Avances Terapéuticos en el Tratamiento de los Tumores Ginecológicos: Papel de la Angiogénesis Ana Oaknin, MD Medical Oncology Department. Gyneacological Cancer Unit Vall d´Hebron University Hospital. Vall d’Hebron Institute of Oncology Barcelona, Spain aoaknin@vhio.net

Gynecologic Tumors(GYNT): Epidemiology GYNT comprises 19% of new cancer cases each year. Endometrial cancer is the GYNT most common. Cervical Cancer is the second most common cancer in women after breast cancer . Ovarian cancer is the fifth cause of cancer death and the most lethal GYNT Source: American Cancer Society. Cancer Facts&Figures 2013.

Tumour Vasculature effect The angiogenic switch in tumour development Tumour Vasculature effect Angiogenic switch Results in over-expression of pro-angiogenic signals, such as VEGF Small tumour (1–2mm) Avascular Dormant Larger tumour Vascular Metastatic potential It is generally accepted that no solid tumour can grow beyond a critical size of 1–2mm or about 106 cells without an adequate vascular supply. The angiogenic ‘switch’ is reviewed in Bergers et al.1 Tumour-associated angiogenesis can be thought to go through two phases, which are separated by the ‘angiogenic switch’. The first is an ‘avascular’ phase, in which tumours are not more than 1–2mm in diameter. These tumours remain dormant, as they are in a steady state between cell proliferation and apoptosis. Some of these small tumours will switch to the second – vascular – phase, in which the tumour grows exponentially. A classic model of the angiogenic switch is that of a pair of scales laden with pro-angiogenic molecules on one side and anti-angiogenic molecules on the other side. An imbalance of these pro- and anti-angiogenic factors triggers the ‘switch’. The onset of angiogenesis or the ‘angiogenic switch’ is a discrete step that can occur at any stage of tumour progression, dependent upon the type of tumour and its micro-environment. Induction of the switch is dependent upon the balance tipping in favour of the pro-angiogenic molecules.   Bergers G, Benjamin LE. Tumorigenesis and the angiogenic switch. Nature Reviews Cancer 2003;3:401–10. Bergers G. et al . Nature Reviews Cancer 2003;3:401–10. Bergers, et al. Nature 2002

Therapeutic Targeting of Angiogenesis One of the most prolific arenas of drug development More than 360 agents in various phases of development Compounds: modeled for direct and/or indirect AA properties Approaches: ligand, receptor, signal, and regulators Targets: endothelial cells, tumor cells, pericytes Approved: 9 (2 others with indirect effects) Most recent Everolimus (3/30/09) in RCC

Ovarian Cancer

Key Signaling Pathways Involved in Ovarian Cancer Angiogenesis The clinical efficacy of angiogenesis inhibitors targeting VEGF marked a milestone in the field of angiogenesis research however overlapping and compensatory alternative angiogenic pathways provide escape mechanisms that likely limit the Angiopoietin-1 (Ang-1) and Angiopoietin-2 (Ang-2) are functional ligands of the Tie2 receptor tyrosine kinase that is expressed on endothelial cells . Ang-1 is expressed by pericytes, smooth muscle cells and fibroblasts and acts in a paracrine manner as a physiological agonist of Tie2. It acts as a maturation factor that stabilizes the mature vasculature by promoting recruitment of pericytes and smooth muscle cells [5]. In contrast, Ang-2 is expressed by endothelial cells and stored in Weibel-Palade-bodies. It acts as an antagonist of Tie2 by blocking Ang-1 dependent Tie2 activation Ang-2 is further described to act as a functional destabilization factor, rendering vasculature in a more plastic state amenable to sprouting (under the influence of other angiogenic cytokines such as VEGF) and is found to be particularly increased in highly vascularized tumors and in pro-angiogenic diseases (e.g. macula degeneration, rheumatoid arthritis, osteoarthritis, psoriasis) [10]. Ang-2 is primarily expressed in endothelial vasculature and only rarely in tumor cells. A shift in the Ang-1-to-Ang-2 expression ratio in favor of Ang-2 was found to be associated with tumor angiogenesis and poor prognosis [11–13]. Ang-2 therefore is considered as a major player of the angiogenic switch in the course of tumor progression. Ang-2 is therefore considered pivotal for the angiogenic switch in the course of tumor progression and, together with VEGF-A, promotes the initiation of angiogenesis and maturation of new vessels. High levels of Ang-2 correlate with increased metastatic and invasive potential in breast cancer, malignant melanoma and lung cancer In contrast to its broad expression in the vasculature of human tumors, Ang-2 shows limited postnatal expression in normal tissue making it a tumor specific target for anti-angiogenic therapies. Recently, different approaches have been described to target the Angiopoietin axis including Tie2-kinase inhibitors, Ang1 binding to Tie2 activates downstream signaling pathways (Fig 1B), which promote blood vessel stability by enhancing interactions between perivascular cells and endothelium, enhancing endothelial cell survival and leading to a more stable vasculature with decreased permeability. Ang2 expression is tightly regulated, and it is synthesized and secreted primarily by endothelial cells at sites of vascular remodeling, in response to proinflammatory stimuli,14 other proangiogenic cytokines (eg, VEGF, insulin-like growth factor 1, and platelet-derived growth factor B), or hypoxia. Ang2 interrupts Ang1 mediated vessel normalization, resulting in impaired pericyte coverage,vessel destabilization, and increased vascular permeability, conditions conducive for sprouting angiogenesis. Studies suggest that for angiogenesis to progress, additional factors, such as VEGF, must be present. Interestingly, Ang2may also play an agonistic role, bfull potential of VEGF monotherapies Yap TA, et al. Nat Cancer Rev 2009; 9:2583; J Clin Oncol, Vol 30, Nº 4, 2012; 441-444

VEGF and Angiopoetin-2 Levels Markers of Poor Prognosis in Ovarian Cancer El porque en cancer de ovario la angigogenesis tumoral es tan importante, es debido a que en estudios experimentales se ha demostrado que una relación directa entre la expresión de VEGF y el grado de angiogenesis tumoral marcado (por la densidad de microvasos) con la mayor formación de ascitis, un menor tiempo a la progresión y menor supervivencia tras recibir la primera línea de tratamiento. Por el contrario, El tratamiento en modelos xenografts de OvCa con un anticuerpo frente al VEGF , ha mostrado una resolución de la ascitis, enlentecimiento de la progresión tumoral, y algo muy importante como vemos en esta diapositiva que es la normalización de la vasculatura aberrante, con lo cual mejoraría la penetración de los citotóxicos en el tumor. Duncan: Using a tissue microarray of 339 primary ovarian cancers, the expression of VEGF was assessed immunohistochemically. Coupled to a comprehensive database of clinicopathologic variables, its effect on these factors and survival was studied. Results: Tumors expressing high levels of VEGF had significantly poorer survival (P = 0.04). Factors shown to predict prognosis independently of each other were age, International Federation of Gynecologists and Obstetricians stage, and the absence of macroscopic disease after surgery. VEGF was independently predictive of prognosis on multivariate analysis (P = 0.02). There was no correlation between VEGF and any clinicopathologic variable. High expression of VEGF was seen in only 7% of the tumors, suggesting that the role of antiangiogenic drugs may be limited to a small subset of patients. Duncan et al; Clin Cancer Res 2008; 14:3030-3035; Sallinen et al; Int Journal Gyn Oncol; 20; 2010. 7

Targeting the VEGF Pathway Strategies Soluble VEGF receptors- aflibercept Antibodies to VEGF Bevacizumab VEGF VEGFR2 VEGFR tyrosine kinase inhibitors: Pazopanib BIBF1120 Cediranib Therefore, we have developed many strategies to inhibit the Angiogenesis. Monoclonal antibodies against VEGF, namely Bevacizumab or small molecules that inhibit the TK domain of the Receptors. To date , the larger amount of clinical evidence in ovarian cancer comes from Bevacizumab. Ribozymes Migration, permeability, DNA synthesis, survival 8

Angiogenesis as a Target in Ovarian Cancer: Anti-VEGF-VEGFR Pathway Anti-vascular endothelial growth factor (VEGF) therapy improves progression-free survival (PFS) GOG 218* Front-line: Bevacizumab HR = 0.72; 95% CI, 0.63–0.821 ICON 7* Front-line: Bevacizumab HR = 0.81; 95% CI, 0.70–0.942 AGO-OVAR12 Front-line: Nintedanib HR = 0.84; 95% CI, 0.72, 0.983 AGO-OVAR16 Maintenance: Pazopanib HR = 0.77; 95% CI, 0.64–0.914 AURELIA** Platinum-resistant, recurrent / 1 or 2 prior regimens: Bevacizumab HR = 0.48; 95% CI, 0.38–0.605 OCEANS* Platinum-sensitive, recurrent / 1 prior regimen: Bevacizumab HR = 0.48; 95% CI, 0.41–0.706 ICON6 Platinum-sensitive, recurrent / 1 prior regimen: Cediranib HR = 0.57; 95% CI, 0.44–0.747 *EMA Approved Burger RA et al. N Engl J Med. 2011;365:2473‒2483. Perren TJ et al . N Engl J Med. 2011;365:2484‒2496. du Bois A et al. J Clin Oncol. 2013;31(18suppl):LBA5503. du Bois A et al. LBA ESGO 2013 Liverpool, UK Pujade-Lauraine E et al. J Clin Oncol. 2012;30(18suppl):LBA5002. Aghajanian C et al. J Clin Oncol. 2012;30:2039‒2045. Ledermann JA et al . Eur J Cancer. 2013;49(suppl):LBA *FDA Approved HR = hazard ratio; 95% CI = confidence interval

Single- Agent BEVACIZUMAB in Ovarian Cancer: Phase II Efficacy Results Parameter, % Cannistra et al.[1] (N = 44) Garcia et al.2] (N = 70) Burger et al.[3] (N = 62) GOG#170 Prior regimens 1 2 3 52 48 100 24 66 1o platinum DFI < 6 mos, % 84 40 60 Dose/schedule 15 mg/kg every 21 days 10mg/kg plus Ciclophosphamide 50mg/day RR, n (%) 7 (16%) 17(24%) 13(21%) ≥ G3 Proteinuria (%) <16 ≥ G3 HTA(%) 14 10 Gastrointestinal perforations (%) 11 6 There is no doubt that the most developed antiangiogenesis agent in Ovarian cancer is Bevacizumab so far. Bevacizumab is very well known for all of you. It is a monoclonal antibody against all VEGF isoforms. By contrast with other solid tumors, Bevacizumab has shown activity as a single agent in ovarian cancer. The first clinical data come from 2 phase II clinical trials. Bevacizumab showed objective response rates that range from 16% to 21% in a very poor prognosis population since the majority of them had a platinum resistant relapse. These results compared well with historic data from GOG trials in resistant relapse in ovarian cancer. Based on these results, Bevacizumab was listed in NCCC guidelines as an active therapy in recurrent ovarian cancer. Following these promising results, the development of BEV in ovarian cancer moved forward both in first line and in the relapsed setting. Today, we will focus only in first line treatment. . 1. Cannistra SA, et al. J Clin Oncol. 2007;25:5180-5186. 3. Burger RA, et al. J Clin Oncol. 2007;25:5165-5171 2.Garcia AA, et al. J Clin Oncol. 2008;26:76-82. 10 10

GOG#218 ICON-7 Two landmark studies have studied the role of Bevacizumab in first line treatment in advanced ovarian cancer. The GOG 218 and ICON- 7 . The GOG was run in the USA and Corea and the ICON-7 in Europe and Canada. Both studies were published in the NEJM in 2011. Let´s start to review the GOG 218 since it is the most solid trial and had served as pivotal study to be submitted to Regulatory Agencies.

GOG#218: Schema I II III Arm 1ºEnd-Point: PFS PFS HR0.77 2008 GOG#218: Schema 15 months Paclitaxel (P) 175 mg/m2 Carboplatin (C) AUC 6 Placebo I Arm Cytotoxic (6 cycles) Maintenance (16 cycles) (CP) Front-line: Epithelial OV, PP or FT cancer Stage III optimal (macroscopic) Stage III suboptimal Stage IV N=1873 RANDOM I Z E 1:1:1 Carboplatin (C) AUC 6 Paclitaxel (P) 175 mg/m2 Placebo BEV 15 mg/kg II (CP + BEV) The GOG 218 is a Phase III , double blinded randomized trial with three arms that analyze if adding bevacizumab to the standard arm improves the results. The eligible patients were patients with Stage IV, stage III with suboptimal surgery and stage II with optimal surgery but with macroscopic residual disease.. As you know the poorest prognostic group in ovarian cancer. The patients were randomized to receive standard treatment with PC + placebo, standard treatment plus concurrent BEVACIZUMAB followed by placebo or paclitaxel/carboplatin concurrent with BVZ and followed by Bevacizumab as maintenance treatment for a total of 15 months. I would like to point out several issues. Bevacizumab was administered at 15 mg/kg. This is the dose used in the GOG Phase II studies. The treatment with Bevacizumab began with the second course of chemotherapy in order to prevent would healing complications. The treatment was stopped in course six cycle in arm 2 and upon 22 courses in Arm III. These data are important to keep in mind since they are important differences with the European study. Well, the primary end-point of the study was Progression Free Survival. The objective of the trial was to show that the incorporation of BEV to standard treatment would be able to reduce the risk of recurrence compare with standard treatment. Remarkably, according to the statistical design the standard arm was compared with both experimental arms, however only if both experimental arms were positive we could establish a comparison between them. The PFS were determined by RECIST Criteria, CA125 and clinical deterioration. BEV 15 mg/kg Carboplatin (C) AUC 6 Paclitaxel (P) 175 mg/m2 III (CP + BEV  BEV) 1ºEnd-Point: PFS PFS HR0.77 M. PFS shift: 14.0 m  18.2M N Engl J Med 2011;365:2473-83., 12 12

GOG-0218: PFS RECIST, global clinical deterioration, or CA-1251 Arm I CP (n=625) Patients with event, n (%) 423 (67.7) Median PFS, months 10.3 Stratified analysis HR (95% CI) One-sided p-value (log rank) Arm II CP + BEV (n=625) 418 (66.9) 11.2 0.908 (0.759–1.040) 0.080a Arm III CP + BEV  BEV (n=623) 360 (57.8) 14.1 0.717 (0.625–0.824) <0.0001a 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 Proportion surviving progression free Let´s see the results Regarding the primary end-point , PFS. When we compare the Arm II to Standard Arm, the incorporation of Bevacizumab only during the chemotherapy, does not improve the results. However, when Bevacizumab is administered concurrent and later as maintenance therapy, the study met its primary end-point. Adding Bevacizumab reduces the risk of recurrence in a 28% with a HR of 0.71. The median PFS was 10 months in standard arm and 14 months in the experimental arm. In addition, I would like to remark that we should not draw erroneous conclusions. I mean, since the concurrent arm was negative, we cannot conclude that the ARM III is positive only because of maintenance phase. We only would have stated that if the trial had had a four arm based on Maintenance Bevacizumab only. I would like to bring your attention to the time when Bevacizumab was stopped. As you can see this is the point were the curves achieved a maximum separation. We will go back over this issue later. CP (Arm I) + BEV (Arm II) ap-value boundary = 0.0116 + BEV → BEV maintenance (Arm III) 0 12 24 36 Months since randomization N Engl J Med 2011;365:2473-83. 13

GOG-0218 CA-125 To Determine Progression Protocol-defined PFS analysis CA-125-censored PFS analysis Median PFS CP (Arm I) 10.3 months 12.0 months CP + BEV  BEV (Arm III) 14.1 months 18.0 months Absolute diff. median PFS 3.8 months 6.0 months Hazard ratio 0.717 0.645 Censored for CA125, % 20 29 Since the GOG 218 was a pivotal study, the Regulatory Agency requested an analysis censoring the progression events exclusively determined by CA125. The results of this analysis even showed better results with a HR OF 0.64. Once again, adding Bevacizumab reduced the risk of recurrence in 36% an improved the PFS by 6 months.

Overall survival at time of final PFS analysis 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 Proportion alive Arm I CP + PLA → PLA (n=625) Arm II CP + BEV → PLA Arm III CP + BEV  BEV (n=623) Patients with events, n (%) 156 (25) 150 (24) 138 (22) Median (months) 39.3 38.7 39.7 HR (stratified) (95% CI) 1.036 (0.827–1.297) 0.915 (0.727–1.152) One-sided p-value 0.361 0.252 1-year OS rate, % 90.6 90.4 91.3 According to the Protocol, the Median Overall Survival was determined at the time of PFS. As you can see in the slide, there were no differences among the three arms. We will analyze the potential reasons for these results later during this talk.   0 6 12 18 24 30 36 42 48 Months since randomization 625 623 442 432 437 173 162 171 46 39 40 No. at risk N Engl J Med 2011;365:2473-83.,

Epithelial ovarian, primary peritoneal or fallopian tube cancer ICON7: Study design Epithelial ovarian, primary peritoneal or fallopian tube cancer High-risk stage I–IIA (grade 3 or clear cell) Stage IIB–IV n=1528 Carboplatin AUC 6* (C) Paclitaxel 175 mg/m2 (P) R 1:1 Carboplatin AUC 6* Paclitaxel 175 mg/m2 (P) Bevacizumab 7.5 mg/kg q3w It is time to move to the European/Canadian Study, The ICON-7 Study. Which are the main differences with the GOG218? First all, the ICON-7 Study is an open label Phase III with 2 arms. The standard arm compared with experimental arm. As you can see in the slide, Bevacizumab was administered at 7.5 mg, half of GOG doses. In addition, the duration of the treatment was only 12 months compared to 15 months in the GOG.   The primary end-point was also the Progression-free survival (PFS) but in this trial exclusively defined by RECIST guidelines .CA-125 elevation alone not defined as disease progression. I think to understand both studies it is important to remark the population enrolled in the ICON-7. This study included a 10% of high-risk early stages , 70% had an optimal surgery according the criteria previous 4th ovarian cancer consensus and a 25% had suboptimal surgery . The study pre-defined a high-risk subgroup that included the Stages III with residual disease greater than 1 cm and Stages IV. This group was a 31% of the global population. In addition, as you know, this is exactly the same population enrolled in the GOG 218. Stratification variables: Stage I–III debulked ≤1 cm vs stage I–III debulked >1 cm vs stage IV and inoperable stage III Intent to start treatment ≤/> 4 weeks after surgery 1º End- Point: PFS only determined by RECIST Criteria 12 months HIGH RISK GROUP( 31% ) : FIGO III>1cm/FIGO IV Debulking Perren TJ et al . N Engl J Med. 2011;365:2484‒2496. 16

Summary of Updated Results High risk: FIGO IV or III with >1cm RD HR 0.68 (95% CI 0.55-0.85) Median 10.5 vs 15.9mo HR = 0.81 (95% CI 0.70–0.94) The study met its primary end-point. The addition of Bevacizumab to standard arm reduces the risk of recurrence with a HR of 0.81. The median PFS was 19.8 months in bev arm vs 17.4 months in the standard therapy( 2.4 months). These results are considered modest. However, when we focused on the high-risk group the benefit is much greater and comparable to the GOG 218. Adding BEV reduces the risk of recurrence in a 32%, with a HR of 0.68. Interestingly, adding bev also brings a benefit in terms of OS HR 0.64 (95% CI 0.48-0.85) Median 36.6 vs 28.8mo HR = 0.81 (95% CI 0.63–1.04) Perren T et al N Engl J Med. 2011 Dec 29;365(26):2484-96.

GOG-0218 and ICON7: Efficacy Across the Trials (Arm I vs III) ICON7 Results of primary analysis n PFS (HR) Median PFS, months PFS ∆ , months p value 1,248 0.72 10.3 vs 14.1 3.8 <0.0001 1,528 0.87 17.4 vs 19.8 2.4 0.04 Results based on RECIST only (regulatory analysis*) 0.65 12.0 vs 18.0 6 Subgroup of advanced disease only (operated FIGO IV/IIIC 1cm+ residuals) N 465 0.68 10.5 vs 15.9 5.4 <0.001 *Censoring CA 125 I would like to summarize the efficacy results from both trials in this slide. As you can see , in both trials the HR for PFS are in the same range ( 0.65 for GOG218 and .68 for ICON7), when we studied the same population, namely the global population in the GOG218 and the High-Risk group in the ICON7 and when the PFS had been determined only by RECIST Criteria.

Is the PFS a valid primary end-point? 4th Ovarian Cancer Consensus Conference. June 25 – 27, 2010 UBC Life Sciences Institute, Vancouver, BC A1-2: What are the appropriate endpoints for different trials: (maintenance, upfront chemotherapy trials including molecular drugs) The recommended primary endpoints for future front line/maintenance clinical trials in ovarian cancer are: Phase II Screening for activity PFS, PFS at defined time point, or Response Phase III Early ovarian cancer - Recurrence free survival (note: recurrence = recurrent disease + deaths from any cause) Advanced ovarian cancer - Both PFS and OS are important endpoints to understand the full impact of any new treatment. Although overall survival is an important endpoint, progression free survival is most often the preferred primary endpoint for trials because of the confounding effect of the post-recurrence/progression therapy on overall survival. Each protocol should specify if PFS or OS is the preferred endpoint. Regardless of which is selected, the study should be designed and powered for both PFS and OS when feasible. Maintenance trials: These criteria should be applied to trials that include maintenance therapy. Int J Gynecol Cancer 2011: 21; 756-762 Then, is the PFS a valid primary end-point in advanced ovarian cancer? The 4th Ovarian Cancer Consensus states that

Magnitude of Clinical Benefit: Is PFS a valid End-Point? Advanced ovarian cancer has a long survival post-progression and subsequent interventions can impact on the OS result, especially the crossover. In GOG 218, 39% of patients received any anti-angiogenic therapy at relapse1. In ICON-7 less than 5% of cross-over could explain the OS When SPP is long enough( >12 months), the chance of expecting a statistically significant prolongation of PFS translating into an OS benefit is extremely low and the number of patients required for demonstrating a survival benefit is extremely large( >2000)2. For clinical trials with a PFS benefit, a lack of statistical significance in OS does not mean a lack of improvement in OS. 1. Chan, et al. SGO 2013: Abstract 292; 2. Broglio et al: J Natl Cancer Inst. 2009 Dec 2;101(23):1642-9 Why no global overall survival benefit has been reported in BEV first line OC trials? This lack of overall survival benefit may be explained by two main facts: the confusion factors: use of subsequent therapies and cross over to experimental therapies and the Long post progression survival. So this take us to the following question: is PFS a valid end-point in first line treatment? Well, let´s go over these different facts

Bevacizumab, in combination with carboplatin and paclitaxel is indicated for the front-line treatment of advanced (FIGO stages III B, III C and IV) epithelial ovarian, fallopian tube, or primary peritoneal cancer Bevacizumab is administered in addition to carboplatin and paclitaxel for up to 6 cycles of treatment followed by continued use of Bevacizumab as single agent until disease progression or for a maximum of 15 months or until unacceptable toxicity, whichever occurs earlier The recommended dose of Bevacizumab is 15 mg/kg of body weight given once every 3 weeks as an intravenous infusion Based on the GOG data supported by ICON.7 the European Agency in 2011 approved Bevacizumab in combination with Carboplatin and Paclitaxel for the front –line treatment of patients with Stage III and IV ovarian cancer. The EMA recommended the dose of 15mg for a maximum of 15 months or until unacceptable toxicity, whichever occurs earlier.

GOG 252: Stage II/III Disease: Small Volume Residual Epithelial Ovarian Cancer Optimal Stage III No prior therapy Phase III PFS primary endpoint Carboplatin AUC=6 (IV) Paclitaxel 80 mg/m2 (d1, 8, 15 3h) Bevacizumab (C2+ C22) x 21 days I Carboplatin AUC=6 (IP) Paclitaxel 80 mg/m2 (d1, 8, 15 3h) Bevacizumab (C2+ C22) x 21 days II Cisplatin 75 mg/m2 (IP d2) Paclitaxel 135 mg/m2 (d1, 3h) Paclitaxel 60 mg/m2 (d8, IP) Bevacizumab (C2+ C22) x 21 days III Open: 27 Jul 2009 Closed: 30 Nov 2011 Accrual: 1100 Study Chair: J Walker ClinicalTrials.gov Identifier: NCT00951496 22

AGO-OVAR 16: Study Design Du Bois A. et al ; Journal Clin Oncol ; 2014.57.4574

Median PFS was 17.9 months (95% CI, 15.9 to 21.8) HR= 0.77; (95% CI, 0.64 to 0.91) 17.9m 12.3m HR=1.08; 95% CI, 0.87 to 1.33; P.499 Median PFS was 17.9 months (95% CI, 15.9 to 21.8) for pazopanib and 12.3 months ( Du Bois A. et al ; Journal Clin Oncol ; 2014.57.4574

Du Bois A. et al ; Journal Clin Oncol ; 2014.57.4574

Recommended Guidelines of Systemic therapy in Relapsed Ovarian Cancer /Cediranib/AMG386

OCEANS: Study schema Platinum-sensitive recurrent OCa No prior BV Measurable disease ECOG 0/1 No prior chemo for recurrent OC No prior BV (n=484) CG + PL C AUC 4 G 1000 mg/m2, d1 & 8 PL q3w until progression CG for 6 (up to 10) cycles C AUC 4 G 1000 mg/m2, d1 & 8 CG + BV BV 15 mg/kg q3w until progression CG + PL (n=242) CG + BV Events, n (%) 187 (77) 151 (62) Median PFS, months (95% CI) 8.4 (8.3–9.7) 12.4 (11.4–12.7) Stratified analysis HR (95% CI) Log-rank p-value 0.484 (0.388–0.605) <0.0001 Carbo/gem from AGO trial (identical dose and schedule), allowed up to 10 cycles in responding pts. Bev continued until progression in contrast to GOG-0218 and ICON7. All pts had measurable disease as defined by RECIST (older criteria). Stratification by cytoreductive surgery was probably not necessary as all pts had measurable disease – driven by thoughts that treatment paradigm was changing in US. Median OS: 18.0 months → 22.8 months 60% power; HR 0.79; 353 events Interim OS at time of PFS analysis; α =0.001 Final OS will be tested at α =0.049 Aghajanian et al. J Clin Oncol; Vol 30;17;2012 28

AURELIA trial design & PFS Results Platinum-resistant OCa ≤2 prior anticancer regimens No history of bowel obstruction/abdominal fistula, or clinical/ radiological evidence of rectosigmoid involvement Chemotherapy Treat to PD/toxicity Optional BEV monotherapyc R BEV 15 mg/kg q3wb + chemotherapy Treat to PD/toxicity Investigator’s choice (without BEV) 1:1 29

ICON 6 Cediranib with Platinum-based Chemotherapy in platinum-sensitive relapsed ovarian cancer Treatment continued to 18 months or until progression (>18 for patients continuing to benefit Continue placebo Switch to Maintenance cediranib Chemotherapy + cediranib Chemotherapy + placebo Arm A (Chemo only) Arm B (Concurrent) Arm C (Maintenance) Relapse > 6 months after completion of first line platinum-based chemotherapy Randomise 2 : 3 : 3 Cediranib: Tirosin Kinase Inhbitor of VEGFR1,2,3 N=456 pts 6 Cycles platinum-based Chemotherapy Carboplatin/paclitaxel Carboplatin/Gemcitabine Single agent platinum Cediranib 800-5000x selectivity for VEGFR2. Phase II trials SA activity Dose reduced 30mg to 20mg due to toxicity in safety phase Reduction in trial size. Due to AZ discontinuing devpt of cediranib. Required a change in primary endpoint from PFS to OS. Main comparison Arm A vs Arm C Baseline characteristics of interest 66% >12month treatment-free interval 90% prior paclitaxel 5% prior bevacizuamb >80% serous carcinoma Chemo- 75% platinum + paclitaxel, 15% platinum-gemcitabine, 11% carboSA 456 patients from 63 centres were enrolled; median age 62 years, ECOG performance status 0/1. Baseline characteristics, e.g. previous treatment interval >12 months (67%) were balanced between arms. PFS comparing reference and concurrent+maintenance using a log-rank test gave a p-value of 0.00001 with an associated Hazard Ratio (HR) of 0.57 (95% CI 0.45-0.74). Adverse events significantly more common in the cediranib-maintenance arm were hypertension, diarrhoea, hypothyroidism, hoarseness, haemorrhage, proteinuria and fatigue. Maintenance phase Ledermann et al ECCO2013

Overall survival Chemo. Maint. Maintenance Restricted mean survival time increases by 2.7 months with maintenance treatment (over two years) Chemotherapy Chemo. Maint. OS events, n (%) 63 (53.3) 75 (45.7) Median, months 20.3 26.3 Log-rank test p=0.042 HR (95% CI) 0.70 (0.51 – 0.99) Test for non-proportionality p=0.0042 Restricted means, months 17.6 Chemo. Maint. Ledermann et al ECCO2013

AMG-386 is a first-in-class recombinant peptide- Fc fusion protein(peptibody). A peptibody fuses a specific peptide with the Fc region of IgG. AMG 386 suppresses tumor angiogenesis by binding to both angiopoietin-1 and -2 (Ang1 and Ang2), thereby preventing their interaction with the Tie2 receptor1. Phase I study demonstrated that AMG 386 was well tolerated and had a safety profile that was distinct from inhibitors of the VEGF pathway2. AMG386

Weekly Paclitaxel + Trebananib Recurrent EOC ≤ 3 prior anticancer regimens Evaluable or measurable disease GOG Performance Status of 0 or 1 PFI < 12 months Treat to PD/toxicity Weekly Paclitaxel + Trebananib Weekly Paclitaxel + Placebo R 1:1 Las pacientes podian haber recibido hasta tres lineas de tratamiento de quimioterapia: 38% 1 linea; 38% 2 lineas y un 25% habian recibido 3 lineas previas. El 53% de las pacientes eran PlatinoResistentes. Prior anti-angiogenic therapy — 8 % en ambos brazos 46% of patients w ILP>6Months 25% 3 previous lines of therapy. Stratification factors Platinum-free interval (PFI) (≤ 6 vs. > 6 months) Measurable disease (Yes/No) Region (North America, Western Europe/Australia, Rest of World) Lancet Oncol, 2014 Jul;15(8), 799-808

TRINOVA- 1 : Progression-free Survival Pac + Placebo (n = 458) Pac + Trebananib (n = 461) Events, n (%) 361 (79) 310 (67) Median PFS, months 5.4 7.2 HR = 0.66 (95% CI, 0.57–0.77) P (stratified log rank) < 0.001 Monk BJ et al; Lancet Oncol, 2014 Jul;15(8), 799-808

TRINOVA-1: Overall Survival (Interim Analysis) Pac + Placebo (n = 458) Pac + Trebananib (n = 461) Events, n (%) 163 (36) 150 (33) Median OS, months 17.3 19.0 HR = 0.86 (95% CI, 0.69–1.08) P (stratified log rank) = 0.19 Monk BJ et al; Lancet Oncol, 2014 Jul;15(8), 799-808

Crossmab technology allows dual binding capabilities Vanucizumab: RO5520985: A2V CrossMab (Anti-human VEGF-A &Anti-human/murine Ang-2) Molecular characteristics Anti-VEGF-A Bevacizumab Anti-Ang-2 LC06 Ang-2 VEGF-A DATA AT ASCO 2015 VH VH VL VL CH1 CL RO5520985 is a bi-specific monoclonal antibody based on a human IgG1 framework. It is comprised of two different heavy chains and two different light chains. One arm of the antibody binds Angiopoietin-2 (Ang2) and one arm binds Vascular Endothelial Growth Factor A (VEGF-A). Point mutations in the CH3 domain (“Knobs-into-holes”) promote the assembly of two different heavy chain Exchange of the constant part of the heavy chain 1 and constant part of the light chain domains in the Ang2 binding Fab promote the correct assembly of the two different light chains called the “CrossMab approach” Roche have developed a new method to construct bispecific antibodies. Standard IgG mAbs, such as Avastin bevacizumab and the anti–angiopoietin 2 (ANG2; ANGPT2) antibody LC06 [a], consist of two identical light chain–heavy chain pairs, which each bind to the same antigen. To construct a bispecific IgG-like CrossMab antibody that could bind to two antigens by using two distinct light chain–heavy chain pairs, a two-step modification process was applied. First, a dimerization interface was engineered into the C-terminus of each heavy chain using previously developed Knob-into-hole (KiH) technology [b]. This ensured that only a heterodimer of two distinct heavy chains, one from Avastin and one from LC06, would be efficiently formed. Next, the constant heavy 1 (CH1) and constant light (CL) domains of LC06 were exchanged [c], keeping the variable heavy (VH) and variable light (VL) domains consistent. The exchange of the CH1 and CL domains ensured that the modified LC06 light chain would only efficiently dimerize with the modified LC06 heavy chain, while the unmodified Avastin light chain would only efficiently dimerize with the Avastin heavy chain. Thus only the desired bispecific CrossMab would be efficiently formed. CH2 CH3 Knobs- into- Holes Crossmab technology allows dual binding capabilities ClinicalTrials.gov Identifier:NCT01688206

Cervical Cancer

Angiogenesis In Cervical Cancer Accumulating evidence supports the concept that angiogenesis plays a central role in cervical carcinogenesis and disease progression Tumor neovascularization is associated with an aggressive course in cervical cancer. Vascular markings seen at colposcopy in women with abnormal Papanicolaou tests are hallmarks for invasive disease, and increased microvessel density and strong immunostaining for the endothelialcell marker, CD31, in cervical cancers suggest a poor prognosis CD31 es una proteina expresada en la superficie celular de células endoteliales Atypical vessels on colposcopy CD31 – Intratumoral microvessel density Tewari KS, Monk BJ. Invasive Cervical Cancer. In: Clinical Gynecologic Oncology, 8th ed. DiSaia PJ, Creasman WT (eds). Mosby, 2012.

Warrant phase III: GOG#240 GOG #227-C: Bevacizumab in the treatment of Recurrent/Metastatic Cervical Cancer Warrant phase III: GOG#240 Monk BJ et al Gynecol Oncol 2008;108:21S abstr 45

Carcinoma of the cervix GOG #240: Incorporation of Bevacizumab in the treatment of Recurrent and Metastatic Cervical Cancer Schema I Paclitaxel 135 or 175 mg/m2 IV Cisplatin 50 mg/m2 IV III Paclitaxel 175 mg/m2 IV Topotecan 0.75 mg/m2 d1-3 Chemo alone Activated: 4/6/09 Closed to accrual: 1/3/12 1:1:1:1 R A N D O M I Z E Carcinoma of the cervix Primary stage IVB Recurrent/persistent Measureable disease GOG PS 0–1 No prior chemotherapy for recurrence (N=452) II Paclitaxel 135 or 175 mg/m2 IV Cisplatin 50 mg/m2 IV Bevacizumab 15 mg/kg IV IV Paclitaxel 175 mg/m2 IV Topotecan 0.75 mg/m2 d1-3 Q21d Rx to PD, toxicity, CR Chemo + Bev 135 is over 24 hrs Stratification factors: Stage IVB vs recurrent/persistent disease Performance status Prior cisplatin Rx as radiation-sensitizer 1º End-Points: If adding BEV to Chemo improves OS If a non-platinum doublet improves OS www.ClinicalTrials.gov Identifier: NCT00803062 41

Chemotherapy + Bev (n=227) GOG #240: OS for Chemo vs Chemo + Bev Chemotherapy + Bev (n=227) 131 (58) 17.0 HR=0.71 (97% CI, 0.54-0.94) P=0.0035 Chemotherapy (n=225) Events, n (%) 140 (62) Median OS, mos 13.3 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 12 36 24 Median follow-up 20.8 mos Proportion Surviving Months on Study Tewari KS, et al. N Engl J Med 2014;370:734-43

Chemotherapy + Bev (n=227) GOG #240: PFS for Chemo vs Chemo + Bev Chemotherapy + Bev (n=227) 183 (81) 8.2 HR=0.67 (95% CI, 0.54-0.82) 2-sided P=0.0002 Chemotherapy (n=225) Events, n (%) 184 (82) Median PFS, mos 5.9 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 36 12 24 RR, % 36 (CR, n=14) 48 (CR, n=28) 2-sided P=0.00807 Proportion Progression-Free The improvement in the primary efficacy endpoint of OS was supported by a statistically significant and clinically meaningful improvement in PFS for patients with persistent, recurrent, or stage IVB carcinoma of the cervix. The Kaplan-Meier plot of PFS showed separation of the curves approximately 2 months after randomization in favor of the Chemo  Bv group which was maintained over time until approximately 24 months when there were few patients at risk Months on Study Tewari KS, et al. N Engl J Med 2014;370:734-43

GOG 240: Chemotherapy in Advanced Cervical Cancer Phase III Studies Median Overall Survival in Months CTX vs. CTX BEV * p< 0.05 13.3 vs. 17.0 * CIS/TOP vs.CIS/PAC 10.3 vs. 12.9 CIS vs.CIS/TOP Months 12 6.5 vs. 9.4 CIS50 vs.CIS100 CIS vs.CIS/IFO 7.0 vs. 7.1 8.0 vs. 8.3 * 1985 1997 2005 2009 2013 Courtesy of: Gottfried Konecny MD

FDA News : For Immediate Release August 14, 2014 FDA approves Bevacizumab to treat patients with aggressive and late-stage cervical cancer First targeted agent licensed for gynecologic malignancy in the USA

Endometrial Cancer GOG#0086P Phase II Randomized 1º End-Point: PFS Arm 1: Paclitaxel 175 mg/m2 IV over 3 hours day 1 Carboplatin AUC = 6 IV day 1 Bevacizumab 15mg/kg IV day 1 Maintenance regimen - Bevacizumab 15mg/kg IV every 21 days until disease progression or prohibition of further therapy. Phase II Randomized 1º End-Point: PFS N= 349 GOG#0086P Arm 2: Paclitaxel 175 mg/m2 IV over 3 hours day 1 Carboplatin AUC = 5 IV day 1 Temsirolimus 25 mg IV days 1 and 8 Maintenance regimen – Temsirolimus 25 mg IV weekly. Days 1,8 and 15 until disease progression or prohibition of further therapy. Eligibility: Stage III or IVA EC measurable disease Stage IVB or Recurrent EC (whether there is measurable disease or not) No prior Chemotherapy Arm 3: Ixabepilone 30 mg/m2 IV over 1 hour day 1 Carboplatin AUC = 6 IV day 1 Bevacizumab 15mg/kg IV day Maintenance regimen - Bevacizumab 15mg/kg IV every 21 days until disease progression or prohibition of further therapy. PI: Carol Aghajanian, M.D. From: 9/14/2009 to 9/9/2014 ClinicalTrials.gov Identifier:NCT00977574

Anti-Angiogenic Therapy in GYNT: Conclusions There is a strong rationale. Four clinical trials incorporating BEV in first line and recurrence : Level 1 evidence( EMA Approved) Meaningful differences in PFS OS benefit in ICON-7 High-Risk group Pazopanib Trial has a poor therapeutic index( benefit/toxicity) Trinova- 1 trial met its 1º End-Point: PFS Results from GOG240 have led to a new standard of care in Metastatic/ Recurrent Cervical Cancer Antiangiogenic Therapy is here to stay.

GRACIAS POR SU ATENCION