br Methods We retrospectively collected pre
Methods: We retrospectively collected pre- and postoperative serum samples as well as tumour tissues and ad-jacent normal tissues from 100 GC patients. Serum samples from non-cancerous patients were served as controls (n = 50). A high-throughput protein detection technology, multiplex proximity extension assays (PEA), was ap-plied to measure levels of over 300 proteins. Alteration of each protein was analysed by univariate analysis. Elastic-net logistic regression was performed to select serum proteins into the diagnostic model.
Findings: We identified 19 serum proteins (CEACAM5, CA9, MSLN, CCL20, SCF, TGF-alpha, MMP-1, MMP-10, IGF-1, CDCP1, PPIA, DDAH-1, HMOX-1, FLI1, IL-7, ZBTB-17, APBB1IP, KAZALD-1, and ADAMTS-15) that together dis-tinguish GC cases from controls with a diagnostic sensitivity of 93%, specificity of 100%, and area under receiver operating characteristic curve (AUC) of 0·99 (95% CI: 0·98–1). Moreover, the 19-serum protein signature pro-vided an increased diagnostic capacity in patients at TNM I-II stage (sensitivity 89%, specificity 100%, AUC 0·99) and in patients with high microsatellite instability (MSI) (91%, 98%, and 0·99) compared to individual pro-teins. These promising results will inspire a large-scale independent cohort study to be pursued for validating the proposed protein signature.
Interpretation: Based on targeted proteomics and elastic-net logistic regression, we identified a 19-serum protein signature which could contribute to clinical GC diagnosis, especially for patients at early stage and those with high MSI.
Fund: This study was supported by a European H2020-Marie Skłodowska-Curie Innovative Training Networks grant (316,929, GastricGlycoExplorer). Funder had no influence on trial design, data evaluation, and interpretation.
© 2019 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http:// creativecommons.org/licenses/by-nc-nd/4.0/).
Abbreviations: ADAMTS-15, A disintegrin and metalloproteinase with thrombospondin motifs 15; APBB1IP, Amyloid beta A4 precursor protein-binding family B member 1-interacting protein; CA9, Carbonic anhydrase 9; CCL20, C-C motif chemokine 20; CDCP1, CUB domain-containing protein 1; CEACAM5, Carcinoembryonic antigen-related cell adhesion molecule 5; DDAH1, dimethylarginine dimethylaminohydrolase 1; FLI-1, Friend leukemia Lovastatin 1 transcription factor; GCNT1, beta-1,3-galactosyl-O-glycosyl-glycoprotein beta-1,6-N-acetylglucosaminyltransferase; HMOX1, Heme oxygenase 1; IGF1, Insulin-like growth factor I; IL-7, Interleukin-7; KAZALD1, Kazal-type serine protease inhibitor domain containing pro-tein 1; MSLN, Mesothelin; MMP-1, Matrix metalloproteinase-1; MMP-10, Matrix metalloproteinase-10; PPIA, Peptidyl-prolyl cis-trans isomerase A; SCF, Stem cell factor/KIT ligand; TGF alpha, Transforming growth factor alpha; ZBTB-17, Zinc finger and BTB domain-containing protein 17.
Corresponding author at: Department of Immunology, Genetics and Pathology, Science for Life Laboratory, Uppsala University, Husargatan 3, SE-751 08 Uppsala, Sweden.
E-mail addresses: [email protected] (Q. Shen), [email protected] (K. Polom), [email protected] (C. Williams), [email protected] (F.M.S. de Oliveira),
[email protected] (M. Guergova-Kuras), [email protected] (F. Lisacek), [email protected] (N.G. Karlsson), [email protected] (F. Roviello), [email protected] (M. Kamali-Moghaddam).
Research in context
Evidence before electrostatic attraction study
The current clinically used diagnostic biomarkers for gastric can-cer (GC) such as carcinoembryonic antigen (CEA), cancer antigen 19-9 (CA19-9), and cancer antigen 72-4 (CA72-4) are neither suf-ficiently sensitive nor specific. The identification of novel reliable markers and development of new technologies are still urgent for GC detection and hence improving the survival of patients. Large-scale proteome screening in non-invasive biological speci-mens is a strategy that provides the possibility of discovering new biomarkers. We performed a PubMed database search for studies focused on GC biomarker discovery published until May 2018 using the following terms: “biomarker” AND “proteomic” AND (“gastric cancer” OR “stomach cancer”). The criteria for inclu-sion were GC diagnostic biomarker studies using any human ma-terials such as serum, plasma, tissue, gastric fluid, urine, circulating cells, or exosomes. Additional studies were identified by surveying the references associated with these primary publi-cations. Studies of biomarkers for prognosis or prediction were not included. Any reports on cell lines or animals were excluded. As a result, almost all studies were based on different types of mass spectrometry or antibody-based tissue microarrays. We have previously developed and optimized in situ proximity ligation assay (PLA) and solid-phase PLA for single protein detection as well as for detection of single protein post-translational modifica-tion such as glycosylation and phosphorylation, which have been applied in GC tissue and serum specimens. Here, we ex-tended our research by applying a targeted proteomics approach, multiplex proximity extension assay (PEA), for large scale protein measurement in GC serum and tissue lysate samples in order to identify potential biomarkers for GC diagnosis.