Long-Term Outcome for Men With Teratoma Found at Postchemotherapy Retroperitoneal Lymph Node Dissection

Transcription

1 Original Article Long-Term Outcome for Men With Teratoma Found at Postchemotherapy Retroperitoneal Lymph Node Dissection Robert S. Svatek, MD 1, Philippe E. Spiess, MD 2, Debasish Sundi, BS 1, Shi-ming Tu, MD 3, Nizar M. Tannir, MD 3, Gordon A. Brown, MD 1, Ashish M. Kamat, MD 1, Christopher G. Wood, MD 1, and Louis L. Pisters, BS 1 BACKGROUND: Patients with pure teratoma within the postchemotherapy retroperitoneal lymph node dissection (PC-RPLND) specimen traditionally have been considered at low risk for disease progression. The objectives of this study were to determine the disease-related outcomes of patients who had pure teratoma identified at the time of PC-RPLND and to examine the prognostic value of clinical variables that were identified previously as important predictors of disease recurrence in these patients. METHODS: Between 1980 and 2003, 97 patients with metastatic nonseminomatous germ cell tumor and pure teratoma histology at the time of PC-RPLND were identified. The medical records of these patients were reviewed retrospectively for pertinent clinical and treatment-related outcomes. RESULTS: At a median follow-up of 7.4 years, 21 patients (22%) developed recurrent disease after PC-RPLND. The 5-year and 10-year probabilities (standard error) of freedom from disease recurrence were 81% 4% and 76% 5%, respectively. The postchemotherapy a-fetoprotein (AFP) level and mediastinal involvement at presentation were statistically significant predictors of disease recurrence on multivariate analysis. Nine of 97 patients (9.3%) died from testis cancer, and 4 patients died from other causes. CONCLUSIONS: In patients with pure teratoma histology at PC-RPLND, mediastinal involvement at presentation and the presence of an elevated AFP level before PC-RPLND predicted an unfavorable outcome. The absence of mediastinal involvement and normal AFP level, however, did not confirm freedom from disease recurrence. Patients who had teratoma at the time of PC-RPLND remained at considerable risk for disease progression because of the unpredictable nature of teratoma and the presence of unrecognized, active germ cell disease outside the retroperitoneum. Cancer 2009;115: VC 2009 American Cancer Society. KEY WORDS: teratoma, postchemotherapy, retroperitoneal lymph node dissection, predictors, outcome. Pure teratoma is identified in 40% of men who undergo retroperitoneal lymph node dissection (RPLND) for metastatic nonseminomatous germ cell tumor (NSGCT) after chemotherapy, whereas viable germ cell tumor (GCT) is discovered in 10% to 15%, and fibrosis or necrosis is discovered in 40% to 50%. 1 Although teratoma histology is considered benign, the natural history of postchemotherapy Corresponding author: Robert S. Svatek, MD, Department of Urology, the University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, TX 77030; Fax: (713) ; rssvatek@sbcglobal.net 1 Departments of Urologic Oncology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas; 2 Department of Interdisciplinary Oncology, H. Lee Moffitt Cancer Center and Research Institute Tampa, Florida; 3 Department of Genitourinary Medical Oncology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas See editorial on pages , this issue. Received: April 18, 2008; Revised: July 26, 2008; Accepted: August 1, 2008 Published online: January 20, 2009, VC 2009 American Cancer Society DOI: /cncr.24145, Cancer March 15, 2009

2 Outcome for Postchemotherapy Teratoma/Svatek et al teratoma is unpredictable because of the potential for malignant degeneration or the development of growing teratoma syndrome. 2-5 In addition, patients who undergo postchemotherapy RPLND (PC-RPLND) may harbor micrometastases that are undetectable with current diagnostic modalities. 6,7 Recurrence rates for patients with teratoma histology at PC-RPLND range from 6% to 39%. 3,4,8-13 Previous series have identified several prognostic variables that may identify patients at high-risk for disease recurrence despite the detection of pure teratoma at PC-RPLND. 3,14 Loehrer et al 3 identified mediastinal primary tumor site, tumor burden, and the presence of immature teratoma histology as significant predictors of disease recurrence on univariate analysis. Carver et al 14 identified International Germ Cell Cancer Collaborative Group (IGCCCG) risk classification 15 and postchemotherapy tumor size as important predictors of disease recurrence. In the current study, we examined the long-term outcome of men who had pure teratoma histology identified at time they underwent PC-RPLND at our institution. In addition, we sought to validate the prognostic value of clinical variables that previously were identified as important predictors of disease recurrence in these patients. MATERIALS AND METHODS Study Design This study was performed after it was approved by the University of Texas M. D. Anderson Cancer Center Institutional Review Board. From the institution s tumor registry, 408 men who were treated for primary testis tumor were identified from January 1980 to July Of the 377 patients for whom complete medical charts were available, 141 had clinical stage I or IS disease and were excluded. From the remaining 236 patients who underwent PC-RPLND, 97 patients (41%) had pure teratoma histology identified at the time of PC-RPLND. Clinical stage was assigned according to the 1997 tumor-lymph node-metastasis (TNM) staging system. Patients were categorized into 3 prognostic groups (good risk, intermediate risk, and poor risk) according to the 1997 IGCCCG guidelines. 15 IGCCCG risk categories could be assigned in 83 patients; the remaining patients (n ¼ 14) could not be assigned to any risk group because of missing tumor marker information. For all patients, a complete medical evaluation was performed before the PC-RPLND, including measurements of serum a-fetoprotein (AFP), human chorionic gonadotropin (hcg), and lactate dehydrogenase (LDH) and radiologic evaluation comprised of chest and abdominal/pelvic computed tomography scans. The largest transverse dimension of the retroperitoneal mass, as determined by abdominal imaging, was noted. Clinical variables, including diagnostic findings, treatment type, and patient outcomes, were recorded along with postoperative surgical variables. All patients received systemic chemotherapy before the underwent PC-RPLND. The chemotherapeutic regimens that were selected for patients varied, depending on the protocols used around the time of diagnosis. In the 1980s, patients received alternating cycles of cyclophosphamide, doxorubicin, and cisplatin (CISCA) and vinblastine plus bleomycin (VB). All patients who had NSGCT received 2 cycles of chemotherapy beyond normalization of serum tumor markers, thereby receiving a minimum of 4 chemotherapy cycles. 16 In the 1990s, patients continued to receive alternating cycles of CISCA and VB with a 20% dose reduction. 17 However, patients in the poor-risk IGCCCG category were given alternating dose-dense chemotherapy. 18 Before 1997, patients with small-volume disease received combined carboplatin, etoposide, and bleomycin. In recent years, 3 cycles of combined bleomycin, etoposide, and cisplatin or 4 cycles of combined etoposide and cisplatin have been given to patients with good-risk NSGCT. For analysis, AFP was analyzed both as a continuous and as a categorical variable (normal AFP, <15 ng/ml), and hcg was analyzed as a continuous or categorical variable (normal hcg, <2.2 miu/ml). 19 A full bilateral-template PC-RPLND was performed in all patients, and all pathologic specimens from the testis primary site and the retroperitoneum or other metastatic sites were reviewed by genitourinary pathologists. The pathologic findings of the PC-RPLND pure teratoma specimens were categorized as either immature teratoma or mature teratoma. Statistical Analysis Continuous variables are reported as medians with the interquartile (IQ) range. The Kaplan-Meier method was Cancer March 15,

3 Original Article used to estimate disease-specific survival and recurrencefree survival, and differences were assessed with the logrank test statistic. Disease-specific death was defined as death resulting from testicular cancer or its treatment in the setting of progressive disease. Patients who died of other causes were censored at the date of their death. Univariate and multivariate survival analyses were performed using the Cox proportional hazards regression model. Cases with missing values were excluded in a list-wise manner during univariate and multivariate analysis. The lowest category was used as the referent category when calculating the hazards ratios for categorical variables. Because of the small number of patients in each subcategory, patients with clinical stage IIA, IIB, and IIC disease were compared against patients with clinical stage III disease during univariate and multivariate analyses. Statistical significance in this study was set as P.050. All reported P values are 2-sided. All analyses were performed with SPSS version 13.0 (SPSS Inc. Chicago, Ill). RESULTS The clinical characteristics of the 97 patients who had pure teratoma identified in the PC-RPLND specimen are listed in Table 1. Primary testis tumor histology was available for 96 patients. The most common histology was mixed NSGCT (83.5%), and the remaining tumor histologic types were pure embryonal carcinoma (9.3%), pure teratoma (5.2%), and pure seminoma (1.0%). The clinical tumor classification of the primary tumor most frequently was T1 (57.7%) followed by T2 (25.8%), and most patients presented with a lymph node status of either N2 (43.3%) or N3 (39.2%). Clinical M1a and M1b disease was present in 37.1% and 13.4% of patients, respectively. The median serum AFP level before PC-RPLND was 2.8 ng/ml (IQ range, ng/ml). The median serum hcg level before PC-RPLND was 1 miu/ml (IQ range, 1-1 miu/ml). The median postchemotherapy retroperitoneal lymph node size was 3 cm (IQ range, cm). A median of 17 lymph nodes was removed per PC- RPLND specimen (IQ range, lymph nodes). Histology at PC-RPLND was available for 91 patients, including 14 (15.4%) who had immature teratoma and 77 (84.6%) who had mature teratoma. No histology was identified as teratoma with malignant degeneration. The Table 1. Clinical Characteristics of 97 Men With Teratoma Identified on Retroperitoneal Histology After Chemotherapy Characteristic Median age [IQR], y 32 [20-37] Median no. of chemotherapy 5 [4-10] cycles prior to RPLND [IQR] Median postchemotherapy size of 3 [2.4-5] retroperitoneal mass [IQR], cm Race No. of Patients (%) White 71 (73.2) Black 24 (24.7) Hispanic 2 (2.1) Site at presentation Retroperitoneum 90 (92.8) Lungs 42 (43.3) Mediastinum 9 (9.3) Supraclavicular lymph node(s) 3 (3.1) Liver 12 (12.4) Clinical stage at presentation IIA 8 (8.2) IIB 27 (27.8) IIC 18 (18.6) III 44 (45.4) IGCCCG risk category Good 37 (38.1) Intermediate 18 (18.0) Poor 28 (28.8) Second-line chemotherapy 38 (39.2) Postchemotherapy AFP level, ng/ml Normal, <15 86 (95.6) Elevated, 15 4 (4.4) Postchemotherapy hcg level, miu/ml Normal, < (94.5) Elevated, (5.5) Teratoma subtype Mature 77 (79.4) Immature 14 (14.4) IQR indicates interquartile range; RPLND, retroperitoneal lymph node dissection; IGCCCG, International Germ Cell Cancer Collaborative Group; AFP, a-fetoprotein; hcg, human chorionic gonadotropin. median follow-up duration was 7.4 years (IQ range, years). The characteristics of all patients who experienced disease recurrence after PC-RPLND are listed in Table 2. Twenty-one patients (21.6%) experienced disease recurrence at a median of 11.4 months (IQ range, months) after PC-RPLND. Eleven patients (52%) experienced disease recurrence within the first year, and 10 patients experienced recurrence >1 year after RPLND. The sites of disease recurrence included the mediastinum 1312 Cancer March 15, 2009

4 Outcome for Postchemotherapy Teratoma/Svatek et al Table 2. Clinical and Pathologic Characteristics of Patients who Developed Disease Recurrence (n¼21) AJCC Stage IGCCCG Subtype Time to Recurrence, mo Site of Recurrence Histology of Recurrence Treatment of Recurrence Time to Death or Last Follow-up, mo Alive, NED III Poor Mature 9.13 Neck Teratoma Surgery IIA Mature Retroperitoneum Viable GCT Chemotherapy and surgery III Poor Mature 18.7 Suprahiatal Teratoma Surgery 88.8 III Good Mature 20.8 Bone (rib) Teratoma Surgery IIB Poor Mature 61.5 Neck Teratoma Surgery IIB Mature Neck Teratoma Surgery IIB Poor Immature 73.5 Mediastinum Teratoma Surgery III Good Mature 11.0 Mediastinum Teratoma Surgery 82.4 Alive, with disease III Poor Mature 31.0 Retroperitoneum and neck Teratoma Surgery 41.2 IIA Good Mature 21.1 Suprahiatal Viable GCT Surgery Dead (other causes) III Poor Mature 11.4 Lungs Teratoma Surgery 48.0 III Poor 1.4 Mediastinum Teratoma Surgery 77.4 Dead of disease III Intermediate Mature 4.7 Mediastinum Teratoma 7.5 III Poor Mature 6.5 Lungs and liver Viable GCT Chemotherapy 17.8 III Poor Mature 9.7 Lungs Teratoma 9.7 III Good Immature 1.7 Lungs Viable GCT 2.8 III Intermediate Mature 6.7 Retroperitoneum Viable GCT Chemotherapy and surgery 39.2 III Poor Mature 12.2 Bone (femur) Teratoma Chemotherapy and surgery IIC Intermediate Mature 7.3 Suprahiatal Viable GCT Chemotherapy and surgery 7.3 III Mature 1.0 Suprahiatal Viable GCT Chemotherapy and radiation 1.0 IIB Good Lungs and suprahiatal Viable GCT Chemotherapy AJCC indicates American Joint Committee on Cancer; IGCCCG, International Germ Cell Cancer Collaborative Group; NED, no evidence of disease; GCT, germ cell tumor; suprahiatal, above the esophageal hiatus. Cancer March 15,

5 Original Article Table 3. Cox Proportional Hazards Model of Clinical and Pathologic Features for the Prediction of Recurrence in 97 Patients Who Had Teratomatous Elements Identified at Postchemotherapy Retroperitoneal Lymph Node Dissection Univariate Model Multivariate Model Characteristic HR 95% CI P HR 95% CI P Age AJCC clinical stage III vs II Presence of teratoma in primary tumor Presence of embyronal in primary tumor Site at presentation Retroperitoneum Lungs Mediastinum < Supraclavicular lymph node(s) Liver IGCCCG risk group Intermediate vs good Poor vs good No. of chemotherapy cycles Postchemotherapy AFP level Continuous Elevated vs nonelevated < Postchemotherapy hcg level Continuous Elevated vs nonelevated Postchemotherapy size of residual mass PC-RPLND histology: Mature vs immature HR indicates hazard ratio; 95% CI, 95% confidence interval; AJCC, American Joint Committee on Cancer; IGCCCG, International Germ Cell Cancer Collaborative Group; AFP, a-fetoprotein; hcg, human chorionic gonadotropin; PC-RPLND, postchemotherapy retroperitoneal lymph node dissection. (n ¼ 4), neck (n ¼ 4), lung(s) (n ¼ 4), suprahiatal (n ¼ 5), bone (n ¼ 2), liver (n ¼ 1), and retroperitoneum (n ¼ 3). Recurrence histology revealed mature teratoma (62%) or viable GCT (38%) in patients with disease recurrence (Table 2). The median number of chemotherapy cycles was similar for those who experienced disease recurrence (median, 5, cycles; IQ range, cycles) and those without disease recurrences (median, 5 cycles; IQ range, 2-8 cycles). The median time from PC-RPLND to disease recurrence was 21 months (IQ range, months) for those who remained alive or who died of other causes at the time they were censored compared with a median of 6.7 months (IQ range, months) for those who died of testis cancer (P ¼.028). The 5-year and 10-year probabilities of freedom from disease recurrence (standard error) were 81% 4% and 76% 5%, respectively. The results of univariate and multivariate analyses for predicting disease recurrence are shown in Table 3. Higher clinical stage, disease site (mediastinum or lung) at presentation, poor IGCCCG risk group, and an elevated postchemotherapy AFP level were significant predictors of disease recurrence on univariate analysis. The presence of mediastinal disease at presentation and postchemotherapy AFP level retained statistical significance as predictors of recurrence after adjusting for covariates. Because clinical stage and disease site at presentation have overlapping features, a separate multivariate analysis was performed after removal of the clinical stage variable. In this model, poor IGCCCG risk group and the involvement of liver and lung at presentation still were not associated significantly with disease recurrence. The presence of mediastinal disease at presentation and postchemotherapy AFP level retained statistical significance as predictors of recurrence, but their hazard ratios changed somewhat (mediastinal involvement: hazards ratio, 7.4 [95% confidence interval (95% CI), ]; elevated vs normal AFP: hazards ratio, 14.5 [95% CI,, ]). To represent the individual and combined abilities of these variables to predict disease recurrence, a cross tabulation is provided (Table 4) Cancer March 15, 2009

6 Outcome for Postchemotherapy Teratoma/Svatek et al Table 4. Cross Tabulation of Disease Recurrence According to Postchemotherapy a-fetoprotein Elevation and the Presence of Mediastinum Involvement at Presentation No. of Patients Variable Recurrence No Recurrence Postchemotherapy AFP level, ng/ml Normal, < Elevated, Disease site at presentation No mediastinum involvement Mediastinum involvement 7 2 Combined variables Normal AFP and no mediastinum involvement Elevated AFP and/or mediastinum involvement AFT indicates a-fetoprotein Of the 97 patients who had pure teratoma at PC- RPLND, 9 patients (9.3%) died from testis cancer, and 4 patients died from other causes. For these 9 patients, the median time from PC-RPLND to death was 9.7 months (IQ range, months). Histology at the time of recurrence demonstrated viable GCT in 6 patients and teratoma was in 3 patients. The 10-year cumulative diseasespecific survival probability (standard error) was 90% 4%. The 5- and 10-year cumulative overall survival probabilities (standard error) were 81% 5% and 76% 6%, respectively. DISCUSSION Patients with pure teratoma histology on PC-RPLND have a variable clinical course that depends on a complete surgical dissection and the presence of unrecognized viable germ cell elements outside the retroperitoneum. At a median follow-up >7 years, this series provides outcomes for patients with the longest follow-up duration published to date. Disease recurrence was identified in 21 of 97 patients (22%), including 8 recurrences with viable GCT. The corresponding 10-year probability of freedom from disease recurrence (standard error) was 76% 5%. Postchemotherapy AFP levels and the presence of mediastinal involvement at presentation were associated significantly with disease recurrence. Although it assumed generally that they have a favorable prognosis, several contemporary series have revealed a high frequency of disease recurrence in patients who have pure teratoma histology after chemotherapy. 3,4,14 Loehrer et al 3 identified 51 patients with pure teratoma on PC-RPLND and reported disease recurrence in 20 (39%). The histology of the recurrence was distributed evenly between pure teratoma and viable disease. Sonneveld et al 4 reported disease recurrence in 9 of 51 patients (9.8%), including 5 recurrences with mature teratoma, 3 with non-gct, and 1 with viable GCT. In what to our knowledge is the largest series published to date, Carver et al 14 observed disease recurrence in 30 of 210 patients (14%) who had retroperitoneal teratoma after chemotherapy. Ten of those 30 patients (33%) had developed disease recurrence with teratoma, 5 (17%) had teratoma with malignant transformation, and 15 (50%) had viable GCT. Disease recurrence for patients with pure teratoma at PC-RPLND is caused either by incomplete retroperitoneal surgical resection or by unrecognized disease metastases at the time of surgery. Similar to other series, 3,14 we observed that most recurrences occurred in extraretroperitoneal sites. Nevertheless, despite undergoing a complete bilateral-template PC-RPLND resection, 3 patients developed recurrent tumor in the retroperitoneum, including 1 patient who subsequently died from disease. This finding emphasizes the need for a meticulous surgical approach to treating these patients regardless of suspected histologic subtype. The possibility of unrecognized disease metastasis at the time of surgery must be considered when following patients after PC-RPLND. Histologic findings of fibrosis and/or necrosis on postchemotherapy retroperitoneal histology have been associated with disease recurrence rates of 5% to 18%. 6,7,20 Active GCT outside the retroperitoneum developed in 19% of our patients despite normal tumor markers and pure teratoma at the time of PC- RPLND. Staging for patients who undergo PC-RPLND clearly is subject to the limitations of current staging modalities for testis cancer. The ability to identify variables that could identify the patients who are likely to develop recurrent disease despite having retroperitoneal teratoma after chemotherapy would be invaluable for prognostication and timely delivery of adjuvant chemotherapy. Loehrer et al 3 identified Cancer March 15,

7 Original Article the initial tumor burden, the presence of immature teratoma, and mediastinal disease site as significant predictors of disease recurrence on univariate analysis, and the presence of immature teratomatous elements and mediastinal disease site were significant on multivariate analyses. In their multivariate analyses, Carver et al 14 reported that increased size of the residual mass after chemotherapy and higher IGCCCG risk classification were significant predictors of disease recurrence. Our analysis also suggests a predictive role for IGCCCG risk stratification, although this predictor was unable to maintain prognostic significance on multivariate analyses. This discrepancy may be caused in part by differences between study populations. Generally, patients from our series represent a more heavily treated population: 39% of men underwent secondline chemotherapy compared with 9% from the Memorial Sloan-Kettering Cancer Center (MSKCC) series. 14 In addition, the MSKCC dataset excluded 5 patients who had progressive, viable GCT despite RPLND. 14 In contrast, our series included all patients with disease recurrence after RPLND, even those who developed a recurrence within 2 months. Although these patients may never have experienced a disease-free interval, RPLND was performed with curative intent and in the absence of obvious metastatic lesions. In the current series, patients with mediastinal disease site at presentation or an elevated serum AFP after chemotherapy were at higher risk for disease recurrence. This is consistent with the finding that, given the early time to recurrence and the absence of teratoma with malignant degeneration in recurrence histology, many of our patients with recurrence likely had unrecognized, viable germ cell elements. Historically elevated tumor marker levels after chemotherapy have been a relative contraindication to RPLND. However, several contemporary series have demonstrated that select patients with elevated tumor marker levels can be cured with surgery Therefore, these patients were not excluded from our analysis. Although elevated serum AFP level before PC-RPLND and the presence of mediastinal disease site at presentation were identified as predictors of disease recurrence in this analysis, these data do not support altering surveillance strategies based on the presence or absence of these features. The confidence intervals for the hazard ratios associated with these 2 prognostic variables are wide, as shown in Table 4, indicating imprecision in the corresponding point estimates. This also is explained by the finding that, although all 4 patients with elevated postchemotherapy AFP levels experienced disease recurrence, most patients with disease recurrence did not have elevated postchemotherapy AFP levels. Furthermore, although 7 of 9 patients with mediastinal disease at presentation suffered disease recurrence after PC-RPLND, approximately 66% of patients who experienced disease recurrence did not present with mediastinal disease involvement. Nine of the patients in our study (9.3%) died from testis cancer, including 3 patients who recurred with teratoma. Loehrer et al 3 also observed that a significant number of patients with postchemotherapy teratoma died from disease. Of their 51 patients, 15.7% died from testis cancer, including 6 who developed disease recurrence with active GCT and 2 who recurred with teratoma. Others have reported lower proportions of cancer-related death among their patients. Sonneveld et al 4 reported a 2% incidence of death from disease with a median followup duration of 51 months. Similarly, Carver et al 14 observed death from disease in 4.7% of their 210 patients. We acknowledge several limitations to the current study. Because the study is retrospective, it suffers from an inherent selection bias, which is common to most RPLND surgical series. 24 The assessment of surgical specimen pathology is subject to differences in reporting practices among our institution s genitourinary pathologists. In addition, our surgical series extended over 25 years; changes occurred in preoperative chemotherapeutic regimens during that period, and the generalizability of these findings is not known. Despite these limitations, this analysis is an important contribution to the existing knowledge of outcomes for patients with retroperitoneal teratoma after chemotherapy. In conclusion, patients with retroperitoneal pure teratoma after chemotherapy are at significant risk of disease progression because of the unpredictable nature of teratoma and the presence of unrecognized, viable germ cell disease outside the retroperitoneum. Although the presence of mediastinal involvement at presentation or an elevated AFP before PC-RPLND help predict which patients will suffer from recurrence, these variables are not sufficiently accurate to be used as the bases for surveillance modification or treatment decisions. Therefore, until improved diagnostic modalities are available to identify patients who harbor micrometastases, patients with pure 1316 Cancer March 15, 2009

8 Outcome for Postchemotherapy Teratoma/Svatek et al teratoma at the time of postchemotherapy RPLND should undergo diligent surveillance and prompt salvage therapy when clinically indicated. Conflict of Interest Disclosures The authors made no disclosures. References 1. Steyerberg EW, Keizer HJ, Fossa SD, et al. Prediction of residual retroperitoneal mass histology after chemotherapy for metastatic nonseminomatous germ cell tumor: multivariate analysis of individual patient data from 6 study groups. J Clin Oncol. 1995;13: Logothetis CJ, Samuels ML, Trindade A, et al. The growing teratoma syndrome. Cancer. 1982;50: Loehrer PJ Sr, Hui S, Clark S, et al. Teratoma following cisplatin-based combination chemotherapy for nonseminomatous germ cell tumors: a clinicopathological correlation. J Urol. 1986;135: Sonneveld DJ, Sleijfer DT, Koops HS, et al. Mature teratoma identified after postchemotherapy surgery in patients with disseminated nonseminomatous testicular germ cell tumors: a plea for an aggressive surgical approach. Cancer. 1998;82: Spiess PE, Kassouf W, Brown GA, et al. Surgical management of growing teratoma syndrome: the M. D. Anderson Cancer Center experience. J Urol. 2007;177: ; discussion Debono DJ, Heilman DK, Einhorn LH, et al. Decision analysis for avoiding postchemotherapy surgery in patients with disseminated nonseminomatous germ cell tumors. J Clin Oncol. 1997;15: Spiess PE, Tannir NM, Brown GA, et al. Recurrence in nonseminomatous germ cell testis tumor patients with no viable tumor at postchemotherapy retroperitoneal lymph node dissection. Urology. 2007;70: Aprikian AG, Herr HW, Bajorin DF, et al. Resection of postchemotherapy residual masses and limited retroperitoneal lymphadenectomy in patients with metastatic testicular nonseminomatous germ cell tumors. Cancer. 1994;74: Davey DD, Ulbright TM, Loehrer PJ, et al. The significance of atypia within teratomatous metastases after chemotherapy for malignant germ cell tumors. Cancer. 1987;59: Donohue JP, Roth LM, Zachary JM, et al. Cytoreductive surgery for metastatic testis cancer: tissue analysis of retroperitoneal masses after chemotherapy. J Urol. 1982;127: Fossa SD, Qvist H, Stenwig AE, et al. Is postchemotherapy retroperitoneal surgery necessary in patients with nonseminomatous testicular cancer and minimal residual tumor masses? J Clin Oncol. 1992;10: Sella A, el Naggar A, Ro JY, et al. Evidence of malignant features in histologically mature teratoma. J Urol. 1991; 146: Tait D, Peckham MJ, Hendry WF, et al. Post-chemotherapy surgery in advanced non-seminomatous germ-cell testicular tumours: the significance of histology with particular reference to differentiated (mature) teratoma. Br J Cancer. 1984;50: Carver BS, Shayegan B, Serio A, et al. Long-term clinical outcome after postchemotherapy retroperitoneal lymph node dissection in men with residual teratoma. J Clin Oncol. 2007;25: [Noauthors listed] International Germ Cell Consensus Classification: a prognostic factor-based staging system for metastatic germ cell cancers. International Germ Cell Cancer Collaborative Group. J Clin Oncol. 1997;15: Logothetis CJ, Samuels ML, Selig D, et al. Improved survival with cyclic chemotherapy for nonseminomatous germ cell tumors of the testis. J Clin Oncol. 1985;3: Fizazi K, Do KA, Wang X, et al. A 20% dose reduction of the original CISCA/VB regimen allows better tolerance and similar survival rate in disseminated testicular non-seminomatous germ-cell tumors: final results of a phase III randomized trial. Ann Oncol. 2002;13: Fizazi K, Prow DM, Do KA, et al. Alternating dose-dense chemotherapy in patients with high volume disseminated non-seminomatous germ cell tumours. Br J Cancer. 2002; 86: Stephenson AJ, Bosl GJ, Motzer RJ, et al. Retroperitoneal lymph node dissection for nonseminomatous germ cell testicular cancer: impact of patient selection factors on outcome. J Clin Oncol. 2005;23: Toner GC, Panicek DM, Heelan RT, et al. Adjunctive surgery after chemotherapy for nonseminomatous germ cell tumors: recommendations for patient selection. J Clin Oncol. 1990;8: Albers P, Ganz A, Hannig E, et al. Salvage surgery of chemorefractory germ cell tumors with elevated tumor markers. J Urol. 2000;164: Beck SD, Foster RS, Bihrle R, et al. Outcome analysis for patients with elevated serum tumor markers at postchemotherapy retroperitoneal lymph node dissection. J Clin Oncol. 2005;23: Murphy BR, Breeden ES, Donohue JP, et al. Surgical salvage of chemorefractory germ cell tumors. J Clin Oncol. 1993;11: Donohue JP, Leviovitch I, Foster RS, et al. Integration of surgery and systemic therapy: results and principles of integration. Semin Urol Oncol. 1998;16: Cancer March 15,