- Email: [email protected]
or bevacizumab in combination with weekly paclitaxel in patients with metastatic triple-negative breast cancer
Annals of Oncology Advance Access published July 22, 2015
original article
Annals of Oncology 00: 1–7, 2015 doi:10.1093/annonc/mdv263
Randomized, phase II, placebo-controlled trial of onartuzumab and/or bevacizumab in combination with weekly paclitaxel in patients with metastatic triple-negative breast cancer V. Diéras1*, M. Campone2, D. A. Yardley3, G. Romieu4, V. Valero5, S. J. Isakoff6, H. Koeppen7, T. R. Wilson7, Y. Xiao7, D. S. Shames7, S. Mocci7, M. Chen8 & P. Schmid 9
Background: Increased hepatocyte growth factor/MET signaling is associated with an aggressive phenotype and poor prognosis in triple-negative breast cancer (TNBC). We evaluated the benefit of adding onartuzumab, a monoclonal antiMET antibody, to paclitaxel with/without bevacizumab in patients with TNBC. Patients and methods: Women with metastatic TNBC were randomized to receive onartuzumab plus placebo plus weekly paclitaxel (OP; n = 60) or onartuzumab plus bevacizumab plus paclitaxel (OBP; n = 63) or placebo plus bevacizumab plus paclitaxel (BP; n = 62). The primary end point was progression-free survival (PFS); additional end points included overall survival (OS), objective response rate (ORR), and safety. This trial was hypothesis generating and did not have power to detect minimum clinically meaningful differences between treatment arms. Results: There was no improvement in PFS with the addition of onartuzumab to BP [hazard ratio (HR), 1.08; 95% confidence interval (CI) 0.69–1.70]; the risk of a PFS event was higher with OP than with BP (HR, 1.74; 95% CI 1.13−2.68). Most patients had MET-negative tumors (88%); PAM50 subtype analysis showed basal-like tumors in 68% of samples. ORR was higher in the bevacizumab arms (OBP: 42.2%; 95% CI 28.6−57.1; BP: 54.7%; 95% CI 41.0−68.4) compared with OP (27.5%; 95% CI 15.9−40.6). Median OS was shorter with OBP (HR, 1.36; 95% CI 0.75−2.46) and OP (HR, 1.92; 95% CI 1.03−3.59), than with BP. Peripheral edema was more frequent in the onartuzumab arms (OBP, 51.8%; OP, 58.6%) versus BP (17.7%). Conclusion: This study did not show a clinical benefit of the addition of onartuzumab to paclitaxel with/without bevacizumab in patients with predominantly MET-negative TNBC. ClinicalTrials.gov: NCT01186991. Key words: bevacizumab, MET, triple-negative breast cancer, onartuzumab, paclitaxel, phase II
introduction Triple-negative breast cancers (TNBCs) are associated with an aggressive clinical course and poorer outcome than other breast cancers [1, 2]. TNBC remains a significant challenge to treat; chemotherapy is effective in some cases, with pathologic complete response (pCR) rates of around 50% reported [3, 4]. However, TNBCs have a worse outcome overall, particularly if pCR is not
*Correspondence to: Dr Véronique Diéras, Department of Clinical Research, Institut Curie Paris & Saint Cloud, 26 rue d’Ulm, 75248 Paris, Cedex 05, France. Tel: +33-1-44-32-46-75; E-mail: [email protected]
achieved [1, 5]. The addition of bevacizumab to chemotherapy in patients with metastatic TNBC significantly increased objective response rate (ORR) and progression-free survival (PFS) versus chemotherapy alone, but the lack of an overall survival (OS) benefit led to controversy about the therapeutic indication [6]. MET signaling plays a vital role in tissue remodeling and its dysregulation has been implicated in many tumors, affecting cell proliferation, invasion, metastasis, and survival [7]. The MET pathway is frequently activated in TNBCs and/or basal-like tumors [8–12]. Onartuzumab (MetMab; Genentech, South San Francisco, CA) is a recombinant, humanized, monovalent, anti-MET monoclonal
© The Author 2015. 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
Received 13 January 2015; revised 1 May 2015 and 2 June 2015; accepted 2 June 2015
Downloaded from http://annonc.oxfordjournals.org/ at University of Cambridge on August 22, 2015
1 Department of Clinical Research, Institut Curie Paris & Saint Cloud, Paris; 2Centre René Gauducheau, Centre Régional de Lutte Contre le Cancer (CRLC) Nantes, Atlantique, France; 3Sarah Cannon Research Institute and Tennessee Oncology, PLLC, Nashville, USA; 4Institut du Cancer de Montpellier (ICM), Montpellier, France; 5 The University of Texas MD Anderson Cancer Center, Houston; 6Massachusetts General Hospital, Boston; 7Genentech, Inc., South San Francisco, USA; 8 Roche Product Development, Shanghai, China; 9Barts Cancer Institute, Queen Mary University of London, London, UK
original article
Annals of Oncology
antibody, which selectively blocks ligand binding and hepatocyte growth factor (HGF)-dependent signaling [13]. Because the HGF/ MET pathway has been implicated in the upregulation of vascular endothelial growth factor (VEGF) expression, dual MET and VEGF inhibition may provide an improved clinical benefit. This phase II study evaluated the efficacy and safety of onartuzumab plus bevacizumab plus paclitaxel (OBP), and onartuzumab plus placebo plus paclitaxel (OP), relative to placebo plus bevacizumab plus paclitaxel (BP) in patients with TNBC.
methods study design
patients Eligibility criteria were ≥18 years; histologically confirmed TNBC; ≤1 prior therapy for metastatic TNBC; measurable/non-measurable metastatic/LR disease not amenable to resection with curative intent; Eastern Cooperative Oncology Group performance status ≤1; consent to providing tumor tissue; and adequate hematologic, renal, and liver function. Exclusion criteria were systemic anticancer therapy within 3 weeks before day 1, cycle 1; previous taxanes for metastatic breast cancer; prior antiangiogenic or anti-MET/antiHGF therapy; prior endocrine or human epidermal growth factor receptor 2-directed therapy; or untreated central nervous system metastases.
assessments Computed tomography scans were obtained at baseline and every two cycles. Disease status per Response Evaluation Criteria in Solid Tumors (RECIST) v1.1 was assessed by investigators. Onartuzumab and bevacizumab serum samples were assessed centrally using validated enzyme-linked immunosorbent assays. Paclitaxel/6-OHpaclitaxel plasma samples were analyzed centrally using a validated liquid chromatography tandem mass spectrometry method. Tumor tissue (archival permitted) was collected for confirmation of TNBC and evaluation of MET expression. MET immunohistochemistry (IHC) was carried out centrally with the CONFIRM SP44 anti-MET rabbit monoclonal primary antibody (Ventana Medical Systems, Inc., Tucson, AZ) on a Ventana Benchmark XT platform. Staining intensity was evaluated using an IHC scoring system [14]. MET positivity was defined as a score of
Assessed for eligibility (N = 250) Excluded (n = 65) Did not meet inclusion/ exclusion criteria (n = 48) Other reason (n = 17) Randomized (n = 185)
Randomly assigned to OBP (n = 63) Treated (n = 61) Not treated (n = 2) Physician decision (n = 1) Patient withdrew consent (n = 1)
Randomly assigned to BP (n = 62) Treated (n = 61) Not treated (n = 1) Lost to follow-up prior to the first dose (n = 1)
Randomly assigned to OP (n = 60) Treated (n = 60)
Discontinued OBP Death Physician decision Withdrawal by patient
Discontinued BP AE Death Lost to follow-up Physician decision
Discontinued OP Death Withdrawal by patient
(n = 31) (n = 27) (n = 1) (n = 3)
Analyzed populations OBP Randomized (n = 63) Safety* (n = 61) PK (n = 61)
(n = 28) (n = 1) (n = 23) (n = 1) (n = 3)
Analyzed populations BP Randomized (n = 62) Safety* (n = 61) PK (n = 61)
(n = 29) (n = 26) (n = 3)
Analyzed populations OP Randomized (n = 60) Safety (n = 60) PK (n = 60)
Figure 1. CONSORT diagram. AE, adverse event; BP, placebo plus bevacizumab plus paclitaxel; OBP, onartuzumab plus bevacizumab plus paclitaxel; OP, onartuzumab plus placebo plus paclitaxel; PK, pharmacokinetics. Asterisk indicates two patients randomized to OBP and one patient randomized to BP did not receive any study treatment and were excluded from the safety population.
| Diéras et al.
Downloaded from http://annonc.oxfordjournals.org/ at University of Cambridge on August 22, 2015
OAM4861g was a randomized, double-blind, placebo-controlled study in patients with metastatic or locally recurrent (LR) TNBC. Treatment comprised 28-day cycles of paclitaxel (90 mg/m2 days 1, 8, and 15), with onartuzumab (10 mg/kg days 1 and 15) or placebo, and bevacizumab (10 mg/kg days 1 and 15) or placebo until disease progression (PD), unacceptable toxicity, or death. No onartuzumab/bevacizumab dose reductions were permitted; paclitaxel dose modifications were allowed per standard practice. Patients who discontinued paclitaxel for toxicity could continue onartuzumab/placebo and bevacizumab/placebo. Patients were randomized to treatment (1:1:1) using an interactive voice/ web-based system stratified according to the number of metastatic sites (<3 versus ≥3), line of therapy (first versus second), and disease-free interval for first-line patients (≤6 versus >6 months). The protocol was approved by the institutional review board of each participating center and the study was conducted in accordance with the
Declaration of Helsinki and Good Clinical Practice guidelines. Patients provided written informed consent.
original article
Annals of Oncology 2+/3+ based on the 50% cutoff developed for non-small-cell lung cancer (NSCLC) [15]. The MET status was determined centrally after randomization and before unblinding. IHC for epidermal growth factor receptor (EGFR), cytokeratin 5, and phosphatase and tensin homolog (PTEN) was carried out on a Ventana autostainer using antibodies 3C6 (Ventana), AF 138 (Covance Biologics, Princeton, NJ), and 138G6 (Cell Signaling Technology, Beverley, MA), with CC1 retrieval conditions. Cases were scored using an H-score algorithm [16]. Formalin-fixed, paraffin-embedded (FFPE) sections were macrodissected to enrich for neoplastic tissue followed by RNA extraction [17] using the high-pure FFPE RNA Micro Kit (Roche Applied Sciences, Indianapolis, IN). NanoString gene expression data were collected using a custom 400-gene breast cancer panel for PAM50 analysis [18].
statistical analysis
results
Characteristic
OBP (N = 63) n (%)
OP (N = 60) n (%)
Median age, years 53.0 54.5 Range 35–79 30–79 ECOG PS 0 35 55.6 35a 60.3 1 28 44.4 23 38.3 Line of therapy First line DFI ≤6 months 10 15.9 15 25.0 DFI >6 months 40 63.5 31 51.7 Second line 13 20.6 14 23.3 Measurable disease Yes 45 71.4 51 85.0 No 18 28.6 9 15.0 MET IHC score (50% cutoff) 0 27 42.9 24 40.0 1+ 25 39.7 22 36.7 2+ 8 12.7 10 16.7 3+ 1 1.6 0 0 Missing 2 3.2 4 6.7 No. of metastatic sites <3 36 57.1 34 56.7 ≥3 27 42.9 26 43.3 Metastatic disease Bone only 5 7.9 0 0 Visceral (no bone) 35 55.6 40 66.7 Bone and visceral 23 36.5 19 31.7 None 0 1 1.7 Prior chemotherapy in the adjuvant setting Yes 31 49.2 25 41.7 No 32 50.8 35 58.3 Prior systemic therapy in the metastatic setting Yes 11 17.5 11 18.3 No 52 82.5 49 81.7
BP (N = 62) n (%) 51.5 37–69 37 25
59.7 40.3
11 35 16
17.7 56.5 25.8
53 9
85.5 14.5
30 28 3 0 1
48.4 45.2 4.8 0 1.6
35 27
56.5 43.5
5 36 20 1
8.1 58.1 32.3 1.6
28 34
45.2 54.8
12 50
19.4 80.6
a
denominator n = 58. BP, placebo plus bevacizumab plus paclitaxel; DFI, disease-free interval; ECOG PS, Eastern Cooperative Oncology Group performance status; IHC, immunohistochemistry; OBP, onartuzumab plus bevacizumab plus paclitaxel; OP, onartuzumab plus placebo plus paclitaxel.
patient demographics From March 2011 to March 2013, 185 patients were randomized to receive OBP (n = 63), OP (n = 60), or BP (n = 62) (Figure 1). Baseline demographics were balanced across the arms (Table 1). The MET status was determined in 178 patients (96%), of whom 22 (12%) had MET-positive tumors (IHC 2+/3+).
efficacy As of 22 March 2013, 133 patients (71.9%) had experienced a PFS event and 17 were receiving treatment (six OBP, two OP, nine BP). Median PFS was 7.3 months with OBP, 5.4 months with OP, and 7.2 months with BP (Table 2). Compared with BP, the risk of a PFS event was higher with OP (stratified HR, 1.74; 95% CI 1.13−2.68; P = 0.011) and similar with OBP (stratified HR, 1.08; 95% CI 0.69−1.70; P = 0.730) (Figure 2A; Table 2).
Objective response was evaluated in 149 patients with measurable disease at baseline. ORR was 42.2% (95% CI 28.6−57.1) with OBP, 27.5% (95% CI 15.9−40.6) with OP, and 54.7% (95% CI 41.0−68.4) with BP (Table 2). At the data cutoff date, 76 deaths (41.1%) were recorded: 27 with OBP, 26 with OP, and 23 with BP. Median OS was 14.7 months, 13.4 months, and 17.4 months, respectively (Table 2). OS was numerically shorter with OBP (stratified HR, 1.36; 95% CI 0.75−2.46; P = 0.316) and OP (stratified HR, 1.92; 95% CI 1.03−3.59; P = 0.038), than with BP (Table 2; Figure 2B).
exploratory end points Exploratory analyses examined the effect of baseline characteristics on PFS (supplementary Figure S1, available at Annals of
doi:10.1093/annonc/mdv263 |
Downloaded from http://annonc.oxfordjournals.org/ at University of Cambridge on August 22, 2015
The primary end point was investigator-assessed PFS, defined as the time from randomization to PD/relapse (RECIST v1.1), or death on study from any cause, whichever occurred first. OS was defined as the time from randomization to death from any cause. Median PFS/OS were estimated by Kaplan–Meier methodology. For each pairwise comparison of the OBP or OP arm versus the BP arm, an estimate of the hazard ratio (HR) and 95% confidence interval (CI) was determined using a Cox regression model stratified by the randomization stratification factors. Log-rank test (stratified) was used to compare survival between the arms. ORR was defined as the proportion of patients with a complete or partial response on two consecutive assessments ≥4 weeks apart (RECIST v1.1); 95% CI were computed using the Blyth-Still-Casella method. Adverse events (AEs) were assessed using National Cancer Institute Common Terminology Criteria for Adverse Events (v 4.0). Onartuzumab concentrations were compared with simulated results based on a population pharmacokinetic (PK) model [19]. Bevacizumab concentrations were compared with predictions from a bevacizumab PK model [20]. The study was designed to accrue ∼180 patients. Efficacy analyses were planned after 126 investigator-assessed PFS events. This trial was hypothesis generating and did not have power to detect minimum clinically meaningful differences between treatment arms. Based on 84 PFS events for the comparison of each onartuzumab-containing arm versus the BP arm, the study had 80% power to detect, at a one-sided significance level of 0.025, an HR of 0.54 for each onartuzumab-containing arm relative to the BP arm. No adjustments for multiple comparisons were made.
Table 1. Baseline patient demographic and disease characteristics
original article
Annals of Oncology
Table 2. Overview of efficacy (randomized population) Efficacy end point
(%)
OP (N = 60) n
(%)
BP (N = 62) n
(%)
39
61.9
47
78.3
47
75.8
7.3 5.68–7.82 1.08 0.69–1.70 0.7300 19 2 17
5.4 3.75–5.55 1.74 1.13–2.68 0.0109 42.2 4.4 37.8
14 0 14
−12.5 −32.2–7.2 9.8 4.67 to NE 27
7.2 5.52–9.26 – – – 27.5 0 27.5
29 6 23
−27.3 −45.4 to −9.1 5.8 3.71 to NE 42.9
26
14.7 10.58 to NE 1.36 0.75–2.46 0.3159
– – 7.5 5.36–11.53 43.3
13.4 9.76 to NE 1.92 1.03–3.59 0.0377
54.7 11.3 43.4
23
37.1 17.4 12.55 to NE – – –
a
Disease progression, relapse, or death. Based on 149 randomized patients with measurable disease at baseline. BP, placebo plus bevacizumab plus paclitaxel; CR, complete response; HR, hazard ratio; NE, not estimable; OBP, onartuzumab plus bevacizumab plus paclitaxel; OP, onartuzumab plus placebo plus paclitaxel; ORR, overall response rate; OS, overall survival; PFS, progression-free survival; PR, partial response. b
Oncology online) and OS. Results were similar across most subgroups. Analysis of the impact of MET IHC was limited by the low number of patients with MET-positive tumors (one IHC 3+, 21 IHC 2+). PAM50 analysis identified 68% of the 165 TNBC samples as basal like. Median PFS in this subgroup was 6.4 months with OBP, 4.4 months with OP, and 6.2 months with BP (supplementary Figure S2, available at Annals of Oncology online). Compared with BP, the risk of a PFS event was higher with OP (stratified HR, 1.61; 95% CI 0.89−2.90) and similar with OBP (stratified HR, 1.05; 95% CI 0.54−2.04). Median OS was 14.7 months with OBP, 10.8 months with OP, and 14.6 months with BP. Compared with BP, the risk of an OS event was higher with OP (stratified HR, 1.72; 95% CI 0.77−3.81) and similar with OBP (stratified HR, 1.07; 95% CI 0.45−2.53), but these differences were not statistically significant. Basal-like tumors showed high EGFR and low PTEN expression (supplementary Figure S3, available at Annals of Oncology online).
pharmacokinetics Mean serum onartuzumab concentrations were similar in the two onartuzumab-containing arms (supplementary Table S1, available at Annals of Oncology online), and were consistent with simulated data (90% CI 154–583 µg/ml steady-state peak;
| Diéras et al.
26.1–149 µg/ml steady-state trough). Observed bevacizumab serum concentrations were similar in the two bevacizumabcontaining arms; at steady-state, most fell within the 90% CI of the predicted concentration–time profile. Plasma paclitaxel/6OH–paclitaxel concentrations were similar among treatment arms (not shown).
safety Median dose intensity was 100% for onartuzumab and bevacizumab. Median paclitaxel dose intensity was 97.8%, 97.2%, and 93.5% in the BP, OP, and OBP arms, respectively. Most common AEs in the onartuzumab-containing arms were peripheral edema, alopecia, and fatigue (Table 3). Peripheral edema was reported by 58.1% and 58.6% of OBP and OP patients, respectively, versus 17.7% of BP patients. Alopecia, epistaxis, and diarrhea were the most frequent AEs in the bevacizumabcontaining arms. Patients receiving OBP were most likely to have an AE leading to treatment withdrawal; most often peripheral edema that led to onartuzumab withdrawal (five patients; 8.1%). Serious AEs occurred in more patients receiving OBP than OP or BP (32% versus 22% versus 24%), as did Grade 3–5 AEs (71% versus 50% versus 61%) (Table 3). Five AEs resulted in death, three (4.8%) with OBP (sepsis, dural fistula, and
Downloaded from http://annonc.oxfordjournals.org/ at University of Cambridge on August 22, 2015
PFS Eventsa Median, months 95% CI Stratified HR versus BP 95% CI Log-rank P value ORRb Responders CR PR ΔORR versus BP 95% CI Median duration of response, months 95% CI OS Patient deaths Median, months 95% CI Stratified HR versus BP 95% CI Log-rank P value
OBP (N = 63) n
original article
Annals of Oncology
A
Probability of survival
1.0 0.8 0.6 0.4 0.2
0
B
3
6
9
12 15 PFS (months)
51 44 53
27 15 35
12 8 21
5 4 7
18
21
24
3 1 4
0 0 4
0 0 2
0 0 0
Probability of survival
1.0 0.8 0.6 0.4 0.2
0 No. at risk: OBP 63 60 OP 62 BP
OBP (n = 63) OP (n = 60) BP (n = 62) 3
6
9
12 OS (months)
15
18
21
24
60 57 60
53 45 54
34 31 43
16 17 24
8 10 9
2 2 5
1 1 3
0 0 0
Figure 2. Kaplan–Meier curves of (A) PFS and (B) OS in the randomized population. BP, placebo plus bevacizumab plus paclitaxel; OBP, onartuzumab plus bevacizumab plus paclitaxel; OP, onartuzumab plus placebo plus paclitaxel.
pulmonary hemorrhage), and two (3.2%) with BP (hemorrhagic tumor necrosis and suicide). Of the known AEs associated with bevacizumab, gastrointestinal perforation occurred in two patients (3.2%) receiving OBP, and venous thromboembolic events were reported in eight patients (12.9%) receiving OBP or BP, and in three patients (5.2%) receiving OP. Most common thromboembolic events were pulmonary embolism (11 patients; 6.0%) or deep vein thrombosis (six patients; 3.3%).
discussion The addition of onartuzumab to BP or to paclitaxel did not improve PFS, OS, or ORR compared with BP in advanced TNBC. Patients in the bevacizumab arms experienced increased ORR and PFS versus the OP arm, in keeping with previous trial results, although bevacizumab is not standard of care for TNBC
in many countries. Analyses by patient baseline characteristics revealed similar results across most subgroups. Analyses according to MET IHC score were inconclusive owing to the low proportion of patients with MET-positive tumors. Comparison of outcomes in the OBP and BP arms suggested no negative impact of onartuzumab in patients with MET-negative disease. The question of a positive treatment effect in patients with MET-positive tumors could not be answered and was not explained by dose intensity, PK, safety, or AE findings. An NSCLC trial of onartuzumab plus erlotinib also demonstrated a benefit for patients with MET-positive tumors but potentially worse outcome in MET-negative disease, which was not explained by patterns of PD, PK, or safety [15]. There are a number of hypotheses for the potentially deleterious effect of onartuzumab in MET-negative tumors. It is conceivable that alternative pathways of MET activation exist such that MET might act as a tumor suppressor in MET-negative disease and as an oncogene in
doi:10.1093/annonc/mdv263 |
Downloaded from http://annonc.oxfordjournals.org/ at University of Cambridge on August 22, 2015
No. at risk: OBP 63 60 OP 62 BP
OBP (n = 63) OP (n = 60) BP (n = 62)
original article
Annals of Oncology
Table 3. Overview of adverse events (safety population) Adverse event
OBP (N = 62) n (%)
OP (N = 58) n (%)
BP (N = 62) n (%)
a One Grade 3 AE (dural fistula) was identified as leading to death. AE, adverse event; BP, placebo plus bevacizumab plus paclitaxel; OBP, onartuzumab plus bevacizumab plus paclitaxel; OP, onartuzumab plus placebo plus paclitaxel; SAE, serious adverse event.
MET-positive disease. Furthermore, the extent of VEGF and MET inhibition may differ between tumors with low and high MET expression. Alternatively, there may be a base level of MET signaling required to trigger apoptosis, but in patients with low MET expression levels there is a greater chance of the MET receptors being blocked by onartuzumab and thereby preventing HGF binding. No drug–drug interaction was evident when onartuzumab and bevacizumab were co-administered with paclitaxel. The incidence and severity of AEs were similar across treatment arms, despite a slightly lower dose intensity of docetaxel in the OBP arm. Peripheral edema was increased in both onartuzumabcontaining arms versus the control arm, as seen in NSCLC [15], although the incidence was much higher in our study (∼60% versus ∼20% NSCLC). A higher rate of AEs leading to withdrawal and of serious AEs were recorded with OBP relative to the other treatments. Most deaths were due to the patients’ underlying TNBC. Exploratory PAM50 analysis revealed that 68% of the tumor samples were basal-like (which is in-line with expectations), with high EGFR expression and low PTEN expression, suggesting activation of alternative signaling pathways. These findings highlight the need for biomarker-guided trials in TNBC to assess the effect of targeted treatment.
| Diéras et al.
acknowledgements The authors thank the patients who participated in this study.
funding This study was supported by Genentech, Inc. Support for thirdparty writing assistance for this manuscript was provided by Genentech, Inc. Grant numbers are not applicable.
disclosure HK, TRW, YX, DSS, SM, and MC are employees of Roche/ Genentech. HK, TRW, YX, and DSS are shareholders of Roche/ Genentech. VD has received honoraria from Roche/Genentech. MC has received honoraria from Novartis, Servier, Menari, and Roche, and research funding from Novartis. DAY has received honoraria from Genentech. VV has received honoraria and research funding from Roche/Genentech. GR has received honoraria from Roche. PS has received research funding from AstraZeneca, Novartis, Genentech, Oncogenex, and Astellas. All remaining authors have declared no conflicts of interest.
references 1. Carey LA, Dees EC, Sawyer L et al. The triple negative paradox: primary tumor chemosensitivity of breast cancer subtypes. Clin Cancer Res 2007; 13: 2329–2334. 2. Dent R, Trudeau M, Pritchard KI et al. Triple-negative breast cancer: clinical features and patterns of recurrence. Clin Cancer Res 2007; 13: 4429–4434. 3. Sikov WM, Berry DA, Perou CM et al. Impact of the addition of carboplatin and/or bevacizumab to neoadjuvant once-per-week paclitaxel followed by dose-dense doxorubicin and cyclophosphamide on pathologic complete response rates in stage II to III triple-negative breast cancer: CALGB 40603 (Alliance). J Clin Oncol 2015; 33: 13–21. 4. Bear HD, Tang G, Rastogi P et al. Bevacizumab added to neoadjuvant chemotherapy for breast cancer. N Engl J Med 2012; 366: 310–320. 5. 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. 6. Miles DW, Diéras V, Cortés J et al. First-line bevacizumab in combination with chemotherapy for HER2-negative metastatic breast cancer: pooled and subgroup analyses of data from 2447 patients. Ann Oncol 2013; 24: 2773–2780. 7. Trusolino L, Bertotti A, Comoglio PM. MET signalling: principles and functions in development, organ regeneration and cancer. Nat Rev Mol Cell Biol 2010; 11: 834–848. 8. Charafe-Jauffret E, Ginestier C, Monville F et al. Gene expression profiling of breast cell lines identifies potential new basal markers. Oncogene 2006; 25: 2273–2284. 9. Graveel CR, DeGroot JD, Su Y et al. Met induces diverse mammary carcinomas in mice and is associated with human basal breast cancer. Proc Natl Acad Sci USA 2009; 106: 12909–12914. 10. Garcia S, Dalès JP, Charafe-Jauffret E et al. Poor prognosis in breast carcinomas correlates with increased expression of targetable CD146 and c-Met and with proteomic basal-like phenotype. Hum Pathol 2007; 38: 830–841. 11. Zagouri F, Bago-Horvath Z, Rössler F et al. High MET expression is an adverse prognostic factor in patients with triple-negative breast cancer. Br J Cancer 2013; 108: 1100–1105.
Downloaded from http://annonc.oxfordjournals.org/ at University of Cambridge on August 22, 2015
Patients with ≥1 AE 62 100.0 57 98.3 62 100.0 AEs occurring in ≥25% of patients in either onartuzumab-containing arm Peripheral edema 36 58.1 34 58.6 11 17.7 Alopecia 30 48.4 30 51.7 37 59.7 Fatigue 29 46.8 24 41.4 30 48.4 Nausea 24 38.7 23 39.7 25 40.3 Diarrhea 33 53.2 13 22.4 31 50.0 Epistaxis 30 48.4 9 15.5 35 56.5 Asthenia 17 27.4 20 34.5 19 30.6 Constipation 21 33.9 16 27.6 23 37.1 Peripheral neuropathy 18 29.0 15 25.9 19 30.6 Headache 17 27.4 10 17.2 18 29.0 Paresthesia 16 25.8 7 12.1 19 30.6 Myalgia 17 27.4 11 19.0 11 17.7 Neutropenia 16 25.8 10 17.2 10 16.1 Patients with ≥1 SAE 20 32.3 13 22.4 15 24.2 Grade 3–5 AE 44 71.0a 29 50.0 38 61.3 Grade 5 AE 2 3.2 0 – 2 3.2 AE leading to withdrawal from: Any study drug 26 41.9 16 27.6 16 25.8 Onartuzumab/placebo 16 25.8 6 10.3 7 11.3 Bevacizumab/placebo 19 30.6 7 12.1 8 12.9 Paclitaxel 23 37.1 14 24.1 14 22.6
In conclusion, the addition of onartuzumab to BP or to paclitaxel did not improve PFS, OS, or ORR compared with BP in patients with advanced TNBC and predominantly METnegative tumors.
Annals of Oncology 12. Ho-Yen CM, Green AR, Rakha EA et al. C-Met in invasive breast cancer: is there a relationship with the basal-like subtype? Cancer 2014; 120: 163–171. 13. Merchant M, Zheng Z, Romero M et al. One-armed 5D5 (OA5D5) is a potent humanized HGF-blocking anti-c-Met monovalent antibody that inhibits HGFdependent activity in vitro and demonstrates anti-tumor efficacy in vivo. In Presented at the American Association of Cancer Research Annual Meeting. Los Angeles, CA, 14–17 April 2007. Abstract LB-372. 14. Koeppen H, Januario T, Filvaroff E et al. Characterization and clinical validation of an immunohistochemical assay for Met in non-small cell lung cancer. In Presented at the Annual Meeting of the United States and Canadian Academy of Pathology. Vancouver, BC, Canada, 17–23 March 2012. Abstract 2001. 15. Spigel DR, Ervin TJ, Ramlau RA et al. Randomized phase II trial of onartuzumab in combination with erlotinib in patients with advanced non-small-cell lung cancer. J Clin Oncol 2013; 31: 4105–4114.
original article 16. Mazières J, Brugger W, Cappuzzo F et al. Evaluation of EGFR protein expression by immunohistochemistry using H-score and the magnification rule: re-analysis of the SATURN study. Lung Cancer 2013; 82: 231–237. 17. Schleifman EB, Desai R, Spoerke JM et al. Targeted biomarker profiling of matched primary and metastatic estrogen receptor positive breast cancers. PLoS ONE 2014; 9: e88401. 18. Parker JS, Mullins M, Cheang MC et al. Supervised risk predictor of breast cancer based on intrinsic subtypes. J Clin Oncol 2009; 27: 1160–1167. 19. Xin Y, Jin D, Eppler S et al. Population pharmacokinetic analysis from phase I and phase II studies of the humanized monovalent antibody, onartuzumab (MetMAb), in patients with advanced solid tumors. J Clin Pharmacol 2013; 53: 1103–1111. 20. Lu JF, Bruno R, Eppler S et al. Clinical pharmacokinetics of bevacizumab in patients with solid tumors. Cancer Chemother Pharmacol 2008; 62: 779–786.
Downloaded from http://annonc.oxfordjournals.org/ at University of Cambridge on August 22, 2015
doi:10.1093/annonc/mdv263 |
original article
Annals of Oncology 00: 1–7, 2015 doi:10.1093/annonc/mdv263
Randomized, phase II, placebo-controlled trial of onartuzumab and/or bevacizumab in combination with weekly paclitaxel in patients with metastatic triple-negative breast cancer V. Diéras1*, M. Campone2, D. A. Yardley3, G. Romieu4, V. Valero5, S. J. Isakoff6, H. Koeppen7, T. R. Wilson7, Y. Xiao7, D. S. Shames7, S. Mocci7, M. Chen8 & P. Schmid 9
Background: Increased hepatocyte growth factor/MET signaling is associated with an aggressive phenotype and poor prognosis in triple-negative breast cancer (TNBC). We evaluated the benefit of adding onartuzumab, a monoclonal antiMET antibody, to paclitaxel with/without bevacizumab in patients with TNBC. Patients and methods: Women with metastatic TNBC were randomized to receive onartuzumab plus placebo plus weekly paclitaxel (OP; n = 60) or onartuzumab plus bevacizumab plus paclitaxel (OBP; n = 63) or placebo plus bevacizumab plus paclitaxel (BP; n = 62). The primary end point was progression-free survival (PFS); additional end points included overall survival (OS), objective response rate (ORR), and safety. This trial was hypothesis generating and did not have power to detect minimum clinically meaningful differences between treatment arms. Results: There was no improvement in PFS with the addition of onartuzumab to BP [hazard ratio (HR), 1.08; 95% confidence interval (CI) 0.69–1.70]; the risk of a PFS event was higher with OP than with BP (HR, 1.74; 95% CI 1.13−2.68). Most patients had MET-negative tumors (88%); PAM50 subtype analysis showed basal-like tumors in 68% of samples. ORR was higher in the bevacizumab arms (OBP: 42.2%; 95% CI 28.6−57.1; BP: 54.7%; 95% CI 41.0−68.4) compared with OP (27.5%; 95% CI 15.9−40.6). Median OS was shorter with OBP (HR, 1.36; 95% CI 0.75−2.46) and OP (HR, 1.92; 95% CI 1.03−3.59), than with BP. Peripheral edema was more frequent in the onartuzumab arms (OBP, 51.8%; OP, 58.6%) versus BP (17.7%). Conclusion: This study did not show a clinical benefit of the addition of onartuzumab to paclitaxel with/without bevacizumab in patients with predominantly MET-negative TNBC. ClinicalTrials.gov: NCT01186991. Key words: bevacizumab, MET, triple-negative breast cancer, onartuzumab, paclitaxel, phase II
introduction Triple-negative breast cancers (TNBCs) are associated with an aggressive clinical course and poorer outcome than other breast cancers [1, 2]. TNBC remains a significant challenge to treat; chemotherapy is effective in some cases, with pathologic complete response (pCR) rates of around 50% reported [3, 4]. However, TNBCs have a worse outcome overall, particularly if pCR is not
*Correspondence to: Dr Véronique Diéras, Department of Clinical Research, Institut Curie Paris & Saint Cloud, 26 rue d’Ulm, 75248 Paris, Cedex 05, France. Tel: +33-1-44-32-46-75; E-mail: [email protected]
achieved [1, 5]. The addition of bevacizumab to chemotherapy in patients with metastatic TNBC significantly increased objective response rate (ORR) and progression-free survival (PFS) versus chemotherapy alone, but the lack of an overall survival (OS) benefit led to controversy about the therapeutic indication [6]. MET signaling plays a vital role in tissue remodeling and its dysregulation has been implicated in many tumors, affecting cell proliferation, invasion, metastasis, and survival [7]. The MET pathway is frequently activated in TNBCs and/or basal-like tumors [8–12]. Onartuzumab (MetMab; Genentech, South San Francisco, CA) is a recombinant, humanized, monovalent, anti-MET monoclonal
© The Author 2015. 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
Received 13 January 2015; revised 1 May 2015 and 2 June 2015; accepted 2 June 2015
Downloaded from http://annonc.oxfordjournals.org/ at University of Cambridge on August 22, 2015
1 Department of Clinical Research, Institut Curie Paris & Saint Cloud, Paris; 2Centre René Gauducheau, Centre Régional de Lutte Contre le Cancer (CRLC) Nantes, Atlantique, France; 3Sarah Cannon Research Institute and Tennessee Oncology, PLLC, Nashville, USA; 4Institut du Cancer de Montpellier (ICM), Montpellier, France; 5 The University of Texas MD Anderson Cancer Center, Houston; 6Massachusetts General Hospital, Boston; 7Genentech, Inc., South San Francisco, USA; 8 Roche Product Development, Shanghai, China; 9Barts Cancer Institute, Queen Mary University of London, London, UK
original article
Annals of Oncology
antibody, which selectively blocks ligand binding and hepatocyte growth factor (HGF)-dependent signaling [13]. Because the HGF/ MET pathway has been implicated in the upregulation of vascular endothelial growth factor (VEGF) expression, dual MET and VEGF inhibition may provide an improved clinical benefit. This phase II study evaluated the efficacy and safety of onartuzumab plus bevacizumab plus paclitaxel (OBP), and onartuzumab plus placebo plus paclitaxel (OP), relative to placebo plus bevacizumab plus paclitaxel (BP) in patients with TNBC.
methods study design
patients Eligibility criteria were ≥18 years; histologically confirmed TNBC; ≤1 prior therapy for metastatic TNBC; measurable/non-measurable metastatic/LR disease not amenable to resection with curative intent; Eastern Cooperative Oncology Group performance status ≤1; consent to providing tumor tissue; and adequate hematologic, renal, and liver function. Exclusion criteria were systemic anticancer therapy within 3 weeks before day 1, cycle 1; previous taxanes for metastatic breast cancer; prior antiangiogenic or anti-MET/antiHGF therapy; prior endocrine or human epidermal growth factor receptor 2-directed therapy; or untreated central nervous system metastases.
assessments Computed tomography scans were obtained at baseline and every two cycles. Disease status per Response Evaluation Criteria in Solid Tumors (RECIST) v1.1 was assessed by investigators. Onartuzumab and bevacizumab serum samples were assessed centrally using validated enzyme-linked immunosorbent assays. Paclitaxel/6-OHpaclitaxel plasma samples were analyzed centrally using a validated liquid chromatography tandem mass spectrometry method. Tumor tissue (archival permitted) was collected for confirmation of TNBC and evaluation of MET expression. MET immunohistochemistry (IHC) was carried out centrally with the CONFIRM SP44 anti-MET rabbit monoclonal primary antibody (Ventana Medical Systems, Inc., Tucson, AZ) on a Ventana Benchmark XT platform. Staining intensity was evaluated using an IHC scoring system [14]. MET positivity was defined as a score of
Assessed for eligibility (N = 250) Excluded (n = 65) Did not meet inclusion/ exclusion criteria (n = 48) Other reason (n = 17) Randomized (n = 185)
Randomly assigned to OBP (n = 63) Treated (n = 61) Not treated (n = 2) Physician decision (n = 1) Patient withdrew consent (n = 1)
Randomly assigned to BP (n = 62) Treated (n = 61) Not treated (n = 1) Lost to follow-up prior to the first dose (n = 1)
Randomly assigned to OP (n = 60) Treated (n = 60)
Discontinued OBP Death Physician decision Withdrawal by patient
Discontinued BP AE Death Lost to follow-up Physician decision
Discontinued OP Death Withdrawal by patient
(n = 31) (n = 27) (n = 1) (n = 3)
Analyzed populations OBP Randomized (n = 63) Safety* (n = 61) PK (n = 61)
(n = 28) (n = 1) (n = 23) (n = 1) (n = 3)
Analyzed populations BP Randomized (n = 62) Safety* (n = 61) PK (n = 61)
(n = 29) (n = 26) (n = 3)
Analyzed populations OP Randomized (n = 60) Safety (n = 60) PK (n = 60)
Figure 1. CONSORT diagram. AE, adverse event; BP, placebo plus bevacizumab plus paclitaxel; OBP, onartuzumab plus bevacizumab plus paclitaxel; OP, onartuzumab plus placebo plus paclitaxel; PK, pharmacokinetics. Asterisk indicates two patients randomized to OBP and one patient randomized to BP did not receive any study treatment and were excluded from the safety population.
| Diéras et al.
Downloaded from http://annonc.oxfordjournals.org/ at University of Cambridge on August 22, 2015
OAM4861g was a randomized, double-blind, placebo-controlled study in patients with metastatic or locally recurrent (LR) TNBC. Treatment comprised 28-day cycles of paclitaxel (90 mg/m2 days 1, 8, and 15), with onartuzumab (10 mg/kg days 1 and 15) or placebo, and bevacizumab (10 mg/kg days 1 and 15) or placebo until disease progression (PD), unacceptable toxicity, or death. No onartuzumab/bevacizumab dose reductions were permitted; paclitaxel dose modifications were allowed per standard practice. Patients who discontinued paclitaxel for toxicity could continue onartuzumab/placebo and bevacizumab/placebo. Patients were randomized to treatment (1:1:1) using an interactive voice/ web-based system stratified according to the number of metastatic sites (<3 versus ≥3), line of therapy (first versus second), and disease-free interval for first-line patients (≤6 versus >6 months). The protocol was approved by the institutional review board of each participating center and the study was conducted in accordance with the
Declaration of Helsinki and Good Clinical Practice guidelines. Patients provided written informed consent.
original article
Annals of Oncology 2+/3+ based on the 50% cutoff developed for non-small-cell lung cancer (NSCLC) [15]. The MET status was determined centrally after randomization and before unblinding. IHC for epidermal growth factor receptor (EGFR), cytokeratin 5, and phosphatase and tensin homolog (PTEN) was carried out on a Ventana autostainer using antibodies 3C6 (Ventana), AF 138 (Covance Biologics, Princeton, NJ), and 138G6 (Cell Signaling Technology, Beverley, MA), with CC1 retrieval conditions. Cases were scored using an H-score algorithm [16]. Formalin-fixed, paraffin-embedded (FFPE) sections were macrodissected to enrich for neoplastic tissue followed by RNA extraction [17] using the high-pure FFPE RNA Micro Kit (Roche Applied Sciences, Indianapolis, IN). NanoString gene expression data were collected using a custom 400-gene breast cancer panel for PAM50 analysis [18].
statistical analysis
results
Characteristic
OBP (N = 63) n (%)
OP (N = 60) n (%)
Median age, years 53.0 54.5 Range 35–79 30–79 ECOG PS 0 35 55.6 35a 60.3 1 28 44.4 23 38.3 Line of therapy First line DFI ≤6 months 10 15.9 15 25.0 DFI >6 months 40 63.5 31 51.7 Second line 13 20.6 14 23.3 Measurable disease Yes 45 71.4 51 85.0 No 18 28.6 9 15.0 MET IHC score (50% cutoff) 0 27 42.9 24 40.0 1+ 25 39.7 22 36.7 2+ 8 12.7 10 16.7 3+ 1 1.6 0 0 Missing 2 3.2 4 6.7 No. of metastatic sites <3 36 57.1 34 56.7 ≥3 27 42.9 26 43.3 Metastatic disease Bone only 5 7.9 0 0 Visceral (no bone) 35 55.6 40 66.7 Bone and visceral 23 36.5 19 31.7 None 0 1 1.7 Prior chemotherapy in the adjuvant setting Yes 31 49.2 25 41.7 No 32 50.8 35 58.3 Prior systemic therapy in the metastatic setting Yes 11 17.5 11 18.3 No 52 82.5 49 81.7
BP (N = 62) n (%) 51.5 37–69 37 25
59.7 40.3
11 35 16
17.7 56.5 25.8
53 9
85.5 14.5
30 28 3 0 1
48.4 45.2 4.8 0 1.6
35 27
56.5 43.5
5 36 20 1
8.1 58.1 32.3 1.6
28 34
45.2 54.8
12 50
19.4 80.6
a
denominator n = 58. BP, placebo plus bevacizumab plus paclitaxel; DFI, disease-free interval; ECOG PS, Eastern Cooperative Oncology Group performance status; IHC, immunohistochemistry; OBP, onartuzumab plus bevacizumab plus paclitaxel; OP, onartuzumab plus placebo plus paclitaxel.
patient demographics From March 2011 to March 2013, 185 patients were randomized to receive OBP (n = 63), OP (n = 60), or BP (n = 62) (Figure 1). Baseline demographics were balanced across the arms (Table 1). The MET status was determined in 178 patients (96%), of whom 22 (12%) had MET-positive tumors (IHC 2+/3+).
efficacy As of 22 March 2013, 133 patients (71.9%) had experienced a PFS event and 17 were receiving treatment (six OBP, two OP, nine BP). Median PFS was 7.3 months with OBP, 5.4 months with OP, and 7.2 months with BP (Table 2). Compared with BP, the risk of a PFS event was higher with OP (stratified HR, 1.74; 95% CI 1.13−2.68; P = 0.011) and similar with OBP (stratified HR, 1.08; 95% CI 0.69−1.70; P = 0.730) (Figure 2A; Table 2).
Objective response was evaluated in 149 patients with measurable disease at baseline. ORR was 42.2% (95% CI 28.6−57.1) with OBP, 27.5% (95% CI 15.9−40.6) with OP, and 54.7% (95% CI 41.0−68.4) with BP (Table 2). At the data cutoff date, 76 deaths (41.1%) were recorded: 27 with OBP, 26 with OP, and 23 with BP. Median OS was 14.7 months, 13.4 months, and 17.4 months, respectively (Table 2). OS was numerically shorter with OBP (stratified HR, 1.36; 95% CI 0.75−2.46; P = 0.316) and OP (stratified HR, 1.92; 95% CI 1.03−3.59; P = 0.038), than with BP (Table 2; Figure 2B).
exploratory end points Exploratory analyses examined the effect of baseline characteristics on PFS (supplementary Figure S1, available at Annals of
doi:10.1093/annonc/mdv263 |
Downloaded from http://annonc.oxfordjournals.org/ at University of Cambridge on August 22, 2015
The primary end point was investigator-assessed PFS, defined as the time from randomization to PD/relapse (RECIST v1.1), or death on study from any cause, whichever occurred first. OS was defined as the time from randomization to death from any cause. Median PFS/OS were estimated by Kaplan–Meier methodology. For each pairwise comparison of the OBP or OP arm versus the BP arm, an estimate of the hazard ratio (HR) and 95% confidence interval (CI) was determined using a Cox regression model stratified by the randomization stratification factors. Log-rank test (stratified) was used to compare survival between the arms. ORR was defined as the proportion of patients with a complete or partial response on two consecutive assessments ≥4 weeks apart (RECIST v1.1); 95% CI were computed using the Blyth-Still-Casella method. Adverse events (AEs) were assessed using National Cancer Institute Common Terminology Criteria for Adverse Events (v 4.0). Onartuzumab concentrations were compared with simulated results based on a population pharmacokinetic (PK) model [19]. Bevacizumab concentrations were compared with predictions from a bevacizumab PK model [20]. The study was designed to accrue ∼180 patients. Efficacy analyses were planned after 126 investigator-assessed PFS events. This trial was hypothesis generating and did not have power to detect minimum clinically meaningful differences between treatment arms. Based on 84 PFS events for the comparison of each onartuzumab-containing arm versus the BP arm, the study had 80% power to detect, at a one-sided significance level of 0.025, an HR of 0.54 for each onartuzumab-containing arm relative to the BP arm. No adjustments for multiple comparisons were made.
Table 1. Baseline patient demographic and disease characteristics
original article
Annals of Oncology
Table 2. Overview of efficacy (randomized population) Efficacy end point
(%)
OP (N = 60) n
(%)
BP (N = 62) n
(%)
39
61.9
47
78.3
47
75.8
7.3 5.68–7.82 1.08 0.69–1.70 0.7300 19 2 17
5.4 3.75–5.55 1.74 1.13–2.68 0.0109 42.2 4.4 37.8
14 0 14
−12.5 −32.2–7.2 9.8 4.67 to NE 27
7.2 5.52–9.26 – – – 27.5 0 27.5
29 6 23
−27.3 −45.4 to −9.1 5.8 3.71 to NE 42.9
26
14.7 10.58 to NE 1.36 0.75–2.46 0.3159
– – 7.5 5.36–11.53 43.3
13.4 9.76 to NE 1.92 1.03–3.59 0.0377
54.7 11.3 43.4
23
37.1 17.4 12.55 to NE – – –
a
Disease progression, relapse, or death. Based on 149 randomized patients with measurable disease at baseline. BP, placebo plus bevacizumab plus paclitaxel; CR, complete response; HR, hazard ratio; NE, not estimable; OBP, onartuzumab plus bevacizumab plus paclitaxel; OP, onartuzumab plus placebo plus paclitaxel; ORR, overall response rate; OS, overall survival; PFS, progression-free survival; PR, partial response. b
Oncology online) and OS. Results were similar across most subgroups. Analysis of the impact of MET IHC was limited by the low number of patients with MET-positive tumors (one IHC 3+, 21 IHC 2+). PAM50 analysis identified 68% of the 165 TNBC samples as basal like. Median PFS in this subgroup was 6.4 months with OBP, 4.4 months with OP, and 6.2 months with BP (supplementary Figure S2, available at Annals of Oncology online). Compared with BP, the risk of a PFS event was higher with OP (stratified HR, 1.61; 95% CI 0.89−2.90) and similar with OBP (stratified HR, 1.05; 95% CI 0.54−2.04). Median OS was 14.7 months with OBP, 10.8 months with OP, and 14.6 months with BP. Compared with BP, the risk of an OS event was higher with OP (stratified HR, 1.72; 95% CI 0.77−3.81) and similar with OBP (stratified HR, 1.07; 95% CI 0.45−2.53), but these differences were not statistically significant. Basal-like tumors showed high EGFR and low PTEN expression (supplementary Figure S3, available at Annals of Oncology online).
pharmacokinetics Mean serum onartuzumab concentrations were similar in the two onartuzumab-containing arms (supplementary Table S1, available at Annals of Oncology online), and were consistent with simulated data (90% CI 154–583 µg/ml steady-state peak;
| Diéras et al.
26.1–149 µg/ml steady-state trough). Observed bevacizumab serum concentrations were similar in the two bevacizumabcontaining arms; at steady-state, most fell within the 90% CI of the predicted concentration–time profile. Plasma paclitaxel/6OH–paclitaxel concentrations were similar among treatment arms (not shown).
safety Median dose intensity was 100% for onartuzumab and bevacizumab. Median paclitaxel dose intensity was 97.8%, 97.2%, and 93.5% in the BP, OP, and OBP arms, respectively. Most common AEs in the onartuzumab-containing arms were peripheral edema, alopecia, and fatigue (Table 3). Peripheral edema was reported by 58.1% and 58.6% of OBP and OP patients, respectively, versus 17.7% of BP patients. Alopecia, epistaxis, and diarrhea were the most frequent AEs in the bevacizumabcontaining arms. Patients receiving OBP were most likely to have an AE leading to treatment withdrawal; most often peripheral edema that led to onartuzumab withdrawal (five patients; 8.1%). Serious AEs occurred in more patients receiving OBP than OP or BP (32% versus 22% versus 24%), as did Grade 3–5 AEs (71% versus 50% versus 61%) (Table 3). Five AEs resulted in death, three (4.8%) with OBP (sepsis, dural fistula, and
Downloaded from http://annonc.oxfordjournals.org/ at University of Cambridge on August 22, 2015
PFS Eventsa Median, months 95% CI Stratified HR versus BP 95% CI Log-rank P value ORRb Responders CR PR ΔORR versus BP 95% CI Median duration of response, months 95% CI OS Patient deaths Median, months 95% CI Stratified HR versus BP 95% CI Log-rank P value
OBP (N = 63) n
original article
Annals of Oncology
A
Probability of survival
1.0 0.8 0.6 0.4 0.2
0
B
3
6
9
12 15 PFS (months)
51 44 53
27 15 35
12 8 21
5 4 7
18
21
24
3 1 4
0 0 4
0 0 2
0 0 0
Probability of survival
1.0 0.8 0.6 0.4 0.2
0 No. at risk: OBP 63 60 OP 62 BP
OBP (n = 63) OP (n = 60) BP (n = 62) 3
6
9
12 OS (months)
15
18
21
24
60 57 60
53 45 54
34 31 43
16 17 24
8 10 9
2 2 5
1 1 3
0 0 0
Figure 2. Kaplan–Meier curves of (A) PFS and (B) OS in the randomized population. BP, placebo plus bevacizumab plus paclitaxel; OBP, onartuzumab plus bevacizumab plus paclitaxel; OP, onartuzumab plus placebo plus paclitaxel.
pulmonary hemorrhage), and two (3.2%) with BP (hemorrhagic tumor necrosis and suicide). Of the known AEs associated with bevacizumab, gastrointestinal perforation occurred in two patients (3.2%) receiving OBP, and venous thromboembolic events were reported in eight patients (12.9%) receiving OBP or BP, and in three patients (5.2%) receiving OP. Most common thromboembolic events were pulmonary embolism (11 patients; 6.0%) or deep vein thrombosis (six patients; 3.3%).
discussion The addition of onartuzumab to BP or to paclitaxel did not improve PFS, OS, or ORR compared with BP in advanced TNBC. Patients in the bevacizumab arms experienced increased ORR and PFS versus the OP arm, in keeping with previous trial results, although bevacizumab is not standard of care for TNBC
in many countries. Analyses by patient baseline characteristics revealed similar results across most subgroups. Analyses according to MET IHC score were inconclusive owing to the low proportion of patients with MET-positive tumors. Comparison of outcomes in the OBP and BP arms suggested no negative impact of onartuzumab in patients with MET-negative disease. The question of a positive treatment effect in patients with MET-positive tumors could not be answered and was not explained by dose intensity, PK, safety, or AE findings. An NSCLC trial of onartuzumab plus erlotinib also demonstrated a benefit for patients with MET-positive tumors but potentially worse outcome in MET-negative disease, which was not explained by patterns of PD, PK, or safety [15]. There are a number of hypotheses for the potentially deleterious effect of onartuzumab in MET-negative tumors. It is conceivable that alternative pathways of MET activation exist such that MET might act as a tumor suppressor in MET-negative disease and as an oncogene in
doi:10.1093/annonc/mdv263 |
Downloaded from http://annonc.oxfordjournals.org/ at University of Cambridge on August 22, 2015
No. at risk: OBP 63 60 OP 62 BP
OBP (n = 63) OP (n = 60) BP (n = 62)
original article
Annals of Oncology
Table 3. Overview of adverse events (safety population) Adverse event
OBP (N = 62) n (%)
OP (N = 58) n (%)
BP (N = 62) n (%)
a One Grade 3 AE (dural fistula) was identified as leading to death. AE, adverse event; BP, placebo plus bevacizumab plus paclitaxel; OBP, onartuzumab plus bevacizumab plus paclitaxel; OP, onartuzumab plus placebo plus paclitaxel; SAE, serious adverse event.
MET-positive disease. Furthermore, the extent of VEGF and MET inhibition may differ between tumors with low and high MET expression. Alternatively, there may be a base level of MET signaling required to trigger apoptosis, but in patients with low MET expression levels there is a greater chance of the MET receptors being blocked by onartuzumab and thereby preventing HGF binding. No drug–drug interaction was evident when onartuzumab and bevacizumab were co-administered with paclitaxel. The incidence and severity of AEs were similar across treatment arms, despite a slightly lower dose intensity of docetaxel in the OBP arm. Peripheral edema was increased in both onartuzumabcontaining arms versus the control arm, as seen in NSCLC [15], although the incidence was much higher in our study (∼60% versus ∼20% NSCLC). A higher rate of AEs leading to withdrawal and of serious AEs were recorded with OBP relative to the other treatments. Most deaths were due to the patients’ underlying TNBC. Exploratory PAM50 analysis revealed that 68% of the tumor samples were basal-like (which is in-line with expectations), with high EGFR expression and low PTEN expression, suggesting activation of alternative signaling pathways. These findings highlight the need for biomarker-guided trials in TNBC to assess the effect of targeted treatment.
| Diéras et al.
acknowledgements The authors thank the patients who participated in this study.
funding This study was supported by Genentech, Inc. Support for thirdparty writing assistance for this manuscript was provided by Genentech, Inc. Grant numbers are not applicable.
disclosure HK, TRW, YX, DSS, SM, and MC are employees of Roche/ Genentech. HK, TRW, YX, and DSS are shareholders of Roche/ Genentech. VD has received honoraria from Roche/Genentech. MC has received honoraria from Novartis, Servier, Menari, and Roche, and research funding from Novartis. DAY has received honoraria from Genentech. VV has received honoraria and research funding from Roche/Genentech. GR has received honoraria from Roche. PS has received research funding from AstraZeneca, Novartis, Genentech, Oncogenex, and Astellas. All remaining authors have declared no conflicts of interest.
references 1. Carey LA, Dees EC, Sawyer L et al. The triple negative paradox: primary tumor chemosensitivity of breast cancer subtypes. Clin Cancer Res 2007; 13: 2329–2334. 2. Dent R, Trudeau M, Pritchard KI et al. Triple-negative breast cancer: clinical features and patterns of recurrence. Clin Cancer Res 2007; 13: 4429–4434. 3. Sikov WM, Berry DA, Perou CM et al. Impact of the addition of carboplatin and/or bevacizumab to neoadjuvant once-per-week paclitaxel followed by dose-dense doxorubicin and cyclophosphamide on pathologic complete response rates in stage II to III triple-negative breast cancer: CALGB 40603 (Alliance). J Clin Oncol 2015; 33: 13–21. 4. Bear HD, Tang G, Rastogi P et al. Bevacizumab added to neoadjuvant chemotherapy for breast cancer. N Engl J Med 2012; 366: 310–320. 5. 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. 6. Miles DW, Diéras V, Cortés J et al. First-line bevacizumab in combination with chemotherapy for HER2-negative metastatic breast cancer: pooled and subgroup analyses of data from 2447 patients. Ann Oncol 2013; 24: 2773–2780. 7. Trusolino L, Bertotti A, Comoglio PM. MET signalling: principles and functions in development, organ regeneration and cancer. Nat Rev Mol Cell Biol 2010; 11: 834–848. 8. Charafe-Jauffret E, Ginestier C, Monville F et al. Gene expression profiling of breast cell lines identifies potential new basal markers. Oncogene 2006; 25: 2273–2284. 9. Graveel CR, DeGroot JD, Su Y et al. Met induces diverse mammary carcinomas in mice and is associated with human basal breast cancer. Proc Natl Acad Sci USA 2009; 106: 12909–12914. 10. Garcia S, Dalès JP, Charafe-Jauffret E et al. Poor prognosis in breast carcinomas correlates with increased expression of targetable CD146 and c-Met and with proteomic basal-like phenotype. Hum Pathol 2007; 38: 830–841. 11. Zagouri F, Bago-Horvath Z, Rössler F et al. High MET expression is an adverse prognostic factor in patients with triple-negative breast cancer. Br J Cancer 2013; 108: 1100–1105.
Downloaded from http://annonc.oxfordjournals.org/ at University of Cambridge on August 22, 2015
Patients with ≥1 AE 62 100.0 57 98.3 62 100.0 AEs occurring in ≥25% of patients in either onartuzumab-containing arm Peripheral edema 36 58.1 34 58.6 11 17.7 Alopecia 30 48.4 30 51.7 37 59.7 Fatigue 29 46.8 24 41.4 30 48.4 Nausea 24 38.7 23 39.7 25 40.3 Diarrhea 33 53.2 13 22.4 31 50.0 Epistaxis 30 48.4 9 15.5 35 56.5 Asthenia 17 27.4 20 34.5 19 30.6 Constipation 21 33.9 16 27.6 23 37.1 Peripheral neuropathy 18 29.0 15 25.9 19 30.6 Headache 17 27.4 10 17.2 18 29.0 Paresthesia 16 25.8 7 12.1 19 30.6 Myalgia 17 27.4 11 19.0 11 17.7 Neutropenia 16 25.8 10 17.2 10 16.1 Patients with ≥1 SAE 20 32.3 13 22.4 15 24.2 Grade 3–5 AE 44 71.0a 29 50.0 38 61.3 Grade 5 AE 2 3.2 0 – 2 3.2 AE leading to withdrawal from: Any study drug 26 41.9 16 27.6 16 25.8 Onartuzumab/placebo 16 25.8 6 10.3 7 11.3 Bevacizumab/placebo 19 30.6 7 12.1 8 12.9 Paclitaxel 23 37.1 14 24.1 14 22.6
In conclusion, the addition of onartuzumab to BP or to paclitaxel did not improve PFS, OS, or ORR compared with BP in patients with advanced TNBC and predominantly METnegative tumors.
Annals of Oncology 12. Ho-Yen CM, Green AR, Rakha EA et al. C-Met in invasive breast cancer: is there a relationship with the basal-like subtype? Cancer 2014; 120: 163–171. 13. Merchant M, Zheng Z, Romero M et al. One-armed 5D5 (OA5D5) is a potent humanized HGF-blocking anti-c-Met monovalent antibody that inhibits HGFdependent activity in vitro and demonstrates anti-tumor efficacy in vivo. In Presented at the American Association of Cancer Research Annual Meeting. Los Angeles, CA, 14–17 April 2007. Abstract LB-372. 14. Koeppen H, Januario T, Filvaroff E et al. Characterization and clinical validation of an immunohistochemical assay for Met in non-small cell lung cancer. In Presented at the Annual Meeting of the United States and Canadian Academy of Pathology. Vancouver, BC, Canada, 17–23 March 2012. Abstract 2001. 15. Spigel DR, Ervin TJ, Ramlau RA et al. Randomized phase II trial of onartuzumab in combination with erlotinib in patients with advanced non-small-cell lung cancer. J Clin Oncol 2013; 31: 4105–4114.
original article 16. Mazières J, Brugger W, Cappuzzo F et al. Evaluation of EGFR protein expression by immunohistochemistry using H-score and the magnification rule: re-analysis of the SATURN study. Lung Cancer 2013; 82: 231–237. 17. Schleifman EB, Desai R, Spoerke JM et al. Targeted biomarker profiling of matched primary and metastatic estrogen receptor positive breast cancers. PLoS ONE 2014; 9: e88401. 18. Parker JS, Mullins M, Cheang MC et al. Supervised risk predictor of breast cancer based on intrinsic subtypes. J Clin Oncol 2009; 27: 1160–1167. 19. Xin Y, Jin D, Eppler S et al. Population pharmacokinetic analysis from phase I and phase II studies of the humanized monovalent antibody, onartuzumab (MetMAb), in patients with advanced solid tumors. J Clin Pharmacol 2013; 53: 1103–1111. 20. Lu JF, Bruno R, Eppler S et al. Clinical pharmacokinetics of bevacizumab in patients with solid tumors. Cancer Chemother Pharmacol 2008; 62: 779–786.
Downloaded from http://annonc.oxfordjournals.org/ at University of Cambridge on August 22, 2015
doi:10.1093/annonc/mdv263 |