This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Albers, P. Articles by Hartmann, M. Search for Related Content PubMed PubMed Citation Articles by Albers, P. Articles by Hartmann, M. Social Bookmarking                            What's this?

Risk Factors for Relapse in Clinical Stage I Nonseminomatous Testicular Germ Cell Tumors: Results of the German Testicular Cancer Study Group Trial

Peter Albers, Roswitha Siener, Sabine Kliesch, Lothar Weissbach, Susanne Krege, Christoph Sparwasser, Harald Schulze, Axel Heidenreich, Werner de Riese, Volker Loy, Erhard Bierhoff, Christian Wittekind, Rolf Fimmers, Michael Hartmann

From the Departments of Urology and Medical Biometry, Bonn University, Bonn; Department of Urology, Münster University, Münster; Department of Urology and Institute of Pathology, Krankenhaus am Urban, Berlin; Department of Urology, Essen University, and Institute of Pathology Essen-Mitte, Essen; Department of Urology, Military Hospital, Ulm; Department of Urology, Städtische Kliniken, Dortmund; Department of Urology, Marburg University, Marburg; Institute of Pathology, Leipzig University, Leipzig; Department of Urology, Military Hospital, Hamburg, Germany; and Division of Urology, Texas Tech University, Lubbock, TX.

Address reprint requests to Peter Albers, MD, Department of Urology, Bonn University, D-53105 Bonn, Germany; email: peter.albers{at}ukb.uni-bonn.de.

	ABSTRACT

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION APPENDIX REFERENCES

 
Purpose: To prospectively assess potential risk factors for relapse in clinical stage I nonseminomatous germ cell tumors of the testis (CS I NSGCT).

Patients and Methods: From September 1996 to May 2002, 200 patients with CS I NSGCT were prospectively assigned to retroperitoneal lymph node dissection (RPLND), and risk factor assessment was performed within a multicenter protocol. One hundred sixty-five patients had an adequate minimum follow-up of 12 months (mean, 34.5 months) or had pathologic stage II.

Results: Pathologic stage II disease was found in 27.9% of patients. Only 0.6% of patients relapsed in the retroperitoneum after confirmation of pathologic stage I disease. With reference pathology, vascular invasion (VI) was most predictive of stage in multifactorial analysis (accuracy, 65.1%). However, the positive predictive value (PPV) of VI to predict patients who have metastatic disease or relapse during follow-up was only 52.7%. With absent VI, low-risk patients had a negative predictive value (NPV) of 76.9%. With a combination of several risk factors, the PPV increased to 63.6% and the negative predictive value increased to 86.5%.

Conclusion: Even with an optimal combination of prognostic factors and reference pathology, more than one third of patients predicted to have pathologic stage II or relapse during follow-up will not harbor metastatic disease and, therefore, would be overtreated with adjuvant therapy. However, patients at low risk may be predicted at an 86.5% level, and thus, surveillance in highly compliant patients would be a valuable option. For high-risk patients, further reduction of adjuvant treatment is necessary.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION APPENDIX REFERENCES

 
APPROXIMATELY ONE fourth of patients with clinical stage I nonseminomatous germ cell tumors of the testes (CS I NSGCT) will have retroperitoneal metastases after surgical staging (pathologic stage II).¹ Another 10% of patients will experience tumor progression outside the retroperitoneum during follow-up.^2–4 Thus, one third of patients with CS I NSGCT without adjuvant treatment will show progressive disease. Adjuvant treatment options for patients with CS I NSGCT include retroperitoneal lymph node dissection (RPLND) with or without chemotherapy for patients with pathologic stage II, adjuvant chemotherapy, and surveillance with chemotherapy at relapse. All options provide cure rates of approximately 99%. However, two thirds of patients are already cured after ablation of the primary testicular tumor.

For more than 10 years, efforts have been made to tailor adjuvant treatment to those patients who are at high risk for relapse and to avoid treatment of patients at low risk of relapse. The high- and low-risk groups can be defined by expert imaging evaluation and by histopathologic and biologic markers of the primary tumor, or by both methods. Previous reports of prospectively performed trials have indicated that a combination of histopathologic parameters² and the combination of percentage of embryonal carcinoma and tumor proliferative activity⁵ are predictive of high- and low-risk patients, respectively. Both models have been confirmed in retrospective series.^6,7

Without having been prospectively assessed, however, vascular invasion (VI) of the primary tumor has gained acceptance as a single risk factor in the latest World Health Organization classification. The positive predictive value (PPV) of this parameter in the largest prospective trial thus far (the Medical Research Council trial), however, was only 48%. More than half of patients at high risk according to the World Health Organization are treated unnecessarily.²

Therefore, since September 1996, the German Testicular Cancer Study Group (GTCSG) has performed a prospective evaluation of risk factors to evaluate possibly improved risk scores and to validate the results of previous investigations of patients with CS I NSGCT in regard to cell cycle–dependent biologic and additional histopathologic risk factors. Preliminary data of this prospective evaluation have been published,⁸ and this article presents the final evaluation of these data.

	PATIENTS AND METHODS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION APPENDIX REFERENCES

 
Patient Population
From September 1996 until May 2002, 416 patients with CS I NSGCT were prospectively enrolled onto the German multicenter trial (DKH 70–7,024, AUO 01/94) to evaluate different treatment options in CS I NSGCT. The trial consists of two parts. In part A, patients are randomly assigned to either RPLND or to one cycle of adjuvant chemotherapy (cisplatin, etoposide, bleomycin [PEB]). The primary end point is to prove a more than 7% difference in the reduction of the recurrence rate after RPLND (10%) by adjuvant chemotherapy with only one cycle (360 patients needed, alpha 5%, beta 20%). In part B, patients are randomly assigned to either RPLND or risk-adapted therapy. The risk-adapted therapy is based on VI, and the primary end point is quality of life. Patients with VI will get two courses of PEB chemotherapy, and patients without VI will be observed on a surveillance protocol. This part of the trial is powered with 5% alpha and 10% beta for a 17% difference in quality-of-life scores measured by the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire C30 questionnaire (120 patients needed). This is a concurrent randomization, and centers had to decide whether to participate in part A or B. The trial involves 66 centers, and 12 of them enrolled 60% of the patients.

Two hundred of these patients were randomly assigned to RPLND and represent the cohort for risk factor evaluation within this trial. Thirteen patients (6.5%) were lost to follow-up or dropped out of the study. Of 187 assessable patients, 165 were pathologic stage II at time of RPLND or were pathologic stage I with a minimum follow-up time of 12 months. More than 80% of relapses after RPLND occur within 1 year after treatment, and therefore, the minimum follow-up of 12 months was considered to be adequate.² One hundred fifty-seven of these patients had a complete histopathologic work-up, including reference pathology and immunohistochemistry. In five patients, however, immunohistochemical staining was not assessable. Flow cytometric evaluation was available in 123 patients. Flow cytometric evaluation was terminated after an interim analysis that showed no additional information compared with immunohistochemical evaluation of proliferation rates. Therefore, of 165 patients with sufficient follow-up, 152 were available with complete data for univariate and multifactorial analysis and represent the cohort for this prospective risk-factor analysis. These patients had a mean age (± SD) of 31.3 years (± 8.3 years), and the mean follow-up time (± SD) of patients with pathologic stage I disease after RPLND was 34.5 months (± 14.11 months; range, 12 to 64 months).

Clinical Staging, Surgery, and Follow-Up Investigations
After ablation of the tumor-bearing testicle, clinical staging was performed using chest and abdominal computed tomography (CT) and tumor markers (alpha-fetoprotein, ß chain of human chorionic gonadotropin, and lactate dehydrogenase). CT scans were evaluated locally. Tumor markers had to decline to normal before surgery (at least three half times for alpha-fetoprotein in case of excessive increase before ablation). RPLND was performed within 3 weeks after histologic diagnosis and staging. Nerve-sparing techniques were applied in 93% of patients, as technically described previously.^9,10 Ipsilateral nerve-sparing RPLND was suggested, which was performed by 39% of surgeons, whereas 61% favored the modified approach of contralateral nerve-sparing RPLND. Bilateral templates were performed in one patient with pathologic stage IIB disease after staging error; in 6% of patients, no data concerning surgical technique were available. After surgery, patients were observed every 2 months by markers and chest x-ray. CT scans were performed three times a year during the first 2 years.

Assessment of Risk Factors
Histopathologic and biologic risk factors have been assessed in all patients and are listed in Tables 1, 2, and 3. After local histopathologic evaluation, the paraffin-embedded tissue blocks containing tumor tissue were sent to the reference pathologists (C.W. and V.L.). The most representative tissue block was evaluated for histopathology, and serial 5-µm sections for immunohistochemical assessment (E.B.) and 50-µm sections for flow cytometry (W.D.) were cut and distributed to the different laboratories. All tissue blocks were coded, and neither the patient’s personal data nor clinical or pathologic stage were given in the code. The selection of the representative tissue block followed the previously published guidelines.⁵ In brief, the tumor tissue–bearing block with the highest amount of tumor with aggressive features like embryonal carcinoma, mitotic figures, and chromatin pattern was selected for further analysis.

Immunohistochemical Assessment of Proliferation Rates
Tumor proliferation was assessed by staining the slides with MIB-1 antibody against the Ki-67 receptor, as previously published.⁷ In brief, 5-µm sections of the selected tumor block were deparaffinized, pretreated by 3 x 10 minutes of microwave heating, and stained with a 1:10 dilution of MIB-1 antibody (Dako, Carpinteria, CA). Negative and positive controls were performed, and the MIB-1 positive cells were detected using the LSAB-kit (peroxidase detection, Dako). MIB-1 evaluation was performed by counting MIB-1 positively stained tumor cells at a magnification of x400 and calculating the percentage of positively stained nuclei of the total number of tumor cells in more than 500 tumor cells (MIB-1 score: MIB-1–positive tumor cells/total number of tumor cells). If possible, counting was performed in embryonal carcinoma only (hot spot counting). In the absence of embryonal carcinoma, the MIB-1 score was evaluated in other nonseminomatous tumor components. The evaluation of the immunohistochemical assessment of MIB-1 scores was performed by a doctor (E.B.) who was not aware of the pathologic or clinical stage of the patient. Selected slides were re-evaluated by the reference pathologist (C.W.) to check for interindividual differences.

Flow Cytometric Assessment of Proliferation Rates
A 50-µm section of the representative tumor block was used for flow cytometry. The preparation of the single-cell solution from the paraffin-embedded section was performed as previously published.^11,12 Flow cytometry was performed with a Becton Dickinson (San Jose, CA) flow cytometer. The single parameters assessed by flow cytometry are listed in Table 3. The gating and evaluation of the results have previously been published, and the method was repeated in the same manner to get comparable results for validation of the previous data.^11,12 The flow cytometry was performed without knowledge of the pathologic or clinical stage of the patient.

Study Coordination
The protocol of the clinical trial was previously approved by the Ethics Committee of the University of Bonn, Bonn, Germany. All participating centers had the local approval of their ethics committee as well. The study was conducted in accordance with the ethical principles stated in the most recent version of the Declaration of Helsinki and the applicable good clinical practice/international conference on hormonization of clinical trails guidelines. The randomization was performed by the study center in Bonn. The correspondence between the local investigators/pathologists and the reference pathologists was coordinated by the study center. Likewise, the data assessment of the reference pathology, immunohistochemical and flow cytometrical analysis, and follow-up data was coordinated by the study center (R.S.). Data monitoring was performed by an independent investigator at the local sites.

Statistical Analysis
All parameters analyzed in this study are listed in Tables 1 to 3. Reference pathology was available in 157 of 165 assessable patients. Thus, VI and the percentages of single tumor components were available in this patient cohort. Complete data of MIB-1 assessment and reference pathology was available in 152 patients. Therefore, comparative analysis and logistic regression analysis was possible in this patient cohort only. In 123 patients, complete data of flow cytometry, in addition to MIB-1 and reference pathology, were available. Therefore, the analysis of flow cytometric data was restricted to this patient cohort.

Means and SDs were computed and compared by the student’s nonpaired t test using SPSS (SPSS Inc, Chicago, IL). Categorical comparisons were performed using the ² analysis. The analysis of the different cell-cycle fractions by flow cytometry was compared between pathologic stage I and pathologic stage II/relapsed patients using the nonparametric Wilcoxon rank sum test because the data were highly skewed. The same test was used for the comparison of pathologic stage II patients and relapsed patients. In addition, stepwise logistic regression analysis was used for multifactorial analysis. All reported P values are two-sided.

	RESULTS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION APPENDIX REFERENCES

 
Study Population: Clinical and Pathologic Staging and Relapse
Of 200 patients who were randomly assigned to surgical staging by RPLND within the GTCSG trial, 13 had to be excluded because of violation of the inclusion criteria or loss to follow-up. Of the remaining 187 patients with CS I NSGCT, 125 (67%) had no metastatic lymph nodes at the time of RPLND (pathologic stage I). Of 187 assessable patients, 165 had pathologic stage II at RPLND, had relapsed, or had pathologic stage I at RPLND with a minimum follow-up of 12 months and, therefore, were judged eligible for further analysis of risk factors. One hundred three (62.4%) of 165 patients had no metastatic lymph nodes at the time of RPLND and remained free of recurrence with a mean (± SD) follow-up of 34.5 months (± 14.1 months; range, 12 to 64 months).

Sixteen (9.7%) of 165 patients later developed metastases, detected at a mean time interval to RPLND of 7.5 months (range, 1 to 40 months). Eight of these patients had a pulmonary relapse; one was found to have an in-field relapse in the retroperitoneum, and seven had the relapse elsewhere (eg, marker elevation only, inguinal). The mean follow-up of the patients with recurrent disease was 26.1 months (range, 1 to 51 months).

Forty-six (27.9%) of 165 patients had metastatic retroperitoneal lymph nodes at the time of RPLND (pathologic stage II). All of these patients had low-volume disease and, according to the protocol, subsequently were treated by two courses of adjuvant PEB chemotherapy. They were then without further relapse at a mean (±SD) follow-up time of 24.9 months (± 14.7 months; range, 1 to 55 months).

Assessment of Risk Factors
The reference pathology was performed in 157 of 165 patients. Another five patients with reference pathology were excluded from comparative analysis because immunohistochemistry was not assessable.

To analyze the predictive power of the different risk factors, patients with pathologic stage II disease at the time of RPLND and patients with pathologic stage I and subsequent relapse during follow-up were grouped together. The univariate statistical analysis, therefore, was performed with 99 patients who had pathologic stage I disease and 58 patients who had pathologic stage II disease or who relapsed (Tables 1 and 2). The multifactorial statistical analysis was performed with 95 patients who had pathologic stage I disease and 57 patients who had pathologic stage II disease or who relapsed (Tables 4 and 5).

MIB-1 Proliferation Rates
The immunohistochemical analysis of proliferation rates with the MIB-1 antibody was available in 152 patients. MIB-1 scores showed significant differences between the groups (Table 1). Patients with pathologic stage I disease had a mean (± SD) MIB-1 score of 57.9% (± 30.5%; range, 3.8% to 97.8%); whereas, patients with pathologic stage II or greater had a mean (± SD) MIB-1 score of 71.9% (± 25.4%; range, 2.4% to 98.2%). Patients with pathologic stage II disease did not significantly differ from patients with subsequent relapse (pathologic stage > II).

Flow Cytometry
The assessment of the different cell-cycle fractions by flow cytometry was performed in 123 patients. The quality of this analysis is given by a low accuracy of computer analysis (RCS) value (RCS range is 0 to 100, and the lowest values refer to histograms with the best computer analysis). Another quality criterion is the coefficient of variation (CV) of the diploid population. A low CV represents a clear histogram analysis without too much debris. Usually, a CV of 15 is the upper border of a histogram that can be evaluated. The median RCS value in the given population (all patients) was 2.6 (1.6 in pathologic stage I patients; 4.6 in pathologic stage II or relapse patients), and the median CV of the diploid population is 10.1 (5.1 in pathologic stage I patients; 16.2 in pathologic stage II and relapse patients). The median number of events analyzed was 15,460, and a median number of 4,856 events were analyzed after gating. In summary, these quality data assure a flow cytometric computer analysis of high quality, with single-cell solutions containing a relatively high amount of debris.

The flow cytometric assessment included the complete cell-cycle fractions (G0, G1, S, G2, and M phase) in diploid and aneuploid cell lines. As can be seen in Table 3, no statistically significant differences have been observed between the different stages. In addition, patients with subsequent relapse after RPLND did not differ in flow cytometric parameters from patients with pathologic stage II disease at the time of RPLND. Therefore, the analysis was terminated at the evaluation of 123 patients. Flow cytometric parameters did not enter the multifactorial analysis of risk factors.

Prediction of Risk Groups Using Cutoff Values
This trial was performed to validate given cutoffs of previously published risk factors.^6,7,11,12 This is why a receiver operating curve analysis of the different cutoffs was not performed. With the exception of flow cytometric parameters, the univariate analysis of the histopathologic and immunohistochemical parameters confirmed the previously established risk factors of VI, percentage of embryonal carcinoma (correspondingly, percentage of teratoma), and MIB-1 score (Tables 1 to 3). The previously published cutoff values of these parameters are absence or no absence of VI (regardless of whether the tumor invades lymphatic or vascular vessels), 50% or more versus less than 50% embryonal carcinoma, and more than 70% versus 70% or less MIB-1. Using these cutoffs, Tables 6 and 7 present the frequency and sensitivity, test specificity, and test accuracy of predictive values. Tables 4 and 5 present the results of the logistic and stepwise logistic regression analysis, respectively, and in Table 8, the parameters were combined to possibly improve the prediction of low- and high-risk groups.

Accordingly, the logistic and stepwise logistic regression analysis favored VI as the best predictor of pathologic stages based on the differences in univariate analysis. The second most important parameter was the MIB-1 score, followed by the percentage of embryonal carcinoma (Tables 4 and 5).

Interestingly, a combination of the parameters following the rank in logistic regression analysis was not able to predict a low- or a high-risk group at the 100% level. The best possible prediction of the low-risk group with a NPV of 86.5% resulted from a combination of patients with absent VI and MIB-1 scores of 70% or less (Table 8). The addition of patients with less than 50% embryonal carcinoma did not improve the NPV but lowered the sample size. Therefore, the combination of VI with MIB-1 was able to predict a low-risk group at the 86.5% level. This group represents 24.3% of all the patients with clinical stage I disease that were assessable with all three parameters.

The best possible prediction of patients with metastases (pathologic stage II) or who relapsed resulted from a combination of positive VI, more than 70% MIB-1, and 50% or greater percentage of embryonal carcinoma. The PPV of these patients with a combination of all three adverse parameters was 63.6%. This high-risk group represents 29.0% of the whole patient cohort with clinical stage I disease. Again, it was not possible to predict patients with metastases at the 100% level.

	DISCUSSION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION APPENDIX REFERENCES

 
Over the last 25 years, the distribution of stage at initial diagnosis of NSGCT has changed, with more patients being diagnosed at earlier stages. Now, approximately 90% of testis cancer patients present with low-stage disease (Lugano/tumor-node-metastasis stages I and II).^13–15

The clinical staging of patients with low-stage testis cancer, however, remains insufficient. Despite new-generation CT scans (helical CT with high resolution) and the introduction of positron emission tomography, the detection of small-volume metastases (< 10 mm) is unreliable.^3,16,17 Still, 25% of patients are understaged by CT scans. Another 10% of patients will develop metastases outside the retroperitoneum during follow-up, mostly in the lungs.^2,18

These figures have been confirmed by the current trial. Of 187 patients with clinical stage I disease, only 125 patients (66.8%) were confirmed with pathologic stage I disease at the time of RPLND and no metastases during follow-up. Forty-six patients (24.6%) were found with pathologic stage II disease, and another 16 patients (8.6%) developed metastases during follow-up after RPLND without retroperitoneal metastases. This indicates that the clinical staging in a multicenter setting is correct in only two thirds of patients. In a single center setting, the staging accuracy might improve, but in a different healthcare system, the referral at initial diagnosis is unlikely.^19,20 Clearly, in this setting, RPLND is beneficial to uncover one fourth of patients with staging error by CT scans and tumor markers. Numerous studies have shown that nodes less than 10 mm may harbor metastatic disease if the CT scan shows multiple nodes in the typical landing zone of testicular tumors. The size criteria are drifting downward, and the arbitrarily chosen cutoff of 10 mm might not be valid any longer.^19,20

In the last 25 years, therefore, more than 1,500 patients have been retrospectively investigated to identify prognostic factors of retroperitoneal metastases and relapse (trials with > 100 patients^2,6,21–27). Only two trials^2,12 prospectively identified risk factors by either surveillance or RPLND with follow-up. The Medical Research Council trial² invented a risk score based on lymphatic and vascular invasion of the primary tumor, the embryonal carcinoma, and the absence of yolk sac tumor to identify patients at high risk of relapse. The Indiana trial¹² identified parameters to predict patients at low risk such as low proliferation rates and low amount of embryonal carcinoma in the primary tumor. In the last 5 years, risk factors, such as absence of VI, have already been used in clinical management, and patients have been observed.^28–31 However, according to the Medical Research Council data, one half of the patients with VI are overtreated, and relapse in the low-risk population occurred in up to 22% of patients. In retrospective series, the percentage of embryonal carcinoma and the proliferation rate of the tumor measured by MIB-1 immunohistochemistry and flow cytometry were suggested to improve the prediction of high- and low-risk patients, respectively.^6,7,11 The predominance of embryonal carcinoma in the primary tumor was proposed to be predictive of pulmonal metastases after RPLND.³² The aim of the current prospective trial of the GTCSG was to validate these newer prognostic factors in a large group of patients with CS I NSGCT who underwent RPLND.

With reference pathology, VI of the primary tumor was most predictive of stage in multifactorial analysis. The overall test accuracy was 65.1%. High-risk patients were predicted with a PPV of 52.7%, compared with the PPV of 48% of three and four risk factors of the prospective Medical Research Council trial.² With VI alone, one half of the patients will remain disease-free despite being at high risk for metastatic disease. If these patients are treated, one half of them will be overtreated. With the addition of another two risk parameters (MIB-1 score > 70% and embryonal carcinoma 50%), the PPV in this series could be improved (63.6%). If this model is applied to the single patient, the minimum risk of overtreatment can be reduced to approximately 35% if all adverse parameters are present. It is of importance that this calculation is based on reference pathology results.

According to the previous trials, the prediction of low-risk patients was better. With absent VI, the NPV was 76.9%. This was lower than expected from the prior studies. With the addition of the MIB-1 score ( 70%), the NPV could be improved (86.5%). If a single patient with CS I NSGCT opts for surveillance because of absent VI, his chance for relapse is approximately 23%. This is in accordance with the already performed clinical trials in Austria and Scandinavia with more than 200 patients and a relapse rate in the low-risk group of 14% to 22%.^28,29 With the addition of a low proliferation rate, the chance for relapse would only be 13.5%. However, the immunohistochemical assessment of the MIB-1 score has an interindividual difference that weakens this parameter. In this series, the MIB-1 results were assessed by the same pathologist who already evaluated this score in the previous series (E.B.). He was not aware of the clinical stage of the patient, and the method of evaluation was defined as previously published.⁷ In comparison with the reference pathologist, however, who re-evaluated a substantial proportion of the slides, the prediction of stages using the same cutoff value of 70% was different in approximately 20% of cases (data not shown).

The flow cytometric assessment of proliferation rates was stopped at 123 patients. The analysis showed that a prediction of the clinical stages was not possible. In a critical evaluation, the quality parameters of the analysis revealed a very high CV of the diploid cell lines. This is usually a result of sample preparation from paraffin-embedded tissue. In this multicenter trial, the fresh tumor tissue was locally fixed and paraffin-embedded by very different methods, and thus, it is difficult to prepare a homogeneous single-cell solution with routinely applied methods. If the single-cell solution contains too much debris, the analysis is weakened. In the previous trials, the CV of the diploid nuclei population averaged 3.5, with an SD of 1.2. The current median CV of the diploid nuclei was 10.1, with a lower quartile of 5.1 and an upper quartile of 16.2. This overlaying problem led to a falsely large proportion of diploid cell lines (77.7%). Consecutively, the amount of aneuploid cells measured by flow cytometry was relatively small, and the possibly important proliferative fractions of the aneuploid cell lines were reduced. This illustrates clearly that flow cytometric analysis is not a routine evaluation. Because of technical difficulties, flow cytometric assessment cannot be recommended in the routine setting.

Taken together, this trial uncovers the limits of histopathologic risk assessment. The evaluation of single tumor components, the judgment on VI, and the assessment of the immunohistochemical proliferation rates are subject to substantial interindividual differences. With great experience in the evaluation of the single parameters, the prediction of stages improves. However, even with an optimal combination of experts and an optimal combination of risk factors, the prediction of the risk groups is far from perfect. At best, one third of patients at high risk are overtreated. The prediction of a low-risk group, however, seems possible. Without risk evaluation, patients in the current trial with modern clinical staging had a 33% chance to develop progressive disease without adjuvant treatment. This chance could be reduced to 23% with absent VI and low MIB-1 score. The combination of both parameters left the patient with a chance of relapse under surveillance of only 13.5%. Because the relapses in this study occurred in 14 (88%) of 16 patients within the first 10 months of follow-up, surveillance of this low-risk group is a valid option. Follow-up investigations may be further reduced in this selected group, and approximately 80% to 90% of patients will remain disease-free with a low burden of uncertainty within the first year. However, late relapse may occur, and patients should be aware of this. This policy was applicable for 52% (82 of 157) of patients without VI, and with an increasing number of low-stage patients, half of them had approximately an 80% to 90% chance of avoiding treatment.

However, the prediction of high-risk patients is not satisfying. With VI alone, the risk for metastatic or progressive disease raises from 33% without risk evaluation to only 52.7%. With all adverse-risk parameters, the risk is approximately 64%. Still, 36% of patients are overtreated if the patients opt for adjuvant therapy. In case of adjuvant chemotherapy, 36% would unnecessarily accept a potential risk of late effects, such as vascular events, alteration in fat metabolism, and secondary malignancies, from cisplatin-containing treatment.^33–37 In this setting, nerve-sparing RPLND without adjuvant chemotherapy in case of micrometastatic disease is an interesting alternative. With an experienced surgeon, this procedure has virtually no long-term toxicity, and it is curative in a substantial cohort of patients (approximately 75% with stage IIA disease). The risk of loosing the antegrade ejaculation with this procedure in clinical stage I patients is less than 2% in experienced hands (single-center experience). In a multicenter setting, the GTCSG found a retrograde ejaculation in 7.4% of 237 patients after nerve-sparing RPLND in clinical stage I patients (unpublished data). The risk of relapse despite negative nodes in the current series is 8.6%. This matches with the large meta-analysis of Sharir et al¹⁸ in more than 600 patients, with a pulmonary relapse rate of 8%. Therefore, the single patient would decrease his 64% relapse rate by RPLND without adjuvant chemotherapy to approximately 25% in case of micrometastatic nodes and to 8% in case of negative nodes. The follow-up would be easy because the chance to experience an in-field relapse in the abdomen is 0.6% (one of 165 patients in this study). In the future, modern molecular markers of the primary tumor and/or serum may improve this prediction of high-risk patients.

Another alternative to solve the problem of the high-risk patients is to further reduce adjuvant chemotherapy. Two cycles of PEB in patients with VI have shown to be highly effective, with a relapse rate of 3%.³⁸ In terms of long-term toxicity, it seems appropriate to reduce the amount of chemotherapy to one cycle of adjuvant PEB only. In advanced cases, it was not possible to alter the regimen (eg, carboplatin instead of cisplatin or treatment without bleomycin). Therefore, to date, only the reduction of cycles may be possible to reduce long-term toxicity. However, patients have to be observed intensively because late relapses may occur, and it is not proven that the reduction of cycles dose-dependently reduces long-term side effects.

In conclusion, this trial, the largest prospective trial of assessment of risk factors in a cohort of testis cancer patients who were treated by RPLND, confirms the possible prediction of patients at very low risk of relapse, and surveillance policies may be offered with a higher safety and, consequently, with a reduction in follow-up visits. However, surveillance policy is only possible with an expert pathologist and a compliant patient. The most important result of the trial confirms the insufficient prediction of high-risk patients. At present, patients at high risk cannot be predicted at a level that would allow the acceptance of the long-term side effects of chemotherapy in the whole cohort. Because clinical staging will probably not improve and better risk factors probably will not show up in the near future, reduction of chemotherapy or surgical treatment remain the only options to avoid unnecessary long-term toxicity in this group of patients.

	APPENDIX

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION APPENDIX REFERENCES

 
The following institutions participated in this sudy: Krankenhaus der Barmherzigen Schwestern, Linz, Austria; Kreiskrankenhaus, Aschersleben; Bundeswehrkrankenhaus, Krankenhaus am Friedrichshain, Krankenhaus Am Urban, Auguste-Viktoria Krankenhaus, Universitätsklinikum Benjamin Franklin, Berlin; St Agnes Krankenhaus, Bocholt; Universitätsklinik, Bonn; Landeskrankenhaus, Coburg; St Vincenz Krankehaus, Datteln; Anhaltinische Diakonische Anstalten, Dessau; Städtische Kliniken, Dortmund; Universitätsklinik, Essen; Städtische Kliniken Höchst, Frankfurt; Universitätsklinik, Greifswald; Allgemeines Krankenhaus, Hagen; Albertinenkrankenhaus, Bundeswehrkrankenhaus, Hamburg; Universitätsklinik, Hannover; Universitätsklinik, Homburg; Bundeswehrkrankenhaus, Städtisches Krankenhaus Kemperhof, Koblenz; Klinikum Holweide, St Elisabeth Krankenhaus, St Hildegardiskrankenhaus, Universitätsklinik, Koln; Kreiskrankenhaus, Lüdenscheid; Klinikum der Universität Heidelberg, Diakonissenkrankenhaus, Mannheim; Universitätsklinik, Marburg; Kreiskrankenhaus, Mechernich; Kreiskrankenhaus, Minden; Städtisches Krankenhaus Bogenhausen, München; Universitätsklinik, Münster; St Elisabeth Krankenhaus, Neuwied; Südharz-Krankenhaus, Nordhausen; Klinikum, Osnabrück; Klinik, Planegg; Klinikum Ernst v. Bergmann, Potsdam; Universitätsklinik, Rostock; Städtisches Krankenhaus, Salzgitter; Klinikum, Schwerin; and Bundeswehrkrankenhaus, Uln, Germany.

	ACKNOWLEDGMENTS

 
We thank Richard S. Foster, MD (Department of Urology, Indiana University, Indianapolis, IN), for his helpful comments.

	NOTES

 
Supported by grants DFG Al 391/3-1 and DFG Al 391/3-2 from the Deutsche Forschungsgemeinschaft and by the Deutsche Krebshilfe, e. V. (trial no. DKH 70-7024).

	REFERENCES

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION APPENDIX REFERENCES

 
1. Donohue JP, Thornhill JA, Foster RS, et al: The role of retroperitoneal lymphandectomy in clinical stage B testis cancer: The Indiana University experience (1965–1989). J Urol 153:85–89, 1995[CrossRef][Medline]

2. Read G, Stenning SP, Cullen MH, et al: Medical Research Council prospective study of surveillance for stage I testicular teratoma. J Clin Oncol 10:1762–1768, 1992[Abstract/Free Full Text]

3. Fernandez EB, Mol JW, Foley JP, et al: Retroperitoneal imaging with third and fourth generation computed axial tomography in clinical stage I nonseminomatous germ cell tumors. Urology 44:548–552, 1994[CrossRef][Medline]

4. Sharir S, Jewett MA, Sturgeon JF, et al: Progression detection of stage I nonseminomatous testis cancer on surveillance: Implications for the followup protocol. J Urol 161:472–475, 1999[CrossRef][Medline]

5. Albers P, Miller GA, Orazi A, et al: Immunohistochemical assessment of tumor proliferation and volume of embryonal carcinoma identify patients with clinical stage A nonseminomatous testicular germ cell tumor at low risk for occult metastasis. Cancer 75:844–850, 1995[CrossRef][Medline]

6. Heidenreich A, Sesterhenn IA, Mostofi FK, et al: Prognostic risk factors that identify patients with clinical stage I nonseminomatous germ cell tumors at low risk and high risk for metastasis. Cancer 83:1002–1111, 1998[CrossRef][Medline]

7. Albers P, Bierhoff E, Neu D, et al: MIB-1 immunohistochemistry in clinical stage I nonseminomatous testicular germ cell tumors predicts patients at low risk for metastasis. Cancer 79:1710–1716, 1997[CrossRef][Medline]

8. Albers P, Siener R, Hartmann M, et al: Risk factors for relapse in stage I non-seminomatous germ-cell tumors: preliminary results of the German Multicenter Trial—German Testicular Cancer Study Group. Int J Cancer 83:828–830, 1999[CrossRef][Medline]

9. Weissbach L, Boedefeld E, Horstmann-Dubral B: Surgical treatment of stage I nonseminomatous germ cell testis tumor. Eur Urol 17:97–106, 1990[Medline]

10. Donohue JP, Thornhill JA, Foster RS, et al: Primary retroperitoneal lymph node dissection in clinical stage A non-seminomatous germ cell testis cancer: Review of the Indiana University experience 1965–1989. Br J Urol 71:326–335, 1993[Medline]

11. De Riese WT, De Riese C, Ulbright TM, et al: Flow-cytometric and quantitative histologic parameters as prognostic indicators for occult retroperitoneal disease in clinical-stage-I non-seminomatous testicular germ-cell tumors. Int J Cancer 57:628–633, 1994[Medline]

12. Albers P, Ulbright TM, Albers J, et al: Tumor proliferative activity is predictive of pathological stage in clinical stage A nonseminomatous testicular germ cell tumors. J Urol 155:579–586, 1996[CrossRef][Medline]

13. Howe HL, Wingo PA, Thun MJ, et al: Annual report to the nation on the status of cancer (1973 through 1998), featuring cancers with recent increasing trends. J Natl Cancer Inst 93:824–842, 2001[Abstract/Free Full Text]

14. Sonneveld DJ, Hoekstra HJ, Van Der Graaf WT, et al: The changing distribution of stage in nonseminomatous testicular germ cell tumours, from 1977 to 1996. Br J Urol Int 84:68–74, 1999

15. Wittekind C, Wagner G: TNM Classification of Malignant Tumours (ed 5). Berlin, Germany, Springer-Verlag, 1997

16. Albers P, Bender H, Yilmaz H, et al: Positron emission tomography in the clinical staging of patients with stage I and II testicular germ cell tumors. Urology 53:808–811, 1999[CrossRef][Medline]

17. Hain SF, O’Doherty MJ, Timothy AR, et al: Fluorodeoxyglucose PET in the initial staging of germ cell tumours. Eur J Nucl Med 27:590–594, 2000[CrossRef][Medline]

18. Sharir S, Foster R, Jewett M: Optimizing follow up for early stage nonseminomatous testicular cancer. J Urol 155:327a, 1996 (abstr 67)[CrossRef]

19. Leibovitch L, Foster RS, Kopecky KK, et al: Improved accuracy of computerized tomography based clinical staging in low stage nonseminomatous germ cell cancer using size criteria of retroperitoneal lymph nodes. J Urol 154:1759–1763, 1995[CrossRef][Medline]

20. Hilton S, Herr HW, Teitcher JB, et al: CT detection of retroperitoneal lymph node metastases in patients with clinical stage I testicular nonseminomatous germ cell cancer: assessment of size and distribution criteria. Am J Roentgenol 169:521–525, 1997[Abstract/Free Full Text]

21. Hoskin P, Dilly S, Easton D, et al: Prognostic factors in stage I non-seminomatous germ-cell testicular tumors managed by orchiectomy and surveillance: Implications for adjuvant chemotherapy. J Clin Oncol 4:1031–1036, 1986[Abstract/Free Full Text]

22. Klepp O, Olsson AM, Henrikson H, et al: Prognostic factors in clinical stage I nonseminomatous germ cell tumors of the testis: Multivariate analysis of a prospective multicenter study—Swedish-Norwegian Testicular Cancer Group. J Clin Oncol 8:509–518, 1990[Abstract]

23. Colls BM, Harvey VJ, Skelton L, et al: Results of the surveillance policy of stage I non-seminomatous germ cell testicular tumours. Br J Urol 70:423–428, 1992[Medline]

24. Sesterhenn IA, Weiss RB, Mostofi FK, et al: Prognosis and other clinical correlates of pathologic review in stage I and II testicular carcinoma: A report from the Testicular Cancer Intergroup Study. J Clin Oncol 10:69–78, 1992[Abstract]

25. Sturgeon JF, Jewett MA, Alison RE, et al: Surveillance after orchidectomy for patients with clinical stage I nonseminomatous testis tumors. J Clin Oncol 10:564–568, 1992[Abstract/Free Full Text]

26. Gels ME, Hoekstra HJ, Sleijfer DT, et al: Detection of recurrence in patients with clinical stage I nonseminomatous testicular germ cell tumors and consequences for further follow-up: A single-center 10-year experience. J Clin Oncol 13:1188–1194, 1995[Abstract]

27. Sogani PC, Perrotti M, Herr HW, et al: Clinical stage I testis cancer: Long-term outcome of patients on surveillance. J Urol 159:855–858, 1998[CrossRef][Medline]

28. Pont J, Albrecht W, Postner G, et al: Adjuvant chemotherapy for high-risk clinical stage I nonseminomatous testicular germ cell cancer: Long-term results of a prospective trial. J Clin Oncol 14:441–448, 1996[Abstract/Free Full Text]

29. Klepp O, Dahl O, Flodgren P, et al: Risk-adapted treatment of clinical stage 1 non-seminoma testis cancer. Eur J Cancer 33:1038–1044, 1997[CrossRef][Medline]

30. Böhlen D, Borner M, Sonntag RW, et al: Long-term results following adjuvant chemotherapy in patients with clinical stage I testicular nonseminomatous malignant germ cell tumors with high risk factors. J Urol 161:1148–1152, 1999[CrossRef][Medline]

31. Ondrus D, Matoska J, Belan V, et al: Prognostic factors in clinical stage I nonseminomatous germ cell testicular tumors: Rationale for different risk-adapted treatment. Eur Urol 33:562–566, 1998[CrossRef][Medline]

32. Sweeney CJ, Hermans BP, Heilman DK, et al: Results and outcome of retroperitoneal lymph node dissection for clinical stage I embryonal carcinoma-predominant testis cancer. J Clin Oncol 18:358–362, 2000[Abstract/Free Full Text]

33. Bokemeyer C, Schmoll HJ: Secondary neoplasms following treatment of malignant germ cell tumors. J Clin Oncol 11:1703–1709, 1993[Abstract/Free Full Text]

34. Travis LB, Curtis RE, Storm H, et al: Risk of second malignant neoplasms among long-term survivors of testicular cancer. J Natl Cancer Inst 89:1429–1439, 1997[Abstract/Free Full Text]

35. Kollmannsberger C, Beyer J, Droz JP, et al: Secondary leukemia following high cumulative doses of etoposide in patients treated for advanced germ cell tumors. J Clin Oncol 16:3386–3391, 1998[Abstract]

36. Gerl A, Muhl-Bayer D, Hansmann G: The impact of chemotherapy on fasting plasma lipids in patients with germ cell tumors (GCT). Proc Am Soc Clin Oncol 18:1252a, 1999 (abstr)

37. Meinardi MT, Gietema JA, van der Graaf WT, et al: Long-term vascular an metabolic sequelae of chemotherapy for metastatic testicular cancer: A comparison with long-term survivors of stage I disease. Proc Am Soc Clin Oncol 19:331a, 2000 (abstr 1302)

38. Cullen MH, Stenning SP, Parkinson MC, et al: Short-course adjuvant chemotherapy in high-risk stage I nonseminomatous germ cell tumors of the testis: A Medical Research Council report. J Clin Oncol 14:1106–1113, 1996[Abstract/Free Full Text]

Submitted July 29, 2003; accepted January 22, 2003.

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

This article has been cited by other articles:

				E. Kikuchi, V. Margulis, P. I. Karakiewicz, M. Roscigno, S. Mikami, Y. Lotan, M. Remzi, C. Bolenz, C. Langner, A. Weizer, et al. Lymphovascular Invasion Predicts Clinical Outcomes in Patients With Node-Negative Upper Tract Urothelial Carcinoma J. Clin. Oncol., February 1, 2009; 27(4): 612 - 618. [Abstract] [Full Text] [PDF]

				C. M. Sturgeon, M. J. Duffy, U.-H. Stenman, H. Lilja, N. Brunner, D. W. Chan, R. Babaian, R. C. Bast Jr., B. Dowell, F. J. Esteva, et al. National Academy of Clinical Biochemistry Laboratory Medicine Practice Guidelines for Use of Tumor Markers in Testicular, Prostate, Colorectal, Breast, and Ovarian Cancers Clin. Chem., December 1, 2008; 54(12): e11 - e79. [Abstract] [Full Text] [PDF]

				P. Albers, R. Siener, S. Krege, H.-U. Schmelz, K.-P. Dieckmann, A. Heidenreich, P. Kwasny, M. Pechoel, J. Lehmann, S. Kliesch, et al. Randomized Phase III Trial Comparing Retroperitoneal Lymph Node Dissection With One Course of Bleomycin and Etoposide Plus Cisplatin Chemotherapy in the Adjuvant Treatment of Clinical Stage I Nonseminomatous Testicular Germ Cell Tumors: AUO Trial AH 01/94 by the German Testicular Cancer Study Group J. Clin. Oncol., June 20, 2008; 26(18): 2966 - 2972. [Abstract] [Full Text] [PDF]

				O. Khan and A. Protheroe Testis cancer Postgrad. Med. J., October 1, 2007; 83(984): 624 - 632. [Abstract] [Full Text] [PDF]

				D. S. Kaufman, M. A. Saksena, R. H. Young, and S. Tabatabaei Case 6-2007 -- A 28-Year-Old Man with a Mass in the Testis N. Engl. J. Med., February 22, 2007; 356(8): 842 - 849. [Full Text] [PDF]

				A. J. Stephenson, G. J. Bosl, R. J. Motzer, M. W. Kattan, J. Stasi, D. F. Bajorin, and J. Sheinfeld Retroperitoneal Lymph Node Dissection for Nonseminomatous Germ Cell Testicular Cancer: Impact of Patient Selection Factors on Outcome J. Clin. Oncol., April 20, 2005; 23(12): 2781 - 2788. [Abstract] [Full Text] [PDF]

				H. J. Schmoll, R. Souchon, S. Krege, P. Albers, J. Beyer, C. Kollmannsberger, S. D. Fossa, N. E. Skakkebaek, R. de Wit, K. Fizazi, et al. European consensus on diagnosis and treatment of germ cell cancer: a report of the European Germ Cell Cancer Consensus Group (EGCCCG) Ann. Onc., September 1, 2004; 15(9): 1377 - 1399. [Abstract] [Full Text] [PDF]

				D. Raghavan New Prognostic Factors for Stage I Testicular Cancer: But Will They Make It to Broadway? J. Clin. Oncol., November 15, 2003; 21(22): 4075 - 4076. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Albers, P. Articles by Hartmann, M. Search for Related Content PubMed PubMed Citation Articles by Albers, P. Articles by Hartmann, M. Social Bookmarking                            What's this?