Anti-VEGF therapy for ovarian cancer
Focusing on GOG 218 and ICON 7
Fernando Cotait Maluf
Diretor do Serviço de Oncologia Clínica
Beneficência Portuguesa
(maluffc@uol.com.br)
VEGF is an early and persistent promoter of
tumour angiogenesis1–4
VEGF
VEGF
bFGF
TGFβ-1
VEGF
bFGF
TGFβ-1
PLGF
VEGF
bFGF
TGFβ-1
PLGF
PD-ECGF
VEGF
bFGF
TGFβ-1
PLGF
PD-ECGF
Pleiotrophin
Continued VEGF expression3
Tumours continually require VEGF to recruit new vasculature5
VEGF continues to be expressed throughout tumour progression, even as
secondary pathways emerge2,3,6,7
1. Bergers, Benjamin. Nat Rev Cancer 2003; 2. Kim, et al. Nature 1993; 3. Folkman. In: DeVita, Hellman, Rosenberg, eds. Cancer: Principles & Practice of
Oncology. Vol 2. 7th ed. Philadelphia, PA: Lippincott Williams & Wilkins;2005; 4. Ferrara, et al. Nat Med 2003
5. Inoue, et al. Cancer Cell 2002; 6. Mesiano, et al. Am J Pathol 1998; 7. Melnyk, et al. J Urol 1999
GL/AVAO/1110/0027a
VEGF is highly expressed in ovarian cancer,
with multiple effects
Switch from benign to
malignant growth
pattern
1
Formation of the
metastases typical of
ovarian cancer on
the peritoneum
3
Accumulation of ascites, by
1. Schumacher et al. Cancer Res 2007
increasing peritoneal blood vessel2. Ramakrishnan et al. Angiogenesis 2005;
3. Zhang et al. Am J Pathol 2002
4. Trinh, et al. Br J Cancer 2009; 5. Belotti, et al. Cancer Res 2003
permeability
1–5
GL/AVAO/1110/0027a
VEGF levels correlate with ascites volume in
preclinical models
 VEGF enhances permeability of
VEGF
750
Ascites (mL)
10
500
5
VEGF (ng/mL)
peritoneal vessels causing ascites
development1–5
 Positive correlation between
ascites volume and VEGF
concentrations in a mouse model
of ovarian cancer5
Ascites
250
0
0
0
4
11
18
24
34
Time (days)
1. Schumacher et al. Cancer Res 2007; 2. Ramakrishnan et al. Angiogenesis 2005
3. Zhang et al. Am J Pathol 2002; 4. Trinh, et al. Br J Cancer 2009; 5. Belotti, et al. Cancer Res 2003
Clinical association of VEGF expression with poor
survival creates rationale for its inhibition
High VEGF levels (n=39)
100
Low VEGF levels (n=31)
p<0.01
Survival (%)
VEGF (–)/(+)
50
(n=31)
VEGF (++)
p<0.01
(n=39)
0
0
1
2
3
4
5
6
7
8
9
10
11
Years
Yamamoto, et al. BJC 1997
Bevacizumab, a humanised monoclonal
antibody, precisely targets VEGF
VEGF
Bevacizumab
VEGF receptor
Bevacizumab prevents binding of VEGF to receptors1,2
Bevacizumab has a long elimination half life (approximately 20 days) which
may contribute to continuous tumour control3
1. Avastin Summary of Product Characteristics; 2. Presta, et al. Cancer Res 1997
3. Avastin prescribing information,
http://www.ema.europa.eu/docs/en_GB/document_library/EPAR__Product_Information/human/000582/WC500029271.pdf
What does bevacizumab’s mechanism of action
suggest about optimal use?
Chemotherapy
Based on the role of VEGF in
ovarian cancer and the precise
VEGF inhibition provided by
bevacizumab
 bevacizumab can be used
early in disease to control
tumour growth, metastasis
and ascites1–5
 bevacizumab can be
combined with
chemotherapy6–9
= chemotherapy
1. O’Connor, et al. Clin Cancer Res 2009; 2. Baluk, et al. Curr Opin Genet Dev 2005; 3. Mabuchi, et al. Clin Cancer Res 2008
4. Hu, et al. Am J Pathol 2002; 5. Huynh, et al. Mol Cancer Ther 2007; 6. Hicklin, Ellis. JCO 2005; 7. Presta, et al. Cancer Res 1997
8. Baka, et al. Expert Opin Ther Targets 2006; 9. Morabito, et al. Oncologist 2006
Single-agent bevacizumab: promising activity
in phase II trials in recurrent ovarian cancer
Prior
Platinum Platinum
OR Median PFS Median OS
n regimens sensitive resistant Study therapy (%) (months) (months)
Single-agent bevacizumab
Burger 20071
62
≤2
Cannistra 20072
44
2–3
Smerdel 20093
38 Median 5



Bevacizumab
21
4.7

Bevacizumab
16
4.4

Bevacizumab
30
5.9
8.6
7.2 (TTP)
16.9
NR
NR
17
Bevacizumab-based combination regimens
Garcia 20084
70
≤3
McGonigle 20085
22
≤2
Kikuchi 20096
22
>1



Bevacizumab +
24
cyclophosphamide

Bevacizumab +
topotecan
22

Bevacizumab +
PLD
36
PLD = pegylated liposomal doxorubicin; NR = not reported
1. Burger, et al. JCO 2007; 2. Cannistra, et al. JCO 2007; 3. Smerdel, et al. ESMO 2009
4. Garcia, et al. JCO 2008; 5. McGonigle, et al. ASCO 2008; 6. Kikuchi, et al. ASCO 2009
63
Phase III Trial of Bevacizumab in the Primary
Treatment of Advanced Epithelial Ovarian,
Primary Peritoneal, or Fallopian Tube Cancer:
A Gynecologic Oncology Group (GOG) Study
R.A. Burger,1 M.F. Brady,2 M.A. Bookman,3
J.L. Walker,4 H.D. Homesley,5 J. Fowler,6
B.J. Monk,7 B.E. Greer,8 M. Boente,9 S.X. Liang10
1Fox
Chase Cancer Center, Philadelphia, PA; 2Gynecologic Oncology Group Statistical
and Data Center, Roswell Park Cancer Institute, Buffalo, NY; 3University of Arizona
Cancer Center, Tucson, AZ; 4University of Oklahoma Health Sciences Center,
Oklahoma City, OK; 5Brody School of Medicine, Greenville, NC; 6James Cancer
Hospital at the Ohio State University, Hilliard, OH; 7University of California, Irvine
Medical Center, Orange, CA; 8Seattle Cancer Care Alliance, Seattle, WA; 9Minnesota
Oncology and Hematology, Minneapolis, MN; 10State University of New York at Stony
Brook, Stony Brook, NY, USA
64
GOG-0218: Schema
Arm
Carboplatin (C) AUC 6
Front-line:
Epithelial OV, PP
or FT cancer
• Stage III optimal
(macroscopic)
• Stage III
suboptimal
• Stage IV
n=1800 (planned)
Paclitaxel (P) 175 mg/m2
R
A
N
D
O
M
I
Z
E
I
(CP)
Placebo
Carboplatin (C) AUC 6
1:1:1
Paclitaxel (P) 175 mg/m2
BEV 15 mg/kg
II
(CP + BEV)
Placebo
Carboplatin (C) AUC 6
Stratification variables:
• GOG performance status
(PS)
• Stage/debulking status
Paclitaxel (P) 175 mg/m2
BEV 15 mg/kg
Cytotoxic
(6 cycles)
Maintenance
(16 cycles)
III
(CP + BEV
 BEV)
15 months
65
GOG-0218: Analysis Plan
• Primary analysis
– Compare investigator-determined progression-free survival
(PFS) for each BEV arm vs control
• If both results positive, compare Arm III (CP + BEV  BEV)
vs Arm II (CP + BEV)
– Disease progression based on: RECIST, global clinical
deterioration, or CA-1251
– Planned sample size of 1800 based on:
• 90% power to detect PFS hazard ratio (HR) 0.77
– Median PFS shift: 14.0 months  18.2 months
• Secondary analyses: Overall survival (OS), safety,
quality of life; correlative laboratory studies
1. Gynecologic Cancer Intergroup Criteria - Rustin et al. J Natl Cancer Inst 2004
66
GOG-0218: Key Eligibility Criteria
• Histologic diagnosis of epithelial OV, PP, or FT cancer
• Following maximal debulking surgery: stage III optimal
(macroscopic residual disease 1 cm) or suboptimal (>1 cm),
or stage IV
• No prior chemotherapy
• 1–12 weeks after initial surgery
• GOG PS 0–2
• No history of significant vascular events
• No evidence of intestinal obstruction requiring parenteral support
• Written informed consent
GOG-0218: Baseline Surgical–Pathologic
Characteristics
67
Arm I
CP
(n=625)
Arm II
CP + BEV
(n=625)
Arm III
CP + BEV  BEV
(n=623)
III optimal (macroscopic)
218 (35)
205 (33)
216 (35)
III suboptimal
254 (41)
256 (41)
242 (39)
IV
153 (25)
164 (26)
165 (27)
543 (87)
523 (84)
525 (84)
Endometrioid
20 (3)
15 (2)
25 (4)
Clear cell
11 (2)
23 (4)
18 (3)
Mucinous
8 (1)
Characteristic, n (%)
Stage/residual size
Histology
Serous
5 (<1)
8 (1)
Tumor grade
3a
412 (66)
435 (70)
430 (69)
2
94 (15)
77 (12)
92 (15)
1
33 (5)
28 (4)
16 (3)
Not specified/pending
86 (14)
85 (14)
85 (14)
Percentages may not total 100% due to rounding or categorization
aGrade 3 includes all clear cell tumors
GOG-0218: Baseline Surgical–Pathologic
Characteristics
68
Arm I
CP
(n=625)
Arm II
CP + BEV
(n=625)
Arm III
CP + BEV  BEV
(n=623)
III optimal (macroscopic)
218 (35)
205 (33)
216 (35)
III suboptimal
254 (41)
256 (41)
242 (39)
IV
153 (25)
164 (26)
165 (27)
543 (87)
523 (84)
525 (84)
Endometrioid
20 (3)
15 (2)
25 (4)
Clear cell
11 (2)
23 (4)
18 (3)
Mucinous
8 (1)
Characteristic, n (%)
Stage/residual size
Histology
Serous
5 (<1)
8 (1)
Tumor grade
3a
412 (66)
435 (70)
430 (69)
2
94 (15)
77 (12)
92 (15)
1
33 (5)
28 (4)
16 (3)
Not specified/pending
86 (14)
85 (14)
85 (14)
Percentages may not total 100% due to rounding or categorization
aGrade 3 includes all clear cell tumors
69
GOG-0218: Select Adverse Events
Onset between cycle 2 and 30 days after date of last treatment
GI eventsa (grade ≥2)
Arm I
CP
(n=601)
7 (1.2)
Arm II
CP + BEV
(n=607)
17 (2.8)
Arm III
CP + BEV  BEV
(n=608)
16 (2.6)
Hypertension (grade ≥2)
43 (7.2)b
100 (16.5)b
139 (22.9)b
4 (0.7)
4 (0.7)
10 (1.6)
Pain (grade ≥2)
250 (41.7)
252 (41.5)
286 (47.1)
Neutropenia (grade ≥4)
347 (57.7)
384 (63.3)
385 (63.3)
Febrile neutropenia
21 (3.5)
30 (4.9)
26 (4.3)
Venous thromboembolic event
35 (5.8)
32 (5.3)
41 (6.7)
Arterial thromboembolic event
5 (0.8)
4 (0.7)
4 (0.7)
CNS bleeding
0
0
2 (0.3)
Non-CNS bleeding (grade ≥3)
5 (0.8)
8 (1.3)
13 (2.1)
RPLS
0
1 (0.2)
1 (0.2)
Adverse event (grade when limited), n (%)
Proteinuria (grade ≥3)
RPLS = reversible posterior leukoencephalopathy syndrome
aPerforation/fistula/necrosis/leak
bp<0.05
70
GOG-0218: Investigator-Assessed PFS
Proportion surviving progression free
1.0
0.9
Patients with event, n (%)
0.8
Arm I
CP
(n=625)
Arm II
CP + BEV
(n=625)
Arm III
CP + BEV  BEV
(n=623)
423
(67.7)
418
(66.9)
360
(57.8)
10.3
11.2
14.1
Median PFS, months
0.7
Stratified analysis HR
(95% CI)
0.6
One-sided p-value (log rank)
0.908
0.717
(0.759–1.040) (0.625–0.824)
<0.0001a
0.080a
0.5
0.4
0.3
0.2
CP (Arm I)
+ BEV (Arm II)
0.1
+ BEV → BEV maintenance (Arm III)
0
0
12
24
Months since randomization
36
ap-value
boundary = 0.0116
Significant PFS improvement, censored for
CA-125 events and non-protocol therapy
1.0
CP + Pl  Pl
(n=625)
CP + Bev15 
Bev15
(n=623)
12.0
18.2
Median PFS (months)
0.8
Stratified analysis HR
(95% CI)
0.62
(0.52–0.75)
PFS estimate
p value one-sided (log rank)
<0.0001*
0.6
0.4
0.2
CP + Pl  Pl
CP + Bev15  Av15
0
0
6
12
*p value boundary = 0.0116
Data cut-off date: September 29, 2009
18
24
PFS (months)
30
36
42
48
Burger et al. NEJM 2011;365 (26):2473–83
© Massachusetts Medical Society
GOG-0218: PFS by disease stage and
debulking status
Arm I
CP + Pl
 Pl
Arm II
CP + B15
 Pl
Arm III
CP + B15
 B15
Randomised patients with stage III optimally debulked disease 2,3
n
219
204
216
Median PFS (months)
12.4
14.3
17.5
0.81 (0.62, 1.05)
0.66 (0.50, 0.86)
Hazard ratio (95% CI)4
Randomised patients with stage III suboptimally debulked disease3
n
253
256
242
Median PFS (months)
10.1
10.9
13.9
0.93 (0.77, 1.14)
0.78 (0.63, 0.96)
Hazard ratio (95% CI)4
Randomised patients with stage IV disease
n
153
165
165
Median PFS (months)
9.5
10.4
12.8
0.90 (0.70, 1.16)
0.64 (0.49, 0.82)
Hazard ratio (95% CI)4
GL/AVAO/1110/0027a
GOG-0218: Final OS results
Deaths, n (%)
Median overall survival
(months)
Hazard ratio (95% CI)
p
CP + Pl
(n=625)
CP + Bev15  Pl
(n=625)
CP + Bev15  Bev15
(n=623)
299
(47.8%)
309
(49.4%)
270
(43.3%)
40.6
38.8
43.8
1.065
(0.908–1.249)
0.879
(0.745–1.038)
0.2197
0.0641
75
GOG-0218: Overall Survival Analysis
At time of final PFS analysis
1.0
0.9
Proportion alive
0.8
0.7
0.6
0.5
0.4
Patients with
events, n (%)
0.3
Median, months
0.2
HRa
0.1
One-sided p-value
Arm I
CP
(n=625)
Arm II
CP + BEV
(n=625)
Arm III
CP + BEV  BEV
(n=623)
156
(25.0)
150
(24.0)
138
(22.2)
39.3
38.7
39.7
1.036
0.915
(0.827–1.297)
(0.727–1.152)
0.361
0.252
(95% CI)
0
No. at
risk
0
12
24
36
Months since randomization
625/625/623
442/432/437
173/162/171
48
46/39/40
aStratified
analysis
GOG-0218: conclusions
 GOG-0218 met the primary objective in the front-line treatment
of advanced ovarian (epithelial OV, PP and FT) cancer
– PFS with CP + Bevacizumab  Bevacizumab single agent (Arm III)
statistically superior to CP (Arm I) alone
 Survival effect may be masked by discontinuation of drug or
subsequent treatment with bevacizumab after trial
 Treatment regimen was generally well tolerated; adverse events
similar to previous bevacizumab studies
 CP + bevacizumab bevacizumab single agent can be
considered as a standard option
ICON7: A phase III Gynecologic
Cancer InterGroup (GCIG) trial of
adding bevacizumab to standard
chemotherapy in women with newly
diagnosed epithelial ovarian, primary
peritoneal or fallopian tube cancer
Jacobus Pfisterer, Tim Perren, Ann Marie Swart,
Jonathan Ledermann, Frederic Selle, Gunnar Kristensen,
Mark Carey, Philip Beale, Andres Cervantes, Amit Oza,
on behalf of GCIG ICON7 collaborators
(AGO-OVAR, MRC/NCRI, GINECO, NSGO, ANZGOG, GEICO, NCIC-CTG)
Schema
Academic-led, industry-supported trial to investigate the use
of bevacizumab and to support licensing
1:1
Carboplatin AUC 5 or 6
Stratification variables:
Paclitaxel 175 mg/m2
• Stage and extent of debulking
I–III debulked ≤1cm vs
I–III debulked >1 cm vs
IV and inoperable stage III
R
Carboplatin AUC 5 or 6
n=1528*
Paclitaxel 175 mg/m2
• Timing of intended treatment
start
≤4 vs >4 weeks after surgery
• GCIG group
Bevacizumab 7.5 mg/kg q3w
18 cycles
*Dec 2006 to Feb 2009
Assessment
Year 1
Years 2–3
Years 4–5
CT
Baseline; after cycles 3 & 6; at 9 & 12 months
Every 6 months
As indicated
CA-125/clinical
assessment
Every chemotherapy cycle; every 6 weeks
during maintenance phase
Every 3 months
Every 6 months
Patient population
• Histologically confirmed epithelial ovarian cancer,
primary peritoneal cancer or fallopian tube cancer
• Prior surgical debulking with the aim of maximal
surgical cytoreduction AND no planned further surgical
debulking before disease progression
• FIGO stage
• I–IIA if high risk: Grade 3 or clear cell histology (10%)
• IIB–IV: All grades and histological subtypes
• Patients with inoperable stage III/IV disease eligible after
biopsy only if no further surgery planned
• ECOG performance status 0–2
Study endpoints
• Primary endpoint: Progression-free survival (PFS)
• Disease progression defined by RECIST guidelines on
radiological, clinical or symptomatic progression
• CA-125 elevation alone not defined as disease progression
• 1520 patients randomised over 2 years (684 events) →
5% significance level, 90% power to detect:
• PFS hazard ratio (HR) of 0.78
• Increase of median PFS from 18 to 23 months
• Secondary endpoints: Overall survival (due 2012),
response rate, toxicity
• Substudies: Quality of life, health economics,
translational research1
1Collinson
et al. IGCS Abstr 1744, Poster session II, Monday 25th October
Baseline characteristics
Control
(n=764)
Research
(n=764)
57 (18–81)
57 (24–82)
ECOG PS, %
0
1
2
47
47
6
45
49
6
Histology, %
Serous
Clear cell
Endometrioid
69
8
7
69
9
8
Grade, %
1
2
3
7
19
74
5
23
71
FIGO stage, %
I/IIA
IIB–IIIB
IIIC/IV
10
21
69
9
20
71
Debulking surgery/residuum, %
Residual tumour ≤1 cm
Residual tumour >1 cm
No surgery
74
26
2
74
26
2
Characteristic
Median age (range), years
ICON7: adverse events (all grades) consistent
with those previously reported with
bevacizumab
CP (n=753)
39.6
40
29.1 28.3
30
Patients (%)
CP + B7.5  B7.5 (n=745)
25.9
20
12.5
11.6
10
9.2
6.2
2.5
0
6.7
5.0
4.4
2.1
1.3 1.7 0.4
1.3
ATE = arterial thromboembolism; CHF = congestive heart failure
RPLS = reversible posterior leucoencephalopathy syndrome
VTE = venous thromboembolism
GL/AVAO/1110/0027a
4.1
1.5 3.6
0.4 0.4
0
0
2.0 2.8
Perren, et al. ESMO 2010
Selected grade ≥3
adverse events
45
Control (n=753)
40
Research (n=745)
Patients (%)
35
30
25
20
18.3
15.1
15
16.5
10
5
0
2.1
4.3
2.7
1.3
1.3
0.9 0.8
1.2 1.7
1.3
0.5
0.4
0.4
0.4 0.3 0
0.3
0.1
0
2.6
3.5
2.0
2.0
ATE = arterial thromboembolism; CHF = congestive heart failure; RPLS = reversible posterior
leucoencephalopathy syndrome; VTE = venous thromboembolism
Progression-free
survival
Academic analysis
Proportion alive without progression
1.00
Control
Research
392 (51)
367 (48)
17.3
19.0
p=0.0041
0.81 (0.70–0.94)
Events, n (%)
Median, months
0.75
Log-rank test
HR (95% CI)
0.50
0.25
Control
17.3
0
No. at risk
Control
Research
Research
0
3
6
9
764
764
723
748
693
715
556
647
19.0
12
15
18
Time (months)
464
585
307
399
216
263
21
24
27
30
143
144
91
73
50
36
25
19
PFS: high-risk patients (FIGO stage III
suboptimal and FIGO stage IV with debulking)
Proportion alive without progression
1.00
Events, n (%)
0.75
Median, months
CP
(n=234)
CP + Av7.5
 Av7.5
(n=231)
173 (74)
158 (68)
10.5
15.9
p<0.001
Log-rank test
Hazard ratio (95% CI)
13.3
Restricted mean
0.50
0.68 (0.55–0.85)
16.5
0.25
CP
CP + Av7.5  Av7.5
10.5
15.9
0
0
Number at risk
CP
CP + Bev7.5
 Bev7.5
234
231
3
6
205
213
9
12
15
18
Time (months)
100
163
63
94
21
24
30
35
27
30
13
13
Perren et al. NEJM 2011;365 (26):2482–2496
© Massachusetts Medical Society
Preliminary analysis of
overall survival
Proportion surviving
1.00
0.75
0.50
Patients with event, n (%)
Control
Research
130 (17)
111 (15)
Log-rank test
0.25
p=0.098
Hazard ratio (95% CI)
0.81 (0.63–1.04)
1-year survival rate, %
93
95
30 (4)
14 (2)
Anti-VEGF after progression, n (%)
0
0
3
6
9
12
15
18
21
24
27
30
252
259
159
162
83
89
33
40
Time (months)
No. at risk
Control
Research
764
764
741
753
724
737
701
716
652
678
486
525
368
404
Update of outcome analysis
Updated PFS
Proportion alive without progression
1.00
Events, n (%)
0.75
CP
CP +
Av7.5 
Av7.5
464 (61)
470 (62)
17.4
19.8
Median, months
Log-rank test
p=0.04
HR (95% CI)
0.87 (0.77–0.99)
0.50
0.25
17.4
19.8
0
0
Number at risk
CP
764
CP + Av7.5
764
 Av7.5
3
6
693
716
Data cut-off date: November 30, 2010
9
12
15
18
21
Time (months)
24
474
599
350
430
221
229
27
30
114
107
33
36
39
27
Perren et al. NEJM 2011;365 (26):2482–2496
© Massachusetts Medical Society
OS: high-risk patients (FIGO stage III
suboptimal and FIGO stage IV with debulking)
1.00
Proportion alive
0.75
0.50
High-risk
subgroup
CP
(n=234)
CP + Av7.5
 Av7.5
(n=231)
Deaths, n (%)
109 (47)
79 (34)
28.8
36.6
Median, months
0.25
Log-rank test
p=0.002
HR (95% CI)
0.64 (0.48–0.85)
1-year OS rate (%)
86
92
0
0
Number at risk
CP
CP + Av7.5
 Av7.5
234
231
3
6
219
222
9
12
194
208
15
18
21
24
Time (months)
166
186
107
134
27
30
46
65
33
36
39
15
18
Perren et al. NEJM 2011;365 (26):2482–2496
© Massachusetts Medical Society
Conclusions
 Bevacizumab combined with chemotherapy and continued
alone (7.5 mg/kg for 12 months) vs chemotherapy
demonstrates
 Continued improvement in PFS with no crossing of curves
 Trend for improved OS continues in the total population
 Suggested a trend for overall survival (OS) improvement but results
were immature (only 34% of the events required for final OS analysis)
 Final analysis of OS is due in 2013
 Treatment effect is greater in patients at high risk of
recurrence, which may be of clinical relevance
Perren, et al. NEJM 2011
Comparison of Results: GOG 218 and ICON 7
GOG 218
No patients
Arm vs Control
BEV dose
Primary endpoint
Further surgery at inclusion
Maximal cytoreduction
Placebo BEV arm
1873
CT + B vs CT + B  B
15mg/kg q 3w
PFS
Permited
Not specified
Yes
ICON 7
1528
CT + B 
B
7.5mg/kg q 3w
PFS
Not Permited
Specified
No
Comparison of Results: GOG 218 and ICON 7
GOG 218
No patients
1873
ICON 7
1528
ΔT Progression-free survival
3.8 months
2.3 months
HR Progression-free survival
HR: 0.71
HR: 0.79
P value
Overall survival
< 0.0001
Similar to control
0.001
Similar to control
Key Points of GOG218 and ICON 7
• Multicentric studies with adequate statistical power
• Relevant questions:
– Is the addition of BEV to induction CT of benefit ? (GOG 218)
– Is the addition of maintenance BEV to induction CT/BEV of benefit ?
(GOG 218; ICON 7)
• Efficacy evaluation among arms at similar time using:
– CA 125
– Measurable disease
Key Points of GOG218 and ICON 7
• Estratification
• Well balanced arms
• Good tolerance
• Adequate dose delivery
• Manageable side effects
Key Points of GOG218 and ICON 7
• Control arm: suboptimal (carboplatin and paclitaxel q3w)
• Carboplatin q3w and paclitaxel weekly
• Amendment to change OS to PFS (GOG 218)
– PFS = OS ?? (FDA/ASCO/AACR)
• First-line induction treatment: possible YES
• Maintenance treatment: unknown
• No Quality of Life Data (yet)
Other key ongoing and published phase II and
III trials of bevacizumab in ovarian cancer
Study
Therapy studied
Phase
n
Start date
GOG 252
IV vs IP chemotherapy (weekly
paclitaxel + carboplatin) + Bev with
maintenance Bev
III
1,250
FPI Q4/2009
OCTAVIA*
Bev + weekly paclitaxel + carboplatin
II
180
FPI Q2/2009
Front line
Platinum sensitive
OCEANS
Carboplatin + gemcitabine ± Bev
to PD
III
480
FPI 04/2007
GOG 213
Carboplatin + paclitaxel ± Bev to PD
III
660
FPI Q4/2007
III
300
FPI Q3/2009
Platinum resistant
AURELIA
Chemotherapy (weekly paclitaxel,
topotecan, or doxorubicin) ± Bev
*completed recruitment
Take Home Messages
 GOG 218 and ICON 7: met the primary objective in the front-line
treatment of advanced ovarian cancer
– CP + Bev  Bev maintenance is superior to CP alone
regarding progression-free survival
 Treatment regimen was generally well tolerated; adverse events
(including GI perforation) similar to previous bevacizumab
studies
 Bevacizumab is the first molecular targeted and anti-angiogenic
agent to demonstrate activity and benefit in this population
 Mature results of survival and quality of life will help to establish
the routine role of Bev in first-line therapy of advanced ovarian
cancer
Obrigado
maluffc@uol.com.br
‘Comparing’ GOG-0218 and ICON7:
design
Trial
Design
ICON7
 Open-label
GOG-0218
 Double-blind, placebo-
controlled
Primary
endpoint
 2 arms
 3 arms
 Bevacizumab for 12 months
 Bevacizumab for 15 months
 Bevacizumab 2.5mg/kg/week
 Bevacizumab 5mg/kg/week
 PFS (RECIST)
 PFS (RECIST + CA-125)
 CA-125-only not allowed)
 Regulatory-required PFS
 No IRC
(censored for CA-125 and nonprotocol therapy)
 Exploratory: IRC-assessed PFS
Patient
population
Early and advanced stages
Advanced stage
including patients w/o residual
tumour
Only patients with macro
residuals
Burger, et al. ASCO 2010
‘Comparing’ GOG-0218 and ICON7:
design
Trial
Design
ICON7
 Open-label
GOG-0218
 Double-blind, placebo-
controlled
Primary
endpoint
 2 arms
 3 arms
 Bevacizumab for 12 months
 Bevacizumab for 15 months
 Bevacizumab 2.5mg/kg/week
 Bevacizumab 5mg/kg/week
 PFS (RECIST)
 PFS (RECIST + CA-125)
 CA-125-only not allowed)
 Regulatory-required PFS
 No IRC
(censored for CA-125 and nonprotocol therapy)
 Exploratory: IRC-assessed PFS
Patient
population
Early and advanced stages
Advanced stage
including patients w/o residual
tumour
Only patients with macro
residuals
Burger, et al. ASCO 2010
Rationale for targeting VEGF in ovarian cancer
 In ovarian cancer, VEGF expression is associated with
– ascites formation1,2
– malignant progression3,4
– poor prognosis5,6
 Bevacizumab has shown promising single-agent activity
in phase II recurrent ovarian cancer studies7,8
 GOG-0218 was designed to study the combination of
bevacizumab with standard chemotherapy in the
front-line treatment of ovarian cancer
1. Kobold, et al. Oncologist 2009; 2. Yoneda, et al. J Natl Cancer Inst 1998
3. Hazleton, et al. Clin Cancer Res 1999; 4. Chen, et al. Gynecol Oncol 2004
5. Carpini, et al. Angiogenesis 2010; 6. Paley, et al. Cancer 1997
7. Burger, et al. JCO 2007; 8. Cannistra, et al. JCO 2007
GOG-0218: subgroup analyses of PFS
Risk factor
Cancer stage and residual lesion size
III, macroscopic ≤1cm
Arm II vs Arm I
Arm III vs Arm I
III, >1cm
Arm II vs Arm I
Arm III vs Arm I
IV
Arm II vs Arm I
Arm III vs Arm I
Histologic type
Serous
Arm II vs Arm I
Arm III vs Arm I
Nonserous
Arm II vs Arm I
Arm III vs Arm I
Tumour grade
1 or 2
Arm II vs Arm I
Arm III vs Arm I
3
Arm II vs Arm I
Arm III vs Arm I
Total no. of patients
Hazard ratio for Avastin (95% CI)
423
434
0.780
0.618
510
496
0.981
0.763
317
318
0.923
0.698
1,066
1,068
0.913
0.701
184
180
0.893
0.713
232
235
1.039
0.578
847
842
0.891
0.700
0.33
0.50
0.67
bevacizumab better
1.00
1.50
2.00
3.00
Control better
Burger et al. NEJM 2011;365 (26):2473–83
© Massachusetts Medical Society
GOG-0218: subgroup analyses of PFS (cont’d)
Risk factor
Total no. of patients
Hazard ratio for Avastin (95% CI)
GOG performance status score
0
Arm II vs Arm I
Arm III vs Arm I
626
616
0.877
0.710
1 or 2
Arm II vs Arm I
Arm III vs Arm I
624
632
0.961
0.690
<60 years
Arm II vs Arm I
Arm III vs Arm I
616
630
0.976
0.680
60–69 years
Arm II vs Arm I
Arm III vs Arm I
414
408
0.892
0.763
≥70 years
Arm II vs Arm I
Arm III vs Arm I
220
210
Age
0.33
0.50
0.67
bevacizumab better
1.00
1.50
2.00
Control better
3.00
0.841
0.678
Burger et al. NEJM 2011;365 (26):2473–83
© Massachusetts Medical Society
GOG-0218: Overall Survival (OS)
Outcome
Deaths, n (%)
1-year survival, %
Arm I
CP
(n=625)
Arm II
CP + BEV
(n=625)
Arm III
CP + BEV  BEV
(n=623)
156
(25.0)
150
(24.0)
138
(22.2)
90.6
90.4
91.3
• Events observed in 24% of patients at time of data lock
• After primary endpoint changed from OS to PFS
– Unblinding to treatment assignment allowed at time of disease
progression
GOG-0218: Overall Survival Analysis
At time of final PFS analysis (January 2010)
1.0
0.9
Proportion alive
0.8
0.7
0.6
0.5
0.4
Patients with
events, n (%)
0.3
Median, months
0.2
HRa
0.1
One-sided p-value
Arm I
CP
(n=625)
Arm II
CP + BEV
(n=625)
Arm III
CP + BEV  BEV
(n=623)
156
(25.0)
150
(24.0)
138
(22.2)
39.3
38.7
39.7
1.036
0.915
(0.827–1.297)
(0.727–1.152)
0.361
0.252
(95% CI)
0
No. at
risk
0
12
24
36
Months since randomization
625/625/623
442/432/437
173/162/171
48
46/39/40
aStratified
analysis
ICON7: A phase III Gynaecologic
Cancer InterGroup (GCIG) trial of
adding bevacizumab to standard
chemotherapy in women with newly
diagnosed epithelial ovarian, primary
peritoneal or fallopian tube cancer
Tim Perren, Ann Marie Swart, Jacobus Pfisterer,
Jonathan Ledermann, Alain Lortholary, Gunnar Kristensen,
Mark Carey, Philip Beale, Andreas Cervantes, Amit Oza
on behalf of GCIG ICON7 collaborators
(MRC/NCRI, AGO-OVAR, GINECO, NSGO, ANZGOG, GEICO, NCIC-CTG)
Patient population
• Histologically confirmed epithelial ovarian cancer,
primary peritoneal cancer or fallopian tube cancer
• Prior surgical debulking with the aim of maximal
surgical cytoreduction undertaken AND no planned
further surgical debulking before disease progression
• FIGO stage
• I–IIA if high risk: Grade 3 or clear cell histology (10%)
• IIB–IV: All grades and histological subtypes
• Patients with inoperable stage III/IV disease eligible after
biopsy only if no further surgery planned
• ECOG performance status 0–2
13
0
13
1
Schema
Academic-led, industry-supported trial to investigate use of
bevacizumab and to support licensing
1:1
Carboplatin AUC6
Stratification variables:
Paclitaxel 175 mg/m2
• Stage & extent of debulking:
I–III debulked ≤1cm vs stage
I–III debulked >1 cm vs stage IV
and inoperable stage III
R
Carboplatin AUC6
n=1528*
Paclitaxel 175 mg/m2
• Timing of intended treatment
start
≤4 vs >4 weeks after surgery
• GCIG group
Bevacizumab 7.5 mg/kg q3w
18 cycles
*Dec 2006 to Feb 2009
Year 1
Years 2–3
Years 4–5
CT
Baseline; after cycles 3 & 6; at 9 & 12 months
Every 6 months
As indicated
CA-125/clinical
assessment
Every chemotherapy cycle; every 6 weeks
during maintenance phase
Every 3 months
Every 6 months
Study endpoints
• Primary endpoint: Progression-free survival (PFS)
• Disease progression defined by RECIST guidelines on
radiological, clinical or symptomatic progression
• CA-125 elevation alone not defined as disease progression
• 1520 patients randomised over 2 years (684 events) →
5% significance level, 90% power to detect:
• PFS hazard ratio (HR) of 0.78
• Increase of median PFS from 18 to 23 months
• Secondary endpoints: Overall survival (due 2012),
response rate, toxicity
• Substudies: Quality of life, health economics,
translational research
13
2
13
3
Baseline
characteristics (1)
Control
(n=764)
Research
(n=764)
57 (18–81)
57 (24–82)
ECOG PS, n (%)
0
1
2
358 (47)
354 (47)
43 (6)
334 (45)
366 (49)
45 (6)
Origin of cancer, n (%)
Ovary (epithelial)
Fallopian tube
Primary peritoneal
Multiple sites
667
29
56
12
(87)
(4)
(7)
(2)
673
27
50
14
Histology
Serous
Clear cell
Endometrioid
Mucinous
Mixed/other
529
60
57
15
103
(69)
(8)
(7)
(2)
(13)
525 (69)
67 (9)
60 (8)
19 (2)
93 (12)
Grade, n (%)
1
2
3
Unknown
56 (7)
142 (19)
556 (74)
10
41 (5)
175 (23)
538 (71)
10
Characteristic
Median age (range)
(88)
(4)
(6)
(2)
13
4
Baseline
characteristics (2)
Characteristic, n (%)
Control
(n=764)
Research
(n=764)
FIGO stage, n (%)
I/IIA
IIB–IIIB
IIIC/IV
75 (10)
160 (21)
529 (69)
67 (9)
155 (20)
542 (71)
Debulking surgery/residuum
Optimal surgery (≤1 cm)
Suboptimal surgery (>1 cm)
No surgery
552 (74)
195 (26)
17 (2)
559 (74)
192 (26)
13 (2)
FIGO stage and residuum*
Stage I–III (≤1 cm)
Stage I–III (>1 cm)
Stage III (inoperable)/IV
508 (66)
150 (20)
106 (14)
518 (68)
140 (18)
106 (14)
Intent to start chemotherapy*
≤4 weeks from surgery
>4 weeks from surgery
328 (43)
436 (57)
326 (43)
438 (57)
*Stratification
variable
13
5
Selected adverse
events (all grades)
45
Control (n=753)
39.6
40
Research (n=745)
Patients (%)
35
30
29.1 28.3
25.9
25
20
15
10
5
0
12.5
11.6
6.2
5.0
4.4
2.5
2.1
1.31.7 0.41.3
6.7
4.1
9.2
3.6
1.5
0.4 0.4 0
0
2.8
2.0
ATE = arterial thromboembolism; CHF = congestive heart failure; RPLS = reversible posterior
leucoencephalopathy syndrome; VTE = venous thromboembolism
13
6
Selected grade ≥3
adverse events
45
Control (n=753)
40
Research (n=745)
Patients (%)
35
30
25
20
18.3
15.1
15
16.5
10
5
0
2.1
4.3
2.7
1.3
1.3
1.7
0.9
1.3
0.8 0.4
0.4 0.3 0
0.3 1.2
0.10.5 0.4
0
2.6
3.5
2.0
2.0
ATE = arterial thromboembolism; CHF = congestive heart failure; RPLS = reversible posterior
leucoencephalopathy syndrome; VTE = venous thromboembolism
13
7
Progression-free
survival
Academic analysis
Proportion alive without progression
1.00
Control
Research
392 (51)
367 (48)
17.3
19.0
p=0.0041
0.81 (0.70–0.94)
Events, n (%)
Median, months
0.75
Log-rank test
HR (95% CI)
0.50
0.25
Control
Research
17.3
0
0
Number at risk
Control
764
Research
764
3
6
9
723
748
693
715
556
647
19.0
12
15
18
Time (months)
464
585
307
399
216
263
21
24
27
30
143
144
91
73
50
36
25
19
13
8
Progression-free
survival
Regulatory analysis
Proportion alive without progression
1.00
Control
Research
392 (51)
367 (48)
16.0
18.3
p=0.0010
0.79 (0.68–0.91)
Events, n (%)
Median, months
0.75
Log-rank test
HR (95% CI)
0.50
0.25
Control
Research
16.0
0
0
Number at risk
Control
764
Research
764
3
6
9
715
733
676
696
529
617
18.3
12
15
18
Time (months)
419
546
247
330
175
232
21
24
27
30
91
100
65
62
26
19
16
11
13
9
PFS: FIGO stage III suboptimal
and FIGO stage IV with debulking
Control
(n=234)
Proportion alive without progression
1.00
Events, n (%)
Median, months
Log-rank test
0.75
Hazard ratio (95% CI)
Restricted mean
Research
(n=231)
173 (74)
158 (68)
10.5
15.9
p<0.001
0.68 (0.55–0.85)
13.3
16.5
0.50
0.25
Control
Research
10.5
0
0
Number at risk
Control
234
Research
231
3
6
205
213
9
15.9
12
15
18
Time (months)
98
159
36
56
21
24
14
10
27
30
2
1
Subgroup analysis of PFS (1)
Origin of cancer
Age
Histology
HR
Hazard ratio (fixed)
<60
202/449
210/450
0.84
60–69
134/242
142/237
0.76
31/73
40/77
0.82
0
154/334
145/358
1.01
1
175/366
210/354
0.66
2
27/45
31/43
0.78
274/525
278/529
0.85
Mucinous
12/19
10/15
0.77
Endometroid
26/60
25/57
0.81
Clear cell
22/67
22/60
0.90
≥70
ECOG PS
No. of events/no. of patients
CP + Av7.5 
CP
Av7.5
Serous
0
0.5
1
CP + Bev7.5  Bev 7.5 better
Age: Trend p=0.69, interaction p=0.83
ECOG: Trend p=0.027, interaction p=0.022
Histology: Interaction test p=0.085
1.5
2
CP better
Perren et al. NEJM 2011;365 (26):2482–2496
© Massachusetts Medical Society
Subgroup analysis of PFS (2)
Origin of cancer
FIGO
Residual
disease
Grade
No. of events/no. of patients
CP + Av7.5 
CP
Av7.5
HR
Hazard ratio (fixed)
I
6/54
9/65
0.73
II
14/83
19/80
0.72
III
277/523
290/522
0.79
IV
70/104
74/97
0.69
Optimal (≤1cm)
226/559
233/552
0.87
Suboptimal (>1cm)
131/192
145/195
0.68
Grade 1
10/41
16/56
0.76
Grade 2
86/175
77/142
0.77
Grade 3
267/538
294/556
0.81
0
0.5
1
CP + Bev7.5  Bev7.5 better
FIGO: Trend p=0.71, interaction p=0.91
Residual disease: Trend p=0.10
Grade: Trend p=0.76, interaction p=0.95
1.5
2
CP better
Perren et al. NEJM 2011;365 (26):2482–2496
© Massachusetts Medical Society
Interim OS analysis, full population
(regulatory request)
1.00
Proportion alive
0.75
CP
CP + Av7.5
 Av7.5
200 (26)
178 (23)
0.50
Deaths, n (%)
Not yet reached
Median, months
0.25
0
Number at risk
CP
CP + Av7.5
 Av7.5
Log-rank test
p=0.11
HR (95% CI)
0.85 (0.69–1.04)
1-year OS rate (%)
92
0
3
6
9
12
15
764
764
741
753
724
737
703
717
672
702
646
680
95
18
21
24
Time (months)
623
657
542
592
421
459
27
30
33
36
39
304
329
212
228
132
129
71
69
26
19
Perren et al. NEJM 2011;365 (26):2482–2496
© Massachusetts Medical Society
OS: high-risk patients (FIGO stage III
suboptimal and FIGO stage IV with debulking)
1.00
Proportion alive
0.75
0.50
High-risk
subgroup
CP
(n=234)
CP + Av7.5
 Av7.5
(n=231)
Deaths, n (%)
109 (47)
79 (34)
28.8
36.6
Median, months
0.25
Log-rank test
p=0.002
HR (95% CI)
0.64 (0.48–0.85)
1-year OS rate (%)
86
92
0
0
Number at risk
CP
CP + Av7.5
 Av7.5
234
231
3
6
219
222
9
12
194
208
15
18
21
24
Time (months)
166
186
107
134
27
30
46
65
33
36
39
15
18
Perren et al. NEJM 2011;365 (26):2482–2496
© Massachusetts Medical Society
Updated PFS
Proportion alive without progression
1.00
Events, n (%)
0.75
CP
CP +
Av7.5 
Av7.5
464 (61)
470 (62)
17.4
19.8
Median, months
Log-rank test
p=0.04
HR (95% CI)
0.87 (0.77–0.99)
0.50
0.25
17.4
19.8
0
0
Number at risk
CP
764
CP + Av7.5
764
 Av7.5
3
6
693
716
Data cut-off date: November 30, 2010
9
12
15
18
21
Time (months)
24
474
599
350
430
221
229
27
30
114
107
33
36
39
27
Perren et al. NEJM 2011;365 (26):2482–2496
© Massachusetts Medical Society
Conclusions
 Bevacizumab combined with chemotherapy and
continued alone (7.5 mg/kg for 12 months) vs
chemotherapy demonstrates
 Continued improvement in PFS with no crossing of
curves
 Trend for improved OS continues in the total
population
 Final analysis of OS is due in 2013
 Treatment effect is greater in patients at high risk of
recurrence, which may be of clinical relevance
Perren, et al. NEJM 2011