Grant E. O’Keefe and J. Perren Cobb
INTRODUCTION
Variability is a hallmark of clinical medicine. Why do patients with seemingly similar injuries or severity of acute illness, receiving comparable and appropriate treatment, often follow different trajectories? For example, one patient recovers uneventfully from massive transfusion for hemorrhagic shock while another follows a prolonged course complicated by nosocomial pneumonia and organ failure. Consider also, the seemingly more straightforward task of preventing or treating deep venous thromboembolic disease. Despite our understanding of the biology of coagulation and pharmacotherapeutic strategies, therapy sometimes fails with fatal consequences. Numerous clinical (environmental) and genetic factors contribute to this variability. The completion of the Human Genome Project provided the foundation on which to build knowledge of genetic variation in both humans and animal models.1 This in turn has been used to establish genotype–phenotype associations for common polygenic diseases, such as diabetes mellitus (metabolic syndrome), cancer, and hypertension. Using this genetic variability to understand disease biology and to better direct therapy (such as drug selection and dosing) are the goals of the field of “Genomic Medicine.”2 At the cellular level, differences in DNA sequence can alter RNA transcription and protein translation, which alters clinical phenotypes. For instance, drug absorption, metabolism, and excretion are all affected by genetic variation (pharmacogenomics). Recent progress in these fields and the application of this knowledge to critically injured patients is the focus of this chapter.
STRUCTURE OF THE GENOME
Since the discovery and publication of the molecular structure of nucleic acids by Watson and Crick in 1954, the genetic basis for many conditions has been determined.3,4 Misconceptions remain, however, regarding the role of genetics and genomics in clinical medicine. Despite the commonly held notion that genetics had little influence on clinical medicine in the past, genetics has, in fact, played an important role in understanding disease for a minority of conditions and patients. As a result of the advances described above, we have entered a period of tremendous growth in our knowledge of genomics that will influence care for all patients.5 In order for clinicians to understand and participate in these advances, we must become “literate” in the language of genetics and genomic medicine. Box 53-1 includes some important definitions of genetic concepts for clinicians, some of which will be more completely developed below.
Box 53-1: Definitions
Allele: One of two or more versions of a genetic sequence at a particular location in the genome.
Base pair (bp): Two nitrogenous bases paired together in double-stranded DNA by weak bonds; specific pairing of these bases (adenine with thymine and guanine with cytosine) facilitates accurate DNA replication; when quantified (e.g., 8 bp), bp refers to the physical length of a sequence of nucleotides.
Complex condition: A condition caused by the interaction of multiple genes and environmental factors. Examples of complex conditions, which are also called multifactorial diseases, are sepsis, cancer, and heart disease.
DNA: Deoxyribonucleic acid, the molecules inside cells that carry genetic information and pass it from one generation to the next.
Exon: The portion of a gene that encodes amino acids.
Gene: The fundamental physical and functional unit of heredity. A gene is an ordered sequence of nucleotides located in a particular position on a particular chromosome that encodes a specific functional product (i.e., a protein or an RNA molecule).
Gene chip: A solid substrate, usually silicon, onto which a microscopic matrix of nucleotides is attached. Gene chips, which can take a wide variety of forms, are frequently used to measure variations in the amount or sequence of nucleic acids in a sample.
Genome: The entire set of genetic instructions found in a cell. In humans, the genome consists of 23 pairs of chromosomes, found in the nucleus, as well as a small chromosome found in the cells’ mitochondria.
Genome-wide association study (GWAS): An approach used in genetics research to look for associations between many (typically hundreds of thousands) specific genetic variations (most commonly SNPs) and particular diseases.
Genotype: A person’s complete collection of genes. The term can also refer to the two alleles inherited for a particular gene.
Haplotype: A set of DNA variations, or polymorphisms, that tend to be inherited together. A haplotype can refer to a combination of alleles or to a set of SNP found on the same chromosome.
HapMap: The nickname of the International HapMap (short for “haplotype map”) Project, an international venture that seeks to map variations in human DNA sequences to facilitate the discovery of genetic variants associated with health. The HapMap describes common patterns of genetic variation among people.
Human Genome Project: An international project completed in 2003 that mapped and sequenced the entire human genome.
Microarray: A technology used to study many genes at once. Thousands of gene sequences are placed in known locations on a glass slide. A sample containing DNA or RNA is deposited on the slide, now referred to as a gene chip. The binding of complementary base pairs from the sample and the gene sequences on the chip can be measured with the use of fluorescence to detect the presence and determine the amount of specific sequences in the sample.
Mutation: A change in a DNA sequence. Germ-line mutations occur in the eggs and sperm and can be passed on to offspring, whereas somatic mutations occur in body cells and are not passed on.
Pharmacogenomics: The branch of pharmacology that deals with the influence of genetic variation on drug response in patients by correlating gene expression or SNPs with a drug’s efficacy or toxicity. By doing so, pharmacogenomics aims to develop rational means to optimize drug therapy, with respect to the patients’ genotype, to ensure maximum efficacy with minimal adverse effects.
Phenotype: The observable traits of an individual person, such as height, eye color, and blood type. Some traits are largely determined by genotype, whereas others are largely determined by environmental factors.
Point mutation: An alteration in DNA sequence caused by a single-nucleotide base change, insertion, or deletion.
Ribonucleic acid (RNA): The molecule synthesized from the DNA template; contains the sugar ribose instead of deoxyribose, which is present in DNA; three types of RNA exist including messenger RNA (mRNA), transfer RNA (tRNA), and ribosomal RNA (rRNA).
Single-nucleotide polymorphism (SNP): A single-nucleotide variation in a genetic sequence; a common form of variation in the human genome.
Systems biology: Research that takes a holistic rather than reductionist approach to understanding organism functions.
Transcription: The synthesis of an RNA copy from a sequence of DNA (a gene); a first step in gene expression.
Translation: During protein synthesis, the process through which the sequence of bases in a molecule of messenger RNA is read in order to create a sequence of amino acids.
Much is known about the human genome. We know that the minority of the 3 gigabases of DNA sequence codes for proteins. It is estimated that only 2% of our DNA sequences code for approximately 20,000 protein coding genes. The function of the remaining 98% of DNA is perhaps the most fascinating aspect of genomics. Figure 53-1 illustrates the general structure of a typical gene. The 5′ control region, often termed the “promoter,” includes DNA sequences that recognize and bind to proteins called transcription factors, whose function is to modify gene expression by controlling transcription. The “start codon” is a series of nucleotides that are recognized by this transcriptional machinery, which initiates generation of messenger RNA (mRNA) from DNA. Not all of this mRNA sequence encodes for protein. Exons are the regions of DNA that are transcribed into RNA to make amino acids. In most genes, multiple exons are separated by introns, which are removed from mRNA prior to translation into protein. The end of transcription is signaled by a series of nucleotides at the end of the coding sequence, referred to as a “stop codon.” The DNA sequence after this end codon, termed the 3′ control region, can also influence the rate of gene transcription and may affect the stability of the mRNA sequence and its translation into protein.
FIGURE 53-1 The general structure of a typical gene.
THE GENETIC BASIS FOR DISEASE
Despite our limited knowledge of the function of much of the genome, our knowledge of the genetic basis for disease is, in fact, extensive. Basic research and clinical observation have elucidated the inheritance of single-gene, Mendelian disorders (transmitted according to Mendel’s laws of inheritance). Most of these are uncommon, and taken together, the most prevalent such as cystic fibrosis or hemochromatosis affect no more than one in several hundred people; however, when the genetic variant is present, its effect on individual patients is substantial. Furthermore, understanding the mechanisms underlying many monogenic disorders has provided pathophysiological information about related and more common disorders. For instance, we have learned a great deal about the pathophysiology of cardiovascular disease from discoveries related to familial hypercholesterolemia, a rare genetic disorder leading to premature atherosclerosis.
The most common type of DNA sequence variation is single base substitutions termed single nucleotide polymorphisms or single-nucleotide polymorphisms (SNPs). SNPs have emerged as powerful genetic markers for studying multifactorial diseases.6 Polymorphisms occur in all individuals and, by strict definition, SNPs exist with a frequency of >1%.7 Therefore, the term SNP does not include numerous other single base substitutions in which the least common allele is present at a frequency of ≤1% (such as “personal mutations” that may be limited to one family), nor does the term encompass other variations such as insertion/deletion polymorphisms. Importantly, SNPs should be distinguished from disease “causing” mutations, which are generally much less common, but have higher penetrance (e.g., sickle cell anemia). Therefore, SNPs typically do not cause disease but in aggregate may influence the risk for developing a disease (e.g., diabetes mellitus, asthma) or the outcome from a disease or condition (e.g., death from sepsis, rate of atherosclerosis in recipients of cardiac transplants).8,9 SNPs that exist in mRNA-coding regions (exons) of DNA can lead to amino acid substitutions, and therefore, may change the structure of the resultant protein. This type of variant potentially has the most profound impact on protein function; however, this type of SNP is the least common.5 SNPs in the regulatory (promoter) region of a gene may influence (increase or, more often, decrease) transcription of that gene, and therefore, may influence the amount of protein available.10 Yet, other SNPs may not directly alter protein abundance or function, but may be important markers for other (unidentified) functional variants—a concept known as linkage.11 The analysis of SNPs has been facilitated by two recent and related developments as follows: (1) the establishment of large data repositories of SNPs; and (2) the availability of relatively affordable “high-throughput” methods for genotyping. Recent techniques have identified at least 10 million SNPs interspersed throughout the human genome, at a frequency of one SNP per ˜300 base-pairs.12 Their role in critical illness, both as risk factors and in determining treatment response, has been investigated for several candidate genes.13 It is hoped that these studies will contribute to the understanding of the biology of critical illness and toward characterizing patient risk, treatment response, and outcome.14
Microsatellite and insertion/deletion polymorphisms represent other genetic variation that may be used to characterize an individual’s risk for disease and response to therapy. Microsatellite or simple sequence repeat (SSR) polymorphisms are created by the presence of short sequences (generally 2–4 base pairs) repeated multiple times in tandem. Insertion/deletion polymorphisms are characterized by the presence or absence of a single base in some cases or a longer fragment in others. Microsatellite and insertion/deletion polymorphisms are generally considered to be markers for other functional variants and are not themselves functional, but in some cases may directly alter gene activity.15,16 Typically, they occur in nonfunctional regions or in gene regulatory regions and not in coding regions as they would likely lead to a truncated, completely nonfunctional protein.
GENE EXPRESSION PROFILING AFTER TRAUMA AND BURNS
As described above, transcription of DNA into mRNA, commonly called “gene expression,” is the first step in the path to new protein production. The specificity of Watson–Crick base pairing allows for the measurement of mRNA abundance using technology similar to that enabling DNA sequencing. Only a short sequence (<100 nucleotides) is required to uniquely identify each gene and its associated mRNA. Small wafers arrayed with minute quantities of short, gene-specific nucleotides (called “microarrays”) are now used routinely to monitor gene expression for all human genes in any tissue of interest. The field of measuring the dynamics of mRNA abundance across the genome is called “gene expression profiling,” an approach used to estimate changes in global gene transcription based upon the good correlation between alterations in relative RNA abundance as measured by microarrays and changes in gene transcription. Of note, the correlation between changes in gene expression (mRNA production) and gene translation (protein production) is not as strong.17
Over the last decade, a team of investigators have been supported by the NIH to study the systemic response to blunt trauma and burn injury—the Inflammation and Host Response to Injury Program (also referred to as the “Glue Grant”). This team has applied these techniques of gene expression profiling to blood and tissue samples from subjects enrolled across the United States.18,19 The application of new computational approaches to gene expression data revealed novel insight into the biology of the response to injury. For example, network-based analysis of circulating white blood cells revealed that acute systemic inflammation induces transient dysregulation of bioenergetic pathways and modulation of leukocyte translational machinery.20
In addition to improving our understanding of pathophysiology, gene expression profiling also holds promise as a novel tool to improve the classification of disease and clinical phenotyping.21 The potential application to diagnostics and therapeutics in critically ill and injured patients is significant.22 If one assigns baseline or “reference” status to the averaged leukocyte gene expression profiles of healthy human volunteers, gene expression profiles from patients 12 hours after injury can be used to compute a difference-from-reference (DFR) score for each patient. In effect, these scores describe an overall response at the mRNA level, which can be compared with standard clinical scores based upon acute physiology, chronic health, and metrics of injury severity.23 These DFR scores early after injury have been associated significantly with adverse outcomes that developed later in the patient’s hospital course, including multiple organ dysfunction, length of mechanical ventilation and hospital stay, and infection rates. These associations remained statistically significant after adjustment for injury severity, indicating that the DFR score provides novel, important clinical information. Similarly, profiling of circulating leukocyte gene expression show promise as a novel tool to measure immune health and to diagnose infection.24 For instance, riboleukograms or graphs of circulating leukocyte changes in relative RNA abundance (gene expression) can be used to track the dynamics of the host response to systemic inflammation and the DFR.25 The diagnostic potential of riboleukograms was reported recently for adult trauma patients who developed ventilator-associated pneumonia and for pediatric patients with acute appendicitis.24,26
GENETIC RISK FACTORS AND PHARMACOGENOMICS
Genomic medicine holds the promise of personalized medicine or individualized treatment based upon a patient’s gene-associated disease risks. Venous thromboembolic disease is one example where understanding genetic risk factors may be important for optimizing care for critically ill patients. Clot formation is determined by a delicate balance of procoagulant and anticoagulant processes interacting via a complex system of cofactors and inhibitor regulated by an elaborate feedback mechanism.27 When this balance is disrupted, abnormal bleeding or clotting may ensue, a common occurrence in the intensive care unit. There are five established genetic defects considered as risk factors for venous thrombosis (Table 53-1).6 Deficiencies in protein C, protein S, and antithrombin III lead to defects in the anticoagulant pathways of blood coagulation and together are found in ˜15% of people with familial thrombophilia.28 The factor V Leiden mutation and increased prothrombin associated with the prothrombin 20210 A allele are much more prevalent and together may account for >60% of familial thrombophilia. Protein C, protein S, and antithrombin deficiencies all involve defects in anticoagulant pathways, whereas the factor V Leiden mutation and the prothrombin gene mutation involve procoagulant factors. Multiple genetic defects are responsible for protein C, protein S, and antithrombin deficiencies whereas both the factor V Leiden and prothrombin defects are caused by single mutations. These mutations are particularly relevant in patients with deep venous thrombosis occurring in the absence of a clear risk factor such as traumatic injury. Up to 50% of such individuals have an underlying defect including these genetic causes, which may contribute to the development of venous thromboembolism even when patients receive appropriate pharmacologic prophylaxis.28 Although it is likely that trauma such as fractures in an extremity or injury to the spinal cord is sufficient to lead to deep venous thrombosis in the absence of prophylaxis, it is possible that failure of prophylaxis may be, in part, due to the aforementioned genetic variants. Nevertheless, it is presently not recommended that patients with so-called provoked deep venous thrombosis undergo screening for these genetic variants. In part, the rationale is due to the notion that treatment will not be affected by the presence of these mutations.29 If the risk of failure of conventional prevention strategies is related to mutations causing thrombophilia, however, it may be possible to provide more aggressive prophylaxis to those at greatest risk for failure.
TABLE 53-1 Genetic “Causes” of Venous Thrombosis
For most patients, treatment for established deep venous thrombosis transitions from heparin-based therapy to warfarin-based therapy. The marked interindividual variability in response and a narrow therapeutic window make safe and effective dosing of warfarin a challenge. Variability in the response to warfarin can be due to genetic differences at a number of steps in its metabolism. The metabolism of warfarin and many other drugs used in critically ill patients is dependent on the cytochrome P450 (CYP) enzyme system. Not only is this system sensitive to influences of diet and medications, genetic differences exist in many of the cytochrome subfamilies, leading to varied pharmacological responses. For instance, cytochrome P450 2C is the subfamily responsible for the metabolism of many drugs and the CYP2C9 isoform is responsible for the catabolism of warfarin.30 Warfarin acts through the interference with vitamin K epoxide reductase, leading to secretion of inactive vitamin K–dependent clotting factors.31 Vitamin K epoxide reductase is encoded by the VKORC1 gene. It is now well recognized that the CYP2C9 genotype, when combined with the VKORC1 genotype, is predictive of dose requirement for oral anticoagulants, a fact that is likely to have clinical utility. In addition to variation in these two genes (which are estimated to contribute to ˜60% of the variation in effective warfarin dose), variation in eight additional genes has been reported to influence warfarin dose, even though to a lesser degree.32 In the near future, pharmacogenomic approaches will likely serve to reduce complications associated with important but risky therapies.33
The arrival of large-scale, individualized, whole-genome sequencing will provide rich opportunities for advancing our knowledge of pharmacogenomics in the intensive care unit beyond treatment for deep venous thrombosis. In the broadest sense, adverse drug events (ADEs) span a wide spectrum that includes errors related to administration to undesired responses occurring at normal drug dosages. There are few situations where the risk for ADEs is higher than in critically ill patients. Drug-induced toxicity in the context of cancer chemotherapy has been investigated using genome-wide approaches.34 These same techniques that identify genetic variation (SNPs) across the entire genome could be used to predict responses to anticoagulants, antibiotics, vasopressors, drugs active in the central nervous system, and the like, in critically ill patients.
GENETIC RISK FACTORS FOR SEPSIS AND ORGAN DYSFUNCTION
Genetic variation influences the risk for and outcome from acute surgical illness and its complications, such as infection and acute lung injury.13,35 The earliest evidence favoring a role for genetic differences in risk of infection and outcome came from an epidemiological study of adoptees and their biologic and adoptive parents. A strong association between early death from infection in adoptees and their biologic, but not adoptive, parents suggested a genetic influence on the risk for and outcome from infection.36 Subsequent studies determined that inflammatory responses in blood were heritable and were associated with outcomes from meningococcal sepsis.37 Following these initial observations, reports have identified associations between specific SNPs and sepsis (reviewed in reference38). Unfortunately, many of the initially striking associations between SNPs and sepsis risk and outcome have not been independently replicated.39,40
Despite the challenges discussed above, genomic variation likely will help us gain a better understanding of the biology of infection and organ failure, help us refine and target our therapies in critically ill patients and help us more accurately estimate prognosis. Along these lines, genetic differences may help explain why “anti-mediator” therapies, often observed to be beneficial in preclinical models and small clinical trials, have not improved outcomes in Phase III clinical trials.41–43 Much of our understanding of the biology of inflammation comes from highly controlled models systems. Examples include the discovery of the roles of tumor necrosis factor-α and toll-like receptor-4 in the inflammatory response.44,45 In genetically diverse populations (like humans), however, genetic variability creates sufficient background “noise” above which it may be difficult to discern true treatment effects (in other words, a low signal to noise ratio). In addition, new discoveries continue to refine our understanding of the pathways involved in the response to infection and inflammation and this improves our understanding of human inflammation and sepsis.46 Also, these pathways are likely influenced by genetic variation that, in turn, will influence clinical outcomes. Coupled with better biomarker-directed classification as described above, identifying important genetic risk factors will likely add to predictive accuracy. Finally, better risk stratification (by clinical and genetic factors) will better inform therapeutic decisions, directing therapy to patients most likely to benefit.18,20,47
Many challenges remain to be addressed before SNPs can be used in the clinical setting. Most importantly, the usefulness of individual SNPs to predict predisposition and response has met with little success, as most SNP–disease associations are not reproduced in subsequent studies.43 Moreover, the largest, most robust genome-wide association study studies published to date on polygenic diseases such as diabetes mellitus, inflammatory bowel disease, heart disease, and cancer indicate that the additional informational value for predicting a given phenotype is small.5 Once the predictive SNPs have been identified, it will be necessary to determine whether directing interventions to particular (“high-risk”) patients improves outcomes.
In sepsis and shock, multiple genetic and environmental factors influence an individual’s risk and clinical course as in more common heart disease, diabetes mellitus, hypertension, and cancer. Within infectious and inflammatory disorders, patient factors, such as age and comorbidity, and disease factors, such as infectious strain, exposure, and concomitant injury, contribute to disease risk and severity making the study of genetic risk factors particularly challenging.
SUMMARY AND IMPLICATIONS FOR CLINICAL PRACTICE
A growing body of information indicates that genomic data will enhance our understanding of common diseases including sepsis and organ failure.18,20 Similarly, genetic influences on many biological responses, such as drug metabolism, contribute to clinically important differences in responses to treatment. The role of genomic medicine in critical illness and injury is potentially great, but remains unanswered. Gene expression studies have provided new genomics tools such as the riboleukogram and significant novel insight that may facilitate more accurate diagnosis and therapy. The successful integration of genomics into caring for critically ill and injured patients, in particular, demands that clinicians become genomically literate.1,28 This chapter provides a framework in which to begin that process.
REFERENCES
1. Guttmacher AE, Collins FS. Realizing the promise of genomics in biomedical research. JAMA. 2005;294(11):1399.
2. Collins FS, Green ED, Guttmacher AE, et al. A vision for the future of genomics research. Nature. 2003;422(6934):835.
3. Watson JD, Crick FH. Molecular structure of nucleic acids; a structure for deoxyribose nucleic acid. Nature. 1953;171(4356):737.
4. Watson JD, Crick FH. A structure for deoxyribose nucleic acid. 1953. Nature. 2003;421(6921):397.
5. Guttmacher AE, Collins FS. Genomic medicine—a primer. N Engl J Med. 2002;347(19):1512.
6. Cardon LR, Bell JI. Association study designs for complex diseases. Nat Rev Genet. 2001;2:91.
7. Brookes AJ. 4th International Meeting on Single Nucleotide Polymorphism and Complex Genome Analysis. Various uses for DNA variations. Eur J Hum Genet. 2002;10(2):53.
8. Mira JP, Cariou A, Grall F, et al. Association of TNF2, a TNF-alpha promoter polymorphism, with septic shock susceptibility and mortality: a multicenter study. JAMA. 1999;282(6):561.
9. Tambur AR, Pamboukian S, Costanzo MR, et al. Genetic polymorphism in platelet-derived growth factor and vascular endothelial growth factor are significantly associated with cardiac allograft vasculopathy. J Heart Lung Transplant. 2006;25(6):690.
10. Wilson AG, Symons JA, McDowell TL, et al. Effects of a polymorphism in the human tumor necrosis factor alpha promoter on transcriptional activation. Proc Natl Acad Sci U S A. 1997;94(7):3195.
11. Feero WG, Guttmacher AE, Collins FS. Genomic medicine—an updated primer. N Engl J Med. 2010;362(21):2001.
12. Palmer LJ, Cardon LR. Shaking the tree: mapping complex disease genes with linkage disequilibrium. Lancet. 2005;366(9492):1223.
13. Wurfel MM, Gordon AC, Holden TD, et al. Toll-like receptor 1 polymorphisms affect innate immune responses and outcomes in sepsis. Am J Respir Crit Care Med. 178(7);710.
14. Wurfel MM. Genetic insights into sepsis: what have we learned and how will it help? Curr Pharm Des. 2008:14(19):1900.
15. Marshall RP, Webb S, Bellingan GJ, et al. Angiotensin converting enzyme insertion/deletion polymorphism is associated with susceptibility and outcome in acute respiratory distress syndrome. Am J Respir Crit Care Med. 2002;166(5):646.
16. Menges T, Hermans PW, Little SG, et al. Plasminogen-activator-inhibitor-1 4G/5G promoter polymorphism and prognosis of severely injured patients. Lancet. 2001;357(9262):1096.
17. Pene F, Courtine E, Cariou A, et al. Toward theragnostics. Crit Care Med. 2009;37(1 suppl):S50–S58.
18. Cobb JP, Mindrinos MN, Miller-Graziano C, et al. Application of genome-wide expression analysis to human health and disease. Proc Natl Acad Sci U S A. 2005;102(13):4801.
19. Feezor RJ, Baker HV, Mindrinos M, et al. Whole blood and leukocyte RNA isolation for gene expression analyses. Physiol Genom. 2004; 19(3):247.
20. Calvano SE, Xiao W, Richards DR, et al. A network-based analysis of systemic inflammation in humans. Nature. 2005;437(7061):1032.
21. Quackenbush J. Microarray analysis and tumor classification. N Engl J Med. 2006;354(23):2463.
22. Cobb JP, O’Keefe GE. Injury research in the genomic era. Lancet. 2004; 363(9426):2076.
23. Warren HS, Elson CM, Hayden DL, et al. A genomic score prognostic of outcome in trauma patients. Mol Med. 2009;15(7–8):220.
24. Cobb JP, Moore EE, Hayden DL, et al. Validation of the Riboleukogram to detect ventilator-associated pneumonia after severe injury. Ann Surg. 2009 Oct;250(4):531–539.
25. Polpitiya AD, McDunn JE, Burykin A, et al. Using systems biology to simplify complex disease: immune cartography. Crit Care Med. 2009; 37(1 suppl):S16–S21.
26. Muenzer JT, Jaffe DM, Schwulst SJ, et al. Evidence for a novel blood RNA diagnostic for pediatric appendicitis: the riboleukogram. Pediatr Emerg Care. 2010;26(5):333.
27. Majerus PW. Human genetics. Bad blood by mutation. Nature. 1994; 369(6475):14.
28. Bertina RM. Factor V Leiden and other coagulation factor mutations affecting thrombotic risk. Clin Chem. 1997;43(9):1678.
29. Weitz JI, Middeldorp S, Geerts W, et al. Thrombophilia and new anticoagulant drugs. Hematol Am Soc Hematol Educ Program. 2004;424.
30. Kumar V, Wahlstrom JL, Rock DA, et al. CYP2C9 inhibition: impact of probe selection and pharmacogenetics on in vitro inhibition profiles. Drug Metab Dispos. 2006;34(12):1966.
31. Wadelius M, Pirmohamed M. Pharmacogenetics of warfarin: current status and future challenges. Pharmacogenomics J. 2007 Apr;7(2):99–111.
32. Wadelius M, Chen LY, Eriksson N, et al. Association of warfarin dose with genes involved in its action and metabolism. Hum Genet. 2007 Mar; 121(1):23–34.
33. Tan GM, Wu E, Lam YY, et al. Role of warfarin pharmacogenetic testing in clinical practice. Pharmacogenomics. 2010;11(3):439.
34. Gamazon ER, Huang RS, Cox NJ, et al. Chemotherapeutic drug susceptibility associated SNPs are enriched in expression quantitative trait loci. Proc Natl Acad Sci U S A. 2010;107(20):9287.
35. Reddy AJ, Christie JD, Aplenc R, et al. Association of human NAD(P)H: quinone oxidoreductase 1 (NQO1) polymorphism with development of acute lung injury. J Cell Mol Med. 2009;13(8B):1784.
36. Sorensen TI, Nielsen GG, Andersen PK, et al. Genetic and environmental influences on premature death in adult adoptees. N Engl J Med. 1988; 318(12):727.
37. Westendorp RG, Langermans JA, Huizinga TW, et al. Genetic influence on cytokine production and fatal meningococcal disease. Lancet. 1997: 349(9046):170.
38. Lin MT, Albertson TE. Genomic polymorphisms in sepsis. Crit Care Med. 2004;32(2):569.
39. Imahara SD, Jelacic S, Junker CE, et al. The TLR4 +896 polymorphism is not associated with lipopolysaccharide hypo-responsiveness in leukocytes. Genes Immun. 2005;6(1):37.
40. O’Keefe GE, Hybki DL, Munford RS. The G→A single nucleotide polymorphism at the –308 position in the tumor necrosis factor-alpha promoter increases the risk for severe sepsis after trauma. J Trauma. 2002; 52(5):817.
41. Opal SM, Cross AS. Clinical trials for severe sepsis. Past failures, and future hopes. Infect Dis Clin North Am. 1999;13(2):285, vii.
42. Vincent JL, Sun Q, Dubois MJ. Clinical trials of immunomodulatory therapies in severe sepsis and septic shock. Clin Infect Dis. 2002;34(8): 1084.
43. Zeni F, Freeman B, Natanson C. Anti-inflammatory therapies to treat sepsis and septic shock: a reassessment. Crit Care Med. 1997;25(7):1095.
44. Beutler B, Poltorak A. Positional cloning of Lps, and the general role of toll-like receptors in the innate immune response. Eur Cytokine Netw. 2000;11(2):143.
45. Tracey KJ, Fong Y, Hesse DG, et al. Anti-cachectin/TNF monoclonal antibodies prevent septic shock during lethal bacteraemia. Nature. 1987; 330(6149):662.
46. Riewald M, Petrovan RJ, Donner A, et al. Activated protein C signals through the thrombin receptor PAR1 in endothelial cells. J Endotoxin Res. 2003;9(5):317.
47. Bernard GR, Wheeler AP, Russell JA, et al. The effects of ibuprofen on the physiology and survival of patients with sepsis. The Ibuprofen in Sepsis Study Group. N Engl J Med. 1997;336(13):912–918.