Management of good risk germ-cell tumours
Ronald de Wit, Erasmus University Medical Center and Daniel den Hoed Cancer Center Rotterdam, the Netherlands.
Following the implementation of cisplatin-containing combined chemotherapy, patients with good-risk metastatic germ-cell cancer have an excellent prognosis. Since the 1980s, bleomycin, etoposide and cisplatin (BEP) have become the standard chemotherapy regimen for these patients. In view of both the high curative potential of BEP chemotherapy and the treatment-related side-effects, trials were carried out in patients with the greatest chance of cure to develop regimens with an improved toxicity profile while maintaining efficacy. Following the results of these trials, the standard chemotherapy in good-risk disease has been reduced from four cycles of BEP (4BEP) to three cycles of BEP (3BEP). Four cycles of etoposide and cisplatin (4EP) is an alternative treatment regimen, with similar efficacy. Studies that explored additional adjustments in the BEP regimen to further decrease toxicity have shown that the lower threshold of efficacy has been reached, and that the efficacy of the chemotherapy is compromised. Especially during the last decade, important long-term side-effects after the treatment of germ-cell cancers have been recognized. Chemotherapy in patients with germ-cell cancer increases the risk of developing cardiovascular disease and second malignant neoplasms. Whether 3BEP or 4EP is the optimal chemotherapy regimen for the future remains to be identified. Possibly differences in acute and late toxicities attributed to chemotherapy might eventually identify the best strategy.
bleomycin, etoposide, carboplatin
bleomycin, etoposide, cisplatin
International Germ Cell Cancer Collaborative Group
cisplatin, vinblastine, bleomycin
second malignant neoplasms
Medical Research Council
European Organization for Research and Treatment of Cancer.
Before the introduction of cisplatin-based chemotherapy, the outcome of the great majority of patients with disseminated germ-cell cancer (GCC) was dismal. In the 1970s, Einhorn and Donohue  reported on a regimen consisting of cisplatin, vinblastine and bleomycin (PVB), yielding impressive results. The durable disease-free survival (DFS) rates obtained with this regimen was a major breakthrough in the treatment of this tumour type. Since then, metastatic germ-cell cancer has become a model for a curable disease.
In 1987, a randomized phase III trial comparing two cisplatin-based therapies was reported by Williams et al.; this study showed that four cycles of the combination of bleomycin, etoposide and cisplatin (4BEP) was superior to four cycles of PVB. Following the report of this study, 4BEP became the standard first-line chemotherapeutic regimen for patients with disseminated GCCs.
Shortly after the introduction of cisplatin-containing chemotherapy it was recognized that patients could be categorized into groups with different prognosis. Variables associated with treatment outcome were identified. Using these prognostic factors clinical trials have focused on attempts either to decrease the toxicity of the standard therapy in patients with a high possibility of cure, or to improve the outcome by intensifying therapy in patients with adverse risk factors .
PROGNOSTIC CLASSIFICATIONS AND THE INTERNATIONAL GERM CELL CANCER COLLABORATIVE GROUP (IGCCCG)
Various prognostic factors have been identified including the primary site (gonadal vs extra-gonadal), histology (seminoma or nonseminoma), extent of disease, localization of metastatic disease and the level of the tumour markers α-fetoprotein, βhCG and lactate dehydrogenase in the serum. A combination of two or more of these factors resulted in several classifications. As large institutions and collaborative groups each used their own criteria of disease extent and serum marker threshold values for determining risk status, comparisons among intergroup study results were hampered. In 1994, the IGCCCG was established to develop one consensus classification. Based on the outcome of nearly 6000 patients with disseminated GCC who had been treated with cisplatin-based combined chemotherapy, the IGCCCG classification was defined, dividing patients into three groups with a good, intermediate or poor prognosis (Table 1). The proportion of patients in the good-prognosis group was 60% for nonseminona and 90% for seminoma . For good-prognosis patients, the overall 5-year survival rate was 91%.
Table 1. The IGCCCG classification*
no non-pulmonary visceral metastases
αFP <1000 µg/L and βhCG <5000 U/L
and LDH <1.5 × ULN
|any primary site|
no non-pulmonary visceral metastases
no non-pulmonary visceral metastases
1000 < αFP < 10 000 µg/L or
5000 < βhCG < 50 000 U/L or
1.5 × N < LDH <10 × ULN
|any primary site|
non-pulmonary visceral metastases
|mediastinal primary or|
non-pulmonary visceral metastases or
αFP >10 000 µg/L or βhCG >50 000 U/L
or LDH > 10 × ULN
In the present review on the management of good-prognosis disease, all studies but one in this particular subgroup were conducted before the introduction of the IGCCCG classification, thus complicating the comparison of the results obtained by using different classifications.
RESULTS OBTAINED BY USING THE STANDARD 4BEP
The BEP regimen consists of bleomycin (30 units) given weekly and both etoposide (100 mg/m2) and cisplatin (20 mg/m2) administered daily for 5 days. Patients received four of these cycles at three-week intervals. Experience with this regimen in the past 20 years has shown a 90% long-term DFS in patients deemed to have a good prognosis [5–7]. The largest prospective data obtained with 4BEP (3BEP and 1EP) in good-prognosis patients by IGCCCG criteria, reported a 2-year progression-free survival (PFS) of 89.4%. These results serve as a reference for interpreting the results reported in studies conducted in the 1980s and 1990s.
SIDE-EFFECTS OF BEP
The first compound in BEP, bleomycin, has long been feared for its induction of pulmonary toxicity (bleomycin-induced pneumonitis, BIP). Reports in the 1980s and 1990s showed clinically significant bleomycin-induced morbidity and mortality rates . Another adverse effect caused by bleomycin is the development of Raynaud’s phenomenon. Both of these side-effects can probably be ascribed to vascular damage. Also, fever and chills, occurring several hours after infusion, and skin toxicity, are well-known side-effects. The second compound in BEP, etoposide, is predominantly characterized by bone-marrow depression leading to febrile neutropenia, bleeding and anaemia. Leukaemia has been ascribed to etoposide, but virtually all cases have occurred after cumulative exposure of >2000 mg/m2. Cisplatin is the most active component of the BEP regimen, but it is also one of the most emetogenic agents known. Other serious side-effects of cisplatin are the induction of nephrotoxicity, myelosuppresion and neurotoxicity. This toxicity manifests as various forms, including neurosensory, but also neuromotor neuropathy, and ototoxicity. Using the standard dose of cisplatin for no more than four cycles, this neuropathy is mostly grade 1 or 2, is mostly reversible, and rarely affects daily life activities in patients.
DECREASING THE TOXICITY OF THE BEP REGIMEN
In view of these treatment-related side-effects and the excellent prognosis of patients with good-risk metastatic GCC, trials carried out in this group aimed to reduce treatment toxicity without compromising efficacy. The studies focused particularly on eliminating bleomycin, reducing the total drug dose by decreasing the number of cycles, and substituting cisplatin by the less toxic component carboplatin.
TRIALS STUDYING THE DELETION OF BLEOMYCIN (EP) FROM TREATMENT
Three studies showed that bleomycin contributes to the effectiveness of the chemotherapy regimen. In 1993 Levi et al. reported a study randomly assigning 218 patients to four cycles of cisplatin and vinblastine with or without bleomycin (PVB vs PV). After ≥4 years of follow-up, the DFS rate was 77% in patients who received PV and 84% in those who received PVB. In a study conducted by Loehrer et al., 178 patients were randomized to receive 3EP (100 mg/m2 on days 1–5) with or without bleomycin, 3BEP; 3BEP was superior to 3EP, as both the failure-free survival (86% vs 69%, P = 0.01) and the overall survival (95% vs 86%, P = 0.01) rates were higher for 3BEP than 3EP. De Wit et al. reported in 1997 on a trial that compared the efficacy of 4EP vs 4BEP in 419 patients with good-prognosis metastatic nonseminomatous testicular cancer. In that study etoposide was given at a total dose of 360 mg/m2 per cycle. The number of complete responses with chemotherapy alone or following surgery after chemotherapy was clearly in favour of the BEP arm (95% vs 87%). After a median follow-up of 7 years, the DFS was 95% in the patients treated with 4BEP vs 90% in those treated with 4EP, whereas overall survival was similar in both groups. In this particular BEP regimen with etoposide at 360 mg/m2 per cycle, even in good-prognosis patients, bleomycin could not be deleted without compromising treatment efficacy, but its use was associated with more toxicity (particularly pulmonary), and efforts to reduce this merit further exploration. In view of the congruent results of these three studies, it can be firmly concluded that bleomycin cannot be omitted without attenuating treatment efficacy from the following regimens: 3BEP, 4PVB or 4BEP with etoposide at 360 mg/m2 per cycle.
TRIALS SUBSTITUTING CISPLATIN BY CARBOPLATIN
As previously mentioned, cisplatin accounts for a substantial part of the toxicity during treatment with BEP. Carboplatin is another platinum analogue with proven activity against GCC. A predominant feature in its toxicity profile is myelosuppression, but carboplatin is less emetogenic and lacks cisplatin-mediated toxicities such as ototoxicity, neurotoxicity and renal deterioration.
Two studies were conducted to investigate whether cisplatin could be substituted by carboplatin. At the Memorial Sloan-Kettering Cancer Center, a trial was conducted that randomized 270 patients to receive 4EP or etoposide and carboplatin (EC, etoposide 500 mg/m2) . The 2-year DFS was 76% in patients treated with EC and 87% treated with EP. In addition, the patients treated with EC were more frequently hospitalized due to neutropenic fever and encountered more thrombocytopenia. The second study was a large collaborative study of the UK Medical Research Council (MRC) and European Organization for Research and Treatment of Cancer (EORTC), where 598 patients were randomly allocated between BEP or BE and carboplatin (BEC) (etoposide 360 mg/m2 in both regimens) . The failure-free survival at 1 year was 91% in the patients allocated to BEP and 77% in those allocated to BEC. Overall survival was significantly greater for the BEP group, with 3-year overall survival rates of 97% and 90% for patients receiving BEP and BEC, respectively.
These data conclusively indicate that carboplatin cannot replace cisplatin for treating patients with good-prognosis metastatic nonseminoma.
TRIALS REDUCING THE NUMBER OF CYCLES
Indiana University was the first to provide evidence that 3BEP might be considered a standard in good-risk patients, instead of 4BEP . To directly compare 4BEP with 3BEP in a true ‘non-inferiority’ study, a trial involving 812 patients, was initiated by the EORTC and the MRC . For patients allocated to four cycles, the final cycle was without bleomycin; etoposide was given at 500 mg/m2. This trial was the first conducted in which patients were entered with good-prognosis disease according to the IGCCCG criteria. The 2-year PFS was 90.4% for 3BEP and 89.4% for 4BEP, showing that these regimens are equivalent in terms of efficacy. Quality of life at 1 year after completing treatment was identical. Frequencies of haematological and non-haematological toxicities were essentially similar, except for both acute and late sensory neuropathy, which were more frequent with 4BEP. This study showed that 3BEP, with etoposide at 500 mg/m2, is sufficient therapy in good-prognosis GCC by IGCCCG criteria.
The study also investigated in a 2 × 2 factorial design if the chemotherapy could be delivered over 3 days instead of the standard 5 days. The 3-day BEP consisted of bleomycin 30 U weekly for 9 weeks, etoposide 165 mg/m2 on days 1–3 (total dose per cycle 495 mg/m2) and cisplatin 50 mg/m2 on days 1 and 2 (total dose per cycle 100 mg/m2). In this way, the total dose administered per cycle was similar for both schedules. The two study arms showed therapeutic equivalence in patients (2-year PFS 88.8% in the 5-day arm and 89.7% in the 3-day arm). Nausea and late ototoxicity occurred more frequently in the 3-day arm in patients receiving 4BEP. Hence, the 3-day regimen can be considered for patients for whom three cycles is sufficient therapy. For those patients who need four cycles, the 5-day regimen is advised.
During the past two decades, investigators at Memorial Sloan Kettering Cancer Center in New York have examined the efficacy and toxicity of 4EP in patients with good-risk metastatic GCC with the standard dose of etoposide (500 mg/m2). A long-term follow-up of 289 patients, reclassified as good-prognosis by IGCCCG criteria and treated with 4EP, has shown a PFS rate of 94% after a median of 7.7 years . These results show that 4EP is effective therapy for treating good-risk metastatic GCC. The results appear to be virtually identical to those obtained in the EORTC/MRC study with 3BEP in IGCCCG good-risk patients . During the past decade these data have served to allow both 3BEP and 4EP to be considered standard treatment options for patients with metastatic good-risk GCC.
DIRECT COMPARISON 3BEP VS 4EP
The only randomized trial to directly compare 3BEP and 4EP was conducted by the French Federation of Cancer Centers Genito-Urinary Group from 1993 to 1999, with a total of 270 patients . The study was originally designed to evaluate therapeutic equivalence, with an expected favourable response of 90% and ≤10% difference in the response proportions between the treatment arms. In view of the curative potential of chemotherapy in testicular cancer, the study should have been designed to exclude a difference of 5% (rather than 10%). Also, the response rate at the completion of chemotherapy does not take into account relapses after chemotherapy and is less suitable to test for therapeutic equivalence. To exclude a difference of 5% in the 2- or 3-year PFS rate would have required 800 patients. In the alternative case that there would be a difference in efficacy between the treatment groups, and the investigators would change their objective by aiming to detect a difference of 10%, a total of 61 events would be required, translating into a total of 408 patients needed. Hence, the 270 patients recruited and 257 evaluable for the efficacy analysis were too few to detect either the superiority of one regimen or non-inferiority. On finding numerical differences in favourable responses and relapses between the treatment arms, the authors changed their primary objective and inadvertently increased the risk of obtaining false-positive results by analysing multiple endpoints . None of the event-free survival and overall survival analyses reached statistical significance, except for a borderline P = 0.052 for the 4-year event-free survival by IGCCCG criteria. In view of the posthoc nature of this analysis, and the multiple endpoint testing, this result cannot be considered conclusive.
In Table 2[8,14,15] the 2-year PFS of several published studies are shown. When the event-free-survival rates of the French study are added to those previously reported results by 3BEP and 4EP (Table 2), there is no indication that there are clinically meaningful differences in the efficacy of the two regimens.
Table 2. The 2-year PFS in good-risk GCC by IGCCCG criteria
||| || |
| 3BEP and 1EP||406||89.4|
||| || |
||| || |
For the time being, 3BEP and 4EP can still be considered standard regimens, which yield excellent long-term efficacy results in IGCCC good risk germ cell cancer.
TOXICITIES IN THE DIRECT COMPARISON OF 3BEP AND 4EP
Thus far, it has remained unclear which chemotherapeutic agent contributes most to toxicity. Many physicians tend to prefer 3BEP over 4EP because they assume that avoiding a fourth (cisplatin) cycle would alleviate some neurological and possibly (late) cardiovascular toxicity, but actually the specific risks have not been assessed. The important contribution of the French study reported by Culine et al. is the comparison of toxicities between the regimens . There were two important differences in toxicity: First, there was an unexpected difference in neurological toxicity in favour of 4EP (all grades; 16% on BEP vs 5% on EP, P = 0.006); second, the incidence of dermatitis, including Raynaud’s phenomenon, was significantly higher in patients treated with 3BEP than in those treated with 4EP (all grades; 29% on BEP vs 8% on EP, P < 0.001). Raynaud’s phenomenon is considered to reflect vascular damage and is associated with late cardiovascular toxicity.
Taken together, adding bleomycin to 3BEP causes greater neurological and vascular toxicity than adding a fourth cycle of EP in the EP regimen. Hence, bleomycin contributes clinically significantly to the acute and potentially late side-effects of the BEP regimen.
CURRENT INSIGHTS IN BIP TOXICITY
Although since the 1980s bleomycin has been feared for its pulmonary toxicity, recent data show that 3BEP, with a cumulative dose of bleomycin of 270 mg, have a minor risk of BIP. The risk of bleomycin-associated death has become <0.2%. Risk factors in good-prognosis patients for bleomycin-induced toxicity are poor renal function, age >40 years, smoking, and the cumulative dose of bleomycin (a total dose of <300 mg rarely causes BIP) [9,17].
Following anecdotal observations in the 1980s of perioperative complications due to BIP, that were ascribed to high inspired-oxygen fractions during surgery, the use of oxygen supplementation was emphatically discouraged after bleomycin treatment. In a more recent study of 77 patients with advanced testicular cancer who had a total of 97 operations to resect residual disease after bleomycin-containing chemotherapy, the importance of intraoperative fractional inspired-oxygen and other potential risk factors was examined . Increased intraoperative fractional inspired-oxygen had no statistical significance in predicting postoperative oxygen problems. The authors concluded that perioperative restrictions on oxygen supplementation in patients treated with bleomycin is not necessary, and that patients should be allowed to receive appropriate inspired-oxygen fractions several months after their last exposure to bleomycin.
Likewise, resuming scuba-diving at 6–12 months after an uncomplicated series of 3 or 4BEP is completely acceptable. Caution should remain for patients who have developed clinical signs of pulmonary function impairment during or shortly after bleomycin treatment .
CURRENT INSIGHTS INTO LONG-TERM SIDE-EFFECTS
Given the long life-expectancy of the young patient after treatment of testicular cancer, the evaluation of long-term complications of the effective therapies has become increasingly important. Several studies showed a higher incidence of second malignant neoplasms (SMN) and cardiovascular diseases (CVD) in patients with GCC than in the general population.
RISK OF SMN
For the risk of SMN the incidence of both leukaemia and solid neoplasms is increased in survivors of GCC. Travis et al. evaluated data from 14 population-based cancer registries in Europe and North-America (1943–2001); >40 000 1-year survivors of GCC were identified and the mean follow-up was 11.3 years. Second solid cancers were diagnosed in nearly 2300 patients (5.6%). Compared to the general population, the risk was about twice as high with chemotherapy or radiotherapy alone, and three times higher with the use of both methods. The absolute risk increased with a longer follow-up and younger age at initial treatment. Similar increased risks of SMNs were reported by Van den Belt et al..
RISK OF CVD
Van den Belt et al. also evaluated the long-term risk of CVD in 5-year survivors of testicular cancer. In 2512 patients who survived for 5 years, who were treated between 1965 and 1995, the incidence of CVD was compared with general population rates. The median follow-up was 18.4 years; 10% of the patients developed coronary heart disease (myocardial infarction/angina pectoris) and 18% developed CVD in general within 20 years after diagnosis. This risk is moderately greater than for the general population. Mediastinal radiotherapy, PVB treatment and recent smoking appeared to be important risk factors for developing CVD. Patients treated with only BEP chemotherapy had a 1.5 times greater risk of overall CVD, but did not have a greater risk of myocardial infarction. Notably, the median follow-up of these patients was relatively short (12.2 years), whereas the median time until myocardial infarction was 14.3 years.
In conclusion, chemotherapy increases the risk of developing major late complication, such as a CVD or SMN. To what extent the individual chemotherapeutic agents contribute to this late toxicity is unclear. As described earlier, bleomycin is associated with Raynaud’s phenomenon and this is related to late cardiovascular toxicity . Future research is warranted to identify which chemotherapeutic agent is the main cause of the late toxicities. These findings imply that physicians should regularly screen survivors of GCC for cardiovascular risk factors. Also, they should advise appropriate risk-reducing strategies to patients, such as treatment of hypertension and hypercholesterolaemia, and lifestyle changes to refrain from smoking, to maintain a healthy body weight and to exercise regularly.
CONCLUSIONS: IS THERE A PREFERRED REGIMEN?
3BEP is considered the standard treatment for patients with metastatic IGCCCG good-risk GCC in Europe; 4EP is an alternative treatment regimen, with identical long-term results. 4EP should be particularly considered for patients at increased risk of bleomycin-induced toxicity (age >40 years, smokers, renal function impairment and pulmonary comorbidities).
Notably, it is clear that with adjustments in the BEP regimen to decrease toxicity, the lowest threshold has been reached, below which the regimen might lose some curative potential. As described previously, cisplatin cannot be replaced by carboplatin, and bleomycin cannot be deleted if a decreased dose of etoposide (360 mg/m2) is being used, or when only three cycles are administered.
To retain optimal efficacy, dose reductions and postponements must be avoided. Hence, when dose reductions or postponements are necessary in patients originally scheduled to receive 3BEP (e.g. because of myelosuppression), a fourth cycle (EP) should be considered. Patients originally scheduled for 3BEP, who develop bleomycin-associated side-effects for which bleomycin administration is postponed, should also continue for a total of four (EP) cycles.
The optimal chemotherapy regimen needs to be identified; it might be the differences in acute and late toxicity caused by the chemotherapy that eventually identifies the optimum regimen .
CONFLICT OF INTEREST