1194P - Tissue and serum biomarker results from the phase II innovations study in non-small cell lung cancer (NSCLC)

Date 29 September 2012
Event ESMO Congress 2012
Session Poster presentation I
Topics Biomarkers
Non-Small-Cell Lung Cancer, Locally Advanced
Presenter Niels Reinmuth
Authors N. Reinmuth1, S.J. Scherer2, R. Penzel3, P.A. Schnabel4, M. Meister5, J. Kiemle-Kallee6, A. Reuss7, J.R. Fischer8, M. Wolf9, M. Thomas5
  • 1Department Of Thoracic Oncology, Thoraxklinik Heidelberg, University of Heidelberg, 69126 - Heidelberg/DE
  • 2Pharma Development Department, F. Hoffmann-La Roche Ltd/Genentech, San Francisco/US
  • 3Institute Of Pathology, University of Heidelberg, Heidelberg/DE
  • 4Department Of Pathology, University of Heidelberg, Section for Thoracic Pathology, Heidelberg/DE
  • 5Department Of Thoracic Oncology, Thoraxklinik Heidelberg, University of Heidelberg, Heidelberg/DE
  • 6Pharma Ag, Roch Pharma AG, Grenzach/DE
  • 7Coordinating Center For Clinical Trials, Phillips-University Marburg, Marburg/DE
  • 8Department Of Medical Oncology, Klinik Löwenstein, DE-74245 - Löwenstein/DE
  • 9Division Of Haematology And Oncology, Klinikum Kassel, Kassel/DE



The phase II INNOVATIONS study was a randomised trial of 224 patients (pts) with advanced non-squamous NSCLC who received 1st-line treatment with either erlotinib (E) plus bevacizumab (B) or cisplatin (P) plus gemcitabine (G) plus B. PGB was superior to EB for progression-free survival (PFS). Pts with EGFR mutation-positive NSCLC were most likely to benefit from EB [1]. This abstract reports exploratory analyses from the optional biomarker (BM) sub-study for a number of candidate BMs commonly examined in previous NSCLC bevacizumab trials.


Tissue (t) markers were determined by immunohistochemistry H-score. Serum (s) markers were analysed on day 0 and 43 using a novel protein array/ELISA. Median baseline values were used to dichotomise pts (low vs high BM) and to correlate BMs with PFS/overall survival (OS) (Cox model). Analyses were not adjusted for multiple testing.


BM population represented 88% of the intent-to-treat population; baseline characteristics were well balanced. Tissue (n = 198) and serum samples (day 0, n = 184; day 43, n = 151) were provided for BM analysis. Soluble VEGFA was not assessed; only serum samples were available. Serum analyses: PFS was longer for the BM population with low baseline IL8 and FLT4 (HR = 0.55, p < 0.05 and HR = 0.50, p < 0.05, respectively); the trend was consistent in EB and PGB treatment arms. Similar results were seen for day 43 analyses (PFS/OS). Low baseline FLT1 was associated with longer PFS in PGB pts (HR = 0.53, p < 0.05). Low ICAM levels were associated with better clinical outcomes (PGB arm). Tissue analyses: None of the three tumour markers tested showed significant association with PFS/OS (p > 0.05). Significant correlations (p < 0.05) were observed between tumour and serum markers: tVEGFR1 and tVEGFR2/tVEGF; tVEGFR2 and tVEGF; sFLT1 and sTie2; sKDR and sICAM; sFLT4 and sICAM.


Several serum baseline candidate BMs showed a statistically significant correlation with OS/PFS, in particular FLT4/IL8. The choice of treatment partner (E/PG) for bevacizumab may influence the value of BMs. Further examination of the data from this and other bevacizumab trials may provide greater insight. [1] Thomas et al. J Clin Oncol 2011;29:7504


N. Reinmuth: Advisory Board, Lilly Corporate sponsored research, F.Hoffmann-La Roche Ltd.

S.J. Scherer: Currently employed by Roche/Genentech.

J. Kiemle-Kallee: Currently employed by Roche.

M. Thomas: Attended advisory boards for Roche and Lilly. Received research funding from Roche and Lilly.

All other authors have declared no conflicts of interest.