In:
Cancer Research, American Association for Cancer Research (AACR), Vol. 75, No. 15_Supplement ( 2015-08-01), p. 582-582
Abstract:
The humanized monoclonal IgG2 antibody abituzumab (EMD 525797, DI17E6) inhibits αν integrin heterodimers. The phase I/II POSEIDON trial (NCT01008475) demonstrated a trend toward improved overall survival (OS) in codon 2 KRAS wild-type mCRC. Integrin ανβ6 expression (immunohistochemistry based) above median identified a poor prognosis patient (pt) subgroup and was predictive of improved OS. Previous exploratory analysis identified individual plasma proteins as negatively prognostic and predictive for OS with abituzumab treatment. We report here the results of exploratory bioinformatic analyses aimed at identifying prognostic or predictive pretreatment plasma protein signatures. Plasma proteins were analyzed (highly protein-specific aptamers [SomaLogic system]) in samples taken at pretreatment screening. Analyses focused on 888 proteins that passed quality control. Groups of genes, termed signatures, with common function or regulation captured in the MSigDB database formed the basis for our analyses. We identified a set of signatures comprising genes with coherent expression patterns that we analyzed for the association of their scores with OS in the treatment arms. Cox models were analyzed per signature to assess the association of continuous and median-thresholded signature scores with OS in the standard of care (SoC) arm to identify prognostic markers, in the biomarker high/low groups, or across all patients to identify predictive markers. Pretreatment plasma samples from 192/216 pts (122 samples from pts treated with abituzumab; 70 from pts treated with SoC alone) with full SomaLogic data were available for analysis of protein levels and signature scores. Sixty-two signatures comprised genes with coherent expression patterns across samples. Among these, we identified a poor-prognosis signature comprising five proteins expressed in neutrophils (CD48, MPO, ELANE, CTSG, PECAM1; HR = 3.28 [1.72-6.28]; uncorrected p = 0.0003; Bonferroni-corrected p = 0.02). Other signatures predicted benefit from abituzumab when scores were higher (or lower) than median, but were not prognostic in the SoC arm. These signatures yielded HRs of 0.40-0.55 and uncorrected p values & lt;0.05 when OS of treated and untreated pts were compared in signature-selected populations (results did not withstand multiple testing corrections). The functional coherence of the identified prognostic signature, i.e. all proteins are expressed in neutrophils, is striking. High neutrophil protein levels identify poor prognosis pts. This signature may indicate an antitumor neutrophilic response in these pts, and may be a surrogate for tumor burden. The predictive plasma protein signatures constitute stratification hypotheses to be assessed in future exploratory biomarker analyses in abituzumab trials in mCRC. Citation Format: Josef Straub, Eike Staub, Miriam Lohr, Giorgio Massimini, Elena Élez, Josep Tabernero. Prognostic and predictive value of plasma protein signatures in a phase I/II trial of abituzumab combined with cetuximab/irinotecan in second-line KRAS wild-type metastatic colorectal cancer (mCRC). [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 582. doi:10.1158/1538-7445.AM2015-582
Type of Medium:
Online Resource
ISSN:
0008-5472
,
1538-7445
DOI:
10.1158/1538-7445.AM2015-582
Language:
English
Publisher:
American Association for Cancer Research (AACR)
Publication Date:
2015
detail.hit.zdb_id:
2036785-5
detail.hit.zdb_id:
1432-1
detail.hit.zdb_id:
410466-3
