(a) Validation of a prognostic protein signature

Prognostic signatures for NSCLC were developed from microarray-based mRNA expression profiling.25–31 These profiles are difficult to translate and reproduce in the clinical setting. However, this problem can be overcome by translating these signatures to their corresponding prognostic protein signature, using IHC to quantify their expression levels in paraffin-embedded tissue samples. The project will develop a protein signature capable of classifying early-stage NSCLC patients into subgroups according to their risk for recurrence.

Based on the findings of earlier studies, we have selected a group of 15–20 proteins that can be used for classifying early-stage NSCLC patients into 2 groups, according to their clinical course. For this purpose, we used an algorithm generated by the biomarker laboratory of the Center for Applied Medical Research of the University of Navarre. The proteins selected include a panel of 5 RNA binding proteins that have been shown to be prognostic markers in cohorts of patients with early-stage NSCLC.32 On this basis, we were able to generate a prognostic protein signature that will be validated in our study cohorts.

(b) ALK, Ros 1 and Pdl-1 determinations

ALK immunohistochemistry assay using two different, validated antibodies, D5F3 and 5a4, will be performed to study the correlation between these genes. All positive results will be confirmed using fluorescence in situ hybridization. IHC study of Ros-1 will be performed to ascertain the percentage of molecular changes. The same technique will be used to determine Pdl-1 and its relationship with survival. Previous studies in Pdl-1 in LC have been based on a limited number of cases; however, studies of Pdl-1 expression in mesothelioma reported a relationship between Pdl-1 expression and patient survival.

(c) Micro-RNA

Micro-RNA (miRNA) are non-coding molecules of 19–25 nucleotides in length. Their function is to regulate the post-transcriptional expression of multiple genes. They are involved in many different cell cycle processes,33 and have been shown to play a leading role in cancer progression.34,35 Tumor-derived miRNAs can enter the bloodstream,36 and their expression profile in serum and plasma is disease-specific.37 Because of this, and because they are stable, easily detectable molecules, miRNAs are now the subject of intense study.38,39 In this project, we will select miRNA to be studied on the basis of the following criteria:

• (1)

  miRNAs that have been identified in at least 3 published studies as potential biomarkers of NSCLC prognosis.

• (2)

  EGFR-, ALK- and KRAS-regulating miRNAs, validated in functional studies, retrieved from a search of a database of miRNA target prediction resources.

• (3)

  miRNAs that regulate other genes of interest in the context of NSCLS that have been studied by researchers taking part in the study (see Supplementary information).

(d) Pentose phosphate pathways

The pentose phosphate pathway plays an important role in the development of cancer because it generates the nitrogenous base needed to synthesize DNA. The non-oxidative phase of the pathway also synthesizes nicotinamide adenine dinucleotide phosphate, which is essential in the synthesis of the fatty acids needed to form cell membranes.

Because of its importance in biomolecular synthesis and the proliferation of the tumor itself, this pathway is regulated by some of the most important proto-oncogenes and tumor-suppressor genes. Research has shown that tumors with a mutated KRAS gene present increased levels of glucose-6-phosphate dehydrogenase (G6PD),40 and cell proliferation via this molecule is directly related with proteins such as TAp73.41 G6PD inhibition can slow down tumor growth,42 while more aggressive tumor growth has been correlated with over-expression of TKTL1.43 In the specific case of LC, recent studies suggest that both TKT and G6PD can be over-expressed,44 which justifies the interest of researchers in pentose phosphate enzymes as potential biomarkers of LC (see Supplementary information).

(e) Inflammatory markers

Chronic inflammation plays a key role in the pathogenesis of NSCLC, particularly in patients with COPD.45,46 Experimental studies have shown increased expression of certain cytokines such as interleukin (IL)-6, IL-8, and IL-10 in patients with COPD and a history of smoking. These proteins promote the lymphocytic inflammatory response by induction of the cyclooxygenase-2 enzyme. These cytokines can inhibit apoptosis, interfere with cell repair mechanisms, and encourage angiogenesis. Other inflammatory mechanisms implicated in LC are mediated by the angiogenesis-regulating vascular endothelial growth factor, the epidermal growth factor, the tumor necrosis factor (TNF)-alpha, the nuclear factor (NF)-kB, and the tumor growth factor-beta (TGF-beta). Chronic inflammation can be important in amplifying mutagenesis, facilitating tumor growth, and increasing the frequency of metastasis.

In this project, IHC assay will be used to quantify the following markers:

• (1)

  Inflammatory cell infiltrate in H&E-stained histological samples.

• (2)

  Specific cell infiltrate, using specific antibodies to identify leukocytes (anti-CD45) and macrophages (anti-CD68).

• (3)

  Levels of inflammatory molecules involved in tumor development processes:

• (3.1)

  Cyclooxygenase-2 (COX-2), involved in angiogenesis, tumor invasion, and immune system suppression.

• (3.2)

  TNF-alpha, a molecule that induces COX-2 expression through mitogen-activated protein kinases (MAPKs) signaling pathways.

• (3.3)

  IL-1beta and IL-8, which induce COX-2 expression and are involved in neutrophil recruitment and tumor progression.

• (4)

  NSCLC-related signaling pathways, such as p65 (NF-kB) and p38 (MAPK).

• (5)

  Determination of the extent of cell proliferation in tumors by quantifying nuclear protein ki-67 expression as a specific marker of tissue growth (see Supplementary information).

(f) Stromal cell activity markers

An important signaling source for cancer cells is stromal change, characterized by an over-abundance of extracellular matrix components and activated fibroblasts, and by abnormal composition of the extracellular matrix. Stromal changes are also observed in fibrogenetic processes. It is important to identify the extent to which signals from activated stroma contribute to LC progression, as this can guide therapeutic strategies. Changes in tumor stroma have been studied in the context of breast cancer, but little is known about their association with LC. Since myofibroblasts are the most abundant cells in activated stroma, we will use IHC to study alpha smooth muscle actin expression (α-SMA). To distinguish myofibroblasts from other cells, we will examine desmin and vimentin expression,47 and the presence of extracellular matrix cicatrizing components will be determined by analyzing collagen and tenascin expression48 (see Supplementary information).

DNA methylation helps regulate gene expression, and studies have shown that aberrant DNA methylation is frequently found in malignancies. Methylation is found in CpG islands, which are concentrated in promoter regions and in the first exon of protein-coding genes. Up to 60% of these genes contain CpG islands and their promoter sequences, and their methylated status is inversely correlated with transcription.49

The following genes: p 16 (cyclin-dependent kinase inhibitor 2) (CDKN2A/p16), H-cadherin (CDH13), adenomatous polyposis coli (APC), Ras association domain family 1 gene (RASSF1A), and death-associated protein kinase 1 (DAPK) are important in the development of NSCLC, and are frequently methylated in tumor tissue.50,51 Methylation of p16, DAPK, CDH13, APC and RASSF1A has been associated with an increased risk from LC,52 and early recurrence.51 It could be important to determine the methylation status of these genes for both diagnosis and prognosis (see Supplementary information).

The characteristics of the inflammatory response in peripheral blood cells, which is partly mediated by the generation of oxidative stress,53 can affect prognosis and therapeutic response in LC. Cancer-related inflammation is also a new therapeutic target,54 and various LC-related inflammatory markers in blood have been described. In this project, we will measure circulating levels of acute phase reactants, inflammatory cytokines, and markers of oxidative stress.

• (1)

  Acute phase reactants:

• (1.1)

  C-reactive protein is a non-specific reactant associated with LC through various mechanisms, including tumor-induced inflammatory response, immune response to tumor antigens, and cytokine production by the tumor cells themselves. High levels of C-reactive protein are predictors of mortality.55

• (1.2)

  Fibrinogen has been associated with increased risk for LC.56

• (2)

  Inflammatory cytokines:

• (2.1)

  IL-6, IL-8 and TNF-α can be elevated in LC due to mechanisms similar to those described for C-reactive protein, and increased circulating levels of these cytokines have been associated with LC.57

• (2.2)

  Over-expression of the inflammatory cytokine IL-10 has been associated with CP progression.58

• (2.3)

  Interferon gamma deficiency has been associated with LC in experimental models,59 and in human series.60

• (3)

  Oxidative stress:

  Circulating oxidative stress has been associated with risk for LC,61 advanced disease,62 and therapeutic response63 (see Supplementary information).