Key points

The rate of diagnosis of germ cell tumors (GCT) has remained fairly constant, and in 2015 it is estimated that 8,430 men will be diagnosed with, and 380 deaths will be attributed to, GCTs. By the International Germ Cell Cancer Consensus Classification, roughly 60% of all metastatic GCTs are classified as good risk. The criteria defining good risk are presented in Table 1 . This group of patients has an excellent prognosis, with greater than 90% expectation of cure. Treatment standards have not changed much in recent years. This article focuses on key concepts in the development of the currently accepted first-line regimens and addresses some evolving areas of interest, if not controversy.

Concept 1: cisplatin combination chemotherapy can cure even advanced germ cell patients

Cisplatin, also known as cis -diamminedichloroplatinum, is a platinum alkylating agent that gained considerable attention in testicular cancer because of its significant activity in refractory disease. In 1974, Einhorn and Donohue at Indiana University investigated cisplatin in combination with vinblastine and bleomycin (PVB) with dosing summarized in Table 2 . A total of 50 patients with disseminated GCT were treated with four cycles of PVB followed by 21 months of maintenance vinblastine, resulting in 74% complete remissions and 26% partial remissions. Five patients with partial remissions were also able achieve disease-free status following surgical resection of residual disease, resulting in an overall 85% disease-free status.

Etoposide, a semisynthetic epipodophyllotoxin derivative, was found to induce complete remissions in patients with cisplatin-refractory GCT, leading to study of this active agent in the first-line setting, in combination with cisplatin. From 1981 to 1984 the Southeastern Cancer Study Group and Mid-Atlantic Oncology Program conducted a phase III study to compare PVB with cisplatin, bleomycin, and etoposide (BEP). A total of 261 patients with disseminated GCTs were randomized to four cycles of cisplatin, 20 mg/m ² intravenous (IV) daily Days 1 to 5; bleomycin, 30 U IV on Days 2, 9, and 16; and either IV vinblastine, 0.15 mg/kg on Days 1 and 2 or IV etoposide, 100 mg/m ² on Days 1 to 5. Among those receiving BEP, 83% achieved complete remission, compared with 74% of those receiving PVB. Survival was higher among patients on the etoposide arm ( P = .048). Both regimens showed similar myelosuppression and pulmonary toxicity, but BEP-treated patients experienced significantly less neuromuscular and gastrointestinal toxicity. As such, four cycles of BEP replaced PVB as the new standard.

It should be noted that resection of residual disease is important toward the cure of GCTs, and often requires coordinated multidisciplinary efforts to achieve optimal outcomes. Incomplete resection is associated with increased risk of relapse. Late relapses are often more resistant to chemotherapy and thus the need for surgical resection becomes even more important. Malignant transformation of residual teratoma also may be resistant to standard chemotherapy. Studies have shown that community low-volume centers have higher rates of recurrence following retroperitoneal lymph node dissection. In contrast, tertiary referral centers achieve the lowest recurrence rates, and thus should be used as a resource for these patients when possible.

Concept 1: cisplatin combination chemotherapy can cure even advanced germ cell patients

Cisplatin, also known as cis -diamminedichloroplatinum, is a platinum alkylating agent that gained considerable attention in testicular cancer because of its significant activity in refractory disease. In 1974, Einhorn and Donohue at Indiana University investigated cisplatin in combination with vinblastine and bleomycin (PVB) with dosing summarized in Table 2 . A total of 50 patients with disseminated GCT were treated with four cycles of PVB followed by 21 months of maintenance vinblastine, resulting in 74% complete remissions and 26% partial remissions. Five patients with partial remissions were also able achieve disease-free status following surgical resection of residual disease, resulting in an overall 85% disease-free status.

Etoposide, a semisynthetic epipodophyllotoxin derivative, was found to induce complete remissions in patients with cisplatin-refractory GCT, leading to study of this active agent in the first-line setting, in combination with cisplatin. From 1981 to 1984 the Southeastern Cancer Study Group and Mid-Atlantic Oncology Program conducted a phase III study to compare PVB with cisplatin, bleomycin, and etoposide (BEP). A total of 261 patients with disseminated GCTs were randomized to four cycles of cisplatin, 20 mg/m ² intravenous (IV) daily Days 1 to 5; bleomycin, 30 U IV on Days 2, 9, and 16; and either IV vinblastine, 0.15 mg/kg on Days 1 and 2 or IV etoposide, 100 mg/m ² on Days 1 to 5. Among those receiving BEP, 83% achieved complete remission, compared with 74% of those receiving PVB. Survival was higher among patients on the etoposide arm ( P = .048). Both regimens showed similar myelosuppression and pulmonary toxicity, but BEP-treated patients experienced significantly less neuromuscular and gastrointestinal toxicity. As such, four cycles of BEP replaced PVB as the new standard.

It should be noted that resection of residual disease is important toward the cure of GCTs, and often requires coordinated multidisciplinary efforts to achieve optimal outcomes. Incomplete resection is associated with increased risk of relapse. Late relapses are often more resistant to chemotherapy and thus the need for surgical resection becomes even more important. Malignant transformation of residual teratoma also may be resistant to standard chemotherapy. Studies have shown that community low-volume centers have higher rates of recurrence following retroperitoneal lymph node dissection. In contrast, tertiary referral centers achieve the lowest recurrence rates, and thus should be used as a resource for these patients when possible.

Concept 2: three cycles of combined cisplatin, bleomycin, and etoposide is associated with high cure rates for good-risk nonseminomatous germ cell tumors

In 1989, Einhorn and colleagues sought to decrease BEP toxicity by reducing the number of total cycles. A phase III study of 184 patients with Indiana criteria good-risk disseminated GCTs were randomized to four cycles of BEP or three cycles of BEP. The disease-free rate was similar between arms; 98% for three cycles of BEP compared with 97% for four cycles of BEP. An additional, noninferiority study was conducted by de Wit and colleagues to compare three cycles of BEP versus 4 cycles in International Germ Cell Cancer Consensus Classification good-risk patients. The study also compared a 5-day versus 3-day schedule for BEP, using a 2 × 2 factorial design. Three cycles of BEP and four cycles of BEP showed equivalence (projected 2-year progression-free survival, 90.4% and 89.4%, respectively) and the 5-day and 3-day regimens were also equivalent, with projected 2-year progression-free survival, 88.8% and 89.7%. A subsequent study by the Australian and New Zealand Germ Cell Trial Group compared standard BEP with a reduced-dose BEP. A total of 166 Memorial Sloan Kettering Cancer Center criteria good-risk patients were randomized to receive either three cycles of standard BEP or four cycles of a modified BEP 3-day regimen (cisplatin, 100 mg/m ² IV daily Day 1; bleomycin, 30 U IV on Day 1; and etoposide, 120 mg/m ² on Days 1–3). Standard BEP showed significantly better overall survival compared with the modified BEP (3 vs 13 deaths, respectively; hazard ratio, 0.22; 95% confidence interval [CI], 0.06–0.77; P = .008) highlighting the importance of the full 500 mg/m ² of etoposide and 90 U bleomycin per cycle. These results increased confidence that three cycles of BEP is adequate therapy for good-risk patients.

Controversy 1: is bleomycin necessary?

Bleomycin, a glycopeptide antibiotic originally isolated from Streptomyces verticillus in 1966, has been shown to have significant antitumor activity in solid and hematologic neoplasms. Its antineoplastic activity results from oxygen-free radical formation and in the induction of double-stranded DNA breakage. A study by the Australasian Germ Cell Trial Group comparing PVB with vinblastine and cisplatin alone (PV) found that cancer-related deaths were significantly greater for patients on PV (15% compared with 5%) compared with PVB ( P = .02). Although 34% of patients receiving bleomycin experienced pulmonary toxicity, the authors pointed out that this risk was greatly outweighed by the improved survival noted for patients who received bleomycin.

The bleomycin-free cisplatin and etoposide (EP) regimen was first tested against an older five-drug regimen: bleomycin, vinblastine, cisplatin, cyclophosphamide, and dactinomycin (VAB-6). Among patients on VAB-6, 96% achieved complete remission at 2 years compared with 93% on EP, with a 12% relapse rate for both arms. Because of its equivalent efficacy and lower toxicity, the authors recommended four cycles of EP as an alternate treatment regimen for good-risk GCT. Subsequently, three cycles of BEP was compared with four cycles of EP in a study conducted by the Genito-Urinary Group of the French Federation of Cancer Centers (GETUG). A total of 270 patients with nonseminomatous germ cell tumors (NSGCT) were randomized to receive three cycles of BEP or four cycles of EP. With median follow-up of 53 months the 4-year event-free survival for the 256 patients categorized as good risk was 93% for BEP compared with 86% for EP, with hazard ratio of 0.46 ( P = .052). Overall survival was 96% for men treated with BEP and 92% for those treated with EP (hazard ratio, 0.42; P = .096). Although this did not reach statistical significance, it contributes to the evidence that bleomycin is an important component of curative therapy for patients with GCT.

Given its excellent cure rates, it is our practice to offer three cycles of BEP as first-line therapy for good-risk NSGCT whenever possible. Although four cycles of EP is accepted by National Comprehensive Cancer Network guidelines and others as an equivalent alternative first-line regimen, we usually reserve four cycles of EP for those patients who are otherwise unfit for bleomycin (ie, those patients with underlying lung disease), or who develop toxicity related to bleomycin after starting BEP.

Concept 3: bleomycin pulmonary toxicity is relatively rare and some factors may facilitate selection of patients at higher risk for alternate regimens

Bleomycin pulmonary toxicity has been a driving concern in the search for an alternative to BEP. Pulmonary toxicity presents with a multitude of syndromes, including bronchiolitis obliterans with organizing pneumonia, eosinophilic hypersensitivity, and interstitial pneumonitis with progression to pulmonary fibrosis. Mortality from bleomycin interstitial pneumonitis is estimated at 3%.

Monitoring for bleomycin pulmonary toxicity has been difficult because of a lack of uniform characteristics and defining criteria. Initial studies identified decreased carbon monoxide diffusing capacity as an indicator for subclinical bleomycin toxicity in the absence of other clinical findings. However, a prospective study comparing 27 patients receiving four cycles of BEP with 27 patients receiving four cycles of EP found that carbon monoxide diffusing capacity was significantly decreased with both regimens, whereas vital capacity and pulmonary capillary blood volume were decreased only in patients treated with BEP. This suggests that it may be more appropriate to use vital capacity and pulmonary capillary blood volume as measures for identifying early bleomycin pulmonary toxicity. Additional methods are needed to optimize the early identification of pulmonary toxicity, which could include high-resolution chest computed tomography.

Early studies suggested that there is a dose-related effect between bleomycin and pulmonary toxicity; fatal bleomycin pulmonary toxicity occurred in less than 1% of patients at lower doses, but increased to 10% in the doses greater than 550 U. However, toxicity has been reported with doses less than 100 U. A retrospective analysis of 194 patients receiving bleomycin for GCT identified additional potential risk factors. Based on five deaths (2.8%) it seemed that bleomycin toxicity was influenced by reduced glomerular filtration rate (GFR) ( P <.001) and by older age ( P <.001), with increased toxicity for each decade of life older than 30 years old. These risk factors were confirmed in a larger series of 835 patients with GCT treated at the Royal Mardsen; in addition to GFR and age, cumulative bleomycin dose greater than 300 U predicted risk of pulmonary toxicity. Additional factors associated with pulmonary toxicity include high-dose chemotherapy, pulmonary surgery, and smoking.

Importantly, the suggestion that use of granulocyte colony–stimulating factor may increase the risk of bleomycin pulmonary toxicity has not been confirmed. A large MRC/EORTC trial of 263 patients receiving BEP or bleomycin, vincristine, cisplatin (BOP)/etoposide, ifosfamide, cisplatin (VIP) showed no increased bleomycin pulmonary toxicity in patients receiving granulocyte colony–stimulating factor. A retrospective study also identified no difference in bleomycin pulmonary toxicity between patients with GCT who did and did not receive granulocyte colony–stimulating factor.

Controversy 2: is the toxicity profile of four cycles of bleomycin-free cisplatin and etoposide better than three cycles of combined cisplatin, bleomycin, and etoposide?

Toxicities associated with the increased cumulative dose of cisplatin and etoposide in four cycles of EP must be weighed against the risk of pulmonary toxicity from bleomycin in three cycles of BEP. To help put this in perspective, three cycles of BEP includes bleomycin, 270 U; cisplatin, 300 mg/m ² ; and etoposide, 1500 mg/m ² . Four cycles of EP includes cisplatin, 400 mg/m ² and etoposide, 2000 mg/m ² .

Studies have shown an increased risk of infertility associated with a cumulative cisplatin dose of 400 mg/m ² . Peripheral neuropathy is a significant dose-limiting toxicity that occurs in as many as 80% of patients following cisplatin-based chemotherapy. Reversible cisplatin-induced neuropathy usually begins at cumulative doses greater than 300 mg/m ² , but persistent neuropathy occurs at cumulative doses greater than 400 mg/m ² . The incidence of cardiovascular events, including myocardial infarction, cerebrovascular accident, and arterial thrombosis, occurring during chemotherapy is estimated to be 0.3%.

The aforementioned GETUG trial of three cycles of BEP compared with four cycles of EP provides direct comparative toxicity data. Patients receiving BEP experienced more neurotoxicity ( P <.006) and dermatitis, and Raynaud phenomenon ( P <.0001). Patients receiving EP experienced more grade 3 to 4 neutropenia ( P = .0002), but there was no statistical difference in neutropenic fever compared with BEP ( P = .44). There was no significant difference in pulmonary toxicity (9% for BEP vs 6% for EP; P = .38), and there were no toxic deaths for either arm. Based on the data currently available, neither regimen seems clearly superior in terms of toxicity at this time.

Concept 4: carboplatin is inferior in achieving cure for nonseminomatous germ cell tumors

The MRC/EORTC conducted a trial from 1989 to 1993, in which 598 patients with good-risk GCT were randomized to four cycles of either BEP or BEC (bleomycin, etoposide, and carboplatin). Etoposide was dosed 120 mg/m ² on Days 1 to 3, and bleomycin was dosed 30 U on Day 2. On the BEP arm, cisplatin was dosed either 20 mg/m ² Days 1 to 5 or 50 mg/m ² Days 1 to 2. On the BEC arm, carboplatin was administered at a dose of area under the curve 5 on Day 1 of each 21-day cycle. BEP showed statistically significant improvement in complete response rates: 94.4%, compared with 87.3% with BEC ( P = .009). The 3-year survival rate for BEP was 97% compared with 90% for BEC.

A multi-institutional study conducted in the United States from 1986 to 1990 randomized 270 Memorial Sloan Kettering Cancer Center criteria good-risk NSGCT patients to receive either four cycles of EP or EC (etoposide, carboplatin). Etoposide was dosed 100 mg/m ² on Days 1 to 5. For the EP arm, cisplatin was dosed 20 mg/m ² Days 1 to 5, whereas on the EC arm, carboplatin was dosed at 500 mg/m ² on Day 1 alone. Complete response rates were 90% and 88% for EP and EC, respectively. Relapse rates were 3% for EP compared with 12% for EC. At median follow-up of 22.4 months, event-free and relapse-free survival were inferior for patients receiving EC ( P = .02 and P = .005, respectively).