Bevacizumab bei mBC.

Präsentation zum Thema: "Bevacizumab bei mBC."—  Präsentation transkript:

1 Bevacizumab bei mBC

2 Avastin - Fachkurzinformation
Avastin® 25 mg/ml Konzentrat zur Herstellung einer Infusionslösung. Qualitative und quantitative Zusammensetzung: Jeder Milliliter enthält 25 mg Bevacizumab. Jede Durchstechflasche enthält 100 mg Bevacizumab in 4 ml bzw. 400 mg in 16 ml, entsprechend 1,4 bis 16,5 mg/ml bei Verdünnung gemäß Empfehlung. Bevacizumab ist ein rekombinanter humanisierter monoklonaler Antikörper, der mittels DNA-Technologie aus Ovarialzellen des chinesischen Hamsters (CHO-Zellen) gewonnen wird. Anwendungsgebiete: Avastin (Bevacizumab) wird in Kombination mit einer Chemotherapie auf Fluoropyrimidin-Basis zur Behandlung von Patienten mit metastasiertem Kolon- oder Rektumkarzinom angewendet. Avastin wird in Kombination mit Paclitaxel zur First-Line-Behandlung von Patienten mit metastasiertem Mammakarzinom angewendet. Zu weiteren Informationen wie auch zum HER2-Status siehe veröffentlichte Fachinformation Abschnitt 5.1 „Pharmakodynamische Eigenschaften“. Avastin wird zusätzlich zu einer Platin-haltigen Chemotherapie zur First-Line-Behandlung von Patienten mit inoperablem fortgeschrittenem, metastasiertem oder rezidivierendem nicht kleinzelligem Bronchialkarzinom, außer bei vorwiegender Plattenepithel-Histologie, angewendet. Avastin wird in Kombination mit Interferon alfa-2a zur First-Line-Behandlung von Patienten mit fortgeschrittenem und/oder metastasiertem Nierenzellkarzinom angewendet. Gegenanzeigen: - Überempfindlichkeit gegen den arzneilich wirksamen Bestandteil oder einen der sonstigen Bestandteile. - Überempfindlichkeit gegen CHO-Zellprodukte oder andere rekombinante humane oder humanisierte Antikörper. - Schwangerschaft (siehe veröffentlichte Fachinformation Abschnitt 4.6 „Fertilität, Schwangerschaft und Stillzeit“). Liste der sonstigen Bestandteile: α,α-Trehalose 2 H2O, Natriumphosphat, Polysorbat 20, Wasser für Injektionszwecke. Inhaber der Zulassung: Roche Registration Limited, 6 Falcon Way, Shire Park, Welwyn Garden City, AL7 1TW, Vereinigtes Königreich. Verschreibungspflicht/Apothekenpflicht: rezept- und apothekenpflichtig, wiederholte Abgabe verboten. Pharmakotherapeutische Gruppe: Monoklonale Antikörper; ATC-Code: L01X C07. Besondere Warnhinweise und Vorsichtsmaßnahmen für die Anwendung, Wechselwirkungen mit anderen Arzneimitteln und sonstige Wechselwirkungen sowie Informationen zu Schwangerschaft und Stillzeit und zu Nebenwirkungen sind der veröffentlichten Fachinformation zu entnehmen.

3 3 positive Phase III Studien: konsistenter Benefit von First-Line Bevacizumab in mBC
E21001,2: open-label phase III trial Combining Avastin with weekly paclitaxel doubled PFS and ORR in patients with mBC, leading to its regulatory approval in this setting AVADO3,4: double-blind, placebo-controlled, phase III trial RIBBON-15: double-blind, placebo-controlled, phase III trial Combining Avastin with standard chemotherapy (taxane, anthracycline-based combination or Xeloda) significantly improved PFS and ORR The benefit of Avastin extends to a wide range of chemotherapy options in the first-line setting Last updated August 31, 2009 The consistently high efficacy and good tolerability of Avastin-containing therapy has been demonstrated in three randomised, phase III trials. The first data showing the significant benefit of combining Avastin with a taxane (weekly paclitaxel) came from E2100 and led to regulatory approval of Avastin in LR/mBC.1,2 These data were subsequently confirmed by results of the AVADO trial3,4 (which led to EU approval of Avastin in combination with docetaxel as first-line therapy) and the RIBBON-1 trial. All of the AVADO data shown in this slide set come from the April 30th 2009 data cut-off. References Miller KD, Wang M, Gralow J, et al. Paclitaxel plus bevacizumab versus paclitaxel alone for metastatic breast cancer. N Engl J Med 2007;357:2666–76. Gray R, Bhattacharya S, Bowden C, et al. Independent review of E2100: a phase III trial of bevacizumab plus paclitaxel versus paclitaxel in women with metastatic breast cancer. J Clin Oncol 2009;27 August (Epub ahead of print). Miles DW, Chan A, Romieu G, et al. Randomized, double-blind, placebo-controlled, phase III study of bevacizumab with docetaxel or docetaxel with placebo as first-line therapy for patients with locally recurrent or metastatic breast cancer (mBC): AVADO. J Clin Oncol 2008;26:1008s (Abstract LBA1011). Avastin SmPC 2009. Robert NJ, Dieras V, Glaspy J, et al. RIBBON-1: Randomized, double-blind, placebo-controlled, phase III trial of chemotherapy with or without bevacizumab (B) for first-line treatment of HER2-negative locally recurrent or metastatic breast cancer (MBC). J Clin Oncol 2009;27(15s):Abstract 1005. 1Miller et al. NEJM 2007; 2Gray et al. JCO 2009; 3Miles et al. ASCO 2008; 4Avastin SmPC 2009; 5Robert et al. ASCO 2009 3

4 E2100 A randomised phase III trial of paclitaxel versus paclitaxel plus bevacizumab as first-line therapy for locally recurrent or metastatic breast cancer KD Miller, M Wang, J Gralow, M Dickler, MA Cobleigh, EA Perez, TN Shenkier, NE Davidson Indiana University Cancer Center, Dana Farber Cancer Institute, Pudget Sound Oncology Consortium, Memorial Sloan Kettering Cancer Center, Rush-Presbyterian-St. Luke’s Medical Center, Mayo Clinic, British Columbia Cancer Agency, Vancouver Cancer Center, Johns Hopkins Oncology Center

5 E2100: Phase III Avastin + Paclitaxel
Previously untreated LR/mBC N=722 Stratification factors: Disease-free interval No. of metastatic sites Adjuvant chemotherapy (yes vs no) ER status Treat to disease progression* Paclitaxela *No crossover permitted Treat to disease progression Paclitaxela + Avastin 10 mg/kg q2w Last updated September 2, 2009 This slide shows the study design for the E2100 phase III trial of weekly paclitaxel with or without Avastin as first-line therapy for LR/mBC. This was a multicentre randomised trial carried out by the Eastern Cooperative Oncology Group (ECOG) in the USA.1 722 patients with chemotherapy-naïve LR/mBC were randomised to receive paclitaxel (90 mg/m2/week for 3 weeks of a 4-week cycle) either alone (n=354) or combined with Avastin (10 mg/kg every 2 weeks; n=368). Randomisation was stratified by disease-free interval (<24 vs ≥24 months), number of metastatic sites (<3 vs ≥3), treatment with chemotherapy in the adjuvant setting (yes vs no) and oestrogen receptor status (positive vs negative vs unknown). Patients were treated until disease progression (as assessed by the investigator) or unacceptable toxicity. Patients in the paclitaxel-alone arm were not permitted to cross over to receive Avastin at disease progression. If patients in the Avastin plus paclitaxel arm discontinued either agent before disease progression, they were allowed to continue with single-agent therapy (Avastin or paclitaxel alone) until progression. The primary objective was to evaluate the efficacy of Avastin plus paclitaxel compared with paclitaxel alone as measured by PFS. This was defined as time from randomisation until disease progression (determined by ECOG review of investigator-reported data) or death from any cause. Secondary objectives included the evaluation of objective response rate, duration of response, overall survival, quality of life and toxicity. All patients were assessed for response until progressive disease, regardless of whether study therapy was continued until this point, and for survival for 5 years after randomisation. Reference Miller KD, Wang M, Gralow J, et al. Paclitaxel plus bevacizumab versus paclitaxel alone for metastatic breast cancer. N Engl J Med 2007;357:2666–76. Primary endpoint: PFS Other endpoints: ORR, OS, quality of life, safety a90 mg/m2 weekly for 3 weeks of a 4-week cycle Miller et al. NEJM 2007 5

6 Phase III Avastin + Paclitaxel (E2100) Einschlusskriterien
Lokal rezidivierendes oder metastasiertes Mammakarzinom HER2+ nur nach vorheriger Behandlung mit Herceptin (Trastuzumab) oder Kontraindikation Keine vorangegangene Chemotherapie in der Indikation MBC adjuvant Taxane erlaubt bei einem disease-free interval >12 Monate ECOG PS 0 oder 1 Keine Antitumor-Therapie innerhalb von 21 Tagen Keine ZNS-Metastasen (CT oder MRI erforderlich) Keine signifikante Proteinurie (>500mg/24 h) Keine therapeutische Antikoagulation To be eligible for this trial patients must have locally or recurrent MBC.1 HER2-positive patients were permitted only if previously treated with Herceptin or if Herceptin was contra-indicated. Additionally, they must not have received prior chemotherapy, although adjuvant taxane was allowed if the disease-free interval was >12 months. Patients must also have had ECOG PS of 0 or 1, and no antitumour therapy within 21 days. Patients with brain metastases, significant proteinuria (>500mg/24 hours) or those taking anticoagulation therapy were not be eligible for this trial. 1. Miller KD, Wang M, Gralow J, et al. A randomized phase III trial of paclitaxel versus paclitaxel plus bevacizumab as first-line therapy for locally recurrent or metastatic breast cancer: a trial coordinated by the Eastern Cooperative Oncology Group (E2100). Breast Cancer Res Treat 2005;94:S6 (Abstract 3). HER = human epidermal growth factor receptor CT = computed tomography MRI = magnetic resonance imaging Miller KD, et al. Breast Cancer Res Treat 2005;94:S6(Abstract 3)

7 E2100: Baseline Characteristics
Paclitaxel (n=354) Avastin + paclitaxel (n=368) Median age, years (range) 55 (27–85) 56 (29–84) ER positive, % 63 61 PgR positive, % 45 HER2 positive, % 1.7 2.4 Disease-free interval, % ≤24 months >24 months 41 59 41 59 Number of metastatic sites, % <3 ≥3 52 48 57 43 Prior taxane therapy, % 19 20 Prior anthracycline therapy, % 51 50 Last updated September 2, 2009 This slide describes the demographic and baseline patient characteristics for the E2100 study population. From December 2001 to May 2004, 722 patients were enrolled onto the study, including 354 randomised to paclitaxel and 368 randomised to Avastin plus paclitaxel.1 The patient characteristics of each group were similar with respect to age, disease-free interval 24 months, metastatic sites 3, adjuvant chemotherapy, taxane therapy and hormone receptor and HER2 status. Reference Gray R, Bhattacharya S, Bowden C, et al. Independent review of E2100: a phase III trial of bevacizumab plus paclitaxel versus paclitaxel in women with metastatic breast cancer. J Clin Oncol 2009;27 August (Epub ahead of print). Gray et al. JCO 2009 7

8 E2100: PFS (confirmed by Independent Review)
PFS by investigator PFS by IRFa 1.0 0.8 0.6 0.4 0.2 0.0 Paclitaxel (n=354) Avastin + paclitaxel (n=368) HR=0.421 p<0.0001 HR=0.483 p<0.0001 5.8 11.3 PFS estimate 5.8 11.4 Publikation Fachinfo Last updated September 2, 2009 This slide shows PFS in trial E2100. A significant increase in PFS, as determined by the trial investigators, was observed in patients receiving Avastin plus paclitaxel compared with paclitaxel alone (HR of 0.42; p<0.0001). Median PFS was 11.4 vs 5.8 months, respectively. The increase in PFS reported by the investigators was confirmed by an IRF, with median values of 11.3 months in the combination arm and 5.8 months in the paclitaxel arm (HR=0.48). The IRF confirmation is robust, as ECOG succeeded in collecting scans from around 90% of patients. Both of these analyses were conducted in accordance with the prespecified and FDA-approved methods in the Statistical Analysis Plan and IRF charter, including: intent to treat analysis of all randomised patients, with the 9 February 2005 data cut-off applied, and using the same PFS definition and censoring rules. These data are not only highly statistically significant, but also clinically meaningful for patients. Reference Gray R, Bhattacharya S, Bowden C, et al. Independent review of E2100: a phase III trial of bevacizumab plus paclitaxel versus paclitaxel in women with metastatic breast cancer. J Clin Oncol 2009;27 August (Epub ahead of print). Months aScans available for 90% of patients Gray et al. JCO 2009 8

9 E2100: Avastin + paclitaxel in context
Docetaxel Chan 1999 Median PFS / TTP 6.5 Doxorubicin Chan 1999 9 months 5.3 Monotherapy Paclitaxel Seidman 2004 6.0 Vinorelbine Muñoz 2006 4.1 Doxorubicin + paclitaxel Jassem 2001 8.3 Xeloda + docetaxel O’Shaughnessy 2002 6.1 Combination chemotherapy Gemcitabine + paclitaxel Albain 2004 5.2 FEC Zielinski 2005 9.0 Epirubicin + docetaxel Pacilio 2006 9.0 Herceptin + chemotherapy Herceptin + docetaxel Marty 2005 11.7 This slide puts the efficacy seen in the ECOG 2100 study into the context of standard regimens used in first-line treatment of mBC. It is observed that single agent or combination chemotherapies of highly active agents do not exceed 9 months PFS, whereas the addition of a biologic compound as seen in the Herceptin study for HER2 positive disease – or now in the E2100 study for HER2 negative disease with Avastin – clearly exceeds 9 months and even approaches 1 year. Docetaxel Marty 2005 6.1 Avastin + paclitaxel E Avastin + chemotherapy (E2100) 11.4 Paclitaxel E 5.8 2 4 6 8 10 12 14 Time (months)

10 E2100: PFS-Benefit in allen Subgruppena
Baseline risk factor Total n Hazard ratio (95% CI) Number of metastatic sites <3 ≥3 (0.41–0.69) (0.38–0.81) Disease-free interval (months) ≤24 months >24 months (0.42–0.79) (0.38–0.67) ER status Positive Negative (0.44–0.78) (0.31–0.61) ER/PR/HER2 combined Negative All others (0.34–0.70) (0.44–0.75) Prior adjuvant chemotherapy Yes No (0.36–0.61) (0.49–1.01) Prior taxane therapy Yes No (0.20–0.54) (0.47–0.76) Prior anthracycline therapy Yes No (0.34–0.62) (0.47–0.86) Avastin + paclitaxel better Paclitaxel better Last updated September 2, 2009 This slide shows PFS in a range of subgroups defined by baseline patient characteristics in trial E2100.1 A consistent increase in PFS was observed across all subgroups in the Avastin plus paclitaxel arm – every group clearly benefited from the addition of Avastin. Patients benefited from Avastin irrespective of prior therapy: adjuvant hormonal therapy (HR=0.56); adjuvant chemotherapy (HR=0.47); adjuvant taxane (HR=0.33); hormonal treatment in the recurrent or metastatic setting (HR=0.53).1 Importantly, patient subgroups that traditionally have a poor prognosis and/or poor response to therapy benefited from the addition of Avastin to paclitaxel, including those with a disease-free interval ≤24 months (HR=0.58). Of particular interest is the group who had received prior therapy with taxanes in the adjuvant setting. It could be argued that such patients have disease that is resistant to taxanes, or recurrence would not have occurred at all, but this subgroup had the largest improvement in PFS of any group evaluated. This is consistent with preclinical data that suggest that concomitant use of anti-VEGF therapy may restore the sensitivity of cancer cells to taxanes. In summary, patient subgroups derived benefit from the addition of Avastin to paclitaxel irrespective of prior therapy, disease-free interval, disease site, tumour burden or hormone receptor status. Reference Gray R, Bhattacharya S, Bowden C, et al. Independent review of E2100: a phase III trial of bevacizumab plus paclitaxel versus paclitaxel in women with metastatic breast cancer. J Clin Oncol 2009;27 August (Epub ahead of print). *IRF assessment Gray et al. JCO 2009 10 10

11 Investigator assessment
E2100: Response Ratea Investigator assessment Patients, % 22% 50% IRF assessment p<0.0001 p<0.0001 Patients, % 48% 23% Last updated September 2, 2009 This slide shows objective response rate for patients with measurable disease at baseline in trial E2100 according to Response Evaluation Criteria in Solid Tumours (RECIST).1,2 The objective response rate more than doubled with the addition of Avastin, from 22% to 50%.2 The objective response rate for patients with measurable disease based on investigator assessments was very similar (23% vs. 48%). The number of patients with progressive disease as their best response was halved in the Avastin-containing arm compared with the control arm (25% vs. 12%, respectively).2 References Gray R, Bhattacharya S, Bowden C, et al. Independent review of E2100: a phase III trial of bevacizumab plus paclitaxel versus paclitaxel in women with metastatic breast cancer. J Clin Oncol 2009;27 August (Epub ahead of print). Klencke B, Bhattacharya S, Samant M, et al. Independent review of E2100 validates progression-free survival (PFS) improvement with the addition of bevacizumab (B) to paclitaxel (P) as initial chemotherapy for metastatic breast cancer (MBC). J Clin Oncol 2008;26(May 20 Suppl.):50s (Abstract 1036). Paclitaxel Avastin + paclitaxel Paclitaxel Avastin + paclitaxel aPatients with measurable disease at baseline Klencke et al. ASCO 2008 11 11

12 E2100: Overall Survival 1.0 Avastin + paclitaxel (n=368): median 26.5 months Paclitaxel (n=354): median 24.8 months 81.4% 0.8 0.8 HR=0.869 (95% CI 0.722–1.046) Log-rank p=0.1374 74.0% p=0.017* 0.6 0.6 OS estimate 0.4 0.4 0.2 0.2 Last updated September 4, 2009 This slide shows mature overall survival data in trial E2100. Median overall survival was 26.5 months with Avastin plus paclitaxel versus 24.8 months with paclitaxel alone; the HR for overall survival was (p=0.1374).1 The substantial PFS benefit in E2100 did not translate into improvements in overall survival, possibly because: The study was powered to detect an overall survival increase from 24 to 31 months, but had only 10–15% power to detect an overall survival increase of 3 months Subsequent therapies may dilute any benefit obtained in the first-line setting (patients typically receive multiple therapies for mBC) Crossover to Avastin may mask an overall survival benefit. In E2100, data on subsequent therapies after first progression were not collected. Following unblinding of E2100 interim results and presentation at the 2005 ASCO annual meeting, patients in the control arm could choose to receive Avastin, which was by then commercially available in the USA.1 Reference Cameron D. Bevacizumab in the first-line treatment of metastatic breast cancer. Eur J Cancer Suppl 2008;6:21–8. 0.0 6 6 12 12 18 18 24 24 30 30 36 36 42 42 48 48 54 54 60 60 Months *Post-hoc Avastin SmPC 2009; Cameron EJC Suppl 2008; Roche data on file 2007 12 12

13 . . . but OS is not statistically significant
Subsequent therapies may dilute any benefit obtained in the 1st-line setting patients typically receive multiple therapies for mBC Post-trial use of experimental compound in control arm patients might mask an overall survival benefit In E2100 data on subsequent therapies after first progression were not collected Avastin was commercially available in the USA at the time the results were presented (ASCO 2005) K.Miller, Bevacizumab Investigator Meeting, Chicago

14 Avastin + paclitaxel (n=363)
E2100: Safety Selected grade 3/4 adverse eventsa, % Paclitaxel (n=348) Avastin + paclitaxel (n=363) Grade 3 Grade 4 Sensory neuropathy 17.0 0.6 23.7 Fatigue 4.9 0.3 10.5 Infection with grade 3/4 neutropenia 1.1 2.8 Hypertension 1.4 15.4 Arterial thromboembolic events†b 1.9 Venous thromboembolic events 2.3 2.0 Bleeding 1.7 Proteinuria Left ventricular dysfunction†c Last updated August 9, 2009 This slide summarises grade 3/4 adverse events in trial E2100. Avastin was generally well tolerated with a favourable safety profile.1 Safety data were reported according to NCI-CTC version 2.0 and collected every 3 months for patients on protocol therapy. After discontinuation of protocol therapy, adverse events were collected every 3 months until 2 years after randomisation, after which collection was every 6 months until 5 years after randomisation.1 All grade 3–5 non-haematological and grade 4/5 haematological adverse events were reported for patients in both treatment arms. Attribution to protocol therapy was determined by the investigator. In addition, selected adverse events were reported through NCI-AdEERS for both treatment arms, although only those from the Avastin plus paclitaxel arm were provided for analysis. Consequently, a direct comparison of the incidence of adverse events between the two study arms cannot be made.1 The toxicities reported in this trial were consistent with those reported previously. Patients received Avastin plus paclitaxel for approximately twice as long as those who received paclitaxel alone due to the doubling of PFS. The median exposure to paclitaxel was greater in those who experienced sensory neuropathy than in those who did not, which may account for the increase in neuropathy and fatigue.1 Grade 4 neutropenia was uncommon in both trial arms.1 NCI-AdEERS = National Cancer Institute Adverse Event Expedited Reporting System References Miles D. Management of toxicity in patients receiving therapy with bevacizumab. Eur J Cancer Suppl 2008;6:29–39. Roche data on file, 2007. aIncludes NCI AdEERS mandatory collection in the Avastin + paclitaxel arm only, which does not allow a valid comparison between the two arms bTwo additional patients died from myocardial infarction in the Avastin + paclitaxel arm cOne additional patient died in the paclitaxel arm Roche data on file 2007; Miles EJC Suppl 2008 14 14

15 E2100: Conclusio Avastin 1st line in Kombination mit Paclitaxel:
Verdoppelung des progressionsfreien Überlebens (11,3 vs. 5,8 Monate) Verdoppelung der Ansprechrate (50% vs. 22%) Ausgezeichnete Verträglichkeit 1 year survival signifikant verlängert (81,2 vs. 73,4%), median OS nicht. (Primärer Studienendpunkt war PFS!)

16 RIBBON-1: Studiendesign
Previously untreated LR/mBC N=1237 Stratification factors: Disease-free interval Previous adjuvant chemotherapy No. of metastatic sites Xeloda, taxane or anthracycline Investigator’s choice of chemotherapy Xeloda + Avastin Treat to disease progression Xeloda + placebo Optional 2nd-line chemo + Avastin Xeloda or taxane/ anthracycline RANDOMISE 2:1 Taxane/anthra + Avastin Taxane/anthra + placebo Xeloda (1000 mg/m2 bid d1–14) Taxane (docetaxel 75–100 mg/m2 or nab-paclitaxel 260 mg/m2 q3w) Anthracycline-based chemotherapy AC (doxorubicin 50–60 mg/m2, cyclophosphamide 500–600 mg/m2) EC (epirubicin 90–100 mg/m2, cyclophosphamide 500–600 mg/m2) FAC (5-FU 500 mg/m2, doxorubicin 50 mg/m2, cyclophosphamide 500 mg/m2) FEC (5-FU 500 mg/m2, epirubicin 90–100 mg/m2, cyclophosphamide 500 mg/m2) Avastin or placebo (15 mg/kg q3w) Last updated September 2, 2009 RIBBON-1, a Genentech-sponsored, placebo-controlled, international, randomised, phase III trial, was designed to determine the benefit of combining Avastin 15 mg/kg every 3 weeks with taxane- or anthracycline-based chemotherapy or Xeloda monotherapy as first-line treatment for patients with HER2-negative LR/mBC. RIBBON-1 is essentially two parallel trials: taxane- or anthracycline-containing therapy in one cohort; Xeloda monotherapy in the other cohort. Each cohort is independently powered to detect a significant PFS benefit when Avastin is combined with chemotherapy. After confirmation of eligibility, investigators declared their planned chemotherapy for each individual patient (taxane- or anthracycline-based chemotherapy or Xeloda). Patients were then randomly assigned in a 2:1 ratio to receive chemotherapy with either Avastin 15 mg/kg every 3 weeks or placebo. Taxane therapy consisted of either docetaxel (75 or 100 mg/m2) or nab-paclitaxel every 3 weeks. Investigators could choose from four anthracycline-containing combination regimens: doxorubicin plus cyclophosphamide (AC); epirubicin plus cyclophosphamide (EC); or either regimen combined with 5-FU (FAC or FEC). The Xeloda dose of 1000 mg/m2 bid represents a dose typical of clinical practice, although this is lower than the approved monotherapy dose. Avastin, taxane and Xeloda therapy were each continued until disease progression, unacceptable toxicity or withdrawal of consent. Anthracycline-based therapy was continued for a maximum of 8 cycles. At disease progression, patients could choose to receive Avastin in combination with chemotherapy in an open-label, post-progression phase, irrespective of whether they were initially randomised to receive Avastin or placebo with their first-line chemotherapy. Reference Robert NJ, Dieras V, Glaspy J, et al. RIBBON-1: Randomized, double-blind, placebo-controlled, phase III trial of chemotherapy with or without bevacizumab (B) for first-line treatment of HER2-negative locally recurrent or metastatic breast cancer (MBC). J Clin Oncol 2009;27(15s):Abstract 1005. Robert et al. ASCO 2009 16

17 RIBBON-1: Studien-Endpunkte
Primary endpoint: PFS (investigator assessed) Secondary endpoints: Overall survival and 1-year overall survival rate ORR PFS (assessed by Independent Review Committee) Safety (NCI CTCAE v3.0; incidence of selected AEs*, serious AEs, AEs leading to treatment discontinuation and clinical cardiotoxicity) Last updated July 9, 2009 The primary objective was to demonstrate superior PFS according to investigator assessment in the two Avastin-containing treatment arms versus the corresponding placebo arms. PFS was defined as the time from randomisation to the first documented disease progression or death from any cause. Patients who received non-protocol-specified anti-cancer therapy before first documentation of disease progression or death were censored at the time of the last tumour assessment before initiation of non-protocol therapy. Secondary objectives included comparison of PFS according to independent review committee (IRC) assessment, objective response rate, duration of response, overall survival, 1-year overall survival rate, safety and clinical cardiotoxicity in the Avastin–anthracycline arm versus the placebo–anthracycline arm. The IRC assessment of PFS was added at the request of the FDA to provide a sensitivity analysis. Reference Robert NJ, Dieras V, Glaspy J, et al. RIBBON-1: Randomized, double-blind, placebo-controlled, phase III trial of chemotherapy with or without bevacizumab (B) for first-line treatment of HER2-negative locally recurrent or metastatic breast cancer (MBC). J Clin Oncol 2009;27(15s):Abstract 1005. *Any grade GI perforation, RPLS Grade ≥2 arterial thromboembolic events, left ventricular systolic dysfunction Grade ≥3 venous thromboembolic events, hypertension, bleeding, proteinuria, sensory neuropathy, wound dehiscence, neutropenia, febrile neutropenia Robert et al. ASCO 2009 17

18 RIBBON-1: Statistical Design
Two independently powered cohorts Statistical and sample size assumptions: Hazard ratio Median PFS (months) Power Sample size Xeloda 0.75 6 8 80% 600 Taxane + anthracycline 0.70 7 10 90% 300/300 Last updated September 2, 2009 Originally the trial was designed as an investigator’s choice trial. However, at the request of the FDA, the trial design was amended so that the Xeloda cohort was sufficiently powered for efficacy, in light of the negative result of the AVF2119 trial of Xeloda with or without Avastin in heavily pretreated mBC. E2100 results showing the efficacy of Avastin combined with paclitaxel were already available and therefore to avoid the need for a very large trial with each cohort individually powered, the taxane and anthracycline cohort remained pooled. In the Xeloda cohort, a sample size of 600 patients and 415 events provided 80% power to detect a hazard ratio of 0.75 (increase in median PFS from 6 to 8 months) with an alpha level of 0.05. In the taxane and anthracycline cohort, a sample size of 600 patients with 405 events provided 90% power to detect a hazard ratio of 0.7 (an increase in median PFS from 7 to 10 months) with an alpha level of 0.05. The study was conducted in 232 centres in Australia, Brazil, Canada, France, Greece, Guatemala, Korea, Mexico, The Netherlands, Norway, Panama, Peru, the Philippines, Russia, Singapore, Spain, Sweden, Taiwan, Ukraine, UK, USA and Uruguay. A study accrual period of 28 months was assumed when the trial was designed. Accrual to the trial was faster than anticipated: 1237 patients were enrolled within 21 months. Reference Robert NJ, Dieras V, Glaspy J, et al. RIBBON-1: Randomized, double-blind, placebo-controlled, phase III trial of chemotherapy with or without bevacizumab (B) for first-line treatment of HER2-negative locally recurrent or metastatic breast cancer (MBC). J Clin Oncol 2009;27(15s):Abstract 1005. Robert et al. ASCO 2009 18

19 RIBBON-1: Baseline Characteristics
Xeloda Taxane Anthracycline Placebo (n=206) Avastin (n=409) (n=104) (n=203) (n=103) (n=212) Mean age, years 57 55 54 (range) (23–88) (28–91) (29–85) (32–88) (34–78) (28–81) ≥65 years, % 26 24 19 13 18 ECOG PS 0, % 53 58 50 47 HR positive, % 74 77 80 79 Triple negative, % 25 22 20 DFI ≤12 months, % 27 52 Adjuvant chemotherapy, % 76 70 71 69 41 39 30 1 60 61 ≥3 metastatic sites, % 45 43 Measurable disease, % 82 89 87 Last updated September 2, 2009 The most marked difference between the chemotherapy groups was in disease-free interval (DFI). Twice as many patients in the anthracycline subgroup than in the other chemotherapy arms had a DFI ≤12 months, presumably reflecting the higher proportion of patients presenting with metastatic disease at first diagnosis and in whom anthracyclines were consequently considered appropriate first-line therapy. In the taxane subgroup, more patients in the Avastin arm than the placebo arm had triple-negative disease (26% versus 19%, respectively). The converse was seen in the anthracycline subgroup (20% versus 25%, respectively). In the Xeloda cohort, approximately three-quarters of patients had received chemotherapy for primary breast cancer. Most had received prior anthracycline and a large proportion had also received taxane therapy. A quite similar pattern was seen in the taxane subgroup, whereas in the anthracycline subgroup, prior anthracycline exposure was not permitted and only 22% of patients had previously received chemotherapy (virtually none with taxane-containing regimens). This again probably reflects the selection of anthracycline therapy in these patients because of initial presentation with metastatic rather than primary breast cancer. References Robert NJ, Dieras V, Glaspy J, et al. RIBBON-1: Randomized, double-blind, placebo-controlled, phase III trial of chemotherapy with or without bevacizumab (B) for first-line treatment of HER2-negative locally recurrent or metastatic breast cancer (MBC). J Clin Oncol 2009;27(15s):Abstract 1005. Roche data on file, 2009. DFI = disease-free interval Robert et al. ASCO 2009; Roche data on file 2009 19

20 RIBBON-1: PFSa Avastin + Xeloda
Placebo + Xeloda (n=206) Avastin + Xeloda (n=409) 1.0 0.8 0.6 0.4 0.2 Investigator assessment Median, months 5.7 8.6 HR (95% CI) 0.69 (0.56–0.84) p=0.0002 IRC assessment Median, months 6.2 9.8 HR (95% CI) 0.68 (0.54–0.86) p=0.0011 8.6 PFS estimate 5.7 Last updated July 9, 2009 The primary objective was met in both cohorts of RIBBON-1, demonstrating significantly improved PFS when Avastin was combined with taxane- or anthracycline-based therapy or with Xeloda. In the Xeloda cohort, the hazard ratio of 0.69 represents a 31% reduction in risk of progression or death. The PFS curves show clear separation at around 2 months (after events in approximately 15% of patients). The curves remain clearly separated until around 18 months, at which point the number of patients at risk is very small and the curves are less reliable. The divergence is particularly pronounced at the median. The IRC-assessed PFS supports these findings. The hazard ratio is almost identical and the magnitude of the difference in median values is very similar in the two analyses. Reference Robert NJ, Dieras V, Glaspy J, et al. RIBBON-1: Randomized, double-blind, placebo-controlled, phase III trial of chemotherapy with or without bevacizumab (B) for first-line treatment of HER2-negative locally recurrent or metastatic breast cancer (MBC). J Clin Oncol 2009;27(15s):Abstract 1005. Months aStratified analysis Robert et al. ASCO 2009 20

21 RIBBON-1: PFS Subgruppen (Xeloda Kohorte)
Baseline risk factor n Median, months HR Favours Avastin + Xeloda Favours placebo + Xeloda Avastin + Xeloda Placebo + Xeloda Age, years <65 ≥65 8.6 9.1 4.7 6.2 Disease-free interval, months ≤12 >12 8.3 8.7 6.9 4.4 Metastatic sites, n <3 ≥3 10.2 6.6 6.4 4.2 Prior adjuvant chemotherapy Yes No 9.2 6.7 Prior taxane therapy 6.1 Visceral disease 8.1 10.6 Hormone receptor status Positive Negative Triple negative Yes No Unknown 12.4 3.9 Last updated September 27, 2009 For all subsets analysed, the hazard ratio favoured the Avastin-containing arm. In the analysis of disease-free interval (DFI) near the top of the slide, DFI is defined as the interval between the last dose of adjuvant chemotherapy and recurrence. Prior taxane therapy appeared to have no impact on the benefit derived from Avastin. The median values in the Xeloda alone arm are quite different (4.2 months in patients who had received prior taxane therapy versus 6.1 months in taxane-naïve patients), but median values in patients receiving Avastin combined with Xeloda appear similar, irrespective of prior taxane exposure. The benefit of Avastin was seen regardless of prior chemotherapy exposure or hormone receptor status. References Dieras V, Glaspy J, Brufsky A, et al. Efficacy in patient subgroups in RIBBON-1, a randomized, double-blind, phase III trial of chemotherapy with or without bevacizumab (B) for first-line treatment of HER2-negative locally recurrent or metastatic breast cancer (MBC). EJC Suppl 2009;7:265 (Abstract 5016). Roche data on file, 2009. 0.2 0.5 1 2 5 Unstratified analyses Dieras et al. ECCO-ESMO 2009; Roche data on file 2009 21 21

22 RIBBON-1: ORR (Avastin + Xelodaa)
Patients, % p=0.0097 35.4 ORR, % 23.6 Last updated September 2, 2009 Objective response rate was significantly higher in the Avastin–Xeloda arm than in the placebo–Xeloda arm. Reference Robert NJ, Dieras V, Glaspy J, et al. RIBBON-1: Randomized, double-blind, placebo-controlled, phase III trial of chemotherapy with or without bevacizumab (B) for first-line treatment of HER2-negative locally recurrent or metastatic breast cancer (MBC). J Clin Oncol 2009;27(15s):Abstract 1005. Placebo + Xeloda (n=161) Avastin + Xeloda (n=325) aPatients with measurable disease at baseline Robert et al. ASCO 2009 22

23 RIBBON-1: Overall Survival (Xeloda Kohorte)
1.0 0.8 0.6 0.4 0.2 Placebo + Xeloda (n=206) Avastin + Xeloda (n=409) Events, n (%) 72 (35.0) 122 (29.8) Median, months 21.2 29.0 HR (95% CI) 0.85 (0.63–1.14) p=0.27 1-year rate, % 74 81 p=0.076 Proportion alive Last updated September 27, 2009 The trial was neither designed nor powered to detect a difference in overall survival. At the time of this analysis, survival results are highly censored. The majority of patients (70%) randomised to Avastin combined with Xeloda were still alive. There was no difference in overall survival, as indicated by a hazard ratio of 0.85. Crossover to Avastin at progression is likely to confound more mature survival analyses. Two-thirds of the patients randomised to Xeloda plus placebo subsequently received Avastin with their second-line treatment in the open-label, post-progression phase of the study. References Robert NJ, Dieras V, Glaspy J, et al. RIBBON-1: Randomized, double-blind, placebo-controlled, phase III trial of chemotherapy with or without bevacizumab (B) for first-line treatment of HER2-negative locally recurrent or metastatic breast cancer (MBC). J Clin Oncol 2009;27(15s):Abstract 1005. Roche data on file, 2009. Avastin at progression, % 69 52 Number at risk: Avastin Placebo Months Robert et al. ASCO 2009; Roche data on file 2009 23

24 RIBBON-1: PFSa (Avastin + Taxane/Anthrazykline)
Placebo + tax/anthra (n=207) Avastin + tax/anthra (n=415) 1.0 0.8 0.6 0.4 0.2 Investigator assessment Median, months 8.0 9.2 HR (95% CI) 0.64 (0.52–0.80) p<0.0001 IRC assessment Median, months 8.3 10.7 HR (95% CI) 0.77 (0.60–0.99) p=0.040 9.2 PFS estimate 8.0 Last updated July 9, 2009 In the taxane/anthracycline cohort, the hazard ratio of 0.64 represents a 36% reduction in the risk of progression or death with Avastin combination therapy. The Kaplan-Meier PFS curves separate early, after events in only a small proportion of patients. The curves are generally well separated throughout the study period, although they come together slightly at the median. The IRC assessment showed a slightly larger increase in median values, driven by a longer median in the Avastin-containing arm but not in the placebo arm between the two analyses. Reference Robert NJ, Dieras V, Glaspy J, et al. RIBBON-1: Randomized, double-blind, placebo-controlled, phase III trial of chemotherapy with or without bevacizumab (B) for first-line treatment of HER2-negative locally recurrent or metastatic breast cancer (MBC). J Clin Oncol 2009;27(15s):Abstract 1005. Months aStratified analysis Robert et al. ASCO 2009 24

25 RIBBON-1: PFS Subgruppen (Taxane/Anthrazykline)
Baseline risk factor n Median, months HR Favours Avastin + tax/anthra Favours placebo + tax/anthra Avastin + tax/anthra Placebo + tax/anthra Age, years <65 ≥65 9.1 10.1 7.9 8.5 Disease-free interval, months ≤12 >12 8.6 9.6 6.4 8.2 Metastatic sites, n <3 ≥3 10.3 6.8 Prior adjuvant chemotherapy Yes No 9.2 8.0 Prior taxane therapy 94 528 6.7 Visceral disease 10.8 6.9 8.7 Hormone receptor status Positive Negative 6.5 6.2 Triple negative Yes No Unknown 12.1 Last updated September 27, 2009 To try to identify subgroups of patients deriving a particularly large (or small) treatment effect with Avastin, exploratory analyses in pre-defined subsets were performed. Factors analysed included stratification factors (DFI, number of metastatic sites, prior adjuvant chemotherapy), age, race, baseline ECOG performance status, menopausal status, sites of involvement, measurable versus non-measurable disease and hormone receptor status. These results show a very consistent picture for Avastin, with the hazard ratio for almost all subgroups overlapping with the hazard ratio for the entire cohort. The hazard ratios in the taxane/anthracycline cohort consistently favour the Avastin-containing arm, demonstrating a benefit in all subgroups analysed. In the analysis of DFI near the top of the slide, DFI is defined as the interval between the last dose of adjuvant chemotherapy and recurrence. References Dieras V, Glaspy J, Brufsky A, et al. Efficacy in patient subgroups in RIBBON-1, a randomized, double-blind, phase III trial of chemotherapy with or without bevacizumab (B) for first-line treatment of HER2-negative locally recurrent or metastatic breast cancer (MBC). EJC Suppl 2009;7:265 (Abstract 5016). Roche data on file, 2009. 0.2 0.5 1 2 5 Unstratified analyses Dieras et al. ECCO-ESMO 2009; Roche data on file 2009 25 25

26 RIBBON-1: ORR (Avastin + Taxane/Anthrazyklinea)
51.3 p=0.0054 37.9 ORR, % Last updated September 2, 2009 In the taxane/anthracycline cohort, objective response rate was significantly higher in the Avastin-containing arm than in the placebo arm. Reference Robert NJ, Dieras V, Glaspy J, et al. RIBBON-1: Randomized, double-blind, placebo-controlled, phase III trial of chemotherapy with or without bevacizumab (B) for first-line treatment of HER2-negative locally recurrent or metastatic breast cancer (MBC). J Clin Oncol 2009;27(15s):Abstract 1005. Placebo + tax/anthra (n=177) Avastin + tax/anthra (n=345) aPatients with measurable disease at baseline Robert et al. ASCO 2009 26

27 RIBBON-1: Overall Survival (Taxane/Anthrazykline)
1.0 0.8 0.6 0.4 0.2 Placebo + tax/anthra (n=207) Avastin + tax/anthra (n=415) Events, n (%) 73 (35.3) 141 (34.0) Median, months 23.8 25.2 HR (95% CI) 1.03 (0.77–1.38) p=0.83 1-year rate, % 81 83 p=0.44 Proportion alive Last updated September 27, 2009 The survival data are highly censored. At the time of this analysis, 66% of patients randomised to Avastin combined with taxane/anthracycline were still alive. There is no difference in overall survival between the two treatment arms in the taxane/anthracycline cohort, shown by a hazard ratio for overall survival close to 1. It is important to emphasise that the trial was neither designed nor powered to show an overall survival benefit. Furthermore, 55% of patients randomised to taxane/anthracycline plus placebo subsequently received Avastin with their second-line treatment in the open-label, post-progression phase of the study, and this crossover to Avastin at progression is likely to confound more mature survival analyses. References Robert NJ, Dieras V, Glaspy J, et al. RIBBON-1: Randomized, double-blind, placebo-controlled, phase III trial of chemotherapy with or without bevacizumab (B) for first-line treatment of HER2-negative locally recurrent or metastatic breast cancer (MBC). J Clin Oncol 2009;27(15s):Abstract 1005. Roche data on file, 2009. Avastin at progression, % 55 50 Number at risk: Avastin Placebo Months Robert et al. ASCO 2009; Roche data on file 2009 27

28 RIBBON-1: Safety Xeloda Taxane Anthracycline Selected grade ≥3 AEs, %
Placebo (n=201) Avastina (n=404) Placebo (n=102) Avastina (n=203) Placebo (n=100) Avastina (n=210) Bleeding 0.5 0.2 5.4 Febrile neutropenia 2.0 7.9 5.0 3.8 GI perforation 1.0 Hypertension 9.4 8.9 10.0 LV systolic dysfunction 2.9b Neutropenia 1.2 4.9 4.0 4.3 Proteinuria 2.2 3.4 1.9 Sensory neuropathy 3.0 8.8 8.4 Venous thromboembolism 3.5 4.8 2.9 AE leading to deathc 2.5 1.5 1.4 Last updated July 9, 2009 The safety profile of Avastin-containing therapy was generally consistent with previous trials of Avastin in breast cancer. Grade 3–5 hypertension occurred in approximately 10% of patients receiving Avastin in each Avastin group. There appeared to be a slight increase in grade 3–5 neutropenia and febrile neutropenia in the Avastin-containing arm in the taxane subgroup, but this effect was not seen in the Xeloda cohort or the anthracycline subgroup. Venous thromboembolic events were observed less frequently in the Avastin-containing arm than in the placebo arm of the taxane subgroup. References Robert NJ, Dieras V, Glaspy J, et al. RIBBON-1: Randomized, double-blind, placebo-controlled, phase III trial of chemotherapy with or without bevacizumab (B) for first-line treatment of HER2-negative locally recurrent or metastatic breast cancer (MBC). J Clin Oncol 2009;27(15s):Abstract 1005. cExcluding AEs related to mBC progression a15 mg/kg q3w; bGrade 3 or 4 events: clinical CHF (grade 3=1, grade 4=1); non-specific symptoms (n=1); measurement error (n=2); progressive disease (n=1); LV = left ventricular Robert et al. ASCO 2009 28

29 RIBBON-1: Conclusio RIBBON-1 ist die dritte positive Phase III Studie zu first-line Avastin plus Standard-Chemotherapie: Avastin plus Taxane, incl. nab-Paclitaxel Avastin plus nicht-taxanhältige Chemotherapie, incl. Xeloda oder Anthrazykline In allen drei Studien (E2100, AVADO und RIBBON-1), konnte Avastin das PFS und die ORR mit allen Chemotherapieregimen signifikant erhöhen Das Nebenwirkungsprofil war mit dem früherer Phase III-Studien konsistent (insbes. keine kardiale Toxizität von Avastin in Kombination mit Anthrazyklinen) Last updated July 9, 2009 RIBBON-1 is the third positive phase III trial evaluating Avastin combined with chemotherapy in breast cancer.1 The results with the Avastin–taxane combination further support findings from E2100 and AVADO. Data from RIBBON-1 also show that the combination of Avastin with either anthracycline-based therapy or Xeloda is active and well tolerated, significantly improving PFS and response rate versus chemotherapy alone. Reference Robert NJ, Dieras V, Glaspy J, et al. RIBBON-1: Randomized, double-blind, placebo-controlled, phase III trial of chemotherapy with or without bevacizumab (B) for first-line treatment of HER2-negative locally recurrent or metastatic breast cancer (MBC). J Clin Oncol 2009;27(15s):Abstract 1005. 29

30 Avastin: Essentielle Komponente der First-Line Therapie des LR/mBC
In drei Phase III Studien im first-line Setting wurde, unabhängig vom Chemotherapie-Partner: Der primäre Endpunkt (signifikante Verlängerung des PFS durch Avastin + Chemotherapie gegenüber Chemotherapie alleine) erreicht1–3 Die Ansprechraten signifikant erhöht Ein konsistenter Benefit aller Patientenssubgruppen demonstriert Gute Verträglichkeit mit geringen Auswirkungen auf das Safety-Profil der Chemotherapie gezeigt Last updated August 9, 2009 In summary, we now have positive data from three randomised, phase III trials (E2100, AVADO and RIBBON-1) demonstrating significantly improved efficacy (PFS and ORR) when Avastin is combined with chemotherapy.1–3 Individually the three trials each provide convincing evidence of the clinically meaningful and statistically significant benefit of Avastin-containing therapy in the first-line setting. Taken together, they provide a remarkably consistent and robust picture and indicate that Avastin is an essential component of first-line therapy for LR/MBC. Avastin-containing regimens have shown good tolerability, particularly in placebo-controlled trials, and Avastin has limited impact on the safety profile of chemotherapy. References Miller KD, Wang M, Gralow J, et al. Paclitaxel plus bevacizumab versus paclitaxel alone for metastatic breast cancer. N Engl J Med 2007;357:2666–76. Miles DW, Chan A, Romieu G, et al. Randomized, double-blind, placebo-controlled, phase III study of bevacizumab with docetaxel or docetaxel with placebo as first-line therapy for patients with locally recurrent or metastatic breast cancer (mBC): AVADO. J Clin Oncol 2008;26:1008s (Abstract LBA1011). Robert NJ, Dieras V, Glaspy J, et al. RIBBON-1: Randomized, double-blind, placebo-controlled, phase III trial of chemotherapy with or without bevacizumab (B) for first-line treatment of HER2-negative locally recurrent or metastatic breast cancer (MBC). J Clin Oncol 2009;27(15s):Abstract 1005. 1Miller et al. NEJM 2007; 2Miles et al. ASCO 2008; 3Robert et al. ASCO 2009 30

31 (3 positive Phase III Studien)
ZUSAMMENFASSUNG LEVEL 1-Evidenz in MBC (3 positive Phase III Studien) Last updated August 9, 2009 In summary, we now have positive data from three randomised, phase III trials (E2100, AVADO and RIBBON-1) demonstrating significantly improved efficacy (PFS and ORR) when Avastin is combined with chemotherapy.1–3 Individually the three trials each provide convincing evidence of the clinically meaningful and statistically significant benefit of Avastin-containing therapy in the first-line setting. Taken together, they provide a remarkably consistent and robust picture and indicate that Avastin is an essential component of first-line therapy for LR/MBC. Avastin-containing regimens have shown good tolerability, particularly in placebo-controlled trials, and Avastin has limited impact on the safety profile of chemotherapy. References Miller KD, Wang M, Gralow J, et al. Paclitaxel plus bevacizumab versus paclitaxel alone for metastatic breast cancer. N Engl J Med 2007;357:2666–76. Miles DW, Chan A, Romieu G, et al. Randomized, double-blind, placebo-controlled, phase III study of bevacizumab with docetaxel or docetaxel with placebo as first-line therapy for patients with locally recurrent or metastatic breast cancer (mBC): AVADO. J Clin Oncol 2008;26:1008s (Abstract LBA1011). Robert NJ, Dieras V, Glaspy J, et al. RIBBON-1: Randomized, double-blind, placebo-controlled, phase III trial of chemotherapy with or without bevacizumab (B) for first-line treatment of HER2-negative locally recurrent or metastatic breast cancer (MBC). J Clin Oncol 2009;27(15s):Abstract 1005. 31

32 Drei positive randomisierte Phase III Studien: Konsistenter Benefit von First-Line Avastin in mBC
E21001,2: open-label Phase III Studie Avastin + Paclitaxel verdoppelte PFS und ORR in mBC-Patientinnen, was zur Zulassung in diesem Setting führte AVADO3,4: doppelblinde, Plazebo-kontrollierte, Phase III Studie RIBBON-15: doppelblinde, Plazebo-kontrollierte, Phase III Studie Avastin + Standard-Chemotherapie (Taxane, Anthrazyklin-basierte Kombination oder Xeloda) mit signifikanter Verbesserung von PFS und ORR Der Benefit durch Avastin wird um viele Chemotherapie-Schemata im first-line Setting erweitert. Last updated August 31, 2009 The consistently high efficacy and good tolerability of Avastin-containing therapy has been demonstrated in three randomised, phase III trials. The first data showing the significant benefit of combining Avastin with a taxane (weekly paclitaxel) came from E2100 and led to regulatory approval of Avastin in LR/mBC.1,2 These data were subsequently confirmed by results of the AVADO trial3,4 (which led to EU approval of Avastin in combination with docetaxel as first-line therapy) and the RIBBON-1 trial. All of the AVADO data shown in this slide set come from the April 30th 2009 data cut-off. References Miller KD, Wang M, Gralow J, et al. Paclitaxel plus bevacizumab versus paclitaxel alone for metastatic breast cancer. N Engl J Med 2007;357:2666–76. Gray R, Bhattacharya S, Bowden C, et al. Independent review of E2100: a phase III trial of bevacizumab plus paclitaxel versus paclitaxel in women with metastatic breast cancer. J Clin Oncol 2009;27 August (Epub ahead of print). Miles DW, Chan A, Romieu G, et al. Randomized, double-blind, placebo-controlled, phase III study of bevacizumab with docetaxel or docetaxel with placebo as first-line therapy for patients with locally recurrent or metastatic breast cancer (mBC): AVADO. J Clin Oncol 2008;26:1008s (Abstract LBA1011). Avastin SmPC 2009. Robert NJ, Dieras V, Glaspy J, et al. RIBBON-1: Randomized, double-blind, placebo-controlled, phase III trial of chemotherapy with or without bevacizumab (B) for first-line treatment of HER2-negative locally recurrent or metastatic breast cancer (MBC). J Clin Oncol 2009;27(15s):Abstract 1005. 1Miller et al. NEJM 2007; 2Gray et al. JCO 2009; 3Miles et al. ASCO 2008; 4Avastin SmPC 2009; 5Robert et al. ASCO 2009 32

33 Positive Phase III Studien: Designs
RIBBON-12 E21001 Xeloda Taxane/ anthracycline Placebo controlled No Yes Chemotherapy Weekly paclitaxel 3-weekly docetaxel/nab-paclitaxel or AC/FAC/EC/FEC Avastin dose 10 mg/kg q2w 15 mg/kg q3w Primary endpoint PFS (inv) IRF review Retrospective Prospective Last updated August 9, 2009 This slide highlights similarities and differences in the trial designs of E2100, AVADO and RIBBON-1. All were conducted in the first-line setting and the primary objective of all three trials was to demonstrate superior PFS with Avastin-containing therapy versus chemotherapy alone. Both AVADO and RIBBON-1 were placebo controlled, whereas E2100, the first of the trials to be completed, was open label with no placebo in the control arm. The trials provide data for Avastin in combination with a range of commonly used first-line chemotherapy regimens: Weekly paclitaxel in E2100 3-weekly docetaxel in AVADO In RIBBON-1, clinicians selected their preferred chemotherapy regimen (docetaxel, nab-paclitaxel, one of four anthracycline-based combination regimens, or Xeloda monotherapy) before randomisation to either Avastin or placebo with chemotherapy. Patients in the anthracycline cohort received one of the following four regimens: AC (doxorubicin 50–60 mg/m2, cyclophosphamide 500–600 mg/m2) EC (epirubicin 90–100 mg/m2, cyclophosphamide 500–600 mg/m2) FAC (5-FU 500 mg/m2, doxorubicin 50 mg/m2, cyclophosphamide 500 mg/m2) FEC (5-FU 500 mg/m2, epirubicin 90–100 mg/m2, cyclophosphamide 500 mg/m2). In all trials, the Avastin dose was the standard breast cancer dose of 5 mg/kg/week equivalent (either 10 mg/kg every 2 weeks or 15 mg/kg every 3 weeks). AVADO included an additional arm evaluating Avastin 7.5 mg/kg every 3 weeks combined with docetaxel. As shown later, the latest results from this trial suggest that the approved breast cancer dose of 15 mg/kg every 3 weeks is the most appropriate, although the AVADO trial was not powered to compare the two Avastin doses. References Miller KD, Wang M, Gralow J, et al. Paclitaxel plus bevacizumab versus paclitaxel alone for metastatic breast cancer. N Engl J Med 2007;357:2666–76. Miles DW, Chan A, Romieu G, et al. Randomized, double-blind, placebo-controlled, phase III study of bevacizumab with docetaxel or docetaxel with placebo as first-line therapy for patients with locally recurrent or metastatic breast cancer (mBC): AVADO. J Clin Oncol 2008;26:1008s (Abstract LBA1011). Robert NJ, Dieras V, Glaspy J, et al. RIBBON-1: Randomized, double-blind, placebo-controlled, phase III trial of chemotherapy with or without bevacizumab (B) for first-line treatment of HER2-negative locally recurrent or metastatic breast cancer (MBC). J Clin Oncol 2009;27(15s):Abstract 1005. 1Miller et al. NEJM 2007; 2Robert et al. ASCO 2009 33

34 Taxane/ anthracycline
Positive Phase III Studien: Patientenpopulationen RIBBON-12 E21001 Xeloda Taxane/ anthracycline N 673 615 622 ER and PgR negative 35% 24% ≥3 metastatic sites 45% 44% Measurable disease 73% 80% 84% Adjuvant chemotherapy 65% 72% Adjuvant taxane 16% 40% 15% Last updated August 9, 2009 The patient populations were generally similar in E2100, AVADO and RIBBON-1.1–3 Prior chemotherapy for LR/mBC was not permitted in any of the trials. The proportion of patients with at least three metastatic sites was very consistent between trials (44–46%). E2100 included a slightly higher proportion of patients with ER- and PgR-negative disease compared with the other trials. The major difference in baseline characteristics was seen in the Xeloda cohort of RIBBON-1, which included a considerably higher proportion of patients previously exposed to adjuvant chemotherapy, including taxane-containing regimens in 40%. This is as expected because Xeloda is an established treatment option in patients pretreated with taxane-containing therapy. References Miller KD, Wang M, Gralow J, et al. Paclitaxel plus bevacizumab versus paclitaxel alone for metastatic breast cancer. N Engl J Med 2007;357:2666–76. Miles DW, Chan A, Romieu G, et al. Randomized, double-blind, placebo-controlled, phase III study of bevacizumab with docetaxel or docetaxel with placebo as first-line therapy for patients with locally recurrent or metastatic breast cancer (mBC): AVADO. J Clin Oncol 2008;26:1008s (Abstract LBA1011). Robert NJ, Dieras V, Glaspy J, et al. RIBBON-1: Randomized, double-blind, placebo-controlled, phase III trial of chemotherapy with or without bevacizumab (B) for first-line treatment of HER2-negative locally recurrent or metastatic breast cancer (MBC). J Clin Oncol 2009;27(15s):Abstract 1005. ER = oestrogen receptor; PgR = progesterone receptor 1Miller et al. NEJM 2007; 2Robert et al. ASCO 2009 34

35 Taxane/ anthracycline
Positive Phase III Studien: PFS und ORR RIBBON-14 E21001,2a Xeloda Taxane/ anthracycline Pacl Avastin + pacl Placebo + Xeloda Avastin + Xeloda Placebo + t/a Avastin + t/a ORR, % 22 50 24 35 38 51 p<0.0001 p=0.0097 p=0.0054 Median PFS, months 5.8 11.3 5.7 8.6 8.0 9.2 HR 0.48 0.69 p=0.0002 0.64 Last updated September 11, 2009 As shown in this slide, the efficacy of Avastin-containing therapy is very consistent across the three trials. All PFS data shown in this slide are from the stratified analyses. In E2100, AVADO and both cohorts of RIBBON-1, PFS was significantly superior in the Avastin-containing arm compared with chemotherapy alone, thus meeting the primary objective of each trial.1–3 Median PFS was approximately 9–11 months in all four Avastin-containing arms.1–3 The response rate with Avastin in combination with a taxane was 49–64%. The response rate was 35% in the Avastin–Xeloda cohort of RIBBON-1.1–3 References Klencke B, Bhattacharya S, Samant M, et al. Independent review of E2100 validates progression-free survival (PFS) improvement with the addition of bevacizumab (B) to paclitaxel (P) as initial chemotherapy for metastatic breast cancer (MBC). J Clin Oncol 2008;26(May 20 Suppl.):50s (Abstract 1036). Gray R, Bhattacharya S, Bowden C, et al. Independent review of E2100: a phase III trial of bevacizumab plus paclitaxel versus paclitaxel in women with metastatic breast cancer. J Clin Oncol 2009;27 August (Epub ahead of print). Avastin SmPC 2009 Robert NJ, Dieras V, Glaspy J, et al. RIBBON-1: Randomized, double-blind, placebo-controlled, phase III trial of chemotherapy with or without bevacizumab (B) for first-line treatment of HER2-negative locally recurrent or metastatic breast cancer (MBC). J Clin Oncol 2009;27(15s):Abstract 1005. aIRF assessment; bPFS censored for non-protocol therapy before disease progression; c15 mg/kg q3w; dExploratory p value 1Klencke et al. ASCO 2008; 2Gray et al. JCO 2009; 3Avastin SmPC 2009; 4Robert et al. ASCO 2009 35

36 Taxane/ anthracycline
Positive Phase III Studien: OS (sek. Endpunkt) RIBBON-14 E21001 Xeloda Taxane/ anthracycline Pacl Avastin pacl Placebo + Xeloda Avastin + Xeloda Placebo + t/a Avastin + t/a Median OS, months 24.8 26.5 21.2 29.0 23.8 25.2 HR 0.87 p=0.14 0.85 p=0.27 1.03 p=0.83 1-year OS rate, %b 74 81 83 p=0.017 p=0.076 p=0.44 Last updated September 26, 2009 There was no significant difference in overall survival in any of the trials. It is important to emphasise that none of the trials was powered to demonstrate an overall survival benefit, and analysis of overall survival is confounded by crossover of a considerable proportion of patients to Avastin after progression or unblinding. In addition, subsequent lines of therapy may obscure potential effects on overall survival. Note that the overall survival data for AVADO are the unstratified analysis, as reported in the SmPC. The overall survival data for RIBBON-1 are highly censored, with at least two-thirds of patients still alive in each treatment arm. References Cameron D. Bevacizumab in the first-line treatment of metastatic breast cancer. Eur J Cancer Suppl 2008;6:21–8. Avastin SmPC 2009. Robert NJ, Dieras V, Glaspy J, et al. RIBBON-1: Randomized, double-blind, placebo-controlled, phase III trial of chemotherapy with or without bevacizumab (B) for first-line treatment of HER2-negative locally recurrent or metastatic breast cancer (MBC). J Clin Oncol 2009;27(15s):Abstract 1005. a15 mg/kg q3w; bExploratory p value 1Cameron EJC Suppl 2008; 2Avastin SmPC 2009; 3Roche data on file 2009; 4Robert et al. ASCO 2009 36

37 Taxane/ anthracycline
Drei positive Phase III Studien: PFS-Subgruppen RIBBON-12b E21001,2a Xeloda Taxane/ anthracycline HR Median, months Median, months Pac Avastin+ pac Pla + Xeloda Avastin + Xeloda Pla + t/a Avastin + t/a Overall n=722 n=615 n=622 0.54 5.8 11.3 0.67 5.7 8.6 0.66 8.0 9.2 Age ≥65 years n=167 n=153 n=124 0.69 7.4 10.4 6.2 9.1 0.83 8.5 10.1 Triple negative n=232 n=137 n=142 0.49 5.3 10.6 0.72 4.2 6.1 0.78 6.5 Adjuvant taxane n=245 n=94 0.33 13.1 0.62 8.7 0.65 6.7 Last updated August 31, 2009 Several subpopulation analyses of all three trials have been conducted and presented. All subgroups analysed have been shown to derive benefit from the addition of Avastin to chemotherapy. Some of the subgroups of particular interest to clinicians are shown in this table. Although there are differences in the magnitude of the benefit, the overall picture of a beneficial effect of Avastin is consistent. Reference Cameron D. Bevacizumab in the first-line treatment of metastatic breast cancer. Eur J Cancer Suppl 2008;6:21–8. Roche data on file 2009. aUnstratified analyses; bStratified analyses 1Cameron EJC Suppl 2008; 2Roche data on file 2009 37 37

38 Bevacizumab in combination with chemotherapy - data from the general oncology practice (ATHENA)

39 Avastin plus taxane-based chemotherapy safety study (ATHENA, MO19391): study design
Avastin (10mg/kg q2w or 15mg/kg q3w) + taxane-based chemotherapy* Previously untreated HER2-negative LR or mBC (n=2,027) Treat to disease progression Last updated: 20 January, 2009 This slide shows the study design and patient recruitment for the single-arm, open-label, multicentre MO19391 study.1 Previously untreated HER2-negative LR or mBC patients received Avastin at either 10mg/kg q2w or 15mg/kg q3w in combination with a taxane-based chemotherapy regimen until disease progression, unacceptable toxicity or withdrawal of patient consent. The choice of taxane and treatment regimen were at the discretion of the investigator. If taxanes were contraindicated or not the standard of care of the treating physician, the patient could receive Avastin in combination with the physician’s standard of care regimen. The use of anthracyclines was not permitted. The primary objective was to assess the safety profile of Avastin in combination with taxanes (as measured by the incidence of National Cancer Institute Common Toxicity Criteria version 3.0 grade 1–5 AEs and serious AEs). Secondary objectives were to assess the efficacy of Avastin in combination with taxanes (as measured by TTP and OS) and to assess the safety of Avastin in patients who develop CNS metastases during, and for 6 months following, the treatment period. As of 29 August 2008, the safety population consisted of 2,027 patients. AEs = adverse events; CNS = central nervous system; HER2 = human epidermal growth factor receptor 2; q2w = every 2 weeks; q3w = every 3 weeks; LR = locally recurrent; mBC = metastatic breast cancer; OS = overall survival; TTP = time to disease progression Smith IE, et al. Primary analysis of study MO19391, a 2000 patient open-label safety study of bevacizumab plus taxane-based therapy as first-line treatment of patients with locally recurrent (LR) or metastatic breast cancer (mBC). Poster presented at SABCS Cancer Res 2009;69(Suppl. 2):292s (Abstract 4118). Primary endpoint: safety secondary endpoints: TTP, OS, safety in patients with CNS metastases *Taxane use according to routine practice or, if taxanes contraindicated, alternative chemotherapy allowed at physician’s discretion Anthracyclines were not permitted Smith, et al. SABCS 2008

40 Avastin plus taxane-based chemotherapy safety study (ATHENA): eligibility
Inclusion criteria age 18 years histologically/cytologically confirmed, HER2-negative, LR or mBC ECOG PS 0–2 Exclusion criteria previous chemotherapy for LR or mBC concomitant hormonal therapy for LR or mBC increased risk of haemorrhage non-healing wound, active peptic ulcer or bone fracture uncontrolled hypertension history of fistula major surgery within 28 days prior to initiation of study therapy Last updated: 20 January, 2009 This slide describes the main eligibility criteria for MO Eligible patients were aged 18 years and had histologically/cytologically confirmed, HER2-negative, LR or mBC. HER2-positive LR or mBC patients could be included if they had progressed after previous Herceptin® therapy. An ECOG PS of 0–2 was required, together with adequate haematological, liver and renal function and a life expectancy 12 weeks. Previous chemotherapy for LR or mBC was not permitted, although prior adjuvant chemotherapy was allowed. Concomitant hormonal therapy for LR or mBC was also not permitted. Patients with increased risk of haemorrhage (use of therapeutic anticoagulants, bleeding diathesis or coagulopathy), non-healing wound, active peptic ulcer or bone fracture, uncontrolled hypertension or history of fistula were excluded as were those patients who had undergone major surgery within 28 days prior to initiation of study therapy. ECOG = Eastern Cooperative Oncology Group; HER2 = human epidermal growth factor receptor 2; LR = locally recurrent; mBC = metastatic breast cancer; PS = performance status Smith IE, et al. Primary analysis of study MO19391, a 2000 patient open-label safety study of bevacizumab plus taxane-based therapy as first-line treatment of patients with locally recurrent (LR) or metastatic breast cancer (mBC). Poster presented at SABCS Cancer Res 2009;69(Suppl. 2):292s (Abstract 4118). Smith, et al. SABCS 2008

41 Athena: chemotherapy regimens
The safety data presented were generated for a review by the Data Safety Monitoring Board, using a data cut-off of April At this time, the safety population consisted of 230 patients and the median duration of therapy was 43 days (range 13–150). Serious adverse events were reported in 37 patients (16.1%). The most frequently occurring of these were neutropenia (12 patients, 5.2%) and febrile neutropenia (eight patients, 3.5%). CNS = central nervous system; mBC = metastatic breast cancer Reference Smith IE, Biganzoli L, Cortes-Funes H, et al. MO19391: an open-label safety study of bevacizumab plus taxane monotherapy or in combination as first-line treatment of patients with locally recurrent or metastatic breast cancer (LR or mBC). Eur J Cancer Suppl 2007;5:221 (Abstr 2123). *Paclitaxel + Xeloda, Gemcitabine, Carboplatin, Doxorubicin, Epirubicin, Vinorelbine **Docetaxel + Xeloda, Gemcitabine, Carboplatin, 5-FU, Cisplatin, Doxorubicin, Cyclophosphamid, Epirubicin, Vinorelbine Smith, et al. SABCS 2008

42 MO19391 (Athena): Efficacy median TTP = 9.5 Monate
Smith, et al. SABCS 2008

43 ATHENA (MO19391): Conclusio
Mit >2000 Pat ist ATHENA die größte Studie mit Avastin + Chemo in 1st line lr/mBC 75% der Patienten erhielten eine taxan-basierte Chemotherapie mit Avastin Safety Daten sind mit denen von AVADO und E2100 konsistent (geringer Impact auf Safety-Profil der Chemo TTP ist mit PFS-Daten der Phase III Studien (E2100 und AVADO) konsistent Bestätigung der Safety und Wirksamkeit von Avastin in Kombination mit Chemotherapie in MBC im „real world clinical setting“ In breast cancer, the efficacy of Avastin has been demonstrated in combination with paclitaxel in a phase III trial1 and with docetaxel in phase II trials.2,3 Data from these studies revealed that Avastin can be effectively combined with taxanes to produced regimens that are well tolerated. Further phase II studies have demonstrated that Avastin can be partnered with other agents that commonly form part of standard of care regimens in mBC.4,5 MO19391 is a large, community-based safety study that will further evaluate the safety and efficacy of Avastin in combination with taxane-based regimens (or other standard of care regimens at the discretion of the investigator) and provide greater numbers of patients with access to this novel agent. PFS = progression-free survival; mBC = metastatic breast cancer References EMEA. Avastin European Public Assessment Report. Available at: Accessed 2 August 2007. Ramaswamy B, Elias AD, Kelbick NT, et al. Phase II trial of bevacizumab in combination with weekly docetaxel in metastatic breast cancer patients. Clin Cancer Res 2006;12:3124–9. Lyons JA, Silverman P, Remick S, et al. Toxicity results and early outcome data on a randomized phase II study of docetaxel ± bevacizumab for locally advanced, unresectable breast cancer. J Clin Oncol 2006;24(20 June suppl.):133s (Abstract 3049). Burstein HJ, Parker LM, Savoie J, et al. Phase II trial of the anti-VEGF antibody bevacizumab in combination with vinorelbine for refractory advanced breast cancer. Breast Cancer Res Treat 2002;79:S115 (Abstract 446). Miller KD, Chap LI, Holmes FA, et al. Randomised phase III trial of capecitabine compared with bevacizumab plus capecitabine in patients with previously treated metastatic breast cancer. J Clin Oncol 2005;23:792–9. Avastin Summary of Product Characteristics (SmPC); 2. Miller, et al. NEJM 2007; 3. Miles, et al. ASCO 2008; Smith, et al. SABCS 2008