- Email: [email protected]
PREPARE trial: a randomized phase III trial comparing preoperative, dose-dense, dose-intensified chemotherapy with epirubicin, paclitaxel, and CMF versus a standard-dosed epirubicin–cyclophosphamide followed by paclitaxel with or without darbepoetin alfa in primary breast cancer—outcome on prognosis
original article
Annals of Oncology 22: 1999–2006, 2011 doi:10.1093/annonc/mdq713 Published online 7 March 2011
PREPARE trial: a randomized phase III trial comparing preoperative, dose-dense, dose-intensified chemotherapy with epirubicin, paclitaxel, and CMF versus a standard-dosed epirubicin–cyclophosphamide followed by paclitaxel with or without darbepoetin alfa in primary breast cancer—outcome on prognosis
Received 31 August 2010; revised 10 November 2010; accepted 12 November 2010
Background: The objective of this study was to compare the effect of dose-intensified neoadjuvant chemotherapy with that of standard epirubicin plus cyclophosphamide followed by paclitaxel in combination with or without darbepoetin on survival in primary breast cancer. Patients and methods: A total of 733 patients received either four cycles of neoadjuvant epirubicin 90 mg/m2 plus cyclophosphamide 600 mg/m2 every 3 weeks followed by four cycles of paclitaxel 175 mg/m2 every 3 weeks (EC/T), or three cycles of epirubicin 150 mg/m2 every 2 weeks followed by three cycles of paclitaxel 225 mg/m2 every 2 weeks followed by three cycles of combination chemotherapy with cyclophosphamide, methotrexate, and fluorouracil (Edd/Tdd/CMF). The patients were randomly assigned to receive darbepoetin or none. The primary objective was to demonstrate a superior disease-free survival (DFS) of Edd/Tdd/CMF compared with EC/T. Results: Estimated 3-year DFS was 75.8% with EC/T versus 78.8% with Edd/Tdd/CMF [hazard ratio (HR) 1.14; P = 0.37] and overall survival (OS) 88.4% versus 91.5% (HR 1.26; P = 0.237). Three-year DFS was 74.3% with darbepoetin versus 80.0% without (HR 1.31; P = 0.061) and OS 88.0% versus 91.8% (HR 1.33; P = 0.139). Patients with a pathologically documented complete response [pathological complete response (pCR)] had a significantly better DFS compared with those without achieving a pCR (estimated 3-year DFS: 89.2% versus 74.9%; HR 2.27; P = 0.001). Conclusion: Neoadjuvant dose-intensified chemotherapy compared with standard chemotherapy did not improve DFS, whereas the addition of darbepoetin might have detrimental effects on DFS. Key words: darbepoetin alfa, neoadjuvant dose-dense chemotherapy, neoadjuvant dose-intensified therapy, primary breast cancer
introduction *Correspondence to: Prof. M. Untch, Department of Obstetrics and Gynecology, Helios Klinikum Berlin-Buch, Schwanebecker Chaussee 50, 13125 Berlin, Germany. Tel: +49-30-9401-53300; Fax: +49-30-9401-53309; E-mail: [email protected]
Neoadjuvant chemotherapy has become an acknowledged alternative to adjuvant chemotherapy not only for locally advanced breast cancer but also for those patients who are
ª The Author 2011. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: [email protected]
original article
1 Department of Obstetrics and Gynecology, Helios Klinikum Berlin-Buch, Berlin; 2German Breast Group, Neu-Isenburg, Germany; 3Division of Hematology and Oncology, David Geffen School of Medicine, University of California, Los Angeles, USA; 4Department of Obstetrics and Gynecology, St Barbara Hospital, Hamm; 5 Hematologic/Oncologic Practice, Wuppertal; 6University Women’s Hospital of Ulm, Ulm; 7Department of Obstetrics and Gynecology, Medical University of Hannover, Hannover; 8University Women’s Hospital of Freiburg, Freiburg; 9Klinikum, Kassel; 10Krankenhaus Bergisch, Gladbach; 11University Breast Center Franconia, Department of Gynecology and Obstetrics, University Hospital Erlangen, Erlangen; 12Hematologic/Oncologic Clinic Mu¨nchen, Munich; 13Department of Obstetrics and Gynecology, Technical University of Munich, Munich; 14Department of Obstetrics and Gynecology, University Hospital Hamburg-Eppendorf, Hamburg; 15University Women’s Hospital of Mainz, Mainz; 16Department of Obstetrics and Gynecology, Ludwig-Maximilians-University Großhadern, Munich; 17Department of Pathology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany
Downloaded from http://annonc.oxfordjournals.org/ by guest on August 26, 2015
M. Untch1*, G. von Minckwitz2, G. E. Konecny3, U. Conrad4, W. Fett5, C. Kurzeder6, H.-J. Lu¨ck7, E. Stickeler8, H. Urbaczyk9, B. Liedtke10, M. W. Beckmann11, C. Salat12, N. Harbeck13, V. Mu¨ller14, M. Schmidt15, S. Hasmu¨ller16, M. Lenhard16, V. Nekljudova2, A. Lebeau17, S. Loibl2 & P. A. Fasching11 on behalf of the Arbeitsgemeinschaft Gyna¨kologische Onkologie PREPARE investigators
original article
Annals of Oncology
preoperatively by CMF compared with a standard anthracycline–taxane-based combination as neoadjuvant treatment without postoperative chemotherapy. In a second randomization, the effect of darbepoetin alfa on anemia and long- as well as short-term outcome was investigated. The primary objective was to examine the DFS additionally to usual objectives like pCR rate and breast-conserving therapy [3].
patients and methods patients This is a phase III, multicenter, prospective, randomized, open-label clinical trial. Patients aged ‡18 and £65 years with histologically confirmed primary breast cancer by core biopsy were enrolled. The primary tumor had to be ‡2 cm based on either clinical or ultrasound measurement. Patients with inflammatory breast cancer were also included. Detailed inclusion and exclusion criteria are described elsewhere [3]. Approval was obtained from the ethics committees of all participating study sites and all patients gave written informed consent to take part in this trial.
treatment plan Eligible patients were randomly assigned to the control arm with epirubicin 90 mg/m2 plus cyclophosphamide 600 mg/m2 every 3 weeks for four cycles followed by paclitaxel 175 mg/m2 every 3 weeks for four cycles (EC/T) or to the experimental arm with sequential dose-dense,
Table 1. Disease-free survival (DFS) and overall survival (OS) estimated at 3 years: overall and according to treatment as well as prognostic subgroups n Overall 714 Darbepoetin 2 369 + 345 Age (years) <50 373 ‡50 341 Tumor grade 1 or 2 236 3 214 Steroid receptor statusa ER and/or PgR positive 310 ER and PgR negative 144 Tumor size (cm) <4 472 ‡4 221 Pathological lymph node status at Involved 277 Not involved 374 pCRa + 117 2 597
EC/T (n = 362) Edd/Tdd/CMF (n = 352) DFS (%) OS (%) DFS (%) OS (%)
Without darbepoetin (n = 369) DFS (%) OS (%)
With darbepoetin (n = 345) DFS (%) OS (%)
75.6
88.4
78.8
91.5
78.0
91.8
74.3
88.0
78.0 73.6
91.0 85.7
82.1 75.2
92.6 90.4
– –
– –
– –
– –
73.0 78.7
86.8 90.1
80.2 77.1
91.2 91.9
77.8 83.0
90.6 93.4
75.3 73.2
87.2 88.8
80.7 69.7
94.7 82.6
86.0 70.3
96.5 86.7
82.7 75.8
94.8 89.0
84.0 63.0
96.5 78.8
74.7 74.0
92.6 78.8
83.5 70.0
94.5 83.3
78.7 77.6
94.0 85.1
79.4 65.7
93.1 76.0
79.8 71.0 surgery 62.6 87.1
92.2 82.0
79.0 82.7
92.4 92.5
77.7 86.2
91.8 91.1
81.4 66.6
92.8 82.7
81.3 95.4
66.6 88.5
83.6 98.3
70.4 88.7
85.1 97.8
58.4 86.8
79.5 95.8
93.6a 72.9
97.9a 86.8
86.0 77.1
95.4 90.6
87.8a 78.4
96.6a 90.8
90.9a 71.1
96.3a 86.4
a
Statistically significant differences between the patients with and without pCR in EC/T, without darbepoetin and with darbepoetin. EC/T, epirubicin plus cyclophosphamide followed by paclitaxel; Edd/Tdd/CMF, epirubicin followed by paclitaxel followed by combination chemotherapy with cyclophosphamide, methotrexate, and fluorouracil; ER, estrogen receptor; PgR, progesterone receptor; pCR, pathologically documented complete response.
2000 | Untch et al.
Volume 22 | No. 9 | September 2011
Downloaded from http://annonc.oxfordjournals.org/ by guest on August 26, 2015
candidates for adjuvant chemotherapy. Combinations of anthracyclines and taxanes are widely used within this indication. Higher response rates turn into higher rates of breastconserving surgery. But it has not conclusively been proven that a higher rate of pathologically documented complete response [pathological complete response (pCR)] will result in an improved long-term outcome. However, pCR is a highly significant predictor for improved disease-free survival (DFS) and overall survival (OS) regardless of treatment [1]. Previously, we have shown that an increased pCR achieved by intensified preoperative chemotherapy was significantly associated with a better DFS and OS [2]. The previous AGO-1 study compared a preoperative dose-dense, dose-intensified, sequential administration of epirubicin and paclitaxel followed [Edd/Tdd surgery /CMF (combination chemotherapy with cyclophosphamide, methotrexate, and fluorouracil)] with a conventionally dosed combination therapy of epirubicin and paclitaxel, both regimens followed postoperatively by standard CMF (epirubicin-taxol /surgery /CMF). The dose-dense chemotherapy showed a significantly higher pCR rate [18% versus 10%; odds ratio (OR) 1.89; P = 0.008] and a significantly better DFS [hazard ratio (HR) 0.71; P = 0.011] and OS (HR 0.69; P = 0.041). The PREPARE (Preoperative Epirubicin, Paclitaxel, Darbepoetin) trial investigated a dose-dense schedule followed
original article
Annals of Oncology
Treatment was reinduced with 50% of the dose, if hemoglobin decreased again to £13.0 g/dl.
end points The primary end point of the PREPARE study was DFS. Secondary end points included the effect of chemotherapy on OS as well as the effects of darbepoetin alfa on DFS and OS. Short-term secondary efficacy end points including pCR and pathohistological lymph node status at surgery have been presented elsewhere [3].
statistical analysis The sample size of 720 patients for this study was chosen to provide 80% power to detect an improvement in DFS at 3 years from the standard therapy with a DFS of 70% to a DFS of 80% in patients receiving dosedense chemotherapy. This equals an HR of 1.4 with a type I error of alpha 5% using a one-sided test. Patients were randomly assigned in a 1 : 1 allocation to receive standard-dose or dose-intensified preoperative chemotherapy. Patients within each treatment arm were further randomly assigned in a 1 : 1 allocation to receive darbepoetin alfa or no darbepoetin alfa therapy.
Figure 1. Disease-free survival (DFS) and overall survival (OS) according to treatment. (A) DFS according to chemotherapy (one-sided log-rank test P = 0.185); (B) DFS according to darbepoetin alfa; (C) OS according to chemotherapy; (D) OS according to darbepoetin alfa. EC/T, epirubicin plus cyclophosphamide followed by paclitaxel; Edd/Tdd/CMF, epirubicin followed by paclitaxel followed by combination chemotherapy with cyclophosphamide, methotrexate, and fluorouracil.
Volume 22 | No. 9 | September 2011
doi:10.1093/annonc/mdq713 | 2001
Downloaded from http://annonc.oxfordjournals.org/ by guest on August 26, 2015
dose-intensified treatment with epirubicin 150 mg/m2 every 2 weeks for three cycles followed by paclitaxel 225 mg/m2 every 2 weeks for three cycles followed by CMF (cyclophosphamide 500 mg/m2, methotrexate 40 mg/m2, and 5-fluorouracil 600 mg/m2) on days 1 and 8 every 4 weeks for three cycles (Edd/Tdd/CMF). In addition, the patients were randomly assigned to receive darbepoetin alfa 4.5 lg/kg body weight every 2 weeks starting with the first dose of epirubicin (day 1) until 14 days after the last dose of paclitaxel and 200 mg oral iron (Fe2+) daily or no darbepoetin [3]. Patients in the Edd/Tdd/CMF arm received pegfilgrastim 6 mg s.c. on day 2 of cycles one to six (Edd/Tdd). Primary prophylactic use of pegfilgrastim during CMF treatment or in the EC/T arm was not mandatory. If a patient had leucopenia for ‡3 days, fever >38.5C, or an active antibiotic-requiring infection, filgrastim (5 lg/kg body weight daily) was administered until recovery. In all consecutive cycles, pegfilgrastim (6 mg) was used prophylactically on day 2. All patients received standard supportive therapy. In the case of tumor progression during chemotherapy, study treatment was discontinued and further treatment was at the discretion of the investigator. To achieve the target hemoglobin level between 12.5 and 13 g/dl for patients receiving darbepoetin alfa, the dose was doubled if the increase was <1 g/dl during the first 4 weeks or discontinued if 14 g/dl was exceeded.
original article The statistical analyses of efficacy end points were carried out on the full analysis set, which included all patients who were randomly assigned to the chemotherapy treatment. Patients who never started therapy were excluded from this analysis. DFS was defined as the time from informed consent to first documentation of relapse or death due to any cause. OS was the time from the date of informed consent to the date of death due to any cause. Local DFS was defined as time in weeks between the date of signing informed consent and date of local recurrence. Patients with no local recurrence reported were censored at the date of the last contact. For univariable analysis, the log-rank test and Cox proportional hazards models were used for group comparisons, and Kaplan–Meier curves were used to estimate DFS and OS probabilities. For multivariable analysis, Cox proportional hazards models for adjusting survival end points were used; full models were fit without any selection of variables. Adjustments were made for age, hormone receptor status, clinical tumor size and nodal status, grade, chemotherapy arm, darbepoetin alfa application, and pCR. Hormone receptor status and grading were captured via the histological reports. All P values are two-sided; except for the primary end point
Annals of Oncology
DFS, a one-sided test was assumed but the two-sided P value is also provided. All calculations were done with SAS version 9.2 under Enterprise Guide 4.1.
results baseline characteristics This trial was conducted in 78 centers in Germany. Between June 2002 and February 2005, 736 patients were enrolled of whom 733 were randomly assigned, 370 in the EC/T group and 363 in the Edd/Tdd/CMF group. Three patients were not randomly assigned due to major inclusion criteria violation. Additionally, the patients were randomly assigned to receive darbepoetin alfa treatment (n = 356) or no darbepoetin treatment (control n = 377). A total of 714 patients who received at least one dose of chemotherapy, 362 receiving EC/T and 352 receiving Edd/Tdd/CMF, were included in this analysis. Downloaded from http://annonc.oxfordjournals.org/ by guest on August 26, 2015
Figure 2. Survival analyses by pCR. (A) Disease-free survival in the whole population by pCR; (B) disease-free survival by pCR stratified according to chemotherapy; (C) overall survival in the whole population by pCR; (D) overall survival by pCR stratified according to chemotherapy. pCR, pathologically documented complete response; EC/T, epirubicin plus cyclophosphamide followed by paclitaxel; Edd/Tdd/CMF, epirubicin followed by paclitaxel followed by combination chemotherapy with cyclophosphamide, methotrexate, and fluorouracil.
2002 | Untch et al.
Volume 22 | No. 9 | September 2011
Annals of Oncology
The median age at randomization was 48 years (23–65 years). Fifty-eight (7.9%) patients had clinical T4 carcinoma including those with inflammatory disease. Baseline characteristics were comparable between the treatment arms (Table 1).
compliance with treatment The flow of patients through the study is illustrated in another publication [3]. Of the 733 randomly assigned patients, 8 patients in the EC/T group and 11 patients in the Edd/Tdd/CMF group did not receive any study treatment. Of 356 patients randomly assigned to receive darbepoetin alfa, 318 received the treatment. Details of dose reductions and treatment discontinuation are reported in another publication [3]. DFS and OS At a median follow-up of 43.5 months, 196 recurrences including deaths without progression were observed. In the
original article EC/T group, 104 events were observed compared with 92 events in the Edd/Tdd/CMF group, corresponding to an estimated 3-year DFS of 75.8% and 78.8%, respectively [HR 1.14; 95% confidence interval (CI) 0.85–1.52; two-sided P = 0.37 (one-sided P = 0.19)]. In the darbepoetin alfa group, 106 events were observed compared with 90 events in the group without darbepoetin alfa, corresponding to an estimated 3-year DFS of 74.3% and 80.0%, respectively (HR 1.31; 95% CI 0.99– 1.74; P = 0.061) (Figure 1A and B). By the same median follow-up, 107 deaths of any reason occurred. In the EC/T group, 60 deaths occurred and 47 in the Edd/Tdd/CMF group, corresponding to 3-year OS probabilities of 88.4% and 91.5%, respectively (HR 1.26; 95% CI 0.86–1.85; P = 0.237). In the darbepoetin alfa group, 59 deaths and in the group without darbepoetin alfa 48 deaths occurred, corresponding to an OS of 88.0% and 91.8%, respectively (HR 1.33; 95% CI 0.91–1.95; P = 0.139) (Figure 1C and D). Downloaded from http://annonc.oxfordjournals.org/ by guest on August 26, 2015
Figure 3. Effects of darbepoetin alfa in different risk subgroups. (A) Effect on disease-free survival (DFS); (B) effect on overall survival (OS). ER, estrogen receptor; PgR, progesterone receptor; pCR, pathologically documented complete response; HR, hazard ratio; CI, confidence interval.
Volume 22 | No. 9 | September 2011
doi:10.1093/annonc/mdq713 | 2003
original article
Annals of Oncology
In an unplanned subgroup analysis, the impact of darbepoetin alfa on DFS and OS was investigated (Figure 3). Patients with either grade 3 (Bloom and Richardson, modified by Elston and Ellis) tumor or a tumor size ‡4 cm had significantly worse DFS when treated with darbepoetin alfa. This effect on OS was only significant for grade 3 tumors.
univariable and multivariable analyses Nodal status at the time of surgery and tumor grade, but not the other baseline factors, as well as pCR were significant and independent prognostic factors for DFS in uni- and multivariable analyses (Table 2). pCR was only a significant prognostic factor for DFS in univariable analysis but did not reach statistical significance in the multivariable analysis (P = 0.070). Yet, recurrence risk was almost reduced by 50% if patients reached a pCR independent of the allocated treatment. Treatment with darbepoetin alfa resulted in a numerical increase in relapses that was statistically not significant (104 with versus 88 without darbepoetin alfa; HR 1.31; P = 0.062). Tumor grade and histopathological nodal status after chemotherapy were significant independent prognostic factors for OS as well (Table 3). Patients with hormone receptorpositive disease had a higher probability to survive, but the effect was statistically not independent of other factors. Patients who reached a pCR showed an almost 80% risk reduction
Table 2. Association of prognostic factors, therapy, and pCR with disease-free survival Events Chemotherapy EC/T Edd/Tdd/CMF Darbepoetin alfa + 2 Age (years) <50 ‡50 Initial tumor size (cm) <4 ‡4 Grading 3 1–2 ER/PgR status Negative/negative Other Nodal status at surgery ypN1–3 ypN0 pCR 2 +
%
103 89
28 25
104 88
30 24
108 84
29 25
119 61
25 28
73 49
34 21
42 79
29 25
113 60
41 16
175 17
29 15
Univariable analysis 95% CI Hazard ratioa
P value
Multivariable analysis (n = 380) Hazard ratioa 95% CI
P value
1.14
0.85–1.52
0.3717
1.23
0.82–1.85
0.3112
1.31
0.99–1.74
0.0621
1.40
0.94–2.1
0.1022
1.16
0.88–1.55
0.2878
1.48
0.98–2.24
0.0639
0.88
0.65–1.20
0.4113
1.10
0.72–1.69
0.6635
1.96
1.36–2.81
0.0003
2.29
1.45–3.60
0.0004
1.27
0.87–1.85
0.2133
1.41
0.86–2.31
0.1707
3.09
2.26–4.26
<0.0001
3.50
2.23–5.51
<0.0001
2.27
1.38–3.73
0.0013
1.86
0.95–3.63
0.0696
pCR, pathologically documented complete response; CI, confidence interval; EC/T, epirubicin plus cyclophosphamide followed by paclitaxel; Edd/Tdd/CMF, epirubicin followed by paclitaxel followed by combination chemotherapy with cyclophosphamide, methotrexate, and fluorouracil; ER, estrogen receptor; PgR, progesterone receptor.
2004 | Untch et al.
Volume 22 | No. 9 | September 2011
Downloaded from http://annonc.oxfordjournals.org/ by guest on August 26, 2015
Overall DFS and OS results did not considerably change when adjusted for baseline factors (age, hormone receptor status, grading, and tumor size) (supplemental Tables S1 and S2, available at Annals of Oncology online). Overall, 71 local relapses were observed. There was no significant difference in the incidence of local relapses between the two chemotherapy groups as well as between patients treated with or without darbepoetin alfa. Patients who experienced a pCR had a significantly better DFS compared with those without achieving a pCR (estimated 3-year probabilities: 89.2% versus 74.9%; HR 2.27; 95% CI 1.38–3.73; P = 0.0013) (Figure 2A). In the EC/T group, patients experiencing a pCR had a better DFS compared with patients not experiencing a pCR (HR 3.66; 95% CI 1.49–8.98; P = 0.005), which was not statistically different in the Edd/Tdd/CMF arm (HR 1.624; 95% CI 0.884–2.984; P = 0.118) (Figure 2B). These results are mirrored in the OS analysis. Again, patients experiencing a pCR at the time of surgery had significantly better OS than those who did not reach a pCR (96.5% versus 88.6%; HR 3.54; 95% CI 1.55–8.07; P = 0.0026) (Figure 2C and Table 3). The better outcome observed for patients who achieved a pCR was also observed in the subgroups treated with or without darbepoetin alfa (DFS: HR 2.38; 95% CI 1.20–4.71; P = 0.013 and OS: HR 2.13; 95% CI 1.03–4.41; P = 0.041; supplemental Figure S1, available at Annals of Oncology online).
original article
Annals of Oncology
Table 3. Association of prognostic factors, therapy, and pCR with overall survival Events
60 47
17 13
59 48
17 13
59 48
16 14
40 19
19 8
27 35
19 11
62 38
13 17
68 26
25 7
101 6
17 5
Univariable analysis 95% CI Hazard ratioa
P value
Multivariable analysis (n = 380) Hazard ratioa 95% CI
P value
1.26
0.86–1.85
0.2372
1.41
0.79–2.51
0.2469
1.33
0.91–1.95
0.1387
1.42
0.81–2.52
0.2253
1.09
0.75–1.60
0.6451
1.45
0.81–2.62
0.2125
2.70
1.56–4.67
0.0004
2.66
1.38–5.13
0.0035
1.90
1.15–3.13
0.0127
1.79
0.93–3.46
0.0837
0.73
0.49–1.10
0.1307
0.94
0.52–1.71
0.8401
3.96
2.52–6.23
<0.0001
3.96
2.05–7.66
<0.0001
3.54
1.55–8.08
0.0026
2.12
0.77–5.81
0.1452
pCR, pathologically documented complete response; CI, confidence interval; EC/T, epirubicin plus cyclophosphamide followed by paclitaxel; Edd/Tdd/CMF, epirubicin followed by paclitaxel followed by combination chemotherapy with cyclophosphamide, methotrexate, and fluorouracil; ER, estrogen receptor; PgR, progesterone receptor.
regarding death, but this effect was also not independent in multivariable analysis.
discussion The PREPARE trial is the first study to investigate a dose-dense, dose-intensified neoadjuvant regimen in comparison with standard epirubicin plus cyclophosphamide followed by paclitaxel with or without an erythropoiesis-stimulating agent (ESA). Our study showed a nonsignificant improvement of DFS and OS for the dose-intensified anthracycline–taxane-containing chemotherapy, although the pCR increases significantly from 12% to 18% (P = 0.0176) [3]. Administration of darbepoetin alfa showed a trend toward a worse DFS, without showing a relevant effect on the clinical and pathohistological response. The results of the PREPARE trial are consistent with those of the NSABP-B27 and the GeparDuo study [4], which showed an improved pCR rate without a significant effect on long-term outcome. Due to the small sample size of a single trial, such outcome differences can probably only be demonstrated in a meta-analysis [5, 6]. The second reason for the observed lack of a significant survival benefit might be that pCR has a different meaning for the various subgroups of breast cancer. Patients with a rather favourable prognosis like the endocrineresponsive tumors have a low probability to achieve a pCR, but their overall outcome is still excellent.
Volume 22 | No. 9 | September 2011
In general, patients who reach a pCR have a significantly improved DFS and OS compared with those without pCR. In the PREPARE trial, patients treated with dose-dense, dose-intensified chemotherapy who achieved a pCR had no improved DFS and OS compared with the standard arm. Our previous study (AGO-1) investigated a similar approach of a dose-dense Edd/Tdd sequence but with CMF given postoperatively. Compared with standard treatments, the dosedense arm of both studies, AGO- 1 and PREPARE, resulted in an increased pCR rate, despite different treatment durations. A significant improvement of DFS, however, was only seen in the AGO-1 trial. One hypothesis of the observed outcome differences between the AGO-1 and the PREPARE trial could be the use of darbepoetin alfa, which did not influence the pCR rate in PREPARE but had a negative impact on survival. In a recent meta-analysis with a different regimen, Erythropoietin stimulating agent (ESA) use was associated with worsened OS and higher mortality during active study treatment [7]. In PREPARE, there is a trend toward a worse DFS in the darbepoetin alfa arm compared with the darbepoetin free arm. Moreover, the absolute DFS difference in the dose-dense arm between patients treated with and without darbepoetin alfa is larger than the difference between the two chemotherapy regimens. An unplanned subgroup analysis of our study revealed poor prognostic factors to be associated with significantly decreased DFS and OS in patients who received darbepoetin.
doi:10.1093/annonc/mdq713 | 2005
Downloaded from http://annonc.oxfordjournals.org/ by guest on August 26, 2015
Chemotherapy EC/T Edd/Tdd/CMF Darbepoetin alfa + 2 Age (years) <50 ‡50 Grading 3 1–2 ER/PgR status Negative/negative Other Tumor size (cm) <4 ‡4 Nodal status ypN1–3 ypN0 pCR 2 +
%
original article
acknowledgements We thank all the patients who participated in the trial and consented to give their data. We thank all the investigators for treating the patients and reporting their data.
funding We thank Amgen, Germany, and Bristol Myers Squibb, Germany, for financial trial support and medication.
2006 | Untch et al.
disclosure PAF is a member of the speakers’ bureau for Novartis, AstraZeneca and Roche; NH received honoraria for consulting and lectures from BMS and Roche; GvM received research funding from BMS and Amgen. All other authors declare no conflict of interest.
references 1. Bear HD, Anderson S, Smith RE et al. Sequential preoperative or postoperative docetaxel added to preoperative doxorubicin plus cyclophosphamide for operable breast cancer. National Surgical Adjuvant Breast and Bowel Project Protocol B-27. J Clin Oncol 2006; 24: 2019–2027. 2. Untch M, Mo¨bus V, Kuhn W et al. Intensive dose-dense chemotherapy compared with conventionally scheduled preoperative chemotherapy for high-risk primary breast cancer. J Clin Oncol 2009; 27: 1–9. 3. Untch M, Fasching PA, Konecny GE et al. Prepare trial: a randomized phase III trial comparing preoperative, dose-dense, dose-intensified chemotherapy with epirubicin, paclitaxel and CMF versus a standard-dosed epirubicin/cyclophosphamide followed by paclitaxel +/2 darbepoetin alfa in primary breast cancer - results at the time of surgery. Ann Oncol 2011; 22: 1988–1998. 4. Kaufmann M, Eiermann W, Schu¨tte M et al. Long-term results from the neoadjuvant GeparDuo trial: a randomized, multicenter, open phase III study comparing a dose-intensified 8-week schedule of doxorubicin hydrochloride and docetaxel (ADoc) with a sequential 24-week schedule of doxorubicin hydrochloride/cyclophosphamide followed by docetaxel (AC-Doc) regimen as preoperative therapy (NACT) in patients (pts) with operable breast cancer (BC). J Clin Oncol 2010; 28: 15S (Abstr 537). 5. von Minckwitz G, Kaufmann M, Ku¨mmel S et al. Integrated meta-analysis on 6402 patients with early breast cancer receiving neoadjuvant anthracyclinetaxane +/- trastuzumab containing chemotherapy. Cancer Res 2009; 69 (Suppl): #79. 6. von Minckwitz G, Untch M, Nu¨esch E et al. Impact of treatment characteristics on response of different breast cancer phenotypes: pooled analysis of the German neo-adjuvant chemotherapy trials. Breast Cancer Res Treat 2011; 125: 145–156. 7. Bohlius J, Schmidlin K, Brillant C et al. Recombinant human erythropoiesisstimulating agents and mortality in patients with cancer: a meta-analysis of randomised trials. Lancet 2009; 373: 1532–1542. 8. Mo¨bus V, Jackisch C, Lu¨ck HJ et al. Intense dose-dense sequential chemotherapy with epirubicin, paclitaxel and cyclophosphamide compared with conventionally scheduled chemotherapy in high-risk primary breast cancer (4+ LN): mature results of an AGO-phase-III study. J Clin Oncol 2010; 28: 2874–2880. 9. Leyland-Jones B, Semiglazov V, Pawlicki M. Maintaining normal hemoglobin levels with epoetin alfa in mainly nonanemic patients with metastatic breast cancer receiving first-line chemotherapy: a survival study. J Clin Oncol 2005; 23: 5960–5972. 10. Bohlius J, Wilson J, Seidenfeld J et al. Recombinant human erythropoietins and cancer patients: updated meta-analysis of 57 studies including 9353 patients. J Natl Cancer Inst 2006; 98: 708–714. 11. Ludwig H, Crawford J, Osterborg A et al. Pooled analysis of individual patient-level data from all randomized, double blind, placebo-controlled trials of darbepoetin alfa in the treatment of patients with chemotherapy-induced anemia. J Clin Oncol 2009; 27: 2838–2847. 12. Untch M, Stoeckl D, Konecny G et al. A multicenter phase II study of preoperative epirubicin, cyclophosphamide (EC) followed by paclitaxel (P) plus trastuzumab (T) in Her2 positive primary breast cancer. San Antonio Breast Cancer Symposium 2005 (Abstr 1064). 13. Untch M, Fasching AP, Konecny GE et al. Pathological complete response after neoadjuvant chemotherapy + trastuzumab treatment predicts survival and detects a patient subgroup at high need for improvement of anti-HER2 therapy. Three year median follow up data of the TECHNO trial. J Clin Oncol 2011; in press .
Volume 22 | No. 9 | September 2011
Downloaded from http://annonc.oxfordjournals.org/ by guest on August 26, 2015
ESA administration, for instance, resulted in patients with tumors >4 cm in an absolute decrease of DFS of >20%. In the dose-dense adjuvant trial by Mo¨bus et al. [8], which randomized for ESA as well, ESA administration did not result in a negative effect on long-term outcome. Thinking of the differences between the published studies and the diverse effect of adding ESA, we hypothesize that the effect of ESA might be dependent on the presence of a relevant tumor load. In the PREPARE trial as well as the BEST study in metastatic breast cancer patients [9], a substantial tumor load is present. Trials in the adjuvant setting with no measurable tumor load did not show an adverse effect of ESA on the prognosis of the patients [10, 11]. Further functional analyses need to specify the mechanism of action and interaction between tumor burden, ESA, and cytotoxic treatment. Our study has several strengths and limitations. To the best of our knowledge, the PREPARE trial is the only one carried out so far in the neoadjuvant setting, evaluating in a prospectively randomized fashion the addition of ESA in chemotherapy-naive patients. Furthermore, the study was designed to detect a difference in DFS but not in the pCR rate. However, the study did not meet the assumptions under which the power calculations were made. DFS in the standard arm was found to be better than assumed, and the sample size calculation was done on the basis of a one-sided test. The pCR did show a significant effect and the DFS did not. These three facts imply that the study might not have had enough power to refute the null hypothesis. As far as patient selection and data capture are concerned, two points must be noted: the vast majority of tumors were HER2 negative because the TECHNO trial for patients with HER2positive disease was recruiting simultaneously and a large number of study sites took part in both studies [12, 13]. Furthermore, only 380 patients could be included into the multivariable analysis because unfortunately grading, hormone receptor status as well as clinical nodal involvement were not captured at baseline for all patients. In conclusion, the PREPARE study confirms that without targeted treatment, the pCR rate of anthracycline–taxanecontaining chemotherapy does not exceed 20% despite increasing the dose intensity. Patients reaching a pCR have an excellent prognosis but the treatment, which increases the pCR rate, does not automatically lead to a survival benefit. Patients should not be treated with ESA in the neoadjuvant setting under the assumption of better tumor oxygenation because a negative influence of darbepoetin alfa on DFS cannot completely be ruled out.
Annals of Oncology
Annals of Oncology 22: 1999–2006, 2011 doi:10.1093/annonc/mdq713 Published online 7 March 2011
PREPARE trial: a randomized phase III trial comparing preoperative, dose-dense, dose-intensified chemotherapy with epirubicin, paclitaxel, and CMF versus a standard-dosed epirubicin–cyclophosphamide followed by paclitaxel with or without darbepoetin alfa in primary breast cancer—outcome on prognosis
Received 31 August 2010; revised 10 November 2010; accepted 12 November 2010
Background: The objective of this study was to compare the effect of dose-intensified neoadjuvant chemotherapy with that of standard epirubicin plus cyclophosphamide followed by paclitaxel in combination with or without darbepoetin on survival in primary breast cancer. Patients and methods: A total of 733 patients received either four cycles of neoadjuvant epirubicin 90 mg/m2 plus cyclophosphamide 600 mg/m2 every 3 weeks followed by four cycles of paclitaxel 175 mg/m2 every 3 weeks (EC/T), or three cycles of epirubicin 150 mg/m2 every 2 weeks followed by three cycles of paclitaxel 225 mg/m2 every 2 weeks followed by three cycles of combination chemotherapy with cyclophosphamide, methotrexate, and fluorouracil (Edd/Tdd/CMF). The patients were randomly assigned to receive darbepoetin or none. The primary objective was to demonstrate a superior disease-free survival (DFS) of Edd/Tdd/CMF compared with EC/T. Results: Estimated 3-year DFS was 75.8% with EC/T versus 78.8% with Edd/Tdd/CMF [hazard ratio (HR) 1.14; P = 0.37] and overall survival (OS) 88.4% versus 91.5% (HR 1.26; P = 0.237). Three-year DFS was 74.3% with darbepoetin versus 80.0% without (HR 1.31; P = 0.061) and OS 88.0% versus 91.8% (HR 1.33; P = 0.139). Patients with a pathologically documented complete response [pathological complete response (pCR)] had a significantly better DFS compared with those without achieving a pCR (estimated 3-year DFS: 89.2% versus 74.9%; HR 2.27; P = 0.001). Conclusion: Neoadjuvant dose-intensified chemotherapy compared with standard chemotherapy did not improve DFS, whereas the addition of darbepoetin might have detrimental effects on DFS. Key words: darbepoetin alfa, neoadjuvant dose-dense chemotherapy, neoadjuvant dose-intensified therapy, primary breast cancer
introduction *Correspondence to: Prof. M. Untch, Department of Obstetrics and Gynecology, Helios Klinikum Berlin-Buch, Schwanebecker Chaussee 50, 13125 Berlin, Germany. Tel: +49-30-9401-53300; Fax: +49-30-9401-53309; E-mail: [email protected]
Neoadjuvant chemotherapy has become an acknowledged alternative to adjuvant chemotherapy not only for locally advanced breast cancer but also for those patients who are
ª The Author 2011. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: [email protected]
original article
1 Department of Obstetrics and Gynecology, Helios Klinikum Berlin-Buch, Berlin; 2German Breast Group, Neu-Isenburg, Germany; 3Division of Hematology and Oncology, David Geffen School of Medicine, University of California, Los Angeles, USA; 4Department of Obstetrics and Gynecology, St Barbara Hospital, Hamm; 5 Hematologic/Oncologic Practice, Wuppertal; 6University Women’s Hospital of Ulm, Ulm; 7Department of Obstetrics and Gynecology, Medical University of Hannover, Hannover; 8University Women’s Hospital of Freiburg, Freiburg; 9Klinikum, Kassel; 10Krankenhaus Bergisch, Gladbach; 11University Breast Center Franconia, Department of Gynecology and Obstetrics, University Hospital Erlangen, Erlangen; 12Hematologic/Oncologic Clinic Mu¨nchen, Munich; 13Department of Obstetrics and Gynecology, Technical University of Munich, Munich; 14Department of Obstetrics and Gynecology, University Hospital Hamburg-Eppendorf, Hamburg; 15University Women’s Hospital of Mainz, Mainz; 16Department of Obstetrics and Gynecology, Ludwig-Maximilians-University Großhadern, Munich; 17Department of Pathology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany
Downloaded from http://annonc.oxfordjournals.org/ by guest on August 26, 2015
M. Untch1*, G. von Minckwitz2, G. E. Konecny3, U. Conrad4, W. Fett5, C. Kurzeder6, H.-J. Lu¨ck7, E. Stickeler8, H. Urbaczyk9, B. Liedtke10, M. W. Beckmann11, C. Salat12, N. Harbeck13, V. Mu¨ller14, M. Schmidt15, S. Hasmu¨ller16, M. Lenhard16, V. Nekljudova2, A. Lebeau17, S. Loibl2 & P. A. Fasching11 on behalf of the Arbeitsgemeinschaft Gyna¨kologische Onkologie PREPARE investigators
original article
Annals of Oncology
preoperatively by CMF compared with a standard anthracycline–taxane-based combination as neoadjuvant treatment without postoperative chemotherapy. In a second randomization, the effect of darbepoetin alfa on anemia and long- as well as short-term outcome was investigated. The primary objective was to examine the DFS additionally to usual objectives like pCR rate and breast-conserving therapy [3].
patients and methods patients This is a phase III, multicenter, prospective, randomized, open-label clinical trial. Patients aged ‡18 and £65 years with histologically confirmed primary breast cancer by core biopsy were enrolled. The primary tumor had to be ‡2 cm based on either clinical or ultrasound measurement. Patients with inflammatory breast cancer were also included. Detailed inclusion and exclusion criteria are described elsewhere [3]. Approval was obtained from the ethics committees of all participating study sites and all patients gave written informed consent to take part in this trial.
treatment plan Eligible patients were randomly assigned to the control arm with epirubicin 90 mg/m2 plus cyclophosphamide 600 mg/m2 every 3 weeks for four cycles followed by paclitaxel 175 mg/m2 every 3 weeks for four cycles (EC/T) or to the experimental arm with sequential dose-dense,
Table 1. Disease-free survival (DFS) and overall survival (OS) estimated at 3 years: overall and according to treatment as well as prognostic subgroups n Overall 714 Darbepoetin 2 369 + 345 Age (years) <50 373 ‡50 341 Tumor grade 1 or 2 236 3 214 Steroid receptor statusa ER and/or PgR positive 310 ER and PgR negative 144 Tumor size (cm) <4 472 ‡4 221 Pathological lymph node status at Involved 277 Not involved 374 pCRa + 117 2 597
EC/T (n = 362) Edd/Tdd/CMF (n = 352) DFS (%) OS (%) DFS (%) OS (%)
Without darbepoetin (n = 369) DFS (%) OS (%)
With darbepoetin (n = 345) DFS (%) OS (%)
75.6
88.4
78.8
91.5
78.0
91.8
74.3
88.0
78.0 73.6
91.0 85.7
82.1 75.2
92.6 90.4
– –
– –
– –
– –
73.0 78.7
86.8 90.1
80.2 77.1
91.2 91.9
77.8 83.0
90.6 93.4
75.3 73.2
87.2 88.8
80.7 69.7
94.7 82.6
86.0 70.3
96.5 86.7
82.7 75.8
94.8 89.0
84.0 63.0
96.5 78.8
74.7 74.0
92.6 78.8
83.5 70.0
94.5 83.3
78.7 77.6
94.0 85.1
79.4 65.7
93.1 76.0
79.8 71.0 surgery 62.6 87.1
92.2 82.0
79.0 82.7
92.4 92.5
77.7 86.2
91.8 91.1
81.4 66.6
92.8 82.7
81.3 95.4
66.6 88.5
83.6 98.3
70.4 88.7
85.1 97.8
58.4 86.8
79.5 95.8
93.6a 72.9
97.9a 86.8
86.0 77.1
95.4 90.6
87.8a 78.4
96.6a 90.8
90.9a 71.1
96.3a 86.4
a
Statistically significant differences between the patients with and without pCR in EC/T, without darbepoetin and with darbepoetin. EC/T, epirubicin plus cyclophosphamide followed by paclitaxel; Edd/Tdd/CMF, epirubicin followed by paclitaxel followed by combination chemotherapy with cyclophosphamide, methotrexate, and fluorouracil; ER, estrogen receptor; PgR, progesterone receptor; pCR, pathologically documented complete response.
2000 | Untch et al.
Volume 22 | No. 9 | September 2011
Downloaded from http://annonc.oxfordjournals.org/ by guest on August 26, 2015
candidates for adjuvant chemotherapy. Combinations of anthracyclines and taxanes are widely used within this indication. Higher response rates turn into higher rates of breastconserving surgery. But it has not conclusively been proven that a higher rate of pathologically documented complete response [pathological complete response (pCR)] will result in an improved long-term outcome. However, pCR is a highly significant predictor for improved disease-free survival (DFS) and overall survival (OS) regardless of treatment [1]. Previously, we have shown that an increased pCR achieved by intensified preoperative chemotherapy was significantly associated with a better DFS and OS [2]. The previous AGO-1 study compared a preoperative dose-dense, dose-intensified, sequential administration of epirubicin and paclitaxel followed [Edd/Tdd surgery /CMF (combination chemotherapy with cyclophosphamide, methotrexate, and fluorouracil)] with a conventionally dosed combination therapy of epirubicin and paclitaxel, both regimens followed postoperatively by standard CMF (epirubicin-taxol /surgery /CMF). The dose-dense chemotherapy showed a significantly higher pCR rate [18% versus 10%; odds ratio (OR) 1.89; P = 0.008] and a significantly better DFS [hazard ratio (HR) 0.71; P = 0.011] and OS (HR 0.69; P = 0.041). The PREPARE (Preoperative Epirubicin, Paclitaxel, Darbepoetin) trial investigated a dose-dense schedule followed
original article
Annals of Oncology
Treatment was reinduced with 50% of the dose, if hemoglobin decreased again to £13.0 g/dl.
end points The primary end point of the PREPARE study was DFS. Secondary end points included the effect of chemotherapy on OS as well as the effects of darbepoetin alfa on DFS and OS. Short-term secondary efficacy end points including pCR and pathohistological lymph node status at surgery have been presented elsewhere [3].
statistical analysis The sample size of 720 patients for this study was chosen to provide 80% power to detect an improvement in DFS at 3 years from the standard therapy with a DFS of 70% to a DFS of 80% in patients receiving dosedense chemotherapy. This equals an HR of 1.4 with a type I error of alpha 5% using a one-sided test. Patients were randomly assigned in a 1 : 1 allocation to receive standard-dose or dose-intensified preoperative chemotherapy. Patients within each treatment arm were further randomly assigned in a 1 : 1 allocation to receive darbepoetin alfa or no darbepoetin alfa therapy.
Figure 1. Disease-free survival (DFS) and overall survival (OS) according to treatment. (A) DFS according to chemotherapy (one-sided log-rank test P = 0.185); (B) DFS according to darbepoetin alfa; (C) OS according to chemotherapy; (D) OS according to darbepoetin alfa. EC/T, epirubicin plus cyclophosphamide followed by paclitaxel; Edd/Tdd/CMF, epirubicin followed by paclitaxel followed by combination chemotherapy with cyclophosphamide, methotrexate, and fluorouracil.
Volume 22 | No. 9 | September 2011
doi:10.1093/annonc/mdq713 | 2001
Downloaded from http://annonc.oxfordjournals.org/ by guest on August 26, 2015
dose-intensified treatment with epirubicin 150 mg/m2 every 2 weeks for three cycles followed by paclitaxel 225 mg/m2 every 2 weeks for three cycles followed by CMF (cyclophosphamide 500 mg/m2, methotrexate 40 mg/m2, and 5-fluorouracil 600 mg/m2) on days 1 and 8 every 4 weeks for three cycles (Edd/Tdd/CMF). In addition, the patients were randomly assigned to receive darbepoetin alfa 4.5 lg/kg body weight every 2 weeks starting with the first dose of epirubicin (day 1) until 14 days after the last dose of paclitaxel and 200 mg oral iron (Fe2+) daily or no darbepoetin [3]. Patients in the Edd/Tdd/CMF arm received pegfilgrastim 6 mg s.c. on day 2 of cycles one to six (Edd/Tdd). Primary prophylactic use of pegfilgrastim during CMF treatment or in the EC/T arm was not mandatory. If a patient had leucopenia for ‡3 days, fever >38.5C, or an active antibiotic-requiring infection, filgrastim (5 lg/kg body weight daily) was administered until recovery. In all consecutive cycles, pegfilgrastim (6 mg) was used prophylactically on day 2. All patients received standard supportive therapy. In the case of tumor progression during chemotherapy, study treatment was discontinued and further treatment was at the discretion of the investigator. To achieve the target hemoglobin level between 12.5 and 13 g/dl for patients receiving darbepoetin alfa, the dose was doubled if the increase was <1 g/dl during the first 4 weeks or discontinued if 14 g/dl was exceeded.
original article The statistical analyses of efficacy end points were carried out on the full analysis set, which included all patients who were randomly assigned to the chemotherapy treatment. Patients who never started therapy were excluded from this analysis. DFS was defined as the time from informed consent to first documentation of relapse or death due to any cause. OS was the time from the date of informed consent to the date of death due to any cause. Local DFS was defined as time in weeks between the date of signing informed consent and date of local recurrence. Patients with no local recurrence reported were censored at the date of the last contact. For univariable analysis, the log-rank test and Cox proportional hazards models were used for group comparisons, and Kaplan–Meier curves were used to estimate DFS and OS probabilities. For multivariable analysis, Cox proportional hazards models for adjusting survival end points were used; full models were fit without any selection of variables. Adjustments were made for age, hormone receptor status, clinical tumor size and nodal status, grade, chemotherapy arm, darbepoetin alfa application, and pCR. Hormone receptor status and grading were captured via the histological reports. All P values are two-sided; except for the primary end point
Annals of Oncology
DFS, a one-sided test was assumed but the two-sided P value is also provided. All calculations were done with SAS version 9.2 under Enterprise Guide 4.1.
results baseline characteristics This trial was conducted in 78 centers in Germany. Between June 2002 and February 2005, 736 patients were enrolled of whom 733 were randomly assigned, 370 in the EC/T group and 363 in the Edd/Tdd/CMF group. Three patients were not randomly assigned due to major inclusion criteria violation. Additionally, the patients were randomly assigned to receive darbepoetin alfa treatment (n = 356) or no darbepoetin treatment (control n = 377). A total of 714 patients who received at least one dose of chemotherapy, 362 receiving EC/T and 352 receiving Edd/Tdd/CMF, were included in this analysis. Downloaded from http://annonc.oxfordjournals.org/ by guest on August 26, 2015
Figure 2. Survival analyses by pCR. (A) Disease-free survival in the whole population by pCR; (B) disease-free survival by pCR stratified according to chemotherapy; (C) overall survival in the whole population by pCR; (D) overall survival by pCR stratified according to chemotherapy. pCR, pathologically documented complete response; EC/T, epirubicin plus cyclophosphamide followed by paclitaxel; Edd/Tdd/CMF, epirubicin followed by paclitaxel followed by combination chemotherapy with cyclophosphamide, methotrexate, and fluorouracil.
2002 | Untch et al.
Volume 22 | No. 9 | September 2011
Annals of Oncology
The median age at randomization was 48 years (23–65 years). Fifty-eight (7.9%) patients had clinical T4 carcinoma including those with inflammatory disease. Baseline characteristics were comparable between the treatment arms (Table 1).
compliance with treatment The flow of patients through the study is illustrated in another publication [3]. Of the 733 randomly assigned patients, 8 patients in the EC/T group and 11 patients in the Edd/Tdd/CMF group did not receive any study treatment. Of 356 patients randomly assigned to receive darbepoetin alfa, 318 received the treatment. Details of dose reductions and treatment discontinuation are reported in another publication [3]. DFS and OS At a median follow-up of 43.5 months, 196 recurrences including deaths without progression were observed. In the
original article EC/T group, 104 events were observed compared with 92 events in the Edd/Tdd/CMF group, corresponding to an estimated 3-year DFS of 75.8% and 78.8%, respectively [HR 1.14; 95% confidence interval (CI) 0.85–1.52; two-sided P = 0.37 (one-sided P = 0.19)]. In the darbepoetin alfa group, 106 events were observed compared with 90 events in the group without darbepoetin alfa, corresponding to an estimated 3-year DFS of 74.3% and 80.0%, respectively (HR 1.31; 95% CI 0.99– 1.74; P = 0.061) (Figure 1A and B). By the same median follow-up, 107 deaths of any reason occurred. In the EC/T group, 60 deaths occurred and 47 in the Edd/Tdd/CMF group, corresponding to 3-year OS probabilities of 88.4% and 91.5%, respectively (HR 1.26; 95% CI 0.86–1.85; P = 0.237). In the darbepoetin alfa group, 59 deaths and in the group without darbepoetin alfa 48 deaths occurred, corresponding to an OS of 88.0% and 91.8%, respectively (HR 1.33; 95% CI 0.91–1.95; P = 0.139) (Figure 1C and D). Downloaded from http://annonc.oxfordjournals.org/ by guest on August 26, 2015
Figure 3. Effects of darbepoetin alfa in different risk subgroups. (A) Effect on disease-free survival (DFS); (B) effect on overall survival (OS). ER, estrogen receptor; PgR, progesterone receptor; pCR, pathologically documented complete response; HR, hazard ratio; CI, confidence interval.
Volume 22 | No. 9 | September 2011
doi:10.1093/annonc/mdq713 | 2003
original article
Annals of Oncology
In an unplanned subgroup analysis, the impact of darbepoetin alfa on DFS and OS was investigated (Figure 3). Patients with either grade 3 (Bloom and Richardson, modified by Elston and Ellis) tumor or a tumor size ‡4 cm had significantly worse DFS when treated with darbepoetin alfa. This effect on OS was only significant for grade 3 tumors.
univariable and multivariable analyses Nodal status at the time of surgery and tumor grade, but not the other baseline factors, as well as pCR were significant and independent prognostic factors for DFS in uni- and multivariable analyses (Table 2). pCR was only a significant prognostic factor for DFS in univariable analysis but did not reach statistical significance in the multivariable analysis (P = 0.070). Yet, recurrence risk was almost reduced by 50% if patients reached a pCR independent of the allocated treatment. Treatment with darbepoetin alfa resulted in a numerical increase in relapses that was statistically not significant (104 with versus 88 without darbepoetin alfa; HR 1.31; P = 0.062). Tumor grade and histopathological nodal status after chemotherapy were significant independent prognostic factors for OS as well (Table 3). Patients with hormone receptorpositive disease had a higher probability to survive, but the effect was statistically not independent of other factors. Patients who reached a pCR showed an almost 80% risk reduction
Table 2. Association of prognostic factors, therapy, and pCR with disease-free survival Events Chemotherapy EC/T Edd/Tdd/CMF Darbepoetin alfa + 2 Age (years) <50 ‡50 Initial tumor size (cm) <4 ‡4 Grading 3 1–2 ER/PgR status Negative/negative Other Nodal status at surgery ypN1–3 ypN0 pCR 2 +
%
103 89
28 25
104 88
30 24
108 84
29 25
119 61
25 28
73 49
34 21
42 79
29 25
113 60
41 16
175 17
29 15
Univariable analysis 95% CI Hazard ratioa
P value
Multivariable analysis (n = 380) Hazard ratioa 95% CI
P value
1.14
0.85–1.52
0.3717
1.23
0.82–1.85
0.3112
1.31
0.99–1.74
0.0621
1.40
0.94–2.1
0.1022
1.16
0.88–1.55
0.2878
1.48
0.98–2.24
0.0639
0.88
0.65–1.20
0.4113
1.10
0.72–1.69
0.6635
1.96
1.36–2.81
0.0003
2.29
1.45–3.60
0.0004
1.27
0.87–1.85
0.2133
1.41
0.86–2.31
0.1707
3.09
2.26–4.26
<0.0001
3.50
2.23–5.51
<0.0001
2.27
1.38–3.73
0.0013
1.86
0.95–3.63
0.0696
pCR, pathologically documented complete response; CI, confidence interval; EC/T, epirubicin plus cyclophosphamide followed by paclitaxel; Edd/Tdd/CMF, epirubicin followed by paclitaxel followed by combination chemotherapy with cyclophosphamide, methotrexate, and fluorouracil; ER, estrogen receptor; PgR, progesterone receptor.
2004 | Untch et al.
Volume 22 | No. 9 | September 2011
Downloaded from http://annonc.oxfordjournals.org/ by guest on August 26, 2015
Overall DFS and OS results did not considerably change when adjusted for baseline factors (age, hormone receptor status, grading, and tumor size) (supplemental Tables S1 and S2, available at Annals of Oncology online). Overall, 71 local relapses were observed. There was no significant difference in the incidence of local relapses between the two chemotherapy groups as well as between patients treated with or without darbepoetin alfa. Patients who experienced a pCR had a significantly better DFS compared with those without achieving a pCR (estimated 3-year probabilities: 89.2% versus 74.9%; HR 2.27; 95% CI 1.38–3.73; P = 0.0013) (Figure 2A). In the EC/T group, patients experiencing a pCR had a better DFS compared with patients not experiencing a pCR (HR 3.66; 95% CI 1.49–8.98; P = 0.005), which was not statistically different in the Edd/Tdd/CMF arm (HR 1.624; 95% CI 0.884–2.984; P = 0.118) (Figure 2B). These results are mirrored in the OS analysis. Again, patients experiencing a pCR at the time of surgery had significantly better OS than those who did not reach a pCR (96.5% versus 88.6%; HR 3.54; 95% CI 1.55–8.07; P = 0.0026) (Figure 2C and Table 3). The better outcome observed for patients who achieved a pCR was also observed in the subgroups treated with or without darbepoetin alfa (DFS: HR 2.38; 95% CI 1.20–4.71; P = 0.013 and OS: HR 2.13; 95% CI 1.03–4.41; P = 0.041; supplemental Figure S1, available at Annals of Oncology online).
original article
Annals of Oncology
Table 3. Association of prognostic factors, therapy, and pCR with overall survival Events
60 47
17 13
59 48
17 13
59 48
16 14
40 19
19 8
27 35
19 11
62 38
13 17
68 26
25 7
101 6
17 5
Univariable analysis 95% CI Hazard ratioa
P value
Multivariable analysis (n = 380) Hazard ratioa 95% CI
P value
1.26
0.86–1.85
0.2372
1.41
0.79–2.51
0.2469
1.33
0.91–1.95
0.1387
1.42
0.81–2.52
0.2253
1.09
0.75–1.60
0.6451
1.45
0.81–2.62
0.2125
2.70
1.56–4.67
0.0004
2.66
1.38–5.13
0.0035
1.90
1.15–3.13
0.0127
1.79
0.93–3.46
0.0837
0.73
0.49–1.10
0.1307
0.94
0.52–1.71
0.8401
3.96
2.52–6.23
<0.0001
3.96
2.05–7.66
<0.0001
3.54
1.55–8.08
0.0026
2.12
0.77–5.81
0.1452
pCR, pathologically documented complete response; CI, confidence interval; EC/T, epirubicin plus cyclophosphamide followed by paclitaxel; Edd/Tdd/CMF, epirubicin followed by paclitaxel followed by combination chemotherapy with cyclophosphamide, methotrexate, and fluorouracil; ER, estrogen receptor; PgR, progesterone receptor.
regarding death, but this effect was also not independent in multivariable analysis.
discussion The PREPARE trial is the first study to investigate a dose-dense, dose-intensified neoadjuvant regimen in comparison with standard epirubicin plus cyclophosphamide followed by paclitaxel with or without an erythropoiesis-stimulating agent (ESA). Our study showed a nonsignificant improvement of DFS and OS for the dose-intensified anthracycline–taxane-containing chemotherapy, although the pCR increases significantly from 12% to 18% (P = 0.0176) [3]. Administration of darbepoetin alfa showed a trend toward a worse DFS, without showing a relevant effect on the clinical and pathohistological response. The results of the PREPARE trial are consistent with those of the NSABP-B27 and the GeparDuo study [4], which showed an improved pCR rate without a significant effect on long-term outcome. Due to the small sample size of a single trial, such outcome differences can probably only be demonstrated in a meta-analysis [5, 6]. The second reason for the observed lack of a significant survival benefit might be that pCR has a different meaning for the various subgroups of breast cancer. Patients with a rather favourable prognosis like the endocrineresponsive tumors have a low probability to achieve a pCR, but their overall outcome is still excellent.
Volume 22 | No. 9 | September 2011
In general, patients who reach a pCR have a significantly improved DFS and OS compared with those without pCR. In the PREPARE trial, patients treated with dose-dense, dose-intensified chemotherapy who achieved a pCR had no improved DFS and OS compared with the standard arm. Our previous study (AGO-1) investigated a similar approach of a dose-dense Edd/Tdd sequence but with CMF given postoperatively. Compared with standard treatments, the dosedense arm of both studies, AGO- 1 and PREPARE, resulted in an increased pCR rate, despite different treatment durations. A significant improvement of DFS, however, was only seen in the AGO-1 trial. One hypothesis of the observed outcome differences between the AGO-1 and the PREPARE trial could be the use of darbepoetin alfa, which did not influence the pCR rate in PREPARE but had a negative impact on survival. In a recent meta-analysis with a different regimen, Erythropoietin stimulating agent (ESA) use was associated with worsened OS and higher mortality during active study treatment [7]. In PREPARE, there is a trend toward a worse DFS in the darbepoetin alfa arm compared with the darbepoetin free arm. Moreover, the absolute DFS difference in the dose-dense arm between patients treated with and without darbepoetin alfa is larger than the difference between the two chemotherapy regimens. An unplanned subgroup analysis of our study revealed poor prognostic factors to be associated with significantly decreased DFS and OS in patients who received darbepoetin.
doi:10.1093/annonc/mdq713 | 2005
Downloaded from http://annonc.oxfordjournals.org/ by guest on August 26, 2015
Chemotherapy EC/T Edd/Tdd/CMF Darbepoetin alfa + 2 Age (years) <50 ‡50 Grading 3 1–2 ER/PgR status Negative/negative Other Tumor size (cm) <4 ‡4 Nodal status ypN1–3 ypN0 pCR 2 +
%
original article
acknowledgements We thank all the patients who participated in the trial and consented to give their data. We thank all the investigators for treating the patients and reporting their data.
funding We thank Amgen, Germany, and Bristol Myers Squibb, Germany, for financial trial support and medication.
2006 | Untch et al.
disclosure PAF is a member of the speakers’ bureau for Novartis, AstraZeneca and Roche; NH received honoraria for consulting and lectures from BMS and Roche; GvM received research funding from BMS and Amgen. All other authors declare no conflict of interest.
references 1. Bear HD, Anderson S, Smith RE et al. Sequential preoperative or postoperative docetaxel added to preoperative doxorubicin plus cyclophosphamide for operable breast cancer. National Surgical Adjuvant Breast and Bowel Project Protocol B-27. J Clin Oncol 2006; 24: 2019–2027. 2. Untch M, Mo¨bus V, Kuhn W et al. Intensive dose-dense chemotherapy compared with conventionally scheduled preoperative chemotherapy for high-risk primary breast cancer. J Clin Oncol 2009; 27: 1–9. 3. Untch M, Fasching PA, Konecny GE et al. Prepare trial: a randomized phase III trial comparing preoperative, dose-dense, dose-intensified chemotherapy with epirubicin, paclitaxel and CMF versus a standard-dosed epirubicin/cyclophosphamide followed by paclitaxel +/2 darbepoetin alfa in primary breast cancer - results at the time of surgery. Ann Oncol 2011; 22: 1988–1998. 4. Kaufmann M, Eiermann W, Schu¨tte M et al. Long-term results from the neoadjuvant GeparDuo trial: a randomized, multicenter, open phase III study comparing a dose-intensified 8-week schedule of doxorubicin hydrochloride and docetaxel (ADoc) with a sequential 24-week schedule of doxorubicin hydrochloride/cyclophosphamide followed by docetaxel (AC-Doc) regimen as preoperative therapy (NACT) in patients (pts) with operable breast cancer (BC). J Clin Oncol 2010; 28: 15S (Abstr 537). 5. von Minckwitz G, Kaufmann M, Ku¨mmel S et al. Integrated meta-analysis on 6402 patients with early breast cancer receiving neoadjuvant anthracyclinetaxane +/- trastuzumab containing chemotherapy. Cancer Res 2009; 69 (Suppl): #79. 6. von Minckwitz G, Untch M, Nu¨esch E et al. Impact of treatment characteristics on response of different breast cancer phenotypes: pooled analysis of the German neo-adjuvant chemotherapy trials. Breast Cancer Res Treat 2011; 125: 145–156. 7. Bohlius J, Schmidlin K, Brillant C et al. Recombinant human erythropoiesisstimulating agents and mortality in patients with cancer: a meta-analysis of randomised trials. Lancet 2009; 373: 1532–1542. 8. Mo¨bus V, Jackisch C, Lu¨ck HJ et al. Intense dose-dense sequential chemotherapy with epirubicin, paclitaxel and cyclophosphamide compared with conventionally scheduled chemotherapy in high-risk primary breast cancer (4+ LN): mature results of an AGO-phase-III study. J Clin Oncol 2010; 28: 2874–2880. 9. Leyland-Jones B, Semiglazov V, Pawlicki M. Maintaining normal hemoglobin levels with epoetin alfa in mainly nonanemic patients with metastatic breast cancer receiving first-line chemotherapy: a survival study. J Clin Oncol 2005; 23: 5960–5972. 10. Bohlius J, Wilson J, Seidenfeld J et al. Recombinant human erythropoietins and cancer patients: updated meta-analysis of 57 studies including 9353 patients. J Natl Cancer Inst 2006; 98: 708–714. 11. Ludwig H, Crawford J, Osterborg A et al. Pooled analysis of individual patient-level data from all randomized, double blind, placebo-controlled trials of darbepoetin alfa in the treatment of patients with chemotherapy-induced anemia. J Clin Oncol 2009; 27: 2838–2847. 12. Untch M, Stoeckl D, Konecny G et al. A multicenter phase II study of preoperative epirubicin, cyclophosphamide (EC) followed by paclitaxel (P) plus trastuzumab (T) in Her2 positive primary breast cancer. San Antonio Breast Cancer Symposium 2005 (Abstr 1064). 13. Untch M, Fasching AP, Konecny GE et al. Pathological complete response after neoadjuvant chemotherapy + trastuzumab treatment predicts survival and detects a patient subgroup at high need for improvement of anti-HER2 therapy. Three year median follow up data of the TECHNO trial. J Clin Oncol 2011; in press .
Volume 22 | No. 9 | September 2011
Downloaded from http://annonc.oxfordjournals.org/ by guest on August 26, 2015
ESA administration, for instance, resulted in patients with tumors >4 cm in an absolute decrease of DFS of >20%. In the dose-dense adjuvant trial by Mo¨bus et al. [8], which randomized for ESA as well, ESA administration did not result in a negative effect on long-term outcome. Thinking of the differences between the published studies and the diverse effect of adding ESA, we hypothesize that the effect of ESA might be dependent on the presence of a relevant tumor load. In the PREPARE trial as well as the BEST study in metastatic breast cancer patients [9], a substantial tumor load is present. Trials in the adjuvant setting with no measurable tumor load did not show an adverse effect of ESA on the prognosis of the patients [10, 11]. Further functional analyses need to specify the mechanism of action and interaction between tumor burden, ESA, and cytotoxic treatment. Our study has several strengths and limitations. To the best of our knowledge, the PREPARE trial is the only one carried out so far in the neoadjuvant setting, evaluating in a prospectively randomized fashion the addition of ESA in chemotherapy-naive patients. Furthermore, the study was designed to detect a difference in DFS but not in the pCR rate. However, the study did not meet the assumptions under which the power calculations were made. DFS in the standard arm was found to be better than assumed, and the sample size calculation was done on the basis of a one-sided test. The pCR did show a significant effect and the DFS did not. These three facts imply that the study might not have had enough power to refute the null hypothesis. As far as patient selection and data capture are concerned, two points must be noted: the vast majority of tumors were HER2 negative because the TECHNO trial for patients with HER2positive disease was recruiting simultaneously and a large number of study sites took part in both studies [12, 13]. Furthermore, only 380 patients could be included into the multivariable analysis because unfortunately grading, hormone receptor status as well as clinical nodal involvement were not captured at baseline for all patients. In conclusion, the PREPARE study confirms that without targeted treatment, the pCR rate of anthracycline–taxanecontaining chemotherapy does not exceed 20% despite increasing the dose intensity. Patients reaching a pCR have an excellent prognosis but the treatment, which increases the pCR rate, does not automatically lead to a survival benefit. Patients should not be treated with ESA in the neoadjuvant setting under the assumption of better tumor oxygenation because a negative influence of darbepoetin alfa on DFS cannot completely be ruled out.
Annals of Oncology