting. Well conducted observational series have identified subgroups of men with LNI who derived a survival benefit from adjuvant RT. 12,18 However, all patients in those series underwent an extended pelvic lymph node dissection (eplnd), a procedure that is not used frequently in the United States. 19 It is unclear whether these results can be generalized to a population that typically undergoes less extensive lymph node sampling. 20 We sought to compare survival outcomes of patients who had pathologic LNI after prostatectomy and received either guideline-supported adjuvant ADT alone versus adjuvant RT plus ADT using a large national cancer database. In addition, we sought to identify specific clinicopathologic subgroups that may derive preferential benefit from the addition of RT. MATERIALS AND METHODS The National Cancer Database (NCDB) is a large, prospectively acquired database that draws on information gathered from Commission on Cancer-accredited cancer centers nationwide. It currently captures 70% of all newly diagnosed malignancies in the United States. The data set includes detailed information on patient characteristics, disease parameters, treatment information, and outcomes. The treatment information contains data not available in other large national databases, including detailed RT information regarding treatment site, treatment source, treatment intent, radiation dose (in grays), and receipt of ADT (with ADT referring to any form of hormone suppression). The NCDB prostate cancer database was queried for patients aged 18 years who were diagnosed between 2003 and Adjuvant RT was defined as pelvic or prostatic radiation delivered with curative intent within 6 months after prostatectomy. Inclusion criteria were as follows: adenocarcinoma histology, pathologic involvement of regional lymph nodes, receipt of RP plus subsequent adjuvant ADT, and 1 lifetime cancer or cases in which the reported tumor was the first of multiple diagnoses. Exclusion criteria were patients those encoded with the following: metastatic disease at diagnosis; patients who were diagnosed at the reporting facility but treated elsewhere; patients who received cobalt-60 teletherapy, stereotactic radiosurgery, brachytherapy, or another nonstandard approach; palliative-intent therapy; and patients with unknown survival information. Patient and disease parameters were examined and included facility type, age, race, insurance status and type, area of residence/rural or urban county, income, education level, Charlson-Deyo comorbidity score, year of diagnosis, histology, tumor grade, tumor (T) classification, regional lymph node count, lymph node ratio (the numbers of positive lymph nodes/lymph nodes examined), Gleason score from prostatectomy specimen, PSA, surgical margin status, chemotherapy, and radiation dose. Income and education level were abstracted by the NCDB based on zip code of residence and were matched to 2012 American Community Survey and 2000 US Census data, respectively. Facility type was determined using Commission on Cancer definitions. Insurance was categorized as none, private, or government (includes Medicare, Medicaid, and other government insurance). The primary endpoint of OS was defined as the number of months between surgery and either death or last follow-up if the patient remained alive. Statistical Methods Statistical analyses were conducted using SAS version 9.4 (SAS Institute, Inc, Cary, NC) and SAS macros developed by the Biostatistics and Bioinformatics Shared Resource at Winship Cancer Institute. 21 Descriptive statistics for each variable were reported. The univariate association with treatment groups (RT vs no RT) was assessed using the chi-square test for categorical covariates and analyses of variance for numerical covariates. The univariate association with OS was assessed using Cox proportional hazards models and log-rank tests. Guarantee time bias was accounted for using an extended Cox model considering RT as a time-varying variable. 22 A plot of extended Kaplan-Meier estimates was generated accordingly. 23 The multivariable models were built using a backward variable-selection procedure with an a 5.2 for removal criteria. All P values were 2 sided, and P <.05 was considered statistically significant. To further reduce treatment selection bias, a propensity score-matching method also was implemented. A logistic regression model predicting RT versus non-rt was used to calculate propensity scores in covariates that were identified as confounders in the final multivariable extended model. Patients from each Cancer February 1,

3 study cohort were matched to each other at ratio of 1:1 based on the propensity score using a greedy 5!1 digit match algorithm. 24 After matching, the balance of covariates between 2 cohorts were evaluated using the standardized differences, and a value of <.1 was considered a negligible imbalance (see Supporting Tables 1 and 2 for standardized differences before and after matching; see online supporting information). The effects were estimated in the matched sample using a Cox model with a robust variance estimator for OS. 25 RESULTS In total, 2569 patients diagnosed between 2003 and 2011 met the study entry criteria (Table 1). Of these, 906 patients (35.3%) received adjuvant RT plus ADT. The median follow-up was 53 months. The median patient age was 62 years (range, years). The majority of patients (65.2%) were aged < 65 years. Most had a Charlson-Deyo score of 0 or 1 (98.1%). The majority of patients had pathologic T3 (pt3) disease (n ; 77.5%). In addition, 55.8% of patients had 1 positive lymph node, 20% had 2 positive lymph nodes, 22.7% had 3 positive lymph nodes, and 1.5% had an unknown number of positive lymph nodes. The median number of lymph nodes involved was 1, and the median number of lymph nodes examined was 8. The median lymph node density was 20%. The majority of patients had a Gleason score from 8 to 10 (56.4%). The median time from surgery to radiation was 3.45 months (range, months), and the median radiation dose used was 67 grays. A comparison of patient, treatment, and disease characteristics by receipt of adjuvant RT is presented in Table 1. RT was more likely to be received by younger patients (P <.001); it was used more frequently in the setting of fewer involved lymph nodes (P <.001), a lymph node density 20% (P <.001), and with involved surgical margin (P <.001). There was increased use of RT over the studied period (P <.001). Patients who had private insurance or nonacademic centers were more likely to receive RT (P <.001). RT receipt did not appear to be associated with higher Gleason score or pretreatment PSA. The actuarial 5-year rate OS was 87% in patients who received RT plus ADT versus 82% in those who received ADT alone (P 5.007) (Fig. 1). The receipt of RT plus ADT was associated with an improvement in OS over ADT alone in multivariate analysis (P <.001; hazard ratio [HR], 1.50; 95% confidence interval [CI], ). Additional factors associated with poorer OS included treatment at an nonacademic or research facility, 3 lymph nodes involved, positive surgical margins, pt3-positive pathologic tumor, and a Gleason score >7 (all P <.05) (see Table 2). Neither the number of lymph nodes removed nor the pretreatment PSA level was independently associated with survival. Patients who had unknown T-classification fared significantly worse than those who had T2 tumors, and men with unknown Gleason scores performed significantly worse than those who had Gleason scores from 2 to 7. This is consistent, because the unknown values are likely a representative sampling of the entire population. A propensity score-matched analysis was performed. Eight hundred twenty-six patients remained in each cohort after matching. The groups were well balanced, and there were no statistically significant differences between groups in any of the tested variables. The actuarial 5-year OS rate was 88% in men who received RT plus ADT versus 81% in those who received ADT alone (P 5.004) (Fig. 2), with an HR of 1.43 (95% CI, ). No interaction was identified between the OS effect of adjuvant treatment type across the tested strata of total number of lymph nodes examined, Gleason score, total number of positive lymph nodes, surgical margin status, lymph node ratio, and pretreatment PSA level. This indicates that the identified benefit of RT plus ADT exists across all strata within these variables. Finally, an optimal cutoff point analysis to dichotomize lymph node ratio by OS identified a value of 27%. DISCUSSION Prostate cancer with LNI after prostatectomy is a frequently encountered clinical scenario without a well defined treatment paradigm. Historically, prostatectomy was aborted with the intraoperative discovery of LNI, because it was considered indicative of a systemic disease. 26 This tenet is carried forth in the American Joint Committee on Cancer (AJCC) staging criteria (AJCC Cancer Staging Manual, seventh edition), in which, regardless of lymph node burden, regional lymph node involvement shares the same stage grouping as distant metastatic disease. 27 However, mounting evidence suggests there is substantial heterogeneity within the N-positive/M0 population, similar to other solid tumor disease sites. Some patients, particularly those with fewer involved lymph nodes, may derive a cure from aggressive locoregional treatment strategies. 3,28,29 Indeed, the results from our study suggest that the addition of RT to ADT after prostatectomy confers a significant OS benefit. In addition, our study provides compelling evidence regarding heterogeneity within the pn-positive/m0 population. 514 Cancer February 1, 2017

4 Adjuvant RT in pn-positive Prostate Ca/Jegadeesh et al TABLE 1. Univariate Association With Radiation No. of Patients (%) Demographic or Clinical Characteristic No Radiation, N Radiation, N P Age, y < (60.8) 657 (39.2) <.001 a (72.12) 249 (27.88) Race White/other 1475 (64.38) 816 (35.62).120 Black 188 (67.63) 90 (32.37) Facility type Community cancer program 111 (53.88) 95 (46.12) <.001 a Comprehensive community cancer program 561 (55.65) 447 (44.35) Academic/research program 991 (73.14) 364 (26.86) Facility location Northeast 251 (60.05) 167 (39.95).018 a South 410 (62.4) 247 (37.6) Midwest 615 (68.03) 289 (31.97) West 387 (65.59) 203 (34.41) Primary payor Not insured 98 (74.24) 34 (25.76) <.001 a Private 954 (60.57) 621 (39.43) Medicaid/other government 63 (68.48) 29 (31.52) Medicare 548 (71.17) 222 (28.83) Median income quartiles <$30, (64.75) 98 (35.25).946 $30,000-$35, (65.12) 150 (34.88) $36,000-$45, (65.92) 243 (34.08) $46, (64.54) 372 (35.46) Charlson-Deyo score (64.42) 780 (35.58) (65.15) 115 (34.85) 2 36 (76.6) 11 (23.4) Year of diagnosis (69.05) 52 (30.95).015 a (66.67) 72 (33.33) (66.67) 80 (33.33) (71.43) 72 (28.57) (67.71) 103 (32.29) (66.45) 102 (33.55) (58.31) 138 (41.69) (61.68) 141 (38.32) (60.65) 146 (39.35) Pathologic tumor classification pt2 213 (68.71) 97 (31.29).182 pt (64.76) 702 (35.24) pt4 95 (60.13) 63 (39.87) Unknown 65 (59.63) 44 (40.37) No. of positive lymph nodes (60.95) 560 (39.05) <.001 a (68.29) 163 (31.71) (71.01) 169 (28.99) Unknown 24 (63.16) 14 (36.84) Lymph node ratio <20% 733 (67.25) 357 (32.75).009 a 20% 887 (63.49) 510 (36.51) Unknown 43 (52.44) 39 (47.56) Gleason score (63.72) 328 (36.28) (65.12) 505 (34.88) Unknown 144 (66.36) 73 (33.64) PSA, ng/ml < (63.41) 412 (36.59) to < (65) 203 (35) (63.93) 211 (36.07) Unknown 198 (71.22) 80 (28.78) Surgical margin status Negative 778 (71.77) 306 (28.23) <.001 a Cancer February 1,

5 TABLE 1. Continued No. of Patients (%) Demographic or Clinical Characteristic No Radiation, N Radiation, N P Positive 840 (59.53) 571 (40.47) Unknown 45 (60.81) 29 (39.19) Regional treatment modality No radiation treatment 1663 (100) 0 (0) <.001 a IMRT 0 (0) 403 (100) 3DCRT 0 (0) 56 (100) Unknown 0 (0) 447 (100) Chemotherapy No 1512 (64.12) 846 (35.88).014 a Yes 101 (76.52) 31 (23.48) Unknown 50 (63.29) 29 (36.71) Regional lymph nodes positive, Mean [median] 2.26 [1] 2 [2].882 Regional lymph nodes examined, Mean [median] [9] 9.36 [7] <.001 a Abbreviations: 3DCRT, 3-dimensional conformal radiation therapy; IMRT, intensity modulated radiation therapy; PSA, prostate-specific antigen. a This P value indicates a statistically significant difference. There is a paucity of level 1 evidence regarding the management of surgically treated pn-positive/m0 patients. Messing et al conducted the only randomized trial in this domain and identified a survival benefit with immediate versus delayed ADT after prostatectomy. 17 Data regarding the role of adjuvant radiotherapy are further limited. At least one-third of pn-positive/m0 patients recur within the pelvis after surgery and thus may benefit from escalated local therapy. 30,31 Three randomized studies demonstrated improved cancer-related outcomes with the addition of adjuvant radiation in the locally advanced but lymph node-negative postprostatectomy setting. It is challenging to extrapolate the results of these studies to the lymph node-positive setting. Several retrospective series have examined the role adjuvant RT in this population. Pooled outcomes analyses from 2 high-volume centers provide much of our understanding of this disease. 12,18,29 In the most recent update, adjuvant RT plus ADT improved OS compared with adjuvant ADT alone. Two risk groups men with 2 involved lymph nodes plus Gleason scores 7 and those with either positive surgical margins or pt3b/pt4 tumors and 3 or 4 involved lymph nodes drove the survival benefit. All patients in that series underwent eplnd. Both the utility and therapeutic value of eplnd is a subject of ongoing debate. 32 Nonetheless, this practice is used only in a minority of patients treated in the United States. 19,33 The median number of lymph nodes removed in our study was only 8 compared with 14 in the study previously discussed study by Abdollah et al. 32 In our series, survival did not appear to be improved by more extensive lymph node dissection. However, with a limited dissection, the total number of involved lymph nodes may not Figure 1. These are Kaplan-Meier (KM) survival (surv.) estimates for all patients by treatment. be truly reflective of disease burden and thus, alone, may not be able to guide decisions regarding prognosis and adjuvant therapy. Previous series have identified a lymph node density between 20% and 30% for stratifying patients at high risk for recurrence. 34,35 Our analysis identified 27% as the most meaningful cutoff point to dichotomize OS. In addition, those who had 3 involved lymph nodes had a significantly worsened OS. However, we did not identify any differential treatment effect of adjuvant RT plus ADT stratified according to the number of positive lymph nodes, the total number of lymph nodes 516 Cancer February 1, 2017

6 TABLE 2. Univariate Association With Overall Survival Demographic or Clinical Characteristic No. HR (95% CI) P Radiation No ( ).007 a Yes 906 Age, y ( ).563 < Race White/other ( ).036 a Black 278 Facility type Community cancer program/ ( ).031 a Comprehensive community cancer program ( ).013 a Academic/research program 1355 Primary payor Private ( ).280 Medicaid/other government ( ).054 Medicare ( ).146 Not insured/unknown 132 Charlson-Deyo score ( ).026 a ( ) Year of diagnosis ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) Pathologic tumor classification pt ( ) <.001 a pt ( ) <.001 a Unknown ( ) <.001 a pt2 310 No. of positive lymph nodes known ( ) ( ) <.001 a ( ) Lymph node ratio Unknown ( ) % ( ) <.001 a <20% 1090 Gleason score Unknown ( ).014 a ( ) <.001 a PSA, ng/ml < ( ) to < ( ) Unknown ( ) Surgical margin status Unknown ( ).052 Positive ( ) <.001 a Negative 1084 Chemotherapy No ( ).971 Unknown ( ).679 Yes 132 Regional lymph nodes positive 1.00 ( ).659 Regional lymph nodes examined 1.00 ( ).653 Radiation dose, Gy 1.00 ( ).357 Radiation days after surgery 1.00 ( ).399 Abbreviations: 3DCRT, 3-dimensional conformal radiation therapy; CI, confidence interval; Gy, grays; HR, hazard ratio; IMRT, intensity-modulated radiation therapy; PSA, prostate-specific antigen. a This P value indicates a statistically significant difference.

7 TABLE 3. Multivariate Association With Overall Survival Demographic or Clinical Characteristic HR (95% CI) P Figure 2. These are Kaplan-Meier (KM) survival (surv.) estimates for all patients by treatment (propensity-matched comparison). examined, or the lymph node ratio. This suggests that, regardless of the overall lymph node burden, adjuvant RT plus ADT improves survival. Three Surveillance, Epidemiology, and End Results population-based analyses have been conducted, and none has identified a benefit from adjuvant RT. 20,36,37 Notably, receipt of ADT was not reported in 2 series and was used only in a minority of patients in the other series. In an exploratory analysis for our study (Supporting Fig. 1), we investigated survival in all patients (n ), not only those who received ADT. In that cohort, no OS benefit was observed from adjuvant RT: the 5-year OS rate was 87.6% with RT versus 85% with no RT (P 5.081). When the analysis was restricted to those who received adjuvant ADT, there was a 5% OS survival benefit at 5 years (P 5.007). In an unplanned subset analysis of Radiation Therapy Oncology Group trial 8531, patients with N1M0 disease had improved OS with RT plus ADT versus RT alone. 38 Roughly 25% of those patients were treated after prostatectomy. We speculate that micrometastatic control may potentiate the survival benefit conferred by improved local control. This pattern is observed in multiple other solid tumor subsites. Our study is clinically impactful for several reasons. We identified a survival benefit with the addition of adjuvant RT plus ADT for patients with pn-positive/m0 prostate cancer. In the propensity-matched analysis, which allows the estimation of a causal effect, the benefit Radiation No 1.50 ( ) <.001 a Yes Facility type Community cancer program/other 1.52 ( ).029 a Comprehensive community 1.38 ( ).006 a cancer program Academic/research program Charlson-Deyo score ( ).026 a ( ) No. of positive regional lymph nodes Unknown 2.54 ( ) ( ) <.001 a ( ) Surgical margin status Unknown 1.30 ( ).339 Positive 1.38 ( ).006 a Negative Pathologic tumor classification pt ( ).004 a pt ( ) <.001 a Unknown 2.88 ( ) <.001 a pt2 Gleason score ( ) <.001 a Unknown 1.61 ( ).010 a 2-7 Abbreviations: 3DCRT, 3-dimensional conformal radiation therapy; CI, confidence interval; HR, hazard ratio; IMRT, intensity modulated radiation therapy. a This P value indicates a statistically significant difference. of RT plus ADT persisted. Patients who did not receive adjuvant RT plus ADT had a 1.4-fold risk of death. This is the largest series to date examining outcomes in this cohort. Data from this study come from a large registry representative of patients treated across the United States. The study was not restricted to those who were treated at high-volume centers or who would meet strict eligibility criteria for clinical trials and thus is generalizable. These factors are significant given the incongruity of surgical and radiation technique and practice in managing this disease. Finally, we provide further evidence of the heterogeneity within men with pn-positive/m0 disease. Patient outcomes related not only to lymph node status but also to clinicopathologic factors of the primary tumor. Outcomes were independently poorer in those who had positive surgical margins, higher Gleason scores, and advanced T-classification. Adjuvant RT plus ADT appears to improve outcomes regardless of the number of involved lymph nodes, the number of lymph nodes 518 Cancer February 1, 2017

8 Adjuvant RT in pn-positive Prostate Ca/Jegadeesh et al dissected, the lymph node ratio, the Gleason score, the surgical margin status, and the pretreatment PSA level. Along with the retrospective nature of this study, our analysis has several meaningful limitations. Similar to other cancer registry-based analyses, there are potential uncaptured confounders that may have led to an imbalance between the comparative groups despite the propensity-matched analysis. We lacked information regarding pretreatment staging, postoperative PSA levels, duration of ADT, and architecture of the involved lymph nodes. The presence of extracapsular extension, lymph node cancer volume, and lymph node size are all potential predictors of patient outcome. 39 Postoperative PSA is a variable missing from all published series examining the question of adjuvant RT. This is unfortunate, because this factor is used clinically to help ascertain undissected lymph node burden and extrapelvic disease and thus to guide the selection of adjuvant therapy. Information regarding RT treatment volume is coded inconsistently within the NCDB; therefore, we did not include it in our analysis. Finally, toxicity information is not captured by the NCDB, so we cannot comment on differences in this outcome between treatment groups. In conclusion, the current results demonstrate that combined adjuvant RT plus ADT is associated with a substantial improvement OS compared with ADT alone in patients who have pn-positive/m0 prostate cancer after prostatectomy. RT plus ADT improves OS across all strata in tested variables, including PSA, lymph Gleason score, surgical margin status, lymph node ratio, the number of lymph nodes examined, and the number involved lymph nodes. Given the growing body of retrospective evidence, there is a pressing need for a prospective, randomized trial evaluating locoregional treatment for lymph node-positive prostate cancer. In the absence of level I evidence, the current results can help guide management within this patient population. FUNDING SUPPORT Research reported in this publication was supported in part by the Biostatistics and Bioinformatics Shared Resource of Winship Cancer Institute of Emory University and by the National Institutes of Health/National Cancer Institute (award P30CA138292). CONFLICT OF INTEREST DISCLOSURES The authors made no disclosures. AUTHOR CONTRIBUTIONS Naresh Jegadeesh: Planning and conducting the study, reporting, and overall guarantor. Yuan Liu: Planning and conducting the study. Chao Zhang: Planning and conducting the study. Jim Zhong: Reporting. Richard J. Cassidy: Reporting. Theresa Gillespie: Reporting. Omer Kucuk: Reporting. Peter Rossi: Reporting. Viraj A. Master: Reporting. Mehrdad Alemozaffar: Reporting. Ashesh B. Jani: Planning and conducting the study, reporting, and overall guarantor. REFERENCES 1. Baker BR, Mohiuddin JJ, Chen RC. The role of radiotherapy in node-positive prostate cancer. Oncology (Williston Park). 2015;29: , Cheng L, Zincke H, Blute ML, Bergstralh EJ, Scherer B, Bostwick DG. Risk of prostate carcinoma death in patients with lymph node metastasis. Cancer. 2001;91: Boorjian SA, Thompson RH, Siddiqui S, et al. Long-term outcome after radical prostatectomy for patients with lymph node positive prostate cancer in the prostate specific antigen era. J Urol. 2007;178: ; discussion Cooperberg MR, Moul JW, Carroll PR. The changing face of prostate cancer. J Clin Oncol. 2005;23: Schroder FH, Hugosson J, Carlsson S, et al. Screening for prostate cancer decreases the risk of developing metastatic disease: findings from the European Randomized Study of Screening for Prostate Cancer (ERSPC). Eur Urol. 2012;62: Drazer MW, Huo D, Eggener SE. National prostate cancer screening rates after the 2012 US Preventive Services Task Force recommendation discouraging prostate-specific antigen-based screening. J Clin Oncol. 2015;33: Bundrick WS, Culkin DJ, Mata JA, Zitman RI, Venable DD. Evaluation of the current incidence of nodal metastasis from prostate cancer. J Surg Oncol. 1993;52: Heidenreich A, Bastian PJ, Bellmunt J, et al. EAU guidelines on prostate cancer. pt 1: screening, diagnosis, and local treatment with curative intent-update Eur Urol. 2014;65: Briganti A, Abdollah F, Nini A, et al. Performance characteristics of computed tomography in detecting lymph node metastases in contemporary patients with prostate cancer treated with extended pelvic lymph node dissection. Eur Urol. 2012;61: Rusthoven CG, Carlson JA, Waxweiler TV, et al. The impact of definitive local therapy for lymph node-positive prostate cancer: a population-based study. Int J Radiat Oncol Biol Phys. 2014;88: Lin CC, Gray PJ, Jemal A, Efstathiou JA. Androgen deprivation with or without radiation therapy for clinically node-positive prostate cancer [serial online]. J Natl Cancer Inst. 2015;107. pii: djv Abdollah F, Karnes RJ, Suardi N, et al. Impact of adjuvant radiotherapy on survival of patients with node-positive prostate cancer. J Clin Oncol. 2014;32: Zagars GK, Pollack A, von Eschenbach AC. Addition of radiation therapy to androgen ablation improves outcome for subclinically node-positive prostate cancer. Urology. 2001;58: 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: Wiegel T, Bartkowiak D, Bottke D, et al. Prostate-specific antigen persistence after radical prostatectomy as a predictive factor of clinical relapse-free survival and overall survival: 10-year data of the ARO trial. Int J Radiat Oncol Biol Phys. 2015;91: Bolla M, van Poppel H, Collette L, et al. Postoperative radiotherapy after radical prostatectomy: a randomised controlled trial (EORTC trial 22911). Lancet. 2005;366: Messing EM, Manola J, Yao J, et al. Immediate versus deferred androgen deprivation treatment in patients with node-positive prostate cancer after radical prostatectomy and pelvic lymphadenectomy. Lancet Oncol. 2006;7: Briganti A, Karnes RJ, Da Pozzo LF, et al. Combination of adjuvant hormonal and radiation therapy significantly prolongs survival of Cancer February 1,

9 patients with pt2-4 pn 1 prostate cancer: results of a matched analysis. Eur Urol. 2011;59: Wang EH, Yu JB, Gross CP, et al. Variation in pelvic lymph node dissection among patients undergoing radical prostatectomy by hospital characteristics and surgical approach: results from the National Cancer Database. J Urol. 2015;193: Rusthoven CG, Carlson JA, Kavanagh BD. External validation of the benefit of adjuvant radiotherapy for pathologic N1M0 prostate cancer. J Clin Oncol. 2015;33: Nickleach D, Liu Y, Shrewsberry A, Ogan K, Kim S, Wang Z. SAS macros to conduct common biostatistical analyses and generate reports [paper PO-5]. Paper presented at: Proceedings of the South- East SAS Users Group; October 20-23, 2013; St Pete Beach, Florida. 22. Giobbie-Hurder A, Gelber RD, Regan MM. Challenges of guarantee-time bias. J Clin Oncol. 2013;31: Snapinn SM, Jiang Q, Iglewicz B. Illustrating the impact of a timevarying covariate with an extended Kaplan-Meier estimator. Am Stat. 2005;59: Reeve BB, Smith AW, Arora NK, Hays RD. Reducing bias in cancer research: application of propensity score matching. Health Care Financ Rev. 2008;29: Lin DY, Wei LJ. The robust inference for the Cox proportional hazards model. J Am Stat Assoc. 1989;84: Schroder FH, Kurth KH, Fossa SD, et al. Early versus delayed endocrine treatment of pn1-3 M0 prostate cancer without local treatment of the primary tumor: results of European Organisation for the Research and Treatment of Cancer a phase III study. J Urol. 2004;172: Greene FL, Page DL, Fleming ID, et al, eds. AJCC Cancer Staging Manual, 6th ed. New York: Springer-Verlag; Briganti A, Karnes JR, Da Pozzo LF, et al. Two positive nodes represent a significant cut-off value for cancer specific survival in patients with node positive prostate cancer. A new proposal based on a 2-institution experience on 703 consecutive N 1 patients treated with radical prostatectomy, extended pelvic lymph node dissection and adjuvant therapy. Eur Urol. 2009;55: Da Pozzo LF, Cozzarini C, Briganti A, et al. Long-term follow-up of patients with prostate cancer and nodal metastases treated by pelvic lymphadenectomy and radical prostatectomy: the positive impact of adjuvant radiotherapy. Eur Urol. 2009;55: Moschini M, Sharma V, Zattoni F, et al. Natural history of clinical recurrence patterns of lymph node-positive prostate cancer after radical prostatectomy. Eur Urol. 2016;69: Nini A, Gandaglia G, Fossati N, et al. Patterns of clinical recurrence of node-positive prostate cancer and impact on long-term survival. Eur Urol. 2015;68: Abdollah F, Gandaglia G, Suardi N, et al. More extensive pelvic lymph node dissection improves survival in patients with nodepositive prostate cancer. Eur Urol. 2015;67: Abdollah F, Sun M, Thuret R, et al. Decreasing rate and extent of lymph node staging in patients undergoing radical prostatectomy may undermine the rate of diagnosis of lymph node metastases in prostate cancer. Eur Urol. 2010;58: Cai T, Nesi G, Tinacci G, et al. Clinical importance of lymph node density in predicting outcome of prostate cancer patients. J Surg Res. 2011;167: Passoni NM, Abdollah F, Suardi N, et al. Head-to-head comparison of lymph node density and number of positive lymph nodes in stratifying the outcome of patients with lymph node-positive prostate cancer submitted to radical prostatectomy and extended lymph node dissection. Urol Oncol. 2014;32:29.e21-29.e Johnstone PA, Assikis V, Goodman M, Ward KC, Riffenburgh RH, Master V. Lack of survival benefit of postoperative radiation therapy in prostate cancer patients with positive lymph nodes. Prostate Cancer Prostatic Dis. 2007;10: Kaplan JR, Kowalczyk KJ, Borza T, et al. Patterns of care and outcomes of radiotherapy for lymph node positivity after radical prostatectomy. BJU Int. 2013;111: Lawton CA, Winter K, Grignon D, Pilepich MV. Androgen suppression plus radiation versus radiation alone for patients with stage D1/pathologic node-positive adenocarcinoma of the prostate: updated results based on national prospective randomized trial Radiation Therapy Oncology Group J Clin Oncol. 2005;23: Cheng L, Bergstralh EJ, Cheville JC, et al. Cancer volume of lymph node metastasis predicts progression in prostate cancer. Am J Surg Pathol. 1998;22: Cancer February 1, 2017