Radiotherapy for stage I and II testicular seminomas: Secondary malignancies and survival

Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎

Original article

Radiotherapy for stage I and II testicular seminomas: Secondary malignancies and survival Hiten D. Patel, M.D., M.P.H.*, Arnav Srivastava, M.P.H., Ridwan Alam, M.P.H., Gregory A. Joice, M.D., Zeyad R. Schwen, M.D., Alice Semerjian, M.D., Mohamad E. Allaf, M.D., Phillip M. Pierorazio, M.D. Department of Urology, The James Buchanan Brady Urological Institute, Johns Hopkins University School of Medicine, Baltimore, MD Received 6 April 2017; received in revised form 26 May 2017; accepted 20 June 2017

Abstract Introduction: Testicular seminoma affects relatively young men with excellent survival outcomes. There has been increasing concern that radiotherapy (RT) leads to secondary malignant neoplasms (SMNs) and subsequent mortality. We evaluated the effect of RT on incidence of SMNs and quantified cancer-related mortality and other causes of death for patients with stage I and II testicular seminoma. Material and methods: A national sample of men (1988–2013) diagnosed with stage IA/IB/IS/IIA/IIB/IIC testicular seminomas from Surveillance, Epidemiology, and End Results were evaluated. Use of RT over time and survival curves (5/10/15-year) was stratified by stage. Log-binomial regression determined relative risk of developing SMNs. Incidence rate ratios (IRR) and age-adjusted Cox proportional hazards models compared overall, cancer-specific survival (CSS), and other cancer-specific survival. Competing-risks regression generated cumulative incidence functions. Prevalence ratios explored excess deaths owing to specific causes. Results: A total of 16,463 men were included with 9,126 (55.4%) undergoing RT with markedly decreased use for stage I seminoma in recent years (o20%) and
50% for stage IIA. RT increased risk of SMNs (relative risk ¼ 1.84 [95% CI: 1.61–2.10, P o 0.01]). Survival rates were excellent (15-year CSS for stage I [Z99%], stage IIA [98.1%], stage IIB-C [96%–97%]). RT was associated with improved CSS for stage IB and IIA, but demonstrated less benefit for stage IA (IRR ¼ 0.63 [95% CI: 0.35–1.14, P ¼ 0.10]) with worse other cancerspecific survival (IRR ¼ 1.80 [95% CI: 0.97–3.59, P ¼ 0.05]). Gastrointestinal, respiratory, urinary, and hematologic malignances accounted for 84% of SMN deaths. Conclusions: RT offers excellent CSS for men with stage I/II seminoma and an increased risk of SMN later in life. Future studies should better evaluate risk-stratification for stage IB patients. r 2017 Elsevier Inc. All rights reserved.

Keywords: Testicular cancer; Causes of death; Survival; Radiotherapy; Secondary malignancy

1. Introduction Testicular seminoma is the most radiosensitive genitourinary malignancy. It affects relatively young men with cancer-specific survival (CSS) rates reported to be 499% for stage I disease regardless of whether adjuvant radiotherapy (RT), adjuvant chemotherapy with carboplatin, or surveillance is implemented during the first course of therapy after orchiectomy [1,2]. Given the long life expectancy among survivors, there has been increasing concern Corresponding author. Tel.: þ1-618-534-4942; fax: þ1-410-955-0833. E-mail address: [email protected] (H.D. Patel). *

http://dx.doi.org/10.1016/j.urolonc.2017.06.051 1078-1439/r 2017 Elsevier Inc. All rights reserved.

regarding the effect of RT on the development of secondary malignant neoplasms (SMNs) and subsequent unintended mortality. A number of experts and guidelines endorse active surveillance as the preferred approach to managing stage I testicular seminoma with the National Comprehensive Cancer Network (NCCN) considering single-agent carboplatin and RT as acceptable alternatives [3,4]. However, a European survey noted that an equal proportion of providers offered first-line surveillance, chemotherapy, or RT as the only option to patients [5]. The reason for this may be that surveillance requires close monitoring with serial imaging and is associated with 8% to 19% relapse rates requiring

2

H.D. Patel et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎

additional therapy based on studies with variable follow-up, the vast majority of which occur within 5 years [2,6]. Balancing relapse with the morbidity of adjuvant therapy is key to informing patients. Early studies failed to show significant increases in SMNs associated with RT, possibly owing to relatively small sample sizes and the need for at least 10 to 15 years of follow-up to tabulate SMNs [7,8]. More recent evidence strongly supports a 1.5- to 2.0-fold increased risk of SMNs with RT and downstream decrease in survival due to SMNs (other cancer-specific survival [OCSS]) [9–11]. However, previous studies have not fully explored the actual causes of death for these patients and comparative survival based on receiving RT. Furthermore, no studies separately evaluated stage IA and stage IB patients, and little evidence is available for stage II seminoma despite the NCCN recommending RT as the preferred strategy for stage IIA seminoma. Therefore, we aimed to evaluate the effect of RT on incidence of SMNs and to quantify survival and actual causes of death for patients with stage I and II testicular seminoma.

receiving initial RT. Use of RT over time was stratified by stage. Log-binomial regression models were constructed to determine the relative risk of developing an SMN associated with RT. Kaplan-Meier curves assessed 5, 10, and 15year survival rates stratified within stage I and stage II seminoma. Incidence rate ratios (IRR) were calculated to evaluate comparative survival by use of RT and to estimate preventative fraction of testicular cancer deaths and attributable fraction of deaths from SMNs. Cox proportional hazards regression models were constructed to further evaluate the age-adjusted relative survival benefit or detriment of RT by stage. Competing-risks regression was implemented to generate cumulative incidence functions for CSS and OCSS. Finally, prevalence ratios were calculated to explore excess deaths owing to specific causes. Statistical analyses were performed using STATA v.12.0 (STATA Corp, College Station, TX, 2011) with two-sided alpha set to 0.05. 3. Results 3.1. Cohort

2. Methods 2.1. Patient population, variables, and outcomes With Institutional Review Board approval, we identified all men with stage I and stage II testicular seminoma from 1988 to 2013 among cancer registries captured by the Surveillance, Epidemiology, and End Results (SEER) Program. Extragonadal germ cell tumors and patients with distant metastases or diagnosis of any prior cancer were excluded. Demographic data, including age, race, and year of diagnosis, were obtained along with postorchiectomy staging data for IA, IB, IS, IIA, IIB, and IIC testicular seminoma (AJCC, seventh edition). Given potential staging limitations of SEER data, extent of disease, lymph node, and tumor marker data were used for patients before 2004 to assign the most appropriate stage rather than using the provided American Joint Committee on Cancer (AJCC) staging categories. AJCC staging assignments for patients after 2004 were also adjusted based on the T and N stage assignments. The independent variable of interest was performance of RT with the first course of therapy as defined by SEER. Patients receiving radiation after failing initial surveillance are not included in the RT group. Outcomes included development of a SMN, overall survival (OS), CSS (deaths due to testicular cancer), and OCSS (deaths due to SMNs). Actual causes of death were tabulated for deaths related to testicular cancer, SMN, or nonmalignant causes. 2.2. Statistical analysis Baseline characteristics were compared between patients receiving RT with the first course of therapy and those not

A total of 16,463 patients diagnosed with primary stage I or stage II testicular seminoma were identified, of which 9,126 (55.4%) were treated with RT during the first course of treatment. Patients receiving RT were more likely to be white, less likely to be Hispanic, and less likely to have stage IIB or IIC disease (Table 1). RT was used in approximately 80% of cases for stage I and stage IIA patients in earlier years, but dramatically decreased to o20% for stage I patients in recent years and remained stable at 50% for stage IIA patients (Fig. 1). 3.2. Risk of SMNs and survival At a mean follow-up of 99.5 months (median ¼ 89 [interquartile range: 40–145]), 661 (7.2%) patients receiving RT compared to 298 (4.1%) not receiving RT developed SMNs for an age-adjusted relative risk of 1.84 (95% CI: 1.61–2.10, P o 0.01) (Table 2). Furthermore, 15-year CSS was approximately 99% or greater for stage I testicular seminoma, 98.1% for stage IIA, and 96% to 97% for stage IIB-C (Table 3). Although differences were small, stage was able to stratify OS and CSS (Fig. 2). Among stage IA patients (RT vs. no RT), the increase in absolute proportion of deaths from SMNs (0.98% vs. 0.28%, P o 0.01) did not appear to be offset by a reduction in absolute cancer deaths (0.55% vs. 0.46%, P ¼ 0.50). Therefore, IRRs were calculated to evaluate the effect of RT accounting for contributed person-time and showed IRR of 0.63 (95% CI: 0.35–1.14, P ¼ 0.10) for CSS and IRR of 1.80 (95% CI: 0.97–3.59, P ¼ 0.05) for OCSS. These corresponded to a preventative fraction of 37.4% for testicular cancer deaths and attributable fraction of 44.5%

H.D. Patel et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎

3

Table 1 Demographics and staging for patients diagnosed with primary testicular seminoma Overall

N Age, mean Race White Black Hispanic Asian Nativea Unknown Year of diagnosis o2000 2000–2004 2005–2009 2010þ Laterality Right Left Unknown Stage IA IB IS IIA IIB IIC

No primary RT

Primary RT

P value

Value

(% or SD)

Value

(% or SD)

Value

(% or SD)

16,463 37.8

– 10.5

7,337 37.7

– 11.1

9,126 37.8

– 10

12,445 454 2,578 609 160 217

75.6% 2.8% 15.7% 3.7% 1.0% 1.3%

5,307 228 1,311 258 78 155

72.3% 3.1% 17.9% 3.5% 1.1% 2.1%

7,138 226 1,267 351 82 62

78.2% 2.5% 13.9% 3.8% 0.9% 0.7%

o0.01

2,963 4,365 4,969 4,166

18.0% 26.5% 30.2% 25.3%

630 1,221 2,342 3,144

8.6% 16.6% 31.9% 42.9%

2,333 3,144 2,627 1,022

25.6% 34.5% 28.8% 11.2%

o0.01

8,736 7,665 62

53.1% 46.6% 0.4%

3,862 3,429 46

52.6% 46.7% 0.6%

4,874 4,236 16

53.4% 46.4% 0.2%

0.5

10,443 2,700 2,027 605 365 323

63.4% 16.4% 12.3% 3.7% 2.2% 2.0%

4,615 1,176 770 281 209 286

62.9% 16.0% 10.5% 3.8% 2.8% 3.9%

5,828 1,524 1,257 324 156 37

63.9% 16.7% 13.8% 3.6% 1.7% 0.4%

o0.01

0.58

SD ¼ standard deviation. Native American or Alaskan.

a

for deaths due to SMNs for RT among stage IA patients. Stage IB and IIA patients experienced fewer events but demonstrated improved CSS (IRR ¼ 0.22 [95% CI: 0.05– 0.83], P ¼ 0.01 for IB; IRR ¼ 0.19 [95% CI: 0.02–1.00], P ¼ 0.03 for IIA), and no statistically significant difference in OCSS (IRR ¼ 0.64 [95% CI: 0.25–1.70], P ¼ 0.31 for IB; IRR ¼ 2.01 [95% CI: 0.16–105.74] for IIA).

90.0%

Ulizaon of Radiotherpy for Tescular Seminoma by Stage

Similar findings were noted with age-adjusted Cox proportional hazards models. RT was associated with improved OS and CSS, although stage-stratified analyses suggested the least benefit in CSS for stage IA patients (hazard ratio ¼ 0.72 [95% CI: 0.41–1.27, P ¼ 0.257]) and Table 2 Models evaluating risk of developing a secondary malignant neoplasm associated with radiotherapy Risk of secondary malignant neoplasm

80.0%

RR

95% CI

P value

70.0%

Low

High

1.78

1.56

2.04

o0.01

1.83

1.60

2.09

o0.01

1.84 1.04 1.51 1.15 1.35 REF 1.46 1.51

1.61 1.04 1.03 0.76 0.89 – 0.84 0.83

2.10 1.05 2.22 1.75 2.06 – 2.54 2.74

o0.01 o0.01 0.037 0.495 0.156 – 0.175 0.156

Proporon

60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0%

Year IA IS

IB IIA

Fig. 1. Use of radiotherapy after orchiectomy over time for patients diagnosed with primary testicular seminoma. (Color version of figure is available online.)

Univariable RT vs. no RT Model 1a RT vs. no RT Model 2a RT vs. no RT Age, per y Stage IA Stage IB Stage IS Stage IIA Stage IIB Stage IIC

RR ¼ relative risk. Model 1 adjusted for age; model 2 adjusted for age and stage.

a

H.D. Patel et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎

4

Table 3 Stage-stratified overall- and cancer-specific survival probabilities at 5, 10, and 15 years N

Overall Stage I Stage IA Stage IB Stage IS Stage IIA Stage IIB Stage IIC

16,463 13,143 10,443 2,700 2,027 605 365 323

OS

CSS

5y

10 y

15 y

5y

10 y

15 y

97.8 98.1 98.2 97.8 97.2 95.8 95.9 94.8

95.4 95.7 95.9 95.3 95.0 94.1 91.7 90.1

92.9 93.1 93.4 91.0 93.4 94.1 87.2 90.1

99.4 99.6 99.6 99.6 99.1 98.1 98.0 96.7

99.3 99.5 99.5 99.5 98.9 98.1 97.2 96.7

99.1 99.4 99.3 99.5 98.8 98.1 95.9 96.7

potentially greater benefit for stage IB, IS, and IIA (Table 4). Although there were few events, no differences in survival were noted for stage IIB and IIC patients. RT was not associated with worse OCSS in the overall cohort, but suggested worse OCSS for stage IA patients (hazard ratio ¼ 1.75 [95% CI: 0.94–3.23, P ¼ 0.076]). Cumulative incidence functions are shown in Fig. 3.

Fig. 2. Survival probabilities for overall survival (A) and cancer-specific survival (B) for patients with testicular seminoma stratified by stage. Scales are adjusted for relative frequency of events. (Color version of figure is available online.)

3.3. Actual causes of death A similar proportion of men in the overall cohort died from SMNs (112 [0.68%]) compared to testicular cancer (113 [0.69%]) while nonmalignant causes of death were the most common (463 [2.81%]) (Table 5; Supplementary Table). Prevalence ratios suggesting excess deaths due to SMNs were accounted for by respiratory (lung or bronchial cancer), gastrointestinal (colorectal and liver), urinary, and hematologic (lymphoma and leukemia) malignancies. Among nonmalignant causes, cardiovascular, pulmonary, and renal etiologies appeared to be increased for patients receiving RT.

4. Discussion We sought to evaluate the effect of RT on incidence of SMNs and survival by stage for patients with stage I and state II testicular seminoma. We found RT was associated with an almost 2-fold increased risk of SMNs compared to absence of RT and contributed to worse OCSS for stage IA patients. Stage IA patients also appeared to derive the least benefit from RT on OS and CSS, whereas stage IB, IS, and IIA patients may derive greater benefit. The present study reports the continued drastic reduction in use of RT in recent years, which is in accordance with established guidelines, but also to suggest a net survival benefit for RT used to treat some stage IB, IS, and IIA patients [3,4]. The reduction in use of RT is a notable finding that reveals a concurrent change in practice patterns with guideline recommendations acknowledging the costbenefits associated with treatment. A previous populationbased SEER analysis included patients with stage I testicular seminoma up to 2001 resulting in a cohort where 78% of patients received RT and were not evaluated for the risk of developing SMNs [12]. They reported standardized mortality ratios (SMR) indicating increased deaths due to SMNs for patients treated with RT (SMR ¼ 1.89 [95% CI: 1.67–2.14]) but reported the same for patients not treated with RT (SMR ¼ 1.46 [95% CI: 1.10–1.96]). No comparison was made within stage I patients based on receipt of RT. In contrast, we included patients through 2013 and report both a relative increase in risk of developing an SMN and worse OCSS for stage IA seminoma that outweighed benefits. No prior analyses, to our knowledge, have separately analyzed stage IA and IB patients or suggested a differential effect by stage. Our findings on the increased risk of developing an SMN associated with RT verify prior studies reporting a similar magnitude of effect [9–11,13]. Notably, a proportion of the comparison cohort not receiving RT may have received chemotherapy with carboplatin rather than surveillance during the initial treatment course. Although data are not available from SEER, chemotherapy may increase the risk of some SMNs but to a lesser degree than RT; in fact,

H.D. Patel et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎

5

Table 4 Survival analyses evaluating the effect of radiotherapy on overall, cancer-specific, and other cancer-specific survival RT vs. no RTa N

OS HR

Overall Stage I Stage IA Stage IB Stage IS Stage IIA Stage IIB Stage IICb

16,463 13,143 10,443 2,700 2,027 605 365 323

0.65 0.70 0.71 0.68 0.56 0.34 0.87 0.87

95% CI

CSS P value

Low

High

0.55 0.59 0.58 0.46 0.36 0.14 0.38 0.24

0.76 0.84 0.86 1.02 0.88 0.80 1.99 3.15

o0.01 o0.01 o0.01 0.061 0.011 0.014 0.735 0.834

HR

0.49 0.65 0.72 0.38 0.38 0.23 1.01 0.88

95% CI

OCSS P value

Low

High

0.34 0.39 0.41 0.11 0.16 0.05 0.27 0.11

0.72 1.09 1.27 1.39 0.93 1.11 3.86 7.01

o0.01 0.100 0.257 0.145 0.035 0.068 0.985 0.900

HR

95% CI

1.16 1.34 1.75 0.70 0.69 1.70 1.28 –

P value

Low

High

0.75 0.82 0.94 0.29 0.20 0.17 0.11 –

1.77 2.19 3.23 1.70 2.42 17.21 14.65 –

0.508 0.242 0.076 0.433 0.563 0.652 0.842 –

HR ¼ hazard ratio. Multivariable Cox proportional hazards models adjusted for age. b Only 2 deaths due to other cancers among stage 2C patients precluding meaningful analysis (both in the no RT group). a

controversy remains whether chemotherapy alone increases the risk of solid malignancies at all [9,11]. Prior data suggest that 10% to 15% of patients may have received

initial chemotherapy, but this would only lead to more conservative estimates for the effect of RT on SMNs and survival [11,14,15]. Therefore, the true risk of SMNs may be underestimated in the present cohort when also considering the expected latency period before SMNs will develop. Table 5 Deaths due to testis cancer, secondary malignancies, and nonmalignant causes for patients with primary testicular seminoma No RT

Fig. 3. Cumulative incidence of death from testicular cancer (A) and secondary malignant neoplasms (B) based on competing-risks regression for patients with stage IA testicular seminoma. (Color version of figure is available online.)

Total cohort Testis cancer death Overall SMN oral/pharynx Gastrointestinal Respiratory Urinary Lymphoma/leukemia Soft tissue/heart Skin Eye/orbit Nervous system Endocrine Myeloma Miscellaneous Nonmalignant COD Cardiac/vascular Infectious disease Accidents Suicide/homicide Gastrointestinal COPD Kidney disease Diabetes Other Not available

7,337 60 30 1 10 8 0 8 2 0 0 1 0 0 0 192 45 15 23 20 13 5 2 6 51 12

RT – 0.82% 0.41% 0.01% 0.14% 0.11% 0.00% 0.11% 0.03% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 2.62% 0.61% 0.20% 0.31% 0.27% 0.18% 0.07% 0.03% 0.08% 0.70% 0.16%

9,126 53 82 0 21 25 7 15 4 4 1 2 1 1 1 271 84 17 40 23 11 13 7 9 50 17

PR – 0.58% 0.90% 0.00% 0.23% 0.27% 0.08% 0.16% 0.04% 0.04% 0.01% 0.02% 0.01% 0.01% 0.01% 2.97% 0.92% 0.19% 0.44% 0.25% 0.12% 0.14% 0.08% 0.10% 0.55% 0.19%

0.71 2.20 0.00 1.69 2.51 NAa 1.51 1.61 NAa NAa 1.61 NAa NAa NAa 1.13 1.50 0.91 1.40 0.92 0.68 2.09 2.81 1.21 0.79 1.14

COD ¼ cause of death; COPD ¼ chronic obstructive pulmonary disease; NA ¼ not applicable; PR ¼ prevalence ratio. a No events for patients without RT for comparison.

6

H.D. Patel et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎

Owing to low prevalence, no prior studies have reported comparative cancer survival for patients with stage IIA testicular seminoma. An analysis of the National Cancer Database noted that stage IIA patients receiving RT appeared to have improved OS at a median of 65 months, but the database included no information on SMNs or causes of death [16]. One systematic review and metaanalysis of 13 studies included 607 patients receiving RT and reported similar relapse rates and OS for RT and chemotherapy [17]. We build on the evidence base and report improved OS and CSS for RT to treat stage IIA patients without a statistically significant difference in OCSS. Although a randomized trial is needed to definitively compare relapse rates of primary RT to chemotherapy for stage IIA patients, the low prevalence and excellent survival rates make it unlikely to have sufficient power to draw conclusions on comparative survival. Notably, the use of RT for patients with IS disease may be owing to suspicion of nodal disease given persistence of tumor markers after orchiectomy and potentially reflective of efficacy for stage IIA disease rather than all patients with IS disease. Despite improvements in relapse rates, RT for stage IA seminoma may no longer be a justifiable treatment option based on numerous prior studies and the similar findings reported from SEER. Although vascular, epididymal, and rete testis invasion have been posited as potential risk factors for relapse of stage I seminoma, validation studies have failed to demonstrate their utility in a risk-adapted model [2,18]. However, 1 potential reason that risk factors for relapse and long-term outcomes have been hard to validate is pooling of stage IA and IB patients with IA patients comprising most of the population. Vascular or lymphatic invasion and extension into the tunica vaginalis, spermatic cord, or scrotum may lead to differences in relapse and survival for stage IB patients compared to IA. Although it may be reasonable to further explore RT or chemotherapy for stage IB patients, the data to date support removal of adjuvant therapy as an option for stage IA patients with active surveillance as the preferred management strategy [4]. Additionally, absolute relapse rates for stage IB patients remain low, and so exploration of RT for stage IB patients may be limited to spermatic cord or scrotal involvement with surveillance still preferred for pT2 disease given the minimal prognostic implications of lymphovascular invasion. A study of SMNs from RT centers in Europe reported increased incidence of gastrointestinal and bladder malignancies for patients with stage I seminoma receiving RT with an SMR of 1.46 (95% CI: 1.30–1.65, P o 0.001) for death from SMNs [10]. A slight increase in leukemia was suggested but not significant (standard incidence ratio ¼ 1.76 [95% CI: 0.92–3.38]). An older study of patients from Europe and North American reported a similar increase in infradiaphragmatic cancers but also reported increases in cancers of the lung and pleura [9]. Our findings show that patients dying from SMNs have a similar distribution of malignancies with gastrointestinal, respiratory, urinary, and

hematologic malignancies accounting for 84% of these deaths. It is uncertain if the increase in supradiaphragmatic cancer deaths is due to potential differences in smoking rates or changes in the radiation field size over time. Additional limitations of the present study include interpretation of the data in the context of cancer registry data. However, we edited the provided SEER AJCC staging classifications based on extent of disease variables to potentially reduce inaccuracies that have been suggested for germ cell tumor data in SEER [19]. As discussed earlier, SEER does not provide data on chemotherapy. Inclusion of patients from more recent years, while allowing for evaluation of RT use over time, reduced the event rate and proportion of overall deaths. Therefore, additional diagnoses of SMNs (up to 13.5% at 20 years for patients receiving RT based on projections from the Danish Testicular Cancer database) and downstream deaths are expected with additional follow-up [11]. Additionally, data to assign appropriate staging was limited in earlier years to differentiate stage IA, IB, IS, IIA, IIB, and IIC patients leading to the decision to use 1988 as a cutoff instead of 1973 as prior SEER analyses have used [12,13]. Finally, a notable proportion of patients had missing tumor size information which did not allow us to validly assess it as a further risk factor for specific causes of death. 5. Conclusions In summary, the use of RT after orchiectomy for stage I testicular seminoma has decreased drastically in recent years and remains at approximately 50% for patients diagnosed with stage IIA testicular seminoma. RT during the first course of therapy was associated with an almost 2-fold increased risk of developing an SMN and contributed to worse OCSS for stage IA patients. Stage IA patients also appeared to derive the least benefit from RT on OS and CSS, while stage IIA patients and some with IB and IS disease may derive greater benefit. Deaths due to SMNs were largely attributed to gastrointestinal, respiratory, and hematologic malignancies. These data add to prior studies that do not support the use of RT for stage IA testicular seminoma and support active surveillance as the preferred treatment strategy. Future studies should better evaluate risk-stratification for stage IB patients. Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at http://dx.doi.org//10.1016/j. urolonc.2017.06.051.

References [1] Mead GM, Fossa SD, Oliver RT, Joffe JK, Huddart RA, Roberts JT, et al. Randomized trials in 2466 patients with stage I seminoma: patterns of relapse and follow-up. J Natl Cancer Inst 2011;103:241–9.

H.D. Patel et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎ [2] Mortensen MS, Lauritsen J, Gundgaard MG, Agerbæk M, Holm NV, Christensen IJ, et al. A nationwide cohort study of stage I seminoma patients followed on a surveillance program. Eur Urol 2014;66: 1172–8. [3] Nichols CR, Roth B, Albers P, Einhorn LH, Foster R, Daneshmand S, et al. Active surveillance is the preferred approach to clinical stage I testicular cancer. J Clin Oncol 2013;31:3490–3. [4] Motzer RJ, Jonasch E, Agarwal N, Beard C, Bhayani S, Bolger GB, et al. Testicular Cancer, Version 2.2015. J Natl Compr Canc Netw 2015;13:772–99. [5] Vossen CY, Horwich A, Daugaard G, van Poppel H, Osanto S. Patterns of care in the management of seminoma stage I: results from a European survey. BJU Int 2012;110:524–31. [6] Tandstad T, Ståhl O, Dahl O, Haugnes HS, Håkansson U, Karlsdottir Å, et al. Treatment of stage I seminoma, with one course of adjuvant carboplatin or surveillance, risk-adapted recommendations implementing patient autonomy: a report from the Swedish and Norwegian Testicular Cancer Group (SWENOTECA). Ann Oncol 2016;27: 1299–304. [7] Chao CK, Lai PP, Michalski JM, Perez CA. Secondary malignancy among seminoma patients treated with adjuvant radiation therapy. Int J Radiat Oncol Biol Phys 1995;33:831–5. [8] Horwich A, Bell J. Mortality and cancer incidence following radiotherapy for seminoma of the testis. Radiother Oncol 1994;30: 193–8. [9] Travis LB, Fosså SD, Schonfeld SJ, McMaster ML, Lynch CF, Storm H, et al. Second cancers among 40,576 testicular cancer patients: focus on long-term survivors. J Natl Cancer Inst 2005;97:1354–65. [10] Horwich A, Fossa SD, Huddart R, Dearnaley DP, Stenning S, Aresu M, et al. Second cancer risk and mortality in men treated with radiotherapy for stage I seminoma. Br J Cancer 2014;110:256–63.

7

[11] Kier MG, Hansen MK, Lauritsen J, Mortensen MS, Bandak M, Agerbaek M, et al. Second malignant neoplasms and cause of death in patients with germ cell cancer: a Danish Nationwide Cohort Study. JAMA Oncol 2016;2:1624–7. [12] Beard CJ, Travis LB, Chen MH, Arvold ND, Nguyen PL, Martin NE, et al. Outcomes in stage I testicular seminoma: a population-based study of 9193 patients. Cancer 2013;119:2771–7. [13] Lewinshtein D, Gulati R, Nelson PS, Porter CR. Incidence of second malignancies after external beam radiotherapy for clinical stage I testicular seminoma. BJU Int 2012;109:706–12. [14] Fung C, Fossa SD, Milano MT, Sahasrabudhe DM, Peterson DR, Travis LB. Cardiovascular disease mortality after chemotherapy or surgery for testicular nonseminoma: a population-based study. J Clin Oncol 2015;33:3105–15. [15] Waingankar N, Handorf E, Smaldone MC, Plimack ER, Wong Y, Haseebuddin M, et al. The use of radiation for stage I seminoma: analysis of the National Cancer Database. J Clin Oncol 2015;33 (Suppl. 7):S385. [16] Glaser SM, Vargo JA, Balasubramani GK, Beriwal S. Stage II testicular seminoma: patterns of care and survival by treatment strategy. Clin Oncol (R Coll Radiol) 2016;28:513–21. [17] Giannatempo P, Greco T, Mariani L, Nicolai N, Tana S, Farè E, et al. Radiotherapy or chemotherapy for clinical stage IIA and IIB seminoma: a systematic review and meta-analysis of patient outcomes. Ann Oncol 2015;26:657–68. [18] Chung P, Daugaard G, Tyldesley S, Atenafu EG, Panzarella T, Kollmannsberger C, et al. Evaluation of a prognostic model for risk of relapse in stage I seminoma surveillance. Cancer Med 2015;4:155–60. [19] Faber KD, Carlos MC, Cortessis VK, Daneshmand S. Validation of Surveillance, Epidemiology, and End Results TNM staging for testicular germ cell tumor. Urol Oncol 2014;32:1341–6.