Functional Genomics Towards Personalized Healthcare and Systems Medicine

Abstract

The 5th Biologie Prospective Santorini Conference explored the themes of systems biology, nutrigenomics and pharmacogenomics, all of which are related to personalized health, personalized therapy and personalized medicine. The conference started with a satellite meeting on genome-wide scan studies where the need for simplified models, the quality of the phenotypes and the input of epigenetics were dominant remarks. All of the omics approaches were then applied during the 3 days‘ sessions to multifactorial diseases (e.g., diabetes, atherosclerosis, cancer and inflammation) and often focused on gene–gene and gene–environment interactions. Afterwards, a fundamental session on drug metabolism, theranostics and pharmacogenetics and their practical aspects showed that the translation to clinical practice is finally happening although much slower than expected.

The 5th Biologie Prospective Santorini Conference gathered 170 participants from 30 countries, originating from academic institutions and industries specialized in genetic-based research, particularly genome-wide association studies (GWAS), with application in chronic diseases (e.g., cardiovascular diseases [CVDs], cancer, inflammation and infection) and in pharmacogenomics and nutrigenomics.

The meeting was organized by the Institut National de la Santé et de la Recherche Médicale (INSERM) team ‘Génétique Cardiovasculaire‘ and by Biologie Prospective with the collaboration of the American Association of Clinical Chemistry (AACC) divisions and the Chinese Society of Laboratory Medicine (CSLM), under the auspices of international organizations (International Federation of Clinical Chemistry and Laboratory Medicine [IFCC], International Atherosclerosis Society [IAS], European Atherosclerosis Society [EAS], International Society of Nutrigenetics/Nutrigenomics [ISNN], International Society of Pharmacogenomics [ISP], Personalized Medicine Coalition [PMC], Asian–Pacific Congress of Clinical Biochemistry [APCCB], European Personalised Medicine Diagnostics [EPEMED] and the European Network for Pharmacogenetics) and national societies (New Society of French Atherosclerosis [NSFA], French Society of Clinical Biology [SFBC], Greek Society of Clinical Chemistry – Clinical Biochemistry [EEKX-KB] and Hellenic Society of Biochemistry and Molecular Biology [HSBMB]).

We thank all our sponsors, in particular the Gold (Bühlmann Laboratories, Randox and Genomatix) and Silver sponsors (Nestlé, Roche, Beckman Coulter Genomics, BioM, Illumina, Future Diagnostics, AbD Serotec, Affymetrix, Bruker, Diasys) and also Protagen, GlaxoSmithKline, Chembio Diagnostic Systems, Servier and Daiichi-Sankyo.

The 3 days of the conference raised intensive discussions on the genetics and omics technologies‘ influences on personalized health:

▪	Day 1: systems biology and infectious diseases;
▪	Day 2: inflammation, CVDs, obesity and nutrigenomics;
▪	Day 3: pharmacogenomics and cancer. A round table on ‘pharmacogenomics and personalized medicine in clinical practice‘ ended the conference.

We briefly present here the most important points made by the speakers during the symposium.

Satellite meeting: GWAS

In its 5th edition, the Santorini Conference – Functional Genomics towards Personalized Health Care – initiated its first satellite meeting devoted to GWAS under the auspices of Philippe Froguel (Unité Mixte de Recherche [UMR] 8199, Centre National de la Recherche Scientifique [CNRS] and Pasteur Institute, Lille, France and Genomic Medicine, Imperial College London, UK) and Sophie Visvikis-Siest (EA4373 ‘Génétique Cardiovasculaire‘, Université Henri Poincaré, Nancy, France). To this occasion, leading experts in this area were invited: John PA Ioannidis (Stanford University, CA, USA), Heather J Cordell (Institute of Human Genetics, Newcastle University, UK), Allen Roses (Deane Drug Discovery Institute, NC, USA), Eleftheria Zeggini (Wellcome Trust Sanger Institute, Cambridge, UK) and Martin Seifert (Genomatix Software, Munich, Germany) to debate the contribution of GWAS in the understanding of common disorders and the future expectations.

The meeting started with a statement of the art by John Ioannidis, recounting the evolution, within a decade, of human genome epidemiology from a discipline of case series or case–control studies of a few dozen participants to the accrual of large-scale teams and consortia including thousands of participants. The scope of findings has varied across different fields: 95 novel loci identified for lipid levels, 40 for Type 2 diabetes (T2D), 45 for Crohn‘s disease and 80 for bone mineral density versus only two, four and one for pancreatic cancer, smoking behavior and bipolar disorder, respectively. Indeed, we have just scratched the surface and very little is known about the implication of nongenetic exposure and how it maps across the phenome.

Two common disorders with a multifactorial component, CVDs and T2D, have been extensively studied since the advent of GWAS. Visvikis-Siest stressed the tremendous findings observed since 2007 in the genetics underpinning CVDs and related quantitative traits. However, these results weighed against the fact that no genetic variant identified by GWAS has been successfully integrated in cardiovascular risk assessment scores to date and even if the insights of personalized medicine based on GWAS findings is challenging, it is thus also premature.

As stated by Philippe Froguel, “GWAS can‘t read the future”. Thus, their integration into personalized medicine should be through their ability to “illuminate the molecular understanding and physiology” of common disorders, such as T2D.

We are definitely witnessing a revolution of genetic epidemiology and yet the previous methodologies should not be abandoned but integrated in ‘network studies‘ including gene–gene–environment interactions assessment and functional validations. Indeed, raw GWAS results do not give clear functional implications relevant to the physiopathology of common disorders and the missing heritability issue remains. This dark matter was consistently discussed by the experts. Philippe Froguel addressed the interest of studying rare/uncommon variants which should have higher penetrance. This ‘rare variants (minor allele frequency < 0.01)/frequent disease‘ hypothesis has been experimented in the Meta-Analyses of Glucose and Insulin-related Traits Consortium (MAGIC) and a novel susceptibility gene for T2D and glucose has been highlighted. However, John Ioannidis still thinks that extending investigations on uncommon variants will lead to small effects anyway. In his opinion and in the opinion of Eleftheria Zeggini, this dark matter is also explained by imperfect phenotype definition and ascertainment; lack of power; epistatic (gene–gene) interactions; gene–environment interactions; differential effects in different populations; imperfect genome coverage for common variants; parent-of-origin specific effects; mutations; pedigree-specific effects and structural variants (rare or common alleles). Tremendous methodological improvement in GWAS should be performed to address all these limitations.

Meanwhile, alternative strategies should be undertaken. Sophie Visvikis-Siest proposed, for example, prospects on pediatric populations in which environmental effects are minimized in order to determine early genetic factors. This is a hint that the EA4373 Cardiovascular Genetics team is currently following on various quantitative traits: lipids, blood pressure and inflammatory markers.

Eleftheria Zeggini also described low-frequency and rare variant association analysis methodology using an alternative multivariate method to combine information across multiple variant sites. Several locus-specific approaches have been proposed and she discussed collapsing and allele-matching methods. These studies could be empowered by focusing on isolated populations, in which rare variants may have increased in frequency and linkage disequilibrium tends to be extended. This is the aim of the UK10k project gathering population-based cohorts with rich phenotype data (4000 whole genomes and 6000 whole exomes).

The following talks focused on post-GWAS on common disorders and Heather Cordell introduced the concept of maternal genotype and/or imprinting effect studying toxoplasmosis and possible genetic predisposition to this infection acquired in utero.

Allen Roses also highlighted phylogenetic analyses defining the evolutionary relations combining the use of rare and common polymorphisms (and not only SNPs) to determine the evolutionary genetics of complex disorders such as Alzheimer‘s disease. After tenfold sequencing of the APOE–TOMM40 linkage disequilibrium region, he identified a polymorphic poly-T variant, rs10524523, in the TOMM40 gene that provides greatly increased precision in the estimation of late-onset Alzheimer‘s disease risk and disease onset for APOE ε3 carriers.

It is clear that the need to upgrade ‘conventional‘ GWAS strategies will succeed through technological advances and, thus, Martin Seifert from Genomatix, Carsten Rosenow from Illumina and Sören Deininger from Bruker discussed their last-generation sequencing tools, the results obtained with the ‘1000 genome project‘ and the progress with mass spectrometry technologies. Martin Seifert considered these upgrades in the core of evidence-based medicine. “As a matter of fact, next-generation sequencing technology is rapidly changing the scientific landscape and we cannot lose sight of the patient‘s profit: the technology‘s increasing efficiency will soon allow epigenomes, genomes and transcriptomes to be sequenced on an individual basis, thus improving knowledge of a patient‘s predisposition to given diseases”. Indeed, treatments based on highly differentiated genetic diagnostics are becoming more feasible. For scientists and patients, “now is the right time to make this vision a reality”.

Session: from genome-wide scan to systems biology

▪ From genome-wide survey of human memory to drug identification & personalized medicine: opportunities & pitfalls

After the GWAS satellite meeting, the first session really in line with the conference spirit was introduced with a very unusual and stimulating presentation by Andreas Papassotiropoulos (University of Basel, Switzerland) on genetic liaison with human memory. He started by giving some examples of emotional memory. Using new tools based on genomic platforms, he demonstrated that genetics can identify molecular and signaling pathways related to human memory. He stressed the importance of precise phenotypes, which could then correlate with neuronal involvement.

For improving memory, he related some experiences with hydroxyfasudil but he preferred testing the simpler model of Caenorhabditis elegans and his knockout variant, which allows the study of the gene–environment interactions. Here again, the phenotype is of primordial importance. This introductory lecture between systems biology and future therapeutic approaches was exactly the theme of the conference.

▪ Optimizing cancer pharmacotherapeutics using mathematical modeling & a systems biology approach

The second speaker, Jean Clairambault (Institut National de Recherche en Informatique et Automatique [INRIA], Paris, France), explored an entirely different aspect of systems biology, trying to convince us of the interest of mathematical equations based on cellular models and pharmacokinetics (PK) data for correcting drug dosage in cancer patients in function of their biological rhythms. Such a systems approach is really very useful to avoid toxic effects on non-malignant cells. He and his coworkers were using ATP-binding-cassette (ABC) transporter-regulated efflux and UDP glucuronosyltransferase genetic variant information on cultured Caco-2 cells with irinotecan as the anticancerous drug. The data obtained could help to optimize the drug infusion time schedules in patients while minimizing toxicity.

▪ Molecular networks & the mechanism of action of drugs

With the third speech of Giulio Superti-Furga (Research Center For Medicine, Austrian Academy of Sciences, Vienna, Austria), we moved on to proteomics, an obligatory element of the systems biology approach. Each protein has multiple domains and functions and Giulio Superti-Furga proposed the ‘proteostasis‘ concept using the allosteric and cooperative data of many proteins. He was applying this concept to Gleevec^®-based intestinal tumor treatment, showing how myristilation is modifying different domains of the BCR–ABL protein complex. Here again, cellular models and pathway analysis are necessary to understand the data obtained by the sophisticated mass spectrometry method. Drug targets are generally not a simple protein but larger multiprotein complexes.

▪ Integration of genomic & epigenomic analysis in T2D

Finally, Christopher Bell (University College London [UCL] Cancer Institute, UCL, UK) introduced epigenetics in the systems biology approach with methylation data on CpG islands obtained in patients suffering from T2D in comparison with the same number of nondiabetic controls. He demonstrated the importance of controlling each technical step. He showed how the epigenetic data are complementary to the genomics data obtained by GWAS. The FTO obesity gene is at the fore of these interesting epigenetic responses, with a clear regulatory function in the brain. Mouse and zebrafish models were used to demonstrate the environmental effects linked to age. The involvement of epigenetics in evolutionary adaptation could then be better studied. However, epigenomic DNA methylation data must also be introduced in chronic disease routes.

Session: infectious diseases, genetics & new biomarkers

Studies on the effects of genetics on infectious disease were poorly known until the last few years, but evidence and characterization begin to be proved thanks to the development of new technologies, including information on the human genome. There is now epidemiological evidence supporting a role of genetics in susceptibility to disease infection. Mutation may be responsible for increasing the severity of a given pathology, but it can also protect from an infection. These studies need to be made in different large populations to distinguish the effect of genetics from that of the environment.

Such an approach should have been presented by Erwin Schurr (McGill Center for Study of Host Resistance, McGill University, Montréal, Canada), who was not able to be present in Santorini. However, the corresponding paper is due to be published in Clinical Chemistry and Laboratory Medicine[1].

The other speakers gave presentations that were more specific.

▪ Progress of HIV-1 infection molecular epidemiology research in China

The complex genetic structure of HIV-1 epidemics can be seen in Asia, where various HIV-1 strains are co-circulating within and among different risk populations. Hong Shang (The First Hospital of China Medical University, Shenyang, China) spoke mainly about the genetic and epidemiologic profile of HIV-1 in different affected Chinese provinces. Some mutations may be responsible for a drug resistance mutation inducing a rise in acute HIV-1 infection. The high genetic diversity of HIV-1 adds considerable complexity to the development of efficacious vaccines.

▪ Novel biomarkers for tuberculosis

Mycobacterium tuberculosis is a pathology affecting more than 2 billion people. There is still no real marker for the detection of this disease as presented by Gerd Michel (Foundation for Innovative New Diagnostics, Geneva, Switzerland). Many of the existing tests are not specific or not sensitive enough. The M. tuberculosis proteome contains 500 different proteins. A multiplex Luminex rapid immunoassay, a tuberculosis proteome microarray technology, has been developed and should allow to fight with efficacy tuberculosis associated or not associated with HIV infection.

▪ Multiplex PCR coupled to biochip array technology for detection of ten sexually transmitted infections

Martin Crockard (Randox Laboratories, Crumlin, UK) developed a biochip array for the simultaneous detection of ten sexually transmitted infections. The array uses dual priming oligonucleotide technology for the multiple amplification of highly specific pathogen sequences, before hybridization on a biochip array.

Session: inflammation as a common denominator of chronic & infectious diseases

▪ The inextricable link between atherosclerosis & inflammatory chronic disease

Claudia Monaco (Kennedy Institute of Rheumatology, Imperial College London, UK) explained some cellular and molecular similarities in atherosclerosis and rheumatoid arthritis and showed that atherosclerosis is the most important complication of rheumatoid arthritis.

She demonstrated that Toll-like receptors (TLRs) drive local inflammation while TLR2 and TLR4 are the best-known members of the TLR superfamily. Their roles have been studied by her scientific group as the ‘inextricable‘ link between chronic inflammatory disorders and CVDs. In addition, cytokine-dependent activation of the NF-κB pathway is a major mechanism involved in both diseases and could represent effective drug targets.

▪ Novel biomarkers for better diagnosis of sepsis

Wei Cui (Peking Union College Hospital, Beijing, China) presented clinical criteria for diagnosis of sepsis and compared some current diagnostic biomarkers including C-reactive protein, procalcitonin and lipopolysaccharide-binding protein for this serious inflammatory condition. She told us that high expression of CD64 has been found to be a better predictive marker for sepsis than procalcitonin and C-reactive protein. As the early diagnosis of sepsis is crucial for treatment and prevention of death, according to recent results, miRNA-150 could also be an early sepsis marker. miRNA-150 is significantly downregulated in sepsis patients compared with healthy controls.

▪ Inflammation & targeted immunotherapy

Jean-Yves Bonnefoy (Transgene, Strasbourg, France) told us that Transgene is developing a viral-based therapeutic vaccine for treatment of cancer and infectious diseases. He presented the TG4010 as an immunotherapeutic agent in cancer therapy, which gives a better survival of patients with normal levels of inflammatory proteins. He ultimately concluded that a combination of anti-inflammatory drugs and therapeutic vaccines may be more beneficial for cancer treatment.

▪ Bioactive lipid mediators & resolution of vascular inflammation

Aksam Merched (INSERM U828, Pessac, France and Baylor College of Medicine, Houston, TX, USA) in an interesting presentation developed the role of bioactive lipid mediators in control of inflammatory homeostasis. With three animal study models including normal, switched off and overexpressed lipoxygenase gene, he demonstrated that 12 of 15 lipoxygenases provide anti-inflammatory signals and protection during normal progression of atherogenesis mediated by downstream products. The local inflammatory response could be controlled by agonist actions of these mediators on macrophages and vascular endothelial cells. As the plasma cholesterol level increased during switching of mice from chow diet to Western diet in this study, he finally proposed a 3D approach including use of lipid-lowering drugs, lipoxygenases, proresolution mediators and polyunsaturated fatty acids for prevention of hyperlipidemia and its treatment.

▪ Genetics of inflammation markers & relationships with lipid metabolism

Ndeye Coumba Ndiaye (EA4373 ‘Génétique Cardiovasculaire‘, Nancy, France) discussed the role of inflammatory molecules, particularly haptoglobin, in CVD given its large genetic heritability. She also mentioned an ongoing GWAS in large pediatric populations, including the Suivi Temporaire Annuel Non Invasif de la Santé des Lorrains Assurés Sociaux (STANISLAS) cohort that has demonstrated so far that a unique new common genetic variant on the HP/HPR cluster explained an important proportion of the haptoglobin level variability.

▪ Potential role of heat shock proteins in CVD: evidence from in vitro & in vivo studies

Majid Ghayour-Mobarhan (Mashhad University of Medical Sciences, Iran) talked about heat shock proteins‘ (HSPs) discovery, their classification, functions and their role in atherosclerosis. He presented some of his epidemiological studies, clinical and animal experiences on HSPs. Anti-HSP60 and 65 antibodies are raised in obese and hyperlipidemic patients. HSP27 antibody titer falls beyond 12 h of postcoronary event.

When considering some animal studies, the role of HSPs in atherogenesis may not be straightforward, with different HSPs potentially having pro- or anti-atherogenic roles.

In addition to a potential role in the initiation of atherosclerosis, they may also be involved in the later stages of disease by inducing a proinflammatory autoimmune response and the recruitment of a HSP-specific inflammatory lymphocyte population.

He finally suggested that we should measure HSP expression in plasma, including serum antigen or antibody concentrations that may be useful as markers of disease susceptibility.

Session: CVDs & gene–environment interactions

▪ Low-density lipoprotein-cholesterol-raising gene variants & future lipid-lowering drug usage

Genetic and environmental factors are important determinants of lipid profiles (low-density lipoprotein-cholesterol [LDL-C], high-density lipoprotein-cholesterol [HDL-C] and triglyceride [TG]). Using a 50K-SNP chip in 5592 participants in the WHITEHALL II study, Steve Humphries (UCL Genetics Institute and British Heart Foundation Laboratories, London, UK) has identified 21 SNPs in PCSK9, SORT1, APOB, HMGCR, APOA5, CETP, LDLR and APOE genes to be associated with LDL-C, 12 SNPs in CETP, LIPC, LPL, APOA4 and BUD13 with HDL-C and 16 SNPs in GCKR, GIMP, TRIB1, RBP4, BAZ1B and LPL with TG. Common alleles in these genes explained 14.6% of LDL-C traits. In combination, these common LDL-C-modifying alleles have substantial effects on LDL-C levels that, combined with Framingham risk score, predicts the use of lipid-lowering (statin) medication.

▪ Interaction of dietary fatty acids with 5-lipoxygenase, ALOX5 activating protein with subclinical markers of coronary artery disease

Population studies report contradictory results in regard to 5 lipoxygenase (ALOX5), lipoxygenase-activating protein (FLAP) and coronary artery disease (CAD). On one hand, it was reported that both loci predispose individuals to CAD. On the other hand it was shown that ALOX5 and FLAP interact with ω3 and 6 fatty acids. Therefore, Michael Y Tsai (Department of Laboratory Medicine and Pathology, University of Minnesota, MN, USA) has investigated the interaction of SNPs in ALOX5 with ω3 and 6 fatty acids in regard to carotid intima–media wall thickness and coronary artery calcification among 2847 participants in the Multi Ethnic Study of Atherosclerosis (MESA). One SNP near the 3´-end of ALOX5 interacted with phospholipid fatty acids in order to modify the association with internal carotid intimal medial wall thickness and the extent of coronary artery calcification. In conclusion, Tsai‘s study confirms that ω3 and 6 fatty acid intake may potentially modify the genetic predisposition of ALOX5 genes to CAD.

▪ Use of phage display screening for biomarker discovery

Proteins secreted by atherosclerotic plaque into the bloodstream consist of new cardiovascular biomarkers. Danielle Hof (Institute for Clinical Chemistry, University Hospital Zurich, Switzerland) proposed to generate antibodies capable of specifically binding to plasmatic targets and to further identify them by mass spectrometry. Using a phage display approach, recombinant antibodies were generated against proteins in the secretomes against post-translational modifications that occur during the inflammatory processes, both taking place in the atherosclerotic plaque. The identification of these proteins is the first step towards the development of an immunoassay usable in routine clinics.

▪ Are triglycerides causally related to CVD?

Triglycerides are inherited complex traits. With LDL-C and HDL-C they are used clinically to evaluate the risk of future CVD. Using a 50K-SNP chip in the WHITEHALL II study, Philippa Talmud (Centre of Cardiovascular Genetics, UCL, UK) found that rs651821 in the APOA5 cluster showed the strongest association with TG and with increased risk of coronary heart disease (CHD). Subsequent analyses demonstrated that rs651821 is in complete linkage disequilibrium with two other SNPs; together the rare haplotype explains 46% of TG variance and leads to decreased APOA5 expression, therefore increasing very low density lipoprotein plasmatic concentrations via reducing its catabolism. In contrast to rs651821 in the APOA5 cluster, TG had no effect on CHD. In conclusion, APOA5 cluster variants explain almost half of TG variance. If the relationship between TG and CHD is further clarified in other populations, APOA5 might be used as a genetic instrument in Mendelian randomization to confirm this relationship.

▪ Plasma plant sterols: a biomarker to predict lipid treatment efficacy?

Michel Krempf (University of Nantes, France) started his presentation by showing that intestinal absorption of cholesterol is estimated for years by using sitisterol and campesterol, plasma plant sterol concentrations, as markers. The status of cholesterol absorption is stable for each individual and poor or high absorbers could be individualized. Sitisterol and campesterol are not synthesized by humans and are poorly absorbed. It is believed that the plasma concentrations are an index of cholesterol absorption and could be a useful tool to identify cholesterol absorption status. Many studies have used this approach and have demonstrated a balance between absorption and synthesis estimated with lathosterol, a precursor of endogenous cholesterol synthesis: lower absorber being higher synthesizer and higher absorber being lower synthesizer. This paradigm was used to explain the large differences observed with cholesterol absorption treatment, such as ezetimide. In addition, this hypothesis is supported by numerous small clinical studies and recent analysis of larger trials has invalidated it. The origin of this discrepancy is still unclear but might be related to the use of sitisterol and campesterol to estimate cholesterol absorption.

Session: nutrigenomic epigenetics

▪ Nutrigenomics: about mechanism, efficacy, disposition & programming

Martin Kussmann (Nestlé Research Center, Lausanne, Switzerland) introduced nutrigenomics, a domain that was recently catalyzed by the revelation of the human genome, thereby showing the first extra large extension to human bacterial and plant genomes. Nutrigenomics is influencing immune, digestive, mental, skeletal, muscular and metabolic diseases. Cohorts are very important in this research domain with nutritional intervention. The mechanisms are often linked to epigenetic regulation and are based on genetic, transcriptomic, proteomic and metabolic approaches. Animal models are also interesting tools for studying the intestinal and gut flora.

▪ Phenotyping approaches & ‘omics‘ applications in nutrition research

Hannelore Daniel (Center for Diet and Disease, University of Munich, Germany) presented an exciting proteomic profiling approach in nutritional human clinical trials, often with healthy volunteers. New proteomic phenotypes could be measured on plasma, saliva, urine and blood cells, which are the only readily available material, but it is also necessary to apply these profiling techniques to dynamic studies, requiring many samples. Mass spectrometry and nuclear magnetic resonance-based profiling are interesting technologies used for intervention studies with soy isoflavone and flaxseed.

The most interesting part of Hannelore Daniel‘s presentation was the human metabolomic study performed in Munich between more than ten institutions on 15 healthy volunteers who stayed 4 days in a study center with two blood samples taken per day. After a fasting period they received different standard liquid diet and were submitted to different tests and stress. She presented results on plasma amino acid charges during fasting and after glucose administration, resulting in opposite variations. The relationship with metabolic pathways are then absolutely necessary for the interpretation of the data and the variation observed. It is then possible to ‘titrate‘ the capacity of adaptation in time and space and to identify deviation from normal.

▪ Long-term influence of diet & environment on gene expression

The Barker hypothesis states that adverse influences early in development and particularly during intrauterine life can result in permanent changes in physiology and metabolism, which result in increased disease risk in adulthood. In this context, obesity and insulin resistance are major public health problems in the 21st century. A better understanding of the factors involved in their development is crucial, especially if one third of obese children will remain so into adulthood. Using an ω3 fatty acid supplementation for murine models, Edgar Delvin (CHU Sainte-Justine, University of Montreal, Canada) has shown that its effect is exercised on first mouse generation and even persists until the second one. Therefore, based on clinical and experimental evidences, Barker‘s hypothesis could be supported. Furthermore, their experimental results have indicated that epigenetic DNA and perhaps histone modifications are involved.

▪ Nutrigenomic mechanisms of the deficiency in methyl donors: experimental-based evidence & population studies

Neurologic and aging disorders are linked to homocysteine and methyl donor deficiency, but their mechanisms remain to be clarified. By using cellular models in a first step, Jean-Louis Guéant (INSERM U724, Université Henri Poincaré and University Hospital of Nancy, France) has demonstrated that methyl donor deficiency in NIE-15 murine neuroblastoma cells reduced proliferation and increased cell differentiation via upregulation of protein phosphatase 2A, pro-nerve growth factor and TNF-α converting enzyme (TACE). In the same direction, folate deficiency in H19-7 rat neuronal progenitor cells affected their proliferation, differentiation and synaptic plasticity through altered cytoskeleton proteins. In a second step, in vitro experiments on rat models were also used to assess the impact of methyl donor deficiency during gestation. The pups displayed increased homocysteine concentrations in the hippocampus and cerebellum, and reduced thickness of the CA1 pyramidal layer that was restored to normal in older rats when applying a normal diet. Homocystein accumulation decreases cystationine-β-synthase in Purkinje cells. Combined with defects in locomotion coordination and in proteins responsible for synaptic migration, vitamin B deprivation in early life may produce selective and persisting brain defects consistent with neurodegeneration. These results could be applied in the Outcome in Ankylosing Spondylitis International Study (OASI) cohort, composed of eldery subjects.

▪ Nutrigenetics in the GWAS era

Nabila Bouatia-Naji (CNRS and Pasteur Institute, Lille, France) presented a large overview on GWAS applications to nutrigenomic studies.

Session: nutrigenomics & gene interaction studies

▪ Self-reported zinc & zinc intake–genetic loci interactions of health status & age-related diseases

Zinc (Zn) is an essential trace mineral for human health; its deficiency can lead to dermatitis and impaired immune function. Infants are particularly vulnerable because of the large amount of Zn required for growth during this period. By defining a Zn level score (significantly correlated to daily Zn intake), George Dedoussis (Harokopio University, Athens, Greece) has evaluated the impact of the -174G/C polymorphism and Zn diet score on plasmatic IL-6 levels. He found this SNP interacted with Zn diet score, as GG genotype was associated with higher plasma IL-6 levels compared with GC/CC genotypes, with increasing Zn diet score. In addition, a significant interaction was observed between another SNP and increased glucose levels; this interaction diminishes as total Zn intake increases (per 1 mg/day). In conclusion, George Dedousis revealed that Zn intake interacts with gene variants to influence IL-6, associated with health status and diabetes

▪ Validation of the use of peripheral blood mononuclear cells as surrogate markers for skeletal muscle tissue in nutrigenomic studies

Peripheral blood monocyte cells (PBMCs) are considered as a substitute of traditional tissue specimens that are not easily accessible. Iwona Rudkowska (Institute of Nutraceuticals and Functional Food, Quebec, Canada) has investigated the possiblity of using PBMCs for gene-expression analysis as an alternative to skeletal muscle tissue (SMT). By carrying out a transcriptomic analysis comparison of PBMCs versus SMT after 8 weeks supplementation with n-3 polyunsaturated fatty acid in 16 insulin-resistant subjects, Iwona Rudkowska found that 88% of transcripts (32341 transcripts) were common in PBMCs and SMT. In addition, a strong correlation (r = 0.84; p < 0.001) was present between transcript expression levels of PBMCs and SMT. In conclusion, PBMCs express the majority of SMT transcripts with comparable expression levels between both tissues.

▪ Genetic variants near IRS1 & their association with T2D, hyperinsulinemia & insulin resistance in UK population-based cohorts & T2D patients

Genome-wide association studies identified a novel SNP, rs2943641, near IRS1 loci that is highly associated with T2D, this association not being replicated in other European populations. Therefore, Nikolaos Yiannakouris (Harokopio University, Athens, Greece) conducted his study to validate the rs2943641C>T association with T2D risk and diabetes-related quantitative traits using data from UK population-based cohorts and T2D patients, and to explore potential associations with the risk of T2D of SNPs within and flanking the IRS1 gene. The rs2943641C was associated with a 14% increased relative risk of T2D (p = 0.006), with CC subjects having an odds ratio of 1.21 compared with T allele carriers. In addition, the C allele of rs2943641 near IRS1 was associated with higher fasting, glucose-stimulated hyperinsulinemia and impaired insulin sensitivity. Their data also suggest that rs2943641 and an IRS1 putative promoter variant (rs6725556) may independently influence T2D risk, although further studies with larger cohorts are needed to confirm the etiological SNPs and to analyze their interactions in different populations.

▪ Adaptive genetic variation & risk of metabolic syndrome & CVDs

Modern human populations have experienced diverse adaptation to environmental changes in the past 100,000 years and this adaptation has been imprinted on the human genome as allele frequency changes in certain genetic variants. By contrast, dietary habits and lifestyles have changed faster than our genomes have adapted to the environment. Chao-Qiang Lai (Nutrition and Genomics Laboratory, JM-USDA Human Nutrition Research Center on Aging at Tufts University, MA, USA) explained that some genetic variants interact with dietary intake and lifestyles to influence the risk of metabolic syndrome. In addition, he explained that the recent rise of metabolic syndrome is partially due to rapid environmental changes that are poorly fitting for genomes adapted to traditional environments.

Session: fundamental pharmacogenomics

▪ From drug metabolism to pharmacogenomics & personalized healthcare

Urs A Meyer (Biozentrum, University of Basel, Switzerland), in an interesting presentation, defined pharmacogenetics as “linking of variation of human genes to clinical responses to drugs”, gave a history and described the effects of pharmacogenetics on the safety and efficacy of drug therapy, highlighting examples of genetic variations that affect the response to treatment. Preprescription genotyping and the use of pharmacogenomic biomarkers, either as required or recommended, help physicians to predict the drug response, prevent adverse drug reactions or determine drug dosing based on the patient‘s drug-metabolizing status. While genotyping for HLA-B*5701 is required to prevent hypersensitivity in HIV patients taking abacavir and is recommended for SLCO1B1 to prevent statin-induced myopathy in patients with hypercholesterolemia, routine genotyping for CYP2D6 in women with breast cancer who are on tamoxifen is not accepted and is controversial. Furthermore, in spite of existing data about the role of CYP2C9 and VKORC1 in the metabolism of warfarin and CYP2C19 in the metabolism of clopidogrel, the clinical utility of these biomarkers of drug response remains controversial. In addition, some side effects such as carbamazepine hypersensitivity are only observed in Chinese populations, which demonstrates the importance of ethnic differences.

▪ Polymorphisms affecting gene transcription in pharmacogenetic candidate genes: detection through allelic expression imbalance

In an exciting lecture, Wolfgang Sadee (The Ohio State University, OH, USA) first showed the general contribution of different types of SNPs – including coding SNPs that alter the encoded amino acids sequence, regulatory SNPs (rSNPs) in untranslated regulatory region of the gene and structural RNA SNPs (srSNPs), which are polymorphisms in the RNA transcript – to changes in protein activity and gene transcription. Then, he elucidated in detail the pharmacogenetic effects of rSNPs and srSNPs on changes in gene expression of drug-metabolizing enzymes, with a focus on rSNPs. rSNPs, which are likely to be more prevalent than coding SNPs, affect gene transcription and are confirmed by allelic expression imbalance, defined as transcription of two allelic variations of the same gene in different amounts due to a polymorphism in a regulatory part of the gene. This is also important for gene–gene interaction (epistasis). These rSNPs considerably affect efficacy and adverse drug effects in drug therapy. In a recent study that has been published in 2010, Wang et al. revealed that rSNPs in CYP3A4 are as frequent as 4–8% and are responsible for two- to five-fold increases in gene expression, which subsequently leads to a decrease in statin dose requirement. rSNPs and srSNPs are now considered as drivers of evolution [2].

▪ Transcriptional regulation & pharmacogenomics

The presentation of George P Patrinos (School of Health Science, University of Patras, Greece) had two relatively separate parts. In the first section, he reported the preliminary results of an ongoing study regarding the role of copy number variants, as a pharmacogenetic biomarker, in response to lithium. The results revealed at least two different chromosomal positions where losses of genomic fragments have been indicative for responsiveness to lithium treatment. In the second part of his speech, based on the results of another study, George Patrinos talked about the correlation between KLF1 (a hematopoietic-specific transcription factor that induces high-level expression of adult β-globin) gene mutation and high levels of fetal hemoglobin in erythroid cells of β-thalassemic patients. KLF1 also indirectly suppresses fetal hemoglobin genes by activating BCL11A, a known suppressor of fetal hemoglobin gene transcription. This genetic variation accounts for different drug therapy in patients with the same β-thalassemia-causing mutation and occasionally an identical chromosomal background.

▪ Relevance of drug transporters for pharmacotherapy

The imperative role played by different transport systems in drug PK and pharmacodynamics (PD) was the subject which Peter Meier-Abt (University of Basel, Switzerland) described in his sophisticated presentation. He showed the importance of drug transporters by reporting acute rejection in two heart transplant patients due to a St John‘s wort-induced intestinal P-glycoprotein (P-gp) drug transporter upregulation which led to a decrease in cyclosporin plasma concentration. The major transport systems involved in the uptake of organic compounds belong to the organic anion transporters and organic cation transporters and the organic anion-transporting polypeptides of the SLCO superfamily of polyspecific drug transporters. The kinetics of this system depend, to a large extent, on the pH of the cell and they are highly substrate specific. On the other hand, ABC (multiple drug resistance [MDR]) proteins are the other major transporter system which transports various molecules across extra- and intra-cellular membranes. Differences in drug transporter expression and activity contribute to variability of the PK and PD of drugs and statins are among the most extensively studied medications. The SLCO1B1*1b gene polymorphism was associated with increased transport activity and decreased systemic bioavailability of statins and the SLCO1B1*5 variant has been considered a major factor in limiting the uptake of statins and increasing statin exposure and, consequently, increasing the risk of myopathy. On the other hand, in a recently published study, MDR3 (ABCB4) and MRP2 (ABCC2) polymorphisms have been associated with drug-induced liver injury and cholestasis in patients who were on risperidone and amoxicillin [3].

▪ SNPs in genes encoding G proteins in pharmacogenetics

G proteins communicate signals from many neurotransmitters and drugs outside the cells, which cause changes inside the cells; therefore, SNPs in genes encoding G proteins are considered, in pharmacogenetic studies, as a “bottleneck in signal transduction pathways”, as described impressively by Winfried Siffert (Institute of Pharmacogenetics, University Hospital Essen, Germany). G proteins consist of α, β and γ subunits. Due to significant effects on the activity level of G proteins, polymorphisms in subunits of this molecule are a matter of interest pharmacogenetically. The GNβ3 C825T polymorphism in the gene of the G-protein β3-subunit is associated with the occurrence of a splice variant termed Gβ3s which is a biologically active Gβ-protein that may mediate the enhanced signal transduction observed in cells with the 825T allele. The vasoconstriction response to endothelin-1, angiotensin II and α2-agonists was significantly enhanced in patients with the C825T polymorphism. T allele carriers of the C825T polymorphism also showed an increased blood pressure reduction in response to clonidine, a central α2-adrenergic receptor stimulant, and showed a better response to selective serotonin reuptake inhibitor antidepressant treatment. On the other hand, the functional G-1211A SNP in the GNAS promoter causes the impairment of transcription factors‘ binding sites, significantly changes Gα protein expression and lipolysis levels in human fatty tissues; therefore, A allele carriers benefit from sibutramine antiobesity therapy, while G allele carriers can lose weight without the drug and are at risk for exaggerated heart rate when using the drug. Overall, functional SNPs in genes encoding G proteins, receptors or downstream effectors are excellent candidates to predict drug response.

Session: pharmacogenomics translational aspects

▪ Translation of technologies into clinical practice: the example of drug metabolism enzymes & transporters

Richard Dean Hockett (Affymetrix, CA, USA) briefly and comprehensively reviewed the genes involved in the drug metabolism pathway. He believed that pharmacogenetic research has been restricted to the search for polymorphisms in one or several genes thought to be involved in a drug‘s disposition. However, now we know that drugs pass through complex metabolic pathways rather than experiencing a single enzymatic modification, it has become clear that pharmacogenetic research must expand beyond the candidate gene approach. Richard Hockett then introduced the well-validated drug metabolism enzymes and transporters (DMET) Plus chip developed by Affymetrix Inc., which can genotype large amounts of variant in numerous genes involved in drug disposition. Finally, he reported the results of their recently published, large-sample study in which the authors comprehensively assessed the allele frequencies of 165 variants comprising 27 DMET genes from 2188 participants across three major ethnic populations (Caucasians, Africans and East Asians) [4].

▪ Systems characterization of drug combination therapies in transgenic mice: focus on hypercholesterolemia

In a nice presentation, Lars Verschuren (TNO Quality of Life, Zeist, The Netherlands), using the results of the study in which he compared the genes affected by combination therapy with statin and ezetimibe versus single drug treatment in transgenic mice, described a ‘systems biology approach‘. In this approach, genes affected by individual drugs are characterized. Considering all data for each single drug therapy may help us to recognize the genes involved in combination therapy and consequently their clinical effects.

▪ Gene-expression meta-analysis of response to infliximab in rheumatoid arthritis

In her short but informative lecture, Alessandra CL Cervino (TcLand expression, Nantes, France) presented a meta-analysis including seven studies (208 patients) on the pharmacogenetic response to infliximab, an anti-TNF agent used in the treatment of rheumatoid arthritis. Interestingly, each study reported completely different sorts of genes as being responsible for the variability observed in response to treatment. Alessandra Cervino explained how she and her colleagues used a six-step process (identification of public data, clinical homogenization, technical homogenization, quality control, probe to gene mapping and algorithm choice) to prepare a standardized meta-analysis. Finally, she presented a list of involved genes as the result of the meta-analysis.

▪ Auto-antibody signatures for disease diagnosis in cancer & autoimmune disease

During their lifespan every person produces antibodies directed against human antigens (auto-antibodies; AAs), which reflect the immune response to a continuous remodeling of cells or tissues caused by a chronic disease process. This fact is the fundamental principle of the technology that UNIarray^® (Protagen AG, Germany) – systematic identification of AA signatures – has been made based on and this was introduced in a nice presentation by Peter Schulz-Knappe (Protagen AG, Dortmund, Germany). UNIarray^® provides a new in vitro diagnostic and companion diagnostic for patient stratification and therapy selection using AAs in the blood. There are numerous advantages of using AAs in serum/plasma compared with genomic based markers in diagnostic applications, including:

▪	AAs are stable in serum and plasma;
▪	AAs are present in serum or plasma in high concentrations;
▪	No enrichment or elaborated sample preparation is required;
▪	AAs are not subject to circadian rhythms or other short-term changes in physiological states.

The companion diagnostic capability of AA detection has particularly been shown in multiple sclerosis and prostate cancer.

▪ Pharmacogenomics translational aspects: thienopyridine genetics

Clopidogrel, the most well-known member of the thienopyridine family and the second top-selling drug in the world, is currently the antiplatelet treatment of choice for preventing secondary cardiac events across the spectrum of acute coronary syndrome and in patients undergoing percutaneous coronary intervention. However, up to 30% of patients do not display adequate antiplatelet response and patients with coronary disease with lesser degrees of platelet inhibition in response to clopidogrel appear to be at increased risk for cardiovascular events. Such variability may arise from genetic alterations in genes involved in the PK and PD of clopidogrel. Using clopidogrel as an example, Sandra Close Kirkwood (freelance consultant, USA) explained, in an exciting lecture, how genetics can be translated into clinical practice. Clopidogrel is an inactive prodrug requiring two-step oxidation by hepatic cytochrome P450 (CYP) to generate its active metabolite, which irreversibly targets the adenosine diphosphate P2Y12 receptor that consequently blocks activation of the glycoprotein IIb/IIIa pathway and prevents platelet aggregation. SNPs in genes involved in the PK/PD of clopidogrel can potentially change the response to this drug. Among the hepatic enzymes involved in the metabolization process of clopidogrel (CYP2C19, 3A4/5, 1A2, 2B6 and 2C9), SNPs of CYP2C19 have been associated with significant changes in response to drug treatment and metabolizing status. Several studies showed that CYP2C19*2/3 and CYP2C19*17 polymorphisms are associated with decreased and increased activity of CYP2C19, respectively. Accordingly, a US FDA Box Warning released in March 2010 warned about the diminished effectiveness of clopidogrel in poor metabolizers (CYP2C19*2/3 SNP carriers).

▪ mRNA flow fish: a novel approach in HIV tropism detection & a valuable tool for the decision over CCR5 blockers

As the scientific director of FlowCytoGen Laboratories Ltd, as well as product manager of Zafiropoulos Co., George Lallas (Zafiropoulos Company, Athens, Greece), in a stimulating and exciting talk, presented HIV Monitoring Assay (HIV Reservoirs Assay ViroTect^®VR and HIV Tropism Assay ViroTect^®Tropism), a fast, quantitative, accurate and cost-effective theragnostic method for personalizing and monitoring virus therapy. Using a combination of these two arrays provides the possibility of determining the efficacy of HIV treatment in individual patients. Following therapy, patients can be completely suppressed or can have persistent replication in T cells or monocytes; therefore, therapy can be tailored to a patient‘s own reservoirs (T cells or monocytes) to achieve better immunologic recovery and prolong the interval to first treatment failure. On the other hand, determining the tropism of HIV (i.e., the cell type that HIV infects) in individual patients may make them candidates for CCR5 inhibitors (CCR5 is used only with a specific strain of HIV-1).

Pharmacogenomics & personalized medicine in clinical practice: introduction to the round table discussions

▪ Changing clinical practice based on evidence generated in the real world: a series of comparative effectiveness study examples

Pharmacogenomics in clinical practice: example from the real world

Bryan Dechairo (Medco Health Solutions Inc., MD, USA) introduced this session by discussing the problems arising when considering the events from scientific innovations to clinical applications. Going by classical examples (warfarin, tamoxifen) he showed that translation from bench to bedside currently takes approximately 14 years. If the patient‘s and private payer‘s acceptance of personalized medicine is greater than that of physicians and government payers, the pharmacogenomics implementation could be quicker. Understanding impact on patient outcomes and cost of test implementation are essential for reimbursement decisions. At least, physician education paired with patient empowerment will be key for future success, which can be summarized in decreasing the time from discovery to clinical practice and in changing clinical practice through evidence of clinical utility.

Impact of ABC transporters in epilepsy & stroke

Alberto Lazarowski (Buenos Aires University, Argentina) showed the importance of ABC transporters (particularly P-gp, the multidrug resistance gene product, but also multidrug-resistance-associated proteins), which are known to play a role in antiepileptic drug extrusion, but are not by themselves sufficient to fully explain the phenomenon of drug-resistant epilepsy. It has been shown that ABC transporters are overexpressed in refractory epilepsy. The study of the MDR-1 gene‘s SNPs (15) could explain the differences in efficacy of antiepileptic drugs.

Application of pharmacogenetics in psychiatry: clinical & technical aspects

Adrian Llerena (CiCAB-Clinical Research Centre, Extremadura University Hospital, Badajoz, Spain) highlighted the necessity of developing clinical psychopharmacology (by independent clinical trials and pharmacovigilance) as knowledge of genetics increases. If personalized medicine (prediction for a subject) is not currently feasible in the field of psychiatry, stratified medicine (prediction for a group) is conceivable. As genetics and physiology are tightly connected, the use of pharmacogenetics in clinical psychiatry must take into account PK of drugs but also therapeutic response, metabolism modulations and adverse reactions.

Cardiopharmacogenomics: application of pharmacogenomics in cardiology

Jerry Yeo (University of Chicago, IL, USA) noticed that, currently, in cardiology, two pharmacogenomic tests were relabeled by the FDA (CYP2C9 and VKORC1 for warfarin and CYP2C19 for clopidogrel. However the pharmacogenomics testing for predicting warfarin responsiveness does not seem to improve health outcomes and, in the Center for Medicaid and Medicare Services (CMS), it is covered only for candidates for anticoagulation therapy and in the context of a prospective, randomized controlled clinical study. Regarding clopidogrel, interindividual variability in response to this drug is multifactorial (lack of compliance, obesity, nature of coronary event, genetic factors in absorption and metabolism, and drug–drug interactions), but the role of CYP2C19*2 and CYP2C19*3 alleles in reduced metabolism of clopidogrel is well recognized even if other alleles could be implicated at a lower level. Another example is given by KIF6, which encodes a kinesin (family of dimeric motor proteins involved in the intracellular transport of organelles, protein complexes and mRNAs). The KIF6 Trp719Arg carrier status (Trp719Arg replaces a nonpolar residue with a basic residue near the coiled-coil structure and affects cargo binding or regulation of the motor domain) affects response to statins and CHD event reduction.

Thus, pharmacogenetic markers will increasingly be used for personalizing drug therapy in cardiology. Their clinical utility is recognized, since they can be used for dose adjustments (warfarin), selection of alternative drugs or dose escalation for loss-of-function carriers (clopidogrel) or selection of patients for aggressive therapy (statins).

Clinical applications: pharmacogenetics of immunosuppressants

Ron Van Schaik (Pharmacogenetics Core Laboratory, Erasmus MC Rotterdam, The Netherlands) showed the interest of pharmacogenetics of immunosuppressants for clinical applications, with examples of tacrolimus, cyclosporin (two calcineurin inhibitors) and prodrug mycophenolate mofetil. Tacrolimus is metabolized by CYP3A4 and CYP3A5 and displays nephrotoxicity. It has been shown that CYP3A5 genotyping can be useful for guiding tacrolimus dosing. Mycophenolate mofetil has also to be dosed facing the possibility of acute rejection and UGT1A9 -275T>A carriers have a lower mycophenolic acid bioavailability and a higher risk of acute rejection. Finally, ABCB1 genotype of the donor seems to be correlated with cyclosporin nephrotoxicity.

▪ Development & use of theranostics biomarkers: needs of guidelines

Genotype-guided dosing of coumarin derivatives: the European Pharmacogenetics of Anticoagulant Therapy Consortium trial

Vangelis G Manopoulos (Democritus University of Thrace, Alexandroupolis and Academic General Hospital of Alexandroupolis, Greece) emphasized the necessity of monitoring the anticoagulant effect of coumarins, due to inter-/intra-patient variability in the response: clinical habits consist of testing the international normalized ratio (INR). Coumarin responses are affected by demographic, clinical and genetic factors, among which are the polymorphisms in two genes: CYP2C9 and VKORC1 (the coumarin target enzyme in the vitamin K cycle, vitamin K epoxide reductase complex subunit 1). CYP2C9*2 and/or *3 allele carriers have increased risk of major bleedings and thus require lower daily doses of coumarins. For VKORC1, 28 SNPs have so far been identified. Patients carrying VKORC1 haplotype A (-1639A/1373T) are more sensitive to coumarins whereas patients carrying the VKORC1 haplotype B (-1639G/1373C) show coumarin resistance. Thus, polymorphisms in the CYP2C9 and VKORC1 genes together account for approximately 40–50% of the variability in coumarin maintenance dose requirements. However, as large prospective randomized clinical trials are missing, the European Pharmacogenetics of Anticoagulant Therapy (EU-PACT) wants to assess, in a single-blinded and randomized controlled trial with a follow-up period of 3 months, the safety and clinical utility of genotype-guided dosing in daily practice for the three main coumarin derivatives used in Europe. The proposed study will be undertaken in seven different European countries at 13 different centers. The aims of the study are first to improve effectiveness of anticoagulation therapy and second to improve the safety of the treatment, the quality of life of the patient and to determine whether genotype-guided dosing is cost effective.

Thus, the EU-PACT trial will evaluate a pharmacogenetic approach to anticoagulation therapy with coumarin derivatives using pharmacogenetic-based dosing algorithms, which include relevant polymorphisms of CYP2C9 and VKORC1 genes.

Development & use of theranostic biomarkers: impact on the industry business model; needs of guidelines & market access issues, the European case

Alain Huriez (European Personalized Medicine Diagnostics [EPEMED] and TcLand expression, Nantes, France) pointed out the necessity of a rigorous strategy for developing personalized medicine without forgetting an important constraint consisting of cost minimization and effectiveness.

Companion diagnostics generate great hopes in the pharmaceutical industry for optimizing each step of the drug value chain, particularly when companion diagnostics will be proposed to current marketed blockbusters. However, a small number of companion diagnostics have reached the market. Europe is far behind the USA when considering practical applications of molecular diagnostics, companion diagnostics or personalized medicine and huge needs remain in terms of education of the various stakeholders in Europe for a better understanding of the recent advances and the needs to bring such innovations to the EU market.

Serving this purpose, EPEMED is a not-for-profit organization bringing together forces in personalized medicine, confronting industry, regulators, payers and insurers as well as governments. It will function as a platform for the harmonization of personalized medicine development and implementation across Europe, focusing on the crucial role of diagnostics.

Conclusion

This conference brought together scientists with backgrounds ranging from human genomics to epidemiology to improve our understanding of human genetic variability and its impact on the prediction, prevention, diagnosis and therapy of multifactorial disorders. In the spirit of this conference, new concepts should help the implementation of molecular biology in clinical practice. Through the different presentations, systems biology, animal and cellular models, metabolomics, proteomics and proteostasis concepts were developed. Blood cells, particularly white cells, are in many cases interesting tools, as are cohorts and biobanks, which are more and more important for studying genetic variations in large populations. For understanding health state evolution, we should remember that dietary habits and lifestyles have changed faster than human genome adaptation to our different environments. This is also true for drug responses. Drugs use pre-existing genes and enzymes for their biotransformation, genes which were devoted to xenobiotics from the surrounding trees or endobiotics from our normal metabolism (such as steroids). The gradual introduction of genetic and genomic biomarkers in personalized therapy over recent years was discussed in a round table [5], which led to the creation of the European Society of Pharmacogenomics and Theranostics (ESPT) [101].

Financial & competing interests disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.