0022-534 7/87 /1386-1393$02.00/0 Vol. 138, December Printed in U.S.A.
THE JOURNAL OF UROLOGY
Copyright© 1987 by The Williams & Wilkins Co.
DIFFICULTIES OF A SURVEILLANCE STUDY OMITTING RETROPERITONEAL LYMPHADENECTOMY IN CLINICAL STAGE I NONSEMINOMATOUS GERM CELL TUMORS OF THE TESTIS GIORGIO PIZZOCARO, FULVIO ZANONI, ROBERTO SALVIONI, ANGELO MILANI, LUIGI PIVA AND SILVANA PILOTTI From the Section of Urologic Oncology and Department of Pathology, Istituto Nazionale per lo Studio e la Cura dei Tumori, Milano, Italy
Between August 1981 and December 1984, 85 consecutive patients with clinical stage I nonseminomatous germ cell tumors of the testis who were suitable for close observation entered a surveillance study after orchiectomy alone. All patients had unequivocally negative chest x-ray, bipedal lymphangiography, and computerized tomography of the abdomen and pelvis, and normal levels of alphafetoprotein and human chorionic gonadotropin before entering the study. Patients were followed closely for 24 to 64 months (median 42 months) with regular chest x-rays, plain films of the abdomen for lymphangiography control, and serum determinations of alpha-fetoprotein and human chorionic gonadotropin but it was difficult to obtain computerized tomography scans of the abdomen at scheduled intervals for such a long period. Followup was closed December 31, 1986. At that date 62 patients (73 per cent) were continuously free of disease after orchiectomy alone and 23 (27 per cent) suffered relapse. · The over-all occurrence rate of retroperitoneal relapses was 16.5 per cent and they usually were detected late, 4 to 36 months (median 10 months) after orchiectomy. Lung metastases were detected much earlier, 2 to 10 months (median 3 months) after orchiectomy. Alpha-fetoprotein and human chorionic gonadotropin elevations preceded the radiographic demonstration of metastases in 8 patients only (35 per cent) and in 1 they were the only sign of relapse. All but 1 patient with relapse were cured with chemotherapy and/or surgery, with an over-all survival rate free of disease of 98.8 per cent. Invasion of the epididymis, rete testis and spermatic cord, primary scrotal surgery, peritumor vascular invasion and embryonal carcinoma were associated with a higher risk for relapse but it was impossible to find clear-cut indications to select patients for adjuvant chemotherapy, retroperitoneal lymphadenectomy or no treatment. Furthermore, the followup of retroperitoneal nodes proved to be much more difficult than expected. Unilateral or modified retroperitoneal lymphadenectomy facilitates management of clinical stage I nonseminomatous germ cell tumors of the testis: only the chest and markers must be followed, the status of the retroperitoneal nodes is known immediately and antegrade ejaculation is preserved in the majority of cases. (J. Ural., 138: 1393-1396, 1987) Surveillance after orchiectomy alone has gained great popularity in the management of clinical stage I nonseminomatous germ cell tumors of the testis. Our purposes were to avoid unnecessary treatment in approximately 70 per cent of the patients who can be cured by orchiectomy alone and to select patients at a high risk of relapse who may benefit from immediate post-orchiectomy therapy. 1 Sine-qua-non conditions to perform the study are careful clinical staging and close followup to detect metastases early, when they are easily curable. Preliminary results were enthusiastic, with reports of 100 per cent survival and 16 to 40 per cent relapse rates. 1- 5 With a longer followup a few deaths have been reported and it became evident that relapses in the retroperitoneal nodes could occur late. 6- 10 We present the end results of our surveillance study in stage I nonseminomatous germ cell tumors of the testis, and we discuss the difficulties of a close and prolonged followup of the retroperitoneal nodes. Preliminary results have been reported previously. 9 •10 PATIENTS AND METHODS
Between August 1981 and December 1984, 90 consecutive patients with clinical stage I nonseminomatous germ cell tuAccepted for publication May 6, 1987, Supported in part by Grant 86.00701.44 Special Project "Oncology" from the Consiglio Nazionale delle Ricerche (C.N.R.), Rome, Italy.
mors of the testis were admitted to our institute. Five patients were not fit for the surveillance study because of a previous contralateral testis tumor (2), synchronous thyroid cancer (1), refusal (1) and a course of chemotherapy given for a falsely increased a-fetoprotein level soon after orchiectomy (1). The remaining 85 patients were suitable for close observation. All patients entered the surveillance study after fully informed consent was obtained. Median patient age was 26 years, with a range of 2 months to 65 years. Only 9 patients (10 per cent) underwent orchiectomy at the institute. The other 76 patients had been referred for clinical staging and further treatment after orchiectomy performed elsewhere. In each case clinical records and slides were reviewed. Histological diagnosis was reassessed in all patients but the T category and peritumoral vascular invasion (V category) could be defined exactly in only 54 and 28 patients (63 and 33 per cent), respectively. Pre-orchiectomy serum titers of a-fetoprotein and of the /)-subunit of human chorionic gonadotropin were available in only 43 patients (50.5 per cent). Sixty-five patients (76 per cent) had undergone primary radical inguinal orchiectomy. The remaining 20 patients (24 per cent) had had transscrotal biopsy or orchiectomy as the first surgical approach. These patients underwent a radical inguinal operation with en bloc removal of the scrotal scar before entering the surveillance study. In all cases the clinical
PIZZOCARO AND ASSOCIATES
staging was based on history, physical examination, post-orchiectomy serum tumor markers, half-life kinetics for a-fetoprotein and human chorionic gonadotropin, posteroanterior and lateral chest x-rays, bipedal lymphangiography, and computerized tomography (CT) of the abdomen and pelvis. Criteria for inclusion in the study were unequivocal absence of any clinical evidence of metastases, including negative serum tumor markers or postoperative half-life kinetics values less than 5 days for a-fetoprotein and less than 2 days for human chorionic gonadotropin, availability for close followup, no second tumor and no further treatment after orchiectomy. No patient had pathological evidence of tumor at the cut end of the spermatic cord. Patients were observed monthly for 1 year, bimonthly for 1 year and every 3 months thereafter. Serum tumor markers of a-fetoprotein and human chorionic gonadotropin, chest x-ray and plain films of the abdomen were repeated at each visit. Lymphangiography was repeated at least once at the disappearance of the contrast medium, usually between 6 and 12 months after orchiectomy. CT of the abdomen and pelvis was scheduled every 2 visits. Patients with recurrent disease entered the ongoing protocols for metastatic disease. Observation was closed on December 31, 1986.
1. Relapse pattern of clinical stage I nonseminomatous germ
cell tumor of the testis Site of First Relapse
No. Pts. With Relapse
% of 85 Pts.
% of 23 Relapses
Scrotum Retroperitoneal nodes only Retroperitoneal and distant Lung only (\'-Fetoprotein levels increased only Totals
1 10 3 8 1
1.2 11.7 3.5 9.4 1.2 27.0
4.3 43.5 13.0 34.8 4.3 100.0
2. Size of metastases at diagnosis and interval free of disease by anatomical sites Metastatic Sites
Retroperitoneal nodes Lung a-Fetoprotein levels increased only
Mos. <2 Cm. 2-5 Cm. >5 Cm. Free of Disease (range) 1 10
3 (2-10) 22 ~ TABLE
Prognostic Factors Duration of symptoms (mos.): l Mos. between orchiectomy and stagin1;: l Type of primary surgery: Inguinal Transscrotal Histology: Embryonal Ca Teratoma ± others Yolk sac tumor T category: Tl-T2 T3-T4a Pre-orchiectomy (\'-fetoprotein and human chorionic gonadotropin: Normal Elevated Peritumor vascular invasion: Absent Present Tumor necrosis: Absent Present
No patient was lost to followup but it was difficult to obtain CT scans of the abdomen at the scheduled times in nearly all patients. After a median observation of 42 months (range 24 to 64 months) 23 patients (27 per cent) had metastases and 62 (73 per cent) remained continuously free of disease. Metastases were diagnosed clinically from 2 to 36 months after orchiectomy. The median interval free of disease was 6 months. Of 23 patients 19 (83 per cent) had recurrence within 1 year, while 4 had evidence of metastases at 17, 22, 24 and 36 months after orchiectomy. The site of the first relapse is illustrated in table 1. The patient with the scrotal relapse had had a transscrotal biopsy followed by a radical inguinal operation performed elsewere. Lung metastases developed 5 months later. Two patients with lung metastases had further recurrences, including 1 in the retroperitoneal nodes. To date 14 of the 85 patients (16.5 per cent) have had retroperitoneal lymph node metastases. The size of the metastases and the interval free of disease are detailed in table 2. It is evident that lung metastases were detected much earlier than retroperitoneal metastases. In particular, all late relapses occurred in the retroperitoneal nodes and, probably, the patient with a late afetoprotein elevation also had a retroperitoneal relapse. The elevation of serum tumor markers preceded the radiographic evidence of metastases in only 8 patients (35 per cent). The duration of symptoms before orchiectomy, interval between orchiectomy and clinical staging, type of primary operation, histology, T category, pre-orchiectomy values of a-fetoprotein and human chorionic gonadotropin, peritumor vascular invasion and tumor necrosis were considered for univariate analysis (table 3). Infiltration of rete testis, epididymis or spermatic cord, primary transscrotal surgery and peritumor intravascular invasion were associated with a significantly increased risk of relapse. Embryonal carcinoma also probably was associated with a higher risk of relapse than teratocarcinoma. The duration of symptoms, interval between orchiectomy and clinical staging, preoperative serum titers of a-fetoprotein and human chorionic gonadotropin, and tumor necrosis did not modify the risk for relapse significantly. A multivariate analysis could not be done because T and V categories were known only in part of the series. However, 1 or more risk factors were present in 17 of 23 patients with relapse, and in 5 of the remaining 6 patients the T and V categories were not known. Relapse occurred in 1 of 9 patients with no risk factor present. Furthermore, no correlation could be found between risk factors and the site of the first relapse.
3. Univariate analysis of prognostic factors No. Relapses/No. Cases(%)
7/34 (20.6) 16/51 (31.4)
5/38 (13.2) 8/16 (50.0)
1/4 (25.0) 10/39 (25.6)
2/18 (11.1) 5/10 (50.0)
1/8 (12.5) 7/26 (26.9)
15/43 (34.9) 8/42 (19.0) 13/65 (20.0) 10/20 (50.0) 10/24 (41.7) 13/58 (22.4) 0/3
RETROPERITONEAL LYMPHADENECTOMY IN TESTIS TUMORS
The patient with minimal retroperitoneal relapse refused therapy and died of widespread disease 7 months later (11 months after orchiectomy). The other 22 patients were rendered free of disease with chemotherapy and/or an operation. Nineteen patients remained continuously free of disease 10 to 54 months (median 32 months) after treatment of the first relapse. The 3 patients who had further relapses were free of disease 26, 29 and 52 months, respectively, after treatment of the last relapse. The over-all survival rate free of disease was 98.8 per cent (84 of 85 patients). DISCUSSION
Despite excellent results we are not satisfied with our surveillance study in clinical stage I nonseminomatous germ cell tumors of the testis. When the study began we supposed that nearly all metastases that had passed unrecognized at careful clinical staging would have been detected by careful followup within a few months. This was true only for lung metastases but not for retroperitoneal metastases, which often were documented late and large in size (table 2). Late relapses occurred in the retroperitoneal nodes also in the Royal Marsden 11 and Danish 12 studies, respectively, 4 and 5 years after orchiectomy. A careful followup with frequent CT scans of the abdomen and pelvis for such a long interval is extremely difficult and it probably is impractical for economic and psychological reasons. As soon as we realized that a much closer and longer observation was necessary for the retroperitoneal nodes, we experienced severe difficulties with our patients. Not only could they not support the cost of frequent CT scans for several years but they also felt insecure when they realized that the period at risk was much longer than had been expected. A type of psychological toxicity developed and the doctor-patient relationship became more difficult. · Our purpose was to detect factors predicting relapses to select patients for adjuvant chemotherapy (high risk for distant metastases), retroperitoneal lymphadenectomy (risk for retroperitoneal metastases only) and close observation (almost no risk for metastases). We were unable to detect clear-cut risk factors to select patients for further treatment or no treatment. Many more cases are necessary for statistical evaluation. Furthermore, all recognized risk factors should be known in every patient to perform a multivariate analysis. This is not possible at a referral center, where 90 per cent of the patients present after orchiectomy. In a larger series Hoskin and associates demonstrated that histology and lymphatic invasion were significant independent prognostic factors. 11 However, they conclude that "it is premature to advocate immediate post-orchiectomy chemotherapy until larger numbers of observation are available to confirm the present findings". Furthermore, we observed 1 relapse among 9 patients with all favorable prognostic factors present. If a 10 per cent relapse rate will be confirmed in patients with favorable prognosis, a careful and prolonged followup also must be maintained in these patients if orchiectomy alone is performed. The retroperitoneal nodes are the weak point of surveillance for clinical stage I nonseminomatous germ cell tumors of the testis, and serum tumor markers of a-fetoprotein and human chorionic gonadotropin are of little help in detecting early metastases. As soon as we realized the impracticality of following patients with repeated lymphangiography and CT scans for 3 to 5 years, we closed our surveillance study, and beginning in January 1985 we returned to retroperitoneal lymphadenectomy for clinical stage I nonseminomatous germ cell tumors of the testis. Unilateral lymphadenectomy is performed in every patient with negative intraoperative findings. 13 Thus, antegrade ejaculation is preserved in the majority of cases and followup is much simpler, since only serum tumor markers and the chest x-rays must be followed. Furthermore, prognostic factors also can be studied easily in surgically resected clinical stage I nonseminomatous germ cell tumors of the testis, with the dual
advantages of immediate knowledge of the retroperitoneal nodes and of an easier followup. 14•15 REFERENCES
1. Peckham, M. J., Barrett, A., Husband, J. E. and Hendry, W. F.: Orchidectomy alone in testicular stage I non-seminomatous germ cell tumours. Lancet, 2: 678, 1982. 2. Read, G., Johnson, R. J., Wilkinson, P. M. and Eddleston, B.: Prospective study of followup alone in stage I teratoma of the testis. Brit. Med. J., 287: 1503, 1983. 3. Sogani, P. C., Whitmore, W. F., Jr., Herr, H. W., Bosl, G. J., Golbey, R. B., Watson, R. C. and DeCosse, J. J.: Orchiectomy alone in the treatment of clinical stage I nonseminomatous germ cell tumor of the testis. J. Clin. Oncol., 2: 267, 1984. 4. Johnson, D. E., Lo, R. K., von Eschenbach, A. C. and Swanson, D. A.: Surveillance alone for patients with clinical stage I nonseminomatous germ cell tumors of the testis: preliminary results. J. Urol., 131: 491, 1984. 5. Jewett, M. A. S., Herman, J. G., Sturgeon, J. F. G., Comisarow, R. H., Alison, R. E. and Gaspodarowicz, M. K.: Expectant therapy for clinical stage A nonseminomatous germ-cell testicular cancer? Maybe. World J. Urol., 2: 57, 1984. 6. Peckham, M. J.: Orchidectomy for clinical stage I testicular cancer: progress report of the Royal Marsden Hospital study. J. Roy. Soc. Med., suppl. 6, 78: 41, 1985. 7. Sogani, P. C., Whitmore, W. F., Jr., Herr, H. W., Morse, M. J., Bosl, G. and Fair, W.R.: Long term experience with orchiectomy alone in treatment of clinical stage I nonseminomatous germ cell tumors of testis (NSGCTT). J. Urol., part 2, 133: 246A, abstract 532, 1985. 8. Schultz, H. P., R111rth, M., van der Maase, H., Sandberg, E. and Pedersen, M.: The Danish Testicular Carcinoma Study. Results from recent clinical trials in non-seminomatous tumours. Acta Urol. Belg., 53: 362, 1985. 9. Pizzocaro, G., Zanoni, F., Milani, A., Salvioni, R., Piva, L., Pilotti, S., Bombardieri, E., Tesoro-Tess, J. D. and Musumeci, R.: Orchiectomy alone in clinical stage I nonseminomatous testis cancer: a critical appraisal. J. Clin. Oncol., 4: 35, 1986. 10. Pizzocaro, G., Zanoni, F., Salvioni, R., Milani, A. and Piva, L.: Surveillance or lymph node dissection in clinical stage I nonseminomatous germinal testis cancer? Brit. J. Urol., 57: 759, 1985. 11. Hoskin, P., Dilly, S., Easton, D., Horwick, A., Hendry, W. and Peckham, M. J.: Prognostic factors in stage I non-seminomatous germ-cell testicular tumors managed by orchiectomy and surveillance: implications for adjuvant chemotherapy. J. Clin. Oncol., 4: 1031, 1986. 12. R~rth, M., von der Maase, H., Nielsen, E. S., Pedersen, M. and Schultz, H.: Orchidectomy alone versus orchidectomy plus radiotherapy in stage I nonseminomatous testicular cancer: a randomized study by the Danish Testicular Carcinoma Study Group. Int. J. Androl., 10: 255, 1987. 13. Pizzocaro, G., Salvioni, R. and Zanoni, F.: Unilateral lymphadenectomy in intraoperative stage I nonseminomatous germinal testis cancer. J. Urol., 134: 485, 1985. 14. Klepp, 0., Fossa, S., Fritjofsson, A., Henrikson, H., MaartmanMoe, H., Persson, B. E., Ranstam, J. and Wicklund, H.: Risk factors for metastases in clinical stage 1 (CS1) nonseminomatous germ cell testicular cancer (NSGT). Read at the 3rd European Conference on Clinical Oncology and Cancer Nursing, abstract 200, Stockholm, June 16-20, 1985. 15. Javadpour, N., Canning, D. A., O'Connell, K. J. and Young, J. D.: Predictors of recurrent clinical stage I nonseminomatous testicular cancer. A prospective clinicopathologic study. Urology, 27: 508, 1986.
EDITORIAL COMMENT This is an important and timely article concerning the problems of a surveillance protocol for patients with clinical stage I nonseminomatous germ cell tumors of the testis. The clinical practice of surveillance has evolved because of the effectiveness of chemotherapy and the capability to salvage patients with stage I disease who have relapse. There currently are at least 5 studies reporting the results of active surveillance protocols, with relapse rates of 20 to 30 per cent but longterm salvage of most patients. In the current study of a carefully selected population with a negative