To estimate the efficacy and toxicity of AMG 386, an investigational peptide-Fc fusion protein that neutralizes the interaction between the Tie2 receptor and angiopoietin-1/2, plus weekly paclitaxel in patients with recurrent ovarian cancer.

Patients with recurrent epithelial ovarian, fallopian tube, or primary peritoneal cancer were randomly assigned 1:1:1 to receive paclitaxel (80 mg/m2 once weekly [QW], 3 weeks on/1 week off) plus intravenous AMG 386 10 mg/kg QW (arm A), AMG 386 3 mg/kg QW (arm B), or placebo QW (arm C). The primary end point was progression-free survival (PFS). Secondary end points included overall survival, objective response, CA-125 response, safety, and pharmacokinetics.

One hundred sixty-one patients were randomly assigned. Median PFS was 7.2 months (95% CI, 5.3 to 8.1 months) in arm A, 5.7 months (95% CI, 4.6 to 8.0 months) in arm B, and 4.6 months (95% CI, 1.9 to 6.7 months) in arm C. The hazard ratio for arms A and B combined versus arm C was 0.76 (95% CI, 0.52 to 1.12; P = .165). Further analyses suggested an exploratory dose-response effect for PFS across arms (Tarone's test, P = .037). Objective response rates for arms A, B, and C were 37%, 19%, and 27%, respectively. The incidence of grade ≥ 3 adverse events (AEs) in arms A, B, and C was 65%, 55%, and 64%, respectively. Frequent AEs included hypertension (8%, 6%, and 5% in arms A, B, and C, respectively), peripheral edema (71%, 51%, and 22% in arms A, B, and C, respectively), and hypokalemia (21%, 15%, and 5% in arms A, B, and C, respectively). AMG 386 exhibited linear pharmacokinetic properties at the tested doses.

AMG 386 combined with weekly paclitaxel was tolerable, with a manageable and distinct toxicity profile. The data suggest evidence of antitumor activity and a dose-response effect, warranting further studies in ovarian cancer.

Angiogenesis has been implicated in the development and progression of ovarian cancer.1,2 Angiopoietin-1 (Ang1) and angiopoietin-2 (Ang2) are ligands of the Tie2 receptor, which is expressed on vascular endothelial cells. The angiopoietin and vascular endothelial growth factor (VEGF) axes are separate pathways that play distinct roles in tumor angiogenesis.3 Evidence suggests that both are involved in pathologic angiogenesis associated with ovarian cancer.46 Treatment with the VEGF inhibitor bevacizumab has been extensively studied and has demonstrated efficacy among patients with advanced ovarian cancer.714 Specifically, bevacizumab maintenance in combination with carboplatin/paclitaxel has shown improved progression-free survival (PFS) over carboplatin/paclitaxel plus placebo13 or carboplatin/paclitaxel alone in newly diagnosed patients.14 However, bevacizumab treatment may increase the risk of bowel perforation and other morbidities among patients with cancer15 and specifically among those with recurrent ovarian cancer.9,1618

AMG 386 is a first-in-class investigational peptide-Fc fusion protein that neutralizes the interaction between the Tie2 receptor and Ang1 and Ang2. In preclinical studies, Ang2 inhibition mediated antiangiogenic and antitumor activity in human tumor xenografts.19 Ang1 antagonism augmented Ang2 antagonism in curbing tumor growth, and dual Ang1/2 inhibition induced vessel regression, whereas selective Ang2 inhibition did not.20 In phase I clinical trials in advanced solid cancers, AMG 386 treatment showed antitumor activity as monotherapy (first-in-human study with doses up to 30 mg/kg once weekly [QW])21 and in combination with chemotherapy (at the selected dose of 10 mg/kg),22 while presenting a distinct adverse event (AE) profile. In the first-in-human study, one patient with advanced refractory ovarian cancer who received AMG 386 at 30 mg/kg QW had a durable partial response per Response Evaluation Criteria in Solid Tumors (RECIST) that persisted until she withdrew from study at week 156.21 The aim of the present study was to evaluate the efficacy, tolerability, and pharmacokinetics of AMG 386 in combination with weekly paclitaxel in patients with recurrent ovarian cancer.


Eligible women (≥ 18 years old) had histologically or cytologically documented epithelial ovarian (International Federation of Gynecology and Obstetrics stage II to IV), fallopian tube, or primary epithelial peritoneal cancer; recurrent measurable or nonmeasurable disease per RECIST (version 1.0)23; or recurrent or nonmeasurable disease per CA-125 progression per Gynecologic Cancer Intergroup (GCIG) criteria.24 Other eligibility criteria were ≤ 3 previous anticancer therapies including ≥ 1 platinum-based regimen and Gynecologic Oncology Group performance status of ≤ 1. Key exclusion criteria were a higher than average risk for bowel perforation (symptoms of partial or complete bowel obstruction, fistula or bowel perforation within 6 months, or requirement for total parenteral nutrition and continuous hydration); ongoing small bowel dysfunction; CNS metastases; prior malignancy (unless curatively treated); peripheral neuropathy grade ≥ 2; previous abdominal radiation therapy; inadequate cardiac, renal, hepatic, or hematologic function; bleeding diathesis within 14 days; and arterial or venous thrombosis within 12 months. Patients previously treated with angiopoietin inhibitors were excluded; prior VEGF/VEGF receptor inhibitor treatment was allowed with a sufficient washout period (60 days for long half-life molecules such as bevacizumab; 21 days for other agents). All patients provided written informed consent, and all study procedures were approved by the appropriate institutional ethics committees.

Study Design and Treatment

This randomized, double-blind, placebo-controlled, phase II study was an international collaboration among 38 centers. The primary end point was PFS using a composite definition (see Statistical Analysis). Secondary end points included overall survival, additional measures of efficacy, CA-125 response rate, change in CA-125 values, incidence of AEs, and pharmacokinetics of AMG 386.

All patients received intravenous paclitaxel 80 mg/m2 QW (3 weeks on/1 week off) and were randomly assigned 1:1:1 using an automated voice response telephone system to also receive intravenous AMG 386 10 mg/kg (arm A), AMG 386 3 mg/kg (arm B) QW, or placebo QW (arm C) until progression, unacceptable toxicity, or withdrawal of consent. Random assignment was stratified by prior anti-VEGF therapy and disease progression on or within 6 months of the last chemotherapy. Treatment assignments were blinded to patients and all study site personnel until the primary analysis. Dose modifications for AMG 386 and placebo were not permitted. If grade ≥ 3 treatment-related toxicity occurred, AMG 386 or placebo was withheld until it resolved to grade ≤ 1 or baseline. Both treatments were discontinued if dosing was withheld for more than 28 days or withheld three times. Patients in arm C were offered open-label AMG 386 10 mg/kg QW monotherapy after disease progression if they remained eligible.

Efficacy Assessments

Computed tomography or magnetic resonance imaging was performed every 8 ± 1 weeks during the study. Tumor responses per RECIST23 were determined by the investigator. Assays for CA-125 were performed by both local and central laboratories. For patients with an assessment of CA-125 progression or response (per GCIG criteria24,25), a confirmatory measurement was required ≥ 28 days after the first criteria for CA-125 progression/response were met. Evaluable patients without confirmation of tumor and/or CA-125 response were considered nonresponders. Patients who discontinued treatment without evidence of progression and who had not withdrawn consent continued with radiographic and CA-125 assessments until progressive disease occurred or a new therapy was initiated.


All AEs were recorded and graded by the investigators according to the National Cancer Institute Common Terminology Criteria for Adverse Events (version 3.0).26


Serum samples for assessment of human anti–AMG 386 binding and neutralizing antibodies (as previously described21) were collected before dose on day 1 of cycles 1 to 3 and every 16 weeks thereafter.


Serum samples were collected at the end of infusion at weeks 1 and 5 (maximum observed concentration [Cmax]) and before infusions at weeks 1, 3, 5, and 9 and every 8 weeks thereafter (minimum observed concentration [Cmin]). Serum AMG 386 concentrations were measured using an enzyme-linked immunosorbent assay; pharmacokinetic parameters were estimated using noncompartmental methods (WINNonlin version 5.1.1 on Citrix; Pharsight Corporation, Mountain View, CA).

Statistical Analysis

The planned enrollment of 150 patients (n = 50 patients per arm) was selected to generate treatment effect estimates. The primary end point was PFS; the primary statistical assessment was estimation of the PFS hazard ratio (HR). The primary analysis was planned when a total of 113 PFS events had occurred. HRs with two-sided 80% and/or 95% CIs were derived from a stratified Cox regression model for comparisons of PFS, PFS per RECIST, and time to progression (TTP) based on the intent-to-treat analysis set between arms A and B combined versus arm C (the same approach was used to estimate HRs for arms A and B separately). With a hypothesized HR for PFS of 0.75, an HR with a two-sided 80% CI with a maximum half-width of 0.22 (comparing arms A and B combined v arm C) was estimated.27,28 Pairwise comparisons between arms were performed using stratified log-rank tests. Although this was an estimation study, for 150 patients, it had approximately 57% power to detect the treatment effect (PFS: HR, 75%) of paclitaxel plus placebo (n = 50) versus paclitaxel plus AMG 386 (n = 100) at a 20% two-sided level of significance (chosen because of the randomized phase II design).

No adjustment for multiple testing was made when calculating CIs. Kaplan-Meier estimates of median survival times were calculated as previously described.29 In prespecified descriptive analyses, dose-response effects for time-to-event end points among treatment groups were explored using stratified Tarone's tests.30

Primary efficacy analyses included the intent-to-treat analysis set (data collected from patients after they had started to receive open-label AMG 386 were excluded from all efficacy analyses except overall survival). Safety analyses included data from the double-blind phase for all patients who received ≥ 1 dose of AMG 386 or placebo. PFS as the primary end point was defined as time from random assignment to disease progression per RECIST, clinical progression (per investigator) or CA-125 progression (per GCIG criteria), or death. PFS defined as time from random assignment to disease progression per RECIST or death from any cause (PFS per RECIST) was a secondary end point. TTP was defined as time from random assignment to disease progression per RECIST, clinical progression, or CA-125 progression.


One hundred sixty-one patients were randomly assigned between July 2007 and January 2009 (arm A, n = 53; arm B, n = 53; arm C, n = 55). One patient in arm A did not receive treatment because of grade 2 asthenia that occurred within 6 days of random assignment; all other randomly assigned patients received ≥ 1 dose. Baseline demographics and clinical characteristics were generally balanced across treatment arms (Table 1). Five percent of all patients had received prior anti-VEGF therapy, and 54% had disease progression on or within 6 months of the last chemotherapy. Median CA-125 levels at baseline in arms A, B, and C were 273 U/mL (range, 4 to 11,696 U/mL), 216 U/mL (range, 4 to 5,835 U/mL), and 157 U/mL (range, 4 to 6,000 U/mL), respectively. At the time of this primary analysis, 16 patients continued to receive AMG 386 or placebo (arms A, B, and C: n = 7, n = 5, and n = 4, respectively; Fig 1). The most common reasons for treatment discontinuation were disease progression (arms A, B, and C: n = 31, n = 25, and n = 37, respectively) and patient's decision to stop treatment/withdrawal of consent (arms A, B, and C: n = 5, n = 11, and n = 8, respectively). Ten patients in arm A, 13 in arm B, and 14 in arm C did not have radiographic or CA-125 progression. Seventeen patients in the placebo arm received AMG 386 in the open-label cross-over portion of the study; three patients have been continuing AMG 386 monotherapy for 2 to 10 months.


Table 1. Baseline Demographics and Clinical Characteristics

Table 1. Baseline Demographics and Clinical Characteristics

Demographic or Clinical Characteristic Arm A (AMG 386 10 mg/kg QW + paclitaxel; n = 53)
Arm B (AMG 386 3 mg/kg QW + paclitaxel; n = 53)
Arm C (placebo + paclitaxel; n = 55)
No. of Patients % No. of Patients % No. of Patients %
    White 41 77 49 92 44 80
    Asian 6 11 2 4 7 13
    Black 3 6 1 2 0 0
    Hispanic 2 4 1 2 2 4
    Other 1 2 0 0 2 4
Age, years
    Median 59 60 62
    Range 27-80 28-85 38-83
GOG performance status
    0 26 49 32 60 30 55
    1 25 47 21 40 25 45
    2-3* 2 4 0 0 0 0
Primary tumor type
    Ovarian cancer 46 87 47 89 44 80
    Primary peritoneal cancer 7 13 6 11 8 15
    Fallopian tube cancer 0 0 0 0 3 5
Tumor histology
    Serous 30 57 23 43 34 62
    Endometrioid 7 13 6 11 3 5
    Clear cell 1 2 0 0 2 4
    Mucinous 0 0 1 2 1 2
    Unclassified 12 23 21 40 13 24
    Unavailable 3 6 2 4 2 4
FIGO disease stage at screening
    I* 1 2 0 0 2 4
    II 1 2 1 2 1 2
    III 22 42 29 55 25 45
    IV 13 25 11 21 17 31
    Unknown/unavailable 16 30 12 23 10 16
No. of prior anticancer therapies
    1 21 40 23 43 17 31
    2 19 37 17 32 26 47
    3 12 23 13 25 12 22
No. of lines of prior platinum therapy
    1 30 58 30 57 26 47
    2 18 35 15 28 22 40
    3 4 8 8 15 7 13
Platinum sensitivity status
    Refractory 5 9 3 6 4 7
    Resistant (PFI < 6 months) 20 38 23 43 20 36
        PFI 6 to 12 months 12 23 21 40 20 36
        PFI > 12 months 15 28 6 11 10 18
    Unavailable 1 2 0 0 1 2
Stratification factors
    Prior anti-VEGF/VEGFR therapy
        Yes 3 6 2 4 3 5
        No 50 94 51 96 52 95
    Disease progression on or within 6 months of the last chemotherapy regimen
        Yes 29 55 29 55 29 53
        No 24 45 24 45 26 47
No. of AMG 386 or placebo infusions
    Median 24 20 13
    Range 1-106 1-55 2-83
Follow-up time, weeks
    Median 66 65 64
    Range 40-120 40-112 40-110

Abbreviations: FIGO, International Federation of Gynecology and Obstetrics; GOG, Gynecologic Oncology Group; PFI, platinum-free interval; QW, once weekly; VEGF, vascular endothelial growth factor; VEGFR, vascular endothelial growth factor receptor.

*Enrollment of patients with GOG performance status ≥ 2 or FIGO stage I disease was considered a protocol violation.

†The PFI was calculated as the time from the last dose of the last platinum-containing regimen until disease progression. Refractory is defined as patients who experienced progression during the first-line platinum-containing chemotherapy regimen; resistant is defined as patients who did not experience progression during the first-line therapy but who experienced disease progression during the last (second or later line) platinum-containing chemotherapy regimen or who experienced progression within 6 months of the last dose of the last (second or later line) platinum-containing regimen.

‡From date of random assignment to the date of data cutoff for the analysis.


At the time of this analysis, 43 patients in arm A, 40 in arm B, and 41 in arm C had experienced a PFS event defined as disease progression per RECIST, clinical progression, CA-125 progression, or death. Median PFS times in arms A and B were 7.2 and 5.7 months, respectively, compared with 4.6 months in arm C (Table 2; Fig 2). In a Cox model analysis of PFS adjusted for the stratification factors, the HR for arms A and B combined versus arm C was 0.76 (95% CI, 0.52 to 1.12; P = .165). Tarone's test and additional analyses exploring dose-exposure suggest a dose-response effect for PFS across the three arms (P = .037; Table 2). A subgroup analysis of PFS suggested that the magnitude of the treatment effect on PFS was greatest among patients who had received one prior therapy and, in arm A, among patients who were platinum refractory (Fig 3). HRs for TTP and PFS per RECIST showed a similar direction relative to placebo as seen with the primary end point results (Table 2). Of the randomly assigned patients, only eight had received prior anti-VEGF therapy. A sensitivity analysis that collapsed the prior anti-VEGF strata provided results consistent with the primary analysis (data not shown). At the time of this analysis, overall survival data were not mature. In an updated analysis (conducted 8 months after the primary analysis), median overall survival times in arms A, B, and C were 22.5, 20.4, and 20.9 months, respectively (Table 2).


Table 2. AMG 386 Efficacy Analysis

Table 2. AMG 386 Efficacy Analysis

End Point Arm A (AMG 386 10 mg/kg QW + paclitaxel; n = 53) Arm B (AMG 386 3 mg/kg QW + paclitaxel; n = 53) Arm C (placebo + paclitaxel; n = 55)
Primary end point: PFS per RECIST, clinical progression, CA-125 progression, or death*
    Kaplan-Meier PFS time, months
        Median 7.2 5.7 4.6
        95% CI 5.3 to 8.1 4.6 to 8.0 1.9 to 6.7
    Stratified Cox model
        PFS v arm C
            HR 0.76 0.75
            95% CI 0.49 to 1.18 0.48 to 1.17
            P .225 .207
        PFS, arms A + B v arm C
            HR 0.76
            95% CI 0.52 to 1.12
            80% CI 0.59 to 0.98
            P .165
    Dose-response test
        Tarone's test, P .037
Secondary end points
    PFS per RECIST or death
        Kaplan-Meier PFS time, months
            Median 7.3 6.8 5.6
            95% CI 5.5 to 8.9 5.3 to 8.7 3.6 to 7.6
        Stratified Cox model
            PFS v arm C
                HR 0.81 0.75
                95% CI 0.51 to 1.30 0.49 to 1.21
                P .384 .237
            PFS, arms A + B v arm C
                HR 0.80
                95% CI 0.53 to 1.20
                P .275
            No. 22 28 28
            % 42 53 51
        Kaplan-Meier OS time, months
            Median 22.5 20.4 20.9
            95% CI 20.4 to NE 17.7 to 28.5 11.3 to 24.2
        Stratified Cox model
            OS v arm C
                HR 0.60 0.77
                95% CI 0.34 to 1.06 0.45 to 1.31
                P .081 .330
    TTP per RECIST, clinical progression (per investigator), or CA-125 progression (per GCIG criteria)
        Kaplan-Meier TTP, months
            Median 7.3 7.4 4.8
            95% CI 5.7 to 8.4 4.6 to 8.4 1.9 to 7.4
        Stratified Cox model
            TTP v arm C
                HR 0.75 0.73
                95% CI 0.48 to 1.19 0.45 to 1.19
                P .223 .207
            TTP arms A + B v arm C
                HR 0.75
                95% CI 0.50 to 1.12
                P .155
Tumor response (per RECIST)
    Patients with measurable disease at baseline
        No. 46 47 52
        % 87 89 95
    Best response assessment
        No. 17 9 14
        % 37 19 27
        Complete response
            No. 2 1 0
            % 4 2 0
        Partial response
            No. 15 8 14
            % 33 17 27
        Stable disease
            No. 20 22 18
            % 43 47 35
        Progressive disease
            No. 6 8 13
            % 13 17 25
            No. 0 0 1
            % 0 0 2
        Not done
            No. 3 8 6
            % 7 17 12
    Confirmed objective response rate, % 37 19 27
        95% CI 23 to 52 9 to 33 16 to 41
    Time to response, months
        Mean 2.9 3.9 2.9
        95% CI 2.2 to 3.7 2.8 to 5.1 2.3 to 3.6
    Duration of response, months
        Median 5.4 5.7 7.1
        95% CI 3.7 to 6.9 4.7 to NE 3.7 to NE
    Best CA-125 response
        Patients evaluable
            No. 41 38 40
            % 77 72 73
        Confirmed CA-125 response
            No. 29 22 11
            % 71 58 28

Abbreviations: CR, complete response; GCIG, Gynecologic Cancer Intergroup; HR, hazard ratio; NE, not estimable; OS, overall survival; PFS, progression-free survival; PR, partial response; QW, once weekly; RECIST, Response Evaluation Criteria in Solid Tumors; TTP, time to progression.

*Defined as the time from random assignment to disease progression per RECIST, clinical progression (per investigator) or CA-125 progression (per GCIG criteria), or death. At the time of this analysis, 10 patients in arm A, 13 patients in arm B, and 14 patients in arm C were censored.

†Defined as the time from random assignment to disease progression per RECIST or death from any cause.

‡At the time of the primary analysis, OS data were not mature. Data are from an updated analysis conducted 8 months after the primary analysis.

§Includes patients with a response assessment of complete response, partial response, or stable disease before the scheduled first assessment of response without an additional assessment of response.

‖Includes patients for whom imaging was not performed at the scheduled assessment of response.

Objective Response Rate and CA-125 Response

Most patients had measurable disease at baseline. The confirmed objective response rate was 37% in arm A, 19% in arm B, and 27% in arm C (Table 2). Duration of response and time to response were similar across treatments. Reductions in the target lesion dimensions from baseline measure according to RECIST occurred in 84%, 79%, and 64% of patients in arms A, B, and C, respectively (Fig 4A).

One hundred nineteen patients were evaluable for CA-125 response. The proportions of patients with a confirmed CA-125 response in arms A, B, and C were 71%, 58%, and 28%, respectively (Table 2). All evaluable patients in arm A had a decrease in serum CA-125 from baseline compared with 84% of patients in arm B and 63% of patients in arm C (Fig 4B).


All patients except two in the placebo arm experienced one or more AE (Table 3). The most frequently occurring toxicities were peripheral edema, fatigue, nausea, alopecia, and diarrhea. Although some AEs were more frequent in arm A than arms B or C (eg, peripheral edema), there was no clear trend in dose-related toxicity. Infusion reactions were rare and were only seen in one patient in arm A (grade 3) and in two patients who received placebo (grade 1). More patients in arms A (11%) and B (17%) had AEs leading to AMG 386/placebo discontinuation, compared with arm C (7%).


Table 3. Incidence of Adverse Events

Table 3. Incidence of Adverse Events

Adverse Event Arm A (AMG 386 10 mg/kg QW + paclitaxel; n = 52)
Arm B (AMG 386 3 mg/kg QW + paclitaxel; n = 53)
Arm C (placebo + paclitaxel; n = 55)
All Grades
Grade ≥ 3
All Grades
Grade ≥ 3
All Grades
Grade ≥ 3
No. % No. % No. % No. % No. % No. %
Patients with any adverse event 52 100 34 65 53 100 29 55 53 96 35 64
Adverse events occurring in ≥ 25% of patients in 1 or more treatment arms
    Peripheral edema 37 71 2 4 27 51 3 6 12 22 2 4
    Fatigue 34 65 1 2 37 70 3 6 26 47 2 4
    Nausea 27 52 1 2 31 58 1 2 17 31 1 2
    Alopecia 26 50 0 0 20 38 1 2 20 36 0 0
    Diarrhea 22 42 1 2 19 36 2 4 15 27 0 0
    Peripheral neuropathy 20 38 5 10 14 26 1 2 17 31 2 4
    Constipation 20 38 1 2 12 23 0 0 15 27 0 0
    Cough 18 35 0 0 16 30 0 0 9 16 1 2
    Abdominal pain 17 33 1 2 17 32 2 4 18 33 3 5
    Pain in extremity 15 29 0 0 15 28 0 0 8 15 0 0
    Headache 15 29 0 0 13 25 0 0 7 13 0 0
    Vomiting 14 27 2 4 14 26 3 6 9 16 2 4
    Dizziness 14 27 0 0 6 11 0 0 10 18 0 0
    Dyspnea 13 25 1 2 20 38 5 9* 12 22 2 4
    Arthralgia 13 25 0 0 8 15 0 0 8 15 0 0
    Urinary tract infection 13 25 0 0 6 11 1 2 7 13 0 0
    Oropharyngeal pain 13 25 0 0 6 11 0 0 8 15 0 0
    Back pain 12 23 2 4 13 25 0 0 10 18 0 0
Adverse events of interest
    GI perforation 0 0 0 0 0 0 0 0 1 2 1 2
    Hypertension 4 8 0 0 3 6 0 0 3 5 0 0
    Arterial thromboembolic events 1 2 1 2 1 2 1 2 0 0 0 0
    Venous thromboembolic events 4 8 3 6§ 3 6 2 4 6 11 5 9
    Hemorrhagic events 14 27 2 4 14 26 1 2 13 24 0 0
    Impaired wound healing 1 2 0 0 1 2 0 0 0 0 0 0
    Proteinuria 3 6 0 0 4 8 0 0 2 4 0 0
    Infusion reactions 1 2 1 2 0 0 0 0 2 4 0 0
    Hypokalemia 11 21 6 12# 8 15 6 11 3 5 2 4

NOTE. Adverse events are reported for all patients who received ≥ 1 dose of AMG 386 or placebo (safety analysis set). Unless otherwise specified, all events are of grade ≤ 3 in severity.

Abbreviation: QW, once weekly.

*Includes one patient with grade 5 dyspnea.

†Grade 4 peritonitis.

‡Grade 4 myocardial infarction.

§Includes two patients with a grade 4 pulmonary embolism.

‖Includes one patient with a grade 5 pulmonary embolism and one with a grade 4 pulmonary embolism.

¶Includes one patient with a grade 4 pulmonary embolism.

#Includes two patients with grade 4 hypokalemia.

The incidence of grade 3 AEs was 54%, 34%, and 53% in arms A, B, and C, respectively; the incidence of grade 4 AEs was 10%, 15%, and 2%, respectively. Serious AEs occurred in 29%, 30%, and 33% of patients in arms A, B, and C, respectively. Grade ≥ 3 AEs occurring more frequently in arms A or B were hypokalemia, peripheral neuropathy, dyspnea (Table 3), anorexia (2%, 6%, and 0% in arms A, B, and C, respectively), and neutropenia (8%, 9%, and 4% in arms A, B, and C, respectively). One patient in arm A and three patients in arm B had grade 5 AEs compared with five patients in arm C. The only AE responsible for more than one death was ovarian cancer (arm C, n = 4). None of the deaths in arms A and B were considered related to AMG 386.

Several AEs generally of interest were noted (Table 3). One patient in arm C had grade 4 peritonitis as a result of bowel perforation; no other perforations occurred. No patients experienced grade ≥ 3 hypertension, proteinuria, or impaired wound healing. The incidence of hemorrhagic events and arterial and venous thromboembolic events was similar across arms. One patient in arm B had a grade 5 pulmonary embolism.

In arms A and B, a total of two patients had pre-existing anti–AMG 386 antibodies, and three patients developed anti–AMG 386 antibodies during the study. No patients developed anti–AMG 386 neutralizing antibodies.


AMG 386 exhibited linear pharmacokinetic properties at doses of 3 and 10 mg/kg. Paclitaxel coadministration did not markedly affect AMG 386 exposure. Among patients who received AMG 386 10 mg/kg, mean Cmin values at weeks 5 (19.2 μg/mL) and 9 (20.6 μg/mL), and every 8 weeks thereafter (27.4 μg/mL) and mean Cmax values at week 5 (277 μg/mL) were consistent with those reported for the same dose in the monotherapy study in advanced solid tumors (26.6 and 249 μg/mL, respectively).21 Similarly, at the 3 mg/kg dose, mean Cmin values at weeks 5 (6.01 μg/mL) and 9 (6.93 μg/mL) and every 8 weeks thereafter (7.02 μg/mL) and mean Cmax values at week 5 (81.9 μg/mL) were similar to those among patients who received the same dose as monotherapy (10.2 and 84.2 μg/mL, respectively).21

Although there was no statistically significant difference between treatment arms, estimates of median PFS obtained in this phase II study were encouraging among women with recurrent ovarian cancer who received weekly AMG 386 in combination with weekly paclitaxel, compared with placebo plus paclitaxel. Similar results were obtained for PFS per RECIST, TTP, objective response rate, and CA-125 response rate. CA-125 response is a widely used marker of response in ovarian cancer trials.31 Lower CA-125 levels during treatment have been associated with longer PFS32,33 and overall survival.33 The consistent improvements in three clinical outcomes including PFS (regardless of definition) suggest meaningful clinical benefits associated with AMG 386 therapy in recurrent epithelial ovarian cancer. Subgroup analyses showed antitumor activity across patients in arm A with platinum-sensitive (PFI, 6 to 12 months), resistant, and refractory disease. Interestingly, among patients receiving AMG 386 10 mg/kg, those with refractory and resistant disease seemed to have derived greatest benefit. Treatment was generally well tolerated, with a distinct and manageable toxicity profile. The pharmacokinetic results, which confirm those from a phase IB study demonstrating that combining AMG 386 with chemotherapy is feasible,22 support repeated coadministration of paclitaxel and AMG 386.

The data also suggested a dose-response effect: AMG 386 10 mg/kg provided a 2.6-month increase in PFS over placebo, whereas AMG 386 3 mg/kg provided a 1.1-month increase. Results from a pharmacokinetic/pharmacodynamic analysis of the data suggest that maximum benefit in PFS may not have been reached with an AMG 386 dose of 10 mg/kg QW and that a dose of 15 mg/kg QW would provide greater exposure and yield further improvements in PFS.34 In a phase I monotherapy study, AMG 386 was tolerable at doses up to 30 mg/kg without apparent changes in toxicity.21

The results of this study and others714,35 indicate that antiangiogenic agents have clinical activity in advanced ovarian cancer. Specifically, two large phase III studies have recently demonstrated that bevacizumab in combination with chemotherapy improves PFS among patients with previously untreated advanced ovarian cancer.13,14 However, some studies have suggested that bevacizumab treatment in recurrent ovarian cancer may result in an increased risk of bowel perforation.13,1618 AMG 386 targets the Tie2-Ang1/2 signaling pathway. Therefore, it is not surprising that its toxicity profile is distinct from that of VEGF pathway inhibitors. The most frequently occurring AEs in the present study were peripheral edema, fatigue, nausea, and alopecia. There were no bowel perforations in either of the AMG 386 treatment arms. Two AEs emerged as likely being associated with AMG 386 treatment; these were peripheral edema and hypokalemia. Both occurred at higher overall incidence rates than with placebo, and there seemed to be more severe events (grade ≥ 3) of hypokalemia in patients who received AMG 386. The incidence rates of hypertension, thromboembolic events, proteinuria, and hemorrhage were similar across the active and placebo arms. Overall, the toxicity associated with AMG 386 treatment was manageable (the incidence of grade ≥ 3 AEs was similar in all arms) and was comparable to that previously reported for AMG 386 monotherapy21 and chemotherapy combination studies.22 No new toxicity signals were identified, and there were no dose-related trends with respect to AEs.

The key limitation of this phase II study is the relatively small number of patients enrolled. Although AMG 386 in combination with paclitaxel (arms A/B) showed clinically meaningful PFS results, the study was designed only to provide an estimation of the HRs for AMG 386 versus placebo. Larger studies will be required to provide a more definitive demonstration of the efficacy of AMG 386 in combination with chemotherapy in advanced recurrent ovarian cancer.

In summary, the results of this estimation study suggest that inhibiting angiogenesis via Tie2/angiopoietin pathway inhibition may offer an effective approach for the treatment of advanced recurrent ovarian cancer. Treatment with once-weekly AMG 386 plus paclitaxel may result in prolonged median PFS. The therapy was well tolerated with a distinct and manageable toxicity profile. The dose-response effect suggests that the maximum clinically effective dose may not have yet been reached. Once-weekly AMG 386 plus paclitaxel for the treatment of recurrent ovarian cancer is being further investigated in a phase III study (TRINOVA-1 [Trial in Ovarian Cancer 1];, NCT01204749).

© 2011 by American Society of Clinical Oncology

See accompanying article on page 441 and editorial on page 345; listen to the podcast by Dr Konstantinopoulos at

Supported by Amgen.

Presented in part at the 46th Annual Meeting of the American Society of Clinical Oncology, June 4-8, 2010, Chicago, IL, and the 35th European Society of Medical Oncology Annual Congress, October 8-12, 2010, Milan, Italy.

The views expressed herein are those of the authors and do not reflect the official policy or position of Brooke Army Medical Center, the US Army Medical Department, the US Army Office of the Surgeon General, the Department of the Army, the Department of Defense, or the US Government.

Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.

Clinical trial information can be found for the following: NCT00479817.

Although all authors completed the disclosure declaration, the following author(s) indicated a financial or other interest that is relevant to the subject matter under consideration in this article. Certain relationships marked with a “U” are those for which no compensation was received; those relationships marked with a “C” were compensated. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.

Employment or Leadership Position: Daniel E. Stepan, Amgen (C); David M. Weinreich, Amgen (C); Marjan Tassoudji, Amgen (C); Yu-Nien Sun, Amgen (C) Consultant or Advisory Role: None Stock Ownership: Daniel E. Stepan, Amgen; David M. Weinreich, Amgen; Marjan Tassoudji, Amgen; Yu-Nien Sun, Amgen Honoraria: Gary E. Richardson, Amgen (advisory board) Research Funding: Beth Y. Karlan, Amgen; Gary E. Richardson, Amgen – AMG 386 Study, Amgen – Denosumab Studies; Vincent L. Hansen, Amgen; John V. Brown III, Amgen; Charles A. Leath III, Amgen Expert Testimony: None Other Remuneration: None

Conception and design: Beth Y. Karlan, Daniel E. Stepan, David M. Weinreich, Marjan Tassoudji, Yu-Nien Sun

Provision of study materials or patients: Amit M. Oza, Diane M. Provencher, Setsuko K. Chambers, John V. Brown III, Allan Covens, Charles A. Leath III, Ignace B. Vergote

Collection and assembly of data: Beth Y. Karlan, Gary E. Richardson, Diane M. Provencher, Vincent L. Hansen, Martin Buck, Charles H. Pippitt Jr, John V. Brown III, Allan Covens, Raj V. Nagarkar, Margaret Davy, Hoa Nguyen, Yu-Nien Sun, Ignace B. Vergote

Data analysis and interpretation: Beth Y. Karlan, Amit M. Oza, Setsuko K. Chambers, Prafull Ghatage, John V. Brown III, Allan Covens, Charles A. Leath III, Daniel E. Stepan, David M. Weinreich, Marjan Tassoudji, Yu-Nien Sun, Ignace B. Vergote

Manuscript writing: All authors

Final approval of manuscript: All authors

1. AA Alvarez, HR Krigman, RS Whitaker , etal: The prognostic significance of angiogenesis in epithelial ovarian carcinoma Clin Cancer Res 5: 587591,1999 MedlineGoogle Scholar
2. PJ Paley: Angiogenesis in ovarian cancer: Molecular pathology and therapeutic strategies Curr Oncol Rep 4: 165174,2002 Crossref, MedlineGoogle Scholar
3. J Holash, PC Maisonpierre, D Compton , etal: Vessel cooption, regression, and growth in tumors mediated by angiopoietins and VEGF Science 284: 19941998,1999 Crossref, MedlineGoogle Scholar
4. AM Martoglio, BD Tom, M Starkey , etal: Changes in tumorigenesis- and angiogenesis-related gene transcript abundance profiles in ovarian cancer detected by tailored high density cDNA arrays Mol Med 6: 750765,2000 Crossref, MedlineGoogle Scholar
5. L Zhang, N Yang, JW Park , etal: Tumor-derived vascular endothelial growth factor up-regulates angiopoietin-2 in host endothelium and destabilizes host vasculature, supporting angiogenesis in ovarian cancer Cancer Res 63: 34033412,2003 MedlineGoogle Scholar
6. K Hata, J Udagawa, R Fujiwaki , etal: Expression of angiopoietin-1, angiopoietin-2, and Tie2 genes in normal ovary with corpus luteum and in ovarian cancer Oncology 62: 340348,2002 Crossref, MedlineGoogle Scholar
7. BJ Monk, E Han, CA Josephs-Cowan , etal: Salvage bevacizumab (rhuMAB VEGF)-based therapy after multiple prior cytotoxic regimens in advanced refractory epithelial ovarian cancer Gynecol Oncol 102: 140144,2006 Crossref, MedlineGoogle Scholar
8. RA Burger, MW Sill, BJ Monk , etal: Phase II trial of bevacizumab in persistent or recurrent epithelial ovarian cancer or primary peritoneal cancer: A Gynecologic Oncology Group Study J Clin Oncol 25: 51655171,2007 LinkGoogle Scholar
9. SA Cannistra, UA Matulonis, RT Penson , etal: Phase II study of bevacizumab in patients with platinum-resistant ovarian cancer or peritoneal serous cancer J Clin Oncol 25: 51805186,2007 LinkGoogle Scholar
10. JD Hurt, DL Richardson, LG Seamon , etal: Sustained progression-free survival with weekly paclitaxel and bevacizumab in recurrent ovarian cancer Gynecol Oncol 115: 396400,2009 Crossref, MedlineGoogle Scholar
11. AA Garcia, H Hirte, G Fleming , etal: Phase II clinical trial of bevacizumab and low-dose metronomic oral cyclophosphamide in recurrent ovarian cancer: A trial of the California, Chicago, and Princess Margaret Hospital phase II consortia J Clin Oncol 26: 7682,2008 LinkGoogle Scholar
12. RT Penson, DS Dizon, SA Cannistra , etal: Phase II study of carboplatin, paclitaxel, and bevacizumab with maintenance bevacizumab as first-line chemotherapy for advanced mullerian tumors J Clin Oncol 28: 154159,2010 LinkGoogle Scholar
13. RA Burger, MF Brady, MA Bookman , etal: Phase III trial of bevacizumab (BEV) in the primary treatment of advanced epithelial ovarian cancer (EOC), primary peritoneal cancer (PPC), or fallopian tube cancer (FTC): A Gynecologic Oncology Group study J Clin Oncol 28: 5s,2010 suppl abstr LBA1 LinkGoogle Scholar
14. T Perren, AM Swart, J Pfisterer , etal: ICON7: A phase III randomised Gynaecologic Cancer Intergroup Trial of concurrent bevacizumab and chemotherapy followed by maintenance bevacizumab, versus chemotherapy alone in women with newly diagnosed epithelial ovarian cancer (EOC), primary peritoneal (PPC) or fallopian tube cancer (FTC) Ann Oncol 21: viii2viii3,2010 abstr LBA4 Google Scholar
15. S Hapani, D Chu, S Wu: Risk of gastrointestinal perforation in patients with cancer treated with bevacizumab: A meta-analysis Lancet Oncol 10: 559568,2009 Crossref, MedlineGoogle Scholar
16. GP Sfakianos, TM Numnum, CB Halverson , etal: The risk of gastrointestinal perforation and/or fistula in patients with recurrent ovarian cancer receiving bevacizumab compared to standard chemotherapy: A retrospective cohort study Gynecol Oncol 114: 424426,2009 Crossref, MedlineGoogle Scholar
17. DL Richardson, FJ Backes, JD Hurt , etal: Which factors predict bowel complications in patients with recurrent epithelial ovarian cancer being treated with bevacizumab? Gynecol Oncol 118: 4751,2010 Crossref, MedlineGoogle Scholar
18. JP Diaz, WP Tew, O Zivanovic , etal: Incidence and management of bevacizumab-associated gastrointestinal perforations in patients with recurrent ovarian carcinoma Gynecol Oncol 116: 335339,2010 Crossref, MedlineGoogle Scholar
19. J Oliner, H Min, J Leal , etal: Suppression of angiogenesis and tumor growth by selective inhibition of angiopoietin-2 Cancer Cell 6: 507516,2004 Crossref, MedlineGoogle Scholar
20. A Coxon, J Bready, H Min , etal: Context-dependent role of angiopoietin-1 inhibition in the suppression of angiogenesis and tumor growth: Implications for AMG 386, an angiopoietin-1/2-neutralizing peptibody Mol Cancer Ther 9: 26412651,2010 Crossref, MedlineGoogle Scholar
21. RS Herbst, D Hong, L Chap , etal: Safety, pharmacokinetics, and antitumor activity of AMG 386, a selective angiopoietin inhibitor, in adult patients with advanced solid tumors J Clin Oncol 27: 35573565,2009 LinkGoogle Scholar
22. AC Mita, CH Takimoto, M Mita , etal: Phase 1 study of AMG 386, a selective angiopoietin 1/2-neutralizing peptibody, in combination with chemotherapy in adults with advanced solid tumors Clin Cancer Res 16: 30443056,2010 Crossref, MedlineGoogle Scholar
23. P Therasse, SG Arbuck, EA Eisenhauer , etal: New guidelines to evaluate the response to treatment in solid tumors: European Organization for Research and Treatment of Cancer, National Cancer Institute of the United States, National Cancer Institute of Canada J Natl Cancer Inst 92: 205216,2000 Crossref, MedlineGoogle Scholar
24. GJ Rustin, M Marples, AE Nelstrop , etal: Use of CA-125 to define progression of ovarian cancer in patients with persistently elevated levels J Clin Oncol 19: 40544057,2001 LinkGoogle Scholar
25. GJ Rustin: Use of CA-125 to assess response to new agents in ovarian cancer trials J Clin Oncol 21: 187s193s,2003 suppl Crossref, MedlineGoogle Scholar
26. Common Terminology Criteria for Adverse Events (CTCAE) version 3.0 Cancer Therapy Evaluation Program, National Cancer Institute Google Scholar
27. D Machin, YB Cheung, MKB Parmar: Survival Analysis: A Practical Approach 2006 ed 2 Chichester, United Kingdom John Wiley & Sons CrossrefGoogle Scholar
28. D Collett: Modelling Survival Data in Medical Research 2003 ed 2 Boca Raton, FL Chapman and Hall/CRC Press Google Scholar
29. R Brookmeyer, J Crowley: A confidence interval for the median survival time Biometrics 38: 2941,1982 CrossrefGoogle Scholar
30. RE Tarone: Tests for trend in life table analysis Biometrika 62: 679682,1975 CrossrefGoogle Scholar
31. GJ Rustin, RC Bast Jr, GJ Kelloff , etal: Use of CA-125 in clinical trial evaluation of new therapeutic drugs for ovarian cancer Clin Cancer Res 10: 39193926,2004 Crossref, MedlineGoogle Scholar
32. JA Bridgewater, AE Nelstrop, GJ Rustin , etal: Comparison of standard and CA-125 response criteria in patients with epithelial ovarian cancer treated with platinum or paclitaxel J Clin Oncol 17: 501508,1999 LinkGoogle Scholar
33. C Tian, M Markman, R Zaino , etal: CA-125 change after chemotherapy in prediction of treatment outcome among advanced mucinous and clear cell epithelial ovarian cancers: A Gynecologic Oncology Group study Cancer 115: 13951403,2009 Crossref, MedlineGoogle Scholar
34. J Lu, E Rasmussen, L Navale , etal: Exposure-response relationships of AMG 386 in combination with weekly paclitaxel in advanced ovarian cancer: Population pharmacokinetic/pharmacodynamic (PK/PD) modeling to facilitate phase III dose selection J Clin Oncol 28: 400s,2010 suppl abstr 5042 LinkGoogle Scholar
35. LS Downs Jr, PL Judson, PA Argenta , etal: A prospective randomized trial of thalidomide with topotecan compared with topotecan alone in women with recurrent epithelial ovarian carcinoma Cancer 112: 331339,2008 Crossref, MedlineGoogle Scholar

Beth Y. Karlan, Cedars-Sinai Medical Center, Los Angeles; John V. Brown III, Gynecologic Oncology Associates, Newport Beach; Daniel E. Stepan, David M. Weinreich, and Yu-Nien Sun, Amgen, Thousand Oaks; Marjan Tassoudji, Amgen, South San Francisco, CA; Vincent L. Hansen, Northern Utah Associates, Ogden, UT; Setsuko K. Chambers, Arizona Cancer Center, Tucson, AZ; Charles H. Pippitt Jr, Piedmont Hematology Oncology Associates, Winston-Salem, NC; Charles A. Leath III, Brooke Army Medical Center, Fort Sam Houston, TX; Hoa Nguyen, Sheridan Healthcare Corporation, Hollywood, FL; Amit M. Oza, Princess Margaret Hospital; Allan Covens, Toronto-Sunnybrook Regional Cancer Centre, Toronto, Ontario; Diane M. Provencher, Centre Hospitalier de l'Université de Montréal–Hôpital Notre-Dame, Montreal, Quebec; Prafull Ghatage, Tom Baker Cancer Centre, Calgary, Alberta, Canada; Gary E. Richardson, Cabrini Hospital, Melbourne, Victoria; Martin Buck, Sir Charles Gairdner Hospital, Perth, Western Australia; Margaret Davy, Royal Adelaide Hospital, Adelaide, South Australia, Australia; Raj V. Nagarkar, Curie Manavata Cancer Centre, Nashik, Maharashtra, India; and Ignace B. Vergote, University Hospital Leuven, Leuven, Belgium.


We thank Rebeca Melara, MS, and Teresa Wong, BS (Amgen), for pharmacokinetic sample and data analyses; Don Zhong, PhD (Amgen), for anti-AMG 386 antibody analysis; and Ali Hassan, PhD (Complete Healthcare Communications), whose work was funded by Amgen, and Beate D. Quednau, PhD (Amgen), for assistance in the preparation of this article.

Downloaded 1,905 times


No companion articles


DOI: 10.1200/JCO.2010.34.3178 Journal of Clinical Oncology 30, no. 4 (February 01, 2012) 362-371.

Published online December 19, 2011.

PMID: 22184370

ASCO Career Center