Surgical Management of Late Relapse on Surveillance in Patients Presenting With Clinical Stage I Testicular Cancer

Surgical Management of Late Relapse on Surveillance in Patients Presenting With Clinical Stage I Testicular Cancer

Oncology Surgical Management of Late Relapse on Surveillance in Patients Presenting With Clinical Stage I Testicular Cancer Kevin R. Rice, Stephen D. ...

234KB Sizes 0 Downloads 11 Views

Oncology Surgical Management of Late Relapse on Surveillance in Patients Presenting With Clinical Stage I Testicular Cancer Kevin R. Rice, Stephen D. W. Beck, Jose A. Pedrosa, Timothy A. Masterson, Lawrence H. Einhorn, and Richard S. Foster OBJECTIVE




To determine survival outcomes in clinical stage I germ cell tumor (GCT) patients requiring retroperitoneal lymph node dissection (RPLND) for late relapse (LR) occurring while on surveillance. The Indiana University Testis Cancer Database was queried from 1985 to 2013 to identify all patients who presented with clinical stage I GCT, elected surveillance, relapsed 2 years after initial diagnosis, and underwent RPLND in treatment of their LR. Clinical, pathologic, and treatment characteristics were reviewed. Twenty-eight patients met inclusion criteria. The mean age at diagnosis was 29.3 years. Testicular primary was pure seminoma in 2, intratubular germ cell neoplasia with scar in 1, nonseminomatous GCT in 24, and unknown in 1 patient. The median time from diagnosis to relapse was 48.5 months (range, 28-321 months). At relapse, serum tumor markers were elevated in 13 patients (46.4%). Nineteen patients were given cisplatin-based chemotherapy at LR. RPLND was initial management of LR in 9. At RPLND, 10, 5, and 13 patients demonstrated fibrosis, teratoma, and viable malignancy, respectively. On the last follow-up, 24 patients (85.7%) were free of disease and 4 patients (14.3%) had died of their disease. When examining outcomes among patients undergoing RPLND at LR of GCT, it appears that patients experiencing LR on surveillance have more favorable histology and survival outcomes than previously reported for unselected patients experiencing LR. UROLOGY 84: 886e891, 2014. Published by Elsevier Inc.


ate relapse (LR) of testicular germ cell tumors (GCT) is defined as disease recurrence 24 months from complete response to primary management in the absence of a second primary GCT. This is a rare phenomenon occurring in 3.6% GCT patients.1 LRs are rarely curable with chemotherapy alone. Thus, surgical resection offers the only chance for durable cure in most cases.2 Additionally, there seems to be a relatively high rate of malignant transformation to somatic-type malignancies such as sarcoma and adenocarcinoma in patients experiencing LR.3,4 These aggressive characteristics at LR have resulted in relatively poor survival outcomes, with reported overall survival rates ranging from 46.9% to 63%.3-5 Pure teratoma at retroperitoneal lymph node dissection (RPLND) and chemotherapy-naïve status have been

associated with improved survival in patients experiencing LR of GCT.2,4 Interestingly, in a review of LR in patients initially presenting with clinical stage I (CS-I) disease, outcomes appeared nearly identical to those reported in patients presenting with metastatic GCT.6 However, nearly two-thirds of the patients in this CS-I study had previously received chemotherapy either as adjuvant treatment for elevated serum tumor marker levels (STMs), pathologic stage-II disease, or for early relapse. With the increasing utilization of surveillance regimens in the management of CS-I GCT, we sought to define the clinical and histologic characteristics of LR in a group of patients who went on to RPLND.

METHODS Financial Disclosure: The authors declare that they have no relevant financial interests. From the Department of Urology, Indiana University School of Medicine, Indianapolis, Indiana; and the Indiana University Melvin and Bren Simon Cancer Center, Indianapolis, Indiana Reprint requests: Kevin R. Rice, M.D., Department of Urology, 8901 Wisconsin Avenue, Building 9, 2nd Floor, Bethesda, MD 20889. E-mail: [email protected] Submitted: February 17, 2014, accepted (with revisions): May 12, 2014


Published by Elsevier Inc.

The Indiana University (IU) RPLND Database was queried from 1985 to 2013 to identify all patients undergoing RPLND 24 months after presenting with stage-I testicular GCT. A chart review was then performed to determine pretreatment characteristics, characteristics at LR, chemotherapeutic regimens, specific pathologic findings, and follow-up data. 0090-4295/14

Table 1. Pretreatment characteristics Characteristic No. of patients Mean age at diagnosis, y Testicular pathology (no. of patients) Seminoma NSGCT ITGCN/regressed GCT Unknown Time to LR, mo, median (range) Elevated STM, n (%) AFP, n (%) HCG, n (%) Mass size, cm, median (range)

Table 2. Indications for primary chemotherapy at LR Value 28 29.3 2 23 1 2 48.5 (28-321) 13 (46.4) 5 (17.9) 9 (32.1) 4.2 (1.5-13)

AFP, alpha fetoprotein; GCT, germ cell tumor; HCG, human chorionic gonadotropin; ITGCN, intratubular germ cell neoplasia; LR, late relapse; No, number; NSGCT, nonseminomatous germ cell tumor; STM, serum tumor marker.

Patients and/or their home physicians were contacted by phone if they had not been seen in the last 12 months. Patients were excluded if they demonstrated evidence of metastatic disease, elevated STM levels, received chemotherapy, received radiotherapy, or underwent RPLND before LR. Patients who were completely lost to follow-up until presenting with LR were also excluded to avoid inclusion of patients with delayed presentation of an early relapse. Patients were referred at various stages in their management of LR, but in general, patients with normal STM levels and easily resectable unifocal retroperitoneal (RP) masses at LR were managed with RPLND. Patients with markedly elevated STM levels, RP masses requiring potentially morbid resection, and/or multifocal recurrences were first administered systemic chemotherapy with post-chemotherapy (PC)-RPLND being reserved for those demonstrating a residual RP mass. Rates are reported for categorical patient features, and measures of central tendency and dispersion are reported for continuous patient features. Kaplan-Meier unadjusted survival analysis curves were used to determine cancer-specific survival for the study group.

RESULTS Twenty-eight patients met inclusion criteria for the study. Clinical characteristics at initial diagnosis as well as at LR are listed in Table 1. Two patients presented with pure testicular seminoma, 23 patients with nonseminomatous germ cell tumor, and 1 patient with intratubular germ cell neoplasia with scar consistent with a regressed GCT. The mean age at initial GCT presentation was 29.3 years. The median time from diagnosis to LR was 48.5 months with a range of 28 months to 27 years. STM levels were elevated in 13 patients (46.4%) at LR, including elevation of human chorionic gonadotropin in 9 and alpha fetoprotein in 5 patients. Twenty-six of 28 patients (93.7%) experienced RP relapse. Two remaining patients demonstrated isolated pelvic recurrences. Four patients experienced relapse in additional sites along with RP recurrence. Information regarding the size of tumor mass at LR was available for 10 patients. The median greatest axial diameter was 4.2 cm (range, 1.5-13 cm). UROLOGY 84 (4), 2014

Indication for Chemotherapy at LR

No. Patients

Elevated tumor markers and high-volume disease* Elevated tumor markers alone Administered at OSH for RP mass alone Stage-III disease Biopsy-proven viable GCT

8 5 4 1 1

OSH, outside hospital; other abbreviations as in Table 1. * Masses >5 cm, multifocal, or stage III at LR.

Primary RPLND was performed in 9 patients with unifocal masses and normal STM levels. All but 1 patient demonstrated viable malignancy, with 4 having seminoma, 2 having embryonal carcinoma, 1 having primitive neuroectodermal tumor (PNET), and 1 having adenocarcinoma. The remaining patient was found to have teratoma at RPLND. Nineteen patients received standard cisplatin-based chemotherapy at LR before RPLND. Indications for primary chemotherapy are listed in Table 2. Most of these patients were diagnosed and initially managed with chemotherapy before referral to IU. In most cases, it appears that chemotherapy was an appropriate initial choice. However, at IU, the patients with normal or minimally elevated STM levels and a single RP mass may have been managed primarily with RPLND depending on the predicted morbidity of the resection. The histology at PCRPLND for these patients is depicted in Figure 1. One patient experienced complete radiographic and serologic remission to chemotherapy. However, this patient relapsed in his retroperitoneum 12 months later and was found to have teratoma at PC-RPLND. Sixteen patients had partial responses. Ten of these patients demonstrated fibrosis only, 2 harbored teratoma, 2 patients had yolk sac tumor, and 1 patient each demonstrated embryonal carcinoma and PNET. Two patients demonstrated progressive disease on chemotherapy. One of these patients with a persistently rising human chorionic gonadotropin level was found to have viable choriocarcinoma and the other with a slowly enlarging mass harbored teratoma only. The histology at RPLND and associated status at last follow-up for the entire study group are demonstrated in Table 3. Of note, the 4 patients who are dead of disease demonstrated either viable GCT or transformation to somatic-type malignancy at RPLND. In addition to standard RPLND, auxiliary procedures and/or resections were required in 7 patients. These included retrocrural resection in 2 patients, pelvic lymphadenectomy in 3 patients, hepatic lobectomy in 1 patient, nephrectomy in 1 patient, excision of peritoneal implants in 1 patient, and complete resection of the abdominal inferior vena cava and aorta with placement of an aortic tube graft in 1 patient. Five patients received some form of postoperative chemotherapy. Three patients received adjuvant chemotherapy after primary RPLND; adjuvant bleomycin, 887

Figure 1. Kaplan-Meier cancer-specific survival analysis. (Color version available online.)

Table 3. Histology at RPLND and outcome Surgery and Histology PC-RPLND (n ¼ 19) Fibrosis Teratoma Viable malignancy Embryonal carcinoma Choriocarcinoma Yolk sac tumor PNET Primary RPLND (n ¼ 9) Teratoma Seminoma Embryonal carcinoma Adenocarcinoma PNET

No. Patients


10 4 5 1 1 2 1


1 4 2 1 1


DOD, dead of disease; NED, no evidence of disease; PNET, primitive neuroectodermal tumor; other abbreviations as in Table 1.

etoposide, and cisplatin  2 was given to 1 patient with seminoma and one patient with embryonal carcinoma in the primary RPLND specimen. Adjuvant PNET-specific chemotherapy was given to one of the patients found to have PNET in his RPLND specimen (the second with PNET at PC-RPLND refused therapy and died of his disease). All 3 of these patients had no evidence of disease (NED) at last follow-up. Adjuvant chemotherapy was not offered to patients with viable GCT at PC-RPLND, as these tumors were thought to most likely represent chemorefractory disease. Oral etoposide was administered to a patient with chemorefractory choriocarcinoma after desperation RPLND. He subsequently underwent reresection of pelvic and RP recurrence followed by salvage gemcitabine. He died of disease 10 months after his initial resection. Another patient received folinic acid, 888

5-fluorouracil, and oxaliplatin for peritoneal carcinomatosis with adenocarcinoma. This patient eventually died of progressive disease. Median follow-up from RPLND for the study group was 21 months (range, 1-237 months). At last follow-up, 24 patients (85.7%) had NED and 4 patients (14.3%) were died of disease. Kaplan-Meier analysis revealed a 2-year cancer-specific survival of 79% (Fig. 1). The 4 patients that died of their disease had viable malignancy at RPLND. These included choriocarcinoma, embryonal carcinoma, PNET, and adenocarcinoma in 1 patient each.

COMMENT LR of testis cancer in patients managed on surveillance is an entity that has not been extensively described in literature. Reported rates of LR in studies evaluating surveillance for CS-I GCT have ranged from 0% to 5.2%.7-14 However, histologic findings and survival outcomes in this subset of patients have not been reported. In a study by Baniel et al, only 2 of 35 CS-I patients experiencing LR did so on a surveillance protocol.6 Dieckmann et al5 included 23 and Sharp et al4 included 8 such patients in 2 large studies on LR. Specific subgroup analysis of CS-I patients experiencing LR on surveillance was beyond the scope of these studies. However, subgroup analysis of 18 chemotherapy-naïve patients was performed in the Memorial Sloan Kettering Cancer Center study revealing a 95% 5-year overall survival. Given the excellent survival demonstrated by these patients, the authors suggested that a history of prior chemotherapeutic management be a prerequisite to define LR.4 Survival in the present study was far better than what would be expected for most patients experiencing LR. UROLOGY 84 (4), 2014

This is almost certainly explained by the chemotherapynaïve status of all 28 patients. Although it is tempting to assume that these patients had a lower-risk biology given that they presented with stage-I disease and took >24 months to demonstrate disease progression or recurrence, stage at presentation has not previously demonstrated prognostic significance for patients experiencing LR.2,6 Previous studies have reported that given the frequent partial responses and high relapse rates after chemotherapy, most patients experiencing LR require surgical resection to achieve durable NED status.3-5 Additionally, finding fibrosis only at PC-RPLND has been exceedingly rare. Although partial radiographic responses were common in the present study, 10 of 16 of these patients were found to have fibrosis only at PC-RPLND. Viable malignancy and transformation to somatic-type malignancy were over-represented in the postchemotherapy group, particularly for a group of patients who initially presented with CS-I disease. It may be that the tumor growth kinetics of a viable tumor that takes >24 months to manifest recurrence are less vulnerable to exploitation by chemotherapy. We feel that these results still support consideration of primary chemotherapy in the management of LR on surveillance in cases with multifocal recurrence, large or unresectable masses, and/or markedly elevated STM levels. Although previously reported to be present in up to 46% of patients experiencing LR,15 only 2 patients (7.1%) in the present study recurred with yolk sac tumor. There was a corresponding relative decrease in the proportion of patients demonstrating elevated alpha fetoprotein level at LR. It has been proposed that yolk sac tumor may have an increased propensity for resistance to cisplatin-based chemotherapy among GCT histologies.16 This characteristic would explain both its prevalence at LR after chemotherapy and its role in conferring chemoresistance at LR. Its relative rarity in the present study supports this hypothesis, as presumably more chemosensitive GCT histologies had not been eliminated by prior exposure the cisplatin-based chemotherapy. Despite the excellent response to chemotherapy demonstrated in this study, there appears to be a role for the safe and effective utilization of primary RPLND in this population. Ideal patients are those with isolated small-volume RP metastases in the primary landing zone demonstrating normal or minimally elevated STM levels. Of the 9 patients managed in this manner, 8 had NED at last follow-up. Three of these patients, including 1 patient found to have RP PNET, received adjuvant chemotherapy because of viable malignancy at RPLND. Although it is our standard practice to administer directed chemotherapy for all patients found to have PNET at RPLND, we typically observe patients found to have pathologic N1 and N2 GCT, reserving chemotherapy for relapse. In the present study, patients treated earlier in the study group were given adjuvant chemotherapy, given the possibility that these were more UROLOGY 84 (4), 2014

aggressive cancers by virtue of their LR. However, the 5 patients treated more recently were not given adjuvant chemotherapy and none have recurred. Thus, when dealing with small-volume non-PNET viable GCT at RPLND, it is most likely safe to reserve postoperative chemotherapy for patients that subsequently recur. One somewhat troubling finding in this study was the development of somatic-type malignancies in 3 patients (10.7%). Only one of these patients had NED at last follow-up. The incidence of somatic malignancies has been relatively high in previous reports of LR.4,15 Increased incidence has been thought to be due to chemoresistance of these transformed histologies as well as the teratoma that many of these tumors are presumed to arise from. It is unclear why there should be an increased incidence in the previously chemotherapy-naïve patients discussed in the present study. It is possible that this finding is simply a function of analyzing a small group of patients at a tertiary referral center. Indeed, in a report by Baniel et al6on 35 stage-I patients experiencing LR published in 1995, no cases of malignant transformation were described. Alternatively, although somatic malignancy in the setting of GCT is rarely curable and often resistant to chemotherapy alone, most of these histologies are unlikely to demonstrate the rapid clinical progression characteristic of untreated GCT. Thus, it may be that the biology of somatic malignancies leads to LR. The management strategy for patients experiencing LR on surveillance at IU is as follows: Patients with normal or minimally elevated STM levels and unifocal RP relapse that can be completely resected with minimal morbidity are managed with primary RPLND. Those demonstrating viable GCT in their RP specimen are observed, with chemotherapy being reserved for those who relapse. Those with PNET are given directed adjuvant chemotherapy. Patients with markedly elevated STM levels, large volume or multifocal recurrences, and/ or concomitant supradiaphragmatic disease are given cisplatin-based chemotherapy, with PC-RPLND reserved for those demonstrating postchemotherapy residual masses. This report has several weaknesses. This is the largest report of CS-I GCT patients experiencing LR on surveillance. However, with just 28 patients meeting inclusion criteria, this is a small study group to draw definitive management recommendations. Additionally, the vast majority of these patients were referred for RPLND after receiving their primary diagnosis, surveillance follow-up, and diagnosis of LR elsewhere. Thus, the proportion of CS-I GCT patients experiencing LR among those managed on surveillance is unknown. Similarly, the number of patients experiencing LR on surveillance and achieving a clinical complete remission to systemic chemotherapy at referring institutions is unknown. Thus, we were unable to determine the true denominator necessary to calculate incidence of LR on surveillance as well as survival outcomes for the entire population. Additionally, it should be noted that the largely favorable outcomes demonstrated by this group of previously 889

chemotherapy-naïve LR patients should not be applied to LR patients in general.

CONCLUSION LR occurring in CS-I patients managed on surveillance is associated with improved survival outcomes when compared with patients having previously received cisplatinbased chemotherapy and/or RPLND. Disease in these patients behaves similarly to primary metastatic disease as well as disease at early relapse, although there is an increased rate of viable malignancy at PC-RPLND. Thus, it can be successfully managed with primary chemotherapy or primary RPLND depending on the extent of disease and STM status. Furthermore, it is likely that in the absence of PNET, patients with pathologic stage N1 and N2 disease can forego immediate adjuvant chemotherapy without any decrement in survival. Lastly, the wide variation in timing of recurrence highlights the importance of lifelong followup in patients with GCT being managed on surveillance. References 1. Oldenburg J, Martin JM, Fossa SD. Late relapses of germ cell malignancies: incidence, management, and prognosis. J Clin Oncol. 2006;24:5503-5511. 2. Baniel J, Foster RS, Gonin R, et al. Late relapse of testicular cancer. J Clin Oncol. 1995;13:1170-1176. 3. George DW, Foster RS, Hromas RA, et al. Update on late relapse of germ cell tumor: a clinical and molecular analysis. J Clin Oncol. 2003;21:113-122. 4. Sharp DS, Carver BS, Eggener SE, et al. Clinical outcome and predictors of survival in late relapse of germ cell tumor. J Clin Oncol. 2008;26:5524-5529. 5. Dieckmann KP, Albers P, Classen J, et al. Late relapse of testicular germ cell neoplasms: a descriptive analysis of 122 cases. J Urol. 2005; 173:824-829. 6. Baniel J, Foster RS, Einhorn LH, et al. Late relapse of clinical stage I testicular cancer. J Urol. 1995;154:1370-1372. 7. 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 10year experience. J Clin Oncol. 1995;13:1188-1194. 8. Hao D, Seidel J, Brant R, et al. Compliance of clinical stage I nonseminomatous germ cell tumor patients with surveillance. J Urol. 1998;160:768-771. 9. Nicolai N, Pizzocaro G. A surveillance study of clinical stage I nonseminomatous germ cell tumors of the testis: 10-year followup. J Urol. 1995;154:1045-1049. 10. Read G, Stenning SP, Cullen MH, et al. Medical Research Council prospective study of surveillance for stage I testicular teratoma. Medical Research Council Testicular Tumors Working Party. J Clin Oncol. 1992;10:1762-1768. 11. Roeleveld TA, Horenblas S, Meinhardt W, et al. Surveillance can be the standard of care for stage I nonseminomatous testicular tumors and even high risk patients. J Urol. 2001;166:2166-2170. 12. Rorth M, Jacobsen GK, von der Maase H, et al. Surveillance alone versus radiotherapy after orchiectomy for clinical stage I nonseminomatous testicular cancer. Danish Testicular Cancer Study Group. J Clin Oncol. 1991;9:1543-1548. 13. Sogani PC, Perrotti M, Herr HW, et al. Clinical stage I testis cancer: long-term outcome of patients on surveillance. J Urol. 1998; 159:855-858. 14. Tandstad T, Cohn-Cedermark G, Dahl O, et al. Long-term followup after risk-adapted treatment in clinical stage 1 (CS1) nonseminomatous germ-cell testicular cancer (NSGCT) implementing


adjuvant CVB chemotherapy. A SWENOTECA study. Ann Oncol. 2010;21:1858-1863. 15. Michael H, Lucia J, Foster RS, et al. The pathology of late recurrence of testicular germ cell tumors. Am J Surg Pathol. 2000;24:257-273. 16. Logothetis CJ, Samuels ML, Trindade A, et al. The prognostic significance of endodermal sinus tumor histology among patients treated for stage III nonseminomatous germ cell tumors of the testes. Cancer. 1984;53:122-128.

EDITORIAL COMMENT SURVEILLANCE AND LIFELONG VIGILANCE The expected survival of patients with clinical stage (CS) I germ cell tumor (GCT) should exceed 99% regardless of the postorchiectomy management strategy undertaken. In an effort to avoid overtreatment and the associated short-, intermediate-, and long-term morbidity associated with chemotherapy, radiation and/or retroperitoneal lymph node dissection (RPLND), surveillance has emerged as the preferred management option for many patients with clinical stage I seminoma and nonseminoma. The cornerstone of this approach is the assumed curability of chemotherapy-naïve patients, given the welldocumented chemosensitivity of GCTs. In this issue of Urology, Rice et al1 report on the Indiana University experience with late relapse (LR) from CS-I GCT. Although the authors correctly point out that the denominator of CS-I patients on surveillance who suffer LR is unknown, and probably unknowable, 4 GCTrelated deaths (14.3%) among 28 patients in this series are sobering, and reminiscent of Dr. David Swanson’s classic editorial “Low Stage Testis Cancer is Still Potentially Lethal.” It was written in response to the dismal survival of CS-I patients suffering LR after suboptimal primary RPLND with or without chemotherapy, and challenged the concept of therapeutic equivalence of modified templates to the standard bilateral template.2 The present report may even underestimate diseasespecific mortality expected in this setting, considering multiple studies have demonstrated superior outcomes for GCT patients managed at high-volume centers such as Indiana University, as compared with national databases. Furthermore, this study should also serve as a reminder that given the relatively short follow-up time of the few randomized and many nonrandomized studies conducted in CS-I GCT, an unknown number of late events are not (and may never be) accounted for and potentially limit definitive conclusions. Several other findings in the present study deserve mention. Although the median time from orchiectomy to LR was 48.5 months, the authors appropriately highlight the wide range of times to recurrence in this group from 28 months to 321 months (almost 27 years) and emphasize the importance of lifelong follow-up.1 Unfortunately, surveillance schedules, patterns of care, and compliance rates vary widely, particularly beyond 3 years of diagnosis. The most common site of LR is the retroperitoneum,3-6 yet there is no consensus regarding the optimal imaging frequency, and many guidelines recommend against obtaining computed tomography scans, laboratory evaluations, or physician visits after 5 years. Suboptimal compliance in this young highly mobile group of patients further compounds the problem, as does the range of providers called on to direct the surveillance program. Follow-up may be provided by urologists, oncologists, internists, or family physicians whose knowledge of testis cancer, patterns of relapse, and LR may also vary widely.7 UROLOGY 84 (4), 2014

Furthermore, the eventual burden of treatment must be heavily considered when deciding between management strategies for CS-I GCT. The goal of surveillance programs is to eliminate the toxicity of unnecessary adjuvant treatment for patients destined to remain disease-free and to identify relapse at the earliest stage possible for those destined to recur. Earlier detection of LR might potentially reduce the burden of disease and treatment required and improve clinical outcome, which encompasses long-term toxicity in addition to disease-free survival. Ideally, all relapsing patients would meet International Germ Cell Cancer Collaborative Group good-risk criteria, yet a patient in this series required hepatic lobectomy, indicating poorrisk disease at relapse. In an effort to avoid including patients with delayed presentation of early relapse, the authors excluded patients who were completely lost to follow-up until presenting with LR. Nevertheless, many patients still presented with very advanced disease. Seven of 19 patients who underwent postchemotherapy RPLND (PC-RPLND) required adjunctive procedures including nephrectomy in one and resection of the inferior vena cava and aorta in another, suggesting residual bulky retroperitoneal disease. This series appears to lend support to the concept of LR as a distinct clinical entity characterized by chemoresistance when compared with patients with early relapsed GCT or de novo metastatic disease. Four of 19 patients (21%) undergoing PCRPLND harbored viable GCT, which is approximately double the expected incidence in contemporary PC-RPLND series. Furthermore, 3 of 28 patients (11%) in this series were found to have somatic transformation of teratoma, which is also considerably higher than the incidence after primary RPLND (<1%) and PC-RPLND (approximately 3%), suggesting unresected teratoma as a probable source. However, it should also be noted that the chemosensitivity of LR patients in this series was superior to historic outcomes among GCT patients experiencing LR after chemotherapy. Additionally, as pointed out by the authors, in contrast to patients with LR after chemotherapy, human chorionic gonadotropin was the predominant tumor marker, and yolk sac tumor histology was rare in the current population. Thus, in our opinion, patients with LR after surveillance represent a distinct entity with more favorable outcome than postchemotherapy LR patients but less favorable than de novo metastatic disease patients. The authors propose a useful algorithm for management of LR after surveillance. They demonstrate the safety and efficacy of primary RPLND in patients with small-volume retroperitoneal disease and normal (ideally) or minimally elevated alpha fetoprotein levels, with chemotherapy dictated by pathologic findings. This approach allows chemotherapy to be avoided altogether in many of these patients, decreasing long-term morbidity. Conversely, patients who present with bulky or multifocal disease, visceral metastasis, and/or elevated serum tumor markers levels should receive International Germ Cell Cancer Collaborative

UROLOGY 84 (4), 2014

Group risk-appropriate chemotherapy followed by PC-RPLND and complete resection of all residual extra-retroperitoneal mass(es). The 35% rate of viable GCT or secondary somatic malignancy after primary chemotherapy at LR underscores the importance of surgical management being incorporated into the management of these patients. This article is yet another reminder that CS-I GCT is still potentially lethal. Survival rates should approach 100% regardless of the therapeutic strategy chosen at the time of initial diagnosis. However, it is critical that each strategy be executed optimally to assure the excellent survival rates expected with the least toxicity. Surveillance is no exception, and the importance of timely detection of relapse cannot be overstated. This requires an understanding of patterns of relapse, strict compliance with surveillance schedules, and educating our patients and our colleagues.8 Given the serious implications for the rare group of patients who relapse many years after orchiectomy, meticulous lifelong vigilance also remains imperative. Joel Sheinfeld, M.D., Urology Service, Department of Surgery, Memorial Sloan Kettering Cancer Center, New York, NY Darren R. Feldman, M.D., Genitourinary Service, Department of Medicine, Memorial Sloan Kettering Cancer Center, New York, NY

References 1. Rice KR, Beck SDW, Pedrosa JA, et al. Surgical management of late relapse on surveillance in patients presenting with clinical stage I testicular cancer. J Urol. 2014;84:886-891. 2. Swanson DA. Low stage testis cancer is still potentially lethal. J Urol. 1995;154:1376-1377. 3. George DW, Foster RS, Hromas RA, et al. Update on late relapse of germ cell tumor: a clinical and molecular analysis. J Clin Oncol. 2003; 21:113-122. 4. Sharp DS, Carver BS, Eggener SE, et al. Clinical outcome and predictors of survival in late relapse of germ cell tumor. J Clin Oncol. 2008;26:5524-5529. 5. Dieckmann KP, Albers P, Classen J, et al. Late relapse of testicular germ cell neoplasms: a descriptive analysis of 122 cases. J Urol. 2005; 173:824-829. 6. Baniel J, Foster RS, Einhorn LH, et al. Late relapse of clinical stage I testicular cancer. J Urol. 1995;154:1370-1372. 7. Vaughn DJ. Editorial comment on “late relapse of testicular germ cell tumors”. Urol Oncol. 2005;23:1078-1439. 8. Kollmannsberger C, Moore C, Chi KN, et al. Non-risk-adapted surveillance for patients with stage I nonseminomatous testicular germ-cell tumors: diminishing treatment-related morbidity while maintaining efficacy. Ann Oncol. 2010;21:1296-1301. UROLOGY 84: 890e891, 2014. Published by Elsevier Inc.