INTRODUCTION — Testicular germ cell tumors (GCTs) have become one of the most curable solid neoplasms because of remarkable treatment advances that began in the late 1970s.
Prior to the development of effective chemotherapy regimens, the five-year survival rate among men with testicular GCTs was 64 percent [1]. Currently, the five-year survival rate is over 95 percent for both seminomas and nonseminomatous germ cell tumors (NSGCTs) in the United States and Europe.
Chemotherapy in patients with extragonadal GCTs (mainly mediastinal or retroperitoneal tumors) is generally based upon the experience in men with testicular GCTs, and these patients are often included in clinical trials. (See "Extragonadal germ cell tumors involving the mediastinum and retroperitoneum".)
Cisplatin-based combination chemotherapy can cure patients with disseminated GCTs, even in the context of widespread visceral metastases, highly elevated serum tumor markers, and other adverse prognostic features. In contrast to the excellent outcomes for men with good-risk advanced testicular GCTs (over 80 percent relapse-free survival following first-line chemotherapy), up to 60 percent of men who have features of intermediate- or poor-risk disease require additional therapy for relapsed disease following first-line chemotherapy (table 1) [2-5]. (See "Initial risk-stratified treatment for advanced testicular germ cell tumors", section on 'Definition of risk'.)
While most men with metastatic GCTs are cured with first line chemotherapy, the cure rate for GCTs that relapse after first line chemotherapy for metastatic disease is much lower. Men who are diagnosed with relapsed or refractory testicular GCTs should be referred to a cancer center with multidisciplinary expertise, and patients should be offered the opportunity to participate in clinical studies whenever possible. The management of men who progress during chemotherapy or relapse following treatment for GCTs will be reviewed here. The use of chemotherapy to treat patients with advanced chemotherapy-naïve disease is discussed elsewhere. (See "Overview of the treatment of testicular germ cell tumors" and "Initial risk-stratified treatment for advanced testicular germ cell tumors".)
DIAGNOSIS OF RELAPSED DISEASE — The diagnosis of relapsed testicular germ cell tumors (GCTs) is typically made by an increase in serum tumor markers or can be suggested by evidence of disease progression on radiographic studies or physical examination. In situations where the evidence of relapse is atypical, biopsy confirmation may also be useful.
Most cases of relapsed testicular GCTs following chemotherapy are discovered during regular posttreatment surveillance. Such surveillance consists of periodic history and physical examinations, serum tumor marker assessments, and radiographic imaging studies. (See "Posttreatment follow-up for men with testicular germ cell tumors".)
Pattern of spread — The pattern of spread for men diagnosed with early-stage testicular GCTs is very predictable:
●Men with clinical stage I disease generally relapse in the retroperitoneum, and thoracic metastases in the absence of retroperitoneal disease are unusual, particularly for pure seminomas. Biopsy may be indicated when serum tumor markers are normal and the pattern of an apparent relapse does not fit with the natural history of testicular GCTs.
●Men whose prior treatment included chemotherapy often present with more disseminated disease. In multiple studies evaluating second-line treatment, over 60 percent of subjects have retroperitoneal disease, 40 to 50 percent have lung metastases, 26 to 32 percent have mediastinal node involvement, 10 to 20 percent have liver metastases, and 2 to 11 percent have bone involvement [6-8].
Serum tumor markers — An increase in serum tumor markers represents the first evidence of disease in over one-half of patients with nonseminomatous germ cell tumors (NSGCTs) who relapse, including some whose marker values were normal at the time of their initial presentation [6,9].
While elevated serum tumor markers are unusual among men with relapsing chemotherapy-naϊve pure seminomas, those who relapse following chemotherapy usually have elevated levels of serum beta-human chorionic gonadotropin (beta-hCG) and/or lactate dehydrogenase (LDH) when relapse is discovered [10-12]. (See "Serum tumor markers in testicular germ cell tumors", section on 'Monitoring response to therapy'.)
Because numerous diverse medical conditions can cause an elevation in LDH, using LDH to monitor for relapse inevitably results in many false-positive results. Thus, using LDH to monitor for relapse is not recommended [13].
Imaging studies — Although serum tumor markers often provide the first clinical evidence of relapse, a substantial minority of men will have normal serum tumor markers at the time of relapse. In this situation, radiographic imaging, particularly computed tomography (CT) scans of the abdomen and pelvis, generally provides the first evidence of relapse. In men who have undergone a retroperitoneal lymph node dissection or paraaortic radiation therapy (RT), relapse in the retroperitoneum is rare. In such patients, relapse is typically detected either as a result of elevated serum tumor markers and/or evidence of metastatic disease on thoracic imaging with a chest radiograph or CT scan.
Role of biopsy — Clinical judgment is required to balance the potential benefit of additional information from a biopsy versus the risks associated with surgical intervention and with delaying treatment. Biopsy confirmation of relapse is usually not indicated but is important when a nodule or mass appears outside the expected pattern of spread and there is no tumor marker evidence of recurrence.
As an example, for men with pure seminomas, normal beta-hCG, no prior treatment with paraaortic RT, and relapse only in the mediastinal and hilar nodes, a biopsy may be obtained to exclude other diagnoses, such as sarcoidosis. It should be noted, however, that there are case reports of men with simultaneous noncaseating granulomas and relapsed testis cancer [14-18].
Conversely, when the pattern of recurrence (eg, enlarged retroperitoneal lymph nodes on imaging, or rising serum beta-hCG or alpha fetoprotein) and the timing of recurrence (eg, first two years after initial treatment) are consistent with relapsed GCT, then salvage treatment should be initiated because the likelihood of biopsy contributing to the management of disease is very low.
CATEGORIES OF PATIENTS — The optimal treatment for men with relapsed germ cell tumors (GCTs) depends upon the initial treatment and response to prior therapy, the location and timing of the relapse, and tumor histology.
●Chemotherapy-naϊve patients – These men typically presented with stage I GCT, which was managed with active surveillance, radiation therapy (RT), retroperitoneal lymph node dissection, or for stage I seminoma, single-agent carboplatin chemotherapy. These patients typically receive first-line cisplatin-based chemotherapy for disseminated disease if they relapse. (See "Initial risk-stratified treatment for advanced testicular germ cell tumors".)
●Men with clinical stage I nonseminoma previously treated with one or two cycles of BEP – Men who received one or two cycles of bleomycin, etoposide, and cisplatin (BEP) chemotherapy for a clinical stage I nonseminoma rarely relapse, so the optimal treatment is not well defined. We treat these men with first-line chemotherapy using either BEP (table 2) or etoposide, ifosfamide, and cisplatin (VIP) (table 3). The choice of chemotherapy regimen depends on their risk category and whether they previously received one or two cycles of BEP. (See 'Chemotherapy-naïve patients' below.)
If relapse is limited to the retroperitoneum and serum tumor markers are normal, retroperitoneal lymph node dissection is an option. (See "Approach to surgery following chemotherapy for advanced testicular germ cell tumors", section on 'Normalized tumor markers with abnormal imaging findings'.)
●Men who relapse after a full course of first-line cisplatin-based chemotherapy – Patients who relapse after first-line chemotherapy generally require additional chemotherapy (ie, second-line chemotherapy). (See 'Treatment after initial chemotherapy' below.)
●Men with platinum-refractory disease – Men who relapse after second-line chemotherapy and those who progress during or within a month of completing their initial cisplatin-based chemotherapy [19] are considered to have platinum-refractory disease and are often treated with high-dose conditioning therapy with autologous hematopoietic cell transplantation (HCT). (See 'Platinum-refractory disease' below.)
●For men with a diagnosis of a nonseminomatous germ cell tumor (NSGCT), the possibility of metastatic teratoma complicates management. Because there is no way to know what histopathological elements are present within a mass prior to resection, growing masses are typically treated with chemotherapy to eradicate non-teratoma components. This is followed by resection of any residual disease. Good clinical judgment and experience are essential here. The possibility of a growing teratoma must be considered in men with a history of NSGCT who have an enlarging mass and normal serum tumor markers [20]. Teratomas are relatively insensitive to chemotherapy and are usually managed with resection. Growing teratoma should be considered in late relapses and when a mass grows during chemotherapy, especially if serum tumor markers are declining. (See "Approach to surgery following chemotherapy for advanced testicular germ cell tumors", section on 'Rationale for resection of residual masses in patients with NSGCT'.)
In contrast, residual masses in men with pure seminoma are usually observed and only treated if they enlarge. For residual masses 3 cm or larger in diameter, fluorodeoxyglucose (FDG) positron emission tomography (PET) scans are sometimes performed. If there is FDG avidity in the residual masses, then open biopsy is typically performed. We recommend obtaining a biopsy to determine whether a residual viable malignancy is present when FDG-avid masses are present, due to the false-positive rate for PET scans in this setting, which has been reported to be as high as 73 percent [21]. If the biopsy is positive, then the patient is typically treated with additional chemotherapy. (See "Treatment of stage II seminoma", section on 'Posttherapy residual masses' and "Approach to surgery following chemotherapy for advanced testicular germ cell tumors", section on 'Seminoma'.)
PROGNOSTIC FACTORS — Men who relapse after completing chemotherapy can be subdivided according to various prognostic factors, including the response to the initial chemotherapy (primary refractory versus recurrence after a complete response), histology (seminoma versus nonseminomatous germ cell tumor [NSGCT]), and the site of the primary germ cell tumor (GCT; testicular versus extragonadal [retroperitoneal or mediastinal]). The heterogeneity of patients included in various studies based upon these and other factors makes it difficult to compare the relative efficacy of treatment for relapse, which includes standard-dose, ifosfamide-based chemotherapy and high-dose conditioning therapy with hematopoietic cell transplantation (HCT).
Duration of initial remission — The response to initial therapy (complete versus partial versus no response) and the duration of remission are important prognostic indicators. Progression of disease either during or within four weeks after completion of cisplatin-based chemotherapy is considered cisplatin-refractory disease. These patients have a very poor prognosis [22]. (See 'Platinum-refractory disease' below.)
Men with NSGCTs who relapse more than two years after their initial chemotherapy were previously considered to have a poor prognosis, but cure rates of up to 68 percent have been reported in contemporary series [23-25]. Treatment with chemotherapy alone has yielded disappointing results in this setting. Therefore, aggressive surgical resection (either as the primary treatment or following chemotherapy) is strongly recommended because the best outcomes have been reported in men who have undergone resection [23,26,27]. (See 'Late relapse' below and "Approach to surgery following chemotherapy for advanced testicular germ cell tumors".)
Late-relapsing seminomas have not been shown to differ from early-relapsing seminomas and should generally be treated with chemotherapy [25,27,28].
Site of primary tumor — Patients with a relapse of a testicular NSGCT appear to have a better prognosis than those treated for an extragonadal NSGCT. Only 10 percent of men who had a primary mediastinal NSGCT, and 30 percent of men who had a primary retroperitoneal NSGCT and who relapse after first-line chemotherapy experience a durable remission with subsequent chemotherapy [29-31]. Slightly better results have been reported with high-dose conditioning therapy with HCT. A study of 59 patients (37 with retroperitoneal and 22 with mediastinal NSGCTs) reported that 15 (41 percent) of the patients with retroperitoneal primary tumors were alive and continuously disease free at the time of the report, compared with only three (14 percent) of the mediastinal NSGCT patients [32]. (See "Extragonadal germ cell tumors involving the mediastinum and retroperitoneum".)
Histology — Patients with recurrent seminoma have better outcomes than those with NSGCTs [33,34]. Standard-dose regimens used for relapsed or refractory disease can result in long-term disease-free survival in approximately one-half of seminoma patients [10,11] compared with approximately 35 percent for NSGCTs [6]. This is also true for seminomas that relapse beyond two years. (See 'Late relapse' below.)
Among patients treated with high-dose conditioning therapy and HCT for recurrent disease, one center reported a two-year progression-free survival rate of 90 percent for seminoma and 52 percent for NSGCTs [35].
Other factors — Various other factors also have a significant impact on prognosis following treatment for relapsed disease. As an example, in one study, the data from 1594 patients were evaluated to develop a prognostic model in patients who progressed following treatment with first-line platinum-based therapy [36]. This model took into account seven factors:
●Response to initial therapy – Complete or Partial Response (CR/PR) with positive markers, CR/PR with negative markers, or Progressive Disease (PD)
●Alpha fetoprotein (AFP) level – Normal, ≤1000, or >1000
●Beta-human chorionic gonadotropin (beta-hCG) level – ≤1000 or >1000
●Presence of bone, liver, or brain metastases
●Duration of initial remission – ≤3 or >3 months
●Site of primary tumor – Gonadal, retroperitoneal, or mediastinal
●Histology – Pure seminoma versus nonseminoma or mixed tumor
Using all of these factors, three-year overall survival ranged from 65 to 89 percent for the lowest-risk group to 4 to 27 percent for the highest-risk group.
CHEMOTHERAPY-NAÏVE PATIENTS — Men who originally present with stage I germ cell tumors (GCTs) are frequently managed with active surveillance, radiation therapy (RT), or retroperitoneal lymph node dissection (RPLND). Stage I seminoma patients may also be treated with single-agent carboplatin chemotherapy. All of these groups typically receive first-line cisplatin-based chemotherapy for disseminated disease if and when they relapse.
Treatment decisions for patients who relapse after being diagnosed with clinical stage I disease depend to some extent on the extent of disease at relapse. When serum tumor markers are normal, relapse is limited to the retroperitoneum, and no nodes are larger than 2 cm in greatest dimension (stage IIA), then chemotherapy and RPLND are both reasonable options. However, RPLND in this setting requires a surgeon with extensive experience with the operation.
Data supporting this approach include a study from Toronto of 580 men with clinical stage I nonseminomatous germ cell tumor (NSGCT) who were initially managed with surveillance, 162 of whom relapsed [37]. Among the 122 men with stage II disease at relapse, 61 received chemotherapy, and 60 were treated with RPLND. Seventy-five percent of those treated with RPLND had stage IIA disease (nodes <2 cm (table 4 and table 5)), whereas 58 percent of the patients treated with chemotherapy had stage IIB/C disease. Elevated serum alpha fetoprotein (AFP) or beta-human chorionic gonadotropin (beta-hCG) at the time of relapse was associated with an increased risk of needing additional treatment after RPLND; 9 of 17 such patients (53 percent) required additional treatment. In contrast, among patients treated with RPLND who had normal tumor markers, 82 percent required no further treatment. Among patients treated with chemotherapy, 61 percent required no further treatment. Of those needing further treatment after chemotherapy, 73 percent underwent resection of residual masses. Overall survival was 99.2 percent in the entire surveillance cohort and 96.9 percent in those who relapsed.
On the other hand, for patients who have elevated serum AFP or beta-hCG, or stage IIB to III disease at relapse (lymph nodes greater than 2 cm or disease outside the retroperitoneum (table 4 and table 5)), chemotherapy is preferred over RPLND. For chemotherapy-naïve patients, the chemotherapy regimen depends on which risk category the patient is in. Good-risk patients should receive three cycles of bleomycin, etoposide, and cisplatin (BEP) chemotherapy (or four cycles of etoposide plus cisplatin), and intermediate- and poor-risk patients should receive four cycles of BEP (or four cycles of etoposide, ifosfamide, and cisplatin [VIP]).
The optimal treatment in patients with clinical stage I nonseminoma who relapse after one or two cycles of BEP chemotherapy is less clear given limited data and is based upon the clinical judgment of our UpToDate experts. Successful results have been reported treating them as chemotherapy-naïve patients, with additional cycles of BEP (table 2) and also with regimens that include ifosfamide and cisplatin (eg, VIP (table 3)) [38,39]. For good-risk patients who received one cycle of BEP, we prefer to administer three additional cycles of BEP at relapse. For men who previously received two cycles of BEP and for those rare patients who have intermediate- or poor-risk disease at relapse, we prefer four cycles of VIP due to concerns of potential chemotherapy resistance to the original regimen (BEP), and to limit the cumulative dose of bleomycin. Regardless of the chosen chemotherapy regimen, these patients clearly have a better prognosis than those who relapse after a full course of first-line chemotherapy. (See "Initial risk-stratified treatment for advanced testicular germ cell tumors".)
For retroperitoneal relapses that occur either after two years of observation or after primary chemotherapy for clinical stage I nonseminoma, there is an increased possibility of teratoma. RPLND, which is both diagnostic and therapeutic, is a reasonable alternative to chemotherapy as the initial intervention. (See "Approach to surgery following chemotherapy for advanced testicular germ cell tumors".)
TREATMENT AFTER INITIAL CHEMOTHERAPY — Most relapses occurring after initial cisplatin-based chemotherapy are seen within the first two years. For these men, we proceed with chemotherapy (ie, second-line therapy). We do not have a preference for either standard-dose chemotherapy or high-dose conditioning therapy with autologous hematopoietic cell transplantation (HCT). However, these patients should be included in clinical trials whenever possible.
Standard-dose chemotherapy — While cisplatin-based regimens are used in this setting, the outcomes are less successful than when they are used in the first-line or adjuvant setting. The three most common standard-dose regimens combine three drugs and include both ifosfamide and cisplatin:
●TIP – Paclitaxel 175 to 250 mg/m2 infused over 24 hours on day 1 followed by ifosfamide 1500 mg/m2 daily for four days (days 2, 3, 4, and 5) and cisplatin 25 mg/m2 daily for four days (days 2, 3, 4, and 5) (table 6) [7].
●VeIP – Vinblastine 0.11 mg/kg per day for two days, ifosfamide 1200 mg/m2 daily for five days, and cisplatin 20 mg/m2 daily for five days [40].
●VIP – Etoposide 75 mg/m2, ifosfamide 1.2 g/m2, and cisplatin 20 mg/m2 daily for five days every three weeks (table 3) [41]. This regimen is only used for those rare patients whose initial chemotherapy did not include etoposide or for those who received one or two cycles of bleomycin, etoposide, and cisplatin (BEP) for clinical stage I nonseminoma.
For patients who previously were treated with VIP as first-line chemotherapy, optimal second-line chemotherapy is not well defined. Our preferred alternative regimens include high-dose etoposide plus carboplatin followed by HCT, and combinations of gemcitabine, paclitaxel, and either cisplatin or oxaliplatin [42-46].
The data to support the use of chemotherapy come from single-arm trials. Examples include the following large series that evaluated TIP:
●In one study, 46 men with relapsed testicular germ cell tumors (GCTs) were treated with the TIP regimen, with granulocyte colony-stimulating factor support [7]. Eligibility in this trial was limited to men with highly favorable prognostic factors, including:
•No prior history of an extragonadal primary tumor
•No more than six cycles of cisplatin-based chemotherapy prior to enrollment
•History of either a complete response or a partial radiographic response plus normalization of serum tumor markers for more than six months
In this highly selected patient population, at a median follow-up of 69 months:
•Chemotherapy alone resulted in a complete and durable response in 29 patients for an overall response rate (ORR) of 63 percent. Among those who also underwent subsequent resection of residual masses, 32 men (70 percent) achieved a durable response.
•At a mean duration of follow-up of 69 months, only 3 of the 32 complete responders relapsed. The two-year progression-free survival rate was 65 percent.
●In a second series of 51 men (with poor- and good-risk disease) who received a modified TIP regimen, 8 and 18 men had a complete and partial response, respectively (ORR 60 percent) [47]. Unlike the prior series, the failure-free survival rate at one year was only 36 percent. When the analysis was limited to good-risk patients, one-year failure-free survival was 43 percent.
Men with seminoma who relapsed following chemotherapy seem to fare better than those with nonseminomatous GCTs (NSGCTs). A review of 24 patients with recurrent seminoma treated at Indiana University (IU) with VeIP had a 54 percent long-term disease-free survival rate [10], while 15 of 27 men (56 percent) with relapsed seminomas at Memorial Sloan Kettering Cancer Center achieved durable remissions with either standard-dose chemotherapy or high-dose conditioning therapy with HCT [11]. These regimens result in relapse-free survival in approximately 35 percent of men with recurrent NSGCTs after first-line therapy [6,47].
High-dose conditioning therapy and hematopoietic cell transplantation — Randomized trials have not reported improvements in outcomes with high-dose conditioning therapy followed by HCT, but nonrandomized studies have reported more favorable results when compared with historical experience with standard-dose salvage chemotherapy [6,48,49]. Not surprisingly, better results have been reported when HCT was used as second-line, rather than third-line, chemotherapy.
In contemporary studies, treatment-related mortality has been less than 5 percent, and long-term disease-free survival is between 40 and 70 percent [34,48,50-54]. Although autologous HCT is considered by many experts to be an appropriate second-line treatment for testicular and extragonadal NSGCTs, clinical studies have often excluded men with poor prognostic features. Consequently, part of the apparent benefit from autologous HCT compared with standard-dose chemotherapy may be attributable to patient selection. (See "Initial risk-stratified treatment for advanced testicular germ cell tumors", section on 'Intermediate- and poor-risk advanced disease'.)
Interpretation of the data from studies is complicated by the use of different high-dose conditioning (preparative) therapy and heterogeneous patient inclusion criteria. The conflicting data on the role of autologous HCT are illustrated by the following reports:
●No difference in outcome was reported when a single cycle of high-dose conditioning therapy followed by autologous HCT was compared with standard-dose chemotherapy in a European trial (IT-94), in which 280 men who relapsed after first-line chemotherapy were randomly assigned to four cycles of VIP/VeIP (chemotherapy alone), or three cycles of VIP/VeIP followed by one cycle of high-dose conditioning therapy (carboplatin, etoposide, cyclophosphamide; CEC) followed by HCT [6]. Compared with standard-dose chemotherapy alone, high-dose conditioning therapy followed by autologous HCT:
•Was not associated with a significant improvement in the three-year event free survival (EFS) rate (42 versus 35 percent, respectively)
•Resulted in a higher toxic death rate (7 versus 3 percent)
The toxic death rate associated with autologous HCT was unexpected and may indicate that high-dose conditioning therapy should not be preceded by multiple cycles of standard-dose chemotherapy. Moreover, only 73 percent of the patients assigned to the HDT arm actually received high-dose conditioning therapy, another finding that throws doubt on the benefit of giving three cycles of VeIP or VIP prior to autologous HCT.
●The most extensive observational data come from an international, multicenter series of 1594 patients with GCTs who relapsed after treatment with a cisplatin-etoposide regimen, followed by either conventional-dose chemotherapy or high-dose conditioning therapy followed by autologous HCT [48]. There was marked heterogeneity between countries regarding the type and frequency of conditioning regimens used. Patients were classified into five prognostic categories, based upon histology, primary tumor site, response to first-line chemotherapy, progression-free interval, serum tumor markers, and sites of metastatic involvement. At a median follow-up of five years, the two-year progression-free survival rate for the entire series was significantly higher with high-dose conditioning therapy and HCT compared with conventional-dose therapy (50 versus 28 percent, hazard ratio [HR] 0.44), as was the five-year overall survival rate (53 versus 41 percent, HR 0.65). Analyses of progression-free survival and overall survival rates based upon prognostic classification found similar results in all groups.
The differences between these two reports may reflect various biases inherent in an observational study or may reflect limitations in the specific regimen used in the European IT-94 trial, which included only one cycle of high-dose conditioning therapy given after three cycles of conventional-dose chemotherapy.
While the administration of a second cycle of high-dose conditioning therapy prior to HCT may have contributed to improved outcomes in some series, a German randomized trial enrolled 216 men and randomly assigned them to treatment with either [55]:
●Arm A – One cycle of standard-dose VIP chemotherapy followed by three cycles of high-dose conditioning therapy with carboplatin and etoposide (EC) followed by autologous HCT with peripheral blood progenitor cells, or
●Arm B – Three cycles of standard-dose VIP followed by one cycle of high-dose conditioning therapy followed by autologous HCT with peripheral blood progenitor cells
The trial originally was designed to enroll 230 men, but was stopped due to excessive toxicity in Arm B. At one year, there were no statistically significant differences between arm A and arm B in terms of:
●EFS (40 versus 37 percent, respectively)
●Progression-free survival (53 versus 49 percent)
●Overall survival (81 versus 62 percent)
Although the one-year overall survival difference may be clinically meaningful, if not statistically significant, the three-year overall survival rates were indistinguishable at 48 percent for arm A versus 46 percent for arm B. Arm A was associated with significantly fewer treatment-related deaths compared with arm B (4 versus 16 percent, respectively). These deaths were predominantly associated with sepsis or cardiotoxicity.
Additional randomized trials are needed to compare conventional-dose chemotherapy (eg, VeIP or TIP (table 6)) with high-dose conditioning therapy followed by autologous HCT as second-line therapy; one phase III trial (TIGER) is ongoing [56]. Because a single course of high-dose conditioning therapy offered no advantage in the IT-94 trial, the central unanswered question is whether two or more courses of high-dose conditioning therapy followed by autologous HCT is superior to four cycles of standard-dose chemotherapy. At this time, if high-dose conditioning therapy is used, at least two cycles should be administered, EC being the preferred regimen.
Surgery may also have a role in the management of patients who have residual masses after second-line chemotherapy. (See "Approach to surgery following chemotherapy for advanced testicular germ cell tumors".)
LATE RELAPSE — A large majority of relapses occur within the first two years following the completion of first-line chemotherapy for advanced testicular germ cell tumors (GCTs). Relapses that occur after two years are considered late relapses. Late relapses are more common in men who have received chemotherapy for disseminated disease (approximately 3 percent incidence) than in earlier-stage patients (<1 percent incidence) [57].
Estimates of the frequency of late relapse come from a pooled analysis of publications between 1989 and 2006 that included 3700 men with nonseminomatous GCTs (NSGCTs) and 2200 with seminomas, which reported that the incidence of late relapse is 3.2 and 1.4 percent, respectively [57]. In addition, the sites of relapse were as follows:
●Late relapses from NSGCT occur most commonly in the retroperitoneum (50 percent), lungs (17 percent), mediastinum (9 percent), neck and supraclavicular region (6.5 percent), and pelvis (4.4 percent). Involvement of other locations (eg, liver or bones) was seen in 12 percent of cases.
●Late relapses from seminomas tended to be located in the nodes, involving the retroperitoneum (55 percent), mediastinum (27 percent), and neck and supraclavicular region (15 percent). The lungs and pelvis were rarely involved (3.2 and 1.6 percent, respectively). In addition, approximately 5 percent of late seminoma relapses occurred in other locations.
Serum tumor markers are useful in detecting late relapses: approximately 50 percent have elevated alpha fetoprotein (AFP), and approximately 25 percent have elevated beta-human chorionic gonadotropin (beta-hCG) at the time of late relapse.
The relative frequency and clinical outcomes of late relapses in patients with clinical stage I testicular GCTs managed with active surveillance are discussed separately. (See "Active surveillance following orchiectomy for stage I testicular germ cell tumors", section on 'Outcome of active surveillance'.)
Treatment — The approach to patients with a late relapse should be discussed in the context of multidisciplinary care because some of these patients may achieve a durable complete remission with combined modality treatment (eg, chemotherapy followed by surgery). The outcomes following treatment are illustrated by several large series:
●One single institution evaluated the outcomes following treatment for late relapses in 83 men (80 with NSGCTs, 3 with seminomas) who had been relapse-free for at least two years following treatment [27]. Two received no therapy, and 69 of the 81 treated patients were rendered free of disease by treatment, including 43 of the 49 who underwent surgery. With a median follow-up of 25 months, 38 (47 percent) were continuously free of disease following therapy. Of the 32 patients treated with chemotherapy alone, only five remained continuously disease-free.
●In another series of 122 men with late relapse after initial therapy, 42 of 48 seminomas (88 percent) but only 35 of 72 patients with NSGCTs (48 percent) were alive and without evidence of disease at last follow-up [25]. For patients with NSGCTs, the highest cure rate (50 percent) was achieved among those who underwent surgery as a component of their treatment. For seminomas, chemotherapy alone was effective in over one-half of cases.
●In a separate institutional series of 75 patients (70 with NSGCTs), the five-year cancer-specific survival (CSS) was 60 percent [23]. The outcome was better in those who underwent complete surgical resection at the time of late relapse (five-year CSS 79 versus 36 percent in those without complete resection). Patients who had not been treated previously with chemotherapy also had a better outcome (five-year CSS 93 versus 49 percent). In a multivariable analysis, unifocal disease and the absence of symptoms were associated with longer CSS.
Lifetime follow-up is necessary for men who achieve disease-free status after a late relapse because of the high frequency of subsequent relapses. In a follow-up series based upon 37 men with late relapse, 9 of the 19 who achieved no evidence of disease (NED) status relapsed a second time, including five with a second late relapse [58,59].
Based upon these and other observations, the following generalizations can be made about late relapses:
●More than one-half of late relapses occur more than five years after completing chemotherapy [24,27,28].
●Risk factors for late relapse include poor-prognosis disease at initial diagnosis, bulky retroperitoneal disease, and prior early relapse [58].
●Most late relapses are malignant, although 10 to 20 percent are teratomas.
●Between 30 and 50 percent of late relapses are associated with normal serum tumor markers. This has led some to advocate long-term annual radiographic surveillance for all patients with a history of GCTs. It is worth noting that the benefit of long-term surveillance computed tomography (CT) scans has not been demonstrated. A population-based study of 1949 men with GCTs reported late relapses in 25, of whom only 11 were asymptomatic at the time relapse was detected. Of those 11, only one had relapse detected initially on CT scans, while the others were diagnosed on physical exam, serum tumor markers, or chest radiograph. Different studies have reported conflicting results on whether patients who are asymptomatic at the time relapse is detected have a better prognosis compared with symptomatic patients [23-25,27]. Therefore, the use of long-term radiologic surveillance for all patients with GCTs is controversial; some experts believe that surveillance should be tailored to the individual patient's risk. Ideally, the benefit of surveillance should be weighed against the risk of second malignancies from the radiation exposure [24,27,28,57,58,60].
●An aggressive surgical approach should be adopted in patients who have late relapse, especially for men with NSGCT. Surgical resection appears to be crucial to achieving long-term survival. (See "Approach to surgery following chemotherapy for advanced testicular germ cell tumors".)
●Late-relapsing seminomas should be treated in the same manner as early-relapsing seminomas.
BRAIN METASTASES — Central nervous system (CNS) metastases occur in approximately 1 percent of men with disseminated germ cell cancer at the time of diagnosis, and between 0.4 and 4 percent subsequently develop brain metastases [61-64]. Brain metastases almost always occur in the setting of concurrent or prior disseminated disease and are associated with a poor prognosis [65]. Brain metastases are common with choriocarcinoma, and these metastases have a tendency to bleed both spontaneously and during treatment with chemotherapy.
For prognostic purposes, patients with brain metastases can be divided into three groups:
●Group 1 – Those who have brain metastases detected prior to first-line chemotherapy
●Group 2 – Those who relapse only in the brain after achieving a complete response to chemotherapy and, if indicated, subsequent surgery
●Group 3 – Those who develop brain metastases after failing to achieve a complete response to chemotherapy and surgery or who develop both brain and extracranial metastases during or after chemotherapy
Cancer-specific survival (CSS) figures range from 43 to 86 percent for group 1, 39 to 44 percent for group 2, and 2 to 26 percent for group 3 [66-72].
The outcomes and influence of other prognostic factors are illustrated by an analysis of 523 men with brain metastases from a germ cell tumor (GCT) collected retrospectively at 46 centers [65]:
●In the 228 men who had brain metastases present at the time of their initial diagnosis, the three-year overall survival rate was 48 percent. Among the 295 men who had brain metastases at the time of relapse, the three-year overall survival rate was only 27 percent.
●Adverse prognostic factors included the presence of liver or bone metastases, a serum beta-human chorionic gonadotropin (beta-hCG) ≥5000 international units/L, or an alpha fetoprotein (AFP) ≥100 ng/mL.
Treatment — The optimal treatment approach for patients with brain metastases is uncertain, and there are no prospective data that adequately address this issue. Systemic chemotherapy, various RT modalities, and/or surgical excision may all be useful in selected circumstances. (See "Epidemiology, clinical manifestations, and diagnosis of brain metastases".)
●Brain metastases prior to initial chemotherapy – Controlling systemic disease is an important priority for most patients who have brain metastases, and cisplatin and etoposide penetrate the brain in the presence of metastases. Thus, standard-dose systemic chemotherapy (bleomycin, etoposide, and cisplatin [BEP] (table 2), or etoposide, ifosfamide, and cisplatin [VIP] (table 3)) is used initially prior to resection or radiation therapy.
If repeat neuroimaging following first-line chemotherapy documents a complete response, we generally prefer careful observation without further treatment. If there is a limited amount of residual tumor following chemotherapy, surgical excision (preferred) and/or focal RT is used to eradicate the residual disease. Patients with numerous brain metastases whose metastases do not resolve with chemotherapy and whose tumors are not amenable to resection or stereotactic radiosurgery should be evaluated for whole-brain radiation.
●Brain metastases at relapse after initial chemotherapy – For patients with brain metastases at time of relapse after first-line chemotherapy, salvage high-dose or standard-dose chemotherapy followed by resection of residual masses is standard. If residual masses cannot be resected, then these patients should be evaluated for either stereotactic radiosurgery or whole-brain RT. For men with a solitary brain metastasis as the only site of relapse, resection followed by stereotactic radiosurgery and/or two cycles of adjuvant chemotherapy are reasonable options. Alternatively, solitary brain relapses can also be treated with standard-dose or high-dose chemotherapy followed by resection of any residual mass, or stereotactic radiosurgery if the residual mass is unresectable.
When possible, whole-brain RT is avoided because of concerns about the late complications of cranial irradiation. Whole-brain RT may be indicated for patients with more extensive residual disease or those with brain metastases that do not respond to systemic chemotherapy. Although whole-brain RT has been combined with concurrent chemotherapy [72,73], this approach may increase the risk of severe, symptomatic white matter changes [66]. (See "Delayed complications of cranial irradiation", section on 'Neurocognitive effects'.)
Patients with brain metastases need to be treated in a center with tertiary neurosurgical support as well as radiation oncology and medical oncology expertise in treating GCTs with brain metastases. The use of chemotherapy for initial management risks CNS hemorrhage due to the highly vascular nature of choriocarcinomas and the high sensitivity and rapid necrotic response of GCTs to chemotherapy.
PLATINUM-REFRACTORY DISEASE — Men with germ cell tumors (GCTs) in second or subsequent relapse and those who progress during or immediately after their initial platinum-based chemotherapy regimen are considered to have platinum-refractory disease. These patients have a poorer prognosis than those treated with chemotherapy for their initial relapse [34,52,74,75]. However, treatment can still be beneficial in select cases. As with other patients with relapsed or refractory disease, patients with platinum-refractory disease should be referred to a cancer center with expertise in GCTs. (See 'Treatment after initial chemotherapy' above.)
For patients with platinum-refractory disease, high-dose conditioning therapy with autologous hematopoietic cell transplantation (HCT) is our preferred treatment option. Data are limited to single-institution experiences, but show that this approach can result in durable complete responses.
For example, the Indiana University Experience included 364 consecutive patients (122 with platinum-refractory disease) who were treated with two cycles of high-dose conditioning therapy followed by HCT between 2004 and 2014 [35]. The median follow-up for the entire cohort was 3.3 years. In the patients with platinum-refractory disease, the two-year progression-free survival was 33 percent, and the two-year overall survival was 37 percent.
The choice of high-dose conditioning therapy for varies among institutions. In the United States and some European centers, high-dose carboplatin and etoposide (EC) is the most commonly used regimen [3,35,49,76,77]. High-dose regimens that incorporate paclitaxel, cisplatin, cyclophosphamide, and thiotepa have also been used [76,78-82]. There are no randomized trials that compare these more complex regimens with EC.
Novel chemotherapy regimens — Regimens incorporating alternative agents such as gemcitabine, paclitaxel, and oxaliplatin show some promise in these settings [43-46,77,83,84]. Despite the poor prognosis for most patients, some individuals achieve durable complete remissions. Examples of such experimental regimens include the following:
●Gemcitabine plus paclitaxel – This combination was evaluated in an Eastern Cooperative Oncology Group (ECOG) phase II study that included 28 patients with chemotherapy-refractory GCTs [46]. Six patients (21 percent) responded, and two of the three complete responders (7 percent of the total) were continuously disease-free at 15 and 25 months, respectively.
In a partially overlapping retrospective series of 32 patients from Indiana University (IU), four (13 percent) achieved durable complete remission with chemotherapy alone (durations 20+ to 57+ months), and a fifth did so with surgery to resect residual disease (duration 63+ months) [44].
●Gemcitabine plus oxaliplatin – Two phase II trials of gemcitabine plus oxaliplatin enrolling a combined total of 59 men have been published. The overall response rates were 32 and 46 percent, respectively, and 11 to 14 percent were in remission at the time of reporting [45,84].
●Paclitaxel, cisplatin, plus gemcitabine – A phase II study that enrolled 22 patients who had received either at least two prior regimens (n = 20) or prior high-dose conditioning therapy (n = 2) reported that the combination of paclitaxel, cisplatin, and gemcitabine resulted in a five-year overall survival of 18 percent. The median follow-up in this study was over six years [42]. Of note, all survivors had surgery following chemotherapy to resect residual masses.
●Paclitaxel, gemcitabine, plus oxaliplatin – This combination was evaluated in a phase II study, in which 21 of 41 patients (51 percent) had an objective response, including two (5 percent) with a complete response [43]. Additional patients achieved a complete response with post-chemotherapy resection of residual masses, and 17 percent of patients were disease-free. However, the median follow-up was only five months.
Subsequently, a registry study using these three drugs in routine clinical practice in 63 patients reported a 44 percent response rate. Eight patients (12 percent) achieved a complete remission: four with chemotherapy alone and four more with chemotherapy plus resection of residual disease. Thirteen (21 percent) lived longer than two years. Median progression-free and overall survival were 4 and 13.3 months, respectively [85].
●Epirubicin plus cisplatin – This combination was evaluated in a phase II study of 30 men with refractory disease [83]. Nine patients were rendered disease-free, in most instances with surgery or radiation to treat post-chemotherapy residual masses. Seven men remained continuously disease-free 25+ to 48+ months after chemotherapy.
Given the overall poor prognosis of men who require third- or later-line treatment, we recommend participation in clinical trials whenever possible.
Next-generation sequencing and targeted therapy — Testicular cancers that are resistant to or relapse after standard chemotherapy may harbor genomic alterations that are potentially druggable, particularly in the clinical trial setting, and genomic profiling can guide clinical research and disclose therapeutic opportunities for these patients. As examples:
●In a series of seven patients treated with nivolumab or pembrolizumab for refractory GCTs, two long-term responses were observed, both in patients whose tumors were highly positive for programmed cell death ligand 1 (PD-L1) staining [86]. By contrast, subsequent phase II trials demonstrated no objective responses for pembrolizumab among patients with relapsed GCTs unselected for PD-L1 expression [87,88].
●Some tumors harbor a high tumor mutational burden of high levels of microsatellite instability (MSI-H), which may predict benefit from immune checkpoint inhibitor immunotherapy [89]. The biology underlying tumor responsiveness to immunotherapy in this setting is discussed in detail elsewhere. (See "Tissue-agnostic cancer therapy: DNA mismatch repair deficiency, tumor mutational burden, and response to immune checkpoint blockade in solid tumors", section on 'Biology of mismatch repair and tumor mutational burden'.)
Although we generally prefer that patients such as these be enrolled in a clinical trial testing new treatment strategies, in May 2017, the US Food and Drug Administration (FDA) approved pembrolizumab for treatment of a variety of advanced solid tumors, regardless of tumor type, that are MSI-H or have deficient mismatch repair (dMMR), that progressed following prior treatment, and for which there are no satisfactory alternative treatment options.
An important point is that MSI-H or dMMR may indicate the presence of Lynch syndrome, an inherited condition that predisposes to several cancers, particularly colorectal cancer. Given that Lynch syndrome is more prevalent than previously thought, all patients with an MSI-H/dMMR solid tumor should be referred for germline genetic assessment for Lynch syndrome, regardless of family history [90]. (See "Lynch syndrome (hereditary nonpolyposis colorectal cancer): Clinical manifestations and diagnosis", section on 'Microsatellite instability testing'.)
PROGNOSIS AND TOXICITY — Patients who have relapsed after their initial therapy can often be salvaged and remain disease-free with second or subsequent lines of therapy. However, subsequent lines of therapy are associated with significantly increased toxicity in long-term survivors. (See "Treatment-related toxicity in men with testicular germ cell tumors" and "Approach to the care of long-term testicular cancer survivors".)
In a retrospective study of 268 patients from the Danish Testicular Cancer database who relapsed after initial radiation therapy (RT) and/or chemotherapy, 136 patients eventually died of their germ cell tumor, and 132 were salvaged with one or more chemotherapy regimens [91]. For patients requiring second or subsequent salvage therapy, the risks of death from other causes, second cancer, gastric ulcer, renal failure, and neurologic disease were significantly increased compared with patients who were disease-free after treatment with one line of bleomycin, etoposide, and cisplatin (BEP (table 2)) therapy (hazard ratios 1.7, 2.6, 3.6, 4.6, and 3.3, respectively).
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Testicular cancer".)
SUMMARY AND RECOMMENDATIONS
●Cisplatin-based combination chemotherapy can cure patients with disseminated germ cell tumors (GCTs), even in the context of widespread visceral metastases, highly elevated serum tumor markers, and other adverse prognostic features. However, treatment requires multidisciplinary input and in our opinion, these patients should be referred to a cancer center with the prerequisite expertise, and patients should be offered the opportunity to participate in clinical studies whenever possible. (See 'Introduction' above.)
●The diagnosis of relapsed disease is made by an increase in serum tumor markers or evidence of disease progression on radiographic studies or physical examination. (See 'Diagnosis of relapsed disease' above.)
•An increase in serum tumor markers represents the first evidence of disease in over one-half of patients with nonseminomatous germ cell tumors (NSGCTs) who relapse, including some whose marker values were normal at the time of their initial presentation.
•While elevated serum tumor markers are unusual among men with relapsing chemotherapy-naϊve pure seminomas, those who relapse following chemotherapy for stage II or III disease usually have elevated levels of serum beta-human chorionic gonadotropin (beta-hCG) and/or lactate dehydrogenase (LDH) when relapse is discovered.
●Men who are chemotherapy-naïve (or who have only received single-agent carboplatin for stage I seminoma) and relapse after their initial treatment typically receive first-line chemotherapy for disseminated disease if they relapse. Retroperitoneal lymph node dissection may be an alternative for men with no elevated tumor markers and a limited relapse (retroperitoneum only and no nodes >2 cm), particularly when relapse occurs more than two years after the initial diagnosis. (See 'Chemotherapy-naïve patients' above and "Initial risk-stratified treatment for advanced testicular germ cell tumors".)
●Men who received one or two cycles of bleomycin, etoposide, and cisplatin (BEP) chemotherapy for a clinical stage I nonseminoma may be treated with first-line chemotherapy (ie, BEP (table 2), or etoposide, ifosfamide, and cisplatin [VIP] (table 3)). The choice of chemotherapy regimen depends on their risk category and whether they previously received one or two cycles of BEP. If relapse is limited to the retroperitoneum and serum tumor markers are normal, retroperitoneal lymph node dissection is an option. (See 'Chemotherapy-naïve patients' above and "Approach to surgery following chemotherapy for advanced testicular germ cell tumors", section on 'Normalized tumor markers with abnormal imaging findings'.)
●For men who relapse within two years after standard initial (or first-line) chemotherapy, we suggest one of three options: standard-dose cisplatin-based chemotherapy, high-dose conditioning therapy with autologous hematopoietic cell transplantation (HCT), or enrollment in a clinical trial (Grade 2C). In patients with NSGCTs, residual masses should be resected following chemotherapy.
•The primary options for standard-dose chemotherapy are either paclitaxel, ifosfamide, plus cisplatin (TIP (table 6)) or vinblastine, ifosfamide, plus cisplatin (VeIP). However, for patients who have not previously received etoposide, we suggest VIP (table 3) (Grade 2C).
•If choosing HCT, we prefer initial high-dose conditioning therapy with two cycles of high-dose carboplatin and etoposide (EC). (See 'Treatment after initial chemotherapy' above.)
●For men with a late-relapsing NSGCT (ie, after a two-year or longer disease-free interval), we suggest combined modality treatment that incorporates an aggressive surgical approach with systemic therapy rather than systemic therapy alone (Grade 2C).
●For patients with brain metastases who have received no prior chemotherapy, we suggest standard-dose cisplatin-based chemotherapy rather than radiation therapy or surgery as initial therapy (Grade 2C). However, select patients may derive additional benefit from these local modalities, and resection of residual masses, when feasible, is recommended for nonseminomas.
For patients relapsing in the brain after receiving chemotherapy, second-line chemotherapy (either standard dose or high dose) is generally preferred as initial treatment, and multimodality therapy for patients relapsing in the brain is often administered. This typically consists of chemotherapy followed by resection of residual masses. The role of whole brain radiation therapy remains controversial and poorly defined, but it is an option for men who relapse with multiple brain metastases. (See 'Brain metastases' above and "Epidemiology, clinical manifestations, and diagnosis of brain metastases".)
●For previously treated patients with platinum-refractory disease, we suggest high-dose conditioning therapy with HCT (Grade 2C). These patients should be referred to centers with expertise in the treatment of men with platinum-refractory GCT. (See 'Platinum-refractory disease' above.)
●Testicular cancers that are resistant to or relapse after standard chemotherapy may harbor genomic alterations that can potentially be targeted, particularly in the clinical trial setting, and genomic profiling can guide clinical research and disclose therapeutic opportunities for these patients. (See 'Next-generation sequencing and targeted therapy' above.)
ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Philip W Kantoff, MD, who contributed to an earlier version of this topic review.
