- Email: [email protected]
Survival after neoadjuvant chemotherapy with or without bevacizumab or everolimus for HER2-negative primary breast cancer (GBG 44–GeparQuinto)†
original articles
Annals of Oncology
Annals of Oncology 25: 2363–2372, 2014 doi:10.1093/annonc/mdu455 Published online 15 September 2014
Survival after neoadjuvant chemotherapy with or without bevacizumab or everolimus for HER2-negative primary breast cancer (GBG 44–GeparQuinto)†
1 Headquarter, German Breast Group, Neu-Isenburg; 2Department of Gynaecology and Obstetrics, University Hospital, Frankfurt; 3Department of Gynaecology and Obstetrics, Klinikum Berlin-Buch, Berlin; 4Department of Gynaecology and Obstetrics, University Hospital, Kiel; 5Breast Center, Luisenkrankenhaus, Düsseldorf; 6 Department of Gynaecology and Obstetrics, University Hospital, Erlangen; 7Department of Medical Oncology, Chop GmbH, Frankfurt; 8Department of Gynaecology and Obstetrics, University Hospital, Magdeburg; 9Department of Gynaecology and Obstetrics, Henriettenstiftung, Hannover; 10Department of Gynaecology and Obstetrics, Praxisklinik, Berlin; 11Department of Gynaecology and Obstetrics, Rot-Kreuz-Klinikum, München; 12Department of Gynaecology and Obstetrics, University Hospital, Ulm; 13 Department of Gynaecology and Obstetrics, Sana-Klinikum, Offenbach; 14Department of Gynaecology and Obstetrics, St Johannes Hospital, Dortmund; 15Department of Gynaecology and Obstetrics, St Gertrauden-Hospital, Berlin; 16Department of Gynaecology and Obstetrics, Hospital, Rheinfelden; 17Department of Gynaecology and Obstetrics, University Hospital, Tübingen; 18Department of Gynaecology and Obstetrics, University Hospital, Rostock, Germany
Received 16 July 2014; revised 8 September 2014; accepted 9 September 2014
Background: The GeparQuinto study showed that adding bevacizumab to 24 weeks of anthracycline–taxane–based neoadjuvant chemotherapy increases pathological complete response ( pCR) rates overall and specifically in patients with triple-negative breast cancer (TNBC). No difference in pCR rate was observed for adding everolimus to paclitaxel in nonearly responding patients. Here, we present disease-free (DFS) and overall survival (OS) analyses. Patients and methods: Patients (n = 1948) with HER2-negative tumors of a median tumor size of 4 cm were randomly assigned to neoadjuvant treatment with epirubicin/cyclophosphamide followed by docetaxel (EC-T) with or without eight infusions of bevacizumab every 3 weeks before surgery. Patients without clinical response to EC ± Bevacizumab were randomized to 12 weekly cycles paclitaxel with or without everolimus 5 mg/day. To detect a hazard ratio (HR) of 0.75 (α = 0.05, β = 0.8) 379 events had to be observed in the bevacizumab arms. Results: With a median follow-up of 3.8 years, 3-year DFS was 80.8% and 3-year OS was 89.7%. Outcome was not different for patients receiving bevacizumab (HR 1.03; P = 0.784 for DFS and HR 0.974; P = 0.842 for OS) compared with patients receiving chemotherapy alone. Patients with TNBC similarly showed no improvement in DFS (HR = 0.99; P = 0.941) and OS (HR = 1.02; P = 0.891) when treated with bevacizumab. No other predefined subgroup (HR+/HER2−; locally advanced (cT4 or cN3) or not; cT1–3 or cT4; pCR or not) showed a significant benefit. No difference in DFS (HR 0.997; P = 0.987) and OS (HR 1.11; P = 0.658) was observed for nonearly responding patients receiving paclitaxel with or without everolimus overall as well as in subgroups. Conclusions: Long-term results, in opposite to the results of pCR, do not support the neoadjuvant use of bevacizumab in addition to an anthracycline–taxane-based chemotherapy or everolimus in addition to paclitaxel for nonearly responding patients. Clinical trial number: NCT 00567554, www.clinicaltrials.gov. Key words: bevacizumab, everolimus, neoadjuvant chemotherapy, disease-free survival, overall survival
*Correspondence to: Dr Gunter von Minckwitz, German Breast Group and Department of Gynaecology and Obstetrics, University Hospital, Frankfurt, c/o GBG Forschungs GmbH, Martin-Behaim-Straße 12, 63263 Neu-Isenburg, Germany. Tel: +49-69-6102-798740; Fax: +49-69-6102-7987440; E-mail: [email protected] † Part of results will be presented at the 37th San Antonio Breast Cancer Symposium, 5–8 December 2014, San Antonio, TX, USA.
introduction Anthracycline-taxane-based neoadjuvant chemotherapy (NACT) of 4–6 months is a valid option in patients with operable breast cancer [1]. A unique aspect of NACT is the opportunity to monitor tumor response to treatment and to individualize
© The Author 2014. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: [email protected].
Downloaded from http://annonc.oxfordjournals.org/ at University of California, San Francisco on December 6, 2014
G. von Minckwitz1,2*, S. Loibl1, M. Untch3, H. Eidtmann4, M. Rezai5, P. A. Fasching6, H. Tesch7, H. Eggemann8, I. Schrader9, K. Kittel10, C. Hanusch11, J. Huober12, C. Solbach2, C. Jackisch13, G. Kunz14, J. U. Blohmer15, M. Hauschild16, T. Fehm17, V. Nekljudova1, B. Gerber18 & for the GBG/AGO-B study groups
original articles
Annals of Oncology
treatment according to this interim response [2]. Patients with a complete or partial interim response are more treatment sensitive and therefore more likely to achieve a pathological complete response (pCR) at surgery [3]; however, patients without an
interim response are considered chemotherapy resistant as they rarely achieve a pCR even after a switch to a noncross-resistant chemotherapy regimen [4]. Therefore, separate treatment strategies might be developed for these patient populations.
Centrally confirmed eligibility of HER2 negative breast cancer n = 1948
Randomized to EC + Bevacizumab n = 974
Started with EC n = 969
Started with EC outside of GeparQuinto study and showed no response n = 32
Started with EC + Bevacizumab n = 956
No response n = 221
Patients with no response to EC +/- Bevacizumab n = 403
No response n = 150
Randomized to Paclitaxel n = 201
Started with Docetaxel + Bevaciczumab n = 969
Started with Docetaxel n = 969
Randomized to Paclitaxel + Everolimus n = 202
Started with Paclitaxel n = 198
Started with Paclitaxel + Everolimus n = 197
Outcome n = 58 DFS events n = 36 Death
Outcome n = 57 DFS events n = 39 Death
n = 74 n = 124
n = 72 n = 125
Outcome n = 195 DFS events n = 118 Death
Outcome n = 202 DFS events n = 116 Death
= intent to treat populations Figure 1. Consort statement: flow of patients throughout the two parts of the study.
| von Minckwitz et al.
Volume 25 | No. 12 | December 2014
Downloaded from http://annonc.oxfordjournals.org/ at University of California, San Francisco on December 6, 2014
Randomized to EC n = 974
original articles
Annals of Oncology
0.36; 95% CI 0.24–1.6; P = 0.34]. Mucosal inflammation, thrombocytopenia, neutropenia, infection and skin rash were more frequent when everolimus was added to paclitaxel [6]. As the required number of relapses or deaths has recently been reached for the HER2-negative cohort, we here report the corresponding survival analyses. Survival analyses of the patients with HER2-positive tumors are planned for 2015 when the required number of events is observed.
patients and methods objectives The primary efficacy end point for both randomized treatment comparisons of patients with HER2-negative breast cancer was pCR, which has been reported earlier together with the related toxicity [4, 5]. We here report secondary objectives of the study comparing survival of the treatment groups using disease-free (DFS), local recurrence-free (LRFS), locoregional recurrence-free (LRRFS), regional recurrence-free (RRFS), distant disease-free,
Table 1. Baseline characteristics of patients by randomized treatment group Characteristic
EC-T (N = 969) n
Age (years) <40 150 40–<50 395 50–<60 257 60+ 167 Median (range) 48 Clinical tumor stage cT4a-c 57 cT4d 60 cT1-3 852 Stage cT1-3 and cN0-2 (operable) 859 cT4 or cN3 (locally advanced) 110 Clinical nodal status cN0 391 cN1 542 Missing 36 Tumor type Ductal invasive 773 Lobular invasive 106 Other 89 Missing 1 Tumor grade I 43 II 507 III 412 Missing 7 Estrogen (ER)/progesterone (PgR) status ER−/PgR− 340 ER+ and/or PgR+ 629 Pretreatment with Bev Yes n.a. No n.a.
Volume 25 | No. 12 | December 2014
%/range
EC-T + Bev (N = 956) n %/range
15.5% 40.8% 26.5% 17.2% 24–78
154 347 286 169 49
16.1% 36.3% 29.9% 17.7% 21–75
26 62 66 44 51
13.1% 31.3% 33.3% 22.2% 27–75
18 70 62 47 50
9.1% 35.5% 31.5% 23.9% 28–76
5.9% 6.2% 87.9%
53 65 838
5.5% 6.8% 87.7%
17 16 165
8.6% 8.1% 83.3%
16 18 163
8.1% 9.1% 82.7%
88.6% 11.4%
843 113
88.2% 11.8%
167 31
84.3% 15.7%
167 30
84.8% 15.2%
41.9% 58.1%
376 554 26
40.4% 59.6%
85 107 6
44.3% 55.7%
79 114 4
40.9% 59.1%
79.9% 11.0% 9.1%
770 102 81 3
80.8% 10.7% 8.5%
158 22 18 0
79.8% 11.1% 9.1%
158 20 19 0
80.2% 10.2% 9.6%
4.5% 52.7% 42.8%
32 503 417 4
3.4% 52.8% 43.8%
9 118 70 1
4.6% 59.9% 35.5%
12 118 66 1
6.1% 60.2% 33.7%
35.1% 64.9%
323 633
33.8% 66.2%
56 142
28.3% 71.7%
53 144
26.9% 73.1%
124 74
62.6% 37.4%
125 72
63.5% 36.5%
n.a. n.a.
Pac (N = 198)
Pac + Eve (N = 197)
n
n
%/range
%/range
doi:10.1093/annonc/mdu455 |
Downloaded from http://annonc.oxfordjournals.org/ at University of California, San Francisco on December 6, 2014
The GeparQuinto phase III study investigated subtype-specific treatment approaches for newly diagnosed, untreated patients with HER2-negative breast cancer [5], HER2-negative breast cancer without response to four cycles of NACT [6] and HER2positive breast cancer [7]. This report focuses on patients with HER2-negative, tumors who were treated with sequential epirubicin/cyclophosphamide and docetaxel (EC-T) NACT and randomized to simultaneous treatment with bevacizumab or no additional therapy or were re-randomized to paclitaxel with or without everolimus if they did not respond to EC [5, 6]. pCR rates increased from 14.9% with EC-T alone to 18.4% with ECT and bevacizumab [odds ratio = 1.29; 95% confidence interval (CI) 1.02–1.65; P = 0.042] and the effect on pCR was more pronounced in patients with triple-negative breast cancer (TNBC; 27.9% and 39.3%; P = 0.003) [8]. Chemotherapy with bevacizumab was associated with a higher incidence of grade 3/4 toxicities [5]. Only 18 (4.6%) of 403 patients nonresponsive to EC achieved a pCR; 7 (3.6%) treated with paclitaxel and everolimus and 11 (5.6%) treated with paclitaxel alone [odds ratio (OR)
original articles
Annals of Oncology
100%
A
90% 80%
60% 50%
+ Censored Logrank P = 0.7837
40%
EC-T 195/969 events ECB-TB 202/956 events
30% 20% 10% 0% 0
12
24
36
48
60
72
EC-T
969
831
696
585
322
78
0
ECB-TB
956
844
700
579
339
77
0
60
72
DFS, months
B
100% 90% 80%
Proportion alive
70% 60% 50%
+ Censored Logrank P = 0.8422
40%
EC-T 118/969 events ECB-TB 116/956 events
30% 20% 10% 0% 0
12
2
6
8
EC-T
969
859
751
629
352
88
0
ECB-TB
956
863
749
651
372
84
0
OS, months Figure 2. Disease-free (A) and overall survival (B) after neoadjuvant anthracycline-taxane based chemotherapy with or without bevacizumab.
| von Minckwitz et al.
Volume 25 | No. 12 | December 2014
Downloaded from http://annonc.oxfordjournals.org/ at University of California, San Francisco on December 6, 2014
Proportion disease-free
70%
original articles
Annals of Oncology
Table 2. Overview on survival end points in both randomized cohorts End point
LRFS LLRFS RRFS DDFS CDFS (only TNBC) DFS OS
Estimated 3-year survival EC-T EC-T + Bevacizumab
HR (95% CI)
P
Estimated 3-year survival Pac Pac + Everolimus
97.3 92.6 95.4 84.3 92.6 81.5 88.7
0.99 (0.59–1.66) 0.99 (0.72–1.37) 0.88 (0.58–1.35) 1.02 (0.82–1.27) 0.95 (0.56–1.62) 1.03 (0.84–1.25) 0.97 (0.75–1.26)
0.978 0.973 0.570 0.853 0.855 0.784 0.842
94.7 87.6 91.4 75.4 n.a. 71.3 83.4
96.6 91.8 95.7 84.0 92.1 80.0 90.7
97.3 91.2 93.1 72.5 n.a. 69.3 81.4
HR (95% CI)
P
0.45 (0.14–1.47) 0.70 (0.36–1.35) 0.87 (0.40–1.89) 0.97 (0.66–1.43)
0.175 0.281 0.734 0.869
1.00 (0.69–1.44) 1.11 (0.70–1.74)
0.987 0.657
cerebral disease-free (CDFS; only in patients with triple-negative disease) and overall survival (OS) as end points.
selection of patients Female patients with previously untreated, unilateral or bilateral, histologically confirmed invasive, nonmetastatic breast carcinoma were enrolled in the study if they provided written informed consent. Patients with any stage of disease deemed to be appropriate for chemotherapy were eligible (e.g. locally advanced tumors with cT4 or cT3, hormone-receptor-negative tumors or hormone-receptor-positive tumors with cN+ (for cT2) or pNSLN+ (for cT1) disease). HER2 status of the tumor had to be negative by HercepTest™ (Dako-score 0 or 1+) or in situ hybridization (Dako-score 2+ and no gene amplification) by the local pathologist. Tumor lesions were required to have a palpable size of ≥2 cm or a sonographical size of ≥1 cm in maximum diameter and had to be measurable in two dimensions, preferably by sonography.
treatment All patients were scheduled to receive four cycles of EC (epirubicin 90 mg/ m2, cyclophosphamide 600 mg/m2, day 1, every 3 weeks) followed by four cycles of T (docetaxel 100 mg/m2, day 1, every 3 weeks). Patients were randomized to receive bevacizumab 15 mg/kg i.v. every 3 weeks starting on day 1 of the first EC cycle or no additional treatment. In patients without a sonographical or clinical response to four cycles of EC ± bevacizumab treatment was stopped if applicable and patients were randomized to weekly paclitaxel given on days 1, 8, 15, every 3 weeks with 80 mg/m2 over four cycles with or without everolimus given orally in a dose of 5 mg per day from day 13 after a dose-escalation starting from 2.5 mg every other day to 5 mg every day. In cases of tumor progression, study treatment was discontinued and further local and/or systemic treatment was permitted at the discretion of the investigator. Patients had to undergo surgery at least 28 days after the last chemotherapy or bevacizumab infusion. Recommended postoperative treatment consisted of radiotherapy to the ipsilateral remaining breast for all patients treated by breast conserving surgery and radiotherapy to the chest wall with or without regional nodes in case of mastectomy for patients with an initial tumor size >5 cm or clinically suspect axillary nodes before treatment or histologically involved nodes after neoadjuvant treatment. Endocrine treatment was given for at least 5 years in patients with HR-positive (≥10% stained cells for estrogen and/or progesterone receptor) tumors. No bevacizumab was given after surgery. Six monthly visits or telephone contacts were scheduled for at least 5 years to collect information on disease status.
Volume 25 | No. 12 | December 2014
assessment of end points DFS (invasive) was defined as time from randomization or histological diagnosis in those patients not being randomized for the bevacizumab question until any invasive locoregional, invasive contralateral or distant recurrence of breast cancer or any second primary invasive nonbreast cancer or death due to any cause [9]. Events counted for LRFS were any local, i.e. ipsilateral invasive or ductal in situ recurrence of the ipsilateral breast; for LRRFS any local or regional, i.e. local or regional lymph node recurrences or any invasive contralateral breast cancer; for RRFS any local, regional or chest wall recurrences; and for OS death of any cause. During NACT, local progression of the breast tumor was not considered as a relapse. Patients without event were censored at the date of last contact. Response to neoadjuvant treatment was assessed by physical examination, ultrasound and/or mammography and/or MRI before surgery as described previously [4]. Pathologic response was assessed locally by examining all surgically excised breast and axillary lymphatic tissue according to TNM classification. pCR was defined as pT0 ypN0. Pathology reports were centrally reviewed at GBG headquarters. End of treatment with the experimental compounds was defined as 4 weeks after last application. Thereafter, adverse events were not any longer recorded.
statistics The GeparQuinto study was an open label, randomized phase III trial with pCR as the primary end point. Allocation to chemotherapy with or without target treatment was carried out by central randomization with a 1 : 1 ratio each and a stratification according to participating site, ER/PgR status (ER or PgR positive versus ER and PgR negative), extent of disease (T4 or N3 versus T1–3 and N0–2) and pretreatment with bevacizumab or not (only for everolimus). Assumptions for sample size calculation have published previously [4, 5]. All analyses were carried out on an intent-to-treat approach including all patients having received one cycle EC for the bevacizumab question and all patients having received one course of paclitaxel for the everolimus question (Figure 1). Survival analysis for the bevacizumab randomization was planned after 379 events observed to show a hazard ratio (HR) of 0.75 between the treatment arms using a two-sided α of 0.05 and 1 − β of 0.2. As the number of patients randomized for the everolimus question was considered too low for confirmative results, it was planned to run this analysis at the same time with the bevacizumab analysis. Time-to-event outcome parameters were estimated using the Kaplan–Meier product-limit method and treatment groups were compared using the log-rank test. Univariate Cox proportional hazards models
doi:10.1093/annonc/mdu455 |
Downloaded from http://annonc.oxfordjournals.org/ at University of California, San Francisco on December 6, 2014
CDFS, cerebral metastasis disease-free survival; CI, confidence interval; DDFS, Distant disease-free survival; DFS, disease-free survival; EC, epirubicin/ cyclophosphamide; HR, hazard ratio; LRFS, local relapse-free survival; LRRFS, locoregional recurrence-free survival; n.a., not assessed; Pac, Paclitaxel; RRFS, regional recurrence-free survival (RRFS); T, taxane; TNBC, triple-negative breast cancer; OS, overall survival.
original articles A
Annals of Oncology
100% 90% 80%
60% 50%
+ Censored Logrank P = 0.9870
40%
Pw 58/198 events PwR 57/197 events
30% 20% 10% 0% Pw PwR
B
0
12
24
36
48
60
72
198 197
173 163
131 125
99 96 DFS, months
57 58
9 17
0 0
100% 90% 80% 70%
Proportion alive
60% 50%
+ Censored Logrank P = 0.6575
40%
Pw 36/198 events PwR 39/197 events
30% 20% 10% 0% Pw PwR
0
12
24
36
48
60
72
198 197
183 172
147 145
114 112
62 65
11 17
0 0
OS, months Figure 3. Disease-free (A) and overall survival (B) after neoadjuvant anthracycline-taxane based chemotherapy with or without everolimus in patients not responding to epirubicin/cyclophosphamide.
| von Minckwitz et al.
Volume 25 | No. 12 | December 2014
Downloaded from http://annonc.oxfordjournals.org/ at University of California, San Francisco on December 6, 2014
Proportion disease-free
70%
original articles
Annals of Oncology
A Subgroup
N
Hazard Ratio
Patients
(95% CI)
Overall
Test for Interaction
1925
1.03 (.844, 1.25)
.784
304 1621
1.05 (.649, 1.70) 1.02 (.819, 1.26)
.843 .881
1690 235
1.09 (.876, 1.35) .772 (.487, 1.22)
.442 .272
767 1096
1.01 (.681, 1.50) 1.07 (.850, 1.35)
.959 .557
1702 223
1.09 (.878, 1.36) .725 (.456, 1.15)
.428 .175
1713 208
1.06 (.860, 1.30) .752 (.365, 1.55)
.605 .440
1085 829
1.01 (.759, 1.36) 1.03 (.790, 1.35)
.921 .813
1262 663
1.10 (.821, 1.47) .990 (.757, 1.29)
.527 .941
320 1605
2.02 (.965, 4.22) .989 (.805, 1.22)
.062 .919
.849
.182
.791
.116
.342
.935
.610
.067
0.4 0.6 0.8 1 1.5 2 HR DFS better with ECB-TB
3
4
DFS better with EC-T
Figure 4. Disease-free survival by subgroups in patients randomized to chemotherapy with or without bevacizumab (A) and in patients randomized to paclitaxel with or without everolimus (B). were used to calculate HRs); 95% CIs are provided. Multivariate Cox model was used to adjust for relevant prespecified baseline characteristics. Univariate Cox model was carried out in subgroups and a Cox model including interaction between subgroup factor and therapy was used to test for heterogeneity across subgroups. All statistical tests were two-sided by default; the significance level was set to 0.05 unadjusted for multiple comparisons.
results Between November 2007 to June 2010, 1948 patients were registered in the HER2-negative part of the study and 1925 patients started with EC (Figure 1). No clinical response after EC treatment was observed in 371 patients who were therefore allocated for the everolimus question (nonresponding cohort). This number was lower than expected. Therefore, the study protocol was amended after the completion of recruitment for the bevacizumab question. Then patients, who received 4 cycles of EC before entering the protocol and who did not respond were allowed to be randomized for the everolimus questions. Despite this amendment, recruitment to complete the nonresponder arm was poor (recruitment of additional 32 patients) and it appeared impossible to recruit the required 566 patients. Therefore, accrual was prematurely closed in June 2011 with 403 patients
Volume 25 | No. 12 | December 2014
randomized into nonresponding cohort (Figure 1). After completion of chemotherapy and surgery, 390 relapses and 229 deaths overall and 112 relapses and 73 deaths in the nonresponding cohort were observed after a median follow-up of 3.8 years. Baseline characteristics of patients were well balanced between treatment groups as previously shown (Table 1) [4, 5]. There was no statistically significant difference neither in DFS nor OS for patients receiving bevacizumab or not together with or after anthracycline–taxane-based chemotherapy (P = 0.7837 for DFS and P = 0.8422 for OS). Two hundred two and 195 events occurred with and without bevacizumab (Figure 2, Table 2). The HR was 1.03 (95% CI 0.84–1.25) for DFS and 0.97 (95% CI 0.75–1.26) for OS, the adjusted HR from multivariate analysis was 1.11 (95% CI 0.91–1.36) for DFS and 1.06 (95% CI 0.82–1.38) for OS. Similarly, no difference in LRFS, LRRFS and RFS was observed between the bevacizumab treatment arms (Table 2). CDFS in patients with TNBC was not different between randomized arms (HR 0.95; 95% CI 0.56–1.62). Subgroup analysis (Figure 4A) could not identify a subgroup for which an effect of bevacizumab on DFS might be postulated. In opposite, patients with a pCR after neoadjuvant treatment showed a trend toward a higher risk for an event if they have received bevacizumab with a HR of 2.02 and a lower 95% CI
doi:10.1093/annonc/mdu455 |
Downloaded from http://annonc.oxfordjournals.org/ at University of California, San Francisco on December 6, 2014
Age Age<40 years Age>=40 years cT cT1–3 cT4 cN cN0 cN+ Extension of disease cT1–3, cNo–2 cT4 or cN3 Histological tumor type Non-lobular Lobular Grading Grase 1–2 Grase 3 Receptor status ER and/or PgR positive ER and PgR negative pCR yes no
P-Value
original articles
Annals of Oncology
B Subgroup
N
Hazard Ratio
Patients
(95% CI)
Overall
Test for Interaction
395
.997 (.691, 1.44)
.987
41 354
1.22 (.443, 3.38) .978 (.660, 1.45)
.697 .912
328 67
1.04 (.678, 1.60) .786 (.388, 1.59)
.847 .504
165 220
1.19 (.638, 2.21) .890 (.559, 1.42)
.589 .624
334 61
.997 (.653, 1.52) .928 (.442, 1.95)
.987 .843
353 42
.945 (.644, 1.39) 1.91 (.478, 7.66)
.773 .360
257 136
1.07 (.618, 1.86) 1.06 (.647, 1.73)
.805 .822
286 109
.853 (.507, 1.44) 1.33 (.795, 2.23)
.551 .276
249 146
.837 (.525, 1.33) 1.32 (.725, 2.41)
.455 .363
.675
.473
.484
.862
.324
.893
.262
.207
0.4 0.60.8 1 1.5 2 HR DFS better with PwR
3 4
6 8
DFS better with Pw
Fig. 4 Continued
bound of 0.965 marginally crossing 1. However, the test for interaction was not significant with P = 0.067). Similar results were obtained for OS (supplementary Figure S1A, available at Annals of Oncology online). Comparing DFS and OS in patients not responsive to EC and treated with paclitaxel with or without everolimus, no statistically significant difference was found (P = 0.9870 for DFS and P = 0.657 for OS) (Figure 3). The adjusted survival analysis revealed an HR of 1.13 (95% CI 0.77–1.65) for DFS and a HR of 1.24 (95% CI 0.77–2.40) for OS (Table 2). Similarly, no difference in LRFS, LRRFS and RFS was observed between the bevacizumab treatment arms (Table 2). Subgroup analysis (Figure 4B) could not identify any subgroup for which an effect of everolimus on DFS might be postulated. Similar results were obtained for OS except for patients pretreated with bevacizumab which showed a higher risk of death when treated with everolimus (supplementary Figure S1B, available at Annals of Oncology online).
discussion Survival analysis of the GeparQuinto study showed that neither bevacizumab nor everolimus improved survival when given simultaneously to NACT for 24 or 12 weeks, respectively. Whereas
| von Minckwitz et al.
these results were expected for everolimus due to its inert effect on pCR, a positive outcome was awaited based on the effect of bevacizumab on pCR. Several hypotheses can be formulated to explain this dissociation of pCR and its effect on survival. The absolute improvement of pCR induced by bevacizumab was only 3.5%. The size of the study was far too small to show a survival effect resulting from those additional 32 patients with a pCR after bevacizumab and a potential switch from a more unfavorable outcome to a more favorable outcome. Despite the absolute difference of pCR rates in patients with TNBC being 11.4%, again only 32 additional patients achieved a pCR. Considering that only one-third of patients had TNBC, statistical power is even lower to show a survival difference. It is ever more of concern that patients with pCR showed a tendency toward a worse outcome when treated with bevacizumab in a way that a pCR induced by chemotherapy and bevacizumab seems less durable than those induced by chemotherapy alone. Such a potential negative rebound effect of bevacizumab was postulated previously e.g. in colon cancer [10]. Survival results of GeparQuinto are in line with those reported from two large adjuvant studies E5103 and BEATRICE randomizing unselected or only patients with TNBC to chemotherapy with or without bevacizumab [11, 12]. The negative
Volume 25 | No. 12 | December 2014
Downloaded from http://annonc.oxfordjournals.org/ at University of California, San Francisco on December 6, 2014
Age Age<40 years Age>=40 years cT cT1–3 cT4 cN cN0 cN+ Extension of disease cT1–3, cNo–2 cT4 or cN3 Histological tumor type Non-lobular Lobular Grading Grase 1–2 Grase 3 Receptor status ER and/or PgR positive ER and PgR negative Bevacizumab pretreatment Without bevacizumab With bevacizumab
P-Value
original articles
Annals of Oncology
funding The trial received funding support from Roche, Novartis and Sanofi Aventis. The funders had no access to the study database and were not involved in the analysis and interpretation of the results. No grant number applicable.
Sanofi Aventis. HJ has consultant/advisor role for Roche and Novartis. All other authors have declared no conflicts of interest.
references 1. Kaufmann M, von Minckwitz G, Mamounas EP et al. Recommendations from an international consensus conference on the current status and future of neoadjuvant systemic therapy in primary breast cancer. Ann Surg Oncol 2012; 19: 1508–1516. 2. von Minckwitz G, Blohmer JU, Costa SD et al. Response-guided neoadjuvant chemotherapy for breast cancer. J Clin Oncol 2013; 31: 3623–3630. 3. von Minckwitz G, Kümmel S, Vogel P et al. Intensified neoadjuvant chemotherapy in early-responding breast cancer: phase III randomized GeparTrio study. J Natl Cancer Inst 2008; 100: 552–562. 4. von Minckwitz G, Kümmel S, Vogel P et al. Neoadjuvant vinorelbine-capecitabine versus docetaxel-doxorubicin-cyclophosphamide in early nonresponsive breast cancer: phase III randomized GeparTrio trial. J Natl Cancer Inst 2008; 100: 542–551. 5. von Minckwitz G, Eidtmann H, Rezai M et al. Neoadjuvant chemotherapy and bevacizumab for HER2-negative breast cancer. N Engl J Med 2012; 366: 299–309. 6. Huober J, Fasching PA, Hanusch C et al. Neoadjuvant chemotherapy with paclitaxel and everolimus in breast cancer patients with non-responsive tumours to epirubicin/cyclophosphamide (EC)±bevacizumab—results of the randomised GeparQuinto study (GBG 44). Eur J Cancer 2013; 49: 2284–93. 7. Untch M, Loibl S, Bischoff J et al. Lapatinib versus trastuzumab in combination with neoadjuvant anthracycline-taxane-based chemotherapy (GeparQuinto, GBG 44): a randomised phase 3 trial. Lancet Oncol 2012; 13: 135–144. 8. Gerber B, Loibl S, Eidtmann H et al. Neoadjuvant bevacizumab and anthracycline-taxane-based chemotherapy in 678 triple-negative primary breast cancers; results from the Geparquinto study (GBG 44). Ann Oncol 2013; 24: 2978–2984. 9. Hudis CA, Barlow WE, Costantino JP et al. Proposal for standardized definitions for efficacy end points in adjuvant breast cancer trials: the STEEP system. J Clin Oncol 2007; 25: 2127–2132. 10. de Gramont A, Van Cutsem E, Schmoll HJ et al. Bevacizumab plus oxaliplatinbased chemotherapy as adjuvant treatment for colon cancer (AVANT): a phase 3 randomised controlled trial. Lancet Oncol 2012; 13: 1225–1233. 11. Miller K, O’Neill A, Dang C et al. Bevacizumab (Bv) in the adjuvant treatment of HER2-negative breast cancer: final results from Eastern Cooperative Oncology Group E5103. J Clin Oncol 2014; 32: (5 suppl); abstr 500). 12. Cameron D, Brown J, Dent R et al. Adjuvant bevacizumab-containing therapy in triple-negative breast cancer (BEATRICE): primary results of a randomised, phase 3 trial. Lancet Oncol 2013; 14: 933–942. 13. von Minckwitz G, Untch M, Blohmer JU et al. Definition and impact of pathologic complete response on prognosis after neoadjuvant chemotherapy in various intrinsic breast cancer subtypes. J Clin Oncol 2012; 30: 1796–1804. 14. Cortazar P, Zhang L, Untch M et al. Pathological complete response and long-term clinical benefit in breast cancer: the CTNeoBC pooled analysis. Lancet 2014; 384: 164–172. 15. Gianni L, Eiermann W, Semiglazov V et al. Neoadjuvant and adjuvant trastuzumab in patients with HER2-positive locally advanced breast cancer (NOAH): follow-up of a randomised controlled superiority trial with a parallel HER2-negative cohort. Lancet Oncol 2014; 15: 640–647.
appendix disclosure vMG: institution received research grants from Roche, Novartis and Sanofi-Aventis. FPA received research grant from Amgen and Novartis and honoraria from Roche, Pfizer, Genomic Health, Novartis, Glaxo Smith-Kline and Amgen. HC has been consultant (advisory board) of Novartis, Sanofi Aventis, Roche, and Pfizer. TH received honoraria from Roche, Pfizer, and
Volume 25 | No. 12 | December 2014
The following centers and additional investigators contributed significantly to the conduct of the trial: Aalen (K. Gnauert); Augsburg (B. Heinrich); Bad Mergentheim (T. Prätz); Bad Nauheim (U. Groh); Bad Reichenhall (H. Tanzer); Baden-Baden (C. Villena); Bayreuth (A. Tulusan); Bergisch Gladbach (B. Liedtke); Berlin (J.-U. Blohmer, K. Kittel, C. Mau, J. Potenberg, J. Schilling); Bielefeld (M. Just); Böblingen (E. Weiss); Bochum
doi:10.1093/annonc/mdu455 |
Downloaded from http://annonc.oxfordjournals.org/ at University of California, San Francisco on December 6, 2014
result of this study was explained by the fact that bevacizumab does not have direct anti-tumor effects and that inhibition of angiogenesis requires gross tumor volumes being dependent from blood supply. Bringing the results of the BEATRICE and GeparQuinto together, it can be hypothesized that bevacizumab supports shrinkage of large tumor volumes even to an undetectable level, but e.g. without eradicating tumor stem cells or distant micrometastases already present at diagnosis which can be the source for later metastatic disease. The NASBP B-40 study has some important differences compared with GeparQuinto that might provide further insight into this hypothesis [12]. Bevacizumab was given for only 18 weeks together with NACT, but was restarted after surgery to complete 1 full year of treatment. Effect of bevacizumab was observed mainly in hormone-receptor-positive/HER2-negative tumors, but with only 1206 patients randomized the statistical power of the study is probably too low as well to demonstrate surrogacy of pCR for survival. In fact, a recent German [13] and an extended global metaanalysis of individual data of 11 955 patients from 11 trials could not prove surrogacy of pCR for survival [14]. Only the NOAH study reported an absolute improvement of pCR by 20% when trastuzumab was added to NACT for patients with HER2positive disease and found this to be predictive for a prolonged better outcome of the trastuzumab-treated patients [15]. Based on our findings, the new registration pathway recently announced by FDA and EMA has to be used with caution. Effect of the new compound on pCR must be substantial (e.g. more than absolute 20%) and there should be certainty that survival data will become available within reasonable time. Evaluation of the everolimus effect might not be final as current duration of follow-up of this phase II GeparQuinto study is still short and delayed treatment effects, especially in the hormone-receptor-positive cohort, cannot be excluded. However, despite the dose of everolimus in combination with chemotherapy was lower than the approved dose in combination with letrozole (5 versus 10 mg), the higher toxicity of this regimen will make a substantial effect on survival mandatory which even at this early point of observation is highly unlikely. In conclusion, results of the GeparQuinto study do not support the use of bevacizumab or everolimus in the neoadjuvant setting in addition to an anthracycline–taxane-based chemotherapy.
original articles
Lübeck (D. Fischer); Ludwigsfelde (A. Kohls); Ludwigshafen (W. Weikel); Magdeburg (J. Bischoff, K. Freese); Mainz (M. Schmidt, W. Wiest); Mannheim (M. Sütterlin); Marktredwitz (M. Dietrich) Minden (M. Grießhammer); München (D.-M. Burgmann, C. Hanusch, B. Rack, C. Salat, D. Sattler); Münster (J. Tio); Mutlangen (E. von Abel); Neuruppin (B. Christensen); Neuss (U. Burkamp); Oldenburg (C.-H. Köhne); Paderborn (W. Meinerz); Quedlinburg (S.-T. Graßhoff); Ravensburg (T. Decker); Recklinghausen (F. Overkamp); Reutlingen (I. Thalmann); Rheinfelden (A. Sallmann); Rosenheim (T. Beck); Rostock (T. Reimer); Rottweil (G. Bartzke); Saarbrücken (M. Deryal); Schweinfurt (M. Weigel); St Gallen, CH (J. Huober, P. Weder); Stade (C.-C. Steffens); Stadthagen (S. Lemster); Stendal (A. Stefek); Stralsund (F. Ruhland); Stuttgart (M. Hofmann, J. Schuster, W. Simon); Traunstein (U. Kronawitter); Trier (M. Clemens); Tübingen (T. Fehm); Ulm (W. Janni); Unna (K. Latos); Villingen-Schwenningen (W. Bauer); Weiden (A. Roßmann); Weinheim (L. Bauer); Weißenfels (D. Lampe); Wiesbaden (V. Heyl, G. Hoffmann, F. Lorenz-Salehi); Witten (J. Hackmann); Würzburg (R. Schlag).
Annals of Oncology 25: 2372–2378, 2014 doi:10.1093/annonc/mdu461 Published online 3 October 2014
The performance of BRCA1 immunohistochemistry for detecting germline, somatic, and epigenetic BRCA1 loss in high-grade serous ovarian cancer J. L. Meisel1, †, D. M. Hyman1,2,8,†*, K. Garg3, Q. Zhou4, F. Dao5, M. Bisogna5, J. Gao6, N. D. Schultz6, R. N. Grisham1,8, M. Phillips5, A. Iasonos4, N. D. Kauff1,5,7,8, D. A. Levine5,8, R. A. Soslow3,8,‡ & D. R. Spriggs1,2,8,‡ 1
Gynecologic Medical Oncology Service; 2Developmental Therapeutics, Department of Medicine; 3Department of Pathology; 4Department of Epidemiology and Biostatistics; 5Gynecology Service, Department of Surgery; 6Computational Biology Program; 7Clinical Genetics Service, Department of Medicine, Memorial Sloan Kettering Cancer Center, New York, USA; 8Weill Cornell Medical College, New York, USA
Received 21 April 2014; revised 8 July 2014, 8 August 2014 and 15 September 2014; accepted 18 September 2014
Background: BRCA1 expression can be lost by a variety of mechanisms including germline or somatic mutation and promotor hypermethylation. Given the potential importance of BRCA1 loss as a predictive and prognostic biomarker in high-grade serous ovarian cancer, we sought to evaluate the utility of BRCA1 immunohistochemistry (IHC) in screening for BRCA1 loss by germline, somatic, and epigenetic mechanisms. Patients and methods: Patients with advanced high-grade serous ovarian cancer who had previously undergone germline BRCA1 testing were identified. Samples from each tumor were stained for BRCA1 and reviewed independently by two pathologists blinded to BRCA status. Tumors with abnormal BRCA1 IHC and wild-type germline testing underwent further evaluation for somatic BRCA1 mutations and promoter hypermethylation. McNemar’s test was used to *Correspondence to: Dr David M. Hyman, Developmental Therapeutics, Department of Medicine, Memorial Sloan Kettering Cancer Center, 300 East 66th Street, Room 1353, New York, NY 10065, USA. Tel: +1-646-888-4544; Fax: +1-646-888-4546; E-mail: [email protected] †
Both authors contributed equally. Both authors contributed equally.
‡
© The Author 2014. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: [email protected].
Downloaded from http://annonc.oxfordjournals.org/ at University of California, San Francisco on December 6, 2014
(U. Bückner); Bonn (M. Wolfgarten); Braunschweig (R. Lorenz); Bremen (G. Doering, S. Feidicker); Chemnitz (P. Krabisch); Cuxhaven (U. Deichert); Deggendorf (D. Augustin); Dortmund (G. Kunz); Dresden (K. Kast); Düsseldorf (von Minckwitz G, C. Nestle-Krämling, M. Rezai); Ebersberg (C. Höß); Eggenfelden (J. Terhaag); Erlangen (P. Fasching); Eschweiler (P. Staib); Essen (B. Aktas); Esslingen (T. Kühn); Frankfurt am Main (F. Khandan, V. Möbus, C. Solbach, H. Tesch); Freiburg (E. Stickeler); Fürstenwalde (G. Heinrich); Fürth (H. Wagner); Gelsenkirchen (A. Abdallah); Gifhorn (T. Dewitz); Göttingen (G. Emons); Greifswald (A. Belau); Hagen (V. Rethwisch); Halle an der Saale (T. Lantzsch, C. Thomssen); Hamburg (U. Mattner, A. Nugent, V. Müller); Hameln (T. Noesselt); Hamm (F. Holms); Hanau (T. Müller); Hannover (J.-U. Deuker, I. Schrader); Herne (D. Strumberg); Hildesheim (C. Uleer); Homburg/Saar (E. Solomayer); Jena (I. Runnebaum); Kaiserslautern (H. Link); Karlsruhe (O. Tomé, H.-U. Ulmer); Kassel (B. Conrad, G. FeiselSchwickardi); Kiel (H. Eidtmann); Köln (C. Schumacher, T. Steinmetz); Landshut (I. Bauerfeind); Lebach (S. Kremers); Leipzig (D. Langanke); Lich (U. Kullmer); Limburg (A. Ober);
Annals of Oncology
Annals of Oncology
Annals of Oncology 25: 2363–2372, 2014 doi:10.1093/annonc/mdu455 Published online 15 September 2014
Survival after neoadjuvant chemotherapy with or without bevacizumab or everolimus for HER2-negative primary breast cancer (GBG 44–GeparQuinto)†
1 Headquarter, German Breast Group, Neu-Isenburg; 2Department of Gynaecology and Obstetrics, University Hospital, Frankfurt; 3Department of Gynaecology and Obstetrics, Klinikum Berlin-Buch, Berlin; 4Department of Gynaecology and Obstetrics, University Hospital, Kiel; 5Breast Center, Luisenkrankenhaus, Düsseldorf; 6 Department of Gynaecology and Obstetrics, University Hospital, Erlangen; 7Department of Medical Oncology, Chop GmbH, Frankfurt; 8Department of Gynaecology and Obstetrics, University Hospital, Magdeburg; 9Department of Gynaecology and Obstetrics, Henriettenstiftung, Hannover; 10Department of Gynaecology and Obstetrics, Praxisklinik, Berlin; 11Department of Gynaecology and Obstetrics, Rot-Kreuz-Klinikum, München; 12Department of Gynaecology and Obstetrics, University Hospital, Ulm; 13 Department of Gynaecology and Obstetrics, Sana-Klinikum, Offenbach; 14Department of Gynaecology and Obstetrics, St Johannes Hospital, Dortmund; 15Department of Gynaecology and Obstetrics, St Gertrauden-Hospital, Berlin; 16Department of Gynaecology and Obstetrics, Hospital, Rheinfelden; 17Department of Gynaecology and Obstetrics, University Hospital, Tübingen; 18Department of Gynaecology and Obstetrics, University Hospital, Rostock, Germany
Received 16 July 2014; revised 8 September 2014; accepted 9 September 2014
Background: The GeparQuinto study showed that adding bevacizumab to 24 weeks of anthracycline–taxane–based neoadjuvant chemotherapy increases pathological complete response ( pCR) rates overall and specifically in patients with triple-negative breast cancer (TNBC). No difference in pCR rate was observed for adding everolimus to paclitaxel in nonearly responding patients. Here, we present disease-free (DFS) and overall survival (OS) analyses. Patients and methods: Patients (n = 1948) with HER2-negative tumors of a median tumor size of 4 cm were randomly assigned to neoadjuvant treatment with epirubicin/cyclophosphamide followed by docetaxel (EC-T) with or without eight infusions of bevacizumab every 3 weeks before surgery. Patients without clinical response to EC ± Bevacizumab were randomized to 12 weekly cycles paclitaxel with or without everolimus 5 mg/day. To detect a hazard ratio (HR) of 0.75 (α = 0.05, β = 0.8) 379 events had to be observed in the bevacizumab arms. Results: With a median follow-up of 3.8 years, 3-year DFS was 80.8% and 3-year OS was 89.7%. Outcome was not different for patients receiving bevacizumab (HR 1.03; P = 0.784 for DFS and HR 0.974; P = 0.842 for OS) compared with patients receiving chemotherapy alone. Patients with TNBC similarly showed no improvement in DFS (HR = 0.99; P = 0.941) and OS (HR = 1.02; P = 0.891) when treated with bevacizumab. No other predefined subgroup (HR+/HER2−; locally advanced (cT4 or cN3) or not; cT1–3 or cT4; pCR or not) showed a significant benefit. No difference in DFS (HR 0.997; P = 0.987) and OS (HR 1.11; P = 0.658) was observed for nonearly responding patients receiving paclitaxel with or without everolimus overall as well as in subgroups. Conclusions: Long-term results, in opposite to the results of pCR, do not support the neoadjuvant use of bevacizumab in addition to an anthracycline–taxane-based chemotherapy or everolimus in addition to paclitaxel for nonearly responding patients. Clinical trial number: NCT 00567554, www.clinicaltrials.gov. Key words: bevacizumab, everolimus, neoadjuvant chemotherapy, disease-free survival, overall survival
*Correspondence to: Dr Gunter von Minckwitz, German Breast Group and Department of Gynaecology and Obstetrics, University Hospital, Frankfurt, c/o GBG Forschungs GmbH, Martin-Behaim-Straße 12, 63263 Neu-Isenburg, Germany. Tel: +49-69-6102-798740; Fax: +49-69-6102-7987440; E-mail: [email protected] † Part of results will be presented at the 37th San Antonio Breast Cancer Symposium, 5–8 December 2014, San Antonio, TX, USA.
introduction Anthracycline-taxane-based neoadjuvant chemotherapy (NACT) of 4–6 months is a valid option in patients with operable breast cancer [1]. A unique aspect of NACT is the opportunity to monitor tumor response to treatment and to individualize
© The Author 2014. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: [email protected].
Downloaded from http://annonc.oxfordjournals.org/ at University of California, San Francisco on December 6, 2014
G. von Minckwitz1,2*, S. Loibl1, M. Untch3, H. Eidtmann4, M. Rezai5, P. A. Fasching6, H. Tesch7, H. Eggemann8, I. Schrader9, K. Kittel10, C. Hanusch11, J. Huober12, C. Solbach2, C. Jackisch13, G. Kunz14, J. U. Blohmer15, M. Hauschild16, T. Fehm17, V. Nekljudova1, B. Gerber18 & for the GBG/AGO-B study groups
original articles
Annals of Oncology
treatment according to this interim response [2]. Patients with a complete or partial interim response are more treatment sensitive and therefore more likely to achieve a pathological complete response (pCR) at surgery [3]; however, patients without an
interim response are considered chemotherapy resistant as they rarely achieve a pCR even after a switch to a noncross-resistant chemotherapy regimen [4]. Therefore, separate treatment strategies might be developed for these patient populations.
Centrally confirmed eligibility of HER2 negative breast cancer n = 1948
Randomized to EC + Bevacizumab n = 974
Started with EC n = 969
Started with EC outside of GeparQuinto study and showed no response n = 32
Started with EC + Bevacizumab n = 956
No response n = 221
Patients with no response to EC +/- Bevacizumab n = 403
No response n = 150
Randomized to Paclitaxel n = 201
Started with Docetaxel + Bevaciczumab n = 969
Started with Docetaxel n = 969
Randomized to Paclitaxel + Everolimus n = 202
Started with Paclitaxel n = 198
Started with Paclitaxel + Everolimus n = 197
Outcome n = 58 DFS events n = 36 Death
Outcome n = 57 DFS events n = 39 Death
n = 74 n = 124
n = 72 n = 125
Outcome n = 195 DFS events n = 118 Death
Outcome n = 202 DFS events n = 116 Death
= intent to treat populations Figure 1. Consort statement: flow of patients throughout the two parts of the study.
| von Minckwitz et al.
Volume 25 | No. 12 | December 2014
Downloaded from http://annonc.oxfordjournals.org/ at University of California, San Francisco on December 6, 2014
Randomized to EC n = 974
original articles
Annals of Oncology
0.36; 95% CI 0.24–1.6; P = 0.34]. Mucosal inflammation, thrombocytopenia, neutropenia, infection and skin rash were more frequent when everolimus was added to paclitaxel [6]. As the required number of relapses or deaths has recently been reached for the HER2-negative cohort, we here report the corresponding survival analyses. Survival analyses of the patients with HER2-positive tumors are planned for 2015 when the required number of events is observed.
patients and methods objectives The primary efficacy end point for both randomized treatment comparisons of patients with HER2-negative breast cancer was pCR, which has been reported earlier together with the related toxicity [4, 5]. We here report secondary objectives of the study comparing survival of the treatment groups using disease-free (DFS), local recurrence-free (LRFS), locoregional recurrence-free (LRRFS), regional recurrence-free (RRFS), distant disease-free,
Table 1. Baseline characteristics of patients by randomized treatment group Characteristic
EC-T (N = 969) n
Age (years) <40 150 40–<50 395 50–<60 257 60+ 167 Median (range) 48 Clinical tumor stage cT4a-c 57 cT4d 60 cT1-3 852 Stage cT1-3 and cN0-2 (operable) 859 cT4 or cN3 (locally advanced) 110 Clinical nodal status cN0 391 cN1 542 Missing 36 Tumor type Ductal invasive 773 Lobular invasive 106 Other 89 Missing 1 Tumor grade I 43 II 507 III 412 Missing 7 Estrogen (ER)/progesterone (PgR) status ER−/PgR− 340 ER+ and/or PgR+ 629 Pretreatment with Bev Yes n.a. No n.a.
Volume 25 | No. 12 | December 2014
%/range
EC-T + Bev (N = 956) n %/range
15.5% 40.8% 26.5% 17.2% 24–78
154 347 286 169 49
16.1% 36.3% 29.9% 17.7% 21–75
26 62 66 44 51
13.1% 31.3% 33.3% 22.2% 27–75
18 70 62 47 50
9.1% 35.5% 31.5% 23.9% 28–76
5.9% 6.2% 87.9%
53 65 838
5.5% 6.8% 87.7%
17 16 165
8.6% 8.1% 83.3%
16 18 163
8.1% 9.1% 82.7%
88.6% 11.4%
843 113
88.2% 11.8%
167 31
84.3% 15.7%
167 30
84.8% 15.2%
41.9% 58.1%
376 554 26
40.4% 59.6%
85 107 6
44.3% 55.7%
79 114 4
40.9% 59.1%
79.9% 11.0% 9.1%
770 102 81 3
80.8% 10.7% 8.5%
158 22 18 0
79.8% 11.1% 9.1%
158 20 19 0
80.2% 10.2% 9.6%
4.5% 52.7% 42.8%
32 503 417 4
3.4% 52.8% 43.8%
9 118 70 1
4.6% 59.9% 35.5%
12 118 66 1
6.1% 60.2% 33.7%
35.1% 64.9%
323 633
33.8% 66.2%
56 142
28.3% 71.7%
53 144
26.9% 73.1%
124 74
62.6% 37.4%
125 72
63.5% 36.5%
n.a. n.a.
Pac (N = 198)
Pac + Eve (N = 197)
n
n
%/range
%/range
doi:10.1093/annonc/mdu455 |
Downloaded from http://annonc.oxfordjournals.org/ at University of California, San Francisco on December 6, 2014
The GeparQuinto phase III study investigated subtype-specific treatment approaches for newly diagnosed, untreated patients with HER2-negative breast cancer [5], HER2-negative breast cancer without response to four cycles of NACT [6] and HER2positive breast cancer [7]. This report focuses on patients with HER2-negative, tumors who were treated with sequential epirubicin/cyclophosphamide and docetaxel (EC-T) NACT and randomized to simultaneous treatment with bevacizumab or no additional therapy or were re-randomized to paclitaxel with or without everolimus if they did not respond to EC [5, 6]. pCR rates increased from 14.9% with EC-T alone to 18.4% with ECT and bevacizumab [odds ratio = 1.29; 95% confidence interval (CI) 1.02–1.65; P = 0.042] and the effect on pCR was more pronounced in patients with triple-negative breast cancer (TNBC; 27.9% and 39.3%; P = 0.003) [8]. Chemotherapy with bevacizumab was associated with a higher incidence of grade 3/4 toxicities [5]. Only 18 (4.6%) of 403 patients nonresponsive to EC achieved a pCR; 7 (3.6%) treated with paclitaxel and everolimus and 11 (5.6%) treated with paclitaxel alone [odds ratio (OR)
original articles
Annals of Oncology
100%
A
90% 80%
60% 50%
+ Censored Logrank P = 0.7837
40%
EC-T 195/969 events ECB-TB 202/956 events
30% 20% 10% 0% 0
12
24
36
48
60
72
EC-T
969
831
696
585
322
78
0
ECB-TB
956
844
700
579
339
77
0
60
72
DFS, months
B
100% 90% 80%
Proportion alive
70% 60% 50%
+ Censored Logrank P = 0.8422
40%
EC-T 118/969 events ECB-TB 116/956 events
30% 20% 10% 0% 0
12
2
6
8
EC-T
969
859
751
629
352
88
0
ECB-TB
956
863
749
651
372
84
0
OS, months Figure 2. Disease-free (A) and overall survival (B) after neoadjuvant anthracycline-taxane based chemotherapy with or without bevacizumab.
| von Minckwitz et al.
Volume 25 | No. 12 | December 2014
Downloaded from http://annonc.oxfordjournals.org/ at University of California, San Francisco on December 6, 2014
Proportion disease-free
70%
original articles
Annals of Oncology
Table 2. Overview on survival end points in both randomized cohorts End point
LRFS LLRFS RRFS DDFS CDFS (only TNBC) DFS OS
Estimated 3-year survival EC-T EC-T + Bevacizumab
HR (95% CI)
P
Estimated 3-year survival Pac Pac + Everolimus
97.3 92.6 95.4 84.3 92.6 81.5 88.7
0.99 (0.59–1.66) 0.99 (0.72–1.37) 0.88 (0.58–1.35) 1.02 (0.82–1.27) 0.95 (0.56–1.62) 1.03 (0.84–1.25) 0.97 (0.75–1.26)
0.978 0.973 0.570 0.853 0.855 0.784 0.842
94.7 87.6 91.4 75.4 n.a. 71.3 83.4
96.6 91.8 95.7 84.0 92.1 80.0 90.7
97.3 91.2 93.1 72.5 n.a. 69.3 81.4
HR (95% CI)
P
0.45 (0.14–1.47) 0.70 (0.36–1.35) 0.87 (0.40–1.89) 0.97 (0.66–1.43)
0.175 0.281 0.734 0.869
1.00 (0.69–1.44) 1.11 (0.70–1.74)
0.987 0.657
cerebral disease-free (CDFS; only in patients with triple-negative disease) and overall survival (OS) as end points.
selection of patients Female patients with previously untreated, unilateral or bilateral, histologically confirmed invasive, nonmetastatic breast carcinoma were enrolled in the study if they provided written informed consent. Patients with any stage of disease deemed to be appropriate for chemotherapy were eligible (e.g. locally advanced tumors with cT4 or cT3, hormone-receptor-negative tumors or hormone-receptor-positive tumors with cN+ (for cT2) or pNSLN+ (for cT1) disease). HER2 status of the tumor had to be negative by HercepTest™ (Dako-score 0 or 1+) or in situ hybridization (Dako-score 2+ and no gene amplification) by the local pathologist. Tumor lesions were required to have a palpable size of ≥2 cm or a sonographical size of ≥1 cm in maximum diameter and had to be measurable in two dimensions, preferably by sonography.
treatment All patients were scheduled to receive four cycles of EC (epirubicin 90 mg/ m2, cyclophosphamide 600 mg/m2, day 1, every 3 weeks) followed by four cycles of T (docetaxel 100 mg/m2, day 1, every 3 weeks). Patients were randomized to receive bevacizumab 15 mg/kg i.v. every 3 weeks starting on day 1 of the first EC cycle or no additional treatment. In patients without a sonographical or clinical response to four cycles of EC ± bevacizumab treatment was stopped if applicable and patients were randomized to weekly paclitaxel given on days 1, 8, 15, every 3 weeks with 80 mg/m2 over four cycles with or without everolimus given orally in a dose of 5 mg per day from day 13 after a dose-escalation starting from 2.5 mg every other day to 5 mg every day. In cases of tumor progression, study treatment was discontinued and further local and/or systemic treatment was permitted at the discretion of the investigator. Patients had to undergo surgery at least 28 days after the last chemotherapy or bevacizumab infusion. Recommended postoperative treatment consisted of radiotherapy to the ipsilateral remaining breast for all patients treated by breast conserving surgery and radiotherapy to the chest wall with or without regional nodes in case of mastectomy for patients with an initial tumor size >5 cm or clinically suspect axillary nodes before treatment or histologically involved nodes after neoadjuvant treatment. Endocrine treatment was given for at least 5 years in patients with HR-positive (≥10% stained cells for estrogen and/or progesterone receptor) tumors. No bevacizumab was given after surgery. Six monthly visits or telephone contacts were scheduled for at least 5 years to collect information on disease status.
Volume 25 | No. 12 | December 2014
assessment of end points DFS (invasive) was defined as time from randomization or histological diagnosis in those patients not being randomized for the bevacizumab question until any invasive locoregional, invasive contralateral or distant recurrence of breast cancer or any second primary invasive nonbreast cancer or death due to any cause [9]. Events counted for LRFS were any local, i.e. ipsilateral invasive or ductal in situ recurrence of the ipsilateral breast; for LRRFS any local or regional, i.e. local or regional lymph node recurrences or any invasive contralateral breast cancer; for RRFS any local, regional or chest wall recurrences; and for OS death of any cause. During NACT, local progression of the breast tumor was not considered as a relapse. Patients without event were censored at the date of last contact. Response to neoadjuvant treatment was assessed by physical examination, ultrasound and/or mammography and/or MRI before surgery as described previously [4]. Pathologic response was assessed locally by examining all surgically excised breast and axillary lymphatic tissue according to TNM classification. pCR was defined as pT0 ypN0. Pathology reports were centrally reviewed at GBG headquarters. End of treatment with the experimental compounds was defined as 4 weeks after last application. Thereafter, adverse events were not any longer recorded.
statistics The GeparQuinto study was an open label, randomized phase III trial with pCR as the primary end point. Allocation to chemotherapy with or without target treatment was carried out by central randomization with a 1 : 1 ratio each and a stratification according to participating site, ER/PgR status (ER or PgR positive versus ER and PgR negative), extent of disease (T4 or N3 versus T1–3 and N0–2) and pretreatment with bevacizumab or not (only for everolimus). Assumptions for sample size calculation have published previously [4, 5]. All analyses were carried out on an intent-to-treat approach including all patients having received one cycle EC for the bevacizumab question and all patients having received one course of paclitaxel for the everolimus question (Figure 1). Survival analysis for the bevacizumab randomization was planned after 379 events observed to show a hazard ratio (HR) of 0.75 between the treatment arms using a two-sided α of 0.05 and 1 − β of 0.2. As the number of patients randomized for the everolimus question was considered too low for confirmative results, it was planned to run this analysis at the same time with the bevacizumab analysis. Time-to-event outcome parameters were estimated using the Kaplan–Meier product-limit method and treatment groups were compared using the log-rank test. Univariate Cox proportional hazards models
doi:10.1093/annonc/mdu455 |
Downloaded from http://annonc.oxfordjournals.org/ at University of California, San Francisco on December 6, 2014
CDFS, cerebral metastasis disease-free survival; CI, confidence interval; DDFS, Distant disease-free survival; DFS, disease-free survival; EC, epirubicin/ cyclophosphamide; HR, hazard ratio; LRFS, local relapse-free survival; LRRFS, locoregional recurrence-free survival; n.a., not assessed; Pac, Paclitaxel; RRFS, regional recurrence-free survival (RRFS); T, taxane; TNBC, triple-negative breast cancer; OS, overall survival.
original articles A
Annals of Oncology
100% 90% 80%
60% 50%
+ Censored Logrank P = 0.9870
40%
Pw 58/198 events PwR 57/197 events
30% 20% 10% 0% Pw PwR
B
0
12
24
36
48
60
72
198 197
173 163
131 125
99 96 DFS, months
57 58
9 17
0 0
100% 90% 80% 70%
Proportion alive
60% 50%
+ Censored Logrank P = 0.6575
40%
Pw 36/198 events PwR 39/197 events
30% 20% 10% 0% Pw PwR
0
12
24
36
48
60
72
198 197
183 172
147 145
114 112
62 65
11 17
0 0
OS, months Figure 3. Disease-free (A) and overall survival (B) after neoadjuvant anthracycline-taxane based chemotherapy with or without everolimus in patients not responding to epirubicin/cyclophosphamide.
| von Minckwitz et al.
Volume 25 | No. 12 | December 2014
Downloaded from http://annonc.oxfordjournals.org/ at University of California, San Francisco on December 6, 2014
Proportion disease-free
70%
original articles
Annals of Oncology
A Subgroup
N
Hazard Ratio
Patients
(95% CI)
Overall
Test for Interaction
1925
1.03 (.844, 1.25)
.784
304 1621
1.05 (.649, 1.70) 1.02 (.819, 1.26)
.843 .881
1690 235
1.09 (.876, 1.35) .772 (.487, 1.22)
.442 .272
767 1096
1.01 (.681, 1.50) 1.07 (.850, 1.35)
.959 .557
1702 223
1.09 (.878, 1.36) .725 (.456, 1.15)
.428 .175
1713 208
1.06 (.860, 1.30) .752 (.365, 1.55)
.605 .440
1085 829
1.01 (.759, 1.36) 1.03 (.790, 1.35)
.921 .813
1262 663
1.10 (.821, 1.47) .990 (.757, 1.29)
.527 .941
320 1605
2.02 (.965, 4.22) .989 (.805, 1.22)
.062 .919
.849
.182
.791
.116
.342
.935
.610
.067
0.4 0.6 0.8 1 1.5 2 HR DFS better with ECB-TB
3
4
DFS better with EC-T
Figure 4. Disease-free survival by subgroups in patients randomized to chemotherapy with or without bevacizumab (A) and in patients randomized to paclitaxel with or without everolimus (B). were used to calculate HRs); 95% CIs are provided. Multivariate Cox model was used to adjust for relevant prespecified baseline characteristics. Univariate Cox model was carried out in subgroups and a Cox model including interaction between subgroup factor and therapy was used to test for heterogeneity across subgroups. All statistical tests were two-sided by default; the significance level was set to 0.05 unadjusted for multiple comparisons.
results Between November 2007 to June 2010, 1948 patients were registered in the HER2-negative part of the study and 1925 patients started with EC (Figure 1). No clinical response after EC treatment was observed in 371 patients who were therefore allocated for the everolimus question (nonresponding cohort). This number was lower than expected. Therefore, the study protocol was amended after the completion of recruitment for the bevacizumab question. Then patients, who received 4 cycles of EC before entering the protocol and who did not respond were allowed to be randomized for the everolimus questions. Despite this amendment, recruitment to complete the nonresponder arm was poor (recruitment of additional 32 patients) and it appeared impossible to recruit the required 566 patients. Therefore, accrual was prematurely closed in June 2011 with 403 patients
Volume 25 | No. 12 | December 2014
randomized into nonresponding cohort (Figure 1). After completion of chemotherapy and surgery, 390 relapses and 229 deaths overall and 112 relapses and 73 deaths in the nonresponding cohort were observed after a median follow-up of 3.8 years. Baseline characteristics of patients were well balanced between treatment groups as previously shown (Table 1) [4, 5]. There was no statistically significant difference neither in DFS nor OS for patients receiving bevacizumab or not together with or after anthracycline–taxane-based chemotherapy (P = 0.7837 for DFS and P = 0.8422 for OS). Two hundred two and 195 events occurred with and without bevacizumab (Figure 2, Table 2). The HR was 1.03 (95% CI 0.84–1.25) for DFS and 0.97 (95% CI 0.75–1.26) for OS, the adjusted HR from multivariate analysis was 1.11 (95% CI 0.91–1.36) for DFS and 1.06 (95% CI 0.82–1.38) for OS. Similarly, no difference in LRFS, LRRFS and RFS was observed between the bevacizumab treatment arms (Table 2). CDFS in patients with TNBC was not different between randomized arms (HR 0.95; 95% CI 0.56–1.62). Subgroup analysis (Figure 4A) could not identify a subgroup for which an effect of bevacizumab on DFS might be postulated. In opposite, patients with a pCR after neoadjuvant treatment showed a trend toward a higher risk for an event if they have received bevacizumab with a HR of 2.02 and a lower 95% CI
doi:10.1093/annonc/mdu455 |
Downloaded from http://annonc.oxfordjournals.org/ at University of California, San Francisco on December 6, 2014
Age Age<40 years Age>=40 years cT cT1–3 cT4 cN cN0 cN+ Extension of disease cT1–3, cNo–2 cT4 or cN3 Histological tumor type Non-lobular Lobular Grading Grase 1–2 Grase 3 Receptor status ER and/or PgR positive ER and PgR negative pCR yes no
P-Value
original articles
Annals of Oncology
B Subgroup
N
Hazard Ratio
Patients
(95% CI)
Overall
Test for Interaction
395
.997 (.691, 1.44)
.987
41 354
1.22 (.443, 3.38) .978 (.660, 1.45)
.697 .912
328 67
1.04 (.678, 1.60) .786 (.388, 1.59)
.847 .504
165 220
1.19 (.638, 2.21) .890 (.559, 1.42)
.589 .624
334 61
.997 (.653, 1.52) .928 (.442, 1.95)
.987 .843
353 42
.945 (.644, 1.39) 1.91 (.478, 7.66)
.773 .360
257 136
1.07 (.618, 1.86) 1.06 (.647, 1.73)
.805 .822
286 109
.853 (.507, 1.44) 1.33 (.795, 2.23)
.551 .276
249 146
.837 (.525, 1.33) 1.32 (.725, 2.41)
.455 .363
.675
.473
.484
.862
.324
.893
.262
.207
0.4 0.60.8 1 1.5 2 HR DFS better with PwR
3 4
6 8
DFS better with Pw
Fig. 4 Continued
bound of 0.965 marginally crossing 1. However, the test for interaction was not significant with P = 0.067). Similar results were obtained for OS (supplementary Figure S1A, available at Annals of Oncology online). Comparing DFS and OS in patients not responsive to EC and treated with paclitaxel with or without everolimus, no statistically significant difference was found (P = 0.9870 for DFS and P = 0.657 for OS) (Figure 3). The adjusted survival analysis revealed an HR of 1.13 (95% CI 0.77–1.65) for DFS and a HR of 1.24 (95% CI 0.77–2.40) for OS (Table 2). Similarly, no difference in LRFS, LRRFS and RFS was observed between the bevacizumab treatment arms (Table 2). Subgroup analysis (Figure 4B) could not identify any subgroup for which an effect of everolimus on DFS might be postulated. Similar results were obtained for OS except for patients pretreated with bevacizumab which showed a higher risk of death when treated with everolimus (supplementary Figure S1B, available at Annals of Oncology online).
discussion Survival analysis of the GeparQuinto study showed that neither bevacizumab nor everolimus improved survival when given simultaneously to NACT for 24 or 12 weeks, respectively. Whereas
| von Minckwitz et al.
these results were expected for everolimus due to its inert effect on pCR, a positive outcome was awaited based on the effect of bevacizumab on pCR. Several hypotheses can be formulated to explain this dissociation of pCR and its effect on survival. The absolute improvement of pCR induced by bevacizumab was only 3.5%. The size of the study was far too small to show a survival effect resulting from those additional 32 patients with a pCR after bevacizumab and a potential switch from a more unfavorable outcome to a more favorable outcome. Despite the absolute difference of pCR rates in patients with TNBC being 11.4%, again only 32 additional patients achieved a pCR. Considering that only one-third of patients had TNBC, statistical power is even lower to show a survival difference. It is ever more of concern that patients with pCR showed a tendency toward a worse outcome when treated with bevacizumab in a way that a pCR induced by chemotherapy and bevacizumab seems less durable than those induced by chemotherapy alone. Such a potential negative rebound effect of bevacizumab was postulated previously e.g. in colon cancer [10]. Survival results of GeparQuinto are in line with those reported from two large adjuvant studies E5103 and BEATRICE randomizing unselected or only patients with TNBC to chemotherapy with or without bevacizumab [11, 12]. The negative
Volume 25 | No. 12 | December 2014
Downloaded from http://annonc.oxfordjournals.org/ at University of California, San Francisco on December 6, 2014
Age Age<40 years Age>=40 years cT cT1–3 cT4 cN cN0 cN+ Extension of disease cT1–3, cNo–2 cT4 or cN3 Histological tumor type Non-lobular Lobular Grading Grase 1–2 Grase 3 Receptor status ER and/or PgR positive ER and PgR negative Bevacizumab pretreatment Without bevacizumab With bevacizumab
P-Value
original articles
Annals of Oncology
funding The trial received funding support from Roche, Novartis and Sanofi Aventis. The funders had no access to the study database and were not involved in the analysis and interpretation of the results. No grant number applicable.
Sanofi Aventis. HJ has consultant/advisor role for Roche and Novartis. All other authors have declared no conflicts of interest.
references 1. Kaufmann M, von Minckwitz G, Mamounas EP et al. Recommendations from an international consensus conference on the current status and future of neoadjuvant systemic therapy in primary breast cancer. Ann Surg Oncol 2012; 19: 1508–1516. 2. von Minckwitz G, Blohmer JU, Costa SD et al. Response-guided neoadjuvant chemotherapy for breast cancer. J Clin Oncol 2013; 31: 3623–3630. 3. von Minckwitz G, Kümmel S, Vogel P et al. Intensified neoadjuvant chemotherapy in early-responding breast cancer: phase III randomized GeparTrio study. J Natl Cancer Inst 2008; 100: 552–562. 4. von Minckwitz G, Kümmel S, Vogel P et al. Neoadjuvant vinorelbine-capecitabine versus docetaxel-doxorubicin-cyclophosphamide in early nonresponsive breast cancer: phase III randomized GeparTrio trial. J Natl Cancer Inst 2008; 100: 542–551. 5. von Minckwitz G, Eidtmann H, Rezai M et al. Neoadjuvant chemotherapy and bevacizumab for HER2-negative breast cancer. N Engl J Med 2012; 366: 299–309. 6. Huober J, Fasching PA, Hanusch C et al. Neoadjuvant chemotherapy with paclitaxel and everolimus in breast cancer patients with non-responsive tumours to epirubicin/cyclophosphamide (EC)±bevacizumab—results of the randomised GeparQuinto study (GBG 44). Eur J Cancer 2013; 49: 2284–93. 7. Untch M, Loibl S, Bischoff J et al. Lapatinib versus trastuzumab in combination with neoadjuvant anthracycline-taxane-based chemotherapy (GeparQuinto, GBG 44): a randomised phase 3 trial. Lancet Oncol 2012; 13: 135–144. 8. Gerber B, Loibl S, Eidtmann H et al. Neoadjuvant bevacizumab and anthracycline-taxane-based chemotherapy in 678 triple-negative primary breast cancers; results from the Geparquinto study (GBG 44). Ann Oncol 2013; 24: 2978–2984. 9. Hudis CA, Barlow WE, Costantino JP et al. Proposal for standardized definitions for efficacy end points in adjuvant breast cancer trials: the STEEP system. J Clin Oncol 2007; 25: 2127–2132. 10. de Gramont A, Van Cutsem E, Schmoll HJ et al. Bevacizumab plus oxaliplatinbased chemotherapy as adjuvant treatment for colon cancer (AVANT): a phase 3 randomised controlled trial. Lancet Oncol 2012; 13: 1225–1233. 11. Miller K, O’Neill A, Dang C et al. Bevacizumab (Bv) in the adjuvant treatment of HER2-negative breast cancer: final results from Eastern Cooperative Oncology Group E5103. J Clin Oncol 2014; 32: (5 suppl); abstr 500). 12. Cameron D, Brown J, Dent R et al. Adjuvant bevacizumab-containing therapy in triple-negative breast cancer (BEATRICE): primary results of a randomised, phase 3 trial. Lancet Oncol 2013; 14: 933–942. 13. von Minckwitz G, Untch M, Blohmer JU et al. Definition and impact of pathologic complete response on prognosis after neoadjuvant chemotherapy in various intrinsic breast cancer subtypes. J Clin Oncol 2012; 30: 1796–1804. 14. Cortazar P, Zhang L, Untch M et al. Pathological complete response and long-term clinical benefit in breast cancer: the CTNeoBC pooled analysis. Lancet 2014; 384: 164–172. 15. Gianni L, Eiermann W, Semiglazov V et al. Neoadjuvant and adjuvant trastuzumab in patients with HER2-positive locally advanced breast cancer (NOAH): follow-up of a randomised controlled superiority trial with a parallel HER2-negative cohort. Lancet Oncol 2014; 15: 640–647.
appendix disclosure vMG: institution received research grants from Roche, Novartis and Sanofi-Aventis. FPA received research grant from Amgen and Novartis and honoraria from Roche, Pfizer, Genomic Health, Novartis, Glaxo Smith-Kline and Amgen. HC has been consultant (advisory board) of Novartis, Sanofi Aventis, Roche, and Pfizer. TH received honoraria from Roche, Pfizer, and
Volume 25 | No. 12 | December 2014
The following centers and additional investigators contributed significantly to the conduct of the trial: Aalen (K. Gnauert); Augsburg (B. Heinrich); Bad Mergentheim (T. Prätz); Bad Nauheim (U. Groh); Bad Reichenhall (H. Tanzer); Baden-Baden (C. Villena); Bayreuth (A. Tulusan); Bergisch Gladbach (B. Liedtke); Berlin (J.-U. Blohmer, K. Kittel, C. Mau, J. Potenberg, J. Schilling); Bielefeld (M. Just); Böblingen (E. Weiss); Bochum
doi:10.1093/annonc/mdu455 |
Downloaded from http://annonc.oxfordjournals.org/ at University of California, San Francisco on December 6, 2014
result of this study was explained by the fact that bevacizumab does not have direct anti-tumor effects and that inhibition of angiogenesis requires gross tumor volumes being dependent from blood supply. Bringing the results of the BEATRICE and GeparQuinto together, it can be hypothesized that bevacizumab supports shrinkage of large tumor volumes even to an undetectable level, but e.g. without eradicating tumor stem cells or distant micrometastases already present at diagnosis which can be the source for later metastatic disease. The NASBP B-40 study has some important differences compared with GeparQuinto that might provide further insight into this hypothesis [12]. Bevacizumab was given for only 18 weeks together with NACT, but was restarted after surgery to complete 1 full year of treatment. Effect of bevacizumab was observed mainly in hormone-receptor-positive/HER2-negative tumors, but with only 1206 patients randomized the statistical power of the study is probably too low as well to demonstrate surrogacy of pCR for survival. In fact, a recent German [13] and an extended global metaanalysis of individual data of 11 955 patients from 11 trials could not prove surrogacy of pCR for survival [14]. Only the NOAH study reported an absolute improvement of pCR by 20% when trastuzumab was added to NACT for patients with HER2positive disease and found this to be predictive for a prolonged better outcome of the trastuzumab-treated patients [15]. Based on our findings, the new registration pathway recently announced by FDA and EMA has to be used with caution. Effect of the new compound on pCR must be substantial (e.g. more than absolute 20%) and there should be certainty that survival data will become available within reasonable time. Evaluation of the everolimus effect might not be final as current duration of follow-up of this phase II GeparQuinto study is still short and delayed treatment effects, especially in the hormone-receptor-positive cohort, cannot be excluded. However, despite the dose of everolimus in combination with chemotherapy was lower than the approved dose in combination with letrozole (5 versus 10 mg), the higher toxicity of this regimen will make a substantial effect on survival mandatory which even at this early point of observation is highly unlikely. In conclusion, results of the GeparQuinto study do not support the use of bevacizumab or everolimus in the neoadjuvant setting in addition to an anthracycline–taxane-based chemotherapy.
original articles
Lübeck (D. Fischer); Ludwigsfelde (A. Kohls); Ludwigshafen (W. Weikel); Magdeburg (J. Bischoff, K. Freese); Mainz (M. Schmidt, W. Wiest); Mannheim (M. Sütterlin); Marktredwitz (M. Dietrich) Minden (M. Grießhammer); München (D.-M. Burgmann, C. Hanusch, B. Rack, C. Salat, D. Sattler); Münster (J. Tio); Mutlangen (E. von Abel); Neuruppin (B. Christensen); Neuss (U. Burkamp); Oldenburg (C.-H. Köhne); Paderborn (W. Meinerz); Quedlinburg (S.-T. Graßhoff); Ravensburg (T. Decker); Recklinghausen (F. Overkamp); Reutlingen (I. Thalmann); Rheinfelden (A. Sallmann); Rosenheim (T. Beck); Rostock (T. Reimer); Rottweil (G. Bartzke); Saarbrücken (M. Deryal); Schweinfurt (M. Weigel); St Gallen, CH (J. Huober, P. Weder); Stade (C.-C. Steffens); Stadthagen (S. Lemster); Stendal (A. Stefek); Stralsund (F. Ruhland); Stuttgart (M. Hofmann, J. Schuster, W. Simon); Traunstein (U. Kronawitter); Trier (M. Clemens); Tübingen (T. Fehm); Ulm (W. Janni); Unna (K. Latos); Villingen-Schwenningen (W. Bauer); Weiden (A. Roßmann); Weinheim (L. Bauer); Weißenfels (D. Lampe); Wiesbaden (V. Heyl, G. Hoffmann, F. Lorenz-Salehi); Witten (J. Hackmann); Würzburg (R. Schlag).
Annals of Oncology 25: 2372–2378, 2014 doi:10.1093/annonc/mdu461 Published online 3 October 2014
The performance of BRCA1 immunohistochemistry for detecting germline, somatic, and epigenetic BRCA1 loss in high-grade serous ovarian cancer J. L. Meisel1, †, D. M. Hyman1,2,8,†*, K. Garg3, Q. Zhou4, F. Dao5, M. Bisogna5, J. Gao6, N. D. Schultz6, R. N. Grisham1,8, M. Phillips5, A. Iasonos4, N. D. Kauff1,5,7,8, D. A. Levine5,8, R. A. Soslow3,8,‡ & D. R. Spriggs1,2,8,‡ 1
Gynecologic Medical Oncology Service; 2Developmental Therapeutics, Department of Medicine; 3Department of Pathology; 4Department of Epidemiology and Biostatistics; 5Gynecology Service, Department of Surgery; 6Computational Biology Program; 7Clinical Genetics Service, Department of Medicine, Memorial Sloan Kettering Cancer Center, New York, USA; 8Weill Cornell Medical College, New York, USA
Received 21 April 2014; revised 8 July 2014, 8 August 2014 and 15 September 2014; accepted 18 September 2014
Background: BRCA1 expression can be lost by a variety of mechanisms including germline or somatic mutation and promotor hypermethylation. Given the potential importance of BRCA1 loss as a predictive and prognostic biomarker in high-grade serous ovarian cancer, we sought to evaluate the utility of BRCA1 immunohistochemistry (IHC) in screening for BRCA1 loss by germline, somatic, and epigenetic mechanisms. Patients and methods: Patients with advanced high-grade serous ovarian cancer who had previously undergone germline BRCA1 testing were identified. Samples from each tumor were stained for BRCA1 and reviewed independently by two pathologists blinded to BRCA status. Tumors with abnormal BRCA1 IHC and wild-type germline testing underwent further evaluation for somatic BRCA1 mutations and promoter hypermethylation. McNemar’s test was used to *Correspondence to: Dr David M. Hyman, Developmental Therapeutics, Department of Medicine, Memorial Sloan Kettering Cancer Center, 300 East 66th Street, Room 1353, New York, NY 10065, USA. Tel: +1-646-888-4544; Fax: +1-646-888-4546; E-mail: [email protected] †
Both authors contributed equally. Both authors contributed equally.
‡
© The Author 2014. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: [email protected].
Downloaded from http://annonc.oxfordjournals.org/ at University of California, San Francisco on December 6, 2014
(U. Bückner); Bonn (M. Wolfgarten); Braunschweig (R. Lorenz); Bremen (G. Doering, S. Feidicker); Chemnitz (P. Krabisch); Cuxhaven (U. Deichert); Deggendorf (D. Augustin); Dortmund (G. Kunz); Dresden (K. Kast); Düsseldorf (von Minckwitz G, C. Nestle-Krämling, M. Rezai); Ebersberg (C. Höß); Eggenfelden (J. Terhaag); Erlangen (P. Fasching); Eschweiler (P. Staib); Essen (B. Aktas); Esslingen (T. Kühn); Frankfurt am Main (F. Khandan, V. Möbus, C. Solbach, H. Tesch); Freiburg (E. Stickeler); Fürstenwalde (G. Heinrich); Fürth (H. Wagner); Gelsenkirchen (A. Abdallah); Gifhorn (T. Dewitz); Göttingen (G. Emons); Greifswald (A. Belau); Hagen (V. Rethwisch); Halle an der Saale (T. Lantzsch, C. Thomssen); Hamburg (U. Mattner, A. Nugent, V. Müller); Hameln (T. Noesselt); Hamm (F. Holms); Hanau (T. Müller); Hannover (J.-U. Deuker, I. Schrader); Herne (D. Strumberg); Hildesheim (C. Uleer); Homburg/Saar (E. Solomayer); Jena (I. Runnebaum); Kaiserslautern (H. Link); Karlsruhe (O. Tomé, H.-U. Ulmer); Kassel (B. Conrad, G. FeiselSchwickardi); Kiel (H. Eidtmann); Köln (C. Schumacher, T. Steinmetz); Landshut (I. Bauerfeind); Lebach (S. Kremers); Leipzig (D. Langanke); Lich (U. Kullmer); Limburg (A. Ober);
Annals of Oncology