DOI: 10.1093/jnci/djq245
Advance Access publication on July 14, 2010.
© The Author 2010. Published by Oxford University Press. All rights reserved.
For Permissions, please e-mail: [email protected].
Article
Comparison of Two Standard Chemotherapy Regimens for
Good-Prognosis Germ Cell Tumors: Updated Analysis of a
Randomized Trial
Peter S. Grimison, Martin R. Stockler, Damien B. Thomson, Ian N. Olver, Vernon J. Harvey, Val J. Gebski, Craig R. Lewis,
John A. Levi, Michael J. Boyer, Howard Gurney, Paul Craft, Amy L. Boland, R. John Simes, Guy C. Toner
Manuscript received April 11, 2009; revised May 17, 2010; accepted June 1, 2010.
Correspondence to: Peter S. Grimison, BSc(Med), MBBS(Hons), MPH, PhD, FRACP, Sydney Cancer Centre, Gloucester House 6, Royal Prince Alfred
Hospital, Missenden Rd, Camperdown, NSW 2050, Australia (e-mail: [email protected]).
Background
The Australian and New Zealand Germ Cell Trials Group conducted a multicenter randomized phase III trial in
men with good-prognosis germ cell tumors of two standard chemotherapy regimens that contained bleomycin,
etoposide, and cisplatin but differed in the scheduling and total dose of cisplatin, the total dose of bleomycin,
and the scheduling and dose intensity of etoposide. The trial was stopped early at a median follow-up of
33 months after a planned interim analysis found a survival benefit for the more dose-intense regimen. The aim
of this analysis was to determine if this survival benefit was maintained with long-term follow-up.
Methods
Between February 1994 and April 2000, 166 men with good-prognosis metastatic germ cell tumors defined by
modified Memorial Sloan-Kettering criteria were randomly assigned to receive 3B90E500P (three cycles, repeated
every 21 days, of 30 kU bleomycin on days 1, 8, and 15; 100 mg/m2 etoposide on days 1–5; and 20 mg/m2
cisplatin on days 1–5; n = 83) or 4B30E360P (four cycles, repeated every 21 days, of 30 kU bleomycin on day 1, 120
mg/m2 etoposide on days 1–3, and 100 mg/m2 cisplatin on day 1; n = 83). Endpoints included overall survival,
progression-free survival, and quality of life and side effects, which were assessed using the Spitzer Quality of
Life Index and the GLQ-8, respectively, before random assignment and during and after treatment. All analyses
were by intention to treat. All P values are two-sided.
Results
The median follow-up was 8.5 years. All but five survivors (3%) were followed up for at least 5 years. Overall
survival remained better in those assigned to 3B90E500P than in those assigned to 4B30E360P (8-year survival: 92%
vs 83%; hazard ratio of death = 0.38, 95% confidence interval = 0.15 to 0.97, P = .037). Progression-free survival
favored 3B90E500P but was not statistically significantly different between the treatment groups (8-year progressionfree survival, 3B90E500P vs 4B30E360P: 86% vs 79%; hazard ratio of progression = 0.6, 95% confidence interval = 0.3 to
1.1, P = .15). At the end of treatment, average scores for most side effect scales favored 3B90E500P. After the completion of treatment, average GLQ-8 scores for numbness (P = .003) and hair loss (P = .04) and the Spitzer
Quality of Life Index (P = .05) favored 3B90E500P.
Conclusion
The survival benefit of 3B90E500P over 4B30E360P was maintained with long-term follow-up.
J Natl Cancer Inst 2010;102:1253–1262
The Australian and New Zealand Germ Cell Trials Group
(ANZGCTG) conducted a prospective randomized phase III trial
comparing two combination chemotherapy regimens consisting of
bleomycin (B), etoposide (E), and cisplatin (P) (ie, BEP) in men
with good-prognosis metastatic germ cell tumors, with accrual
from February 1994 to April 2000. When the trial was designed in
the early 1990s, there was uncertainty about the appropriate dose,
dose intensity, and duration of BEP chemotherapy in this population. Some of the trial group’s participating sites were using regimens that were developed in North America and included three
cycles of BEP with etoposide dosing at 500 mg/m2 per cycle. Other
sites were using regimens that were developed in the United
jnci.oxfordjournals.org Kingdom and Europe, which included four cycles of BEP with
etoposide dosing at 360 mg/m2 per cycle. There was also uncertainty about the importance of bleomycin. An earlier trial by our
group (1) showed an advantage to including bleomycin with cisplatin and vinblastine, but there was continued concern about
bleomycin toxicity. Therefore, the concept of testing a BEP
regimen with a lower dose of bleomycin was attractive.
In the ANZGCTG trial (2), patients were randomly assigned to
receive one of the two BEP regimens, which differed in the
number of cycles, the scheduling and total dose of cisplatin, the
scheduling and dose intensity of etoposide, and the total dose of
bleomycin (Table 1). One regimen—three cycles, repeated every
JNCI
|
Articles 1253
CONTE X T A N D C AVE A T S
Prior knowledge
A multicenter randomized phase III trial in men with good-prognosis
germ cell tumors of two standard bleomycin-, etoposide-, and
cisplatin-containing chemotherapy regimens that differed in the
scheduling and total dose of cisplatin, the total dose of bleomycin,
and the scheduling and dose intensity of etoposide found a
survival benefit for the more dose-intense regimen after being
stopped early at a median follow-up of 33 months after a planned
interim analysis.
Study design
An updated analysis of data from this randomized trial after a
median of 8.5 years of follow-up that examined overall and
progression-free survival and health-related quality of life.
Contribution
Overall survival remained better in those assigned to the more
dose-intense regimen. Progression-free survival also favored the
more dose-intense regimen but was not statistically significantly
different between the treatment groups. The main difference in
health-related quality of life between regimens was in feelings of
numbness or pins and needles after the completion of treatments,
which favored the more dose-intense regimen.
Implications
The dose intensity of etoposide and the total dose of bleomycin are
critical in determining the outcome of therapy.
Limitations
The effects of treatment on survival may have been overestimated
because of early stoppage of the trial after a planned interim
analysis. The treatment groups were not balanced with regard to
International Germ Cell Consensus Classification prognosis criteria, which were not available when the trial was designed.
From the Editors
21 days, of 30 kU bleomycin on days 1, 8, and 15; 100 mg/m2
etoposide on days 1–5; and 20 mg/m2 cisplatin on days 1–5
(3B90E500P)—was based on a regimen that was used at Indiana
University since the mid-1980s (3), except that in the trial, each
dose of bleomycin was given 1 day earlier. The other regimen—
four cycles, repeated every 21 days, of 30 kU bleomycin on day 1,
120 mg/m2 etoposide on days 1–3, and 100 mg/m2 cisplatin on day
1 (4B30E360P)—was based on the control arm in a Medical Research
Council and European Organization for Research and Treatment
of Cancer (MRC/EORTC) trial that was under way when the
ANZGCTG trial was being designed; the MRC/EORTC trial
tested the substitution of carboplatin for cisplatin (4), and bleomycin was given on day 2 and cisplatin was given for a period of 2 or
5 days. The results of the MRC/EORTC trial showed an overall
survival rate of 97% at 3 years with the BEP control arm vs 90%
in the experimental arm, supporting the use of the BEP control
arm as a comparator in the ANZGCTG trial.
Patients in the ANZGCTG trial were classified as good prognosis according to modified Memorial Sloan-Kettering Cancer
Center (MSKCC) criteria (5), which was the best prognostic
classification available at the time for predicting complete response
to initial treatment (6). The modification of this classification was
based on subsequently published research from MSKCC (7) and
allowed patients with retroperitoneal primary nonseminoma to be
potentially eligible. When the International Germ Cell Consensus
Classification (IGCCC) for prognosis became available 3 years
after this trial began (8), the trial was modified to exclude patients
with poor-prognosis disease according to IGCCC criteria.
Of a planned 260 patients, 166 were recruited to the trial and
randomly assigned to receive 3B90E500P or 4B30E360P (83 patients
per arm). The trial was stopped early after a second planned interim analysis at a median follow-up of 33 months showed a substantial survival benefit for 3B90E500P compared with 4B30E360P
(three vs 15 deaths; overall survival at 3 years: 90% vs 81%; hazard
ratio [HR] of death = 0.22, 95% confidence interval
[CI] = 0.06 to 0.77, P = .008) (2). The aim of this updated analysis
was to determine if the survival benefit was maintained with longterm follow-up. We also report for the first time on patterns of
relapse, treatment received after relapse, response to subsequent
treatment, and effects on health-related quality of life (HRQL).
Table 1. Comparison of chemotherapy treatment regimens*
Treatment regimen
Overall characteristics
Number of cycles
Cycle duration
Duration of treatment
Cisplatin (P)
Dose and schedule per cycle
Planned dose intensity
Planned total dose
Etoposide (E)
Dose and schedule per cycle
Planned dose intensity
Planned total dose
Bleomycin (B)
Dose and schedule per cycle
Planned dose intensity
Planned total dose
3B90E500P
4B30E360P
3
21 d
9 wk
4
21 d
12 wk
20 mg/m2 on days 1–5
100 mg/m2 per 3 wk
300 mg/m2
100 mg/m2 on day 1
100 mg/m2 per 3 wk
400 mg/m2
100 mg/m2 on days 1–5
500 mg/m2 per 3 wk
1500 mg/m2
120 mg/m2 on days 1–3
360 mg/m2 per 3 wk
1440 mg/m2
30 000 units on days 1, 8, 15
90 000 units per 3 wk
270 000 units
30 000 units on day 1
30 000 units per 3 wk
120 000 units
* Reprinted with permission from Elsevier from Toner, GC, Stockler MR, Boyer MJ, Jones M, Thomson DB, Harvey VJ, Olver IN, Dhillon H, McMullen A, Gebski
CJ, Levi JA, Simes RI. Comparison of two standard chemotherapy regimens for good prognosis germ-cell tumours: a randomised trial. Lancet. 2001:739–745.
B = bleomycin; E = etoposide; P = cisplatin.
1254 Articles
|
JNCI
Vol. 102, Issue 16 | August 18, 2010
Methods
Trial Design
This trial is registered in the Australian New Zealand Clinical
Trials Registry (Identifier ACTRN12605000142639). Full details
of the trial design have been described previously (2). In brief,
eligible patients had histologically confirmed germ cell tumors
with measurable disease or elevated serum tumor markers and
were classified as having good-prognosis disease according to
modified MSKCC criteria as described above. Patients with a nonseminoma arising in the mediastinum, with brain metastases, or
with a history of another cancer were excluded. Patients with a
histological diagnosis of seminoma who had an elevated serum
level of alpha-fetoprotein were classified as having nonseminoma.
The trial protocol was approved by the institutional review board
at each participating center, and written informed consent was
obtained from all patients in accordance with national guidelines.
Patients were stratified by institution and primary histology (seminoma vs nonseminoma) and randomly assigned by the coordinating center to 3B90E500P or 4B30E360P. Recommendations in the
protocol for other therapy, including post-chemotherapy surgery
and salvage chemotherapy, were the same for both regimens.
Patients were assessed after each cycle of chemotherapy, at the
completion of study treatment, every 6 months until 5 years after
random assignment, and then annually thereafter. Information
about patients who did not attend follow-up assessments was
obtained by telephone and by searching cancer registries where
possible.
HRQL was assessed by means of questionnaires that were completed by the patients and clinicians before random assignment
(ie, at baseline); at weeks 3, 6, and 9 after random assignment (ie,
during treatment); at week 12 after random assignment (ie, at the
completion of treatment); and at months 6, 9, and 12 after random
assignment (ie, after the completion of treatment). All patients
completed the GLQ-8 (9), an eight-item questionnaire that
measures the following side effects of chemotherapy on linear
analogue self-assessment scales: “feeling anxious or depressed,”
“feeling sick (nausea or vomiting),” “numbness or pins and
needles,” “loss of hair,” “tiredness,” “appetite or sense of taste,”
“sexual interest or ability,” and “thought of actually having treatment.” Potential scores range from 0 (“None”) to 100 (“Worst
I can imagine”). The last 96 patients to be recruited also completed
the Utility-Based Questionnaire-Cancer (10), a 30-item questionnaire that measures global health on two scales: “general health” (one
item with four response categories ranging from 4 [“Excellent”]
to 1 [“Poor”]) and “health status” (a thermometer with scores
ranging from 100 [“Full health”] to 0 [“Death”]), and specific
aspects of HRQL that are grouped within a physical function subscale (sum of three items with scores ranging from 4 [least impairment] to 18 [worst impairment]), a social activities subscale (sum of
five items with scores ranging from 5 [least impairment] to 20
[worst impairment]), a self-care subscale (sum of four items with
scores ranging from 4 [least impairment] to 12 [worst impairment]), and a distresses subscale (sum of 20 items about distresses
because of physical and psychological symptoms relevant to cancer
and its treatment with scores ranging from 0 [least impairment] to
200 [worst impairment]). Clinicians completed the Spitzer Quality
jnci.oxfordjournals.org of Life Index (11,12), a five-item questionnaire that measures
a patient’s activity, daily living, health, support, and outlook.
Potential scores are the sum of the five items and range from 5
(least impairment) to 15 (worst impairment).
Statistical Analysis
Full details of statistical analysis have been previously described (2).
Endpoints reported here include overall survival, progression-free
interval, and HRQL. Other trial endpoints [all previously reported
(2)] were response proportions, failure-free survival, and toxicity. All
survival times were measured from the date of random assignment.
Overall survival was defined as the time to death from any cause.
Progression-free interval was defined as the time to progression for
refractory patients, to relapse for initially responding patients, or to
death from any cause without progression or relapse.
All efficacy analyses were performed on the basis of intention to
treat. Sensitivity analyses of overall survival and progression-free
interval were performed. First, analyses were restricted to patients
who were classified as having good prognosis according to IGCCC
criteria. Second, analyses included all patients with adjustment for
IGCCC prognostic group. Survival distributions were displayed by
Kaplan–Meier curves and compared by using the log-rank test. A
Cox proportional hazards model was used to estimate hazard ratios
and 95% confidence intervals and for multivariable analysis of survival that adjusted for IGCCC prognostic group. All data conformed
to proportional hazards assumptions as confirmed by using the proportional hazards test on each individual variable and globally.
HRQL ratings between treatment groups at baseline were compared using t tests. Generalized estimating equations were used to
compare HRQL ratings between treatment groups during treatment (at weeks 3, 6, and 9 after random assignment) and after treatment (at months 6, 9, and 12 after random assignment). All P values
are two-sided and were not adjusted for multiple comparisons.
Statistical significance was defined as P less than or equal to .05.
Results
Patients
The trial profile is shown in Figure 1. A total of 83 patients were
randomly assigned to each chemotherapy regimen (Figure 1). All but
five survivors (3%) were followed up for at least 5 years. The median
follow-up was 8.5 years (range = 2 months to 12.9 years). There were
potentially important protocol violations for three patients. One
patient with a mediastinal primary tumor who was allocated to
4B30E360P experienced disease progression and died before completing
treatment. Subsequent pathology review following commencement of
protocol treatment indicated that this patient had an anaplastic carcinoma rather than a seminoma as was originally diagnosed. One
patient in each treatment group inadvertently received the opposite
treatment: A patient who was assigned to receive 4B30E360P but who
actually received 3B90E500P subsequently died of progressive disease,
whereas a patient who was assigned to receive 3B90E500P but who actually received 4B30E360P remained alive and disease free. Nine
patients—six assigned to receive 3B90E500P and three assigned to
receive 4B30E360P—had a nonseminoma and a residual mass but did
not undergo resection of the residual mass, contrary to recommendations in the clinical trial protocol.
JNCI
|
Articles 1255
166 randomly assigned
Figure 1. Trial profile for patients allocated to
chemotherapy treatment regimens containing
bleomycin, etoposide, and cisplatin (BEP) at
differing doses, dose intensities, and schedules.
3B90E500P = three cycles, repeated every 21 days, of
30 kU bleomycin on days 1, 8, and 15; 100 mg/m2
etoposide on days 1–5; and 20 mg/m2 cisplatin
on days 1–5; 4B30E360P = four cycles, repeated
every 21 days, of 30 kU bleomycin on day 1,
120 mg/m2 etoposide on days 1–3, and 100 mg/m2
cisplatin on day 1.
83 assigned to 3B90E500P
83 assigned to 4B30E360P
82 received allocated treatment
1 received other study treatment
82 received allocated treatment
1 received other study treatment
77 alive at last follow-up
<2 y: 1
2 to <5 y: 2
5 to <10 y: 52
10 y: 22
68 alive at last follow-up
<2 y: 0
2 to <5 y: 2
5 to <10 y: 48
10 y: 18
Patients’ characteristics at baseline were reasonably balanced
between the two treatment groups (Table 2). However, when
patients were reclassified according to the IGCCC criteria, more
patients with intermediate- or poor-prognosis disease were assigned
to 4B30E360P than to 3B90E500P (16 vs 12 patients, respectively).
Two patients assigned to 4B30E360P had extrapulmonary visceral
metastases (one to liver and the other to bone) compared with
none assigned to 3B90E500P.
Table 2. Patient characteristics*
Characteristic
Median age (range), y
Tumor type
Pure seminoma
Nonseminoma
Primary site, No.
Testis
Retroperitoneum
Mediastinum (seminoma
only)
IGCCC prognostic group, No.
Good
Intermediate
Poor
Sites involved†, No.
Unknown‡
Abdominal or pelvic nodes
Lung
Mediastinum
Neck nodes
Bone or liver
3B90E500P (n = 83)
4B30E360P (n = 83)
28 (14–60)
32 (17–62)
14
69
20
63
81
2
0
81
1
1
71
10
2
67
12
4
8
57
20
5
4
0
7
60
27
8
7
2
* Reprinted with permission from Elsevier from Toner, GC, Stockler MR, Boyer
MJ, Jones M, Thomson DB, Harvey VJ, Olver IN, Dhillon H, McMullen A,
Gebski CJ, Levi JA, Simes RI. Comparison of two standard chemotherapy
regimens for good prognosis germ-cell tumours: a randomised trial. Lancet.
2001:739–745. B = bleomycin; E = etoposide; IGCCC = International Germ
Cell Consensus Classification; P = cisplatin.
† Patients could have more than one site.
‡ Elevated serum tumor markers without radiologically evident disease.
1256 Articles
|
JNCI
Survival
Overall survival remained better in patients assigned to 3B90E500P
than in those assigned to 4B30E360P (six vs 15 deaths; 8-year survival: 92% vs 83%; HR of death = 0.38, 95% CI = 0.15 to 0.97, P =
.037) (Figure 2, A, and Table 3). Two deaths in each group were
attributed to protocol treatment. Three deaths in the 3B90E500P
group and 10 deaths in the 4B30E360P group were attributed to
progression of the germ cell tumor. Four deaths were attributed to
other causes: One patient in the 3B90E500P group who refused a
blood transfusion for religious reasons died of postoperative blood
loss after resection of a large residual mass in the retroperitoneum,
and the three deaths in the 4B30E360P group were because of a fatal
acute myocardial infarction, a suicide, and a second malignancy
(metastatic large cell carcinoma) that was not considered to be
related to treatment. No deaths were attributed to bleomycin.
Sensitivity analyses that accounted for imbalances in prognostic
factors gave consistent results. For example, overall survival
remained better in patients assigned to 3B90E500P than in those
assigned to 4B30E360P when the analysis was restricted to the 138
patients who were classified with good-prognosis disease according
to IGCCC criteria (three vs 10 deaths; 8-year survival: 96% vs
85%; HR of death = 0.38, 95% CI = 0.15 to 0.98, P = .035) (Figure 2,
B, and Table 3). Analysis of all 166 patients with adjustment for
IGCCC prognostic group showed a trend in overall survival that
favored 3B90E500P and was of similar magnitude to that without
adjustment but was not statistically significantly different between
the treatment groups (HR of death = 0.40, 95% CI = 0.16 to 1.03,
P = .058).
Progression and Outcomes After Progression
Progression-free survival favored 3B90E500P but was not statistically
significantly different between the treatment groups (3B90E500P vs
4B30E360P: 11 vs 18 events; 8-year progression-free survival: 86% vs
79%; HR of progression = 0.6, 95% CI = 0.3 to 1.1, P = .15)
(Figure 2, C). There was less refractory disease, fewer early relapses,
and fewer deaths because of other causes in the 3B90E500P group vs
Vol. 102, Issue 16 | August 18, 2010
1.0
0.8
Proportion surviving
B
3B90E500P
4B30E360P
Proportion surviving
A
0.6
Hazard ratio = 0.38 (95% CI = 0.15 to 0.97)
Log-rank P = .037
0.4
3B90E500P
0.8
4B30E360P
0.6
Hazard ratio = 0.38 (95% CI = 0.15 to 0.98)
Log-rank P = .035
0.4
0.2
0.2
0.0
1.0
0
2
4
6
8
0.0
10
0
2
Years from randomization
4
6
8
10
Years from randomization
Number at risk
Numbers at risk
3B90E500P
83
80
78
68
46
22
3B90E500P
71
68
67
57
38
17
4B30E360P
83
73
69
63
44
18
4B30E360P
67
60
57
52
36
12
C
1.0
Proportion not progressing
3B90E500P
0.8
4B30E360P
0.6
Hazard ratio = 0.6 (95% CI = 0.3 to 1.1)
Log-rank P = .146
0.4
Adjusted hazard ratio = 0.60 (95% CI = 0.28 to 1.26)
Wald P = .175
0.2
0.0
0
Number at risk
3B90E500P
83
4B30E360P
83
2
4
6
8
10
Years from randomization
74
73
62
43
21
69
66
60
43
18
Figure 2. Kaplan–Meier plots of overall survival and progression-free
survival for 3B90E500P (bleomycin, etoposide, cisplatin) and 4B30E360P
after median follow-up of 8.5 years. A) Overall survival for all patients
(n = 166). B) Overall survival for 138 patients with good-risk disease by
International Germ Cell Cancer Collaborative Group (International
Germ Cell Consensus Classification [IGCCC]) criteria. C) Progressionfree survival for all patients adjusted for IGCCC risk group. All statistical
tests were two-sided. CI = confidence interval.
the 4B30E360P group; however, rates of late relapse and deaths
because of treatment were identical in the two groups (Table 3).
Three of the nine patients with nonseminoma and a residual
mass who did not undergo resection of the residual mass subsequently
relapsed. Two of these patients were in the 3B90E500P group, and one
of these patients was in the 4B30E360P group (Table 4).
Treatment received after relapse and outcomes are presented in
Table 4. Both patients with refractory disease died. Of the 15
patients with an early relapse, seven have died and eight are alive
at a median follow-up of 81 months (range = 69–122 months). All
four patients with late relapse died. Compared with patients allocated to 4B30E360P, those allocated to 3B90E500P were less likely to
die after developing refractory disease or early relapse (Table 3).
scores for eight side effects associated with chemotherapy as
assessed by the GLQ-8 (Supplementary Figure 1, available
online), for six scales of the Utility-Based Questionnaire-Cancer,
or for the Spitzer Quality of Life Index (Supplementary Table 1,
available online). During treatment (weeks 3, 6, 9, and 12 after
randomization), the average scores for all scales did not differ statistically significantly between patients allocated to 4B30E360P and
those allocated to 3B90E500P (P > .05; Figure 3 and Supplementary
Table 1 and Supplementary Figures 2–4, available online).
However, at 12 weeks after randomization, the average scores for
most scales were higher (ie, the side effect was worse) for patients
allocated to 4B30E360P than for those allocated to 3B90E500P. By
6 months after randomization (ie, after completion of treatment),
average scores for all scales except numbness or pins and needles
had returned to baseline levels or lower (Figure 3 and Supplementary
Figures 2–4, available online). At months 6, 9, and 12 after randomization (ie, after completion of treatment), compared with
patients allocated to 3B90E500P, those allocated to 4B30E360P had
higher mean scores for numbness or pins and needles (P = .003)
Health-Related Quality of Life
HRQL data were available for 149 (90%) of the 166 patients who
were randomly assigned. At baseline (ie, before treatment), there
were no differences between the treatment groups in average
jnci.oxfordjournals.org JNCI
|
Articles 1257
Table 3. Outcome of treatment*
All patients (n = 166)
Outcome
Progression
Total, No. (%)
Refractory (<3 mo), No.
Early relapse (≥3 mo to <2 y), No.
Late relapse (≥2 y), No.
Outcome of progression, No. (%)
Alive (salvaged)
Dead
Death
Total, No. (%)
Due to protocol treatment, No.
Due to progression of germ cell tumor, No.
Refractory (<3 mo)
Early relapse (≥3 mo to <2 y)
Late relapse (≥2 y)
Due to other cause, No.
IGCCC good-prognosis patients (n = 138)
3B90E500P (n = 83)
4B30E360P (n = 83)
3B90E500P (n = 71)
4B30E360P (n = 67)
8 (10)
0
6
2
13 (16)
2
9
2
6 (8)
0
5
1
9 (13)
2
5
2
5 (63)
3 (38)
3 (23)
10 (77)
5 (83)
1 (17)
2 (22)
7 (78)
6 (7)
2
3
0
1
2
1
15 (18)
2
10
2
6
2
3
3 (4)
1
1
0
0
1
1
10 (15)
1
7
2
3
2
2
* B = bleomycin; E = etoposide; IGCCC = International Germ Cell Consensus Classification; P = cisplatin.
and loss of hair (P = .04) and for the Spitzer Quality of Life Index
(P = .05). Mean scores for all other scales after completion of treatment did not differ statistically significantly between treatment
groups (Figure 3 and Supplementary Table 1 and Supplementary
Figures 2–4, available online).
Discussion
This updated analysis of a randomized trial of first-line chemotherapy for patients with good-prognosis germ cell tumors showed
that the survival benefit of 3B90E500P over 4B30E360P was maintained
with long-term follow-up. Progression-free survival favored
3B90E500P but was not statistically significantly different between
the treatment groups. There were few differences in quality of life
between treatment groups. However, at 12 weeks after random
assignment, the average scores for most side effect scales favored
3B90E500P. After the completion of treatment, the average scores
favored 3B90E500P for numbness, hair loss, and Spitzer Quality of
Life Index (P < .05). After the completion of treatment, the average
scores for all other scales had returned to baseline or better.
Optimal treatment of metastatic germ cell tumors is critical
because these tumors are lethal, occur at a young age, and are
highly curable (13). The majority of patients with metastatic
disease have good-prognosis disease by IGCCC criteria (8). Since
the introduction of BEP and EP chemotherapy in the 1980s, the
cure rate for patients with good-prognosis disease is 90% or higher
(3,5). Since then, a series of randomized trials have aimed to maintain this high cure rate while reducing toxicity. These trials have
demonstrated that three cycles of BEP is as effective as four (14–
16), that cisplatin has similar efficacy and toxicity when the
same total dose is given for a period of 3 days rather than 5 days
(16), and that carboplatin is less effective than cisplatin (4,17).
Randomized trials have also demonstrated that excluding bleomycin from cisplatin-based regimens is associated with worse survival
when using three cycles (18) or when using vinblastine instead of
etoposide (1) and with an inferior response rate when using four
cycles with etoposide at 360 mg/m2 per cycle (19).
1258 Articles
|
JNCI
The purpose of this trial was to compare three cycles of a BEP
regimen similar to that developed at Indiana University (3) with
four cycles of a less dose-intense BEP regimen that was being
used in the United Kingdom and Europe when this trial was
designed. The findings of this trial suggest that the dose intensity
of etoposide and the total dose of bleomycin are critical in determining the outcome of therapy.
Our results confirm that the less dose-intense regimen was inferior but do not allow us to confidently identify which of the many
differences between the regimens is most important to explain the
differences in outcome. However, it is unlikely that differences in
the schedule of administration of cisplatin are responsible for the
differences in outcome because a previous trial (16) showed that
cisplatin has similar efficacy and toxicity when the same total dose
is given for a period of 3 days rather than for a period of 5 days. In
addition, the total dose of cisplatin was lower in the 3B90E500P arm
than in the 4B30E360P arm, which rules out this difference as an explanation for the better outcomes in patients who received 3B90E500P.
There were major differences between the treatment regimens in
the dose intensity and total dose of bleomycin. Therefore, it is possible that the variation in bleomycin dose between treatment arms in
this trial is responsible for the observed differences in survival.
Although there was only a minor difference in the total received
dose of etoposide between the two regimens, the dose intensity of
etoposide was 28% lower in the 4B30E360P arm compared with the
3B90E500P arm. We consider this difference in etoposide dose
intensity to be the most likely explanation for the better outcome
observed with the more dose-intense regimen, and we recommend
that etoposide at 500 mg/m2 per cycle should be the standard dose
intensity used in all adult germ cell treatment programs.
Patients were less likely to die following progression after treatment with 3B90E500P than after treatment with 4B30E360P, and this
difference may partly explain the benefit of 3B90E500P in overall
survival but not in progression-free survival. However, we cannot
exclude the possibility that patients allocated to 4B30E360P received
second-line treatment for refractory disease or relapse that was
inferior to that received by patients allocated to 3B90E500P.
Vol. 102, Issue 16 | August 18, 2010
jnci.oxfordjournals.org JNCI
|
Articles 1259
M
NS
S
NS
NS
NS
S
NS
S
NS
NS
NS
Good
Good
Intermediate
Good
Poor
Good
T
T
NS
S
NS
S
Intermediate
Good
RP
RP, lung
RP
RP, lung, M
RP
RP
RP, lung
RP, lung
Lung, M
None
RP
Lung, M
RP, CLN
RP, lung
RP
Neck
RP
Lung, liver
None
Lung
RP
Site of
metastases
3B90E500P
4B30E360P
4B30E360P
3B90E500P
3B90E500P
3B90E500P
4B30E360P
4B30E360P
4B30E360P
3B90E500P
3B90E500P
3B90E500P
3B90E500P
4B30E360P
4B30E360P
4B30E360P
4B30E360P
4B30E360P
4B30E360P
4B30E360P
4B30E360P
Chemotherapy
allocated
None
None
None
None
None
RPLND
None
None
None
None
RPLND
None
RPLND
Lung
resection
RPLND
None
None
None
None
None
RPLND
Surgery
Residual mass†
CR after chemo
CR after chemo
Residual mass†
CR after chemo
CR after chemo
CR after chemo
CR after chemo
CR after
chemo +
surgery
CR after
chemo +
surgery
Residual mass†
CR after
chemo
Incomplete
response
Residual
mass†
CR after
chemo
CR after
chemo +
surgery
Incomplete
response
requiring
chemo for CR
CR after
chemo +
surgery
CR after chemo
Incompletely
resected
Progressive
disease
Response to
treatment
M, neck
RP (including
pelvis)
RP, M, lung
RP, lung,
brain
RP
RP (pelvis)
M, lung,
liver
RP, lung
RP, neck
Neck
RP
Lung
RP
M, lung,
liver, brain
RP
M, bone
RP
T, lung, liver
Bone
Lung, M,
neck
RP
Site(s) of
relapse
Scan + markers
Scan
Clinical
Clinical
Scan only
Scan only
Scan + markers
Clinical
Clinical
Scan + markers
Scan only
Scan only
Clinical
Marker only
Marker only
Scan only
Scan + markers
Marker only
Scan + markers
Clinical
Clinical
Evidence of
relapse
71
103
26
29
15
12
9
9
9
9
11
8
8
8
7
7
6
6
4
2.7
2
Time to
relapse, mo
Surgery‡
Chemo
Chemo
Chemo
Chemo, surgery‡
Chemo, HDCT
Chemo, HDCT
Chemo, RT
Chemo
Chemo
Chemo, surgery‡
Chemo
Chemo
HDCT
HDCT, RT, surgery‡
Chemo, HDCT
HDCT
Chemo, RT
Chemo, HDCT
RT
Chemo
Treatment at
relapse
DOD
DOD
DOD
DOD
AWD
AWD
AWD
AWD
AWD
AWD
AWD
AWD
DOD
DOD
DOD
DOD
DOD
DOD
DOD
DOD
DOD
Current
status
14
0
11
2
83
75
73
93
69
112
99
78
17
8
17
12
12
10
3
0
10
Months since
relapse
‡ Surgical resection of residual mass after chemotherapy.
† Residual mass evident on imaging studies with normal serum tumor markers.
chemotherapy with stem cell transplantation; IGCCC = International Germ Cell Consensus Classification; M = mediastinum; NS = nonseminoma; P = cisplatin; RP = retroperitoneum; RPLND = retroperitoneal lymph node
dissection; RT = radiotherapy; S = pure seminoma; T = testis.
* AWD = alive without disease; B = bleomycin; chemo = standard-dose chemotherapy; CLN = cervical lymph node dissection; CR = complete response; DOD = died of disease (germ cell tumor); E = etoposide; HDCT = high-dose
T
T
T
NS
Good
Late relapse (≥2 y)
Good
Good
T
S
T
T
T
T
T
T
T
T
T
T
Good
Intermediate
Good
Good
Good
Good
T
NS
Poor
T
T
Early relapse (≥3 mo to <2 y)
Intermediate
NS
S
Good
Primary
site
T
Histology
Refractory disease (<3 mo)
Good
S
IGCCC prognostic
group
Table 4. Characteristics and outcomes of patients with refractory disease or relapse*
A
B
60
50
40
Mean score
Mean score
50
60
30
4B30E360P
20
40
30
4B30E360P
20
10
10
3B90E500P
0
3B90E500P
0
3
0
6
9 12
26
39
0
52
3
6
D
60
26
39
52
60
50
50
40
40
Mean score
Mean score
C
9 12
Weeks from randomization
Weeks from randomization
30
4B30E360P
20
30
4B30E360P
20
10
10
3B90E500P
0
3B90E500P
0
3
0
6
9 12
26
39
52
Weeks from randomization
0
3
6
9 12
26
39
52
Weeks from randomization
Figure 3. Quality of life for patients assigned to 3B90E500P and
4B30E360P. Health-related quality of life was assessed by GLQ-8 at
baseline (ie, before treatment); at weeks 3, 6, and 9 after random
assignment (ie, during treatment); at week 12 (ie, at completion of
treatment); and at months 6, 9, and 12 after random assignment (ie,
after completion of treatment). A higher score indicates greater impairment. A) Tiredness. B) Feeling sick. C) Feeling anxious or
depressed. D) Numbness or pins and needles. Error bars correspond
to 95% confidence intervals. B = bleomycin; E = etoposide; P =
cisplatin.
The survival rates in both treatment arms were lower than
those reported in other trials of first-line chemotherapy for goodprognosis metastatic germ cell tumors (15–19). This difference
probably reflects the inclusion in this trial of 28 patients with
intermediate- and poor-risk disease according to the more recent
IGCCC criteria (8). The prognostic classification that we originally used to select patients for this trial resulted in our inclusion
of a larger number of patients in the good-prognosis category
compared with other classifications that were available when we
designed the study (6). A second possible explanation for the
lower survival rates in this study is the failure of patients to
undergo resection of residual masses after first-line chemotherapy
(which occurred in nine patients in this trial) or at relapse.
Surgery to resect residual masses is important but is not done
uniformly or at all treatment centers in Australia, New Zealand,
and Europe.
The occurrence of relapses in both study arms more than
2 years after the completion of treatment affirms the importance of
long-term follow-up. Late relapses in this study were associated
with uniformly poor outcomes.
We found few differences in HRQL between regimens. The
main difference was in feelings of numbness or pins and needles,
which were worse in the 4B30E360P arm than in the 3B90E500P arm.
The 4B30E360P arm included a single day of cisplatin each cycle and
a higher total dose of cisplatin compared with the 3B90E500P arm
(400 vs 300 mg/m2). We found that average scores for most scales
had returned to baseline levels or better after treatment, which is
reassuring and consistent with results of most other studies
(16,20,21). Although we did not study long-term survivorship issues, such as sexual function, potential cardiac and vascular toxicity
(22), and psychological morbidity, we recognize their importance
and the need for further study of these issues (23,24).
The strengths of this study are its randomized design, intention-totreat analysis, and the fact that the results after long-term follow-up
support the results of the initial analysis. A limitation of the trial is
that the effects of treatment on survival may have been overestimated because the trial was stopped early after a planned interim
analysis met predefined stopping rules. Results from analyses that
prompt early stopping tend to exaggerate the effects of treatment
(25). Another limitation was the imbalance between treatment
groups by IGCCC criteria, which were not available when the trial
was designed. However, our sensitivity analyses that adjusted for
this imbalance yielded results and conclusions similar to those of
the primary analysis.
1260 Articles
|
JNCI
Vol. 102, Issue 16 | August 18, 2010
The BEP regimen that was shown to be superior in this trial,
which included etoposide at 500 mg/m2 per cycle and bleomycin at
30 000 international units weekly, is the regimen that seems to be
most widely used for good-prognosis metastatic germ cell tumors
at specialized treatment centers around the world, as is reflected in
international guidelines (26,27). Excellent cure rates of 90% or
higher are expected with this regimen (13,14). Future studies in
men with good-prognosis metastatic germ cell tumors should
focus on better predictors of prognosis and response, including
molecular markers such as DNA methylation (28), pharmacogenomic profiles such as variants of genes encoding cytochrome
P450 3A5 (29) and bleomycin hydrolase (30), and slower-thanexpected decline of serum tumor markers human chorionic gonadotropin and a-fetoprotein (31); on efforts to better prevent,
predict, and manage acute toxicity, including emesis, pneumonitis,
infection, and fatigue; and on long-term survivorship issues, including sexual function, cardiovascular toxicity, and psychological
morbidity. Patients with germ cell tumors should be referred to
high-volume specialized treatment centers, where they are more
likely to receive optimal therapy (32–34). Future phase III trials in
metastatic testicular cancer should involve international collaborations to ensure that sufficient numbers of patients with this rare
condition are recruited in a timely period.
Supplementary Data
Supplementary data can be found at http://www.jnci.oxfordjournals
.org/.
References
1. Levi JA, Raghavan D, Harvey V, et al. The importance of bleomycin in
combination chemotherapy for good-prognosis germ cell carcinoma.
Australasian Germ Cell Trial Group. J Clin Oncol. 1993;11(7):1300–1305.
2. Toner GC, Stockler MR, Boyer MJ, et al. Comparison of two standard
chemotherapy regimens for good-prognosis germ-cell tumours: a randomised
trial. Australian and New Zealand Germ Cell Trial Group. Lancet.
2001;357(9258):739–745.
3. Williams SD, Birch R, Einhorn LH, et al. Treatment of disseminated
germ-cell tumors with cisplatin, bleomycin, and either vinblastine or
etoposide. N Engl J Med. 1987;316(23):1435–1440.
4. Horwich A, Sleijfer DT, Fossa SD, 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. 1997;15(5):
1844–1852.
5. Bosl GJ, Geller NL, Bajorin D, et al. A randomized trial of etoposide +
cisplatin versus vinblastine + bleomycin + cisplatin + cyclophosphamide +
dactinomycin in patients with good-prognosis germ cell tumors. J Clin
Oncol. 1988;6(8):1231–1238.
6. Bajorin D, 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. 1988;6(5):786–792.
7. Toner GC, Panicek DM, Heelan RT, et al. Adjunctive surgery after chemotherapy for nonseminomatous germ cell tumors: recommendations for
patient selection. J Clin Oncol. 1990;8(10):1683–1694.
8. International Germ Cell Cancer Collaborative Group. International
Germ Cell Consensus Classification: a prognostic factor-based staging
system for metastatic germ cell cancers. International Germ Cell Cancer
Collaborative Group. J Clin Oncol. 1997;15(2):594–603.
9. Coates A, Glasziou P, McNeil D. On the receiving end—III. Measurement of
quality of life during cancer chemotherapy. Ann Oncol. 1990;1(3):213–217.
jnci.oxfordjournals.org 10. Grimison P, Simes RJ, Hudson HM, et al. Deriving a patient-based utility
index from a cancer-specific quality of life questionnaire. Value Health.
2009;12(5):800–807.
11. Spitzer WO, Dobson AJ, Hall J, et al. Measuring the quality of life of
cancer patients: a concise QL-Index for use by physicians. J Chronic Dis.
1981;34(12):585–597.
12. Sloan JA, Loprinzi CL, Kuross SA, et al. Randomized comparison of four
tools measuring overall quality of life in patients with advanced cancer.
J Clin Oncol. 1998;16(11):3662–3673.
13. Horwich A, Shipley J, Huddart R. Testicular germ-cell cancer. Lancet.
2006;367(9512):754–765.
14. Einhorn LH, Williams SD, Loehrer PJ, et al. Evaluation of optimal duration of chemotherapy in favorable-prognosis disseminated germ cell
tumors: a Southeastern Cancer Study Group protocol. J Clin Oncol. 1989;
7(3):387–391.
15. 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. 1998;16(2):702–706.
16. de Wit R, Roberts JT, Wilkinson PM, et al. 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
Council. J Clin Oncol. 2001;19(6):1629–1640.
17. Bajorin DF, Sarosdy MF, Pfister DG, et al. Randomized trial of etoposide
and cisplatin versus etoposide and carboplatin in patients with good-risk germ
cell tumors: a multiinstitutional study. J Clin Oncol. 1993;11(4):598–606.
18. Loehrer PJS, Johnson D, Elson P, et al. Importance of bleomycin in
favorable-prognosis disseminated germ cell tumors: an Eastern Cooperative
Oncology Group trial. J Clin Oncol. 1995;13(2):470–476.
19. 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.
1997;15(5):1837–1843.
20. Fleer J, Hoekstra HJ, Sleijfer DT, et al. Quality of life of survivors of
testicular germ cell cancer: a review of the literature. Support Care Cancer.
2004;12(7):476–486.
21. Dahl A, Mykletun A, Fossa S. Quality of life in survivors of testicular
cancer. Urol Oncol. 2005;23(3):193–200.
22. Nuver J, Smit AJ, Sleijfer DT, et al. Left ventricular and cardiac autonomic function in survivors of testicular cancer. Eur J Clin Invest. 2005;
35(2):99–103.
23. Luckett T, Butow PN, King MT, et al. Psycho-social issues in long-term
survivors of testicular cancer: directions for future research. Asia Pac J Clin
Oncol. 2008;4(3):125–131.
24. Rudberg L, Carlsson M, Nilsson S, et al. Self-perceived physical, psychologic, and general symptoms in survivors of testicular cancer 3 to 13 years
after treatment. Cancer Nurs. 2002;25(3):187–195.
25. Pocock. When to stop a clinical trial. BMJ. 1992;305(6847):235–240.
26. Albers P, Albrecht W, Algaba F, et al. Guidelines on Testicular Cancer.
Arnhem, the Netherlands: European Association of Urology; 2009. http://
www.uroweb.org. Accessed May 15, 2010.
27. National Comprehensive Cancer Network. NCCN Clinical Practice
Guidelines in Oncology for Testicular Cancer 2010. Fort Washington, PA:
NCCN; 2009. www.nccn.org. Accessed May 15, 2010.
28. Lind GE, Skotheim RI, Lothe RA. The epigenome of testicular germ cell
tumors. APMIS. 2007;115(10):1147–1160.
29. Zhuo X, Zheng N, Felix CA, et al. Kinetics and regulation of cytochrome
P450-mediated etoposide metabolism. Drug Metab Dispos. 2004;32(9):
993–1000.
30. de Haas EC, Zwart N, Meijer C, et al. Variation in bleomycin hydrolase
gene is associated with reduced survival after chemotherapy for testicular
germ cell cancer. J Clin Oncol. 2008;26(11):1817–1823.
31. Fizazi K, Culine S, Kramar A, et al. Early predicted time to normalization
of tumor markers predicts outcome in poor-prognosis nonseminomatous
germ cell tumors. J Clin Oncol. 2004;22(19):3868–3876.
JNCI
|
Articles 1261
32. Toner GC, Neerhut GJ, Schwarz MA, et al. The management of testicular cancer in Victoria, 1988-1993. Urology Study Committee of the
Victorian Co-operative Oncology Group. Med J Aust. 2001;174(7):
328–331.
33. Collette L, Sylvester RJ, Stenning SP, et al. Impact of the treating institution
on survival of patients with “poor-prognosis” metastatic nonseminoma.
J Natl Cancer Inst. 1999;91(10):839–846.
34. Feuer EJ, Sheinfeld J, Bosl GJ. Does size matter? Association between
number of patients treated and patient outcome in metastatic testicular
cancer. J Natl Cancer Inst. 1999;91(10):816–818.
Funding
This work was supported by the National Health and Medical Research Council
to NHMRC Clinical Trials Center (Unit Grant), New South Wales Cancer
Council (Program Grant), and Apex Foundation and Apex Clubs of Australia
(Donation).
Notes
No authors had a conflict of interest. All authors gave their approval for
publication. The authors had full responsibility for designing and analyzing
the study, interpreting the data, writing the manuscript, and submitting it for
publication.
The following investigators contributed to this trial: D. B. Thomson,
E. Walpole, D. Jackson (Princess Alexandra Hospital, Brisbane, Australia);
V. J. Harvey, P. Thompson, T. Vaughan (Auckland Hospital, Auckland,
New Zealand); C. R. Lewis, M. Friedlander, K. McDonald (Prince of Wales
Hospital, Sydney, Australia); J. A. Levi, D. Bell, H. Wheeler, M. Glass (Royal
North Shore Hospital, Sydney, Australia); G. C. Toner, V. Walcher (Peter
MacCallum Cancer Institute, Melbourne, Australia); M. J. Boyer, M. Tattersall,
A. Coates, N. Teriana (Royal Prince Alfred Hospital, Sydney, Australia); P. Craft,
J. May (Canberra Hospital, Canberra, Australia); H. Gurney, R. Kefford,
P. Harnett, N. Wilcken, S. Luke (Westmead Hospital, Sydney, Australia);
E. Bayliss, H. Weinstein (Royal Perth Hospital, Perth, Australia); I. N. Olver,
T. Marafioti (Royal Adelaide Hospital, Adelaide, Australia); S. Ackland,
1262 Articles
|
JNCI
J. Stewart, T. Bonaventura, S. Brew (Newcastle Mater Misericordiae Hospital,
Newcastle, Australia); D. Grimes, D. Wyld, J. Campbell (Royal Brisbane
Hospital, Brisbane, Australia); D. Goldstein, E. Moylan, A. Goldrick, D. Burns
(Liverpool Hospital, Sydney, Australia); D. Boadle, B. Sundstrup, C. Carter
(W P Holman Clinic, Launceston, Australia); R. Bell, R. McLennan, K. White,
A. Woollett (Geelong Hospital, Geelong, Australia); D. Kotasek, B. Dale,
J. Norman, K. Pittman, S. Moldovan (Queen Elizabeth Hospital, Adelaide,
Australia); W. I. Burns, R. Snyder, R. Kennedy (St Vincent’s Hospital,
Melbourne, Australia); R. Lowenthal, R. Kimber, N. Le Mottee (Royal Hobart
Hospital, Hobart, Australia); E. Abdi, K. Dunne (Bendigo Base Hospital,
Bendigo, Australia); C. Underhill, K. Clarke, N. McMonigle (Albury Base and
Murray Valley Private Hospitals, Albury, Australia); J. Shapiro, M. Schwarz,
M. D’Astoli (The Alfred Hospital, Melbourne, Australia); and M. R. Stockler,
P. Beale, L. Truong (Concord Repatriation General Hospital, Sydney, Australia).
We thank the patients for their commitment to the study; the principal investigators, coinvestigators, and study coordinators at the participating centers
for their dedication and enthusiasm; and K. Mann (Statistician) and other staff at
the National Health and Medical Research Council Clinical Trials Centre who
were responsible for central coordination and data management of the trial.
Affiliations of authors: NHMRC Clinical Trials Centre, University of Sydney,
Sydney, Australia (PSG, VJG, ALB, RJS); Sydney Cancer Centre, RPA and
Concord Hospitals, Sydney, Australia (MRS, MJB); Department of Medical
Oncology, Princess Alexandra Hospital, Brisbane, Australia (DBT); Cancer
Council Australia, Sydney, Australia (INO); Regional Cancer Service,
Department of Medical Oncology, Auckland Hospital, Auckland, New
Zealand (VJH); Department of Medical Oncology, Prince of Wales Hospital,
Sydney, Australia (CRL); Department of Medical Oncology, Royal North
Shore Hospital, Sydney, Australia (JAL); Department of Medical Oncology,
Westmead Hospital, Sydney, Australia (HG); Department of Medical
Oncology, Canberra Hospital, Canberra, Australia (PC); Department of
Medical Oncology, Peter MacCallum Cancer Centre, Melbourne, Australia
(GCT).
Vol. 102, Issue 16 | August 18, 2010