Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Incidence of Metachronous Testicular Cancer in Patients With Extragonadal Germ Cell Tumors

Incidence of Metachronous Testicular Cancer in Patients With Extragonadal Germ Cell Tumors Abstract Background: The frequency of subsequent testicular cancer (referred to as metachronous testicular cancer) in men who have had previous testicular cancer is relatively high. The rate of metachronous testicular cancer in men with extragonadal germ cell tumors (EGCTs), however, is largely unknown. We conducted a retrospective study of EGCT patients to determine the incidence, cumulative risk, and specific risk factors for metachronous testicular cancers. Methods: A standardized questionnaire about patient characteristics, the extent of EGCT disease, any second malignancies, and treatments received was completed for 635 patients with EGCTs identified from the medical records of 11 cancer centers in Europe and the United States from 1975 through 1996. Comparisons with age group-specific data from the Saarland, Germany, population-based cancer registry were used to calculate the standardized incidence ratio (SIR). The Kaplan-Meier method was used to analyze survival data and cumulative risk. All statistical tests were two-sided. Results: Sixteen EGCT patients (4.1%) developed metachronous testicular cancers, with a median time between diagnoses of 60 months (range, 14–102 months). The risk of developing metachronous testicular cancers was statistically significantly increased in patients with EGCTs (observed = 16; expected = 0.26; SIR = 62; 95% confidence interval [CI] = 36 to 99) and in subsets of EGCT patients with mediastinal location (SIR = 31; 95% CI = 8 to 59), retroperitoneal location (SIR = 100; 95% CI = 54 to 172), and nonseminomatous histology (SIR = 75; 95% CI = 43 to 123). The cumulative risk of developing a metachronous testicular cancer 10 years after a diagnosis of EGCT was 10.3% (95% CI = 4.9% to 15.6%) and was higher among patients with nonseminomatous EGCTs (14.3%; 95% CI = 6.7% to 21.9%) and retroperitoneal EGCTs (14.2%; 95% CI = 5.6% to 22.8%) than among patients with seminomatous EGCTs (1.4%; 95% CI = 0.0% to 4.2%) and mediastinal EGCTs (6.2%; 95% CI = 0.1% to 12.2%). Conclusions: Patients with EGCTs, particularly those with retroperitoneal or nonseminomatous tumors, but also those with primary mediastinal EGCTs, are at an increased risk of metachronous testicular cancer. Testicular cancer is the most common malignancy in men aged 15–35 years (1,2). Approximately 5%–7% of these germ cell tumors are of extragonadal origin (3), with the most common sites being the mediastinum and the retroperitoneum. Extragonadal germ cell tumors (EGCTs) are histologically similar to gonadal germ cell tumors; however, there is no testicular abnormality detectable either by palpation or by high-resolution ultrasonography in patients with EGCTs. Twenty-five years after initial treatment for testicular cancer, the cumulative risk of developing a contralateral testicular cancer is approximately 5%. The risk is higher for patients with nonseminomatous cancers or for patients younger than 30 years of age at the time of diagnosis (4,5). Overall, the relative risk of developing a second testicular cancer is 25-fold higher than in the age-matched population. The relative risk increase is probably a result of the existence of a precursor lesion for testicular cancer (the so-called “testicular intraepithelial neoplasia” [TIN] lesion) in testicular cancer patients. The assumed prevalence of a contralateral TIN lesion in patients with testicular cancer is approximately 5% (6). Although a precursor lesion for the development of EGCTs has not been established, the occurrence of TIN lesions has been observed in 20%–40% of patients with retroperitoneal EGCTs (7) but is very rare in patients with a primary mediastinal EGCT (8). To our knowledge, the subsequent development of testicular cancer (referred to as metachronous testicular cancer) after the patient has been given a diagnosis of a retroperitoneal or mediastinal EGCT has been reported for only 12 patients (9–15). Assessing the incidence of metachronous testicular cancer in patients with EGCTs is complicated by the fact that both events are rare. In this study, we reviewed a large, multicenter database that included information on 635 patients with EGCTs who had been treated during the cisplatin-based chemotherapy era to determine the incidence of metachronous testicular cancer in patients with EGCTs. Patients and methods Data Collection We evaluated the medical records of all patients with EGCTs consecutively diagnosed and/or treated at 11 cancer centers in the United States and Europe from 1975 through 1996. The contributing centers and their locations, the population sample at each center, and the time period for each center are, respectively, as follows: Indiana University, Indianapolis, IN (n = 216, 1989 through 1996); Institut Gustave-Roussy, Villejuif, France, and Centre Léon-Berard, Groupe d'Etude des Tumeurs Urologiques et Genitales, Lyon, France (n = 93, 1975 through 1996); Medical Center II, Eberhard-Karls-University Tuebingen, Germany (n = 13, 1986 through 1993); Hannover University Medical School, Hannover, Germany (n = 88, 1978 through 1995); The Norwegian Radium Hospital, Oslo, Norway (n = 48, 1980 through 1995); Klinikum Groβhadern, Munich, Germany (n = 63, 1979 through 1996); The Royal Marsden Hospital, Sutton, U.K. (n = 65, 1979 through 1994); Kaiser-Franz Josef Spital, Vienna, Austria (n = 19, 1975 through 1996); and Virchow-Klinikum, Berlin, Germany (n = 30, 1987 through 1994). Patients were referred to the participating cancer centers because of their diagnosis of EGCT. The inclusion criterion for this study was evidence of an EGCT, defined as the presence of an extragonadal mass in the mediastinum, retroperitoneum, or elsewhere in the absence of testicular abnormalities at physical examination or ultrasonography. If testicular abnormalities were noted, a biopsy specimen was collected to exclude a diagnosis of primary testicular cancer. A TIN lesion or a scar found at the time the biopsy specimen was taken was not a cause for exclusion from the study. Overall, 637 patients were identified for this study, and 635 patients were eligible. Two patients were excluded from the analysis because chart review revealed invasive testicular cancer. We obtained the status of all living patients as of February 1998. For the data collection, a standardized questionnaire was sent to each center, where it was completed by one of the coinvestigators. We acquired detailed information on patient characteristics, such as the location and histology of the primary EGCT; the extent of the disease, including serum tumor marker concentrations of human β-chorionic gonadotropin (β-HCG), α-fetoprotein (AFP), and lactate dehydrogenase (LDH); history of testicular abnormalities; details on diagnostic methods, treatments, responses to treatment, and follow-up periods; and second testicular cancers, other second cancers, or treatment-related leukemia. We collected information regarding levels of β-HCG, AFP, and LDH because elevation of AFP and/or β-HCG levels with evidence of a mediastinal or retroperitoneal mass is considered specific for the diagnosis of a nonseminomatous EGCT. These tumor markers allow distinctions to be made between nonseminomatous EGCTs (elevated β-HCG and AFP) and other mediastinal or retroperitoneal tumors, including seminomatous germ cell tumors (no elevated markers). The completed questionnaires were checked for plausibility and data consistency at Eberhard-Karls-University Tuebingen Medical Center. The questionnaires were returned to the principal investigator at each center if any important data were missing. For this report, the clinical records of all patients with EGCTs who developed metachronous testicular cancer were reviewed in detail. All histopathologic slides of these cancers were diagnosed and classified by each center's pathology department. Because all patients' data and chart reviews were obtained anonymously and retrospectively, no institutional approval was necessary from any of the participating institutions. Statistical Analysis The duration of follow-up was calculated from the date of the EGCT diagnosis. Survival calculations were performed from the time of metachronous testicular cancer diagnosis and from the time of EGCT diagnosis. All data were entered in a personal computer at the Eberhard-Karls-University Tuebingen, Medical Center II, Germany. To calculate the standard incidence ratio (SIR), data were transferred to the Institute for Medical Information Processing, Eberhard-Karls-University Tuebingen. To determine the number of new cases of metachronous testicular cancer expected in the study group, we used the age group-specific data (classified in 5-year age groups) of the cancer registry of Saarland, Germany, from 1975 through 1996. This cancer registry is a population-based registry that covers the federal state of Saarland located in the southwest region of Germany (16). Results were standardized on the basis of age and follow-up duration (patients' years of risk) and expressed as the SIR with the associated 95% confidence intervals (CIs), which were calculated on the assumption of a Poisson distribution (17). The SIR was considered to be statistically significant if the 95% CI did not include the value 1.0. The SIR calculation was performed with the use of the SAS system (SAS, version 6.11 for Windows; SAS Institute, Inc., Cary, NC). The patients' years at risk, the expected number of cases for every age group, and the SIR were calculated according to Breslow and Day (17). All other statistical analyses were performed with the use of the SPSS system (SPSS for Windows, version 8.0; SPSS Inc., Chicago, IL). The Kaplan–Meier method was used to calculate survival data and to determine the cumulative risk of developing metachronous testicular cancer (18). Cox proportional hazards models were performed to identify variables that could predict the occurrence of metachronous testicular cancers in patients with EGCTs, including categorical variables, such as histology of the primary tumor, age grouping (25 years and older versus younger than 25 years), serum tumor marker concentrations at diagnosis (elevated; yes versus no), presence of additional metastatic sites (yes versus no), and type of treatment, and continuous variables, such as age and tumor marker concentrations at diagnosis (19). All statistical tests were two-sided. Results Because the occurrence of metachronous testicular cancer in patients with EGCTs is a rare event, we identified patients from 11 cancer centers throughout Europe and the United States. In total, 635 patients with EGCTs, with a median age of 30 years (range, 14–79 years), were included in the analysis. Of these 635 patients, 341 (53.7%) had primary mediastinal EGCTs, 283 (44.6%) had primary retroperitoneal EGCTs, and one had primary cervical lymph node involvement. The primary tumor location could not be determined in 10 patients (1.6%) because of widespread disease (including pulmonary metastases) and simultaneous midline involvement of the mediastinum and the retroperitoneum. Histology reports were available for 628 patients, of whom 524 (82.5%) had nonseminomatous EGCTs and 104 (16.4%) had seminomatous EGCTs. Tumor histology was not available for seven patients (1.1%) who had elevated AFP and/or β-HCG levels. Of the 635 patients with EGCTs, 14 (2.2%) received unilateral orchiectomy before or at the end of the treatment for the EGCT. None of the patients underwent bilateral orchiectomy (Table 1). Median follow-up time for surviving patients was 55 months (95% CI = 50 to 60 months). Table 2 classifies all patients according to site of disease and histology. Sixteen of 635 patients developed metachronous testicular cancer, with a median time interval between both diagnoses of 60 months (range, 14–102 months). Another patient developed an EGCT after having testicular cancer. Of the 283 patients who had had a primary retroperitoneal EGCT, 12 (4.2%) developed metachronous testicular cancer (Tables 1 and 2), with a median time interval between both diagnoses of 42 months (95% CI = 27 to 56 months). By contrast, of 341 patients who had had a primary mediastinal EGCT, four (1.2%) developed metachronous testicular cancer, with a median time interval between both diagnoses of 69 months (95% CI = 50 to 88 months) (Table 3). Two of these four patients had 12-cm- and 10-cm-sized mediastinal masses at initial presentation, which was the only disease presentation. The two remaining patients had a mediastinal tumor mass and metastatic disease extending to the abdominal lymph nodes outside the midline structures, with either lung metastases or cervical lymph node/thyroid gland involvement. The median age at diagnosis of EGCT in patients developing metachronous testicular cancer was 29 years (range, 19–49). Serum concentrations of the tumor markers AFP, β-HCG, and LDH were elevated in eight, six, and nine of the 16 patients, respectively (Tables 1 and 3). Of the 16 patients who developed metachronous testicular cancer, the majority were initially diagnosed with nonseminomatous EGCTs (n = 15 [94%]) (Table 3). By contrast, the majority of patients who developed metachronous testicular cancer were diagnosed with classical seminoma (n = 11 [69%]). Three (19%) patients had mixed nonseminomatous germ cell tumors, one patient had embryonal cell carcinoma, and one had differentiated teratoma (6% each). The tumor histologies for the EGCT and subsequent metachronous testicular cancer were comparable for one patient, with both tumors containing mixed nonseminomatous germ cell tumor with embryonal cell carcinoma, seminomatous, and teratomatous components. In this patient, the time interval between diagnosis of EGCT and the occurrence of the metachronous testicular cancer was 84 months. Five of 16 patients had had a bilateral testicular biopsy before treatment for the EGCT. The biopsy specimens from four of five patients were found to have normal histology. One patient, who was diagnosed with a TIN lesion, had a primary mediastinal seminoma and developed a subsequent metachronous testicular cancer 41 months after the completion of chemotherapy. Of the 16 patients with metachronous testicular cancer, all had previously received cisplatin-based combination chemotherapy (cisplatin, etoposide, and bleomycin, n = 10; cisplatin, vinblastine, and bleomycin, n = 3; etoposide, cisplatin, bleomycin, andcyclophosphamide, n = 1; cisplatin, etoposide, and ifosfamide, n = 1; carboplatin, vinblastine, and bleomycin, n=1) for the treatment of their EGCT, and 11 were allocated to postchemotherapy surgery because of residual tumor masses (Table 1). The response status could be evaluated in 15 patients. All patients had attained complete remission after chemotherapy and/or secondary surgery. Patients' characteristics are summarized in detail in Tables 1 and 3. Treatment for metachronous testicular cancer consisted of orchiectomy alone for those with stage I disease (11 of the 16 patients with metachronous testicular cancer [69%]) and orchiectomy plus chemotherapy and/or secondary surgery for those with metastatic disease of the metachronous testicular cancers (five patients [31%]). After a median follow-up time of 51 months (range, 1–154 months) after treatment for metachronous testicular cancer, all 16 patients were alive, without evidence of disease. The median overall survival after diagnosis of EGCT was 109 months (range, 46–203 months). In comparison with an age-matched population, there was a statistically significant risk of developing metachronous testicular cancer among the 635 patients with EGCTs during 2089 person-years at risk (observed [O] = 16; expected [E] = 0.26; SIR = 62; 95% CI = 36 to 99). Only 0.26 cases of de novo testicular cancer were expected during the study time period for a population this size. In addition, the risk of developing metachronous testicular cancers was statistically significantly increased for the following subsets of EGCT patients: those with primary mediastinal EGCTs (SIR = 31 [O = 4; E = 0.13; 95% CI = 8 to 59]); those with primary retroperitoneal EGCTs (SIR = 100 [O = 12; E = 0.12; 95% CI = 54 to 172]); and those with nonseminomatous EGCTs (SIR = 75 [O = 15; E = 0.2; 95% CI = 43 to 123]). Patients with seminomatous EGCTs did not have an increased risk of developing metachronous testicular cancers (SIR = 20; [O = 1; E = 0.05; 95% CI = 0.5 to 111]). No particular characteristics such as patient age, type of treatment received, histology of the primary tumor, presence of additional metastatic sites, or serum tumor marker concentrations at EGCT diagnosis could be identified that predicted the occurrence of metachronous testicular cancer in patients with EGCTs. We used the Kaplan–Meier method to estimate the cumulative risk of developing metachronous testicular cancer within 10 years after the initial diagnosis of the EGCT and found that the cumulative risk was 10.3% (95% CI = 4.9% to 15.6%). The cumulative risk of developing metachronous testicular cancer was 14.3% (95% CI = 6.7% to 21.9%) if the EGCT was nonseminomatous and 1.4% (95% CI = 0.0% to 4.2%) if the EGCT was seminomatous (P = .10). A retroperitoneal EGCT was associated with a cumulative risk of 14.2% (95% CI = 5.6% to 22.8%), and a primary mediastinal location was associated with a risk of 6.2% (95% CI = 0.1% to 12.2%), a statistically nonsignificant difference (P = .18). Inspection of the cumulative risk estimates suggested that there might be two time periods during which patients with EGCTs seem more likely to develop metachronous testicular cancers; at approximately 2.5–4 years and at approximately 6–7.5 years after the EGCT diagnosis (Fig. 1). Discussion The occurrence of metachronous testicular cancer in patients with retroperitoneal and mediastinal EGCTs appears to be an infrequent event, with only 16 of 635 patients who were previously diagnosed and effectively treated for EGCTs developing metachronous testicular cancer. There are only 12 reported cases of metachronous testicular cancer in patients with EGCTs available from the literature (9–15). This report is the first systematic investigation to provide data about the incidence and cumulative risks of the development of metachronous testicular cancer after treatment for an EGCT in a representative large cohort of patients by comparison with an age-matched general population. In addition, this investigation tries to identify specific risk factors for the development of subsequent metachronous testicular cancers in patients with EGCTs. The average time between diagnosis of EGCT and diagnosis of metachronous testicular cancer in the 12 patients reported in the literature (9–15) was approximately 7–8 years (range, 35–168 months). However, we observed a median time of 5 years between the two diagnoses (range, 19–142 months). Our results may not conflict with those reported in the literature (9–15), because it appears that there might be two time periods, the first at approximately 2.5–4 years and the second at approximately 6–7.5 years, during which patients are at an increased risk of developing metachronous testicular cancer (Fig. 1). Our results are similar in two ways to those of the 12 cases reported in the literature to date (9–15). First, the majority of patients had nonseminomatous EGCTs before subsequently developing metachronous testicular cancers of seminomatous histology. Comparable histologic results of both the EGCTs and the metachronous testicular cancers were found in only one of the 16 patients. Second, in our cohort of 635 patients with EGCTs, the majority of patients (n = 12) who developed metachronous testicular cancer had had a primary retroperitoneal EGCT. In addition, we note that four of 341 patients with primary mediastinal EGCTs subsequently developed metachronous testicular cancer. On the basis of available literature (9–15) and our series, we conclude that patients with EGCTs who develop metachronous testicular cancer have an excellent prognosis. In our series, all five patients receiving chemotherapy for metastatic disease have achieved long-lasting remissions. Eleven patients with metachronous testicular cancer confined to the testis were treated with orchiectomy alone. After a median follow-up period of 51 months (range, 1–154 months) after treatment for metachronous testicular cancers, no treatment failures or recurrences have occurred and all patients are alive without disease. The median overall survival since diagnosis of EGCT was 109 months (range, 46–203 months). Our observed survival data are in accordance with the available literature (9–15). No metachronous testicular cancer-related deaths have so far been reported, suggesting that the metachronous testicular cancers in patients with EGCTs seem to possess a favorable biologic behavior similar to that of de novo testicular germ cell cancers. To calculate the SIR, we used the Saarland Cancer Registry because the age group-specific data available from the registry consider the changes in the incidences of metachronous testicular cancer during the approximately 20 years that make up the study time period. Despite some differences in the incidence of testicular cancer between the participating centers in Europe and the United States, the overall incidence of testicular cancer is comparable among the countries included in the analysis. In general, the differences among the participating countries were smaller than those within the different regional cancer registries of individual countries, particularly when considering age-specific data and time periods (16). We found that the risk for developing metachronous testicular cancer among patients with EGCTs affects mainly those patients with retroperitoneal EGCT (SIR = 100; 95% CI = 54 to 172) or with nonseminomatous EGCT (SIR = 75; 95% CI = 43 to 123). The markedly elevated SIR clearly indicates that those patients with EGCTs are at an increased risk for metachronous testicular cancer, despite the use of chemotherapy in all patients with EGCTs who later developed metachronous testicular cancer. Although this risk translates to a 10% chance of developing metachronous testicular cancer during a 10-year period, there was no mortality observed in those patients with subsequent metachronous testicular cancer. The observed incidence is clearly higher than that assumed from the available literature (9–15). A possible explanation for the discrepancy may be that we were persistent in our follow-up of the study patients. Considering the SIR for contralateral cancer in testicular germ cell tumor patients (4), the risk for metachronous testicular cancer is approximately two to four times higher in the population with EGCTs than in those patients without EGCTs. We identified statistically significant differences in the occurrence of metachronous testicular cancers in different subsets of patients with EGCTs. It is interesting that even the subset of patients with EGCTs with primary mediastinal tumors had an elevated SIR of 31 (95% CI = 8 to 59). Only patients with seminomatous EGCTs did not have a statistically significantly increased risk of developing metachronous testicular cancers when compared with the age-matched population. It has to be kept in mind, however, that our observations are based on a relatively small number of metachronous testicular cancers among patients with EGCTs. Using the Cox proportional hazards model, we could not identify particular characteristics such as patient age, type of treatment received, histology of the primary tumor, serum tumor marker concentrations at diagnosis, or the presence of additional metastatic sites that predicted for the occurrence of metachronous testicular cancer in patients with EGCTs. It remains unclear why only some patients with nonseminomatous EGCTs located in the retroperitoneum develop metachronous testicular cancers, whereas the majority of patients are unaffected. Histologic examination of biopsy specimens from the testes of patients with presumed EGCT has also not provided any clear-cut explanation. Some investigators have found microscopic tumor foci in the testes of numerous patients with EGCTs, whereas others have done so in a minority of patients with EGCTs or only in those patients with evidence of retroperitoneal disease (7,20–22). The histologic finding of a scar in the testis in a patient with an EGCT suggested that a primary testicular tumor might have regressed to form a so-called burned-out germ cell tumor (7). One possible origin of metachronous testicular cancer is the TIN lesion, which has been hypothesized to be the precursor lesion of seminomatous and nonseminomatous testicular germ cell cancers. The prevalence of contralateral TIN lesions in patients with testicular germ cell cancer is believed to be approximately 5% (6). The observation of a simultaneous occurrence of retroperitoneal EGCTs and TIN lesions provides some evidence for the hypothesis that EGCTs are of testicular origin and that the EGCTs are confused with metastases (7). The simultaneous occurrence of TIN lesions and EGCTs, however, does not necessarily imply that EGCTs and TIN lesions have a common malignant progenitor cell. In patients with primary mediastinal germ cell tumors, TIN lesions have been described only in two patients so far (8,14). In both patients, histologic examination of the testes did not reveal an invasive germ cell tumor or a scar. One patient with a nonseminomatous mediastinal EGCT died shortly after diagnosis (8). The other patient with a seminomatous mediastinal EGCT developed a metachronous testicular cancer of nonseminomatous histology 41 months later (14). Thus, the association of EGCTs and TIN lesions may also represent an independent event indicating a generalized biologic defect in gonadal and extragonadal germ cells. The development of metachronous testicular cancers despite initial cisplatin-based chemotherapy for the successful treatment of EGCTs has several implications: First, an approximately 60-fold increased biologic risk for subsequent testicular cancer exists in EGCT patients. This is a germ cell-specific phenomenon because a generally elevated risk for other solid tumors in patients with EGCTs has been excluded (23). Second, the histogenesis of metachronous testicular cancers in patients with EGCTs remains an enigma even with histopathologic examination studies (7,20–22). In our study, metachronous testicular cancers have also been observed in patients with mediastinal EGCTs, a finding that has been reported only once before (14). This may indicate that the mediastinal EGCT is a primary neoplasm with a concomitant in situ lesion in the testis, suggesting—in the opinion of the authors—a more general defect of germ cells. This finding also supports a possible existence of true midline retroperitoneal germ cell tumors regardless of the presence of TIN lesions in the testes of several patients. Another indication of the existence of a retroperitoneal EGCT is the observation of the different outcomes of these patients when compared with those patients who have gonadal germ cell tumors with retroperitoneal metastases, despite comparable prognostic characteristics (Hartmann, J. T: unpublished data). Third, although chemotherapy may reduce the risk of TIN lesions and subsequent testicular cancer (24), it appears that it cannot completely abrogate the risk of developing metachronous testicular cancer. Fourth, in contrast with the well-known association of primary mediastinal nonseminomatous germ cell tumors and hematologic disorders (25), there appears to be no negative impact on survival because of the development of metachronous testicular cancers in patients with EGCTs. Both the biology of EGCTs and the histogenesis of subsequent testicular cancer should be areas of further investigation. Table 1. Characteristics of all 635 patients with EGCT and 16 patients who later developed metachronous testicular cancers*   Patients (n = 635)  Patients† ( n = 16)  *EGCT = extragonadal germ cell tumors; CI = confidence interval.  †Percentage of each population of EGCT patients.  ‡Additional lymph nodes outside of midline.  §Cut point for the detection of elevated levels.  Location of primary EGCT, No. of patients (%)          Mediastinum  341 (53.7)  4 (1.2)      Retroperitoneum  283 (44.6)  12 (4.2)      Cervical lymph nodes  1 (0.2)        Not determined  10 (1.6)    Histology, No. of patients (%)          Nonseminomatous  524 (82.5)  15 (2.9)      Seminomatous  104 (16.4)  1 (1.0)      Not determined  7 (1.1)    Age, y          Median  30  29      Range  14–79  19–49  Sites of metastases, No. of patients (%)          Bone  25 (3.9)  0      Lung  200 (31.5)  4 (2.0)      Central nervous system  22 (3.5)  0      Liver  87 (13.7)  0      Abdominal lymph nodes‡  123 (19.4)  7 (5.7)      Cervical lymph nodes  79 (12.4)  4 (5.1)      Other  75 (11.8)  3 (4.0)      None  24 (3.8)  0  Serum tumor marker at diagnosis          α-Fetoprotein              Elevated, >10, No. of patients (%)§  336 (52.9)  8 (2.4)          Median (range) in ng/mL  183 (1–500 000)  63 (10–11 282)      Human β-chorionic gonadotropin              Elevated, >5, No. of patients (%)§  318 (50.1)  6 (1.9)          Median (range) in mIU/mL  27 (1–9 999 998)  5 (5–17 000)      Lactate dehydrogenase              Elevated, >200, No. of patients (%)§  346 (54.5)  9 (2.6)          Median (range) in IU/L  583 (80–9970)  692 (180–2980)  Type of treatment of EGCT, No. of patients, (%)          Chemotherapy and/or surgery  611 (96.2)  16 (2.6)      Primary irradiation  15 (2.4)  0      Surgery alone  7 (1.1)  0      Not evaluable  2 (0.3)  0  Follow-up status, No. of patients (%)          Alive  377 (59.4)  16 (4.2)      Dead of disease  199 (31.3)  0      Dead of other cause  44 (6.9)  0      Recurrence  234 (36.9)  0      Lost to follow-up  15 (2.4)  0  Estimated survival, % (95% CI)          At 5 y  58 (54 to 62)  100      At 8 y  56 (51 to 60)  100    Patients (n = 635)  Patients† ( n = 16)  *EGCT = extragonadal germ cell tumors; CI = confidence interval.  †Percentage of each population of EGCT patients.  ‡Additional lymph nodes outside of midline.  §Cut point for the detection of elevated levels.  Location of primary EGCT, No. of patients (%)          Mediastinum  341 (53.7)  4 (1.2)      Retroperitoneum  283 (44.6)  12 (4.2)      Cervical lymph nodes  1 (0.2)        Not determined  10 (1.6)    Histology, No. of patients (%)          Nonseminomatous  524 (82.5)  15 (2.9)      Seminomatous  104 (16.4)  1 (1.0)      Not determined  7 (1.1)    Age, y          Median  30  29      Range  14–79  19–49  Sites of metastases, No. of patients (%)          Bone  25 (3.9)  0      Lung  200 (31.5)  4 (2.0)      Central nervous system  22 (3.5)  0      Liver  87 (13.7)  0      Abdominal lymph nodes‡  123 (19.4)  7 (5.7)      Cervical lymph nodes  79 (12.4)  4 (5.1)      Other  75 (11.8)  3 (4.0)      None  24 (3.8)  0  Serum tumor marker at diagnosis          α-Fetoprotein              Elevated, >10, No. of patients (%)§  336 (52.9)  8 (2.4)          Median (range) in ng/mL  183 (1–500 000)  63 (10–11 282)      Human β-chorionic gonadotropin              Elevated, >5, No. of patients (%)§  318 (50.1)  6 (1.9)          Median (range) in mIU/mL  27 (1–9 999 998)  5 (5–17 000)      Lactate dehydrogenase              Elevated, >200, No. of patients (%)§  346 (54.5)  9 (2.6)          Median (range) in IU/L  583 (80–9970)  692 (180–2980)  Type of treatment of EGCT, No. of patients, (%)          Chemotherapy and/or surgery  611 (96.2)  16 (2.6)      Primary irradiation  15 (2.4)  0      Surgery alone  7 (1.1)  0      Not evaluable  2 (0.3)  0  Follow-up status, No. of patients (%)          Alive  377 (59.4)  16 (4.2)      Dead of disease  199 (31.3)  0      Dead of other cause  44 (6.9)  0      Recurrence  234 (36.9)  0      Lost to follow-up  15 (2.4)  0  Estimated survival, % (95% CI)          At 5 y  58 (54 to 62)  100      At 8 y  56 (51 to 60)  100  View Large Table 2. Two-way table of 635 patients with extragonadal germ cell tumors according to site of disease and histology*   EGCT histology†    Metachronous testicular cancer  Not determined  Seminomatous  Nonseminomatous  Total  *EGCT = extragonadal germ cell tumor.  †Chi-square, P = .25.  ‡Includes one patient with primary cervical lymph node involvement.  No, EGCT location              Mediastinal  3  50  284  337      Retroperitoneal  4  52  215  271      Not determined  —  1  10  11‡          Total  7  103  509  619  Yes, EGCT location              Mediastinal  —  1  3  4      Retroperitoneal  —  —  12  12      Not determined  —  —  —  —          Total  —  1  15  16    EGCT histology†    Metachronous testicular cancer  Not determined  Seminomatous  Nonseminomatous  Total  *EGCT = extragonadal germ cell tumor.  †Chi-square, P = .25.  ‡Includes one patient with primary cervical lymph node involvement.  No, EGCT location              Mediastinal  3  50  284  337      Retroperitoneal  4  52  215  271      Not determined  —  1  10  11‡          Total  7  103  509  619  Yes, EGCT location              Mediastinal  —  1  3  4      Retroperitoneal  —  —  12  12      Not determined  —  —  —  —          Total  —  1  15  16  View Large Table 3. Summary of clinicopathologic features of 16 EGCT patients who later developed MTC* Patient No.  Age, y†  EGCT location  EGCT histology  AFP,† ng/mL, <10‡  β-HCG,† mIU/mL, <5‡  Treatment for EGCT (No. of cycles)  MTC histology  Stage of MTC§  Interval between EGCT and MTC diagnoses, mo  Survival since diagnosis of EGCT, mo  *AFP = α-fetoprotein; EGCT = extragonadal germ cell tumor; β-HCG = human β-chorionic gonadotropin; med = mediastinal; MTC = metachronous testicular cancers; n.e. = not evaluable; nonsem = nonseminomatous; rp = retroperitoneal; sem = seminomatous; PEB = cisplatin, etoposide, and bleomycin; PEI = cisplatin, etoposide, and ifosfamide; PVB = cisplatin, vinblastine, and bleomycin; ECBC = etoposide, cisplatin, bleomycin, and cyclophosphamide; CEB = carboplatin, etoposide, and bleomycin.  †Age at diagnosis of EGCT.  ‡Cut point for detection of elevated levels.  §According to Lugano classification (26).  ∥Previously published in reference (14).  16  28  rp  nonsem  <10  500  PEB (4)  sem  Metastatic  74  125  26  19  med  nonsem  63  5  PEB (4)  sem  Testis only  80  115  32  33  rp  nonsem  50  50  PEB (4)  nonsem  Metastatic  30  70  41  40  rp  nonsem  <10  <5  PVB (3)  sem  Metastatic  100  166  67  49  rp  nonsem  <10  <5  PEB/PVB (4)  sem  Testis only  88  203  105∥  24  med  sem  <10  81  PEB (4)  nonsem  Metastatic  41  92  113  29  rp  nonsem  5150  38  PEB (4)  nonsem  Testis only  30  132  171  22  rp  nonsem  602  <5  CEB (4)  sem  Testis only  48  52  216  23  rp  nonsem  11 282  <5  PEB (4)  sem  Testis only  14  88  218  26  med  nonsem  47  293  PEB (4)  nonsem  Testis only  84  102  226  48  med  nonsem  <10  <5  PEB (5)  sem  Testis only  72  73  331  34  rp  nonsem  404  5  PEI (4)  sem  Testis only  35  103  335  n.e.  rp  nonsem  2800  <5  ECBC (6)  sem  Testis only  102  124  359  30  rp  nonsem  <10  <5  PVB (4)  sem  Metastatic  42  195  512  34  rp  nonsem  <10  <5  PVB (4)  sem  Testis only  78  186  527  22  rp  nonsem  <10  17 000  PEB (3)  nonsem  Testis only  39  46  Patient No.  Age, y†  EGCT location  EGCT histology  AFP,† ng/mL, <10‡  β-HCG,† mIU/mL, <5‡  Treatment for EGCT (No. of cycles)  MTC histology  Stage of MTC§  Interval between EGCT and MTC diagnoses, mo  Survival since diagnosis of EGCT, mo  *AFP = α-fetoprotein; EGCT = extragonadal germ cell tumor; β-HCG = human β-chorionic gonadotropin; med = mediastinal; MTC = metachronous testicular cancers; n.e. = not evaluable; nonsem = nonseminomatous; rp = retroperitoneal; sem = seminomatous; PEB = cisplatin, etoposide, and bleomycin; PEI = cisplatin, etoposide, and ifosfamide; PVB = cisplatin, vinblastine, and bleomycin; ECBC = etoposide, cisplatin, bleomycin, and cyclophosphamide; CEB = carboplatin, etoposide, and bleomycin.  †Age at diagnosis of EGCT.  ‡Cut point for detection of elevated levels.  §According to Lugano classification (26).  ∥Previously published in reference (14).  16  28  rp  nonsem  <10  500  PEB (4)  sem  Metastatic  74  125  26  19  med  nonsem  63  5  PEB (4)  sem  Testis only  80  115  32  33  rp  nonsem  50  50  PEB (4)  nonsem  Metastatic  30  70  41  40  rp  nonsem  <10  <5  PVB (3)  sem  Metastatic  100  166  67  49  rp  nonsem  <10  <5  PEB/PVB (4)  sem  Testis only  88  203  105∥  24  med  sem  <10  81  PEB (4)  nonsem  Metastatic  41  92  113  29  rp  nonsem  5150  38  PEB (4)  nonsem  Testis only  30  132  171  22  rp  nonsem  602  <5  CEB (4)  sem  Testis only  48  52  216  23  rp  nonsem  11 282  <5  PEB (4)  sem  Testis only  14  88  218  26  med  nonsem  47  293  PEB (4)  nonsem  Testis only  84  102  226  48  med  nonsem  <10  <5  PEB (5)  sem  Testis only  72  73  331  34  rp  nonsem  404  5  PEI (4)  sem  Testis only  35  103  335  n.e.  rp  nonsem  2800  <5  ECBC (6)  sem  Testis only  102  124  359  30  rp  nonsem  <10  <5  PVB (4)  sem  Metastatic  42  195  512  34  rp  nonsem  <10  <5  PVB (4)  sem  Testis only  78  186  527  22  rp  nonsem  <10  17 000  PEB (3)  nonsem  Testis only  39  46  View Large Fig. 1. View largeDownload slide Cumulative risk for the occurrence of 16 metachronous testicular cancers in 635 patients with extragonadal germ cell tumors. The x-axis represents the months since extragonadal germ cell tumor diagnosis. The y-axis represents the cumulative risk of metachronous testicular cancer expressed as a percentage. Fig. 1. View largeDownload slide Cumulative risk for the occurrence of 16 metachronous testicular cancers in 635 patients with extragonadal germ cell tumors. The x-axis represents the months since extragonadal germ cell tumor diagnosis. The y-axis represents the cumulative risk of metachronous testicular cancer expressed as a percentage. Presented in part at the annual meeting of the European Society of Medical Oncology, Hamburg, Germany, October 2000. References 1 Hartmann JT, Kanz L, Bokemeyer C. Diagnosis and treatment of patients with testicular germ cell cancer. Drugs  1999; 58: 257–81. Google Scholar 2 Bosl GJ, Motzer RJ. Testicular germ-cell cancer. N Engl J Med  1997; 337: 242–53. Google Scholar 3 Pottern LM, Goedert JJ. Epidemiology of testicular cancer. In: Javadpour N, editor. Principles and management of testicular cancer. New York (NY): Thieme; 1986; p. 107–19. Google Scholar 4 Osterlind A, Berthelsen JG, Abildgaard N, Hansen SO, Hjalgrim H, Johansen B, et al. Risk of bilateral testicular germ cell cancer in Denmark: 1960–1984. J Natl Cancer Inst  1991; 83: 1391–5. Google Scholar 5 Wanderas EH, Fossa SD, Tretli S. Risk of a second germ cell cancer after treatment of a primary germ cell cancer in 2201 Norwegian male patients. Eur J Cancer  1997; 33: 244–52. Google Scholar 6 Dieckmann KP, Loy V. Prevalence of contralateral testicular intraepithelial neoplasia in patients with testicular germ cell neoplasms. J Clin Oncol  1996; 14: 3126–32. Google Scholar 7 Daugaard G, von der Maase H, Olsen J, Rorth M, Skakkebaek NE. Carcinoma-in-situ testis in patients with assumed extragonadal germ-cell tumours. Lancet  1987; 2: 528–30. Google Scholar 8 Hailemariam S, Engeler DS, Bannwart F, Amin MB. Primary mediastinal germ cell tumor with intratubular germ cell neoplasia of the testis—further support for germ cell origin of these tumors: a case report. Cancer  1997; 79: 1031–6. Google Scholar 9 Lokich J. Metachronous gonadal and extragonadal primary germ cell tumors: two case reports. Cancer Invest  1994; 12: 406–8. Google Scholar 10 Allaway M, Nseyo UO. Primary testicular seminoma in a patient with a history of extragonadal non-seminomatous germ cell carcinoma. Urology  2000; 55: 949–50. Google Scholar 11 Hayashi T, Mine M, Kojima SI, Sekine H. Extragonadal germ cell tumor followed by metachronous testicular tumor. A case report. Urol Int  1996; 57: 194–6. Google Scholar 12 Gonzalez Quintela A, Lopez Bonet E, Roman J, Aramburo P. Testicular germ cell tumor seven years after a retroperitoneal germ cell tumor. Eur Urol  1991; 19: 336–8. Google Scholar 13 Gerl A, Clemm C, Lamerz R, Wilmanns W. Cisplatin-based chemotherapy of primary extragonadal germ cell tumors. A single institution experience. Cancer  1996; 77: 526–32. Google Scholar 14 Dueland S, Stenwig AE, Heilo A, Hoie J, Ous S, Fossa SD. Treatment and outcome of patients with extragonadal germ cell tumours—the Norwegian Radium Hospital's experience 1974–94. Br J Cancer  1998; 77: 329–35. Google Scholar 15 Daniel C, Fizazi K, Culine S, Zelek L, Wibault P, Theodore C. Metachronous gonadal and extragonadal primaries, or late relapse of germ cell tumor? Urol Oncol  2001; 6: 41–4. Google Scholar 16 Parkin DM, Muir CS.. Cancer incidence in five continents. Comparability and quality of data. IARC Sci Publ  1992; 120: 45–173. Google Scholar 17 Breslow NE, Day NE. Statistical methods in cancer research. Volume II—The design and analysis of cohort studies. IARC Sci Publ  1987; 82: 1–406. Google Scholar 18 Kaplan EL, Meier P. Non-parametric estimation from incomplete observation. J Am Stat Assoc  1958; 53: 457–81. Google Scholar 19 Cox DR. Regression models and life table analysis. J R Stat Soc Ser B  1972; 34: 187–220. Google Scholar 20 Asif S, Uehling DT. Microscopic tumor foci in testes. J Urol  1968; 99: 776–9. Google Scholar 21 Cox JD. Primary malignant germinal tumors of the mediastinum. A study of 24 cases. Cancer  1975; 36: 1162–8. Google Scholar 22 Luna MA, Valenzuela-Tamariz J. Germ-cell tumors of the mediastinum, postmortem findings. Am J Clin Pathol  1976; 65: 450–4. Google Scholar 23 Hartmann JT, Nichols CR, Droz JP, Horwich A, Gerl A, Fossa SD, et al. The relative risk of second non germinal malignancies in patients with extragonadal germ cell tumors. Cancer  2000; 88: 2629–35. Google Scholar 24 Christensen TB, Daugaard G, Geertsen PE, von der Maase H. Effect of chemotherapy on carcinoma in situ of the testis. Ann Oncol  1998; 9: 657–60. Google Scholar 25 Hartmann JT, Nichols CR, Droz JP, Horwich A, Gerl A, Fossa SD, et al. Hematologic disorders associated with primary mediastinal nonseminomatous germ cell tumors. J Natl Cancer Inst  2000; 92: 54–61. Google Scholar 26 Cavalli F, Monfardini S, Pizzocaro G. Report on the International Workshop on Staging and Treatment of Testicular Cancer. Eur J Cancer  1980; 16: 1367–72. Google Scholar © Oxford University Press http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png JNCI: Journal of the National Cancer Institute Oxford University Press

Incidence of Metachronous Testicular Cancer in Patients With Extragonadal Germ Cell Tumors

Download PDF
6 pages

Loading next page...
 
/lp/oxford-university-press/incidence-of-metachronous-testicular-cancer-in-patients-with-tSyRF0f8nf

References (31)

Publisher
Oxford University Press
Copyright
© Oxford University Press
ISSN
0027-8874
eISSN
1460-2105
DOI
10.1093/jnci/93.22.1733
Publisher site
See Article on Publisher Site

Abstract

Abstract Background: The frequency of subsequent testicular cancer (referred to as metachronous testicular cancer) in men who have had previous testicular cancer is relatively high. The rate of metachronous testicular cancer in men with extragonadal germ cell tumors (EGCTs), however, is largely unknown. We conducted a retrospective study of EGCT patients to determine the incidence, cumulative risk, and specific risk factors for metachronous testicular cancers. Methods: A standardized questionnaire about patient characteristics, the extent of EGCT disease, any second malignancies, and treatments received was completed for 635 patients with EGCTs identified from the medical records of 11 cancer centers in Europe and the United States from 1975 through 1996. Comparisons with age group-specific data from the Saarland, Germany, population-based cancer registry were used to calculate the standardized incidence ratio (SIR). The Kaplan-Meier method was used to analyze survival data and cumulative risk. All statistical tests were two-sided. Results: Sixteen EGCT patients (4.1%) developed metachronous testicular cancers, with a median time between diagnoses of 60 months (range, 14–102 months). The risk of developing metachronous testicular cancers was statistically significantly increased in patients with EGCTs (observed = 16; expected = 0.26; SIR = 62; 95% confidence interval [CI] = 36 to 99) and in subsets of EGCT patients with mediastinal location (SIR = 31; 95% CI = 8 to 59), retroperitoneal location (SIR = 100; 95% CI = 54 to 172), and nonseminomatous histology (SIR = 75; 95% CI = 43 to 123). The cumulative risk of developing a metachronous testicular cancer 10 years after a diagnosis of EGCT was 10.3% (95% CI = 4.9% to 15.6%) and was higher among patients with nonseminomatous EGCTs (14.3%; 95% CI = 6.7% to 21.9%) and retroperitoneal EGCTs (14.2%; 95% CI = 5.6% to 22.8%) than among patients with seminomatous EGCTs (1.4%; 95% CI = 0.0% to 4.2%) and mediastinal EGCTs (6.2%; 95% CI = 0.1% to 12.2%). Conclusions: Patients with EGCTs, particularly those with retroperitoneal or nonseminomatous tumors, but also those with primary mediastinal EGCTs, are at an increased risk of metachronous testicular cancer. Testicular cancer is the most common malignancy in men aged 15–35 years (1,2). Approximately 5%–7% of these germ cell tumors are of extragonadal origin (3), with the most common sites being the mediastinum and the retroperitoneum. Extragonadal germ cell tumors (EGCTs) are histologically similar to gonadal germ cell tumors; however, there is no testicular abnormality detectable either by palpation or by high-resolution ultrasonography in patients with EGCTs. Twenty-five years after initial treatment for testicular cancer, the cumulative risk of developing a contralateral testicular cancer is approximately 5%. The risk is higher for patients with nonseminomatous cancers or for patients younger than 30 years of age at the time of diagnosis (4,5). Overall, the relative risk of developing a second testicular cancer is 25-fold higher than in the age-matched population. The relative risk increase is probably a result of the existence of a precursor lesion for testicular cancer (the so-called “testicular intraepithelial neoplasia” [TIN] lesion) in testicular cancer patients. The assumed prevalence of a contralateral TIN lesion in patients with testicular cancer is approximately 5% (6). Although a precursor lesion for the development of EGCTs has not been established, the occurrence of TIN lesions has been observed in 20%–40% of patients with retroperitoneal EGCTs (7) but is very rare in patients with a primary mediastinal EGCT (8). To our knowledge, the subsequent development of testicular cancer (referred to as metachronous testicular cancer) after the patient has been given a diagnosis of a retroperitoneal or mediastinal EGCT has been reported for only 12 patients (9–15). Assessing the incidence of metachronous testicular cancer in patients with EGCTs is complicated by the fact that both events are rare. In this study, we reviewed a large, multicenter database that included information on 635 patients with EGCTs who had been treated during the cisplatin-based chemotherapy era to determine the incidence of metachronous testicular cancer in patients with EGCTs. Patients and methods Data Collection We evaluated the medical records of all patients with EGCTs consecutively diagnosed and/or treated at 11 cancer centers in the United States and Europe from 1975 through 1996. The contributing centers and their locations, the population sample at each center, and the time period for each center are, respectively, as follows: Indiana University, Indianapolis, IN (n = 216, 1989 through 1996); Institut Gustave-Roussy, Villejuif, France, and Centre Léon-Berard, Groupe d'Etude des Tumeurs Urologiques et Genitales, Lyon, France (n = 93, 1975 through 1996); Medical Center II, Eberhard-Karls-University Tuebingen, Germany (n = 13, 1986 through 1993); Hannover University Medical School, Hannover, Germany (n = 88, 1978 through 1995); The Norwegian Radium Hospital, Oslo, Norway (n = 48, 1980 through 1995); Klinikum Groβhadern, Munich, Germany (n = 63, 1979 through 1996); The Royal Marsden Hospital, Sutton, U.K. (n = 65, 1979 through 1994); Kaiser-Franz Josef Spital, Vienna, Austria (n = 19, 1975 through 1996); and Virchow-Klinikum, Berlin, Germany (n = 30, 1987 through 1994). Patients were referred to the participating cancer centers because of their diagnosis of EGCT. The inclusion criterion for this study was evidence of an EGCT, defined as the presence of an extragonadal mass in the mediastinum, retroperitoneum, or elsewhere in the absence of testicular abnormalities at physical examination or ultrasonography. If testicular abnormalities were noted, a biopsy specimen was collected to exclude a diagnosis of primary testicular cancer. A TIN lesion or a scar found at the time the biopsy specimen was taken was not a cause for exclusion from the study. Overall, 637 patients were identified for this study, and 635 patients were eligible. Two patients were excluded from the analysis because chart review revealed invasive testicular cancer. We obtained the status of all living patients as of February 1998. For the data collection, a standardized questionnaire was sent to each center, where it was completed by one of the coinvestigators. We acquired detailed information on patient characteristics, such as the location and histology of the primary EGCT; the extent of the disease, including serum tumor marker concentrations of human β-chorionic gonadotropin (β-HCG), α-fetoprotein (AFP), and lactate dehydrogenase (LDH); history of testicular abnormalities; details on diagnostic methods, treatments, responses to treatment, and follow-up periods; and second testicular cancers, other second cancers, or treatment-related leukemia. We collected information regarding levels of β-HCG, AFP, and LDH because elevation of AFP and/or β-HCG levels with evidence of a mediastinal or retroperitoneal mass is considered specific for the diagnosis of a nonseminomatous EGCT. These tumor markers allow distinctions to be made between nonseminomatous EGCTs (elevated β-HCG and AFP) and other mediastinal or retroperitoneal tumors, including seminomatous germ cell tumors (no elevated markers). The completed questionnaires were checked for plausibility and data consistency at Eberhard-Karls-University Tuebingen Medical Center. The questionnaires were returned to the principal investigator at each center if any important data were missing. For this report, the clinical records of all patients with EGCTs who developed metachronous testicular cancer were reviewed in detail. All histopathologic slides of these cancers were diagnosed and classified by each center's pathology department. Because all patients' data and chart reviews were obtained anonymously and retrospectively, no institutional approval was necessary from any of the participating institutions. Statistical Analysis The duration of follow-up was calculated from the date of the EGCT diagnosis. Survival calculations were performed from the time of metachronous testicular cancer diagnosis and from the time of EGCT diagnosis. All data were entered in a personal computer at the Eberhard-Karls-University Tuebingen, Medical Center II, Germany. To calculate the standard incidence ratio (SIR), data were transferred to the Institute for Medical Information Processing, Eberhard-Karls-University Tuebingen. To determine the number of new cases of metachronous testicular cancer expected in the study group, we used the age group-specific data (classified in 5-year age groups) of the cancer registry of Saarland, Germany, from 1975 through 1996. This cancer registry is a population-based registry that covers the federal state of Saarland located in the southwest region of Germany (16). Results were standardized on the basis of age and follow-up duration (patients' years of risk) and expressed as the SIR with the associated 95% confidence intervals (CIs), which were calculated on the assumption of a Poisson distribution (17). The SIR was considered to be statistically significant if the 95% CI did not include the value 1.0. The SIR calculation was performed with the use of the SAS system (SAS, version 6.11 for Windows; SAS Institute, Inc., Cary, NC). The patients' years at risk, the expected number of cases for every age group, and the SIR were calculated according to Breslow and Day (17). All other statistical analyses were performed with the use of the SPSS system (SPSS for Windows, version 8.0; SPSS Inc., Chicago, IL). The Kaplan–Meier method was used to calculate survival data and to determine the cumulative risk of developing metachronous testicular cancer (18). Cox proportional hazards models were performed to identify variables that could predict the occurrence of metachronous testicular cancers in patients with EGCTs, including categorical variables, such as histology of the primary tumor, age grouping (25 years and older versus younger than 25 years), serum tumor marker concentrations at diagnosis (elevated; yes versus no), presence of additional metastatic sites (yes versus no), and type of treatment, and continuous variables, such as age and tumor marker concentrations at diagnosis (19). All statistical tests were two-sided. Results Because the occurrence of metachronous testicular cancer in patients with EGCTs is a rare event, we identified patients from 11 cancer centers throughout Europe and the United States. In total, 635 patients with EGCTs, with a median age of 30 years (range, 14–79 years), were included in the analysis. Of these 635 patients, 341 (53.7%) had primary mediastinal EGCTs, 283 (44.6%) had primary retroperitoneal EGCTs, and one had primary cervical lymph node involvement. The primary tumor location could not be determined in 10 patients (1.6%) because of widespread disease (including pulmonary metastases) and simultaneous midline involvement of the mediastinum and the retroperitoneum. Histology reports were available for 628 patients, of whom 524 (82.5%) had nonseminomatous EGCTs and 104 (16.4%) had seminomatous EGCTs. Tumor histology was not available for seven patients (1.1%) who had elevated AFP and/or β-HCG levels. Of the 635 patients with EGCTs, 14 (2.2%) received unilateral orchiectomy before or at the end of the treatment for the EGCT. None of the patients underwent bilateral orchiectomy (Table 1). Median follow-up time for surviving patients was 55 months (95% CI = 50 to 60 months). Table 2 classifies all patients according to site of disease and histology. Sixteen of 635 patients developed metachronous testicular cancer, with a median time interval between both diagnoses of 60 months (range, 14–102 months). Another patient developed an EGCT after having testicular cancer. Of the 283 patients who had had a primary retroperitoneal EGCT, 12 (4.2%) developed metachronous testicular cancer (Tables 1 and 2), with a median time interval between both diagnoses of 42 months (95% CI = 27 to 56 months). By contrast, of 341 patients who had had a primary mediastinal EGCT, four (1.2%) developed metachronous testicular cancer, with a median time interval between both diagnoses of 69 months (95% CI = 50 to 88 months) (Table 3). Two of these four patients had 12-cm- and 10-cm-sized mediastinal masses at initial presentation, which was the only disease presentation. The two remaining patients had a mediastinal tumor mass and metastatic disease extending to the abdominal lymph nodes outside the midline structures, with either lung metastases or cervical lymph node/thyroid gland involvement. The median age at diagnosis of EGCT in patients developing metachronous testicular cancer was 29 years (range, 19–49). Serum concentrations of the tumor markers AFP, β-HCG, and LDH were elevated in eight, six, and nine of the 16 patients, respectively (Tables 1 and 3). Of the 16 patients who developed metachronous testicular cancer, the majority were initially diagnosed with nonseminomatous EGCTs (n = 15 [94%]) (Table 3). By contrast, the majority of patients who developed metachronous testicular cancer were diagnosed with classical seminoma (n = 11 [69%]). Three (19%) patients had mixed nonseminomatous germ cell tumors, one patient had embryonal cell carcinoma, and one had differentiated teratoma (6% each). The tumor histologies for the EGCT and subsequent metachronous testicular cancer were comparable for one patient, with both tumors containing mixed nonseminomatous germ cell tumor with embryonal cell carcinoma, seminomatous, and teratomatous components. In this patient, the time interval between diagnosis of EGCT and the occurrence of the metachronous testicular cancer was 84 months. Five of 16 patients had had a bilateral testicular biopsy before treatment for the EGCT. The biopsy specimens from four of five patients were found to have normal histology. One patient, who was diagnosed with a TIN lesion, had a primary mediastinal seminoma and developed a subsequent metachronous testicular cancer 41 months after the completion of chemotherapy. Of the 16 patients with metachronous testicular cancer, all had previously received cisplatin-based combination chemotherapy (cisplatin, etoposide, and bleomycin, n = 10; cisplatin, vinblastine, and bleomycin, n = 3; etoposide, cisplatin, bleomycin, andcyclophosphamide, n = 1; cisplatin, etoposide, and ifosfamide, n = 1; carboplatin, vinblastine, and bleomycin, n=1) for the treatment of their EGCT, and 11 were allocated to postchemotherapy surgery because of residual tumor masses (Table 1). The response status could be evaluated in 15 patients. All patients had attained complete remission after chemotherapy and/or secondary surgery. Patients' characteristics are summarized in detail in Tables 1 and 3. Treatment for metachronous testicular cancer consisted of orchiectomy alone for those with stage I disease (11 of the 16 patients with metachronous testicular cancer [69%]) and orchiectomy plus chemotherapy and/or secondary surgery for those with metastatic disease of the metachronous testicular cancers (five patients [31%]). After a median follow-up time of 51 months (range, 1–154 months) after treatment for metachronous testicular cancer, all 16 patients were alive, without evidence of disease. The median overall survival after diagnosis of EGCT was 109 months (range, 46–203 months). In comparison with an age-matched population, there was a statistically significant risk of developing metachronous testicular cancer among the 635 patients with EGCTs during 2089 person-years at risk (observed [O] = 16; expected [E] = 0.26; SIR = 62; 95% CI = 36 to 99). Only 0.26 cases of de novo testicular cancer were expected during the study time period for a population this size. In addition, the risk of developing metachronous testicular cancers was statistically significantly increased for the following subsets of EGCT patients: those with primary mediastinal EGCTs (SIR = 31 [O = 4; E = 0.13; 95% CI = 8 to 59]); those with primary retroperitoneal EGCTs (SIR = 100 [O = 12; E = 0.12; 95% CI = 54 to 172]); and those with nonseminomatous EGCTs (SIR = 75 [O = 15; E = 0.2; 95% CI = 43 to 123]). Patients with seminomatous EGCTs did not have an increased risk of developing metachronous testicular cancers (SIR = 20; [O = 1; E = 0.05; 95% CI = 0.5 to 111]). No particular characteristics such as patient age, type of treatment received, histology of the primary tumor, presence of additional metastatic sites, or serum tumor marker concentrations at EGCT diagnosis could be identified that predicted the occurrence of metachronous testicular cancer in patients with EGCTs. We used the Kaplan–Meier method to estimate the cumulative risk of developing metachronous testicular cancer within 10 years after the initial diagnosis of the EGCT and found that the cumulative risk was 10.3% (95% CI = 4.9% to 15.6%). The cumulative risk of developing metachronous testicular cancer was 14.3% (95% CI = 6.7% to 21.9%) if the EGCT was nonseminomatous and 1.4% (95% CI = 0.0% to 4.2%) if the EGCT was seminomatous (P = .10). A retroperitoneal EGCT was associated with a cumulative risk of 14.2% (95% CI = 5.6% to 22.8%), and a primary mediastinal location was associated with a risk of 6.2% (95% CI = 0.1% to 12.2%), a statistically nonsignificant difference (P = .18). Inspection of the cumulative risk estimates suggested that there might be two time periods during which patients with EGCTs seem more likely to develop metachronous testicular cancers; at approximately 2.5–4 years and at approximately 6–7.5 years after the EGCT diagnosis (Fig. 1). Discussion The occurrence of metachronous testicular cancer in patients with retroperitoneal and mediastinal EGCTs appears to be an infrequent event, with only 16 of 635 patients who were previously diagnosed and effectively treated for EGCTs developing metachronous testicular cancer. There are only 12 reported cases of metachronous testicular cancer in patients with EGCTs available from the literature (9–15). This report is the first systematic investigation to provide data about the incidence and cumulative risks of the development of metachronous testicular cancer after treatment for an EGCT in a representative large cohort of patients by comparison with an age-matched general population. In addition, this investigation tries to identify specific risk factors for the development of subsequent metachronous testicular cancers in patients with EGCTs. The average time between diagnosis of EGCT and diagnosis of metachronous testicular cancer in the 12 patients reported in the literature (9–15) was approximately 7–8 years (range, 35–168 months). However, we observed a median time of 5 years between the two diagnoses (range, 19–142 months). Our results may not conflict with those reported in the literature (9–15), because it appears that there might be two time periods, the first at approximately 2.5–4 years and the second at approximately 6–7.5 years, during which patients are at an increased risk of developing metachronous testicular cancer (Fig. 1). Our results are similar in two ways to those of the 12 cases reported in the literature to date (9–15). First, the majority of patients had nonseminomatous EGCTs before subsequently developing metachronous testicular cancers of seminomatous histology. Comparable histologic results of both the EGCTs and the metachronous testicular cancers were found in only one of the 16 patients. Second, in our cohort of 635 patients with EGCTs, the majority of patients (n = 12) who developed metachronous testicular cancer had had a primary retroperitoneal EGCT. In addition, we note that four of 341 patients with primary mediastinal EGCTs subsequently developed metachronous testicular cancer. On the basis of available literature (9–15) and our series, we conclude that patients with EGCTs who develop metachronous testicular cancer have an excellent prognosis. In our series, all five patients receiving chemotherapy for metastatic disease have achieved long-lasting remissions. Eleven patients with metachronous testicular cancer confined to the testis were treated with orchiectomy alone. After a median follow-up period of 51 months (range, 1–154 months) after treatment for metachronous testicular cancers, no treatment failures or recurrences have occurred and all patients are alive without disease. The median overall survival since diagnosis of EGCT was 109 months (range, 46–203 months). Our observed survival data are in accordance with the available literature (9–15). No metachronous testicular cancer-related deaths have so far been reported, suggesting that the metachronous testicular cancers in patients with EGCTs seem to possess a favorable biologic behavior similar to that of de novo testicular germ cell cancers. To calculate the SIR, we used the Saarland Cancer Registry because the age group-specific data available from the registry consider the changes in the incidences of metachronous testicular cancer during the approximately 20 years that make up the study time period. Despite some differences in the incidence of testicular cancer between the participating centers in Europe and the United States, the overall incidence of testicular cancer is comparable among the countries included in the analysis. In general, the differences among the participating countries were smaller than those within the different regional cancer registries of individual countries, particularly when considering age-specific data and time periods (16). We found that the risk for developing metachronous testicular cancer among patients with EGCTs affects mainly those patients with retroperitoneal EGCT (SIR = 100; 95% CI = 54 to 172) or with nonseminomatous EGCT (SIR = 75; 95% CI = 43 to 123). The markedly elevated SIR clearly indicates that those patients with EGCTs are at an increased risk for metachronous testicular cancer, despite the use of chemotherapy in all patients with EGCTs who later developed metachronous testicular cancer. Although this risk translates to a 10% chance of developing metachronous testicular cancer during a 10-year period, there was no mortality observed in those patients with subsequent metachronous testicular cancer. The observed incidence is clearly higher than that assumed from the available literature (9–15). A possible explanation for the discrepancy may be that we were persistent in our follow-up of the study patients. Considering the SIR for contralateral cancer in testicular germ cell tumor patients (4), the risk for metachronous testicular cancer is approximately two to four times higher in the population with EGCTs than in those patients without EGCTs. We identified statistically significant differences in the occurrence of metachronous testicular cancers in different subsets of patients with EGCTs. It is interesting that even the subset of patients with EGCTs with primary mediastinal tumors had an elevated SIR of 31 (95% CI = 8 to 59). Only patients with seminomatous EGCTs did not have a statistically significantly increased risk of developing metachronous testicular cancers when compared with the age-matched population. It has to be kept in mind, however, that our observations are based on a relatively small number of metachronous testicular cancers among patients with EGCTs. Using the Cox proportional hazards model, we could not identify particular characteristics such as patient age, type of treatment received, histology of the primary tumor, serum tumor marker concentrations at diagnosis, or the presence of additional metastatic sites that predicted for the occurrence of metachronous testicular cancer in patients with EGCTs. It remains unclear why only some patients with nonseminomatous EGCTs located in the retroperitoneum develop metachronous testicular cancers, whereas the majority of patients are unaffected. Histologic examination of biopsy specimens from the testes of patients with presumed EGCT has also not provided any clear-cut explanation. Some investigators have found microscopic tumor foci in the testes of numerous patients with EGCTs, whereas others have done so in a minority of patients with EGCTs or only in those patients with evidence of retroperitoneal disease (7,20–22). The histologic finding of a scar in the testis in a patient with an EGCT suggested that a primary testicular tumor might have regressed to form a so-called burned-out germ cell tumor (7). One possible origin of metachronous testicular cancer is the TIN lesion, which has been hypothesized to be the precursor lesion of seminomatous and nonseminomatous testicular germ cell cancers. The prevalence of contralateral TIN lesions in patients with testicular germ cell cancer is believed to be approximately 5% (6). The observation of a simultaneous occurrence of retroperitoneal EGCTs and TIN lesions provides some evidence for the hypothesis that EGCTs are of testicular origin and that the EGCTs are confused with metastases (7). The simultaneous occurrence of TIN lesions and EGCTs, however, does not necessarily imply that EGCTs and TIN lesions have a common malignant progenitor cell. In patients with primary mediastinal germ cell tumors, TIN lesions have been described only in two patients so far (8,14). In both patients, histologic examination of the testes did not reveal an invasive germ cell tumor or a scar. One patient with a nonseminomatous mediastinal EGCT died shortly after diagnosis (8). The other patient with a seminomatous mediastinal EGCT developed a metachronous testicular cancer of nonseminomatous histology 41 months later (14). Thus, the association of EGCTs and TIN lesions may also represent an independent event indicating a generalized biologic defect in gonadal and extragonadal germ cells. The development of metachronous testicular cancers despite initial cisplatin-based chemotherapy for the successful treatment of EGCTs has several implications: First, an approximately 60-fold increased biologic risk for subsequent testicular cancer exists in EGCT patients. This is a germ cell-specific phenomenon because a generally elevated risk for other solid tumors in patients with EGCTs has been excluded (23). Second, the histogenesis of metachronous testicular cancers in patients with EGCTs remains an enigma even with histopathologic examination studies (7,20–22). In our study, metachronous testicular cancers have also been observed in patients with mediastinal EGCTs, a finding that has been reported only once before (14). This may indicate that the mediastinal EGCT is a primary neoplasm with a concomitant in situ lesion in the testis, suggesting—in the opinion of the authors—a more general defect of germ cells. This finding also supports a possible existence of true midline retroperitoneal germ cell tumors regardless of the presence of TIN lesions in the testes of several patients. Another indication of the existence of a retroperitoneal EGCT is the observation of the different outcomes of these patients when compared with those patients who have gonadal germ cell tumors with retroperitoneal metastases, despite comparable prognostic characteristics (Hartmann, J. T: unpublished data). Third, although chemotherapy may reduce the risk of TIN lesions and subsequent testicular cancer (24), it appears that it cannot completely abrogate the risk of developing metachronous testicular cancer. Fourth, in contrast with the well-known association of primary mediastinal nonseminomatous germ cell tumors and hematologic disorders (25), there appears to be no negative impact on survival because of the development of metachronous testicular cancers in patients with EGCTs. Both the biology of EGCTs and the histogenesis of subsequent testicular cancer should be areas of further investigation. Table 1. Characteristics of all 635 patients with EGCT and 16 patients who later developed metachronous testicular cancers*   Patients (n = 635)  Patients† ( n = 16)  *EGCT = extragonadal germ cell tumors; CI = confidence interval.  †Percentage of each population of EGCT patients.  ‡Additional lymph nodes outside of midline.  §Cut point for the detection of elevated levels.  Location of primary EGCT, No. of patients (%)          Mediastinum  341 (53.7)  4 (1.2)      Retroperitoneum  283 (44.6)  12 (4.2)      Cervical lymph nodes  1 (0.2)        Not determined  10 (1.6)    Histology, No. of patients (%)          Nonseminomatous  524 (82.5)  15 (2.9)      Seminomatous  104 (16.4)  1 (1.0)      Not determined  7 (1.1)    Age, y          Median  30  29      Range  14–79  19–49  Sites of metastases, No. of patients (%)          Bone  25 (3.9)  0      Lung  200 (31.5)  4 (2.0)      Central nervous system  22 (3.5)  0      Liver  87 (13.7)  0      Abdominal lymph nodes‡  123 (19.4)  7 (5.7)      Cervical lymph nodes  79 (12.4)  4 (5.1)      Other  75 (11.8)  3 (4.0)      None  24 (3.8)  0  Serum tumor marker at diagnosis          α-Fetoprotein              Elevated, >10, No. of patients (%)§  336 (52.9)  8 (2.4)          Median (range) in ng/mL  183 (1–500 000)  63 (10–11 282)      Human β-chorionic gonadotropin              Elevated, >5, No. of patients (%)§  318 (50.1)  6 (1.9)          Median (range) in mIU/mL  27 (1–9 999 998)  5 (5–17 000)      Lactate dehydrogenase              Elevated, >200, No. of patients (%)§  346 (54.5)  9 (2.6)          Median (range) in IU/L  583 (80–9970)  692 (180–2980)  Type of treatment of EGCT, No. of patients, (%)          Chemotherapy and/or surgery  611 (96.2)  16 (2.6)      Primary irradiation  15 (2.4)  0      Surgery alone  7 (1.1)  0      Not evaluable  2 (0.3)  0  Follow-up status, No. of patients (%)          Alive  377 (59.4)  16 (4.2)      Dead of disease  199 (31.3)  0      Dead of other cause  44 (6.9)  0      Recurrence  234 (36.9)  0      Lost to follow-up  15 (2.4)  0  Estimated survival, % (95% CI)          At 5 y  58 (54 to 62)  100      At 8 y  56 (51 to 60)  100    Patients (n = 635)  Patients† ( n = 16)  *EGCT = extragonadal germ cell tumors; CI = confidence interval.  †Percentage of each population of EGCT patients.  ‡Additional lymph nodes outside of midline.  §Cut point for the detection of elevated levels.  Location of primary EGCT, No. of patients (%)          Mediastinum  341 (53.7)  4 (1.2)      Retroperitoneum  283 (44.6)  12 (4.2)      Cervical lymph nodes  1 (0.2)        Not determined  10 (1.6)    Histology, No. of patients (%)          Nonseminomatous  524 (82.5)  15 (2.9)      Seminomatous  104 (16.4)  1 (1.0)      Not determined  7 (1.1)    Age, y          Median  30  29      Range  14–79  19–49  Sites of metastases, No. of patients (%)          Bone  25 (3.9)  0      Lung  200 (31.5)  4 (2.0)      Central nervous system  22 (3.5)  0      Liver  87 (13.7)  0      Abdominal lymph nodes‡  123 (19.4)  7 (5.7)      Cervical lymph nodes  79 (12.4)  4 (5.1)      Other  75 (11.8)  3 (4.0)      None  24 (3.8)  0  Serum tumor marker at diagnosis          α-Fetoprotein              Elevated, >10, No. of patients (%)§  336 (52.9)  8 (2.4)          Median (range) in ng/mL  183 (1–500 000)  63 (10–11 282)      Human β-chorionic gonadotropin              Elevated, >5, No. of patients (%)§  318 (50.1)  6 (1.9)          Median (range) in mIU/mL  27 (1–9 999 998)  5 (5–17 000)      Lactate dehydrogenase              Elevated, >200, No. of patients (%)§  346 (54.5)  9 (2.6)          Median (range) in IU/L  583 (80–9970)  692 (180–2980)  Type of treatment of EGCT, No. of patients, (%)          Chemotherapy and/or surgery  611 (96.2)  16 (2.6)      Primary irradiation  15 (2.4)  0      Surgery alone  7 (1.1)  0      Not evaluable  2 (0.3)  0  Follow-up status, No. of patients (%)          Alive  377 (59.4)  16 (4.2)      Dead of disease  199 (31.3)  0      Dead of other cause  44 (6.9)  0      Recurrence  234 (36.9)  0      Lost to follow-up  15 (2.4)  0  Estimated survival, % (95% CI)          At 5 y  58 (54 to 62)  100      At 8 y  56 (51 to 60)  100  View Large Table 2. Two-way table of 635 patients with extragonadal germ cell tumors according to site of disease and histology*   EGCT histology†    Metachronous testicular cancer  Not determined  Seminomatous  Nonseminomatous  Total  *EGCT = extragonadal germ cell tumor.  †Chi-square, P = .25.  ‡Includes one patient with primary cervical lymph node involvement.  No, EGCT location              Mediastinal  3  50  284  337      Retroperitoneal  4  52  215  271      Not determined  —  1  10  11‡          Total  7  103  509  619  Yes, EGCT location              Mediastinal  —  1  3  4      Retroperitoneal  —  —  12  12      Not determined  —  —  —  —          Total  —  1  15  16    EGCT histology†    Metachronous testicular cancer  Not determined  Seminomatous  Nonseminomatous  Total  *EGCT = extragonadal germ cell tumor.  †Chi-square, P = .25.  ‡Includes one patient with primary cervical lymph node involvement.  No, EGCT location              Mediastinal  3  50  284  337      Retroperitoneal  4  52  215  271      Not determined  —  1  10  11‡          Total  7  103  509  619  Yes, EGCT location              Mediastinal  —  1  3  4      Retroperitoneal  —  —  12  12      Not determined  —  —  —  —          Total  —  1  15  16  View Large Table 3. Summary of clinicopathologic features of 16 EGCT patients who later developed MTC* Patient No.  Age, y†  EGCT location  EGCT histology  AFP,† ng/mL, <10‡  β-HCG,† mIU/mL, <5‡  Treatment for EGCT (No. of cycles)  MTC histology  Stage of MTC§  Interval between EGCT and MTC diagnoses, mo  Survival since diagnosis of EGCT, mo  *AFP = α-fetoprotein; EGCT = extragonadal germ cell tumor; β-HCG = human β-chorionic gonadotropin; med = mediastinal; MTC = metachronous testicular cancers; n.e. = not evaluable; nonsem = nonseminomatous; rp = retroperitoneal; sem = seminomatous; PEB = cisplatin, etoposide, and bleomycin; PEI = cisplatin, etoposide, and ifosfamide; PVB = cisplatin, vinblastine, and bleomycin; ECBC = etoposide, cisplatin, bleomycin, and cyclophosphamide; CEB = carboplatin, etoposide, and bleomycin.  †Age at diagnosis of EGCT.  ‡Cut point for detection of elevated levels.  §According to Lugano classification (26).  ∥Previously published in reference (14).  16  28  rp  nonsem  <10  500  PEB (4)  sem  Metastatic  74  125  26  19  med  nonsem  63  5  PEB (4)  sem  Testis only  80  115  32  33  rp  nonsem  50  50  PEB (4)  nonsem  Metastatic  30  70  41  40  rp  nonsem  <10  <5  PVB (3)  sem  Metastatic  100  166  67  49  rp  nonsem  <10  <5  PEB/PVB (4)  sem  Testis only  88  203  105∥  24  med  sem  <10  81  PEB (4)  nonsem  Metastatic  41  92  113  29  rp  nonsem  5150  38  PEB (4)  nonsem  Testis only  30  132  171  22  rp  nonsem  602  <5  CEB (4)  sem  Testis only  48  52  216  23  rp  nonsem  11 282  <5  PEB (4)  sem  Testis only  14  88  218  26  med  nonsem  47  293  PEB (4)  nonsem  Testis only  84  102  226  48  med  nonsem  <10  <5  PEB (5)  sem  Testis only  72  73  331  34  rp  nonsem  404  5  PEI (4)  sem  Testis only  35  103  335  n.e.  rp  nonsem  2800  <5  ECBC (6)  sem  Testis only  102  124  359  30  rp  nonsem  <10  <5  PVB (4)  sem  Metastatic  42  195  512  34  rp  nonsem  <10  <5  PVB (4)  sem  Testis only  78  186  527  22  rp  nonsem  <10  17 000  PEB (3)  nonsem  Testis only  39  46  Patient No.  Age, y†  EGCT location  EGCT histology  AFP,† ng/mL, <10‡  β-HCG,† mIU/mL, <5‡  Treatment for EGCT (No. of cycles)  MTC histology  Stage of MTC§  Interval between EGCT and MTC diagnoses, mo  Survival since diagnosis of EGCT, mo  *AFP = α-fetoprotein; EGCT = extragonadal germ cell tumor; β-HCG = human β-chorionic gonadotropin; med = mediastinal; MTC = metachronous testicular cancers; n.e. = not evaluable; nonsem = nonseminomatous; rp = retroperitoneal; sem = seminomatous; PEB = cisplatin, etoposide, and bleomycin; PEI = cisplatin, etoposide, and ifosfamide; PVB = cisplatin, vinblastine, and bleomycin; ECBC = etoposide, cisplatin, bleomycin, and cyclophosphamide; CEB = carboplatin, etoposide, and bleomycin.  †Age at diagnosis of EGCT.  ‡Cut point for detection of elevated levels.  §According to Lugano classification (26).  ∥Previously published in reference (14).  16  28  rp  nonsem  <10  500  PEB (4)  sem  Metastatic  74  125  26  19  med  nonsem  63  5  PEB (4)  sem  Testis only  80  115  32  33  rp  nonsem  50  50  PEB (4)  nonsem  Metastatic  30  70  41  40  rp  nonsem  <10  <5  PVB (3)  sem  Metastatic  100  166  67  49  rp  nonsem  <10  <5  PEB/PVB (4)  sem  Testis only  88  203  105∥  24  med  sem  <10  81  PEB (4)  nonsem  Metastatic  41  92  113  29  rp  nonsem  5150  38  PEB (4)  nonsem  Testis only  30  132  171  22  rp  nonsem  602  <5  CEB (4)  sem  Testis only  48  52  216  23  rp  nonsem  11 282  <5  PEB (4)  sem  Testis only  14  88  218  26  med  nonsem  47  293  PEB (4)  nonsem  Testis only  84  102  226  48  med  nonsem  <10  <5  PEB (5)  sem  Testis only  72  73  331  34  rp  nonsem  404  5  PEI (4)  sem  Testis only  35  103  335  n.e.  rp  nonsem  2800  <5  ECBC (6)  sem  Testis only  102  124  359  30  rp  nonsem  <10  <5  PVB (4)  sem  Metastatic  42  195  512  34  rp  nonsem  <10  <5  PVB (4)  sem  Testis only  78  186  527  22  rp  nonsem  <10  17 000  PEB (3)  nonsem  Testis only  39  46  View Large Fig. 1. View largeDownload slide Cumulative risk for the occurrence of 16 metachronous testicular cancers in 635 patients with extragonadal germ cell tumors. The x-axis represents the months since extragonadal germ cell tumor diagnosis. The y-axis represents the cumulative risk of metachronous testicular cancer expressed as a percentage. Fig. 1. View largeDownload slide Cumulative risk for the occurrence of 16 metachronous testicular cancers in 635 patients with extragonadal germ cell tumors. The x-axis represents the months since extragonadal germ cell tumor diagnosis. The y-axis represents the cumulative risk of metachronous testicular cancer expressed as a percentage. Presented in part at the annual meeting of the European Society of Medical Oncology, Hamburg, Germany, October 2000. References 1 Hartmann JT, Kanz L, Bokemeyer C. Diagnosis and treatment of patients with testicular germ cell cancer. Drugs  1999; 58: 257–81. Google Scholar 2 Bosl GJ, Motzer RJ. Testicular germ-cell cancer. N Engl J Med  1997; 337: 242–53. Google Scholar 3 Pottern LM, Goedert JJ. Epidemiology of testicular cancer. In: Javadpour N, editor. Principles and management of testicular cancer. New York (NY): Thieme; 1986; p. 107–19. Google Scholar 4 Osterlind A, Berthelsen JG, Abildgaard N, Hansen SO, Hjalgrim H, Johansen B, et al. Risk of bilateral testicular germ cell cancer in Denmark: 1960–1984. J Natl Cancer Inst  1991; 83: 1391–5. Google Scholar 5 Wanderas EH, Fossa SD, Tretli S. Risk of a second germ cell cancer after treatment of a primary germ cell cancer in 2201 Norwegian male patients. Eur J Cancer  1997; 33: 244–52. Google Scholar 6 Dieckmann KP, Loy V. Prevalence of contralateral testicular intraepithelial neoplasia in patients with testicular germ cell neoplasms. J Clin Oncol  1996; 14: 3126–32. Google Scholar 7 Daugaard G, von der Maase H, Olsen J, Rorth M, Skakkebaek NE. Carcinoma-in-situ testis in patients with assumed extragonadal germ-cell tumours. Lancet  1987; 2: 528–30. Google Scholar 8 Hailemariam S, Engeler DS, Bannwart F, Amin MB. Primary mediastinal germ cell tumor with intratubular germ cell neoplasia of the testis—further support for germ cell origin of these tumors: a case report. Cancer  1997; 79: 1031–6. Google Scholar 9 Lokich J. Metachronous gonadal and extragonadal primary germ cell tumors: two case reports. Cancer Invest  1994; 12: 406–8. Google Scholar 10 Allaway M, Nseyo UO. Primary testicular seminoma in a patient with a history of extragonadal non-seminomatous germ cell carcinoma. Urology  2000; 55: 949–50. Google Scholar 11 Hayashi T, Mine M, Kojima SI, Sekine H. Extragonadal germ cell tumor followed by metachronous testicular tumor. A case report. Urol Int  1996; 57: 194–6. Google Scholar 12 Gonzalez Quintela A, Lopez Bonet E, Roman J, Aramburo P. Testicular germ cell tumor seven years after a retroperitoneal germ cell tumor. Eur Urol  1991; 19: 336–8. Google Scholar 13 Gerl A, Clemm C, Lamerz R, Wilmanns W. Cisplatin-based chemotherapy of primary extragonadal germ cell tumors. A single institution experience. Cancer  1996; 77: 526–32. Google Scholar 14 Dueland S, Stenwig AE, Heilo A, Hoie J, Ous S, Fossa SD. Treatment and outcome of patients with extragonadal germ cell tumours—the Norwegian Radium Hospital's experience 1974–94. Br J Cancer  1998; 77: 329–35. Google Scholar 15 Daniel C, Fizazi K, Culine S, Zelek L, Wibault P, Theodore C. Metachronous gonadal and extragonadal primaries, or late relapse of germ cell tumor? Urol Oncol  2001; 6: 41–4. Google Scholar 16 Parkin DM, Muir CS.. Cancer incidence in five continents. Comparability and quality of data. IARC Sci Publ  1992; 120: 45–173. Google Scholar 17 Breslow NE, Day NE. Statistical methods in cancer research. Volume II—The design and analysis of cohort studies. IARC Sci Publ  1987; 82: 1–406. Google Scholar 18 Kaplan EL, Meier P. Non-parametric estimation from incomplete observation. J Am Stat Assoc  1958; 53: 457–81. Google Scholar 19 Cox DR. Regression models and life table analysis. J R Stat Soc Ser B  1972; 34: 187–220. Google Scholar 20 Asif S, Uehling DT. Microscopic tumor foci in testes. J Urol  1968; 99: 776–9. Google Scholar 21 Cox JD. Primary malignant germinal tumors of the mediastinum. A study of 24 cases. Cancer  1975; 36: 1162–8. Google Scholar 22 Luna MA, Valenzuela-Tamariz J. Germ-cell tumors of the mediastinum, postmortem findings. Am J Clin Pathol  1976; 65: 450–4. Google Scholar 23 Hartmann JT, Nichols CR, Droz JP, Horwich A, Gerl A, Fossa SD, et al. The relative risk of second non germinal malignancies in patients with extragonadal germ cell tumors. Cancer  2000; 88: 2629–35. Google Scholar 24 Christensen TB, Daugaard G, Geertsen PE, von der Maase H. Effect of chemotherapy on carcinoma in situ of the testis. Ann Oncol  1998; 9: 657–60. Google Scholar 25 Hartmann JT, Nichols CR, Droz JP, Horwich A, Gerl A, Fossa SD, et al. Hematologic disorders associated with primary mediastinal nonseminomatous germ cell tumors. J Natl Cancer Inst  2000; 92: 54–61. Google Scholar 26 Cavalli F, Monfardini S, Pizzocaro G. Report on the International Workshop on Staging and Treatment of Testicular Cancer. Eur J Cancer  1980; 16: 1367–72. Google Scholar © Oxford University Press

Journal

JNCI: Journal of the National Cancer Institute – Oxford University Press

Published: Nov 21, 2001

There are no references for this article.