|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2001 American Society for Clinical Oncology Equivalence of Three or Four Cycles of Bleomycin, Etoposide, and Cisplatin Chemotherapy and of a 3- or 5-Day Schedule in Good-Prognosis Germ Cell Cancer: A Randomized Study of the European Organization for Research and Treatment of Cancer Genitourinary Tract Cancer Cooperative Group and the Medical Research CouncilFrom the Rotterdam Cancer Institute and University Hospital, Rotterdam; University Hospital, Nijmegen, the Netherlands; Northern Centre for Cancer Treatment, Newcastle upon Tyne; Christie Hospital, Manchester; Royal South Hants Hospital, Southampton; Medical Research Council Cancer Trials Office, London, United Kingdom; The Norwegian Radium Hospital, Oslo, Norway; and The European Organization for Research and Treatment of Cancer Data Center, Brussels, Belgium. Address reprint requests to Ronald de Wit, MD, PhD, Department of Medical Oncology, Rotterdam Cancer Institute (Dr Daniel den Hoed Kliniek) and University Hospital, PO Box 5201, 3008 AE Rotterdam, the Netherlands; email: wit{at}onch.azr.nl
PURPOSE: To test the equivalence of three versus four cycles of bleomycin, etoposide, and cisplatin (BEP) and of the 5-day schedule versus 3 days per cycle in good-prognosis germ cell cancer. PATIENTS AND METHODS: The study was designed as a 2 x 2 factorial trial. The aim was to rule out a 5% decrease in the 2-year progression-free survival (PFS) rate. The study included the assessment of patient quality of life. A cycle of BEP consisted of etoposide 500 mg/m2, administered at either 100 mg/m2 days 1 through 5 or 165 mg/m2 days 1 through 3, cisplatin 100 mg/m2, administered at either 20 mg/m2 days 1 through 5 or 50 mg/m2 days 1 and 2. Bleomycin 30 mg was administered on days 1, 8, and 15 during cycles 1 through 3. The randomization procedure allowed some investigators to participate only in the comparison of three versus four cycles. RESULTS: From March 1995 until April 1998, 812 patients were randomly assigned to receive three or four cycles: of these, 681 were also randomly assigned to the 5-day or the 3-day schedule. Histology, marker values, and disease extent are well balanced in the treatment arms of the two comparisons. The projected 2-year PFS is 90.4% on three cycles and 89.4% on four cycles. The difference in PFS between three and four cycles is -1.0% (80% confidence limit [CL], -3.8%, +1.8%). Equivalence for three versus four cycles is claimed because both the upper and lower bounds of the 80% CL are less than 5%. In the 5- versus 3-day comparison, the projected 2-year PFS is 88.8% and 89.7%, respectively (difference, -0.9%, (80% CL, -4.1%, +2.2%). Hence, equivalence is claimed in this comparison also. Frequencies of hematologic and nonhematologic toxicities were essentially similar. Quality of life was maintained better in patients receiving three cycles; no differences were detected between 3 and 5 days of treatment. CONCLUSION: We conclude that three cycles of BEP, with etoposide at 500 mg/m2, is sufficient therapy in good-prognosis germ cell cancer and that the administration of the chemotherapy in 3 days has no detrimental effect on the effectiveness of the BEP regimen.
THE TREATMENT of metastatic germ cell cancer with modern cisplatin-based chemotherapy results in the cure of approximately 70% to 80% of patients.1 Factors associated with treatment failure include the extent of metastatic disease and serum levels of the ß-subunit of human chorionic gonadotrophin (HCG),  -fetoprotein (AFP), and lactate dehydrogenase (LDH).2 On the basis of these prognostic factors, clinical trials have focused on attempts either to decrease the toxicity of the standard four cycles of bleomycin, etoposide, and cisplatin (BEP) chemotherapy in patients with a high possibility of cure or to improve the results by intensifying therapy in patients with adverse risk factors. The expected long-term progression-free survival (PFS) rate for good-prognosis patients, according to the International Germ Cell Cancer Collaborative Group Consensus Classification (IGCCC) criteria, is approximately 90%.3 One approach to decrease toxicity in these patients has been to simply delete one cycle of BEP. The South Eastern Cancer Study Group conducted a trial that was designed to show a less than 10% difference in the response rate for three cycles of BEP compared with four cycles of BEP, and 184 patients entered the study.4 At the time of the report, with a median follow-up of 19 months 92% of patients on each arm were disease-free. Hence, it was concluded that administration of three cycles of BEP was sufficient therapy. It has since been questioned whether this trial had provided sufficient evidence to conclude that three cycles of BEP is equivalent to four cycles. Although the sample size was sufficient to detect a 10% difference at a power of 85%, this size did not allow the detection of a clinically meaningful difference of between 5% and 9%. In 1995, the Genitourinary Group of the European Organization for Research and Treatment of Cancer (EORTC) and the Testicular Tumor Working Party of the Medical Research Council (MRC) began a prospective, randomized trial that compared three cycles of BEP with four cycles (three cycles of BEP, and one cycle of EP) in IGCCC good-prognosis germ cell cancer patients. This study was designed to rule out a difference of 5% or more in the PFS rate at 2 years with 90% power. Using a 2 x 2 factorial design, patients were also randomly assigned to receive BEP for the standard 5 days, or the same total dosages of the chemotherapy during 3 days. The rationale for this second randomization was to test the equivalence of the administration of the chemotherapy for 3 days, which is increasingly being used in the United Kingdom. The study included an assessment of patient quality of life.
Patients Patients were eligible for the study if they had metastatic germ cell cancer and all of the following characteristics for good-prognosis germ cell cancer according to the IGCCC classification:3 (1) for nonseminoma, all of the following: (a) a testis or retroperitoneal primary tumor, (b) AFP less than 1,000 U/L, (c) HCG less than 5,000 U/L (1,000 ng/mL), (d) LDH less than 1.5 times the upper limit of normal, and (e) no liver, bone, brain, or other nonpulmonary visceral metastases; and (2) for pure seminoma, (a) either a stage greater than or equal to II C according to Royal Marsden Classification5 or relapse after radiotherapy; (b) no liver, bone, or brain metastases, but any primary site; and (c) HCG or LDH level. Patients were not accepted if they had received prior chemotherapy, had a creatinine clearance less than 40 mL/min. (unless this was caused by obstructive uropathy that could be relieved by nephrostomy), or were older than 65 years of age. All patients gave written informed consent. Patients were stratified at randomization by histology (pure seminoma or nonseminoma) and by hospital. They were then randomly assigned to one of four groups: three cycles of BEP (3BEP) for 3 days, 3 BEP for 5 days, three cycles of BEP and one of EP (3BEP–1EP) for 3 days, or 3BEP–1EP for 5 days. Several centers in the United Kingdom were able to randomize only for the number of cycles and administer the chemotherapy for 3 days as their local standard.
Treatment
Response Assessment After completion of the allocated chemotherapy, patients with normal levels of tumor markers and no clinical or radiologic evidence of residual masses were classified as complete responders and were monitored without further therapy. Patients in whom markers normalized, but who showed evidence of residual tumor mass, underwent debulking surgery, unless the initial histologic diagnosis was pure seminoma. The protocol advised complete macroscopic resection of all tumor remnants. These patients were classified as complete responders if the histologic examination showed no viable cancer cells. If viable malignancy was found and it was considered that it had been resected completely, the patients were classified as having been rendered disease-free by chemotherapy plus surgery. Patients in whom the surgical resection of residual disease was incomplete, those who had continuing elevation of tumor markers, or those who had disease progression while receiving chemotherapy or within 2 months after the completion of chemotherapy were classified as incomplete responders. Rising tumor markers or an increase in tumor volume (unless this was caused by mature teratoma that was completely resectable) was considered progression of disease. Patients with residual masses who did not undergo debulking surgery were classified nonassessable for response and included in disease-free (progression-free) survival and survival analyses. Events in the PFS analysis were incomplete response, progression of disease, and death.
Sample Size and Statistical Analysis
The main end point was the duration of PFS with the duration of survival being the secondary end point. It was estimated that the PFS rate in each of the two control groups (3BEP–1EP, 5 days of administration) would be 92% at 2 years. To show the noninferiority of the experimental arms (3BEP, 3 days BEP), a difference of 5% or more in the PFS rate at 2 years had to be ruled out. When we set The Statistical Analysis Software was used. Analyses were performed on an intent-to-treat basis. That is, patients were analyzed according to their allocated treatment, irrespective of whether or not they had actually received that treatment. All efficacy comparisons of the two chemotherapy regimens (3BEP–1EP versus 3BEP) were stratified by group (EORTC versus MRC) and for the intensification of the treatment (5 days versus 3 days versus not randomized). All efficacy comparisons of the 5-day schedule with the 3-day schedule were stratified by group (EORTC versus MRC) and for chemotherapy regimen (3BEP–1EP versus 3BEP). All patients entered in the trial were used for the purpose of comparing 3BEP–1EP with 3BEP. The patients who were not randomized between the 3-day and the 5-day schedules were excluded from the evaluation of that question. Time-to-event analyses. Time-to-event curves were estimated using the Kaplan-Meier technique. The following procedures were applied when comparing PFS and overall survival. The noninferiority was tested using the one-sided modified test for equivalence by means of the Cox proportional hazards regression model,6 stratified as described earlier for group and number of days of treatment. The significance level for these tests was 10%, according to the trial design. Stratified hazard ratios are presented together with their two-sided 80% confidence intervals (CI), as they provide the same boundaries as one-sided 90% CI. Confidence intervals for a difference in PFS rate were calculated using the standard formula for a difference in binary proportions with the standard errors estimated using Greenwood’s formula.
Response rate and toxicity.
The comparison of the response rate between 3BEP–1EP and 3BEP (and the comparison of 3 days with 5 days) was made using the Mantel-Haenszel statistic for stratified 2 x 2 tables. Homogeneity of the odds ratios over the stratums was tested using a Breslow-Day test for homogeneity. Estimation of the common odds ratio over the pooled table was performed using the Mantel-Haenszel method. Toxicities were compared by means of Follow-up after the completion of treatment. The protocol advised pulmonary vital capacity or carbon dioxide diffusion capacity (DLCO) measurements and audiometry threshold testing at 4,000 Hz before the start of treatment and repeated within 3 months after completion of treatment. All patients were seen monthly during the first year, twice monthly during the second year, three times monthly during the third year, every 6 months in the fourth year, and yearly thereafter. At all visits, serum levels of tumor markers were determined. Chest x-rays were performed every other visit, computed tomographic scans were repeated at 6, 12, 24, and 36 months. Residual masses after chemotherapy for pure seminoma were observed by computed tomographic scans at 3-month intervals during the first year and at 6-month intervals during the second and third years, unless regression occurred. Quality-of-life assessment. The EORTC QLQ-C30, supplemented by a testicular cancer module, was used in this study.7 Patients were required to fill out the forms at the following intervals: (1) at randomization (before chemotherapy), (2) 3 months after randomization (3 or 6 weeks after chemotherapy), and (3) 6 months, 1 year, and 2 years after randomization. To visualize the evolution of the various quality-of-life scores as a function of time in the treatment arms, the means and confidence intervals of the scores (expressed on a 0% to 100% scale) of the patients at each assessment time were computed by treatment arm using a linear mixed model that accounted for the serial correlations between observations and allowed intermittent missing forms. Tests for overall treatment effect and for overall time effect were computer based on a reduced model without an interaction term whenever the interaction effect was not significant on the main model.
Between April 1995 and April 1998, 318 patients were randomly assigned at the EORTC and 494 were randomly assigned at the MRC. Of these 812 patients, 681 were randomly assigned both between the number of cycles and the number of days. Of the other 131, 66 received three cycles for 3 days and 65 received four cycles for 3 days without being randomly assigned for chemotherapy between 3 and 5 days. For the comparison of the number of cycles these 131 patients were grouped with those who had been randomly assigned to the corresponding treatment arms. The median duration of follow-up is 25 months, with a minimum follow-up of 2 years available for 93% of the patients.
Comparison of the Number of Cycles The pretreatment characteristics of all patients are listed in Table 2. The two treatment arms were well balanced with respect to age, histology, and disease extent.
Treatment administered. Three hundred eighty-nine patients (95.8%) on three cycles and 388 patients (95.6%) on four cycles received the allocated number of cycles of chemotherapy (Table 3). Four treatment cycles were administered to patients randomly assigned to receive three cycles for the following reasons: ineligibility (more advanced disease), three patients (1%); dose reductions in previous cycles, two patients (0.5%); and treatment failure (progressive disease), one patient (0.2%). No more than one patient (0.2%) received a fourth cycle because of not yet achieving a radiologically complete response after three cycles. The reasons to withhold a fourth cycle in patients randomly assigned to receive four cycles were toxicity (six patients) and patient refusal (two patients).
The relative dose-intensity of the agents for all cycles is listed in Table 4. The total doses per square meter for all cycles delivered and the relative dose-intensity of cisplatin, bleomycin, and etoposide were not different between the two treatment groups; the median relative dose-intensity of cisplatin for all cycles was 99.7% in four cycles and 99.9% in three cycles. For bleomycin these percentages were 100% on both arms. For etoposide, these percentages were 98.6% in four cycles and 98.6% in three cycles. Cycle postponements occurred in 13% of patients on three cycles and in 17% of patients on four cycles. With few exceptions, in both groups, postponement was performed in no more than one cycle.
Toxicity. Hematologic toxicities (worst toxicity throughout all cycles) are listed in Table 5. The frequency of toxicities was not different between the two treatment arms. The incidence of leukopenic fever throughout the course of the chemotherapy was also similar (P = .448). Acute nonhematologic toxicities and late toxicities are also listed in Table 5. Acute sensory neuropathy (P = .001) was observed more frequent in patients receiving four cycles, and this difference remained during follow-up as late sensory neuropathy (P = .019). Acute pulmonary toxicity was observed more frequent in patients receiving four cycles (P = .047). Pulmonary toxicity was also reported during follow-up in patients who received four cycles, but the numbers were too small to detect actual differences. Of note, the measured changes in pulmonary function that were available for a subgroup of the patients were not significantly different; the median percent decrease of pulmonary vital capacity before and after treatment, available for 98 patients on both arms, was 4.1% on three cycles and 3.6% on four cycles (P = .473). For the DLCO, available for 51 patients on three cycles and 54 patients on four cycles, the median percent decrease was 19.1% and 19.6%, respectively (P = .223).
Audiometric testing revealed no difference in threshold at 4,000 Hz before and after treatment (113 observations on three cycles and 116 on four cycles); median percent change was 0 dB on both arms. Surgery. After chemotherapy, residual masses were found in 200 patients (49%) on three cycles and 195 (48%) on four cycles. Postchemotherapy surgery was performed in 73 patients (34%) on three cycles and 84 patients (43%) on four cycles. The most frequent reason not to perform surgery for residual masses was pure seminoma at entry. Histologic findings were essentially the same for the two treatment groups; overall, viable malignancy was found in 12% of the surgical specimens, mature teratoma in 59%, and necrosis or fibrosis in 29%. Response. Responses to chemotherapy with or without surgery are listed in Table 6. Of 792 eligible patients, 281 of 397 (70.8%) on three cycles achieved a complete response to the chemotherapy. For patients on four cycles, this number was 290 of 395 (73.4%). In addition, nine patients on three cycles (2.3%) and six patients on four cycles (1.5%) had viable malignancy completely resected at surgery. Therefore, 290 patients on three cycles (73.1%) and 296 on four cycles (74.9%) achieved a status of no evidence of disease (NED; P = .411) corresponding to an odds ratio of 1.10 in favor of four cycles (95% CI, 0.80 to 1.52). The odds ratio of responses was independent of the number of days of treatment and the randomization group (Breslow-Day test of heterogeneity, P = .683).
Ninety-eight patients on three cycles (24.7%) and 90 on four cycles (22.8%) had nonassessable disease; most were residual masses in patients with pure seminoma. Another frequent reason was residual lesions judged too small to justify resection by the investigator (nodal disease < 1 cm and pulmonary lesions < 0.5 cm). Both on three cycles and on four cycles, there were nine immediate treatment failures (2.3%): on three cycles, these were seven incomplete responses and two progressions during therapy; on four cycles, there were six incomplete responses and three progressions during therapy. Progression-free survival and survival. On three cycles, of the 281 patients who achieved complete response to chemotherapy, 11 exhibited progression, and of the nine patients rendered NED by chemotherapy plus surgery, two progressed. Of the 98 nonassessable patients, 11 progressed. In addition, there were three failures in ineligible patients (one incomplete response, two relapses). On four cycles, of the 290 patients with complete response to chemotherapy, 11 progressed; of the six patients who achieved NED, four progressed; and of the 90 nonassessable patients, eight progressed. There were no events in the ineligible patients on four cycles (Table 6). Therefore, including the immediate treatment failures, 33 eligible patients (8.3%) and 36 of all patients randomly assigned to receive three cycles (8.9%) eventually had a treatment failure. Of the patients randomly assigned to receive four cycles, eventually 32 (7.9%) had a treatment failure. Of the 12 deaths among patients receiving three cycles, five died of malignant disease, two of infection, one of pulmonary toxicity, and four of non–disease-related reasons. Of the 12 deaths among patients receiving four cycles, three died of cancer, two of infection, one of pulmonary toxicity, and six of non–disease-related reasons, of which several were traffic accidents. The Kaplan-Meier projections of the primary end point, PFS by number of cycles on all patients, are presented in Fig 1. The 2-year PFS rate for eligible patients was 90.7% (95% CI, 87.8% to 93.6%) on three cycles and 89.1% (95% CI, 85.9% to 92.3%) on four cycles. The 2-year PFS of all patients (intent to treat) was 90.4% (95% CI, 87.4% to 93.3%) on three cycles and 89.4% (95% CI, 86.2% to 92.5%) on four cycles.
Therefore, at 2 years the difference in PFS between 3BEP and 3BEP–1EP was -1.0% (intent to treat) with 80% confidence limits (CL) of -3.8% to +1.8%. Hence, noninferiority is claimed as the upper bound of the 80% CL is less than 5%. In addition, because the lower bound of the CL is also greater than -5%, equivalence can be claimed as well. The hazard ratio for PFS, stratified for intensification of treatment and randomization group, was 0.93 (80% CI, 0.71 to 1.24), the P value of the stratified test for noninferiority was .021.
Comparison of the Number of Days Treatment administered. All 681 patients were treated according to the randomization code of 3 days or 5 days. The relative dose-intensity of the agents for all cycles is listed in Table 7. The total doses per square meter for all cycles and the relative dose-intensity of cisplatin, etoposide, and bleomycin were not different between the two treatment groups; the median relative dose-intensity of cisplatin for all cycles was 100% for 3 days and 99.6% for 5 days. For bleomycin the percentages were 100% on both arms. For etoposide, these percentages were 98.5% for 3 days and 98.7% for 5 days. However, looking at the relative dose-intensity of etoposide expressed in categories, it appeared that fewer patients on 5 days of treatment had 91% to 110% of the dose (78% v 83% for 3 days). The percentage of patients receiving less than 90% versus the percentage receiving greater than or equal to 90% of the dose of etoposide reached a borderline significance (P = .088). In accord with this observation, dose reductions of etoposide (at any time throughout the three or four cycles) were more frequent for 5 days, 16.3% v 4.4% for 3 days (P = .001). Most of these reductions were for reasons of hematologic toxicity, ie, values at start of a new cycle, 15.5% for 5 days versus 4.1% for 3 days (P = .001). Likewise, postponements of next chemotherapy cycles (at any time throughout three or four cycles) were significantly more frequent when administering the chemotherapy for 5 days, as compared with for 3 days, 20% v 10%, respectively (P = .001).
Toxicity. Hematologic toxicities (worst toxicity observed) are listed in Table 8. Although etoposide dose reductions and cycle postponements occurred slightly more frequently in patients receiving chemotherapy for 5 days, from the blood values at day 1, the frequencies of grade 3 and 4 nadir values of leukocytes and platelets, and the frequencies of leukocytopenic fever were similar between the two arms. Acute nonhematologic toxicities and late toxicities are listed in Table 8. Nausea was reported more frequently in patients receiving cisplatin 50 mg/m2 days 1 and 2 (3-day regimen) as compared with 20 mg/m2 for 5 days (P = .003). Neurotoxicity was not increased in the condensed regimen. Ototoxicity was reported slightly more frequently for the 3-day regimen, but this reached only borderline statistical significance (P = .056). There was no indication of increased late neurotoxicity. However, late ototoxicity was reported more frequently in patients on the 3-day arm (P = .002).
The median percentage decrease of pulmonary vital capacity before and after treatment for 93 patients on 3 days and 103 patients on 5 days in whom data were available was 2.4% and 4.2%, respectively (P = .98). For DLCO, data were available for 49 patients on 3 days and 56 patients on 5 days; in these patients the median percentage decrease was 18.9% and 19.6%, respectively (P = .85). Although grade 1 and 2 ototoxicity was reported more frequently in the patients receiving the chemotherapy for 3 days, in the subset of patients for whom we had data available of audiometric testing at 4,000 Hz before and after treatment (108 on 3 days and 120 on 5 days), there was no decrease; median percent change was 0 dB on both arms. Response. Responses to chemotherapy are listed in Table 9. Of 662 eligible patients, 242 of 333 (72.7%) receiving the chemotherapy for 3 days achieved a complete response to the chemotherapy. For 5 days, this number was 234 of 329 (71.1%). Of the 62 patients on 3 days who underwent postchemotherapy surgery, viable cancer was found in six, and this was completely resected in five (1.5%). In patients receiving 5 days of chemotherapy, surgery was performed in 75 cases; viable cancer was found in eight, and this was completely resected in six patients (1.8%). Hence, 247 patients on 3 days (74.2%) and 240 on 5 days (72.6%) achieved a status of NED (P = .718), corresponding to an odds ratio of responses of 0.94 (95% CI, 0.66 to 1.33) in favor of 3 days. The odds ratio was not influenced by the number of cycles nor by the randomization group (Breslow-Day test for heterogeneity, P = .432).
There were eight immediate treatment failures (five incomplete responses, three progressions) on 3 days, and eight falures (six incomplete responses, two progressions) on 5 days. In addition, there were three failures in ineligible patients on 5 days. Progression-free survival and survival. On 3 days of administration, of the 242 patients who achieved complete response to chemotherapy, 10 relapsed, and of the five patients rendered NED by chemotherapy plus surgery, two relapsed (Table 9). Of the 78 nonassessable patients, nine relapsed. On the 5-day schedule, of the 234 patients with complete response to chemotherapy, eight relapsed; of the six who achieved NED, three relapsed; and of the 81 nonassessable patients, nine relapsed. In addition, there were three failures in the ineligible patients on 5 days of administration. Hence, including the immediate treatment failures, 29 (8.7%) patients randomly assigned to receive the chemotherapy for 3 days eventually had a treatment failure. Of the patients randomly assigned to 5 days, 28 (8.5%) eligible patients and 31 (9.4%) of all patients eventually had a treatment failure. Of the 11 deaths on the 3-day schedule, five were because of malignant disease, three of treatment-related toxicity, and three of non–disease-related reasons. Of the nine deaths on the 5-day schedule, one was because of cancer, two of treatment-related toxicity, and six of other reasons (including several traffic accidents). Figure 2 displays the Kaplan-Meier projections of PFS by number of days on all patients. The 2-year PFS rate of eligible patients was 89.5% (95% CI, 86.1% to 92.9%) on 3 days of treatment and 89.0% (95% CI, 85.5% to 92.5%) on 5 days. The 2-year PFS of all patients (intent to treat) was 89.0% (95% CI, 86.3% to 93.0%) on 3 days and 88.8% (95% CI, 85.3% to 92.2%) on 5 days.
Therefore, at 2 years the difference in PFS (intent to treat) between 5 days and 3 days was -0.9% with 80% CL of -4.1% to + 2.2%. Hence, also for this comparison, noninferiority is claimed as the upper bound of the 80% CI is less than 5% and equivalence is claimed because, in addition, the lower boundary is greater than -5%. The stratified hazard ratio for PFS was 1.05 (80% CI, 0.78 to 1.41), and the P value for noninferiority was .008.
Quality-of-life analysis.
For 666 patients, both baseline and follow-up information was available, with a total of 2,209 forms. The compliance was 82% at baseline, 54% at 3 months, 65% at 6 months, 60% at 1 year, and 45% at 2 years. The In the comparison of three versus four cycles, there were significant differences in favor of three cycles for physical functioning (P = .008), role functioning (P = .03), cognitive functioning (P = .04), fatigue (P = .003), nausea and vomiting (P = .008), appetite loss (P = .001), and overall quality of life (P = .03). For all scales, the difference was most striking at 3 months, and decreased afterward to no difference at 1 year. Among the testicular tumor–specific questions, it appeared that 3BEP was associated with less pain, numbness, and tingling in hands and feet, less tinnitus, and better sexual functioning. These findings were also most striking at 3 months and gradually disappeared. For the comparison of 3 days versus 5 days, no statistically significant difference was observed between the two regimens in terms of quality of life. However, the analysis suggested a trend in favor of 5 days for the following items: less nausea and vomiting on 5 days at 3 months; less appetite loss at 3 months; less pain, numbness, and tingling of hands and feet at 3 months; and less ringing in the ears at 3, 6, and 12 months. No statistically significant interaction was detected between the number of cycles and the number of days of treatment for any of the quality-of-life items of the QLQ-C30 or for the items of the testicular tumor module, except for ringing in the ears (tinnitus, P = .0193); there was significantly more tinnitus with 3 days as compared with 5 days in patients treated with four cycles (P = .0002). This difference did not appear in patients treated with three cycles (P = .3687).
In the past decade a number of trials have investigated the optimal chemotherapeutic regimen in good-prognosis germ cell cancer. In an attempt to decrease the toxicity of standard BEP therapy, the substitution of carboplatin for cisplatin, the deletion of bleomycin, and the optimal duration of therapy have been investigated. Randomized trials have convincingly shown that carboplatin is inferior to cisplatin in germ cell cancer,8 and that bleomycin cannot be deleted, at least not if therapy is confined to three cycles or when using the European BEP regimen with etoposide doses of 360 mg/m2.9 The most important remaining question was whether or not the Southeastern Cancer Study Group (SECSG) trial in 184 patients had convincingly shown that 3BEP with etoposide scheduled at 500 mg/m2 (hereafter BE500P) was sufficient therapy in good-prognosis germ cell cancer. In a recent long-term follow-up report on 118 patients, there was still no suggestion of a difference,10 but the lack of power to answer the question of whether a clinically meaningful difference in overall or disease-free survival of between 5% and 10% was excluded remained. Although four cycles of etoposide 500 mg/m2 plus cisplatin (hereafter E500P) and three cycles of BE500P have become conventional American treatment options in patients with good-prognosis disease, controversy remained in Europe.11 This was partly because of the lower dose of etoposide to justify the implementation of only three cycles of BE360P, and partly because equivalence had not been established by the SECSG study results. In 1995 the EORTC and MRC began the present large randomized trial that compared three cycles of BE500P versus four cycles (3BE500P–1E500P) that was aimed to rule out a difference of 5% or more in the PFS rate at 2 years with 90% power. Using a 2 x 2 factorial design, patients were also randomly assigned to receive the standard BEP for 5 days, or the same total dosages of the chemotherapy per cycle for 3 days. The purpose of this second randomization was to test the equivalence of the shortened regimen that was increasingly being used in the United Kingdom. Eight hundred twelve patients were randomly assigned to receive 3BEP and 3BEP–1EP, and of these, 681 were also randomly assigned to receive administration of the chemotherapy for 5 days or for 3 days. At the time of the analysis, the median duration of follow-up was 25 months, with a minimum follow-up of 2 years available for 93% of the patients. The 2-year PFS rate (intent to treat) was 90.4% on three cycles and 89.4% on four cycles; a difference of -1.0% (80% CL, -3.8% to +1.8%). Therefore, equivalence is claimed as both the upper and lower bounds of the 80% CL are less than 5%. In the comparison for the numbers of days of administration, the 2-year PFS rate was 89.7% on the 3-day schedule and 88.8% on 5-days, a difference of -0.9% (80% CL, -4.1% to +2.2%). Hence, also for this comparison, equivalence is claimed. Nausea was reported more frequently in patients receiving the condensed regimen with cisplatin 50 mg/m2 on days 1 and 2, as compared with those receiving 20 mg/m2 for 5 days. Neurotoxicity was not increased by the 3-day schedule. Ototoxicity, especially tinnitus, was reported more frequently, but this reached only borderline statistical significance. Late ototoxicity was reported more frequently on the 3-day schedule, but in the subset of patients for whom we had data available of audiometric testing before and after chemotherapy there was no measurable hearing impairment; the median percent decrease at 4,000 Hz was 0 dB on both arms. With respect to quality of life, in the comparison of three versus four cycles, there were significant differences in favor of three cycles for physical functioning, role functioning, cognitive functioning, fatigue, nausea and vomiting, appetite loss, and overall quality of life, as well as less pain, numbness, and tingling in hands and feet; less ringing in the ears; and better sexual functioning. For all scales, the difference was most striking at 3 months, and decreased afterward to no difference at 1 year. For the comparison of 3 days versus 5 days, no statistically significant differences were observed between the two regimens, except for the item ringing in the ears when four cycles were administered, in which case 3 days was worse than 5 days. We thus conclude that equivalence in terms of efficacy is demonstrated between 3BE500P and 3BE500P–1E500P. Obviously, with the establishment of the therapeutic equivalence of 3BE500P, the better quality-of-life measures, and the risk of adding toxicity by a fourth cycle, we believe that 3BE500P should be regarded as the optimal chemotherapy regimen in good-prognosis germ cell cancer. Consequently, as 3BE500P is appropriate therapy, BE360P can be eliminated from the existing regimens, as it would require four cycles, and the use of four cycles of E500P remains a matter of opinion, mainly depending on the degree of concern for bleomycin-induced pulmonary toxicity. Because bleomycin is an essential component in the BEP regimen,9,11 four cycles are clearly warranted if bleomycin cannot be given for reasons of pulmonary toxicity. Also, in case of substantially impaired chemotherapy dose-intensity achieved during the initial cycles of chemotherapy, for whatever reason, the extension of a fourth cycle should be considered. In addition, we conclude that the same total dose of the chemotherapy per cycle can be safely administered for 3 days, without detrimental effect on the effectiveness of the BEP regimen.
APPENDIX
Supported by grant nos. 5U10 CA11488-24 through 2U10 CA11488-29 from the National Cancer Institute, Bethesda, MD.
The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of the National Cancer Institute.
1. Einhorn LH: Treatment of testicular cancer: A new and improved model. J Clin Oncol 8: 1777-1781, 1990[Abstract]
2.
Bajorin DF, Katz A, Chan E, et al: Comparison of criteria for assigning germ cell tumor patients to "good risk" and "poor risk" studies. J Clin Oncol 6: 786-792, 1988
3.
International Germ Cell Cancer Collaborative Group: International Germ Cell Consensus Classification: A prognostic factor-based staging system for metastatic germ cell cancers. J Clin Oncol 15: 594-603, 1997 4. Einhorn LH, Williams SD, Loehrer PJ, et al: Evaluation of optimal duration of chemotherapy in favourable-prognostic disseminated germ cell tumors: Bleomycin and either vinblastine or etoposide. N Engl J Med 316: 1435-1440, 1987[Abstract] 5. Peckham MJ: Investigations and staging: general aspects and staging classification, in Peckham MJ (ed): The Management of Testicular Tumors. London, United Kingdom, Edward Arnold, 1971, pp 89-101 6. Com-Nougue C, Rodary C, Patte C: How to establish equivalence when the data are censored: A randomized trial of treatments for B non-Hodgkin lymphoma. Stat Med 12: 1353-1364, 1993[Medline] 7. Fosså SD, Moyniham C, Serbouti S: Patients’and doctors’perception of long-term morbidity in patients with testicular cancer stage I: A descriptive pilot study. Support Care Cancer 4: 118-128, 1996[Medline]
8.
Horwich A, Sleijfer DT, FossåSD, et al: Randomized trial of bleomycin, etoposide, and cisplatin compared with bleomycin, etoposide, and carboplatin in good-prognosis metastatic nonseminomatous germ cell cancer: A multiinstitutional Medical Research Council/European Organization for Research and Treatment of Cancer Trial. J Clin Oncol 15: 1844-1852, 1997
9.
de Wit R, Stoter G, Kaye SB, et al: Importance of bleomycin in combination chemotherapy for good-prognosis testicular nonseminoma: A randomized study of the European Organization for Research and Treatment of Cancer Genitourinary Tract Cancer Cooperative Group. J Clin Oncol 15: 1837-1843, 1997 10. Saxman SB, Finch D, Gonin R, et al: Long-term follow-up of a phase III study of three versus four cycles of bleomycin, etoposide, and cisplatin in favorable-prognosis germ-cell tumors: The Indian University experience. J Clin Oncol 16: 702-706, 1998[Abstract]
11.
Bajorin DF, Bosl GJ: Bleomycin in germ cell tumor therapy: Not all regimens are created equal. J Clin Oncol 15: 1717-1719, 1997 Submitted July 27, 2000; accepted November 20, 2000.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2001 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
2 tests. 
