A Surveillance Study of Clinical Stage I Nonseminomatous Germ Cell Tumors of the Testis: 10-Year Followup

A Surveillance Study of Clinical Stage I Nonseminomatous Germ Cell Tumors of the Testis: 10-Year Followup

~22-5347/95/1543-1045$03.00/0 THEJOURNAL OF UROLOGY Copyright 0 1995 by AMERICAN UROUX;~C~U. AS~OCUTION,hc. Vol. 164,1045-1049. September 1996 Printe...

724KB Sizes 0 Downloads 2 Views

~22-5347/95/1543-1045$03.00/0 THEJOURNAL OF UROLOGY Copyright 0 1995 by AMERICAN UROUX;~C~U. AS~OCUTION,hc.

Vol. 164,1045-1049. September 1996 Printed in U S A .

A SURVEILLANCE STUDY OF CLINICAL STAGE I NONSEMINOMATOUS GERM CELL TUMORS OF THE TESTIS: 10-YEARFOLLOWUP NICOLA NICOLAI

GIORGIO PIZZOCARO

AND

From the Division of Urology,Istituto Nazionnlo Turnon, Milan, Italy

ABSTRACT

Purpose: We evaluate the 10-year results of a surveillance study of clinical stage I nonseminomatous germ cell tumors of the testis. Materials and Methods: Between 1981 and 1984 we recruited 85 consecutive evaluable patients with nonseminomatous germ cell tumors of the testis and normal post-orchiectomy physical examination, chest x-rays, bipedal lymphangiography, abdominal scans and serum tumor markers. The patients were followed for at least 10 years after orchiectomy alone, which was performed elsewhere in 90% of the cases. Results: The interval between visits was twice as long as it was scheduled. Relapses occurred in 25 patients (29.4%) after a median disease-free interval of 7 months (range 2 to 68). Five patients had further relapses and 3 (3.5%) died of cancer. Retroperitoneal relapses (19%) occurred later than lung relapses, and they were diagnosed when larger than 5 cm.in 7 patients. The percentage of embryonal carcinoma within the tumor associated with relapse (p = 0.008),T category (p = 0.023),scrotal violation (p = 0.042)and vascular invasion (p = 0.063)had a weak correlation but data on T category and vascular invasion were available for only some patients. Conclusions: Surveillance is a difEcult type of study and missing data may compromise the therapeutic program based on prognostic factors. KEY WORDS:carcinoma, testidar neoplasms, survival, neoplasm metastasis

tute. The remaining 76 men (90%)were referred to us after orchiectomy performed elsewhere. Of these 76 patients 20 (26.3%) had undergone transscrotal biopsy or transsrrotal orchiectomy initially, and they all underwent en bloc removal of the funicular stump and scrotal scar before starting the surveillance program. No pathological 6nding other than evidence of tumor at the cut end of the funicular stump was considered for exclusion criteria. Clinical staging in all cases was based on clinical history, complete physical examination, determination and half-life kinetics of serum a-fetoprotein (AFP)and fi subunit of human chorionic gonadotropin (EHCG),front and lateral chest x-rays, bipedal lymphangiography,and computerized tomography (CT)of the abdomen and pelvis. Every radiogram was jointly reviewed by radiologists and urologists. All clinical and radiographic examinations had to be unequivocallynegative for metastasis (in particular, no visible para-aortic or pre-aortocaval nodes on CT). Serum tumor markers had to be strictly normal according to half-life kinetics of 5 days for AFP and 1 day for BHCG (markers were repeated at least twice following orchiectomy). Fully informed consent, availability for close clinical observation, no second tumor, no further treatment beyond radical orchieetomyand no pathe logical evidence of tumor at the cut end of the funicular stump (following radical surgery) completed the criteria for PATIENTS AND METHODS inclusion into the study. The clinical records and slides for the 76 patients who A surveillance program was offered to all patients with clinical stage I nonseminomatous germ cell tumors of the underwent surgery elsewherewere reviewed to evaluate clintestis hospitalized between August 1981 and December 1984. ical history and reassess histology. Duration of symptom Of the 90 consecutive eligible patients 5 were excluded be- before orchiectomy,interval between orchiectomyand d e h cause of a previous contralateral testicular tumor (2), a con- itive clinical staging, type of primary surgery, pre-orchiectemporary thyroid cancer (I),patient refusal (1)and 1cycle tomy titers of AFP and FHCG, histology of the primary of chemotherapy administered for a falsely elevated n-feto- tumor, percentage of embryonal carcinoma, necrosis witbin protein level after orchiectomy (1).To date 85 patients (me- the tumor, T category and pen-tumoral vascular invasion dian age 26 years) were evaluable. were recorded, whenever possible, and assayed as predictors Only 9 patients (10%)underwent orchiectomy at our kti- of metastases. Followup was once a month during year 1,every 2 month8 during year 2, every 3 months up to year 5 and 6 Accepted for publication February 24,1996.

The standard treatment of clinical stage I nonseminomatous germ cell tumors of the testis is retroperitoneal lymph node dissectionl.2 or close observation after orchiectomy alone.3.4 Both therapies are commonly considered adequate and equivalent in terms of definitive cure rates. Surveillance gained more and more success as a consequence of improved diagnostic procedures and identification of risk factors, mainly vascular invasion.4.6 On the other hand, with introduction of the nerve sparing technique, retroperitoneal lymphadenectomyhas become an acceptable option to treat the retroperitoneal nodes.2.6.7 Careful consideration of risk factors as well as of the adequacy and length of followup is of paramount importancefor proper evaluation of these therapeutic strategies. Recently, attention has been focused on the impact of late (more than 2 years) recurrences in the global outcome of patients with clinical stage I testicular cancer in a surveillance program8 and adjuvant chemotherapy programs have been i n t r o d u d in high risk patients.9-11 Stimulated by these observations, we reanalyzed the results in our patients who were placed on dose observation &r orchiectomy done for clinical stage I nonseminomatous germ cell tumors of the testis more than 10 years ago. Middle-term results have already been reported.12

1046

1046

SURVEILLANCE IN CLINICAL STAGE I TESTICULAR CANCER

months thereafter. Serum tumor markers, chest x-ray and plain films of the abdomen for lymphangiographic control were repeated at each visit. Lymphangiography was repeated at least once after disappearance of the contrast medium, generally 6 to 12 months &r the initial examination. CT of the abdomen and pelvis was scheduled for every 2 visits. After year 5 ultrasonography of the abdomen and pelvis could replace CT. Patients with relapse underwent ongoing therapeutic protocols. Followup was closed on May 31, 1994. RESULTS

Median followup for the entire series was 132 months (range 114 to 156). Actually, on average patients were seen every 2 months during the first 2 years, every 4 months during year 3, every 6 months until year 5 and annually thereafter. CT was performed every 2 visits during the first 5 years. Re-orchiectomy serum AFP and &HCG levels were available for only 43 patients (50%). Histology was reassessed according to the available slides in all cases. T category and vascular invasion could be defined precisely in 54 (63%) and 28 (33%) patients, respectively. The percentage of embryonal carcinoma within the tumor was defhed in 81 patients (95%). A total of 25 (29.4%) patients had relapse at a median of 7 months after orchiectomy (range 2 to 68). Of the total relapses 19 (76%) occurred within 1 year, 3 (12%) during year 2 and 3 (12%)thereafter (2at 36 months and 1 at 68 months). Table 1 shows initial relapses according to site and interval of occurrence. Three patients had synchronous retroperib neal and distant metastases (lung in 2 and supraclavicular in 1).The patient with a scrotal recurrence underwent surgery alone and lung metastases developed 5 months later. Four other patients had further relapses. The size of retroperitoneal and lung metastases is shown in figure 1. Of the 11 lung metastases 9 (83%)were identified when smaller than 2 cm.versus only 1 of 14 retroperitoneal metastases (7%). In particular, 50% of the retroperitoneal metastases were larger than 5 cm.when first diagnosed. The median tumor-free interval for patients with lung metastases was 4 months (range 2 to 68)compared to 12months (range 4 to 36) for those with retroperitoneal metastases. The interval for diagnosis of retroperitoneal or lung metastases only is shown in figure 2. This different pattern was weakly significant by the Mann-Whitney rank sum test (z sub t = 2.407, p = 0.016). Elevation of at least 1 of the 2 serum tumor markers preceded a measurable lesion in 7 of the 25 relapses (28%), including 1 patient (4%) with a relapse detected by AFP elevation only who probably suffered an unidentified retroperitoneal metastasis. Ten patients (40%) had a relapse not accompanied by any marker elevation. Marker elevation occurred in 2 of the 3 patients with late relapses. Table 2 shows clinical and pathological factors assayed at

RG.1. Size of metastases according to site of first relapse in patients with metastases in retroperitoneal nodes only (ll),in lun onl (g), in retroperitoneal nodes and lung (2), and m retropentonei andrsupraclavicular nodes (1).

FIG.2. Months at diagnosis of first relapse according to site in 20 atients with retro ritoneal or lung metastases only (p = 0.016, Rlann-mtney r a Zs u m test).

univariate analysis as potentially predictive for metastasis. The percentage of embryonal carcinoma, available in 81 of 86 cases (95%),demonstrated a significant correlation with metastases (p = 0.008, chi-square test). With the limitations due No.Pta. With 96 26 Median Mas' Site of Relapae Relapse (40) Relapsea at Diwoeia to the lack of data in many patients, T category (p = 0.023) (range) and open transscrotal biopsy (p = 0.042)showed some c o r n 11' (12.9) 44 lO(4-36) Retmperitoneal nodes only lation with metastases by the chi-square test. Vascular inva9t (10.6) 36 4 (2-68) Lung only sion (p = 0.069)showed only a trend correlation but this Retmperitoneal and distant 3$ (3.6) 12 7 (7-15) AFT elevation only 1 (1.2) 4 22 information was available in only 28 patients. The lack of If (1.2) 4 - 3 scrotum much data did not afford a multivariate analysis. However, Totals 25 (29.4) 100 7 (2-68) the relapse rate in patients with missing data was never * One patient had relapse again in the lung and mediastinum. lower than in those with available data. t Three patients had relapse again: 1 in the lung, 1 in the lung and retm (herall, 82 of our 85 patients (96.5%)are currently disperitoneum, and 1 in the brain. ease-free, since 22 of 25 with relapse (88%)could be cured (3 Lung in 2 patients and supraclavicular area in 1. 0 Followed by lung relapse 5 months later. with retroperitoneal lymph node dissection alone, and 19 TABLE1. Site and interval of the initial relapse in 86 evaluable patients

1047

SURVEILLANCE IN CLINICAL STAGE I TESTICULAR CANCER TABLE2. Clinical and pathological featwes assayed 08 potential predictors of metostotic spread at univariate analysis

-

Feature g, Embryonal Ca: 100 50 to 99 Less than 50 Not reassessable

No.Relapid chi-squant Total ( 8 )

6/10 (60) lY30 (36.7) 6/41 (14.6) 1/4 (60.0)

Rstrop.ritowun

w Msdiutinum and neck 8aotum

9.741

0.008

6.196

0.029

6.343

0.042

3.319

0.069

(0)

6.144

0.076

16'43 (37.2) 9/42 (21.4)

1.846

Not significant

7/34 (20.6) 18/51 (36.3)

1.476

Not significant

T category: 1

More than 1 Not reassessable Type of primary surgery: Inguinal orchiectomy Transscrotal omhiedomy Open t r a n m t a l biopsy Vascular invasion: Absent Present Not reassessable Histology: Embryonal Ca with or without others* Teratoma with or without others Yolk sac tumor Mos. between orchiectomy and staging: 1 or less More than 1 Mos. symptoms: 1 or less More than 1 Pre-orchiectomy s e w tumor markers: Normal Elevated Not reassessable Necrosis within primary tumor: Absent Present Not reassessable * No teratoma.

6/38 (16.9) 8/16 (60) lYgl(36.6) 15/66 (23.1) 3/8 (37.6) 7/12 (68.3) 1/18 (11.1) 5/10 (60) 18/57 (31.6) 1y24 (46.8) 14/66 (24.1)

om

sita

p Vdw

No.Evanta

9bEbh.

8330ila

16

17.6

1.9.

16.9

45.6 98.4

2t 1 1

2.4 1.2 1.2

Brain 1 AFP ohvation only Tot& s9 'Two pathta had repeated lung t h o ruplrcl.Viarluand 1 mediutuul.

-

1.2 30.0

6.1 3.0 8.0 3.0

100

(2.3%). One patient was treated with cisplatin, etoponide and bleomycin for a late retroperitoneal relapse, and sUgered a right categvry T2NOM0, grade 1 renal cell carcinoma 87 months aRer entering the surveillance study (47 month after starting chemotherapy). He had no clinical evidence of disease 116 and 29 months aRer o c ~ u l t ' ~ n of c ethe tasticular and renal tumor, respectively. One untreWd patiemt (1.7%) had a contralateral clinical stage I nonseminomatoua germ cell tumor of the testis 112 montbs aftar orchiectomy and underwent unilateral nerve sparing retroperibneal lymph node dissection according to the new ongoing protocol. He was disease-be 139 months aRer the initial orchiectomy and 27 months after the secondary te~ticulartumor. DISCUSSION

Long-term (29.4%) and mid-term (27%)" rates 1/4 (26) remain comparable. However, some additional comiderlff39 (25.6) 0.329 Not significant ations can be made. The main reault of our surveillance etudy 14/42 (33.3) wae failure to identify clearly risk factors for relapse and to adhere to the scheduled followup program. For these r e ~ ~ o p ~ , Ye (12.6) in January 1985 we lwlumed performing retroperitoneal 7/26 (26.9) 0.133 Not significant lymphadenectomy for clinical stage I nonseminoma.18 17/61 (33.3) Ofour patients 90% had undergone orchiectomy elsewhere and 20 scrotal violations were recorded. Significant praspoetic information, such as pre-orchiectomy AFP and HCG titers, T category and vascular invasion, was not available in with cisplatin-based chemotherapy with or without subse- most patients (table 2). Thie was not a selection bias, since quent surgery). Two of the latter 19 patients had further 1p relapses in patients with missing data were not significantly lapse, requiring salvage therapy for cum: 1 for a lung Apse 13 M e r e n t from those obeerved in patients with available data months after the initiallune metaetaeis (16months after orchi- but we could not wrform a multivariate analmb of risk ectomy),and 1for lung andk$mperitoneal recurrenceSmonths factors. only the paircentage of ernbryod car&mm, which could be aammed in after the initial lung relapse (11 months aRer orchiectomy). was not considered in a previous t i t u d y , ~ Median followup of the 22 apparently d patienta was 122 95% of the casesand it was Bigaificantlyassociated with the risk of metaetesee. This result is consistant with that oftha months since the last relapse (range 77 to 151). A total of 3 patients (3.5%)died of the disease. One patient Teeti& Cancer Intergroup Study, which met similar but had lymphangiographic evidence of a retroperitoneal metas- leas severe difficulties in the central pathology review.'* We found that a dose ahemation regimen was difEcult to tasis smaller than 2 cm.6 month &r orchieetomy, refused any immediate treatment and died of progressive dieease 7 follow, not only for objective r8880138, such aa m p h i c a l months later (13 months after orchiectomy). One patient area or poor availability of CT in our country early in the presented with a 10 cm. large retroperitoneal 1111188 17 198oe, but also for psychological reaeons, since many pamonths after orchiectomy that regressed completely with tients became anxious and they could not tolerate close folchemotherapy alone. He was treated successfully for a lung lowup. Furthermore, the oceurreneaof late relapses forced ua recurrence after 55 months (72 months aRer orchiectomy) to maintain long-term followup, rendering the d i B 4 t i e s but mediastinal relapse developed a few months later (78 with the patienta more severe. We must note that a patient months after orchiectomy)and he died at 106 months despite refused immediate therapy for a curable small volume refurther salvage therapy. One patient had a 5 cm.diw peritoneal relapse and he died. Also, Young et a l reported lung metastasis &r 68months, which was treated SU~O~SS- poor compliance with surveillance therapy after orchiectamy fully with chemotherapy and surgery but he suffered a brain alone in a small series of patients with clinical stage I nonmetastasis 27 months later (96 months after orchiectomy) seminomatous germ cell tumors of the testie.15 Furthermore, in our series lung metaetaaee were diagnosed earlier and at and died during salvage therapy. Table 3 illustrates aU metastatic events (initialand s u b a smaller size than retroperitoneal lesions. "his obrvation quent relapses) recorded in the entire series. If we include was significant, and we suppose that it depended on the the case of AFP elevation only as a putative unidentified greater di€ticultiesin evaluating the retroperitoneum due to retroperitoneal metastasis, retmperitoneal nodes were in- the relatively low accuracy of CT and lymphangiography,18 volved with metastases in 19%, representing nearly 50% of and because of our difE4ties in regularly scheduling theee examinations. all metastatic sites. Nevertheless, the overall recurrence rate in our eerieS was Secondary malignancies developed in 2 of the 85 patienta

1048

SURVEILLANCE IN CLINICAL STAGE I TESTICULAR CANCER

tumor marker elevation preceded a measurable lesion in only of 132 months, which agrees with results of other long-term 28% of our cases and 40% of the relapses were not accompasurveillance studies. The Medical Research Council Testicu- nied by any marker elevation at all. Similarly, of the 100 lar Tumors Working Party reported a 27% (16% in the ret- relapses (27%) recorded in a large British series 13 had roperitoneum) recurrence rate among 373 patients with clin- marker elevation only and 39 had no marker elevation at all.4 ical stage I nonseminoma who were recruitedin a prospective study and followed for an average of 5 years.' The Danish CONCLUSIONS Testicular Cancer Study Group reported a 30% (20% in the Surveillance after orchiectamy alone certainly represents a retroperitoneum) recurrence rate among 77 patients underlargely accepted alternative to nerve sparing retroperitoneal going surveillance with a median followup of 64 months.16 Recurrence rates of surveillance studies seem to compare lymphadenectomy for clinical stage I nonseminomatous germ favorably with metastatic rates r e p o d in the largest sur- cell tumors of the testis. Followingidentification of predictive gical series. Donohue et al reported 112 retroperitoneal me- factors for metastasis, high risk patients are undergoing tastases(29.6%) in 378clinical stage I nonseminoma patients adjuvant chemotherapy in several European countries with undergoing retmperitond lymphadenectomy between 1979 promising preliminary [email protected] Nevertheless, when seand 1989.2 Of the 266 patients with negative nodes 31 had lecting the management of clinical stage I nonseminomatous relapse, for an overall metastatic rate of 37.8%. All 378 testis cancer to date we also must consider (and know) the clinical stage I cancer patients underwent determination of basic questions of risk versus benefit and cost versus benefit tumor marker levels before lymphadenectomy and in 87% of each treatment modality in addition to the specific risk of the retroperitoneal disease was staged with CT. "he Swed- relapse in the individual patient. The Medical Research ish-Norwegian Testicular Cancer Group reported 75 retro- Council and the Swedish-Norwegian Testicular Cancer peritoneal metastases (26.9%) and 30 relapses (15%) in 204 Group studiesfailed to identify specific risk factors for either . ~ risk stage I patients with negative nodes among 279 with clinical stage I retroperitoneal or distant m e t a s t a s e ~ . ~Good nonseminomatous germ cell tumors of the testis undergoing cancer patients can be considered for surveillance but patient retroperitoneal lymph node dissection between 1981 and compliance and adequate availability of medical resources 1986.The final metastases rate was 37.6%. All patients un- are essential for a long observation program. High risk paderwent marker determination before lymphadenectomy, tients may be offered nerve sparing retroperitoneal lymphadenectomyz or 2 adjuvant courses of cisplatin-based chemoand 95% underwent CT of the abdomen and pelvis.6 This 8 to 10% difference in the metastatic rate between therapy.9-11 Fertility may be preserved2 or it may recoverg-11 clinical and surgical series is not easy to understand. Differ- following both treatment modalities but careful observation ent series of patients are not necessarily comparable because is necessary following lymphadenectomy to detect distant of possible referral bias, and some of the surveillance pa- metastases. Also, patients treated with adjuvant chemothertients with lung metastases might also have had retroperi- apy must be followed for possible recurrent teratoma,9.10 or t o n d tumors too small to be detected on CT but that re- occasional refractory carcinoma.1° Only a comparative study solved with chemotherapy (1 of our patients had a second among these 3 Merent treatment modalities in stratified relapse in the retroperitoneum after the initial lung relapse). risk patients could answer the basic questions on risk versua However, it cannot be excluded that, since patients entering benefit and cost versus benefit in the management of clinical surveillance are offered only expectant therapy, physicians stage I nonseminoma. are much more committed to clinical staging of these tumors Dr. Silvana pilotti reviewed the histological slides, thanin those who are to be treated actively. For instance, we dstrict criteria to defme clinical stage I cancer and, as a REFERENCES result, the overall percentage of negative nodes in clinical stage IIA cancer patients undergoing surgery during that 1. Pizzocaro, G.: Retroperitoneal lymphadenectomy in clinical period was as high as q096.17 Besides, Sturgeon et al noted stage I nonseminomatousgerminal testis cancer. Eur. J. Surg. Oncol., 12 25, 1986. the importance of accuracy in the interpretation of abdomi2. Donohue, J. P., Thornhill,J. A., Foster, R. S., Rowland, R. G. and nopelvic CT and lymphangiographic radiograms.1S In their Bihrle, R.: Retroperitoneal lymphadenectomyfor clinical stage surveillance study they found a 46.6% progression rate in the A testis cancer (1965to 1989): modifications of technique and first 46 patients and a 26.6% progression rate in the next 60 impact on ejaculation. J. Urol., 149 237, 1993. due to the different accuracies in evaluating CT and lym3. Peckham, M.J.,Barrett, A., Husband, J. E. and Hendry, W. F.: phangiography during the 2 periods. Orchidedomy alone in testicular stage I non-seminomatous Attention has been paid recently to late recurrences more germ cell tumours. Lancet, 2 678, 1982. than 2 years after orchiectomy.8 In our series there were 3 4. Read, G., Stenning, S. P., Cullen, M. H., Parkinson, M. C., late recurrences, representing 122 of all relapaes. The 2 Horwich, A, &ye, S. B. and Cook, P. A.: Medical Research Council prospective study of surveillance for stage I testicular retroperitoneal recurrences were 36 months after orchiecteratoma. Medical Research Council Testicular Tumors Worktomy. One recurrence was preceded by marker elevation and ing Party.J. Clin. Oncol., 1 0 1762,1992. with equivocal CT of the abdomen. The patient underwent 5. Klepp, O.,Olsson, A. M., Henrikson, H., Aass, N., Dahl, 0.. lymphadenectomy and 29 small metastatic nodes were Stenwig, A. E., Persson, B. E., Cavallin S M , E., Fossl, S. D. found. In 1 case a 14 cm.retroperitoneal mass was found 8 and Wahlqvist, L.: Prognostic factors in clinical stage I nonmonths after a normal CT. In 1 case a 5 cm.solitary lung seminomatous germ cell tumors of the testis: multivariate metastasis developed 68 months after orchiectomy. Chest analysis of a prospective multicenter study. Swedish-Norwex-rays were normal 6 months previously. This patient subsegian Testicular Cancer Group. J. Clin. Oncol., 8: 509, 1990. quently died of the disease. Late relapse has been observed 6. Pizzocam, G.,Salvioni, R. and Zanoni, F.: Unilateral lymphadenectomy in intraoperative stage I nonseminomatous germiby others in surveillance studies. After a median followup of nal testis cancer. J. Urol., 134: 485, 1985. 64 months, the Danish Testicular Cancer Study Group reP., Goldberg, S. D., Sturgeon, J. F. G., ported 4 late relapses (5%)in 77 patients, 1 of which occurred 7. Jewett, M.A.,Kang, Y.-S. Thomas,G. M., Alison, R. E. and Gospodarowicz, M. K: Re& at 62 months.16 Among the 373 patients recruited in the roperitoneal lymphadenectomy for testis tumor with nerveMedical Research Council prospective surveillance study 8 sparing for ejaculation. J. Urol., 139: 1220,1988. late relapses (2%) were reported.' Recent reports suggested 8. Lowe, B. A.: Surveillance versus nerve-sparing retroperitoneal that late recurrencea also are not uncommon after primary lymphadenectomy in stage I nonseminomatous germ-cell tuchemothera~y.~9.2'J Moreover, serum tumor markers seem to mors. Urol. Clin. N. Amer., u): 75,1993. be of little help for early diagnosis of relapse, since a serum 9. Studer, U.E.,Fey, M. F., Calderoni, k,Kraft. R.,MazzuccheUi

29.4% (19% in the retmperitoneum) after a median followup

SURVEILLANCE IN CLINICAL STAGE I TESTICULAR CANCER L. and Sonntag, R. W.: Adjuvant chemotherapy after orchiectomy in high-risk patients with clinical stage I non-seminomatous testicular cancer. Eur. Urol., 25: 444,1993. 10. Hoeltl, W., Pont, J., Kosak,D., Honetz, N. and Marberger, M.: Treatment decision for stage I non-eeminomatous germ cell tumours based on the risk factor 'vascular invasion". Brit. J. Urol., 6 9 83, 1992. 11. Cullen, M. H. on behalf of the Medical Research Council Teaticular Tumour Working Party: Short c ~ u r a eadjuvant chemotherapy in high risk stage I non-seminomatous gem cell tumows of the testis (NSCClT): an MRC study report. In: Advances in the Biosciences. Germ Cell Tumoura HI. Edited by E. G. Jones, P. H. Amden and I. Appleyard. Oxford: Pergamon, vol. 91, pp. 201-202, 1994. 12. Pizzocaro, G., Zanoni, F., Salvioni, R., Milani, A, Piva, L. and Pilotti, S.: Difficulties of a surveillance study omitting reperitoneal lymphadenectomy in clinical stage I nonaeminomatous germ cell tumors of the testis. J. Urol., 1% 1393, 1987. 13. Pizzocaro, G., Nicolai, N., Salvioni, R., Piva, L.,Faustini, M., Zanoni, F. and Milani, A: Comparison between clinical and pathological staging in low stage noneeminomatous germ cell testicular tumors. J. Urol., 148: 76, 1992. 14. Sesterhenn, 1. A., Weiss, R. B., Mostofi. F. E, Stablein, D.M., Rowland, R. G., Falkson, G., Rivkind, 5. E. and Vogelzang, N. J.: Prognosis and other clinical correlates of pathologic review in stage I and I1 testicular carcinoma: a report from the

1049

T d c u l a r Cancer Intergroup Study. J. Clin. Oncol.. 101 69, 1992. 16. Young,B. J., Bulk, B. D., Rueaell. J. A and Trew, M.6.:Com-

pliance with follow-up of patients treated for nonaeminomatow testicular cancsr. Brit. J. Cancer, 6& 806,1991. 16. Mrth, M.,Jawbeen, G. E, von der Maaee, H., Madsen, E. L.. Nieleen, 0. S.,Pedenen. M. and Schultz, H.:S u r v e i h m done v e m radiothereipy after orchieaomy for clinical I noneeminomatma testiculnr cnncer. Danieh Teati& Cancer Study Croup. J. Clia Oncol.. 9: 1543.1991. 17. Pizzoeam. G.: Retroperitoneal lymph node dierection in clinical stage IIA and IIB noneeminomatow germ cell tumours of the teetie. Int. J. A n b l . , 10: 269, 1987. 18. Sturgeon, J. F.. Jewett, M. A, Alieon, R E.,G o s p h ~ i C C . M. K., Blend, R, Herman, S., Richmond, H., " h ~G.,, Duncan, W. and Munro, A: Surveillance after orchidectomy for patients with clinieal &age I noneeminomatous testie tumors. J. Clin. Onwl.. 10: 684,1992. 19. Nichols, C., Baniel, J., Foster, R.S..Donohue, J. P.and Einhorn, L.H.:Late relapee of germ cell turnore.Proc.Amer. 8oe. C h Oncol., 19: 234,abstract 722,1904. 20. Gerl, A, Clemm. C., Hartaneteia,R, Kohl,P.and W h M M . w.: Late relapee of non eeminomatou cell tumorn (NSGCT) after &platin baeed chemotherapy. PNIC.Amer. SOC. Clin. Oncol., 1 s 229. abetract 703,1994.