Multicenter study evaluating a dual policy of postorchiectomy surveillance and selective adjuvant single-agent carboplatin for patients with clinical stage I seminoma

Original article

Annals of Oncology 14: 867–872, 2003 DOI: 10.1093/annonc/mdg241

Multicenter study evaluating a dual policy of postorchiectomy surveillance and selective adjuvant single-agent carboplatin for patients with clinical stage I seminoma J. Aparicio1*, X. García del Muro2, P. Maroto3, L. Paz-Ares4, E. Alba5, A. Sáenz6, J. Terrasa7, A. Barnadas8, D. Almenar9, J. A. Arranz10, M. Sánchez11, A. Fernández12, J. Sastre13, J. Carles14, J. Dorca15, J. Gumà16, A. L. Yuste1 & J. R. Germà2 On behalf of the Spanish Germ Cell Cancer Cooperative Group (GG)† 1

Hospital Universitario La Fe, Valencia; 2Institut Catalá d’Oncologia, Barcelona; 3Hospital de Sant Pau, Barcelona; 4Hospital 12 de Octubre, Madrid; Hospital Clínico Universitario, Málaga; 6Hospital Clínico, Zaragoza; 7Hospital Son Dureta, Mallorca; 8Hospital Germans Trias i Pujol, Badalona; 9 Hospital Doctor Peset, Valencia; 10Hospital Gregorio Marañón, Madrid; 11Hospital Donostia, San Sebastián; 12Hospital General, Albacete; 13 Hospital Clínico San Carlos, Madrid; 14Hospital del Mar, Barcelona; 15Hospital Josep Trueta, Girona; 16Hospital Universitari Sant Joan, Reus, Spain 5

Background: After decades of irradiation as standard therapy for clinical stage I testicular seminoma, alternative treatment approaches have emerged including postorchiectomy surveillance and adjuvant chemotherapy. This study was performed to assess a dual policy of surveillance and selective single-agent carboplatin (for high-risk cases) in a multicenter setting. Patients and methods: From 1994 to 1999, 203 patients with stage I seminoma were included. Sixty (29.6%) were considered poor-risk cases (i.e. with vascular invasion and/or pathological tumor stage pT2 or greater) and received two courses of adjuvant carboplatin, whereas 143 (70.4%) without risk criteria underwent close surveillance. Results: Median follow-up was 52 months (range 14–92). Relapses were observed in two (3.3%) patients treated with carboplatin and in 23 patients (16.1%) on surveillance, with a median time to recurrence of 11 months (range 3.9–39.6). All relapsing patients were rendered disease-free, mainly with cisplatin-based chemotherapy. Four patients died from tumor-unrelated causes. Actuarial 5-year overall survival was 96.7% and cause-specific survival was 100%. Five-year disease-free survival was 83.5% for patients on surveillance, and 96.6% for those receiving carboplatin. Conclusions: This dual treatment policy is feasible in a multicenter setting and preserves 70% of patients from adjuvant chemotherapy. Single-agent carboplatin is effective in reducing the relapse rate in patients with high-risk stage I seminoma. A better definition of local risk features would probably improve patient selection, thus minimizing the incidence of recurrences on surveillance. Key words: adjuvant carboplatin, prognostic factors, stage I seminoma, surveillance

Introduction Testicular germ-cell cancer is a model for curable malignancy, with an overall survival rate of 90–95% when considering all stages. Current therapeutic trials are aimed at maintaining these favorable results while reducing treatment-related toxicity, particularly in non-advanced disease [1]. Until recently, irradiation to the ipsilateral pelvic and para-aortic nodes was the standard

*Correspondence to: Dr J. Aparicio, Servicio de Oncología Médica, Hospital Universitario La Fe, Avda. Campanar 21, E-46009 Valencia, Spain. Tel: +34-6-197-31-38; Fax: +34-6-197-31-38; E-mail: [email protected] †Members of the Spanish Germ Cell Cancer Cooperative Group are listed in the Acknowledgements. © 2003 European Society for Medical Oncology

therapy for patients with stage I seminoma after orchiectomy. Reported relapse rates have ranged from 3% to 5%, and seminoma deaths occur in about 2% of cases [2]. However, a small but significant increased risk for second malignancies and peptic ulcers with this treatment approach has been confirmed. In addition, recent improvements in clinical staging (particularly high-resolution computed tomography scanning) have made stage I seminoma a rather different disease from that whose 60% relapse rate led to use of prophylactic radiotherapy nearly 30 years ago. Alternative treatment options have been investigated in this patient subset, including surveillance [3] and adjuvant chemotherapy [4]. Although no randomized clinical trial has been published yet, long-term patient survival does not seem to be compromised in comparison with traditional irradiation. How-

Downloaded from http://annonc.oxfordjournals.org/ at Gumberg Library on July 12, 2015

Received 21 October 2002; revised 16 December 2002; accepted 23 January 2003

868

Patients and methods Patients with histologically proven pure seminoma, clinical stage I disease and resection margins free of tumor were included in this study after orchiectomy performed at any of the GG centers. Routine staging consisted of clinical history, physical examination, chest X-ray films, computed tomography (CT) of abdomen and pelvis, ultrasonography of the contralateral testicle, whole blood cell counts, serum chemistries including lactate dehydrogenase, α-fetoprotein (AFP) and β-human chorionic gonadotropin (BHCG). Patients with UICC (1987 edition) tumor stage pT2 or greater (i.e. tumors infiltrating albuginea, epididymis, spermatic cord or scrotum) [5] or with venous or lymphatic vascular invasion (considered local risk factors) received adjuvant single-agent carboplatin (400 mg/m2 every 28 days, for two courses). Treatment was given on an outpatient basis, within 2 h. Systematic antiemetic prophylaxis with dexamethasone and 5-hydroxytryptamine-3 antagonists was employed. Com-

Results From January 1994 to June 1999, 203 cases were included in 202 patients (one patient with a bilateral seminoma was registered twice) from 29 centers. Median patient age was 35 years (range 15–81) and median of maximum tumor diameter was 44 mm (range 10–120). One hundred and ninety-nine patients underwent inguinal orchiectomy, two underwent scrotal orchiectomy and one had abdominal surgery for an undescended testicle. Preoperative serum BHCG levels were positive (i.e. >5 mU/ml) in 30 cases (14.8%) (median 32 mU/ml; range 10–3730). By definition, all patients were pre- and postorchiectomy AFP negative, and postorchiectomy BHCG negative. Sixty patients (29.6%) with risk factors received two courses of carboplatin, whereas 143 (70.4%) were managed with surveillance. Main patient characteristics and distribution of prognostic factors among treatment arms are summarized in Table 1. Median time from orchiectomy to chemotherapy start was 37 days (range 9–70). Time interval between courses was 21, 28 and 35 days in four, 50 (86%) and four cases, respectively. All treatment delays were due to asymptomatic myelotoxicity. No dose reductions were performed. Except for emesis, no episodes of grade 3–4 adverse events were noted. Chemotherapy toxicity is summarized in Table 2. At least 20 patients (10%) are known to have fathered after treatment (13 on surveillance, seven after carboplatin). At the time of the present analysis, median follow-up time was 52 months (range 14–92), with 164 patients (80%) being followed for >3 years and only five patients (2.5%) being lost to follow-up. Relapses were observed in two (3.3%) patients treated with adjuvant carboplatin and in 23 (16.1%) patients on surveillance. Five-year DFS was 87.3% (95% CI 82.5% to 92%) for all patients, 83.5% (95% CI 77.2% to 89.8%) for those on surveillance and 96.6% (95% CI 92% to 100%) for chemotherapy-treated cases

plete blood cell counts, serum biochemistry and toxicity assessment (World Health Organization criteria) were mandatory on day 28 of the first chemotherapy course, whereas toxicity of the second course was recorded mainly in a retrospective form in absence of symptoms. Patients without risk factors were managed by clinical surveillance. Chest X-rays, AFP and BHCG were scheduled at 3, 6, 9, 12, 18, 26, 36, 48, 60 and 72 months after orchiectomy, and abdominal CT scans were performed at 6, 12, 18, 26, 36, 48, 60 and 72 months [6]. Tumor relapses in both groups were treated primarily with etoposide and cisplatin chemotherapy (E400P) [7]. Potential prognostic factors for relapse were prospectively recorded, including patient age (≤30 years versus >30 years), tumor diameter (≤40 mm versus >40 mm), histological subtype (classical versus anaplastic), pT stage (pT1 versus pT2–4), vascular invasion, rete testis invasion and preoperative BHCG levels (negative versus positive). Histological features were reviewed locally. Overall survival (OS) and disease-free survival (DFS) were estimated from the date of orchiectomy with the Kaplan–Meier method [8]. Comparison of resulting curves and univariate analysis of prognostic factors were performed with the log-rank test [9]. Ninety-five per cent confidence intervals (95% CI) are given when appropriate.

Figure 1. Actuarial disease-free survival for the entire study population according to treatment allocation: adjuvant carboplatin (upper curve, n = 60) versus surveillance (lower curve, n = 143). Cross marks indicate censored time points.

Downloaded from http://annonc.oxfordjournals.org/ at Gumberg Library on July 12, 2015

ever, surveillance is associated with relapse rates of 15–20% (most of them can be salvaged with chemotherapy or irradiation), considerable psychological stress and incremental costs. On the other hand, adjuvant chemotherapy with single-agent carboplatin, although generally well tolerated, implies treating all patients with stage I seminoma, while 80% of them are thought to be cured with orchiectomy. In 1994, the Spanish Germ Cell Cancer Cooperative Group (GG) designed a protocol to evaluate a dual treatment policy for stage I seminoma in which only high-risk patients (with an expected relapse rate >20%) were assigned to adjuvant carboplatin whereas the remainder (whose expected relapse rate is <15%) were managed with surveillance. If this strategy was feasible, most patients would be preserved from postorchiectomy treatment and the relapse rate both on surveillance and after chemotherapy would be significantly decreased. Mature results of this protocol are presented here. The objective of this study was to assess the following: (i) the percentage of patients needing adjuvant chemotherapy; (ii) the relapse rate in patients treated with either carboplatin or surveillance; and (iii) long-term overall and disease-free survival rates in both groups.

869 Table 1. Patient characteristics Feature

Surveillance (n = 143) (%)

Carboplatin (n = 60) (%)

All patients (n = 203) (%)

Age (years) ≤30

47 (32.9)

18 (30.0)

65 (32.0)

>30

96 (67.1)

42 (70.0)

138 (68.0)

Tumor diameter (mm)a ≤40

66 (50.4)

20 (37.7)

86 (46.7)

>40

65 (49.6)

33 (62.3)

98 (53.3)

Yes

17 (11.9)

13 (21.7)

30 (14.8)

No

126 (88.1)

47 (78.3)

173 (85.2)

138 (96.5)

56 (93.3)

194 (95.6)

5 (03.5)

4 (06.7)

9 (04.4)

pT1

143 (100)

16 (26.7)

159 (78.3)

pT2

–

38 (63.3)

38 (18.7)

pT3

–

6 (10.0)

6 (03.0)

Yes

–

35 (60.3)

35 (17.9)

No

137 (100)

23 (39.7)

160 (82.1)

Preoperative serum BHCG positive

Histological subtype Classical

Downloaded from http://annonc.oxfordjournals.org/ at Gumberg Library on July 12, 2015

Anaplastic Staging (pT)

Vascular invasionb

Rete testis invasionc Yes

27 (19.1)

23 (39.0)

50 (25.0)

No

114 (80.9)

36 (61.0)

150 (75.0) 143 (70.4)

Local risk criteria None

143 (100)

–

Vascular invasion

–

16 (26.7)

16 (07.9)

>pT1

–

25 (41.7)

25 (12.3)

Both criteria

–

19 (31.6)

19 (09.4)

Data available in 184 (90.6%)a, 195 (96.1%)b and 200 (98.5%)c patients. BHCG, β-human chorionic gonadotropin.

(Figure 1). Median time to relapse for patients with tumor recurrences was 11 months (range 3.9–39.6). Relapses were located mainly in the retroperitoneum (84%), whereas one case each presented in mesenteric nodes (this patient had received a transcrotal orchiectomy), pelvis, lung and as serum BHCG increase. Only two cases (8%) relapsed with symptoms of disease, the rest being detected through routine CT scans (76%) or increase in serum BHCG levels (16%). Median tumor size at relapse was 32.5 mm (range 0–60). All patients were rendered disease-free with chemotherapy (four cases with three courses of classic BEP, 18 cases with four courses of E400P), irradiation (two cases) or surgery (one case). However, four patients (2%) died of causes unrelated to seminoma or its treatment: one traffic accident, one heroine overdose, one hepatic cirrhosis and one non-Hodgkin’s lymphoma. Four patients (2%) had second germ-cell malignancies: two metachronous non-seminomatous germ-cell tumors, one synchronous and one metachronous testicular seminomas. One other patient developed acute leukemia. These latter five patients with second

Table 2. Acute toxicity in 60 patients treated with 120 courses of carboplatin Toxicity

WHO grade (% of courses) 1

2

3

4

Neutropenia

8.3

3.3

–

–

Thrombocytopenia

3.3

1.6

–

–

Anemia

2.5

–

–

–

Vomiting

15.0

7.5

5.0

–

Mucositis

–

0.8

–

–

–

–

–

Fatigue

4.2

Forty-three (71.6%) patients presented no adverse effects at all. WHO, World Health Organization.

malignancies are living disease-free. Of note, all six malignancies in this series occurred in patients on surveillance. Five-year OS was 96.7% (95% CI 92.8% to 100%) and cause-specific OS was 100%.

870 A univariate analysis of prognostic factors for DFS was performed. In the whole series, only adjuvant therapy (favoring carboplatin versus surveillance, P = 0.0142) reached statistical significance. No predictive features could be identified in patients on surveillance, whereas positive serum preoperative BHCG levels were of adverse prognostic value (P = 0.0064) in patients treated with adjuvant carboplatin. However, in all groups, younger age was associated with a non-significant increased risk of recurrence. Table 3 depicts the distribution of relapses among prognostic categories.

Discussion

Table 3. Distribution of seminoma relapses among prognostic categories Feature

Surveillance (%) 23/143 (16.1)

Carboplatin (%) 2/60 (3.3)

All patients (%) 25/203 (12.3)

≤30

9/47 (19.1)

1/18 (5.5)

10/65 (15.4)

>30

14/96 (14.6)

1/42 (2.4)

15/138 (10.9)

≤40

9/66 (13.6)

0/20 (–)

9/86 (10.5)

>40

11/65 (16.9)

2/33 (6.1)

13/98 (13.3)

Yes

2/17 (11.8)

2/13 (15.4)

4/30 (13.3)

No

21/126 (16.7)

0/47 (–)

21/173 (12.1)

Age (years)

Tumor diameter (mm)a

Preoperative serum BHCG positive

Histological subtype Classical

21/138 (15.2)

2/56 (3.6)

23/194 (11.8)

Anaplastic

2/5 (40)

0/4 (–)

2/9 (22.2)

pT1

23/143 (16.1)

1/16 (6.3)

24/159 (15.1)

pT2

–

1/38 (2.6)

1/38 (2.6)

pT3

–

0/6 (–)

0/6 (–)

Yes

–

2/35 (5.7)

2/35 (5.7)

No

22/137 (16.1)

0/23 (–)

22/160 (13.7)

Staging (pT)

Vascular invasionb

Rete testis invasionc Yes

8/27 (29.6)

1/23 (4.3)

9/50 (18.0)

No

14/114 (12.3)

1/36 (2.8)

15/150 (10.0)

None

23/143 (16.1)

–

23/143 (16.1)

Vascular invasion

–

1/16 (6.2)

1/16 (6.2)

>pT1

–

0/25 (–)

0/25 (–)

Both criteria

–

1/19 (5.3)

1/19 (5.3)

Local risk criteria

Data available in 184 (90.6%)a, 195 (96.1%)b and 200 (98.5%)c patients.

Downloaded from http://annonc.oxfordjournals.org/ at Gumberg Library on July 12, 2015

Approximately 75% of patients with testicular seminoma present with stage I disease at diagnosis. Over the past 30 to 40 years, standard treatment of these patients involved radical inguinal orchiectomy and routine adjuvant radiation therapy with 25–30 Gy to the retroperitoneal and ipsilateral pelvic lymph nodes. The

results of this modality have been excellent, with more than 2600 patients being reported from 1980 to 1994, consistent long-term DFS rates of 96% and OS rates of 98% [2]. However, recent concerns about a 6–9% incidence of peptic ulcers [10] and especially a two- to three-fold increased risk of second malignancies in irradiated fields [11, 12] have resulted in the evaluation of alternative therapies. The first of these attempts has been the successful reduction of either the radiation doses or the radiation fields [13]. However, it is becoming increasingly evident that radiation therapy to all stage I seminoma patients may represent overtreatment. In fact, amounting experience with postorchiectomy surveillance has demonstrated that nearly 80% of these patients can expect to be cured without adjuvant treatment, and almost all recurrences can be successfully rescued. The development of imaging of retroperitoneal nodes with CT has allowed a more accurate assessment of disease extent and evaluation of treatment results, and has facilitated follow-up evaluation on surveillance programs. More than 900 patients managed by this approach have already been reported

871 a protective effect of chemotherapy against a second contralateral testicular tumor (in comparison with surgery alone) can not be ruled out, as suggested by Oliver et al. [4]. The simplification of treatment according to our multicenter protocol means that there would be reduced referral of early stage testis cancer to specialist centers to discriminate between high- and low-risk patients. Although our simplified surveillance policy seems adequate, it should be possible to minimize costs by further increasing the intervals for follow-up investigations. In conclusion, although there is need for improvement, this dual treatment policy for patients with stage I seminoma is feasible, preserves 70% of cases from adjuvant treatment (and 59% from any form of postorchiectomy therapy) without compromising longterm results, and it could be used as a risk-adapted alternative to irradiation, surveillance and carboplatin.

Acknowledgements This work was presented in part at the 2001 ASCO Meeting, San Francisco, CA, USA. Participant members of the Spanish Germ Cell Cancer Cooperative Group are as follows: Jorge Aparicio, Ana L. Yuste (Hospital Universitario La Fe, Valencia); Xavier García del Muro, José R. Germà (Institut Catalá d’Oncologia, Barcelona); Pablo Maroto (Hospital de Sant Pau, Barcelona); Luis Paz-Ares (Hospital 12 de Octubre, Madrid); Emilio Alba (Hospital Clínico Universitario, Málaga); Alberto Sáenz (Hospital Clínico, Zaragoza); Josefa Terrasa (Hospital Son Dureta, Mallorca); Agustí Barnadas (Hospital Germans Trias i Pujol, Badalona); Daniel Almenar (Hospital Doctor Peset, Valencia); José A. Arranz (Hospital Gregorio Marañón, Madrid); Miguel Sánchez (Hospital Donostia, San Sebastián); Antonio Fernández Aramburu (Hospital General, Albacete); Javier Sastre (Hospital Clínico San Carlos, Madrid); Joan Carles (Hospital del Mar, Barcelona); Joan Dorca (Hospital Josep Trueta, Girona); Josep Gumà (Hospital Universitari Sant Joan, Reus); Sonia Del Barco (Hospital Clínico San Cecilio, Granada); Alberto Rodríguez (Hospital J.R. Jiménez, Huelva); Enrique Barrajón (Hospital General, Elche); Romà Bastús (Mutua de Terrassa); Severina Domínguez (Hospital Txagorritxu, Vitoria); Carmen Crespo (Hospital Ramón y Cajal, Madrid); Marta López-Brea (Hospital Marqués de Valdecilla, Santander); Adolfo Murias (Hospital Insular de Gran Canaria); Enrique Gallardo (Hospital de Terrassa); Purificación Martínez (Hospital de Basurto, Bilbao); Francisco J. Dorta (Hospital Nª Sª de la Candelaria, Tenerife); María Lomas (Hospital Infanta Cristina, Badajoz); Antonio Colmenarejo (Hospital del Aire, Madrid).

References 1. Dearnaley DP, Huddart RA, Horwich A. Managing testicular cancer. Br Med J 2001; 322: 1583–1588. 2. Zagars GK. Management of stage I seminoma: radiotherapy. In Horwich A (ed.): Testicular Cancer: Investigation and Management, 2nd edition. London, UK: Chapman & Hall 1996; 99–121. 3. Warde P, Gospodarowicz MK, Panzarella T et al. Stage I testicular seminoma: results of adjuvant irradiation and surveillance. J Clin Oncol 1995; 13: 2255–2262.

Downloaded from http://annonc.oxfordjournals.org/ at Gumberg Library on July 12, 2015

since 1990 [3, 14–16], with 5-year DFS rates of 80–85% and OS approaching 100%. The disadvantages of clinical surveillance are incremental costs, need of patient compliance to the follow-up protocol, psychological stress and the chance of late relapses [17]. A third treatment option is adjuvant chemotherapy with singleagent carboplatin. To date, nearly 550 patients so treated have been reported [4, 18–20]. Acute toxicity is mild and long-term side-effects have not been described. Moreover, 5-year DFS is 97–100%, with a 100% cause-specific OS. However, once again, all patients being treated need to show a disease with very good prognosis. An ongoing phase III randomized study (MRC-TE19) is comparing adjuvant radiotherapy with a single cycle of carboplatin, although recent experience suggests that two chemotherapy courses are more effective [18]. To our knowledge, this is the first report evaluating a dual policy of postorchiectomy surveillance and selective adjuvant chemotherapy for stage I seminoma. Our results suggest that this schedule is feasible in a nationwide multicenter setting, and longterm results are equivalent to those achieved with other modalities (i.e. no seminoma-related mortality was seen). Considering patients treated with adjuvant chemotherapy and after tumor relapse, 120 of 203 cases (59.1%) were safely preserved from any form of therapy. With only two of 60 relapses (3.3%), we confirm two courses of carboplatin as an effective adjuvant treatment even for patients with high-risk criteria. Moreover, it seems evident that starting chemotherapy within 10 days after orchiectomy is not mandatory [18–20]. The main limitation of this study was the absence of universally accepted prognostic factors for relapse in patients managed by surveillance at the time of its design. In a single center experience, tumor diameter, microvascular invasion and age reached statistical significance [21]. In the Royal Marsden Hospital series of 103 patients managed by surveillance, the only significant prognostic factor for relapse was the presence of lymphatic and/or vascular invasion (9% versus 17% relapse rate) [15]. More recently, in a joint study with 638 patients from three large series of surveillance, only tumor size and rete testis invasion entered the multivariate prognostic model [22]. We chose pTNM and vascular invasion because of its widespread use and general acceptance [23]. However, a 16.1% relapse rate on surveillance (similar to that expected in unselected stage I patients) makes us question the predictive value of our criteria. None of the prognostic factors studied in this report achieved statistical significance; a fact that could be related to the small number of events (relapses) and/or the impact of selective adjuvant chemotherapy. Thus, in 1999 the Spanish GG started a second dual policy study using the new selection criteria (tumor size >40 mm and rete testis invasion) [22], employing carboplatin with the more adequate area-under-the-curve (AUC 7) dosing, and using a 21-day interval between cycles. We have not specifically addressed other important issues such as long-term carboplatin toxicity, fertility, quality of life aspects and economic costs both on surveillance and after carboplatin because all of them were investigated in previous studies [3, 4, 14–19]. However, it is noteworthy that we observed in this series six second malignancies (four of them germ-cell tumors), but none of them occurred in the group treated with carboplatin. Then,

872 14. von der Maase H, Specht L, Jacobsen GK et al. Surveillance following orchidectomy for stage I seminoma of the testis. Eur J Cancer 1993; 29A: 1931–1934. 15. Horwich A, Alsanjari N, A’Hern R et al. Surveillance following orchidectomy for stage I seminoma. Br J Cancer 1992; 65: 775–778. 16. Germà JR, Climent MA, Villavicencio H et al. Treatment of stage I testicular tumours. Br J Urol 1993; 71: 473–477. 17. Sharda NN, Kinsella TJ, Ritter MA. Adjuvant radiation versus observation: a cost analysis of alternate management schemes in early-stage testicular seminoma. J Clin Oncol 1996; 14: 2933–2939. 18. Dieckmann K-P, Brüggeboes B, Pichlmeier U et al. Adjuvant treatment of clinical stage I seminoma: is a single course of carboplatin sufficient? Urology 2000; 55: 102–106. 19. Reiter WJ, Brodowicz T, Alavi S et al. Twelve-year experience with two courses of adjuvant single-agent carboplatin therapy for clinical stage I seminoma. J Clin Oncol 2001; 19: 101–104. 20. Steiner H, Höltl L, Wirtenberger W et al. Long-term experience with carboplatin monotherapy for clinical stage I seminoma: a retrospective single-center study. Urology 2002; 60: 324–328. 21. Warde P, Gospodarowicz MK, Banerjee D et al. Prognostic factors for relapse in stage I testicular seminoma treated with surveillance. J Urol 1997; 157: 1705–1710. 22. Warde P, Specht L, Horwich A et al. Prognostic factors for relapse in stage I seminoma managed by surveillance: a pooled analysis. J Clin Oncol 2002; 20: 4448–4452. 23. Hoeltl W, Kosak D, Pont J et al. Testicular cancer: prognostic implications of vascular invasion. J Urol 1987; 137: 683–685.

Downloaded from http://annonc.oxfordjournals.org/ at Gumberg Library on July 12, 2015

4. Oliver RTD, Edmonds PM, Ong JYH et al. Pilot studies of two and one course carboplatin as adjuvant for stage I seminoma: should it be tested in a randomized trial against radiotherapy? Int J Radiat Oncol Biol Phys 1994; 29: 3–8. 5. Hermanek P, Sobin LH. UICC: TNM Classification of Malignant Tumours, 4th edition. Berlin, Germany: Springer 1987. 6. Germà-Lluch JR on behalf of the Spanish Germ-Cell Cancer Group (GG). Adjuvant treatment for stage I germ-cell testicular tumours: preliminary experience of the Spanish Germ-Cell Cancer Group. In Jones WG, Appleyard I, Harnden P, Joffe JK (eds): Germ Cell Tumours IV. London, UK: John Libbey 1998; 139–142. 7. Arranz JA, García del Muro X, Gumà J et al. E400P in advanced seminoma of good prognosis according to the International Germ Cell Cancer Collaborative Group (IGCCCG) classification: the Spanish Germ Cell Cancer Group experience. Ann Oncol 2001; 12: 487–491. 8. Kaplan E, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958; 53: 457–481. 9. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 1966; 50: 163–170. 10. Fossa SD, Aass N, Kaalhus O. Radiotherapy for testicular seminoma stage I: treatment results and long-term post-irradiation morbidity in 365 patients. Int J Radiat Oncol Biol Phys 1989; 16: 383–389. 11. Bokemeyer C, Schmoll HJ. Treatment of testicular cancer and the development of secondary malignancies. J Clin Oncol 1995; 13: 283–292. 12. Chao CK, Lai PP, Michalski JM et al. Secondary malignancy among seminoma patients treated with adjuvant radiation therapy. Int J Radiat Oncol Biol Phys 1995; 33: 831–835. 13. Fossa SD, Horwich A, Russell JM et al. Optimal planning target volume for stage I testicular seminoma: a Medical Research Council randomized trial. J Clin Oncol 1999; 17: 1146–1154.