Chapter 2.1.1

Mario Plebani, Martina Zaninotto,  and Giuseppe Lippi

The rapid advances in medical research that have occurred over the past few years have allowed us to dissect molecular signatures and functional pathways that underlie disease initiation and progression, as well as to identify molecular profiles related to disease subtypes in order to determine their natural course, prognosis, and responsiveness to therapies (Dammann and Weber, 2012). The “omics” revolution of the past 15 years has represented the most compelling stimulus in personalized medicine that, in turn, should be simply defined as “getting the right treatment to the right patient at the right dose and schedule at the right time” (Schilsky, 2009). As a matter of fact, among the 20 most-cited papers in molecular biology and genetics that have been published in the past decade, 13 entail omics methods or applications (Ioannidis, 2010).

Omics is an English-language neologism that refers to a field of study in biology focusing on large-scale and holistic data, as derived from its root of Greek origin which refers to wholeness or to completion. Initially, the suffix omics had been used in the word genome, a popular word for the complete genetic makeup of an organism, and later, in the term proteome. Genomics and proteomics succinctly describe a new way of holistic analysis of complete genomes and proteomes, and the success of these terms led to more emphasis in the trend of using omics as a convenient term to describe holistic ways of looking at complex systems, particularly in biology.

Fields with names like genomics (genetic complement), transcriptomics (gene expression), proteomics (protein synthesis and signaling), metabolomics (concentration and fluxes of cellular metabolites), metabonomics (systemic profiling through the analysis of biological fluids), and cytomics (the study of cell systems—cytomes—at a single cell level) have been introduced in medicine with increasing emphasis (Plebani, 2005). However, beyond these terms, multiple “omics” fields, with names like epigenomics, ribonomics, epigenomics, oncopeptidomics, lipidomics, glycomics, spliceomics, and interactomics, have been similarly explored regarding molecular biomarkers for the diagnosis and prognosis of human diseases.

Each of these emerging disciplines grouped under the umbrella of the term omics shares the simultaneous characterization of dozens, hundreds, or thousands of genes (genomics), gene transcripts (transcriptomics), or proteins (proteomics) and other molecules, that in aggregate and in parallel should be coupled with sophisticated bioinformatics to reveal aspects of biological function that cannot be culled from traditional linear methods of discovery (Finn, 2007). While an increasing body of literature has been produced to prove that “omics” will irrevocably modify the practice of medicine, that change has yet to occur and its precise details are still unclear. The reasonable assumption that the application of “omics” research will be riddled with difficulties has led to a much better appreciation of concepts of knowledge translation, translational research, and translational medicine.

The paradigm of obstacles in translating new “omics” insights into clinical practice is a study reporting that a blood test, based on pattern-recognition proteomics analysis of serum, was nearly 100% sensitive and specific for detecting ovarian cancer and was possibly useful for screening (Petricoin et al., 2002a). The approach involved the analysis of a drop of blood using mass spectrometry, resulting in a large number of mass-to-charge ratio peaks (15,000 to 300,000 peaks, depending on technology), that were then subjected to pattern-recognition analysis to derive an algorithm that discriminates patients with cancer from those without. Which substances cause the peak (e.g., proteins, peptides, or something else) was yet unknown, as it was unclear whether these substances were released by tumor cells or by their microenvironment.

After this and further supporting work (Adam et al., 2002; Drake et al., 2003; Petricoin et al., 2002b; Vlahou et al., 2001; Zhu et al., 2003), commercial laboratories planned to market a test in late 2003 or early 2004, but plans were delayed by the U.S. Food and Drug Administration (FDA). Questions were raised about whether technological results were reproducible and reliable enough for widespread application in clinical practice.

During the past few years, a large number of scientists have been able to identify other candidate protein disease biomarker profiles using patient research study sets and to achieve high diagnostic sensitivity and specificity in blinded test sets. Nevertheless, translating these research findings to useful and reliable clinical tests has been the most challenging accomplishment. Clinical translation of promising ion fingerprints has been hampered by “sample collection bias, interfering substances, biomarker perishability, laboratory-to-laboratory variability, surface-enhanced laser desorption ionization chip discontinuance and surface lot changes, and the stringent dependence of the ion signature on the subtleties of the reagent composition and incubation protocols” (Liotta and Petricoin, 2008, p.3). Systematic biases arising from preanalytical variables seem to represent a relevant issue. Examples of non-disease-associated factors include (1) within-class biological variability, which may comprise unknown subphenotypes among study populations; (2) preanalytical variables, such as systematic differences in study populations and/or sample collection, handling, and preprocessing procedures; (3) analytical variables, such as inconsistency in instrument conditions, resulting in poor reproducibility; and (4) measurement imprecision (Hortin et al., 2006). Biological variability, in particular, may entail potential diurnal variation in protein expression, thus making standardization of sample collection time virtually mandatory. An evaluation of the effects of gender, age, ethnicity, pathophysiological conditions, and benign disorders is also crucial for understanding other possible effects on protein profiling expression. Regarding preanalytical conditions (e.g., collection practices, sample handling, and storage), these may differ from institution to institution, thus influencing the detection of proteins present in biological fluids. Standardization and use of specimens from multiple institutions are hence necessary to reliably demonstrate efficiency and reproducibility of protein profiling (Lippi et al., 2006; Banks, 2008). Although these preanalytical influences have been recognized for a long time, their impact is likely to be greater in proteomics studies, given the simultaneous analysis of several proteins, resolution of multiple forms of proteins, and detection of peptide fragments arising from active cleavage processes. Moreover, relatively few studies have been performed in such a way that quality control, an essential and quality-related feature, should be incorporated in proteomic experimental protocols (Hortin, 2005). Reproducibility studies performed with adequate control materials are prerequisites for safe introduction of proteomic techniques in clinical laboratory practice. Table 2.1 summarizes the major problems in translating proteomics insights into clinical practice.

It is now clearly accepted that the lack of standardization in how specimens are collected, handled, and stored represents one of the major hurdles to progress in the hunt for new and effective biomarkers (Poste, 2011). Nevertheless, the significance of assay technical quality has recently been underpinned. Diamandis, for example, has elegantly demonstrated that the assay for a new promising marker for prostate carcinoma (Diamandis, 2007) was strongly affected by severe methodological drawbacks, including its dependence on the total protein content, namely the albumin concentration in serum. The major limitations of this assay are even more important when considering the apparently spectacular clinical results that have been highly publicized to the media, whereas potential following failures were not. A review of the literature on translational research in oncology has revealed that most of the 939 publications on prognostic factors for patients with breast cancer that have appeared over a 20-year period were based on research assays with poor evidence of robustness or analytical validity (Simon, 2008). This fact should lead journal editors to ask for more robustness of the analytical techniques used for quantification of novel, putative biomarkers (Anderson et al., 2013), since problems such as data manipulation, poor experimental design, reviewer’s bias, and overinterpretation of results are reported with increasing frequency (Diamandis, 2006).

Current limitations and open questions regarding clinical proteomics reflect a lack of appreciation of the many steps involved, thus including evaluation of pre-, intra-, and postanalytical issues; inter-laboratory performance; standardization; harmonization; and quality control, which are all needed to progress from method discovery to clinical practice (Plebani and Laposata,...