198P - Biomarker (BM) results from GOG-0218, a phase 3 trial of front-line bevacizumab (BV) + chemotherapy (CT) for ovarian cancer (OC)

Date 30 September 2012
Event ESMO Congress 2012
Session Poster presentation II
Topics Ovarian Cancer
Biomarkers
Presenter Michael Birrer
Authors M.J. Birrer1, H. Lankes2, R.A. Burger3, R. Mannel4, H. Homesley5, V. Henschel6, M. Sovak7, S.J. Scherer8, S. de Haas9, C. Pallaud10
  • 1Hematology/oncology, Massachusetts General Hospital, Cancer Center, 02114-2696 - Boston/US
  • 2Gynecologic Oncology Group, Statistical & Data Center, Roswell Park Cancer Institute, 14263 - Buffalo/US
  • 3Department Of Surgical Oncology, Fox Chase Cancer Center, 19111 - Philadelphia/US
  • 4Department Of Obstetrics And Gynecology, University of Oklahoma Health Sciences Center, 73104 - Oklahoma City/US
  • 5Department Of Obstetrics And Gynecology, Wake Forest University School of Medicine, 27157-1065 - Winston-Salem/US
  • 6Biostatistics, F Hoffmann-La Roche Ltd, 4070 - Basel/CH
  • 7Pharma Development, Genentech, Inc., South San Francisco/US
  • 8Pharma Development, Genentech, Inc., 94080 - South San Francisco/US
  • 9Pharma Development, F. Hoffmann-La Roche Ltd, 4070 - Basel/CH
  • 10Pharma Development Oncology Department, F. Hoffmann-La Roche Ltd., Basel/CH

Abstract

Background

GOG-0218 showed significantly improved progression-free survival (PFS) in patients (pts) receiving BV 15 mg/kg q3w concurrently with CT and continued alone until progression or for up to 15 mo. GOG-0218 includes extensive BM evaluation to identify pts benefitting most from BV. Analysis of plasma VEGF-A and VEGFR-2 was prioritised based on encouraging findings in BV trials in several tumour types.

Table: 198P

PFS OS
Median, mo HR Median, mo HR
Subgroup PL BV PL BV
VEGF-A
≤ median 12.3 18.6 0.52 43.0 48.5 0.87
> median 11.0 17.5 0.70 33.9 41.8 0.78
≤ Q3 12.3 18.6 0.59 45.1 48.5 0.89
> Q3 9.7 13.8 0.67 28.6 37.7 0.72
VEGFR-2
≤ median 12.0 18.0 0.68 35.1 40.6 0.90
> median 10.4 18.2 0.53 38.4 48.5 0.69
≤ Q3 11.0 17.3 0.63 38.2 41.8 0.87
> Q3 12.5 22.1 0.46 38.6 0.59

Q = quartile

Methods

Pts with newly diagnosed stage IV or macroscopic optimal stage III OC were randomised to 6 cycles (c) of CT with: placebo (PL) c2−22 (arm A); BV c2−6 → PL c7−22 (arm B); or BV c2−22 (arm C). Post-surgery pre-CT plasma samples were analysed using a multiplex ELISA. Baseline (BL) BM levels were used to dichotomise pts. Potential interactions between BM levels and PFS (1° endpoint) and overall survival (OS; 2° endpoint) were tested using log-rank testing and Cox regression approaches.

Results

Post-surgery samples were available from 582 of 1248 pts in arms A and C. Median BL VEGF-A and VEGFR-2 levels were 144.3 pg/mL and 14.7 ng/mL, respectively. No significant interaction was seen at α = 0.05. Exploratory analyses with other cut-offs are hypothesis generating for potential predictive (VEGF-A and VEGFR-2) or prognostic (VEGF-A: OS) value. Exploratory analyses revealed no correlation between plasma VEGF-A and time since surgery.

Conclusions

The potential prognostic (VEGF-A) and predictive (VEGF-A, VEGFR-2) value seen in breast, pancreatic and gastric cancers was not apparent in post-surgery samples from GOG-0218 using a median cut-off. Results with other cut-offs provide a rationale for further investigation of potential prognostic and predictive value. Findings may reflect differing biology and interplay between VEGF-A isoforms across tumour types. The possible impact of pre- vs post-surgery samples is also being investigated.

Disclosure

R.A. Burger: RB has served on Advisory Boards for Roche/Genentech.

R. Mannel: RM has served on Advisory Boards for Genentech.

V. Henschel: VH is employed by and holds shares in Roche.

M. Sovak: MS is an employee of Genentech.

S.J. Scherer: SS is an employee of Genentech Inc.

S. De Haas: SdH is an empoloyee of F Hoffmann-La Roche Ltd.

C. Pallaud: CP is an employee of F Hoffmann-La Roche Ltd.

All other authors have declared no conflicts of interest.